Dementia in sub-Saharan Africa: Challenges and opportunities.

Abstract

We have updated our estimates of the prevalence of dementia and numbers living with
the condition in the sub-Saharan Africa (SSA) region, the incidence of dementia, and
the economic impact of dementia, supplemented by new reviews of the evidence on
the mortality risk for people with dementia and of risk factors in SSA. We have also
systematically reviewed qualitative evidence on the experience of living with dementia
in SSA, providing important insights into the everyday reality in this region. Finally, we
examined briefly the development of social protection schemes in SSA and assessed their effects on the health of older people.

Full text

Dementia in sub-Saharan Africa
Challenges and opportunities

Authors
Dr Maëlenn Guerchet
The Global Observatory for Ageing and Dementia Care, King’s College
London (UK)
Dr Rosie Mayston
Centre for Global Mental Health, King’s College London (UK)
Prof Peter Lloyd-Sherlock
School of International Development, University of East Anglia (UK)
Prof Martin Prince
The Global Observatory for Ageing and Dementia Care, King’s College
London (UK)
Dr Isabella Aboderin
Aging and Development Program, Africa Population and Health
Research Center, Nairobi (Kenya)
Dr Rufus Akinyemi
Neuroscience and Ageing Research Unit, Institute for Advanced
Medical Research and Training, College of Medicine, University of
Ibadan (Nigeria)
Dr Stella-Maria Paddick
Institute for Ageing and Health, Newcastle University (UK)
Prof Anders Wimo
Department of Neurobiology, Care sciences and Society, Karolinska
Institute, Stockholm (Sweden)
Dr Mary Amoakoh-Coleman
Department of Epidemiology, Noguchi Memorial Institute for Medical
Research, University of Ghana
Dr Richard Uwakwe
Nnamdi Azikiwe University Teaching Hospital, Nnewi, Anambra State
(Nigeria)
Dr Peter Ezeah
Department of Sociology/Anthropology, Nnamdi Azikiwe University,
Awka (Nigeria)
Alzheimer’s Disease International
Contributors
Dr Declare Mushi
Kilimandjaro Christian Medical University College (Tanzania)
Dr A. Matthew Prina
The Global Observatory for Ageing and Dementia Care, King’s College
London (UK)
Acknowledgements
Prof Pierre-Marie Preux and Dr Harielle Samba (University of
Limoges, France), Prof Adesola Ogunniyi (University of Ibadan,
Nigeria) as well as the 10/66 Dementia Research Group
(London, UK) for providing data or manuscripts from their
population-based studies.
All the Alzheimer associations who have provided stories to
illustrate our report.
Chapter 1, 3, 4 photos – Leah Beach – www.leahbeach.com
Cover and chapter 2, 5 photos – iStock
Design – David O’Connor – www.daviddesigns.co.uk
Published by Alzheimer’s Disease International (ADI), London.
September 2017.
Copyright © Alzheimer’s Disease International.
Chapter 1
Dr Maëlenn Guerchet, Dr Isabella Aboderin
Chapter 2
Dr Maëlenn Guerchet, Dr Rufus Akinyemi, Dr Stella-Maria
Paddick, Prof Anders Wimo, Prof Martin Prince
Chapter 3
Dr Rosie Mayston, Dr Richard Uwakwe, Dr Peter Ezeah,
Prof Martin Prince, Prof Peter Lloyd-Sherlock
Chapter 4
Prof Peter Lloyd-Sherlock, Dr Mary Amoakoh-Coleman
Chapter 5
Dr Maëlenn Guerchet, Prof Martin Prince, Alzheimer’s Disease
International
This report was independently researched and authored by
Dr Maëlenn Guerchet, Dr Rosie Mayston, Prof Peter Lloyd-
Sherlock, Prof Martin Prince, Dr Isabella Aboderin, Dr Rufus
Akinyemi, Dr Stella-Maria Paddick, Prof Anders Wimo,
Dr Mary Amoakoh-Coleman, Dr Richard Uwakwe, Dr Peter
Ezeah, with contributions from others as listed. The evidence
reported and the inferences drawn are the responsibility of the
authors alone.
Personal stories
Stories included in boxes throughout the report were
provided by Alzheimer’s associations (members of
ADI) in sub-Saharan Africa. They are to illustrate how
living with dementia in different parts of the African
continent can be, including the journey to diagnosis
and care. We have specifically chosen stories
highlighting how associations have been supporting
the people with dementia and their carers. These
are raw reports of personal experiences and are in
no way generalizable to the population of either the
country they are coming from or the region.

3DEMENTIA IN SUB-SAHARAN AFRICA
Foreword
Sub-Saharan Africa represents today one of the biggest challenges for Alzheimer’s Disease
International (ADI). To understand the scale of this challenge it is sufficient to say that after 37 years
of incessant work to build capacity in the civil society sector globally, we only have seven member
nations out of a possible 49 in the region.
My discussions in the last few months with both our members and high level officials have brought
to the fore a number of stark facts.
The scale of stigma in sub-Saharan Africa is huge. It was very difficult to find people with dementia
willing to speak at our conference this September and, by doing that, to admit that they have
the condition. There is talk of older people with mental illness being branded witches and being
restrained forcefully, or worse. This report reassesses the scale of this problem, but we should not
underestimate it, as few are willing to speak up.
At policymaking level, it is clear to me that mother and child health issues and infectious diseases
are the top priority in this region and that non-communicable diseases (NCDs) and older people’s
issues are not really on the radar of many sub-Saharan countries.
This report is therefore extremely timely. It not only provides new quantitative and qualitative
evidence and adjusts upwards most of our previous estimates – thereby confirming the scale of
the epidemic is worse than previously suggested – but it examines many more issues. From risk
prevention and reduction, to social and health care, health insurance and pensions, the report
pragmatically raises possible solutions to an impending crisis.
The report also quantifies the obvious. The patterns of care in sub-Saharan Africa are changing;
ageing population, lower fertility, more women working, rural-urban migration and migration more
generally are all contributing factors to a change in the way care is delivered traditionally – within
the family. With extended families increasingly losing this role, the potential for a crisis is evident.
Yet, older people are seen as not contributing to economies, and their voices and rights are
routinely excluded or ignored. Inequality is also a huge issue.
We obviously need more data and more research in virtually every area described in this report.
There are large gaps in our knowledge and we need more evidence coming from countries in the
region.
We must act now. Whereas in other regions we need more awareness of dementia as a disease,
not a normal part of ageing, in sub-Saharan Africa we need to go deeper and challenge the
prejudices faced by those who have the disease. Governments also need to tackle this from the
viewpoint of risk reduction and prevention and in the context of NCDs more broadly. There is
a pressing need to educate about dementia and this needs to be taken on by the public health
functions of all countries.
The approval of the World Health Organisation (WHO) Global Action Plan on Dementia in May
2017 allows us to put greater pressure on governments to take the issue with the seriousness it
deserves and act. The African Union adopted in January 2016 a ‘Protocol on the Rights of Older
Persons in Africa’ which is legally binding. African states agreed on the Sustainable Development
Goals framework, which includes older people, and recently all African countries approved the
WHO’s Global Action Plan. This includes the monitoring systems set up by the Global Dementia
Observatory and the understanding that data should begin to arrive soon.
ADI and its members will continue in their work, challenging governments, proposing solutions,
strengthening civil society, representing the voices of those most at risk and tackling inequality. If
you are reading this report and think action should be taken, join us.
Paola Barbarino, Chief Executive Officer

ALZHEIMER’S DISEASE INTERNATIONAL4
Alzheimer’s Disease International
Dementia in sub-Saharan Africa
Challenges and opportunities
The Global Observatory for
Ageing and Dementia Care
The Global Observatory for Ageing and Dementia Care, hosted at the Health Service
and Population Research Department, King’s College London, was founded in 2013.
Supported by Alzheimer’s Disease International, and King’s College London, the
Observatory has a tripartite mission:
1. To build upon ADI’s 10/66 Dementia Research Group program of population-based
and intervention research in low and middle income countries, maximising the
impact that research findings from our data can have upon policy and practice.
2. To develop, evaluate, and promote primary care and community interventions for
people with dementia.
3. To synthesise global evidence for policymakers and public, in particular, continuing
and developing our role in the preparation of high impact evidence-based reports
for Alzheimer’s Disease International (World Alzheimer Reports 2009, 2010, 2011,
2013, 2014, 2015, 2016 and Nutrition and dementia), the World Health Organization
(Dementia: a public health priority, 2012) and other relevant intergovernmental
organisations.

5DEMENTIA IN SUB-SAHARAN AFRICA
Contents
Summary 6
CHAPTER 1
Introduction – Ageing in sub-Saharan Africa 10
Demographic ageing in sub-Saharan Africa 11
The rise of non-communicable diseases:
epidemiological transition and triple burden 12
Policy context and implications for research 14
CHAPTER 2
The epidemiological evidence:
prevalence, costs, incidence, mortality and risk factors 16
Introduction 16
Prevalence 17
Costs 24
Incidence 25
Mortality associated with dementia 28
Risk factors 29
Conclusion and recommendations 34
Personal story – South Africa 41
CHAPTER 3
Living with dementia in sub-Saharan Africa:
systematic review of qualitative evidence 42
Methodology 42
Search results and characteristics of selected studies 43
Results 44
Discussion 47
Recommendations 49
Strengths and limitations of our approach 52
CHAPTER 4
Social protection and the health of older people
in sub-Saharan Africa 54
Introduction 54
Pensions for older people 54
Health insurance: does Ghana show the way forward? 56
The effects of social protection on the health status of older people 57
Conclusion 57
Personal story – Mauritius 59
CHAPTER 5
Conclusions and recommendations 60
Summary 60
Global Action Plan on Dementia 62
Concluding remarks and call to action 68
Appendix A 69
Personal story – Zimbabwe 70

ALZHEIMER’S DISEASE INTERNATIONAL6
Summary
CHAPTER 1
Introduction – Ageing in sub-Saharan Africa
1. We have updated our estimates of the prevalence of dementia and numbers living with
the condition in the sub-Saharan Africa (SSA) region, the incidence of dementia, and
the economic impact of dementia, supplemented by new reviews of the evidence on
the mortality risk for people with dementia and of risk factors in SSA. We have also
systematically reviewed qualitative evidence on the experience of living with dementia
in SSA, providing important insights into the everyday reality in this region. Finally, we
examined briefly the development of social protection schemes in SSA and assessed their
effects on the health of older people.
2. The phenomenon known as population ageing or ‘demographic transition’ is occurring
globally – much more rapidly in some developing middle income countries. Despite having
the youngest populations in the world, sub-Saharan African countries also have the fastest
growing rates of older people in the world. The population of older people aged 60+ is
projected to reach 161 million by 2050. However, there are clear sub-regional differences
in ageing profiles.
3. Societal implications of population ageing and the effects of social change on experiences
of old age in SSA remain poorly understood while being varied and complex. Families are
a key social group, particularly in the absence of comprehensive public welfare systems
and provision. Major concerns exist about changes in SSA family structures driven by
urbanization, migration, global agendas and persistent economic strain.
4. As people live longer and changes in lifestyles and behaviours predisposing towards
certain risk factors occur, chronic diseases become more common. Sub-Saharan Africa
is not spared by this epidemiologic transition, and a rise in non-communicable diseases
(NCDs) has been described, causing a triple burden of disease in SSA which could have a
devastating impact on a region already facing significant resource constraints.
5. Health inequalities among SSA’s older adults emerge along several axes, including
gender, socio-economic position and age, but also place of residence and type of health
condition.
6. The policy architecture on ageing and health in SSA has been consolidated in recent
years, with an emergence of regional and national frameworks as well as African countries’
commitments to global agendas (including the United Nations’ Sustainable Development
Goals and the WHO Global Strategy on Ageing and Health). Despite such policy advances,
comprehensive programmatic actions on older people’s health remain sparse. There is an
urgent need for a concerted research effort to help promote and inform consideration of
dementia in national and regional health agendas.
CHAPTER 2
The epidemiological evidence: prevalence, costs, incidence,
mortality and risk factors
1. We provide an overview of the epidemiology and impact of dementia in SSA, using
essentially the same methods as previously used in the World Alzheimer Reports,
conducting new reviews of the studies regarding dementia prevalence, incidence and
mortality. We present updated estimates of the numbers of people living with dementia in
SSA, as well as costs, and evidence on risk factors of dementia, with a specific note on
HIV-associated neurocognitive disorders.
2. Our regional estimate of age-specific dementia prevalence in people aged 60 years
and over is now 6.38%, doubling with every 7.2 year increment in age. This prevalence
is slightly higher than that presented in the World Alzheimer Report 2015, due to the
inclusion of three additional studies in the meta-analysis.

7DEMENTIA IN SUB-SAHARAN AFRICA
3. We estimate that 2.13 million people were living with dementia in SSA in 2015. This
number will reach 3.48 million by 2030, and 7.62 million in 2050, with the most important
increases in Eastern and Central SSA. The increase in these new estimates compared
to the World Alzheimer Report 2015 is driven by both population ageing and the slightly
higher estimate of prevalence of dementia in SSA.
4. The total costs of dementia in sub-Saharan Africa were an estimated US$6.2 billion in
2015. The pattern of distribution of costs between the three major sub-categories (direct
medical, social care and informal care) has not changed, with two-thirds to three-quarters
of the total costs attributed to informal care.
5. We estimated over 367,000 new cases of dementia in 2015 in SSA, a slightly lower
estimate than that in the World Alzheimer Report 2015. However, previous estimates were
generated by using the evidence on the incidence of dementia estimated in all low and
middle countries (including one study in Nigeria) while we have now been able to rely on
the evidence on incidence of dementia in SSA only.
6. A two-fold increased mortality risk for people living with dementia in SSA was estimated.
Relatively similar mortality hazards have been reported in other low and middle countries.
7. Studies of risk factors of dementia in SSA have shown consistent association of age and
female sex with dementia. The evidence around other risk factors is more patchy, and
this needs to be more thoroughly investigated. Education and the APOE ε4 allele showed
very mixed effects. The strongest evidence on modifiable risk factors is the association
of dementia with vascular risk factors, while the evidence on early-life, psychological,
psychosocial and lifestyle factors remains tentative.
8. An important gap in the evidence has been identified regarding HIV-associated dementia
(prevalence or risk factors) among older people in SSA. Existing, albeit limited, data relate
almost exclusively to young adults.
9. The evidence on dementia prevalence, incidence and mortality has considerably expanded
since the World Alzheimer Report 2009. We are therefore slowly moving away from single
estimates for the whole sub-Saharan region for most of the epidemiological indicators.
However, the evidence on risk factors of dementia in the region is still not very strong,
mainly relying on cross-sectional studies rather than longitudinal ones.
10. Although the overall quality of dementia studies in SSA is high, strengthened by the
systematic use of multi-domain cognitive batteries and informant interviews during
dementia assessment, there is still room for improvement on the correct implementation
of multistage designs and their analysis, and adequate description of the studies when
disseminating results.
11. A major current limitation regarding our estimates of costs of dementia in sub-Saharan
Africa is that they are mainly based on imputations, as no original data has been published
yet on resource use and costs of dementia in SSA. Descriptions of care systems and
amounts of informal care provided for at least one country in each of the sub-regions
would contribute to more reliable estimates.
CHAPTER 3
Living with dementia in sub-Saharan Africa
1. Understanding beliefs, perceptions and experiences relating to illness, care, help-seeking
and treatment is critical to understanding the nature of the problem and designing
interventions and services that are person-centred and which meet locally defined needs.
We carried out a systematic review of the scientific qualitative evidence on the experience
of living with dementia in SSA.
2. Four over-arching themes emerged from the data: beliefs (understandings of dementia
and beliefs about causality), experiences (symptoms, stigma, impact of caring for a person
living with dementia), responses (care arrangements, help-seeking) and unmet needs
(emotional support, knowledge, economic support and services).
3. No equivalent term for dementia was identified in any local languages, while there was a
general lack of awareness of dementia among the community. The most prominent belief
regarding underlying cause of dementia was that although not experienced by everyone,
dementia-like symptoms were part of “normal ageing”.

ALZHEIMER’S DISEASE INTERNATIONAL8
4. Participants most frequently reported problems with memory and communication, but
problems with overall functioning were also reported, as well as problems typically
associated with co-morbidities. When people with dementia were perceived to be witches,
fear inevitably resulted in discrimination, isolation and instances of violence. Where
participants stated that there was no stigma, as dementia-like symptoms were accepted
as a “normal part of old age”, people with dementia were routinely referred to as childlike.
5. Care, mainly delivered by families, was shaped by a need to show gratitude for the care
provided to the current carers by people with dementia earlier in their lives. Women tended
to be the hands-on carers while men often played a role in decision-making about care.
Caring was often a collective responsibility with different components assigned to different
family members. When used, hired help was perceived as a difficult but necessary
component of the care strategy.
6. Families were pluralistic in their approaches to treatment, using both biomedical and
traditional or faith healing to treat the problems of older people. The condition was always
recognised as incurable if the acknowledged cause was ageing. The proportion of people
with dementia who seek help from non-biomedical healers in SSA was significant.
7. Emotional support for carers was felt to be essential to enable them to continue to deliver
care. Education was also deemed to be essential, both to counteract beliefs related to
witchcraft and to improve the level of support for carers in their role.
8. People living with dementia and carers were dependent upon families and communities
to support the costs of care, with costs of transportation to clinics felt to be particularly
problematic. Care was managed by families with no involvement or support from
government sources.
9. The evidence identified covered only five countries, with all sub-regions but Southern SSA
represented by only one or two studies. Given the heterogeneity and diversity of cultures,
languages and beliefs within SSA, the generalisability of those results to other countries or
settings than those where they originated is debatable.
CHAPTER 4
Social protection and the health of older people in sub-
Saharan Africa
1. Social protection, which seeks to reduce individuals’ vulnerability, is a necessity for
older people across SSA, especially for people living with dementia. We examined the
development of social protection in SSA, mainly based on cash-transfers and health
insurance, and assessed its effects on the health of older people by looking at South
Africa and Ghana, for which data were available.
2. A number of countries, mainly located in southern SSA, have relatively extensive
and generous social pension schemes, usually combined with separate contributory
arrangements. In the rest of the region, the coverage of pension schemes and the value of
benefits is considerably more limited. Some countries have developed pilot social pension
schemes, usually with the support of external funding. Social protection in SSA is highly
inequitable and crowds out public expenditure for more vulnerable groups of older people.
3. South Africa has an established system of non-contributory pensions, disability benefits
and “Grant-in-Aid” benefit for older people aged over 60-65, adults with chronic
conditions or who “require regular care”. However, the geographical distribution of the
benefits remains uneven. A system of non-contributory cash transfers also exists and has
been put forward as an example of “best practice” for other developing countries. It has
been claimed that it does not just benefit older people, but the entire households, which
may increase the authority of older people and hence the potential willingness of family
members to care for frail older relatives.
4. In SSA, only a small number of health insurance schemes have achieved good coverage
of their populations. Coverage is usually highly concentrated on more affluent social
groups. In this context, Ghana’s national health insurance scheme, free for people aged
over 69 years, is unique and has been presented as a potential model for other resource-
constrained countries.

9DEMENTIA IN SUB-SAHARAN AFRICA
5. The availability of and access to health services across most of sub-Saharan Africa being
extremely limited, the capacity of social pensions to have a significant impact on the
health status of older people will remain very limited unless substantial improvements are
made to the suitability, quality and accessibility of these services.
6. The direct effect of health insurance and pensions on enhancing health outcomes for
older people is contingent on the generosity of the scheme, health awareness and the
availability of suitable services. Simply being enrolled in health insurance or utilising health
services should not in themselves be considered positive outcomes.
CHAPTER 5
Conclusions and recommendations
1. We estimate that 2.13 million people were living with dementia in sub-Saharan Africa in
2015, with numbers projected to nearly double every 20 years, increasing to 3.48 million
by 2030 and 7.62 million by 2050. There were over 367,000 new cases of dementia in
2015, while people with dementia had a two-fold increased mortality risk compared
to people not affected by the condition. The total costs of dementia in SSA are now
estimated at US$ 6.2 billion in 2015, with two-thirds to three-quarters of the total costs
and cost per person attributed to informal care costs.
2. In May 2017, the World Health Organization’s (WHO’s) Global Plan of Action on Dementia
was unanimously adopted at the 70th session of the World Health Assembly in Geneva.
As yet, no dementia plans or strategies have been established by countries in SSA. The
WHO Global Plan, which includes specific actions for Member States, international,
regional and national level partners, provides indicators and targets to evaluate levels
of implementation, progress and impact. For the first time, governments can be held
accountable on the advancements of dementia awareness, risk reduction, diagnosis, care
and treatment, support for care partners and research in their countries.
3. Improving dementia awareness and friendliness, access to services (through Universal
Health Coverage) and dementia risk reduction are challenges that countries in sub-
Saharan Africa must tackle. Considering the evident lack of ethical, social and legal
protection of people with dementia in SSA, it is essential that rights are recognized,
respected and protected to empower people living with dementia, those who support
them and the community as a whole. The gap in evidence for almost all domains of
the dementia field in SSA is striking. The Global Plan must be used to support the
development of dementia research globally and guide nations to define adequate research
agendas, supporting national and regional health agendas.
4. ADI and its focused, energetic and committed member associations are committed to
finding solutions to a very real and tangible problem. Associations are forging ahead,
creating hope and activity were there was little, mobilizing goodwill and resources, and
trying to sensitise their governments. Governments in the region have a real opportunity
right now – thanks to the excellence of these grassroots organisations – to join forces with
them and by adopting and implementing the seven key areas of the WHO Global Plan, to
show the rest of Africa the way.

ALZHEIMER’S DISEASE INTERNATIONAL10
CHAPTER 1
Introduction – Ageing in
sub-Saharan Africa
Introduction
This report comprises an overview of current knowledge about the evolution, spread and patterns
of dementia in sub-Saharan Africa (SSA). To offer a complete description of the epidemiology of the
condition in the region, we provide updated estimates of the prevalence and incidence of dementia,
the numbers of people living with the condition, and its economic impact. We additionally review
evidence on key risk factors for, and the mortality risk of, people living with dementia in SSA. In
order to provide insights on lived realities of the condition in the region we offer a systematic review
of qualitative evidence of the experiences and perspectives of people with dementia, their families
and those offering services to them in SSA contexts. Finally, we discuss key developments in the
expansion of health- or old age-focused social protection schemes in SSA and their effects on
the health status of older adults. Social protection measures are an increasingly salient element
in countries’ efforts to ensure the well-being of vulnerable populations. As such, such schemes
are particularly relevant to meeting the growing care and service needs associated with a rising
dementia burden.
As with previous reports, our focus is on older people aged 60 years and over. This is
notwithstanding the possibly elevated proportion of early onset dementia in SSA (higher than the
global estimate of 2-8% of all cases) due to the high seroprevalence of HIV infection1. Evidence on
HIV-associated dementia in the region will be assessed separately.
The geographical focus of this report is on the region of sub-Saharan Africa, as commonly defined
by key international organizations such as the United Nations or World Bank. SSA comprises 49
countries and four sub-regions with highly diverse social, economic, physical and cultural profiles.
Nonetheless, certain features and challenges of SSA’s demography and health context cut across
nations. Key among these aspects are (i) an expected sharp rise in the absolute size of the older
population, (ii) a triple burden of disease driven by persistent communicable- and emergent non-
communicable conditions as well as injury and trauma and, (iii) weak or poorly developed health
and social care systems.
We therefore begin this report – in the remainder of this chapter – with a brief description of broad
demographic and epidemiological trends in SSA, key challenges of ageing, health and dementia
that arise within this context, and the policy frames that presently exist to respond to them. We
conclude with a reflection on emerging implications for research.

11DEMENTIA IN SUB-SAHARAN AFRICA
1.1 Demographic ageing in sub-Saharan Africa
The phenomenon known as population ageing – a last stage in the so called ‘demographic
transition’ marked by a rise in the proportion of older people in the total population – is occurring
globally. It results from shifts from high fertility / high mortality states to low fertility / low mortality
states driven by improvements in public health, economic and social development. Such advances
typically lead, first, to declines in child mortality, giving rise to high rates of population growth.
Subsequent declines in fertility rates (people having fewer children) often to ‘replacement level’
(two births per woman) then slow population growth. Parallel improvements in adult health lead to
decreased mortality in later life. This further increases life expectancy and, together with reduced
fertility rates, causes a steady rise in the total population share of older adults.
The pace of such population ageing is much more rapid in some low and middle income countries
(LMIC) than was the case in the ‘old’, now industrialised world. While, for example, it took between
45 and 115 years for some high income countries (HIC) to double their total population aged 65
years and over (from 7% to 14%), it is estimated that many LMIC will accomplish the transition in
less than 25 years2.
Figure 1.1
Projected trends in percentage of population over 60 years: SSA and other world regions
Table 1.1
Projected trends in absolute numbers (in thousands) of older people (60+): SSA and other world regions
2015 2020 2025 2030 2035 2040 2045 2050
Sub-Saharan
Africa
46,455 54,072 63,293 74, 5 0 4 89,098 108,085 132,347 161, 077
Latin America and
the Caribbean
70,922 85,353 102,6 50 120,9 5 9 139,69 3 159,785 180,4 47 200,031
Asia 5 0 7, 9 5 4 599,433 711,622 844,487 969,231 1,066,63 5 1,170,148 1,2 9 3,710
Europe 176 ,513 191,371 206,338 217, 2 20 2 26,010 232,527 238,084 242,001
Northern America 74, 5 89 86,537 9 7, 74 3 104,79 9 110 ,403 113, 741 118, 4 42 122,67 9
Oceania 6,481 7,4 9 3 8,610 9,553 10,6 08 11,3 4 7 12 ,274 13, 204
Within this overall picture there are clear sub-regional differences in ageing profiles. Central and
eastern Africa, as Table 1.2 shows, will see the largest increases in the population of older people
while Southern Africa will see the smallest. The highest rise in life expectancy at birth will take
place in Central Africa (from 55.6 years in 2010-2015 to 67.7 years in 2045-2050). Life expectancy
at 65, presently highest in Eastern Africa, will grow by the largest margin in this sub-region as
well as in Southern Africa. Southern Africa also has the highest proportion of older adults in the
population and, together with Eastern Africa, will see the highest rise - a doubling - in this share by
mid-century. Much lower increases in the population proportion of older people are expected in
Western and Central Africa.
0.00
5.00
10.00
15.00
20.00
25.00
30.00
35.00
40.00
Percentage population 60+
SSA
Latin America and the Caribbean
Asia
Europe
Northern America
Oceania

ALZHEIMER’S DISEASE INTERNATIONAL12
Table 1.2
Demographic profiles of ageing in SSA sub-regions between 2015 and 20503
60+ (thousands) Proportionate
increase (%)
Percentage of
population over
60 (%)
Life expectancy at birth Life expectancy at 65
2015 2050 2015-2050 2 015 2050 2010-2015 2045-2050 2010-2015 2045-2050
West Africa 16,006 52,752 230 4.53 6.61 55.0 65.5 11. 3 12.9
Central Africa 6,901 24 ,411 254 4.54 6.61 55.6 67.7 13 .2 15.0
East Africa 18,86 8 72,436 284 4.78 8.25 60.5 7 2.1 14.2 16.4
Southern
Africa
4,680 11, 47 7 14 5 7. 4 7 14.71 5 7.1 68.2 13. 2 15.5
The societal implications of population ageing and the effects of social change on experiences
of old age in SSA remain poorly understood and are undoubtedly varied and complex. A critical
factor in both, however, is families. Families are understood as the key social group in which the
lives of, and support for, older people in SSA are embedded. The relevance of families is arguably
particularly pronounced within contexts, as in most SSA countries, marked by a broad absence of
comprehensive public welfare systems and provision4-6.
Major concerns exist about changes in SSA family structures and bonds driven by urbanization,
migration, global agendas and persistent economic strain - and the impacts of such shifts on
the roles of, security for, and attitudes toward older adults. Additional concerns focus on gaps in
available care for older adults, as well as an expansion of new care responsibilities on them - as a
result of the HIV/AIDS epidemic particularly in Southern and Eastern Africa6.
There remains considerable debate about the ways in which major social trends are shaping
experiences of families and later life, and the extent to which a commonly assumed erosion of
customary family care systems and values for elders is, indeed, occurring6.
However, there is little doubt that for many older adults in SSA, later life is marked by a vulnerability
to poverty, social exclusion, ill-health and disability. At the same time, there is growing recognition
that older adults continue to play a range of key roles and that they wield relevant influences
whether positive or negative, active or latent, at family, community or societal levels7.
1.2 The rise of non-communicable diseases: epidemiological
transition and triple burden
To a large extent the burden of ill-health and disability faced by SSA’s older adults results from
chronic, non-communicable diseases (NCD), which are expanding in SSA as a function of a so-
called ‘epidemiological transition’. As people live longer and lifestyles and behaviours change,
such conditions become more prevalent with bigger impacts on the health of a population. Such a
transition involves specific processes in which:
i) Infectious diseases, including malaria, tuberculosis, and polio are brought under
control, and childhood and maternal health improves as a result of economic and social
development. Together, such shifts reduce mortality at young ages, as well as preventable
childhood or lifelong disability.
ii) Simultaneously, changes in behaviour and lifestyles occur in tandem with urbanization,
globalization and industrialisation – towards a ‘western’ pattern of sedentarism, high
dietary consumption of salt, fat and sugars, smoking and alcohol use. Such risk factors
drive an increase in the incidence of certain chronic diseases, specifically cardiovascular
diseases, cancers, diabetes and obesity. They are also linked to dementia, however,
suggesting likely rises in the condition in populations undergoing this transition.
A clear sign of an ongoing epidemiological transition in sub-Saharan Africa is the rise in morbidity
and mortality caused by NCD over the last decade8-10. In 2015, over 2.7 million deaths due to NCDs
occurred in SSA, a 63% increase from 199011. Findings of the latest Global Burden of Disease

13DEMENTIA IN SUB-SAHARAN AFRICA
(GBD) Study estimate the number of Disability Adjusted Life Years (DALYs) caused by NCD in
SSA at over 153 million, representing almost 30% of the burden caused by all diseases. Among
the older population (aged 70 years and over), this proportion rises to 72%, with over 15 million
DALYs11. The contribution of NCDs to the burden of disability in SSA is even higher. They account
for 63% of all Years Lived with Disability (YLD) in the total population and 83% of YLD in the
older11.
The most prominent NCDs in SSA at present are cardio-vascular diseases (CVDs) and neoplasms
as the first causes of both death and disability in the overall population. This pattern remains the
same for the older population, where neurological disorders rank as the 5th cause of deaths and
8th cause of DALYs11.
Dementia emerged globally as one of the 15 leading causes of DALYs among older people in
201012 and in the oldest age groups (75 and over), Alzheimer’s and other dementias now rank as
one of the top ten leading causes of disability in 201513. In SSA, dementias such as Alzheimer’s
are still less prominent. Nonetheless they pose a considerable disease burden on the region’s
older population that is possibly underestimated in present GBD analysis. In 2010 in all of SSA,
Alzheimer’s disease (AD) was described as the 33rd leading cause of death in people aged 70 and
older14. However, authors acknowledged that the GBD mortality estimates for AD had to be viewed
with caution given a lack of data from brain autopsy, non-invasive diagnostic and imaging methods
used and a lack of expert diagnosticians. Revised estimates for 2015 draw on updated reviews and
computations but remain hampered by minimal evidence on dementia in SSA15. They identify AD as
causing 2.4% of DALYs and 3% of deaths in the region’s population aged 70 years and older11.
Concomitant with expanding NCD, SSA continues to face a major burden of communicable
diseases, including HIV/AIDS, as well as injury and trauma. This triple burden of disease implies
a critical challenge for SSA’s already under-resourced health systems16, 17. Given the constrained
health sector capacities in the region, services and responses for the prevention and management
of NCD remain highly circumscribed18. The lack of requisite health care may be particularly
pronounced among older adults who face a number of supply- but also demand-side barriers to
service use19, 20. Thus, for example, awareness of hypertension in the older population was found
to be extremely low in Ghana, South Africa and Central Africa, with rates of treatment and control
even lower21-23.
A crucial, yet often overlooked, factor in assessing older adults’ health and access to requisite care
in SSA is the major disparities that exist in this regard – and that reflect a wider heterogeneity in
the region’s older population. Research suggests that health inequalities among SSA’s older adults
emerge along several axes, including gender, socio-economic position and age, but also the place
of residence and the type of health condition. Studies such as the WHO Study on Global Ageing
and Adult Health (SAGE) in collaboration with the INDEPTH network of Health and Demographic
Surveillance sites (HDSS)24 or the HAALSI study (Health and Aging in Africa: longitudinal studies
of INDEPTH communities) point to a number of patterns in this regard. For example, in rural South
Africa older adults with chronic communicable disease and non-communicable disease were found
to report higher health care use than those with acute conditions25. In Ghana, higher wealth and
multimorbidity were significant predictors of public and private outpatient healthcare utilization
among older adults, with health insurance coverage a predictor only for the use of public facilities26.
Older Ghanaians with higher levels of functional limitations were more likely to rate their health
as being poorer, while wealthier individuals were more likely to rate their health as good27. In rural
South Africa, higher mortality among older adults was associated with lower socio-economic
status (assessed by household assets scores) as well as lower age (with greater death rates among
those aged 50-69 as compared with those aged 70-79 years)28. The HAALSI study reported worse
physical performance among older adults living in South Africa compared to those in the US,
Mexico and China, as well as a high cardiometabolic risk in this South African population29,30.
Over the longer term, the progressing epidemiological transition in SSA combined with thus far
deficient health systems’ responses to it portend a mounting burden of NCD, including dementia, in
the region’s expanding older population.
At the same time, social and economic change will likely entrench or modify patterns of health
inequality among older adults and will shape the regional, national, and local contexts within which
responses to unfolding health care and service needs must be forged.

ALZHEIMER’S DISEASE INTERNATIONAL14
1.3 Policy context and implications for research
The policy architecture on ageing and health in SSA has seen a remarkable consolidation in recent
years, with an emergence of relevant regional and national frameworks as well as African countries’
commitments to a number of overarching global agendas.
Chief among the global frames embraced by SSA governments are (i) the United Nations
Sustainable Development Goals (SDG), including SDG 3 to ‘ensure healthy lives for all at all ages’31,
and (ii) the recently adopted World Health Organization (WHO) Global Strategy on Ageing and
Health32. Grounded in the SDG, the Strategy commits Member States to action, among others, to
align health systems to the needs of older persons and to forge sustainable and equitable systems
for providing long-term care33.
Key Africa-regional instruments, which uphold the rights of older adults to health and access
to care include the African Union (AU) Protocol to the African Charter on Human and People’s
Rights on the Rights of Older Persons34 as well as the broader AU Agenda 206335, which sets
out an overall vision of development in the continent. Broad recommendations on action toward
enhancing the health and care of older adults are incorporated in the AU Policy Framework and
Plan of Action on Ageing (2002)36 as well as in the more recently adopted Common Africa Position
on Long-Term Care Systems (2017).
At country level such recommendations are broadly echoed in national policies or policy
frameworks on ageing that have been adopted in a growing number of SSA States. A few
countries, moreover, have enshrined key rights or entitlements of older citizens (including to health
and care) in their Constitutions or in dedicated legislation. Some States, finally, have included
varying provisions regarding older persons in sectoral plans or strategies on health or other
areas37, 38.
Despite such policy advances, comprehensive programmatic action on older people’s health
has remained sparse. Where they exist, interventions have centred on measures – such as fee
exemption schemes or social health insurance – intended to reduce financial barriers to older
adults’ service access. In some countries, community-based care schemes to address older
people’s health needs are being explored, often in conjunction with civil society-led initiatives.
A key impediment to the forging of more robust responses on older people’s health, and dementia
specifically, is the lack of consideration of such issues in priority agendas for health at both regional
and national levels. A case in point is the virtual absence of any reference to ageing or older
populations in the AU Africa Health Strategy 2016-2030.
In line with such gaps, SSA’s health systems remain oriented toward serving children or
reproductive age adults with little capacity to address old age-related health needs. Crucially, such
orientation extends to emerging NCD agendas in the region. These centre on primary, lifestyle-
focused prevention earlier in life of the ‘big four’ diseases (CVD, neoplasms, diabetes and chronic
obstructive lung disease), with little focus on the management of established conditions in later life,
and virtually no consideration of other diseases, such as dementia, that are of key concern to older
populations20, 39.
The present policy context suggests an urgent need for a concerted research effort to help
promote and inform a fuller consideration of dementia and other ageing and health challenges in
mainstream national and regional health agendas.
Two principal kinds of evidence are required to this end. First, evidence that helps to ‘clarify the
case’ on why SSA countries ought to address the health of older people, including dementia,
as part of their core health and development interests. Second, evidence to ‘clarify the specific
directions’ - in terms of target groups and settings, approaches or practice models that any such
responses should take. Specific areas and questions for research are highlighted in this report.

15DEMENTIA IN SUB-SAHARAN AFRICA
References
1. World Health Organization, Deme ntia: a public health
priority. 2012, World Health Organization: Geneva. p.
112.
2. Kinsella, K. and D. Phillips, T he Challenge of Globa l
Aging. Population Bulletin, 2005. 60(1).
3. United Nations, Department of Economic and Social
Affairs, and Population Division, World Population
Prospe cts: The 2015 Revision, Key Findings a nd
Advance Tables. 2015.
4. Keating, N., Critica l reflections on famil ies of old er
adults. Adv Gerontol, 2011. 24(2): p. 343-9.
5. Therborn, G., Afr ican fa milie s in a global context. 2006,
Afrikainstitutet: Uppsala, Sweden: Nordiska
6. Aboderin, I. and J. Hoffman, Families, Intergenerational
Bonds, a nd Agin g in Sub- Saha ran Afr ica. Can J Aging,
2015. 34(3): p. 282-9.
7. Aboderin, I., Ageing Africa. Opportu nities for
development, in Global age ing: per il or pro mise?, W.E.
Forum., Edito r. 2012: Geneva. p. 69-73.
8. Dalal, S., et al., Non-communicable diseases in sub-
Sahar an Afri ca: what we k now now. Int J Epidemiol,
2011. 40(4): p. 885-9 01.
9. Atun, R., et al., Diabetes in sub-Saharan Africa:
from cli nical c are to health po licy. Lancet Diabetes
Endocrinol, 2017. 5(8): p. 622-667.
10. Cappuccio, F.P. and M.A. Miller, Cardiovascular
disease and hypertension in sub-Saharan Africa:
burden, risk and interventions. Intern Emerg Med,
2016. 11(3): p. 299-30 5.
11. Global Burden of Disease Study, GBD Results Tool.
2015: Washington.
12. Prince, M.J., et al., The burden of diseas e in olde r
people and implicatio ns for he alth policy and practic e.
Lancet, 2015. 385(9967): p. 549-62.
13. G.B.D. Disease Injur y Incidence Prevalence
Collaborators, Global, regional, and national incidence,
prevalence, and year s lived with disability for 310
disea ses and injuries, 1990 -2015: a systematic a nalysis
for the Global Bu rden of D isease Study 2015. Lancet,
2016. 388(10053): p. 1545-1602.
14. Rowe, J.S., et al., Alzheimer’s disease and other
dementia in sub-Saharan Africa and the Caribbean. J
Am Geriatr Soc, 2014. 62(10): p. 1991-2.
15. Prince, M., et al., World Alz heime r Report 2015 - The
Global Impact of dementia: an analysis of prevalence,
incidence, costs and trends. 2015: London.
16. African Union, Africa Health Strategy, 2016-2030. 2016.
17. Nyirenda, M.J., Non-communicable diseases in
sub-S ahar an Africa: unde rsta nding th e drivers of the
epidemic to inform intervention strategies. Int Health,
2016. 8(3): p. 157-8.
18. Checkley, W., et al., Man agement of NCD in l ow- and
middle-income countries. Glob Heart, 2014. 9(4): p.
431-43.
19. Gomez-Olive, F.X., et al., Self-reported health
and hea lth car e use in an ageing p opula tion in th e
Aginco urt sub-district of rura l South Africa. Glob
Health Ac tion, 2013. 6: p. 19305.
20. Ab oderin, I.A. a nd J.R. Beard, Old er peop le’s health
in sub-Saharan Africa. Lancet, 2015. 385(9968): p. e9-
e11.
21. Lloyd-Sherlock, P., et al., Hypertension among
older adults in l ow- and middle- incom e countries:
prevalence, awareness and control. Int J Epidemiol,
2014. 43(1): p. 116-28.
22. Pilleron, S., et al., Prevalence, awareness, treatment,
and control of hypertension in older people in Centra l
Africa: the EPIDEMCA study. J Am Soc Hypertens,
2017. 11(7): p. 449-460.
23. Gomez-Olive, F.X., et al., Stark Regional and S ex
Differences in the Prevalence and Awareness of
Hyper tension: An H3Africa AWI- Gen Study Acros s 6
Sites in Sub-Saharan Africa. Glob Heart, 2017.
24. Kowal, P., et al., Data resou rce profile: the Wor ld Heal th
Organ izatio n Study on g lobal AGEing and adult health
(SAGE). Int J Epidemiol, 2012. 41(6): p. 1639-49.
25. Ameh, S., et al., Pr edictors of he alth ca re use by adults
50 years and over in a r ural S outh Afr ican setting. Glob
Health Ac tion, 2014. 7(1): p. 24771.
26. Awoke, M.A., et al., Predictors of public and private
healthcare utilization and associated health system
responsiveness among older adults in Ghana. Glob
Health Ac tion, 2017. 10(1): p. 1301723.
27. Debpuur, C., et al., Self-reported health and functional
limitations am ong older peop le in the Ka ssena-
Nankana District, Ghana. Glob Health Action, 2010. 3.
28. Gomez-Olive, F.X., et al., Social conditions and
disab ility r elate d to the mor tality of old er people in
rural South Afr ica. Int J Epidemiol, 2014. 43(5): p. 1531-
41.
29. Payne, C.F., et al., Physical Fu nction in an Aging
Population in Rural Sou th Afric a: Findi ngs From
HAALSI a nd Cros s-National C omparisons With HRS
Sister Studies. J Gerontol B Psychol Sci Soc Sci, 2017.
30. Ga ziano, T.A., et al., Cardiomet abolic risk in a
population of older adults with multiple co-morbidities
in rura l south africa: the HAALSI (Health and Ag ing in
Africa: longitudinal studie s of INDEPT H commu nities)
stu dy. BMC Public Health, 2017. 17(1): p. 206.
31. United Nations, Transforming o ur world: the 2030
agenda for sustainable development. 2015, United
Nations: New York.
32. World Health Organization, D raft g lobal strategy and
action pl an on ageing and health. 2015: Geneva,
Switzerland.
33. World Health Organization, Multisectoral action for a
life cou rse ap proac h to healthy ageing: global strategy
and plan of action o n ageing and health 2016, World
Health Organization: Geneva.
34. African Union, Proto col to the Africa Charter on Hum an
and Peop les’ Rights on the R ights of O lder Pe rsons in
Africa. 2016 .
35. African Union, Agenda 2063: The Africa we want. 2 013.
36. African Union, AU Policy Framewo rk and Plan of Actio n
on Agein g 2002.
37. Ab oderin, I., Under standing and advanc ing the health
of older populations in sub-saharan Africa: policy
perspectives and evidence needs. . Public Health
Reviews 2011. 32(2): p. 357-376.
38. HelpAge/UNFPA, Ageing in the 21st Century.
A celebration a nd a challenge. 2012, HelpAge
International and UNFPA: London and New York.
39. Juma, P.A., et al., Analysis of Non-communicable
disea se preve ntion po licie s in five Sub -Sah aran Af rican
countries: Study protocol. Arch Public Health, 2016.
74: p. 25.

ALZHEIMER’S DISEASE INTERNATIONAL16
CHAPTER 2
The epidemiological evidence:
prevalence, costs, incidence,
mortality and risk factors
2.1 Introduction
In 2005, ADI convened a group of international experts to generate up-to-date evidence-based
estimates for the prevalence and numbers of people living with dementia in all regions of the
world1. Those figures have provided a good foundation for global policymaking and assisted
national Alzheimer’s associations in raising awareness of the upcoming challenge that would be
faced by current and future generations.
At that time, the evidence on dementia was incomplete and scanty in many regions of the world,
especially in low-and-middle income countries (LMIC). Only three epidemiological studies were
identified for the entire African continent. As a result, the prevalence estimate for the African region
was influenced by the one good quality study, performed in Nigeria, which reported a strikingly low
prevalence of dementia2. The consensus therefore estimated the prevalence of dementia for the
WHO AFRO (D/E) region to be 1.6% at age over 60 years, representing 0.5 million people living with
dementia in this region in 2001. The projected increase over time was however important: 234%
between 2001 and 2040, leading to an estimated 1.6 million people living with dementia in Africa in
2040. Africa was thus, as many other LMIC, starting from a low base but expected to experience
a rapid increase in numbers. The rate of increase was generally 3-4 times higher in developing
compared to developed regions1.
Hypotheses have been generated to explain these low prevalences (and the gap between
developing/developed regions), including a low survival with dementia, low incidence (i.e. the rate
at which new cases develop), a lower level of exposure to environmental risk factors (particularly
cardiovascular risk and hyperlipidaemia), and a higher mortality in early life (conferring a ‘survival
advantage’, potentially protecting against neurodegeneration or delaying the onset of clinical
manifestations). Given the lack of evidence, the need for more research in Africa was obvious.

17DEMENTIA IN SUB-SAHARAN AFRICA
A few years later, the 2009 World Alzheimer Report3 gave a further overview and analysis of the
situation, based on the research data available at that time. Systematic reviews of the evidence
were performed and global estimates for dementia prevalence and numbers were generated and
updated. The evidence available on dementia prevalence in sub-Saharan Africa (SSA) had not
changed much; coverage was still sparse and prevalence estimates were still primarily relying on
the consensus estimates from 2005. However, despite relatively low standardised prevalence,
higher estimates of the number of people living with dementia in SSA were provided (0.71 million
in 2010, increasing to 2.54 million in 2050)3, mainly driven by the demographic ageing of the
population. The WHO report (‘Dementia: a public health priority’) which followed in 2012 completed
the picture by reviewing the evidence on the incidence of dementia. At that time, a single incidence
study was identified, conducted in Nigeria4. It was then estimated that 525,818 new cases of
dementia were arising each year in Africa, with more than half of them (293,191) living in SSA5.
In 2013, in preparation for the G8 Dementia Summit held in London, the Global Observatory for
Ageing and Dementia Care and ADI worked in a joint effort on a policy brief reassessing and
updating the evidence on the scale and the distribution of the global dementia epidemic6. The
evidence from the sub-Saharan regions had significantly increased and prevalence estimates were
higher than previously reported, with an age-standardised prevalence in SSA regions from 2.07
- 4.00% to 4.67%. Hence, it was estimated that in 2013, 1.3 million older people were living with
dementia in SSA, reaching 5.05 million in 20506.
Pursuing their efforts, a complete update of the global impact of dementia was published in the
2015 World Alzheimer Report7. The significant improvement in evidence coverage on dementia
prevalence in African regions since 2009 was highlighted, as was the high quality of the new
evidence generated. Prevalence and numbers of people living with dementia in Africa did not vary
much between 2013 and 2015 as most of the new data on dementia prevalence were published
between 2009 and 2013. However, accounting for the revised population estimates, the numbers of
people living with dementia in SSA were estimated at 1.69 million in 2015, increasing to 2.64 million
in 2030 and 5.72 million in 2050 (representing a 145 to 300% proportionate increase between
2015-2050).
The coverage regarding the incidence of dementia was not different from the previous WHO review,
with still a single study representing SSA. However, considering revised population estimates, the
estimated number of incident cases of dementia in SSA was 446,568 in 2015, representing 4.5% of
the global number of incident cases7.
Although several reviews on the prevalence and epidemiology of dementia in SSA have been
published8-11, their methods and inclusion/exclusion criteria were not always consistent, hence
providing varying estimates. In this chapter, we therefore provide an overview of the epidemiology
and impact of dementia in SSA, using essentially the same methods as previously used in the
World Alzheimer Reports, conducting new reviews of the studies regarding dementia prevalence,
incidence and mortality. We will also present updated estimates of the numbers of people living
with dementia in this region, as well as costs. The evidence on risk factors of dementia will also be
reviewed, with a specific note on the HIV-associated neurocognitive disorders.
2.2 Prevalence
2.2.1 Methods
We conducted a systematic review of the literature on the prevalence of dementia in different
databases following the strategy below.
• Search date: May 2017
• Databases: Embase, Global Health, Ovid MEDLINE® In-Process & Other Non-Indexed
Citations and Ovid MEDLINE®, PsycARTICLES Full Text & PsycINFO
• Search terms: (Africa OR individual countries of Africa) AND (epidemiolog* or prevalence)
AND (dement* or alzheime*)
Inclusion criteria: population-based studies of the prevalence of dementia among people aged
60 years and over (according to DSM-IV or ICD-10 criteria, or similar clinical criteria), for which the
fieldwork started on or after 1st January 1980.

ALZHEIMER’S DISEASE INTERNATIONAL18
Exclusion criteria (related to sampling, case ascertainment procedures, and outcome definitions):
A. Sampling design
• Studies of prevalence from the follow-up phase (rather than the inception phase) of a
population cohort
• Studies sampling from an out-of-date population register (prepared more than three years
prior to the survey)
• Studies of nursing home or residential care populations, primary care attendees or other
unrepresentative service-user populations
B. Ascertainment/outcome definition
• Studies in which the ascertainment of dementia depended upon help-seeking and/or receipt
of dementia care services
• Studies in which ‘dementia’ was diagnosed purely on the basis of cognitive impairment, for
example according to a cut point on the MMSE
• Two phase studies, in which screening procedures were clearly inadequate, especially all
large-scale screening studies of people of all ages for all neurological disorders
• Studies of the prevalence of Alzheimer’s disease or other subtypes of dementia, or restricted
to young onset dementia
• Procedures
Two independent reviewers (MG and AMP) read the abstracts of all publications identified on the
electronic databases, excluding only those that clearly did not meet the above criteria. In the next
stage, printed copies of the remaining publications were read, and a consensus was made on
those that met all criteria. We read studies published in English and French. All eligible studies were
systematically coded for their study design and quality according to the following criteria:
1. Country
2. WHO/Global Burden of Disease World Region (see Appendix A for list of countries and
regions)
3. Inclusion of residents of long term care institutions
4. Start and finish dates for fieldwork, and census dates if provided
5. Lower and upper age limits
6. Sampling (simple random, stratified random, whole population, other)
7. Design (one phase/two phase/three phase)
8. Overall sample size (first phase)
9. Numbers interviewed (first phase) and proportion responding
10. For two-phase surveys only
a. Numbers selected for the second phase (for two phase surveys)
b. Numbers interviewed (second phase) and proportion responding
c. Screen negatives sampled for the second phase (yes/no)
d. Screen negatives given same assessment as screen positives (yes/no)
e. Weighting back carried out (no weighting back/ appropriate weighting back/ no weighting
back, but no false positives)
f. Time interval between first and second phase
g. Screening instrument/s
11. Diagnostic criteria (not specified, ICD, DSM, GMS/AGECAT, CAMDEX, other clinical
criteria)
12. Use of multidomain cognitive assessment, informant interview, disability assessment,
neuroimaging
An overall quality score was derived by summing scores for the following elements:
Sample size – <500, 0.5 points; 500-1499, 1 point; 1500-2999, 1.5 points; >=3000 2 points
Design – Two-phase study with no sampling of screen negatives, 0 points; two-phase study with
sampling of screen negatives but no weighting back, 1 point; one-phase study or two-phase study
with appropriate sampling and weighting, 2 points

19DEMENTIA IN SUB-SAHARAN AFRICA
Response proportion – <60%, 1 point; 60-79%, 2 points; >=80%, 3 points
Diagnostic assessment – one point each for multidomain cognitive test battery, formal disability
assessment, informant interview and clinical interview
Data extraction
For studies reporting unweighted prevalence, we extracted either numerator and denominator, or
prevalence and denominator, or prevalence and standard error, or prevalence and 95% confidence
intervals. Numerator and denominator could then be calculated from any of these combinations.
For studies reporting weighted prevalence we extracted either weighted prevalence and weighted
standard error or weighted 95% confidence intervals. Effective numerators and denominators
(taking account of the design effect) could then be calculated.
Prevalence estimates were stratified differently in different publications. Age-specific prevalence
data could generally be calculated from age- and sex-specific estimates. If studies provided age-
specific estimates not stratified by sex, sex-specific estimates not stratified by age or an overall
prevalence stratified neither by age nor sex, we contacted the authors to request age- and sex-
specific estimates. We could therefore model the effect of age on dementia prevalence for all
included studies, and the effects of age and sex for the subset of studies that had provided age-
and sex-specific estimates.
Meta-analytical methods for estimating dementia prevalence
We used a random effect exponential (Poisson) model to assess the effects of age and sex on the
prevalence of dementia in SSA. Random effects are assumed to have a gamma distribution – the
alpha coefficient is an estimate of overdispersion and an index of between study heterogeneity.
Age was coded as the mean for each age group reported, this was calculated from the mean
observed in 4 population-based studies in West and Central Africa for which individual data was
available12-14. We ran two models, one for the effect of age, and one for the main effects of age and
sex, and an interaction between age and sex. We then applied the relevant mean ages and sex
codings to the coefficients estimated from the models, to estimate prevalence in five year age-
bands from 65-84 years, and for those aged 85 and over, for both sexes combined.
To facilitate comparison between regions, we have calculated overall age-specific and age- and
sex-specific standardised prevalence for all those aged 60 and over, using Western Europe as the
standard population.
Estimation of the numbers of people with dementia
As described earlier, we generated both age-specific and age- and sex-specific meta-analysed
dementia prevalence estimates for sub-Saharan Africa (including West, East, Central and Southern
regions). Numbers of people living with dementia were estimated by applying the age- and sex-
specific prevalence of dementia for each region to the UN population estimates for 2015 and the
population projections until 205015.
2.2.2 Results
The initial search yielded abstracts for 1007 publications, of which 328 were discarded as
duplicates leaving 678 abstracts for review. After reading the abstracts, 653 publications were
excluded as clearly ineligible (not carried out in SSA and/or before 1980, reviews, case or hospital-
based studies, studies on other neurological or psychiatric disorders), leaving 21 publications for
further review. We obtained copies of the full published version of each paper, which were then
carefully assessed against inclusion/exclusion criteria and the citation of other relevant studies in
SSA.
A further 10 publications were excluded at this stage (8 were already reported elsewhere, 1 was
not matching the criteria, and 1 was a conference abstract only), leaving 8 publications (describing
10 studies) that were eligible for inclusion in the review. Two further studies were identified through
conference abstracts and/or personal contacts with the 10/66 Dementia Research Group.
For 1 of these publications, we requested more information from the authors in order to be able to
use the data in the meta-analysis in another form than in which it was provided in the publication.
Finally, 12 studies were fully eligible for inclusion.

ALZHEIMER’S DISEASE INTERNATIONAL20
Coverage
Reasonable to good coverage was identified for West SSA (6 studies) and Central Africa (4 studies).
Sparse coverage was achieved for East and Southern SSA, still represented by a single study each
(Tanzania and South Africa) (Table 2.1).
Although improvements in coverage have been seen since the 2009 World Alzheimer Report for
West SSA and Central SSA, Eastern and Southern SSA are still lacking research relative to their
population size.
Quality
The principal characteristics of the included studies are given in Table 2.3, by region.
Most of the studies carried out in SSA have used a multistage design (2-3 stages). If multiphase
surveys are reputed to be more cost- and time-efficient in epidemiological research, it can be
a threat to quality when not implemented appropriately. The common error is to fail to submit
a random sample of participants who have scored above the cutpoint during the screening
assessment (‘screen-negatives’) to the same diagnostic assessment as those who scored below
the cutpoint (‘screen-positives’). The correct procedure, estimating the false positive rate before
‘weighting back’, allows the overall prevalence to be calculated accounting for the different
sampling proportions of screen positive and screen negatives. Unfortunately, in half of the
multiphase studies identified for the region, investigators have failed to apply the correct procedure
including the assessment of screen negatives. In those studies, the results will tend towards an
under-estimation of true dementia prevalence and an over-estimation of precision.
Multiphase designs can also generally be affected by participant loss to follow-up between the
different stages of the study (screening and diagnostic assessment). This would tend to over- or
under- estimate the true prevalence. Such bias can be avoided by the use of a single stage design
where dementia diagnosis is performed through a computerized algorithm, like the one developed
by the 10/66 Dementia Research Group. Although specifically developed for low- and middle-
income countries where low levels of education are common and specialists able to diagnose
dementia are lacking, this one-phase design has been little used so far in SSA.
Evidence of cognitive impairment (and decline compared to previous level of functioning) in
memory and other domains of cognitive function, as well as consequent social or occupational
impairment, are key in establishing a dementia diagnosis. Other potential causes of social and
cognitive impairment, such as psychological comorbidities (depression, psychosis, or delirium)
have to be excluded. Including multi-domain cognitive testing, disability assessment, and clinical as
well as informant interviews is therefore necessary in a thorough dementia diagnostic assessment.
All the studies included have done so, using a range of different instruments and scales. Although
the effect of applying such diagnostic assessments on dementia prevalence is not known, the
general quality of the studies is improved.
Most of the eligible studies (n=8) had a sample size of between 500-1499, with Central Africa being
the region where the sample size was consistently around 500 participants. Sample sizes tended to
be larger in West Africa. Although sample size should not bias prevalence estimates in a systematic
way, larger studies will estimate prevalence with a better precision, and are therefore preferable.
Studies with higher proportions of participants responding provide more accurate prevalence
estimates. Participants included in a survey might differ from those who could not be contacted
Table 2.1
Coverage, by region, with respect to size of the older population
Region Over 60
year old
population
(millions)
Number
of eligible
dementia
prevalence
studies
Number of
studies/
10 million
population
Tot al
population
studied
Total population
studied/ million
population
SSA, Central 5.21 4 7.7 3020 580
SSA, East 20.08 1 0.5 1198 60
SSA, Southern 5.36 11.9 206 38
SSA, West 17.70 63.4 5925 335

21DEMENTIA IN SUB-SAHARAN AFRICA
or did not consent to take part in the survey. High response proportions may avoid two different
situations where bias would affect the estimation of dementia prevalence. People with dementia
could be under-represented in the sample if relatives are reluctant for them to participate or
because they find it more difficult to complete the interview / assessment. On the other hand,
people with dementia could be over-represented due to an increased availability at the time of the
survey (less likely to be out of the house working or carrying out other activities). Participation rates
in the included studies are very high, mostly over 90%, exceeding that achieved in other world
regions7.
As already highlighted in the 2015 World Alzheimer Report, overall study quality was especially high
in sub-Saharan Africa, despite the lack of appropriate use of multiphase design and the somewhat
small sample size in some regions.
Prevalence
We considered the evidence-base to be sufficient and generalizable to conduct a meta-analysis for
the sub-Saharan African region combining the central, southern, eastern, and western sub-regions.
The prevalence of dementia increased exponentially with age, doubling with every 7.2 year
increment in age. For both sexes combined, it increased from 1.74% at 60-64 years to 23.31%
at 85+ in SSA (Table 2.2). Prevalence of dementia tended to be lower for men in SSA, however
this effect wasn’t significant. Significant heterogeneity in age-specific and age- and sex-specific
prevalence between studies within SSA was found (alpha=0.43/0.31).
Both the age- and the age- and sex-specific dementia prevalences generated were slightly higher
than the ones presented in 2015. Standardised prevalence for those aged 60 and over was also
higher than previously described, 6.38% instead of 4.63% for the age-specific and 7.23% instead
of 5.47% for the age and sex-specific prevalence.
The meta-analysis of dementia prevalence included three additional studies compared to the one
realised in 2015, two from Nigeria and one from South Africa. The cross-culturally validated 10/66
dementia criteria were used in two of those additional studies, while all the other studies used the
DSM (III or IV) criteria. A higher prevalence of dementia was recorded in studies using the 10/66
dementia criteria compared to the DSM-IV diagnostic criteria. When considering only the studies
which diagnosed dementia according to the DSM (III or IV) criteria, the age- and age- and sex-
specific prevalence of dementia were slightly lower, closer to those estimated in 2015 (Table 2.2).
Heterogeneity in age-specific and age- and sex-specific prevalence between studies was reduced
(alpha=0.26/0.18). The standardised prevalence of dementia was similar to our previous estimates
(5.47%)7.
Estimation of the number of people with dementia
Applying the age- and sex-specific prevalence generated earlier (which should provide the most
precise overall prediction of regional prevalence) to the UN population projections15, we estimate
that 2.13 million people were living with dementia in 2015 (Table 2.4). This number will reach 3.48
million by 2030, and 7.62 million in 2050. Between 2015 and 2050, numbers in the SSA region will
increase by 257%, with the most important increases in East and Central SSA.
Table 2.2
Estimates of dementia prevalence (%) in sub-Saharan Africa
Sub-Saharan
Africa
60-64 65-69 70-74 75 -79 80-84 85+ Age-standardised
prevalence for all
those aged 60 years
and over
All studies
(n=12)
Male 1.3 4 2.02 3.09 5.05 7. 7 7 17. 2 2 7.23 *
Female 2.43 3.68 5.66 9.30 14. 35 32.07
All 1.74 2.64 4.07 6.71 10 .3 8 23.31 6.38
DSM criteria
only (n=10)
Male 1.16 1.70 2.51 3.95 5.86 12.20 5.50*
Female 1.74 2.66 4.16 6.94 10.86 24.90
All 1.23 1. 89 2.94 4.91 7.6 8 17.60 4.71
*2 studies were not included in the meta-analysis because they did not provide age- and sex- specific prevalence

ALZHEIMER’S DISEASE INTERNATIONAL22 23DEMENTIA IN SUB-SAHARAN AFRICA
Table 2.3
Study characteristics and quality, by region
Reference, by
GBD region
Country,
region/city
Rural /
urban area
Design Sampling Residents of
institutions
Lower
& upper
age
limits
Dates of
fieldwork
Sample
size
Numbers
interviewed
(proportion
responding)
Screening
instruments
Diagnostic
criteria
Second phase Multidomain cognitive
assessment? Disability?
Neuroimaging?
Time
interval
Numbers
selected
Number
interviewed
Screen
negatives
Screen
negatives
Assessment
Weighting
back
SSA West                 
Hendrie et al.,
199 5 2
Nigeria,
Ibadan
urban 2 stage Door-to-door No ≥ 65 01/05/92-
31/03/94
2535 2494 (98.4%) CSI-D DSM III-R/
ICD10
10
months
279 (423) 214 Ye s Same Yes Clinical assessment / ADL /
CT scans
Uwakwe
and 10/ 66
Dementia
Research
Group**
Nigeria,
Anambra
rural 1Catchment area* No ≥ 65 2003-
2007
123 8 53 8 (43.4%
of complete
interviews)
CSI-D, GMS,
10 words
10/66 n/a n/a n/a n/a n/a n/a WHO-DAS
Guerchet et al.,
200912
Benin, Djidja rural 2 stage Catchment area* No ≥ 65 03 / 07-
04/07
514 5 0 2 ( 9 7. 6 % ) CSI-D DSM-IV 1 month 52 47 Yes Same No Clinical assessment /
Psychometric tests / ADL / CDR
Yusuf et al.,
20 1116
Nigeria,
Zaria
urban 1 stage Systematic
random
sampling
No ≥ 65 0 3/07-
10/07
322 322 (100 %) CSI-D, CER AD,
BDS
DSM-IV /
ICD10
n/a n/a n/a n/a n /a n/a CER AD / S tick Design Test
Paraiso et al.,
20 1114
Benin,
Cotonou
urban 2 stage Random
sampling
(Proportional)
No ≥ 65 05/08-
09/08
1162 1139 ( 9 8.0% ) CSI-D DSM-IV 2 months 154 147 No n/a n/a Clinical assessment /
Psychometric tests / ADL / CDR
Ogunniyi et al.,
201617
Nigeria,
Lalupon
rural 2 stage Catchment area* No ≥ 65 05 /13-
02 /14
642 642 (100 %) IDE A DSM-IV 2-8
months
175 146 Ye s Same No Clinical assessment / CHIF
SSA Central                 
Guerchet et al.,
201013
CAR, Bangui urban 2 stage Catchment area* No ≥ 65 10/09 -
12/0 9
509 49 6 ( 9 7. 4%) CSI-D / Five-
word test
DSM-IV 1 month 188 16 5 No n/a n/a Clinical assessment /
Psychometric tests / ADL / NPI
/ CDR
Congo,
Brazzaville
urban 2 stage Catchment area* No ≥ 65 01/10-
03/10
546 520 (95.2%) CSI-D / Five-
word test
DSM-IV 1 month 148 13 3 No n/a n /a Clinical assessment /
Psychometric tests / ADL / NPI
/ CDR
Guerchet et al.,
2013 18
CAR, Nola rural 2 stage Catchment area* No ≥ 65 11 /11-
03 /12
501 473 (9 4.4%) CSI-D DSM-IV 2 months 18 2 117 No n/a n/a Clinical assessment /
Psychometric tests / ADL / NPI
/ CDR
CAR, Bangui urban 2 stage Random
sampling
(Proportional)
No ≥ 65 01/12-
03 /12
514 5 0 0 ( 9 7. 3 %) CSI-D DSM-IV 1 month 161 115 No n/a n/a Clinical assessment /
Psychometric tests / ADL / NPI
/ CDR
Guerchet et al.,
2013 18
Congo,
Gamboma
rural 2 stage Catchment area* No ≥ 65 08 /12-
12/12
529 520 (94.3%) CSI-D DSM-IV 1.5
months
284 193 No n/a n/a Clinical assessment /
Psychometric tests / ADL / NPI
/ CDR
Congo,
Brazzaville
urban 2 stage Random
sampling
(Proportional)
No ≥ 65 0 9/12 -
12/12
537 500 (9 3 .1% ) CSI-D DSM-IV 3 months 148 131 No n/a n/a Clinical assessment /
Psychometric tests / ADL / NPI
/ CDR
SSA East                 
Longdon et al.,
2013 19
Tanzania,
Hai
rural 2 stage Catchment area* No ≥ 70 Not stated 1260 1198 (95.1%) CSI-D DSM-IV Not
stated
312 296 Yes Same Yes CER AD
SSA South                  
Vanderpoel et
al., 2013 20
South
Africa,
Muangang
urban 1Catchment area* No ≥ 65 2003-
2007
230 206 (89.6%) CSI-D, GMS,
10 words
10/66 n/a n/a n/a n/a n/a n/a WHO-DAS
* Door-to-door, ** Unpublished, 2004 – 20 06
MMSE = Mini Mental St ate Examination (-il for the illiterate version)
CSI-D = Community Screening Interview for Dementia
GMS = Geriatric Mental State
DEA = Identification and Intervention for Dementia in Elderly Africans
CERAD = Consortium to Establish a Registry for Alzheimer’s Disease
BDS = Blessed Dementia Scale
CHIF = Clinician home-based assessment of impairment in functioning
ADL = Activities of Daily Living
NPI = Neuropsychiatric Inventory
WHO-DAS = WHO Disability Assessment Scale
CDR = Clinical Dementia Rating scale
BOMC = Blessed Orientation-Memory-Concentration Test
BDRS-CER AD = Blessed Dementia Rating Scale - Consortium for the
Establishment of a Registry of Alzheimer’s Disease version
ADAS = Alzheimer’s Disease Assessment Scale
Psychometric tests = Free and Cued Reminding Test, Zazzo’s cancellation
Task, Isaac’s Set Test of verbal fluency

ALZHEIMER’S DISEASE INTERNATIONAL24
These new estimates are slightly higher than those reported in the 2015 World Alzheimer Report
(1.63 million in 2015, 2.65 million in 2030 and 5.66 million in 2050)7. This increase is driven by both
population ageing and the slightly higher prevalence of dementia estimated for SSA. The African
region started from a low base but will continue to experience a rapid increase in numbers.
2.3 Costs
The global cost of dementia, estimated at US$ 818 billion in 2015, is unevenly distributed with 87%
incurred in high income countries while 58% of people with dementia were living in low and middle
income countries7. The increase of the global societal costs of dementia reported between 2010
and 2015 was given multiple plausible explanations, including the increases in numbers of people
with dementia and/or increases in per person costs.
As estimated costs of dementia depend on the estimated numbers of people living with dementia
in each region, we revised the estimation of the costs of dementia in sub-Saharan Africa in 2015.
2.3.1 Methods
The estimates of the societal economic cost of dementia provided in this report have been
generated using the same general approach as for the 2010 and 2015 World Alzheimer Reports.
Costs are estimated at the country level and then aggregated in various combinations to
summarise cost by Global Burden of Disease region. For each country, there is a cost per person
(per capita) estimate which is then multiplied by the number of people estimated to be living with
dementia in that country. The per capita costs are divided into three cost sub-categories: direct
medical costs, direct social care costs (paid and professional home care, and residential and
nursing home care) and costs of informal (unpaid) care. The base option for costing informal care is
an opportunity cost approach, valuing hours of informal care by the average wage for each country.
As in 2015, for countries with no cost data, cost estimations are derived by imputation. The
assumption for the imputation is that there is a relationship between a country’s per capita GDP
and annual per capita direct costs of dementia. For low and middle income countries (LMIC), the
partitioning of the imputed total direct costs into direct medical and social care sector costs was
derived from a relevant study from China21 and more information available from the 10/66 Cost
of Illness (COI) studies (China, India, Cuba, Peru, Venezuela, Dominican Republic and Mexico)22,
where two-thirds of the direct costs were medical and one-third derived from the social care
sector. These proportions were used as a basis for imputation in many African countries. Data
from SSA are still lacking. Compared to the estimates provided in 2015, the update of the costs of
dementia in SSA is therefore purely based on the new prevalence figures presented earlier.
To make a forecast on future trends in the societal cost of dementia, we need to estimate trends
in the numbers of people living with dementia and trends in the per person costs. Considering the
lack of empirical data from SSA regarding resources use and costs, providing forecasts of costs of
dementia in the future would only rely on imputations, which is why no projections have been made
for this report.
Table 2.4
Total population over 60, crude estimated prevalence of dementia (2015), estimated number of people with de-
mentia (2015, 2030, 2050) and proportionate increases (2015-2030, 2015-2050)
GBD Region Over 60
population
(millions,
2015)
Crude estimated
prevalence (%,
2015)
Number of people with
dementia (millions)
Proportionate increases
(%)
2015 2030 2050 2015-2030 2015-2050
SSA, Central 5.21 4.41 0.23 0.39 0.85 70 269
SSA, East 20.08 4.53 0.91 1.57 3.62 72 298
SSA, Southern 5.36 5.41 0.29 0.41 0.76 41 162
SSA, West 17.70 3.95 0.70 1.11 2.39 59 241
Sub-Saharan Africa 48.35 4.40 2.13 3.48 7.6 2 63 257

25DEMENTIA IN SUB-SAHARAN AFRICA
2.3.2 Results
The total costs of dementia in sub-Saharan Africa were estimated at US$ 6.2 billion in 2015 (Table
2.5). Our previous estimation was slightly lower (US$ 4.9 billion) - the difference is only explained
by the revision of the numbers of people living with dementia in SSA.
The pattern of distribution of costs between the three major sub-categories (direct medical,
social care and informal care) has not changed. As often reported for low or lower-middle income
countries, the proportional contribution of direct medical care is modest (19.7% overall) and the
contribution of social care costs is low (9.7%). The major contribution to the total costs is attributed
to informal care costs (two-thirds to three-quarters). Some of the greatest contributions of informal
care were already reported in the African region in the 2010 and 2015 World Alzheimer Reports3, 7.
Per person costs of dementia did not change compared to those presented in the 2015 World
Alzheimer Report, as they do not relate directly to the update of the numbers of people living with
dementia.
As for societal costs of dementia, the greater contributions to the per person costs were those of
informal care costs (more than half for all SSA regions).
Although the costs of dementia (by sub-category and per person) seem higher in Southern Africa
than other regions, those figures need to be interpreted with caution as they are driven by a higher
GDP per capita (which is used for imputations) in South Africa. However, income and wealth in
South Africa are very inequitably distributed.
2.4 Incidence
2.4.1 Methods
We recently updated the systematic review of the world literature conducted for the 2015 World
Alzheimer Report7, identifying population-based studies of the incidence of dementia (defined
according to DSM-IV, ICD-10 or similar clinical criteria), including people aged 60 years and over,
and for which the field work for the baseline phase started on or after 1st January 1980. The search
followed the strategy below:
• Search date: February 2017
Table 2.5
Sub-category costs of dementia in 2015 (billion US$, and percent of total costs), by GBD region
Direct medical costs Direct social sector
costs
Informal care costs Total
costs
US$
(billions)
Per cent US $
(billions)
Per cent US $
(billions)
Per cent US $
(billions)
SSA, Central 0.1 28.5% 0.1 14.2% 0.2 57.3% 0.4
SSA, East 0.4 20.8% 0.2 10.3% 1.3 68.9% 1.9
SSA, Southern 0.5 16.4% 0.2 8.0% 2 .1 75.6% 2.8
SSA, West 0.2 22.8% 0.1 11. 2 % 0.7 66.0% 1.0
Sub-Saharan Africa 1.2 19.7 % 0.6 9.7% 4.3 70.6% 6.2
Table 2.6
Costs of dementia per person, in categories, in 2015 (billion US$, and percent of total costs), by GBD region
Direct medical costs Direct social sector
costs
Informal care costs Tota l co sts
US$ Per cent US$ Per cent US$ Per cent US$
SSA, Central 536 28.5% 266 14 .1% 1078 5 7. 4 % 1880
SSA, East 442 20.8% 219 10.3% 1460 68.9% 212 0
SSA, Southern 15 52 16.4% 768 8.1% 717 0 75.5% 9490
SSA, West 338 22.8% 167 11.2% 978 66.0% 1480

ALZHEIMER’S DISEASE INTERNATIONAL26
• Databases: EMBASE, Global Health, MEDLINE, PsychExtra and PsychInfo
• Search terms: dementia AND (incidence OR epidemiology)
Again, all stages of the search were completed by two independent reviewers (MG and AMP).
Papers were excluded at this stage only when the abstract clearly demonstrated that the paper
did not meet the above criteria. After a consensus was made on studies that met all criteria,
publications from sub-Saharan Africa were selected for inclusion in this report.
Data extraction
All eligible studies were systematically coded for their study design and characteristics according
to the following criteria:
1. Country
2. WHO / Global Burden of Disease World Region (see Appendix A for list of countries and
regions)
3. Inclusion of urban or rural areas
4. Start and finish dates for fieldwork
5. Lower and upper age limits
6. Sampling strategy (whole population, catchment area, random sampling, stratified random
sampling)
7. Overall sample size
8. Response rate
9. Case ascertainment (community survey only or community + institution survey)
10. Diagnostic criteria (not specified, ICD, DSM, GMS/AGECAT, CAMDEX, other clinical
criteria)
11. Presence of clinical diagnosis
12. Diagnostic Instruments (GMS/AGECAT, CAMDEX, MMSE, Dementia Differential
Scale, Hachinski Ischemic Index, consensus panel, physical/neurological examination,
standardized questionnaire, clinical evaluation, other).
Incidence data was extracted from the studies as follows. According to the data presented in the
paper, we extracted numerator (case) and denominator (person-years), incidence and standard
error, or incidence and 95% confidence intervals.
Incidence estimates were stratified differently in different publications. To maximise the precision of
our meta-analysis, we required incidence estimates in five-year age-bands. If this was not available
in the publication, we wrote to the authors to request age-specific incidence data. We could
therefore model the effect of age on dementia incidence for all included studies.
Meta-analytical methods for estimating dementia incidence
As for the meta-analysis of prevalence data, we used a random effect exponential (Poisson) model
to assess the effect of age on the incidence of dementia. The alpha coefficient is an estimate of
over-dispersion and an index of between study heterogeneity. Age was coded as the mean for
each age group reported.
2.4.2 Results
While the 2015 World Alzheimer Report included only two studies from a single SSA country:
Nigeria4, 23, three new potentially eligible incidence studies were identified. Two were published
between 2015 and 2017 and another study was identified through a conference abstract. Overall,
5 studies (representing 6 cohorts) from two countries, Nigeria and Congo (Republic of) were fully
eligible. The characteristics of those studies are given in Table 2.7.
SSA West was the best represented sub-region, with 4 studies, while there was still no evidence
regarding dementia incidence for the SSA Eastern and Southern regions.
After data extraction, only four of those studies24, 25, 27 could be included in the meta-analysis, case
(numerator) and person-years (denominator) could not be extracted for the remaining two4, 26.
Collectively, the meta-analysed studies included 7,035 older people ‘at risk’ representing over
20,214 person-years of follow-up. Cohort size and length of follow-up were both higher in Nigerian

27DEMENTIA IN SUB-SAHARAN AFRICA
Table 2.7
Characteristics of the incidence studies in sub-Saharan Africa
Study, reference Country,
region/
city
Rural /
urban
Dates of
fieldwork
Lower
age
limit
Sampling
strategy
Case
ascertainment
Diagnostic
Criterion
Clinical
diagnosis
Diagnostic instruments Mid-year
of baseline
survey
Follow-
up time
(mean or
*median)
Cohort
at risk
Person-
years
Indianapolis-Ibadan
Dementia Project 4
Nigeria,
Ibadan
Urban 1992
onwards
65 Door-to-door Communit y
survey
DSM-III/
ICD10
Yes CSI-D, CERAD, CAMDEX,
clinical exam
199 4 5.1 2459 Not
stated
Indianapolis-Ibadan
Dementia Project 24
Nigeria,
Ibadan
Urban 1992
onwards
70 Door-to-door Communit y
survey
DSM-III/
ICD10
Yes CSI-D, CERAD, CAMDEX,
clinical exam
199 2 NS 1174 6158
2001
onwards
2001 NS 1895 8 712
Ibadan Study of
Aging25
Nigeria,
Ibadan
Urban 11/ 0 3-
08/04
65 Cluster
random
sample
Community
survey
§Yes 10-WDRT, CHIF § 2004 3.3 122 5 3888
Ibadan Study of
Aging26
Nigeria,
Ibadan
Urban 11/ 0 3-
08/04
65 Cluster
random
sample
Community
survey
§Yes 10-WDRT, CHIF § 2003 NS 1894 6502
EPIDEMCA-FU**27 Congo,
Gamboma
and
Brazzaville
Rural &
urban
07/12 –
09 /15
65 Catchment
area /
random
sampling
Community
survey
DSM-IV Yes Free and Cued Reminding
Test, Zazzo’s cancellation
Task, Isaac’s Set Test of
verbal fluency, Geriatric
Depression Scale
2012 2847 14 56
* NS = not stated. ** EPIDEMCA-FU = Epidemiology of Dementia In Central Africa – Follow-up
§ Diagnostic relied on logistic regression models (drawing on data derived from the Ibadan Indianapolis Dementia Project, using the DSM-III / ICD10 criteria)
to classify subjects with dementia from those without based on the performance on the 10-WDRT af ter adjusting for sex and age.

ALZHEIMER’S DISEASE INTERNATIONAL28
studies than in Congo. For one study, the research diagnostic criteria were not clearly specified; all
the others had applied the DSM- III or IV criteria (combined with ICD-10 in one study).
Estimation of the incidence of dementia
The incidence of dementia increases exponentially with increasing age. For all studies combined,
the incidence of dementia doubles with every 7.7 year increase in age, from 4.1/ 1000 person years
(pyr) at age 60-64 to 59.8/1000 pyr at age 90+ (see Table 2.8).
Based on the current evidence available, the incidence of dementia in SSA appears to be relatively
similar to the incidence estimated for low and middle income countries in the 2015 World Alzheimer
Report (according to DSM criteria). Those figures are lower than the dementia incidence estimated
in countries with high incomes7.
There was significant heterogeneity in the incidence estimates of those four studies combined
(alpha = 0.16, 95%CI: 0.00-0.19).
Estimation of the number of incident cases of dementia per year
The number of annual incident cases of dementia that occurred in SSA was generated by
estimating first the numbers of older people at risk (total population in each age group, minus
numbers with prevalent dementia), and then applying the incidence rates estimated earlier.
The numbers of new cases increases and then declines with increasing age in each region. In SSA
peak incidence is among those aged 70-74 years, while it is for 75-79 in Southern SSA and among
those aged 65-69 in West SSA (Table 2.9). We estimated about 367,698 new cases of dementia
each year in SSA. These new estimates are slightly lower than annual numbers of new cases
estimated in the 2015 World Alzheimer Report. However, the previous estimates were generated by
using the age-specific incidence of dementia estimated in all LMIC (including one study in Nigeria)
whilst we have now been able to use the age-specific incidence of dementia in SSA (including only
studies in SSA).
2.5 Mortality associated with dementia
People developing dementia see their lives shortened by the condition. Although median survival
times have been estimated at 7.1 years for people with Alzheimer’s disease and 3.9 years for those
with vascular dementia28, survival times with dementia can vary significantly across individuals.
The independent contribution of dementia to mortality is difficult to assess. Dementia is rarely
considered as a direct or underlying cause of death, making death certificates unreliable. People
with dementia are also likely to have comorbid health conditions that may or not be related to the
Table 2.8
Meta-analysed estimates of dementia incidence, generated from Poisson random effects models
Region Number
of studies
included in
meta-analysis
Age group Age- and sex-
standardised
incidence, for those
aged 60+ (using
Western Europe
as the standard
population)
60-64 65-69 70-74 75 -79 80-84 85-89 90+
Sub-Saharan
Africa
44.1 6.1 9.2 14. 4 22.6 34.4 59.8 13.26
Table 2.9
Estimated annual numbers of incident cases of dementia, by age group and region
Region 60-64 65-69 70 -74 75-7 9 80-84 85-89 90+ To tal
SSA Central 9,907 10, 860 11, 2 7 2 10,33 9 7, 5 4 6 3,381 1,510 54, 815
SSA East 26,238 29, 5 74 30,486 28,669 23,336 11, 24 3 5,406 154,95 5
SSA Southern 6,567 6,693 6,824 7, 516 6,524 5,002 1,6 81 40,807
SSA West 24,523 26,561 2 6 ,18 0 21,350 12,928 4,394 1,18 5 117,12 1
Sub-Saharan Africa 6 7, 2 3 4 73,688 74,76 2 6 7, 8 7 5 50,334 24,020 9,78 5 3 6 7, 6 9 8

29DEMENTIA IN SUB-SAHARAN AFRICA
dementia process and may precipitate death. Therefore, death of people with dementia cannot be
attributed automatically to dementia.
2.5.1 Methods
We aimed at reviewing the evidence on mortality risk for people with dementia in sub-Saharan
African populations.
We identified studies estimating mortality risk among older people with dementia in longitudinal
cohort studies carried out in SSA, through the results of the systematic reviews conducted for
the purpose of this report (prevalence and incidence) and by conducting a scoping review of the
literature using the search terms ‘dementia AND mortality AND sub-Saharan Africa’.
We extracted estimates for each eligible study, provided as hazard ratios and their 95% confidence
interval. A random effect meta-analysis was used to pool estimates. Heterogeneity was examined
using Chi2 test and quantified using Higgins I2.
2.5.2 Results
We identified 4 studies estimating the risk of death among people with dementia, two from
Nigeria26, 29, one from Tanzania30, and one from the Republic of Congo31. All were based on cohort
studies previously described in the incidence section (see Table 2.7).
Among all of them, an increased mortality risk was reported for people with dementia (from HR=1.5
(95%CI: 1.2-1.8) in the Ibadan Study of Ageing26 to HR=6.3 (95%CI: 3.2-12.6) in Tanzania30).
Overall, a two-fold increased mortality risk for people with dementia (HR=2.3, 95%CI: 1.0-3.5) was
estimated in the meta-analysis (Figure 2.1). Relatively similar mortality hazards have been reported
in other low and middle countries, with a pooled HR at 2.8 (95%CI: 2.5-3.1) in the 10/66 Dementia
Research Group studies in Latin America and China32.
In two studies, the effect of dementia severity was investigated. For the Yorubas in Nigeria, the risk
for low severity (CDR ≤1) was lower (HR=2.1, 95%CI: 0.7-6.3) than for the high severity group
(CDR≥2) (HR=4.5, 95%CI: 0.8-25.1)29. In Congo, greater dementia severity (CDR ≤1 vs. CDR≥2)
was also strongly associated with increased mortality risk (HR=1.9, 95%CI: 1.2-3.0)31.
The role of the most common genetic
risk factor for dementia, the APOE
ε4 allele, was also examined in the
Indianapolis Ibadan Dementia Project in
Nigeria. Having APOE ε4 alleles was not
significantly associated with mortality in
Ibadan participants under 75 (HR=1.0,
95%CI: 0.8-1.4) or over 75 (HR=1.2,
95%CI: 0.8-1.7)33.
2.6 Risk factors
A large number of potential risk and
protective factors for dementia and
cognitive impairment have been
investigated in epidemiological studies
worldwide34. Research findings suggest
that dementia has a lifelong trajectory,
with risk clustering around specific
periods and often accumulating, sometimes even decades before the onset of the first clinical
symptoms35. Some risk factors may have an effect in early-life, other may have an influence later in
life and interact with the environment over the life course.
In the following section, we will review the evidence on the association of several potential risk
factors with the onset of dementia among sub-Saharan populations. Studies were identified
through both systematic reviews for prevalence and incidence presented earlier, as well as a
scoping review using the search terms ‘risk factor AND dementia AND sub-Saharan Africa’. We
principally focused this review on population-based studies, where the participants were living in
the community.
NOTE: Weights are from random effects analysis
Overall (I-squared = 52.4%, p = 0.098)
EPIDEMCA-FU (Congo)
Study
ID
Ibadan Study of Ageing (Nigeria)
Indianapolis Ibadan Dementia Project (Nigeria)
Hai (Tanzania)
2.28 (1.02, 3.54)
2.53 (1.42, 4.49)
HR (95% CI)
2.83 (1.10, 7.27)
6.33 (3.19, 12.58)
100.00
29.61
%
Weight
51.41
12.65
6.32
-12.6
0
12.6
1.50 (1.20, 1.80)
Figure 2.1
Forest plot for studies estimating the mortality risk for people
with dementia in SSA

ALZHEIMER’S DISEASE INTERNATIONAL30
2.6.1 Genetic factors
The Apolipoprotein E (APOE) gene appears to have been studied more than other dementia–
related genes/loci in Africa36. Existing in 3 allelic forms (APOE ε2, APOE ε3, APOE ε4), APOE ε2 is
protective whereas APOE ε4 is associated with age–at–onset and increases the risk of Alzheimer’s
Disease in a dose–dependent manner among Caucasians. Frequencies of APOE ε4 vary widely
in sub-Saharan Africa: 19.8% (Yoruba Nigerians), 32% (Kikuyu Kenyans) and 40% (in the
Pygmies of Central African Republic)37. Nevertheless, APOE ε4 showed a lack of association with
Alzheimer’s Disease (AD) among Yoruba Nigerians38 and Kikuyus of East Africa39 or with cognitive
impairment in Benin12. More recent longitudinal data from an enriched cohort of the Yoruba
subjects demonstrated a significant association between APOE ε4 homozygosity and AD and a
weak association between one allele ε4 and incident AD and cognitive decline40. The possibility
of gene - gene and/or gene-environment interactions which have varied over time offers plausible
explanations for this observed trend.
While other mutations have been reported in the Amyloid Precursor Protein (APP), Presenilin 1
(PSEN1) and Presenilin 2 (PSEN2) gene among North African populations41, a study of a large
South African Xhosa family with early onset AD (affecting 12 individuals across 4 generations)
yielded a novel mutation in the PSEN1 gene and novel CHMP2B polymorphisms have been
described in another South African family with Frontotemporal dementia (FTD)42.
2.6.2 Age and sex
Increasing age is the most consistent risk factor for dementia worldwide. It was also the most
consistent across studies undertaken in SSA where an increased risk of dementia with advancing
age was also described2, 12, 14, 16, 19, 43-47. While prenatal stress, early life exposure to infectious
diseases of poverty, malnutrition and violence might cause early death or influence the ageing
process and reduce longevity for people in developing countries, life expectancy has been rising in
SSA in addition to exponential population growth, impacting the future burden of dementia in the
region.
Dementia and AD were also significantly associated with the female sex in several SSA studies.
The female sex is associated with a 2- to 8-fold increased risk14, 25, 43, 46, largely due to their longer
lifespan. It conforms to nearly all studies in LMIC confirming that women are marginally more likely
to develop dementia and AD, particularly in very old age, on the basis of the greater longevity in
women48.
2.6.3 Vascular disease risk factors
Eighty-five percent of the global burden of non-communicable diseases (NCDs) occurs in LMICs
including those of SSA49 and cardiovascular diseases (CVDs) account for most of this burden. The
current epidemiologic transition from infectious diseases to non-communicable but modifiable
chronic disorders has resulted from gradual adoption of a Western lifestyle that includes excessive
caloric intake, unhealthy diet, and decreased physical activity. Hypertension, one of the major
NCDs, has the highest global burden in Africa with a prevalence of 46% according to the WHO50.
The burdens of type 2 diabetes, obesity, hyperinsulinemia and dyslipidaemia are also rising in SSA
with attendant increasing rates of stroke, a major outcome of vascular brain injury.
Accumulating evidence suggests that stroke injury and vascular factors increase risk for AD and
other dementias51. Hypertension, hypercholesterolemia and peripheral arterial disease in particular
have demonstrated significant associations with dementia and AD in SSA studies43, 52-54, with
increased odds between 1.5 and 2.5 to develop dementia (Table 2.10).
2.6.4 Literacy and education
Robust evidence has accrued from prospective cohort studies, systematic reviews and meta-
analyses demonstrating illiteracy or low educational achievement to be a robust risk factor
for dementia worldwide34. Although more education does not necessarily protect individuals
from developing neurodegenerative and vascular neuropathology, it does mitigate the impact
of pathology on the clinical expression of cognitive decline and dementia55. This explains the
‘Cognitive Reserve’ hypothesis as a metric of educational attainment.
The findings from studies on the association between education and dementia are mixed in SSA.
While several studies demonstrate an inverse association56-58, some show a lack of

31DEMENTIA IN SUB-SAHARAN AFRICA
Box 1
HIV-associated neurocognitive impairment in
sub-Saharan Africa
Worldwide, over 35 million individuals are infected with HIV, of whom over 24 million live in
sub-Saharan Africa71. In high-income countries (HICs), combination antiretroviral therapy
(cART) has greatly reduced opportunistic infections and near-normalised life expectancy,
but long term complications including cognitive impairment and dementia are becoming
increasingly prevalent72.
HIV-associated neurocognitive disorders (HAND) are defined by American Academy of
Neurology (AAN) criteria reflecting the range of cognitive impairments currently seen in people
with HIV73. These are:
• HIV-associated dementia (HAD),
• HIV-associated minor neurocognitive disorder (MND),
• Asymptomatic Neurocognitive Impairment (ANI).
Mechanisms are complex and include opportunistic infections of the central nervous system
(CNS), direct neurotoxic effect of the HIV virus, long term inflammatory processes in the CNS
and neurotoxic effect of cART74, 75. Access to cART reduces the prevalence and severity of
HAND, but does not prevent it76-78.
In HIC, with widespread access to cART incidence of HAND is decreasing but prevalence
increasing due to near-normalised life expectancy. HAD is becoming uncommon, and less
severe HAND (ANI and MND) predominate72. This ageing of populations living with HIV has
resulted in new challenges. Both HIV infection and long term cART treatment are associated
with premature ageing, including accelerated Alzheimer’s disease and vascular disease79, 80.
Older individuals with HIV are therefore likely to be at increased risk of these dementias as well
as HAND.
As the population of SSA continues to age, it is likely that greater numbers of older adults will
be affected by HIV. The scale-up of cART in SSA has led to an increase in HIV prevalence due
to increasing life expectancy in treated individuals81. The proportion of HIV positive individuals
aged 50 and over in SSA is predicted to triple by 2040, and to increase from one in seven, to
one in four by 204082, 83.
Epidemiological studies of neurodegenerative dementia amongst the growing population of
older adults in SSA, have not generally screened for HIV or HAND84. All existing published data
on HAND relate to younger adults (mean age ranging from 29.5 to 42.7)85-87 and many studies
specifically exclude older adults.
Few studies of HAND in Africa report prevalence using AAN criteria. Early studies of HAD
(previously known as the AIDS dementia complex) describe a subcortical dementia occurring
in the context of advanced HIV infection, opportunistic infections and low CD4 count prior
to the availability of cART and associated with poor prognosis. These hospital-based clinical
prevalence studies, conducted during the original AIDS epidemic, report prevalence of
between 11.4% and 41% in young adult inpatients, reflecting differing clinical diagnostic
criteria88-90. The international WHO study reported prevalence of 19% in Zaire (now Democratic
Republic of Congo) during this epidemic91 using detailed neuropsychological assessment.
More recent clinic-based studies using AAN criteria report high, but varying HAND prevalence
across different SSA settings (see Table 2.A)92-96, 100. One retrospective hospital-based study
of 175 HIV+ individuals aged 55 and over in Congo (Republic of) reported a high prevalence of
neuropsychiatric disorder (49.7%) but did not assess for cognitive impairment or confusion99.

ALZHEIMER’S DISEASE INTERNATIONAL32
Most studies report HAND using the International HIV Dementia Scale (IHDS), a brief screening
tool originally developed and validated in Uganda100. Reported prevalence of HAND using
the IHDS is generally higher than that using standard diagnostic criteria but varies markedly
between studies, ranging from 14% to 85% in studies completed in Uganda101, 102, Central
African Republic (CAR)87, Ethiopia103, Malawi104, Botswana105 and Cameroon106. The IHDS was
validated in a relatively well-educated sample100 unlikely to be typical of other areas of SSA.
A recent meta-analysis reported the diagnostic accuracy of the IHDS for HAND to be limited
(sensitivity 64.3%, specificity 49.6%)107.
Existing SSA studies consistently report cognitive impairment in individuals with HIV compared
with HIV negative controls, especially in cognitive domains known to be most affected in HIV
(executive function, processing and motor speed and verbal learning)93, 98, 108-110. As in HICs,
there is evidence that cART improves but may not normalise cognitive function111.
Some initial evidence suggests that HAND prevalence may be affected by HIV clade112 and
that cardiovascular disease increases risk. Increasing age and lower educational level are
relatively consistently associated with HAND92, 105, 106. HIV-related factors such as disease
stage, CD4 count and viral load are associated with HAND in some studies92, 97, 113-115 but not
others92, 104, 114.
Generally, existing SSA studies (of prevalence or risk factors for HAND) are difficult to
compare. The cohorts studied differ markedly in the diagnostic criteria used, proportion of
individuals on cART, HIV disease stage and educational background85, 86. These existing, albeit
limited data relate almost exclusively to young adults. Our current knowledge of prevalence
and risk factors for HAND in older adults is limited to data from settings outside SSA. Studies
of HAND in older people in SSA have not yet taken place and this is a current and urgent
knowledge gap.
Table 2.A
Studies of HAND by AAN (2007) criteria in SSA
Author,
Year
Location Population Age HAND prevalence Other/risk factors
Kelly et al.,
2014
Blantyre,
Malawi
HIV+ adults on
cART
Med 39(18-71) 3% HAD
12% MND
55% ANI
Symptomatic HAND not
associated with symptom
score
Joska et
al ., 2 011
South Africa,
(primary care)
170 HIV +
adults not
on cART
(commencing
treatment)
Mn 29.5 (3.65)
Range 18-35
42.4% MND
25.4% H AD
Significant association of
older age and low education
with HIV dementia. Trend
towards lower CD4
Yusuf et
al., 2017
Nigeria 418 H IV+
adults on cART
for at least one
year
Mn age 37.2
SD 9.3
21.5%
9.6% MND/HAD
Sacktor et
al., 2 015
Uganda
(rural)
299 HIV+ not
on cART
Mn age 36
(SD 9)
27% HAD D and A clades. No
difference in dementia
prevalence between HIV +/-
groups, only MND/ANI
Oshinaike
et al., 2012
Nigeria,
(hospital)
208 HIV+ Mn age 36.8 42.3% HAND
Yakasai et
al., 2 015
Nigeria
(teaching
hospital)
80 HIV + (50%
on cART)
40 HIV -
HAD 23% off –cART
(5% on cART)
Symptomatic HAND
40% in cART – and
30% in cART +
Comprehensive neuropsych.
battery and functional
assessment (but possibly
no clinical interview) used
for AA N HAND diagnosis

33DEMENTIA IN SUB-SAHARAN AFRICA
association19, 43, 46, 59. Low literacy is often linked to poverty or lower socioeconomic status, which
is also associated with poorer health, lower access to health care, and increased risk of dementia.
However in African communities where formal education systems may be less developed, informal
and traditional systems of instruction ensure that individuals are well trained to understand and
cope with the environmental and socioculturally determined demands that life brings their way. This
may explain the lack of association between literacy / ‘formal’ educational attainment and dementia
in some studies. Formal education may not be a good indicator of cognitive ability in this context43.
2.6.5 Early-life factors
Early-life risk factors including factors related to perinatal circumstances (such as birth weight
and intrauterine environment) as well as factors affecting the growth and development of the
brain or body, such as nutrition, have been shown to influence the risk of dementia in later life.
Although well-researched risk factors in longitudinal studies in HIC, early-life risk factors have
been understudied for now in the SSA region. As early-life events can easily be subjects to recall
bias, the establishment of a causal link between those and the onset of dementia in late-life can be
challenging.
For now, only cross-sectional studies in Central Africa have investigated the association between
early-parental death and dementia. In the EDAC (Epidemiology of Dementia in Central Africa) study,
Guerchet and colleagues found a significant association between the death of one parent during
childhood and dementia in late life43 (Table 2.10). No stressful life event occurring during childhood
was associated with dementia in the same countries a few years later60.
2.6.6 Lifestyle factors
Lifestyles including dietary habits, physical activity, smoking and alcohol use are often interrelated
and are usual targets for optimal cardiovascular and brain health as well as dementia prevention.
The relationship of alcohol with dementia has been described as a ‘J-’ or ‘U-’ shaped relationship
with abstainers and heavy drinkers having higher risk61. The findings from African studies on the
relationship between alcohol and dementia are mixed with Pilleron et al. showing protective effect62
while Gureje et al. demonstrated a deleterious effect44 (Table 2.10).
Undernutrition is a major problem among older people in SSA with up to 96 % of participants
demonstrating nutritional vulnerability and 18% undernourished in a Nigerian study on nutritional
assessment in elderly Yorubas63. Nutritional factors associated with dementia in African populations
include reduced body mass index – as an index of undernutrition64, 65 and lower consumption
of oleaginous foods62 while pre-stroke daily fish intake was found to be protective against early
cognitive impairment among stroke survivors56. The low incidence of dementia in the Yoruba
Nigerians was consistent with their traditional low calorie and low fat diet consisting of grains, yam
tubers (Dioscorea rotundata), vegetables, and some fish48.
2.6.7 Psychosocial and psychological factors
There is a weight of evidence from previous studies linking depression and subsequent onset
of dementia, even though it remains to be disentangled whether the depression is a prodromal
syndrome which comes before the full blown disease (dementia) or whether depression is an
independent risk factor for dementia66.
The EDAC study in Central Africa found a significant cross-sectional association between
depressive symptoms at time of diagnosis and dementia (OR=3.50, 95%CI: 1.56-7.82)43.
There is also a body of evidence suggesting that personality types that demonstrate proneness to
psychological distress may be more at risk of dementia. It was found in the Indianapolis – Ibadan
Dementia Project that personality changes could predict future dementia irrespective of changes
in cognition or functional status67. An association between Dependent Personality Disorders was
described among older people in Congo but not in Central African Republic68 (Table 2.10).
Investigating social network in the older populations, living with others was found protective against
dementia69 in Nigeria while social isolation or a poor social network were risk factors of prevalent or
incident dementia in SSA25, 26, 70.

ALZHEIMER’S DISEASE INTERNATIONAL34
2.6.8 Summary
Overall, studies of risk factors of dementia have a similar coverage to that of prevalence and
incidence studies, with the bulk of the evidence coming from West SSA (especially Nigeria)
and Central SSA. Although age and sex seem to be consistently associated with dementia in
SSA studies, the evidence around other risk factors is more patchy, and this needs to be more
thoroughly investigated. Education and the APOE ε4 allele, strong risk factors of dementia in HIC,
showed very mixed effects in SSA populations, which might be explained by several hypotheses
(gene-gene and/or gene–environment interactions, lack of power, cultural factors). The strongest
evidence on modifiable risk factors is probably the association of dementia with vascular risk
factors, while the evidence on early-life, psychological, psychosocial and lifestyle factors remains
tentative. Most importantly, this evidence relies mainly on studies with a cross-sectional design,
which cannot establish any causal relationship between dementia and those potential risk factors.
2.7 Conclusion and recommendations
This chapter assembles, for the first time, evidence on the incidence as well as the prevalence of
dementia, the societal economic cost of dementia, mortality risk, and the potential for prevention
in sub-Saharan Africa. This work has benefited from the recent expansion of the evidence base on
the epidemiology of dementia in the region. Considering the direct evidence of the accumulated
data, the earlier impression that the prevalence of dementia is much lower in sub-Saharan Africa
compared to high income countries is now fading. The current evidence base, and the detailed
estimates provided, while still provisional, constitute the best currently available basis for policy-
making, planning and allocation of health, welfare and population prevention resources.
Thoroughly reviewing the research carried out up to now on dementia epidemiology in population
surveys, we have identified a new set of studies for the sub-Saharan region. While the evidence
on dementia prevalence has considerably expanded since the 2009 World Alzheimer Report, the
evidence on dementia incidence and mortality have also improved. We are therefore slowly moving
away from single estimates for the whole sub-Saharan region for most of the epidemiological
indicators. However, the evidence on risk factors of dementia in the region is not very strong
for now, remaining tentative for most of the modifiable risk factors and mainly relying on cross-
sectional studies rather than longitudinal ones.
These updated reviews suggest that our previous estimates certainly underestimated the scale
of the dementia epidemic in sub-Saharan countries. The upward revision of both prevalence of
dementia and numbers of people living with dementia does not imply that prevalence has changed
recently. Rather, improvements in the extent and quality of the available evidence have improved
the precision of our estimates.
Overall, our reviews suggest:
• A modest increase in dementia prevalence, from 5.5% to 7.2% for the age- and sex-
standardised dementia prevalence for all four sub-Saharan regions, coming closer to the
prevalence reported for high income countries,
• A consequently modest increase in costs of dementia in SSA, with a great contribution to
informal care costs for all regions,
• An incidence of dementia for all 4 SSA regions slightly lower that was estimated for all LMIC
in 2015,
• A 2-fold mortality risk for people with dementia in SSA, relatively similar to other LMIC.
However, our reviews highlight continued deficiencies in research evidence. It is therefore
important to underline the main limitations.
Coverage
Despite a modest improvement in coverage for the West and Central SSA regions over the last
decade, Southern and East SSA are still represented by single studies which cannot provide
an accurate description of the epidemic compared to their population size. SSA is expected to
experience rapid population ageing. Prevalence, incidence and mortality risk are almost, if not
entirely, unknown for some regions and they may change over time considering the simultaneous
rise in non-communicable diseases in the region. Studies of risk factors of dementia have a similar
coverage to that of prevalence and incidence studies; consequently almost no evidence is available
for the East and Southern SSA.

35DEMENTIA IN SUB-SAHARAN AFRICA
The need for descriptive studies is still huge. It may seem elementary but such studies are
necessary to increase awareness among policy makers and stakeholders, and to improve the
accuracy and recency of figures for planning and allocation of resources.
Quality
Although the overall quality is high in dementia studies in SSA, the misapplication of the
multiphase study design was seen in 50% of the studies, with no screen negatives included at the
second stage and/or no weighting back to estimate the correct prevalence. Therefore, previous
recommendations can be reiterated. Research funders, ethics committees and journals should
encourage the application of the correct procedure by not funding or approving faulty designs
or not publishing findings if the analyses are not correctly performed or limitations not clearly
acknowledged. The effect of incorrect application of two-stage designs on dementia prevalence is
difficult to quantify here although it should lead to an under-estimation of true dementia prevalence.
When multistage designs are implemented, measures should also be taken to prevent the loss
of follow-up between screening and diagnostic stages, the main one being to minimise the delay
between the two stages.
The proportion of studies that had to be excluded from the meta-analyses for dementia prevalence
or incidence because numerators (cases) and denominators (total population or person-years) were
not reported was lower than in other regions. However, characteristics of the studies (sampling,
diagnostic criterion, follow-up time) were also not always clearly reported. Despite the existence of
reporting guidelines (like the STROBE116) to help increase the quality of published health research, it
appears that authors, peer-reviewers and journal editors are not yet all following those.
Some recent incidence studies have included sensitivity analyses to examine the robustness of
their estimates taking into account the changes in cohort composition, length of follow-up or loss
of follow-up24, 27. Conducting verbal autopsies during follow-up to assess whether the older people
have developed memory problems / dementia-like symptoms before death is useful for estimating
the impact of attrition and the number of possible incident cases of dementia who died before
diagnosis can therefore be estimated, and can be modelled to provide an alternative incidence
estimate. Considering the high mortality rates observed among older populations in low and middle
income countries, we would recommend the use of sensitivity analyses when possible in future
studies.
Altogether, study quality was strengthened by the systematic use of multi-domain cognitive
batteries and informant interviews during dementia assessment but there is still room for
improvement on the correct implementation of multistage designs and their analysis, and adequate
description of the studies when the time to disseminate the results comes.
Heterogeneity
Considering the limited evidence available, modelling prevalence and incidence in this review
assumed that both indicators were uniform within the sub-Saharan region. However, significant
heterogeneity was found in both prevalence and incidence estimates across sub-Saharan Africa.
Given the diversity of cultures, languages, levels of development, health and risk profiles, lifestyles,
as well as genetic background existing in the SSA region, this is not surprising. Beyond the
statistical heterogeneity, the level of similarity of methods and instruments to assess dementia
used in these studies is certainly high compared to other GBD regions. The Community Screening
Interview for Dementia (CSI-D) was a consistent screening instrument in prevalence studies
while the majority of prevalence and incidence studies followed the DSM-IV criteria for dementia
diagnosis. Standardisation of study procedures, which would reduce methodological variability,
seems like an achievable goal in this region. The new DSM-V criteria and the use of one-phase
designs might have to be considered but cross-cultural adaptations might still be necessary and
validity of both in this context has to be established. The study of temporal trends in the coming
decades will require maintaining a constant methodology over time. As research on dementia in
SSA is actively developing, it might be timely to reach a consensus regarding standardisation of
study methods.
Risk factors and potential for prevention
Despite growing evidence on HIV-associated dementia (HAND) in Southern SSA, only younger
populations have been included and a current and urgent knowledge gap has been identified
for older populations in SSA. When feasible, including HIV screening in the future generations of
population-based studies of dementia would help in filling this gap.

ALZHEIMER’S DISEASE INTERNATIONAL36 37DEMENTIA IN SUB-SAHARAN AFRICA
Psychosocial factors Early - life factors Education Vascular Factors
Table 2.10
Characteristics of the epidemiological studies on the modifiable risk factors of dementia in SSA
Reference Country, region Design Population, sample size Exposure Outcome Measurement Effect size (95% CI)
Hall et al., 200653 Nigeria, Ibadan Longitudinal ≥ 70 yrs, n=1075 Cholesterol levels and LDL Dementia: DSM-III-R/ICD-10; AD: NINCDS-ADRDA Increasing levels of cholesterol levels and LDL associated with increasing risk of
AD in individuals without the APOE-ε4 allele (respectively OR=1.015, p=0.025 and
OR=1.017, p=0.023) but not in those with APOE-ε4 in Yoruba subjects
Ogunniyi et al., 201154 Nigeria, Ibadan Longitudinal ≥ 65 yrs, n=2494 Hypertension Dementia: DSM-III-R/ICD-10; AD: NINCDS-ADRDA aOR=1.52 (95%CI: 1.01-2.30)
Guerchet et al., 201243 Central Africa (CAR / Congo) Cross-sectional ≥ 65 yrs, n=977 Hypertension Dementia: DSM-IV aOR=1.72 (95%CI: 1.03-2.85)
Guerchet et al., 201352 Central Africa (CAR / Congo) Cross-sectional ≥ 65 yrs, n=977 Peripheral artery disease
(measured by Ankle Brachial Index)
Dementia: DSM-IV aOR=2.37 (95%CI: 1.31-4.26)
Ochayi et al., 200659 Nigeria, Central Cross-sectional ≥ 65 yrs, n=280 Education (none vs primary) Dementia: Community Screening Interview for Dementia (cutpoint) Lack of association between level of education and dementia (OR not shown)
Ogunniyi et al., 200646 Nigeria, Ibadan Longitudinal ≥ 65 yrs, n=1255 Education (in years) Dementia: DSM-III-R/ICD-10; AD: NINCDS-ADRDA aOR=0.81 (95%CI: 0.53–1.07)
Chen et al., 201039 Kenya, Nyeri Case-control ≥ 70 yrs, n=184 (84 cases
and 100 controls)
Education (in years) Dementia : DSM IV and ICD-10 No association between years of education and dementia status.
Paraïso et al., 201114 Benin, Cotonou Cross-sectional ≥ 65 yrs, n=1137 Primary school vs. none Dementia Diagnosis: DSM-IV, NINCDS-ADRDA Differences in level of education did not affect dementia prevalence
Guerchet et al., 201243 Central Africa (CAR / Congo) Cross-sectional ≥ 65 yrs, n=977 No education (vs primary) Dementia: DSM-IV For men: aOR=0.52 (95%CI: 0.14-1.93); for women: aOR=2.41 (95%CI: 0.99-
5.85)
Toure et al., 201247 Senegal, Dakar Cross-sectional ≥ 65 yrs, n=507 Illiteracy Dementia: Aging in Senegal test (cutpoint) aOR=2.76 (95%CI: 1.23–6.20)
Longdon et al., 201319 Tanzania, Hai Cross-sectional ≥ 70 yrs, n=1198 Education (none vs primary) Dementia: DSM-IV aOR=0.69 (95%CI: 0.40-1.20)
Paddick et al., 201358 Tanzania, Hai Cross-sectional ≥ 70 yrs, n=1198 Education (none vs primary) Dementia: 10/66 criteria aOR=2.76 (95%CI: 1.23-6.10)
Akinyemi et al., 201456 Nigeria, Ibadan Cross-sectional Stroke survivors: ≥ 45 yrs,
n=143
Low education Post -stroke cognitive impairment: ASA/AHA and DSM IV criteria aOR=5.09 (95%CI: 2.17-11.95)
Guerchet et al., 201243 Central Africa (CAR / Congo) Cross-sectional ≥ 65 yrs, n=977 Death of one parent during childhood Dementia: DSM-IV aOR=1.72 (95%CI: 1.01-2.93)
Lifestyle factors
Ochayi et al., 200659 Nigeria, Central Cross-sectional ≥ 65 yrs, n=280 Undernutrition: BMI < 18.5 kg/m2Dementia: Community Screening Interview for Dementia (cutpoint) aOR=3.5 (95%CI: 1.2-9.9)
Gureje et al., 200644 Nigeria, Southwest Cross-sectional ≥ 65 yrs, n=2152 Lifetime history of alcohol Dementia: 10-word delayed recall test and CHIF aOR=1.92 (95%CI: 1.38-2.79)
Ogunniyi et al., 201164 Nigeria, Ibadan Longitudinal ≥ 65 yrs, n=2494 Body Mass Index Dementia: DSM-III-R/ICD-10; AD: NINCDS-ADRDA Subjects with incident dementia declined 0.19 kg/m2 per year more than the
normal
Guerchet et al., 201243 Central Africa (CAR / Congo) Cross-sectional ≥ 65 yrs, n=977 Body Mass Index <18.5 (undernutrition) Dementia: DSM-IV aOR=1.77 (95%CI: 1.02-3.06)
Pilleron et al., 201562 Central Africa (CAR / Congo) Cross-sectional ≥ 65 yrs, n=2002 Lower consumption of oleaginous food Dementia: DSM-IV; MCI: Petersen’s criteria MCI: aOR=3.7 (95%CI: 1.4-9.9) & dementia: aOR=2.8 (95%CI: 1.0-7.7)
Alcohol Consumption aOR=0.3 (95%CI: 0.1-0.8)
Akinyemi et al., 201456 Nigeria, Ibadan Cross-sectional Stroke survivors: ≥ 45 yrs,
n=143
Fish intake Post -stroke cognitive impairment: ASA/AHA and DSM IV criteria Pre-stroke daily fish intake was protective: aOR=0.39 (95%CI: 0.15-0.89)
Pilleron et al., 201565 CAR Cross-sectional ≥ 65 yrs, n=2002 Undernutrition: mid-upper arm circumference
(MUAC) <24 cm, BMI<18.5, or arm muscular
circumference (AMC) < 5th percentile
Dementia: DSM-IV; MCI: Petersen’s criteria CAR: MUAC - aOR=1.8 (95%CI: 1.4-2.4), BMI - aOR=2.3 (95%CI: 1.6-3.1), AMC -
aOR=2.3 (95%CI: 1.1-4.6)
Congo Congo: MUAC - aOR=1.8 (95%CI: 0.9-3.7), BMI - aOR=2.2 (95%CI: 0.9-5.8), AMC
- aOR=2.6 (95%CI: 0.8-9.0)
Ogunniyi et al., 200069 Nigeria, Ibadan Cross-sectional ≥ 65 yrs, n=2494 Living with others Dementia: DSM-III-R/ICD-10; AD: NINCDS-ADRDA aOR=0.06 (95%CI: 0.01-0.49)
Smith-Gamble et al.,
200267
Nigeria, Ibadan Longitudinal ≥ 65 yrs, n=2494 Personality change Dementia: DSM-III-R/ICD-10; AD: NINCDS-ADRDA OR=2.0
Gureje et al., 201125 Nigeria, Southwest Longitudinal ≥ 65 yrs, n=2149 Poor social engagement (no participation
in family or communities activities)
Dementia: 10-word delayed recall test and CHIF aOR=2.0 (95%CI: 0.8-5.1)
Guerchet et al., 201243 Central Africa (CAR /
Congo)
Cross-sectional ≥ 65 yrs, n=977 Depressive symptoms (Goldberg’s scale) Dementia: DSM-IV aOR=3.50 (95%CI: 1.56-7.82)
Toure et al., 201247 Senegal, Dakar Cross-sectional ≥ 65 yrs, n=507 Low social network (3 or less contacts
with family per week)
Dementia: Aging in Senegal test (cutpoint) aOR=4.37 (95%CI: 1.80-10.61)
Pilleron et al., 201468 Congo Cross-sectional ≥ 65 yrs, n=2002 Dependent personnality disorder (PDQ-4+) Dementia: DSM-IV; MCI: Petersen’s criteria Congo: aOR=4.8 (95%CI: 2.0-11.7)
CAR No association in CAR
Pilleron et al., 201560 Central Africa (CAR /
Congo)
Cross-sectional ≥ 65 yrs, n=2002 Stressful life events Dementia: DSM-IV; MCI: Petersen’s criteria Associations with MCI: total number of SLEs (OR = 1.1, 95%CI: 1.0–1.2),
number of SLEs from the age of 65 (OR = 1.2, 95%CI: 1.0–1.3), and serious
illness in a child when the participant was aged 65 or more (OR = 2.8, 95%CI:
1.6–4.6). No association with dementia
Ojagbemi et al., 201626 Nigeria, Ibadan Longitudinal ≥ 65 yrs, n=2149 Low occupational status Dementia: 10-word delayed recall test and CHIF HR=3.2 (95%CI: 1.3-8.0)

ALZHEIMER’S DISEASE INTERNATIONAL38
Longitudinal studies of dementia in the United States, Rotterdam (Netherlands), Stockholm
(Sweden) and Britain (CFAS) have reported recent decreases in prevalence and incidence rates117.
The reasons adduced for the observations include better prevention/control of vascular diseases
and decline in stroke incidence, higher educational attainment and healthier lifestyles. Given the
fact that the profile of vascular risk factors for dementia in Africa might largely be similar to those
reported from these high income countries, and in absence of stronger evidence in SSA, African
countries as other LMIC regions might benefit from adopting measures aiming at controlling
vascular factors as well as improving education and nutrition in order to stem the looming dementia
epidemic predicted for these regions.
Absence of evidence on costs
The main limitation regarding the estimation of costs of dementia in sub-Saharan Africa is that
those are mainly based on imputations (see methods), as no original data has been published yet
on resource use and costs of dementia in sub-Saharan Africa. Descriptions of care systems and
amounts of informal care provided for at least one country in each of the regions would greatly
contribute to having more reliable estimates, on which stakeholders and policymakers could
confidently rely.
References
1. Ferri, C.P., et al., Global prevalence of deme ntia: a
Delphi consensus study. Lancet, 2005. 366(9503): p.
211 2-7.
2. Hendrie, H.C., et al., Prevalence of Alzheimer’s disease
and dementia in two communities: Nigerian Africans
and African Americans. Am J Psychiatry, 1995. 152(10):
p. 14 8 5- 92.
3. Alzheimer’s Disea se International, World Alzheimer
Repor t. 2009.
4. Hendrie, H.C., et al., Incidence of dementia and
Alzheimer disease in 2 communities: Yoruba residing
in Ibadan, Nigeria, and African Americans residing in
Indianapolis, Indiana. Jama, 2001. 285(6): p. 739-47.
5. World Health Organization, Dementia: a public health
priority. 2012, World Health Organization: Geneva. p.
112.
6. Prince, M., et al., Policy Brief for Heads of Government:
The global impact of Dementia 2013-2050. 2013,
Alzheimer’s Disease International: London. p. 7.
7. Prince, M., et al., World Alzheimer Repor t 2015 - The
Global Impact of dementia: an analysis of prevalence,
incidence, costs and trends. 2015: London.
8. George -Carey, R., et al., An estimate of the prevalenc e
of dementia in Africa: A systematic analysis.[Erratum
appear s in J Glob Health. 2012;2(2):020401corr].
Journal of Global Health, 2012. 2(2): p. 020401.
9. Lekoubou, A., J.B. Echouffo-Tcheugui, and A.P.
Kengne, Epidemiology of neurodegenerative diseases
in sub-Saharan Africa: a systematic review. BMC
public health, 2014. 14(pp 653).
10. Mavrodaris, A., J. Powell, and M. Thorogood,
Prevalences of dementia and cognitive impairment
among older people in sub-Saharan Africa: A
systematic review. Bulletin of the World Health
Organiz ation, 2013. 91(10): p. 773-783.
11. Olayinka, O.O. and N.N. Mbuyi, Epidemiology
of Dementia among the Elderly in Sub-Saharan
Africa. International Journal of Alzheimer’s Disease,
2014(pagination).
12. Guerchet, M., et al., Cognitive impairment and
dementia in elderly people living in rural Benin, West
Africa. Dementia and Geriatric Cognitive Disorders,
2009. 27(1): p. 34-41.
13. Guerchet, M., et al., Prevalence of de mentia in elderly
living in two cities of Central Africa: The EDAC sur vey.
Dementia and Ger iatric Cognitive Disorders, 2010.
30(3): p. 261-268.
14. Par aiso, M.N., et al., Prevale nce of dementia among
elderly people living in cotonou, an urban area of benin
(West Africa). Neuroepidemiology, 2011. 36(4): p. 245-
251.
15. UN Department of Economic and Social Affairs
Population Division, World Population Prospects: The
2015 Revision, Key Findings and Advance Tables.
2015.
16. Yusuf, A .J., et al., Prevalence of dementia and
dementia subtypes among community-dwelling
elderly people in northern Nigeria. International
Psychogeriatrics, 2011. 23(3): p. 379-386.
17. Ogunniyi, A., et al., Prevalence estimates of
major neurocognitive disorders in a rural Nigerian
community. Brain and Behavior, 2016. 6(7).
18. Guerchet, M., et al., Comparison of rural and urban
dementia prevalences in two countries of central
africa: T he epidemca study. Alzheimer’s and Dementia,
2013. Conference: p. Alzheimer’s Association
International Conference 2013. Boston, MA United
States. Conference Publication: (var.pagings). 9 (4
SUPPL. 1) (pp P688).
19. Longdon, A .R., et al., The prevale nce of dementia in
rural Tanzania: A cross-sectional community-based
study. International Journal of Geriatric Psychiatry,
2013. 28(7): p. 728-737.
20. Vanderpoel R., Heyns M., and and the 10/66 Dementia
Research Group. A lgorithmic case prediction of
dementia in relation to local clinicain diagnosis in
an indegenous South Africa population.;. in 27th
International Conference of Alzheimer’s Disease
International. 2012. London.
21. Wang, H., et al., Care giver Time and Cost of Home
Care for Alzheimer’s Disease: A Clinic-based
Obser vational Study in Beijing, China. Age ing
International, 2010. 35(2): p. 153-165.
22. Zhaorui, L., Economic Costs of Dementia in Low and
Middle Income Countries, in Centre for the Economics
of Mental a nd Physical Health (CEMPH), Institute of
Psychiatr y at King’s College London. 2012, King´s
college: London.
23. Gureje, O., et al., Incidence of and risk factors for
dementia in the Ibadan study of aging. J Am G eriatr
Soc, 2011. 59(5): p. 869-74.
24. Gao, S., et al., Dementia incidence declined in
African-Americans but not in Yoruba. Alzheimer’s and
Dementia, 2016. 12(3): p. 244-251.
25. Gureje, O., et al., Incidence of and risk factors for
dementia in the Ibadan study of aging. Journal of the
American Geriatrics Societ y, 2011. 59(5): p. 869-874.
26. Ojagbemi, A ., T. Bello, and O. Gureje, Cognitive
Reserve, Incident Dementia, and Associated Mortality
in the Ibada n Study of Ageing. Journal of the American
Geriatrics Society, 2016. 64(3): p. 590 -595.
27. Samba, H., et al., Incidence of dementia among older
adults in central africa: First results from the republic
of congo in the EPIDEMCA-FU study. Alzheimer’s and
Dementia, 2015. 1): p. P221-P222.
28. Dewey, M.E. and P. Saz, Dementia, cognitive
impairment and mortality in pe rsons aged 65 and over
living in the community: a systematic review of the
literature. Int J Geriatr Psychiatry, 2001. 16(8): p. 751-
61.

39DEMENTIA IN SUB-SAHARAN AFRICA
29. Per kins, A .J., et al., Risk of mor tality for dementia in a
developing country: the Yoruba in Nigeria. Int J Geriatr
Psychiatr y, 2002. 17(6): p. 566-73.
30. Paddick, S.M., et al., Mortality rates in communit y-
dwelling Tanzanians with dementia and mild cognitive
impairment: a 4-yea r follow-up study. Age Age ing,
2015. 44(4): p. 636-41.
31. Samba, H., et al., Dementia-associated mortality and
its predictors among older adults in sub-Saharan
Africa: results from a 2-year follow-up in Congo (the
EPIDEMCA-FU study). Age Ageing, 2016. 45(5): p. 681-
7.
32. Prince, M., et al., Dementia incide nce and mortality
in middle-income countries, and associations with
indicators of cognitive reserve: a 10/66 Dementia
Research Group population-based cohort study.
Lancet, 2012. 380(9836): p. 50-8.
33. Lane, K.A., et al., Apolipop rotein E and mortality in
African-Americans a nd Yoruba. J Alzheimers Dis,
2003. 5(5): p. 383-90.
34. Prince, M., et al., World Alzheimer Report 2014 -
Dementia Risk Reduction: An analysis of prote ctive
and modifiable risk factors. 2014: London.
35. Whalley, L.J., F.D. Dick, and G. McNeill, A life-course
approach to the aetiology of late-onset dementias.
Lancet Neurol, 20 06. 5(1): p. 87-96.
36. Ak inyemi, R.O., et al., Neurogenomics in Africa:
Perspectives, progress, possibilities and priorities. J
Neurol Sci, 2016. 366: p. 213-223.
37. Ogunniyi, A .A., R.O.; , Epidemiology and genetics of
Alzheimer’s disease with special reference to Africa ns.
Niger. J. Health Sci., 2003. 3: p. 1-6.
38. Gureje, O., et al., APOE epsilon4 is not associated with
Alzheimer’s disease in elderly Nige rians. Ann Neurol,
2006. 59(1): p. 182-5.
39. Chen, C.H., et al., A comparative study to screen
dementia and APOE genotypes in an ageing East
African population. Neurobiol Aging, 2010. 31(5): p.
732-40.
40. Hendrie, H.C., et al., APOE epsilon4 and the risk for
Alzheimer disease and co gnitive decline in African
Americans and Yoruba. Int Psychogeriatr, 2014. 26(6):
p. 977-8 5.
41. Rassas, A .A., et al., High APOE epsilon 4 allele
frequencies associated with Alzheimer diseas e in a
Tunisian population. Neurol Sci, 2012. 33(1): p. 33-7.
42. Heckmann, J.M., et al., Novel presenilin 1 mutation
with profound neurofibrillary pathology in an
indigenous Southern African family with early-onset
Alzheimer’s disease. Brain, 2004. 127(Pt 1): p. 133-42.
43. Guerchet, M., et al., Factors associated with dementia
among elderly people living in two cities in Central
Africa: the EDAC multicenter study. J Alzheimers Dis,
2012. 29(1): p. 15-24.
44. Gureje, O., A. Ogunniyi, and L. Kola , The profile and
impact of probable dementia in a sub-Saharan African
community: Results from the Ibadan Study of Aging. J
Psychosom Res, 200 6. 61(3): p. 327-33.
45. Hall, K., et al., Risk factors and Alzheimer’s disease: A
comparative study of two communities. Australian and
New Zealand Journal of Psychiatry, 1998. 32(5): p. 698-
706.
46. Ogunniyi, A ., et al., Risk factors for incident
Alzheimer’s disease in African Americans and Yoruba.
Metabolic brain disease, 2006. 21(2-3): p. 235-240.
47. Toure, et al., Risk factors for dementia in a senegalese
elderly population aged 65 years and over. Dement
Geriatr Cogn Dis E xtra, 2012. 2: p. 160-8.
48. Kalaria, R.N., et al., Alzheimer’s disease and vascular
dementia in developing countries: prevalence,
management, and risk factors. Lancet Neurol, 200 8.
7(9): p. 812-26.
49. World Health Organization. Available from: http://www.
who.int /ncds/en/.
50. World Health Organization. Global Health Observatory
(GHO) Data. Raised Blood Pressure-situation and
trends. . Available from: http://www.who.int/gho/ncd/
risk_factors/blood_pressure_prevalence_text/en/. .
51. Ak inyemi, R.O., et al., Vascular risk factors and
neurodegeneration in ageing related dementias:
Alzheimer’s disease and vascular dementia. Curr
Alzheimer Res, 2013. 10(6): p. 642-53.
52. Guerchet, M., et al., Association between a low
ankle-brachial index and dementia in a general elderly
population in Central Africa (Epidemiology of Dementia
in Centra l Africa Study). J Am Geriatr Soc, 2013. 61(7):
p. 113 5-40 .
53. Hall, K., et al., Cholesterol, APOE genotype, and
Alzheimer disease: An epidemiologic study of Nigerian
Yoruba. Neurology, 2006. 66(2): p. 223-227.
54. Ogunniyi, A., et al., Hypertension and incident
dementia in community-dwelling elderly Yoruba
Nigerians. Acta Neurologica Scandinavica, 2011.
124(6): p. 396-402.
55. Members, E.C.C., et al., Education, the brain and
dementia: neuroprotection or compensation? Brain,
2010. 133(Pt 8): p. 2210-6.
56. Ak inyemi, R.O., et al., Profile and determinants of
vascular cognitive impair ment in African stroke
survivors: the CogFAST Nigeria Study. J Neurol Sci,
2014. 346(1-2): p. 241-9.
57. Paddick, S.-M., et al., The as sociation between
educational level and deme ntia in rural Tanzania .
Dementia & Neuropsychologia, 2014. 8(2): p. 117-125.
58. Paddick, S.M., et al., Dementia prevalence estimates
in sub-Saharan Africa: comparison of two diagnostic
criteria. Glob Health Action, 2013. 6(1): p. 19646.
59. Ochayi, B. and T.D. Thacher, Risk factors for de mentia
in central Nigeria. Aging and Mental Health, 2006.
10(6): p. 616-620.
60. Pilleron, S., et al., Association between Stressful
Life Events and Cognitive Disorders in Central
Africa: Results from the EPIDEMCA Program.
Neuroepidemiology, 2015. 44(2): p. 99-107.
61. Ronksley, P.E., et al., Association of alcohol
consumption with selected cardiovascular disease
outcomes: a systematic review and meta-analysis.
BMJ, 2011. 342: p. d671.
62. Pilleron, S., et al., Diet, Alcohol Consumption and
Cognitive Disorders in Ce ntral Africa: A Study from the
EPIDEMCA Program. J Nutr Health Aging, 2015. 19(6):
p. 657- 67.
63. Olayiwola, I.O. and A.O. Ketiku, Socio-demographic
and nutritional assessment of the elderly Yorubas in
Nigeria. Asia Pac J Clin Nutr, 2006. 15(1): p. 95-101.
64. Ogunniyi, A ., et al., Weight loss and incident dementia
in elderly Yoruba Nigerians: a 10-year follow-up study.
Int Psychogeriatr, 2011. 23(3): p. 387-94.
65. Pilleron, S., et al., Association between mild cognitive
impairment and dementia and undernutrition among
elderly people in Central Africa: Some results from
the EPIDEMCA (Epidemiology of Dementia in Central
Africa) programme. British Journal of Nutrition, 2015.
114(2): p. 306-315.
66. Diniz, B.S., et al., Late-life depre ssion and risk of
vascular dementia and Alzheimer’s disease: systematic
review and meta-analysis of community-based cohort
studies. Br J Psychiatry, 2013. 202(5): p. 329-35.
67. Smith-Gamble, V., et al., Informant reports of changes
in personality predict dementia in a population-based
study of elderly african americans and yoruba. Am J
Geriatr Psychiatry, 2002. 10(6): p. 724-32.
68. Pilleron, S., et al., Is dependent pe rsonality disorder
associated with mild cognitive impairment a nd
dementia in Central Afric a? A result from the
EPIDEMCA programme. Int Psychogeriatr, 2014: p.
1-10.
69. Ogunniyi, A ., et al., Epidemiology of dementia in
Nigeria: results from the Indianapolis-Ibadan study. Eur
J Neurol, 20 00. 7(5): p. 485-90.
70. Toure, K., et al., Risk factors for dementia in a
Senegalese elderly population. African Journal of
Neurological Sciences, 2009. 28(1).
71. ( WHO), W.H.O. Number of people (all ages) living
with HIV: Estimates by WHO region. Global He alth
Obser vatory data repositor y 2017 27/01/2017[cited
2017 28/07/2017].
72. De eks, S.G., S.R . Lewin, and D.V. Havlir, The end of
AIDS: HIV infection as a chronic disease. The Lancet,
2013. 382( 99 03): p. 15 25 -1533.
73. Antinori, A ., et al., Updated research nosology for HIV-
associated neurocognitive disorders. Neurology, 2007.
69 (18) : p . 17 89-179 9.
74. Clifford, D.B. and B.M. Ances, HIV-associated
neurocognitive disorder. The Lancet Infectious
Diseases, 2013. 13(11): p. 976-986.

ALZHEIMER’S DISEASE INTERNATIONAL40
75. Nightingale, S., et al., Controversies in HIV-associated
neurocognitive disorders. The Lancet Neurology, 2014.
13 (11 ): p . 11 3 9 -1151.
76. Sacktor, N., et al., HIV-associated cognitive impairment
before and after the advent of combination therapy.
Journal of Neurovirology, 2002. 8(2): p. 136-42.
77. Heaton, R.K., et al., HIV-associated neurocognitive
disorde rs before and during the era of combination
antiretroviral therapy: differences in rates, nature, and
predictors. Journal of Neurovirology, 2011. 17(1): p.
3-16.
78. Harezlak, J., et al., Persistence of HIV-associated
cognitive impairment, inflammation, and neuronal
injury in era of highly active antiretroviral treatment.
AIDS, 2011. 25(5): p. 625-33.
79. Valcour, V. and R. Paul, HIV infection a nd dementia in
older adults. Clinical Infe ctious Diseases, 200 6. 42(10):
p. 14 49-54 .
80. Scott, J.C., et al., Neurocognitive consequences of HIV
infection in older adults: an evaluation of the “cor tical”
hypothesis. AIDS and Behavior, 2011. 15(6): p. 1187-
119 6 .
81. Negin, J. and R.G. Cumming, HIV infec tion in older
adults in sub-Saharan Africa: extrapolating prevalence
from existing data . Bulletin of the World Health
Organiz ation, 2010. 88(11): p. 847-853.
82. Negin, J., E.J. Mills, and T. Bärnighausen, Aging with
HIV in Africa: the challenge s of living longer. AIDS
(London, England ), 2012. 26(0 1): p. S1.
83. Hontelez, J.A.C., et al., The impact of antiretroviral
treatment on the age composition of the HIV epidemic
in sub-Saharan Africa. AIDS (London, England), 2012.
26( 0 1).
84. Lekoubou, A., J.B. Echouffo-Tcheugui, and A.P.
Kengne, Epidemiology of neurodegenerative diseases
in sub-Saharan Africa: a systematic review. BMC
Public Health, 2014. 14: p. 653.
85. Habib, A.G., et al., Neurocognitive impairment in HIV-
1-infected adults in Sub-Saharan Africa: a systematic
review and meta-analysis. International Journal of
Infectious Diseases, 2013. 17(10): p. e820-31.
86. Lekoubou, A., J.B. Echouffo-Tcheugui, and A.P.
Kengne, Epidemiology of neurodegenerative diseases
in sub-Saharan Africa: a systematic review. BMC
public health, 2014. 14(1): p. 1.
87. Pas cal, M., et al., Determinants of Neurocognitive
Impairment in HIV in a Cohor t of Patients on
Antiretroviral Therapy Followed in Bangui (Central
African Republic). Neuroscience and Medicine, 2016.
7(01): p. 1.
88. Howlett, W.P., et al., Neurological disorders in AIDS
and HIV disease in the northern zone of Tanzania. Aids,
1989. 3(5): p. 289-296.
89. Per riëns, J.H., et al., Neurological complications
of HIV-1-seropositive internal medicine inpatients
in Kinshasa, Zaire. Journal of acquired immune
deficiency syndromes, 1991. 5(4): p. 333-340.
90. Bélec, L., et al., Low prevalence of neuro-psychiatric
clinical manifestations in centr al African patie nts with
acquired immune deficiency syndrome. Transactions
of the Royal Society of Tropical Medicine and Hygiene,
1989. 83(6): p. 844-846.
91. Maj, M., et al., WHO Ne uropsychiatric AIDS Study,
cross-sectional phase II: neuropsychological and
neurolo gical findings. A rchives of general psychiatry,
1994. 51(1): p. 51-61.
92. Joska, J.A ., et al., Cha racter ization of HIV-Associated
Neurocognitive Disorders among individuals starting
antiretroviral therapy in South Africa. AIDS & Behavior,
2011. 15(6): p. 1197-203.
93. Yusuf, A .J., et al., Prevalence of HIV-associated
neurocognitive disorder (HAND) among patients
attending a tertiary health facility in Northern Nigeria.
Journal of the International Association of Providers of
AIDS Care (JIAPAC), 2017. 16(1): p. 48-55.
94. Yakasai, A.M., et al., Prevalence and correlates of
HIV-associated neurocognitive disorders (HAND)
in Northwestern Nigeria. Neurology research
international, 2015. 2015.
95. Oshinaike, O.O., et al., Comparison of the Minimental
State Examination Scale and the International HIV
Dementia Scale in assessing cognitive function in
Nigerian HIV patients on antiretroviral therapy. AIDS
research and treatment, 2012. 2012.
96. Kelly, C.M., et al., HIV Associated Neurocognitive
Disorders (HAND) in Malawian Adults and Effect on
Adherence to Combination A nti-Retroviral Therapy: A
Cross Sectional Study. PloS one, 2014. 9(6): p. e98962.
97. Wong, M.H., et al., Frequency of and risk factors for
HIV dementia in an HIV clinic in sub-Saharan Africa.
Neurology, 2007. 68(5): p. 350-355.
98. Holguin, A., et al., HIV-1 effects on neuropsychological
performance in a resource-limited country, Zambia.
AIDS & Behavior, 2011. 15(8): p. 1895-901.
99. Ibara, J.R., et al.,[Acquired immunodeficiency
syndrome in elderly persons in a tropical zone.
Apropos of 175 cases in the Congo]. Bulletin de la
Societe de Pathologie Exotique, 2002. 95(2): p. 100-2.
100. Sack tor, N.C., et al., The Inte rnational HIV De mentia
Scale: a new rapid screening test for HIV dementia.
AIDS, 2005. 19(13): p. 1367-74.
101. Nakasujja, N., et al., Depression symptoms and
cognitive function among individuals with advanced
HIV infection initiating HA ART in Uganda. BMC
Psychiatr y, 2010. 10: p. 44.
102. Nakku, J., E. Kinyanda, and S. Hoskins, Prevalence
and factors associated with probable HIV dementia
in an African population: A cross-sectional study of
an HIV/AIDS clinic population. BMC psychiatr y, 2013.
13(1): p. 126.
103. Belete, T., G. Medfu, and E. Yemiyamrew, Prevale nce
of HIV associated neuroco gnitive deficit among HIV
positive people in Ethiopia: a cross se ctiona l study
at Ayder Referral Hospital. Ethiopian journal of health
sciences, 2017. 27(1): p. 67-76.
104. Patel, V.N., et al., High prevalence of suspected HIV-
associated dementia in adult Malawian HIV patients.
International Journal of STD & AIDS, 2010. 21(5): p.
356-8.
105. Lawler, K., et al., Neurocognitive impairment among
HIV-positive individuals in Botswana: a pilot study.
Journal of the International AIDS Society, 2010. 13: p.
15.
106. Atashili, J., et al., Prevalence, characteristics and
correlates of a positive-dementia screen in patients on
antiretroviral therapy in Bamenda, Cameroon: a cross-
sectional study. BMC Neurology, 2013. 13: p. 86.
107. Haddow, L.J., et al., A systematic review of the
screening accuracy of the HIV Dementia Scale and
International HIV Dementia Scale. PloS one, 2013. 8(4):
p. e 61826.
108. Kanmogne, G.D., et al., HIV-associated neurocognitive
disorde rs in sub -Saha ran Africa: a pilot study in
Cameroon. BMC Neurology, 2010. 10: p. 60.
109. Clifford, D.B., et al., Neurological eva luation of
untreated human immunodeficiency virus infected
adults in Ethiopia. Journal of Neurovirology, 2007.
13(1): p. 67-72.
110. Robertson, K.R., et al., Pattern of neuropsychological
performance among HIV positive patients in Uganda.
BMC neurology, 2007. 7(1): p. 8.
111. Sack tor, N., et al., Antiretroviral therapy improves
cognitive impairment in HIV+ individuals in sub-
Saharan Africa. Neurology, 2006. 67(2): p. 311-4.
112. Sacktor, N., et al., HIV-as sociated cognitive impairment
in sub-Saharan Africa—the potential effect of clade
diversity. Nature Clinical Practice Neurology, 2007.
3(8): p. 436- 443.
113. Sunmonu, T.A., et al., Intellectual impairment in
patients with newly diagnosed HIV infection in
southwestern Nigeria. BioMed research international,
2015. 2015.
114. Njamnshi, A.K ., et al., Risk factors for HIV-associate d
neurocognitive disorders (HAND) in sub-Saharan
Africa: the case of Yaounde-Cameroon. Journal of the
Neurological Science s, 2009. 285(1-2): p. 149-53.
115. Royal, W., et al., Clinical features and preliminary
studies of virological correlate s of neurocognitive
impairment among HIV-infected individuals in Nigeria.
Journal of neurovirology, 2012. 18(3): p. 191-199.
116. Vandenbroucke, J.P., et al., Strengthening the
Reporting of Observational Studies in Epidemiology
(STROBE): explanation and elaboration. Epidemiology,
2007. 18(6): p. 805-35.
117. Wu, Y.T., et al., Dementia in western Europe:
epidemiological evidence and implications for policy
making. L ancet Neurol, 2016. 15(1): p. 116-24.

41DEMENTIA IN SUB-SAHARAN AFRICA
Personal story
South Africa – Mr C
Mrs C and her husband live in a house in Umlazi, a township outside
Durban. They have two sons and a daughter. Around two years ago,
Mr C’s behaviour started to change. He was repeating himself and
kept telling stories of his youth, and on many occasions he would
get lost. On one such occasion, he was travelling with a friend to Durban. When
his friend was busy, he slipped away, asking a stranger for a lift to Inchanga - the
area where he grew up. When they arrived in Inchanga, the man who gave him a
lift soon realised that Mr C was confused and did not recognise the area. He took
him to Pietermaritzburg Police station.
In the meantime, the friend whom Mr C had travelled with realised he had
disappeared, and alerted Mrs C. That whole day and night, she and her children
searched for Mr C, stopping at various hospitals, the railway station, and the
morgue – frantically trying to find him. The following day, the Pietermaritzburg
Police contact the Umlazi Police station informing them of Mr C’s whereabouts.
The family were extremely relieved, but knew they had to take measures to ensure
his safety. They had a family meeting, and it was decided that Mrs C would quit
her job to become a full-time carer for her husband.
Around this time, in 2015, she contacted Alzheimer’s South Africa for information
and assistance. She attended a ‘Caring for the person with dementia’ course at
the Alzheimer’s South Africa office in Pinetown. The family employed three carers
to assist Mrs C, and to take care of Mr C’s daily and nightly needs. Mr C spends
much of his time in bed and needs assistance to eat, but he goes out into the
yard, with assistance, to exercise his legs and retain as much mobility as possible.
The family follows a holistic approach to his care, with support from a neurologist,
minister of religion, physiotherapist, nurse, community development worker, as
well as family and friends.
Having realised that other families in the area are facing similar challenges,
Mrs C has opened her home to allow a community development worker from
Alzheimer’s South Africa to run a support group for carers and family members of
people affected by Alzheimer’s disease, which meets on a monthly basis. Mr C’s
experience shows how a loving and caring family, with support, can ensure that a
loved one affected by Alzheimer’s disease can live a higher quality life, retaining
their dignity and self-respect.

ALZHEIMER’S DISEASE INTERNATIONAL42
CHAPTER 3
Living with dementia in
sub-Saharan Africa: systematic
review of qualitative evidence
3.1 Methodology
The value of qualitative evidence alongside findings from quantitative methodologies is increasingly
recognised in health sciences. Qualitative methods offer the opportunity to put flesh on the
bones of statistics by foregrounding individual real-life experiences of people living with illness
and their caregivers. Understanding beliefs, perceptions and experiences relating to illness, care,
help-seeking and treatment are critical to understanding the nature of the problem and designing
interventions and services that are person-centred and which meet locally defined needs.
Qualitative evidence about dementia in SSA is therefore an essential complement to quantitative
understanding of the scale of the problem.
We carried out a systematic review of the scientific qualitative evidence on the experience of living
with dementia in SSA. We were interested in all kinds of qualitative research: findings based on
studies using any kind of qualitative methodologies. This included: open-ended in-depth interviews,
focus-group discussions and ethnographic methods. These methodologies differ from quantitative
research in that they aim to capture the voices, perceptions and experiences of participants
through more open-ended interactions with researchers, which are often more naturalistic in tone.
Research participants were: people living with dementia (PLWD), family caregivers of PLWD, paid
caregivers (delivering treatment or care in the homes of PLWD or at public or private day care or
residential facilities), healthcare workers, traditional and faith healers with experience of treating
PLWD, community members.

43DEMENTIA IN SUB-SAHARAN AFRICA
We used the following search strategy to identify relevant articles. We used the search terms
“(dementia OR Alzheimers)” AND “(qualitative OR anthrop* OR ethnog* OR cultural OR understand*
OR belie* OR experien* OR phenomenol*) AND “(Africa OR sub-Saharan Africa OR {list of countries
included in SSA according to World Bank definition})”. We searched four databases: Embase,
MEDLINE, Global Health and PsychInfo. Following this, we checked the reference lists of all
selected articles. Chapter authors, experts in ageing in SSA/qualitative methodologies, reviewed
the list of selected articles and made additional suggestions for possible inclusions.
Inclusion criteria:
• Studies reporting results of primary qualitative research, carried out in SSA countries
• Studies published in peer-reviewed scientific journals
• Studies focussed upon dementia, i.e. with participants who: have a diagnosis of dementia or
have symptoms consistent with this diagnosis; carers of these people; healthcare workers/
healers involved in the care of this group; community members talking about this group
• English language publications
Exclusion criteria:
• Editorials and opinion pieces, letters where no primary data is presented
• Studies carried out among samples not resident in sub-Saharan African countries
• Quantitative studies
• Studies focussed more generally upon older people or older people with other chronic
conditions
In order to combine and summarise the results of the different studies identified, we used a
“thematic synthesis” approach1 to our analysis. This is a recognised approach widely used by
qualitative researchers to examine consistencies and contradictions across different studies,
with the aim of achieving insights which go beyond simple description of individual study results.
This approach required the extraction of the results sections of included articles, which were
then entered into a qualitative software programme (OpenCode 4.02), in which format they were
then coded. This involved free coding of excerpts of text which were organised into ‘descriptive’
themes; this was followed by further interpretation and development of higher level ‘analytical’
themes.
3.2 Search results and characteristics of selected studies
The search retrieved 1129 abstracts. On the basis of an initial screening, we removed 295 records
due to duplication. One article not identified in the database search (accepted for publication by
a peer reviewed journal but not yet published) was added to the screening list. Eight hundred and
thirty-five titles and abstracts were screened by RM for relevance and 28 full-texts were assessed
for eligibility. Articles were excluded for the following reasons: 475 because they were not clearly
focussed on dementia; 309 did not appear to report the results of primary qualitative research; 37
were duplicates not previously identified as such; 3 studies were not published in peer reviewed
publications (1 x dissertation; 2 x conference abstracts). One study was not carried out in SSA. Ten
articles were ultimately selected for inclusion in the synthesis.
The selected studies were carried out in South Africa (n=5), Tanzania (n=2), Nigeria (n=1), Ghana
(n=1) and Congo (n=1). Family caregivers of PLWD were the most common participant group
included (n=8)3-10, followed by PLWD (n=4)3, 5, 8, 10 (see Table 3.1). Three out of four studies carried
out among paid caregivers were conducted among residential facility staff, mostly nurses, in South
Africa7, 9, 11. The final study was carried out with hospital workers in the Republic of Congo5. A study
carried out in Tanzania included traditional and faith healer participants, whilst the only Nigerian
study was carried out with community members8. In-depth interviews were the most common
methodology, used in all but the Nigerian study, which was based on data from focus-group
discussions12. Authors’ research questions focussed on experiences of symptoms, understandings
of dementia and causality, caregiving arrangements and management strategies.

ALZHEIMER’S DISEASE INTERNATIONAL44
3.3 Results
Four over-arching themes encompassing ten sub-themes emerged from the data:
1. Beliefs
• Understandings of dementia
• Beliefs about causality
2. Experiences
• Symptoms
• Stigma
• Impact of caring for a person living with dementia
3. Responses
• Care arrangements
• Help-seeking
4. Unmet needs
• Emotional support
• Knowledge
• Economic support and services
Beliefs
Understandings of “dementia”
“I have never heard of [dementia]. Nevertheless, I know that when you
get older, you become like a small child. You have problems, worries,
and memory loss.”
Relative of PLWD, Republic of Congo5
When prompted by interviewers, no equivalent term for dementia was identified in any local
languages3, 5, 8, 10. In general, there was felt to be a lack of awareness of dementia among the
community10, a nurse in South Africa described how, before she started working in residential care,
she perceived dementia as a “white person’s disease”9. Health problems of older people were
rather characterised as problems associated with having “grown very old… ‘hangyena’” in Ghana3
or “ugonjwa wa wazee” disease of old people in Tanzania10. Often these general descriptors
encompassed dementia-like symptoms as well as symptoms of other chronic health problems3,5.
However, in Tanzania, in some cases, participants highlighted specific symptoms “ugonjwa
wa kusahau” (memory loss disease/disease of forgetting)10 within this syndrome. In Ghana,
participants recognised the specific role of the deterioration of the brain as a component of overall
bodily decline3.
Beliefs about causality
“You know when you buy something new and you use it for a long
time it becomes old, it does not remain like you bought it. That is how
the human brain is: when you are young it works well, but when you
grow old it does not work well – so when someone becomes old then
every part of the body too becomes old”
Grandson of PLWD, Ghana3
Five papers reported results related to the underlying cause of dementia-like symptoms. The most
prominent belief regarding underlying cause of dementia was that although not experienced by
everyone, dementia-like symptoms were part of “normal ageing”. Metaphors from nature and from
material objects were used to describe decline in functionality over time3, 8, 12. Both stress5, 8 and
grief3 were perceived to be associated with dementia-like symptoms. Witchcraft was understood
to be a cause for a minority of participants in the studies carried out in Ghana and Tanzania3, 10. In
Tanzania, one faith healer ascribed dementia to punishment by God8. A minority of participants in
the Mushi et al study were unsure as to the cause of dementia10.

45DEMENTIA IN SUB-SAHARAN AFRICA
Experiences
Symptoms
“At first she was able to go to the farm and gather firewood. Even
when you tried to stop her, she would not mind you- she was very
strong. She would fetch all the firewood a tractor could bring, as if
three young men went to fetch that firewood. Now she cannot even
come out of her room, not even to talk of going to the farm to gather
firewood. When you ask her, she complains of joint pains and that she
cannot walk, that is how I got to know that she has changed. Then I
already mentioned to you that when I tell her something immediately I
come back to ask again she has forgotten about it. So, that too I have
seen that she has changed”
Daughter of PLWD, Ghana3
Participants most frequently reported problems with memory and communication, e.g. forgetting
appointments, forgetting where they put things, repeating themselves3, 5, 8, 10. Problems with overall
functionality were also reported, as well as problems typically associated with co-morbidities
(pain, sight/hearing problems) in Ghana and Tanzania3, 8. Caregivers in Congo reported behavioural
problems such as irritability and suspicion5.
Stigma
“When persons act in strange ways the talk starts in villages and
townships: they say ‘that one has been bewitched’…they go to these
old people, vandalise their houses, burn their houses... they accuse
them of being witches”
Residential facility nurse, South Africa9
Symptoms of dementia meant that PLWD were the subject of teasing within their communities
in Ghana, South Africa and Nigeria3, 4, 12. The meaning of this was interpreted differently by
participants. In Ghana, this was not perceived to transgress the boundaries of respect for older
people3, however, in South Africa, the fact that a PWD was perceived as a joke by the community
was interpreted as an indirect threat to the older person as this meant that their safety and
wellbeing were not taken seriously4. When PLWD were perceived to be witches, as reported in
Nigeria, fear inevitably resulted in discrimination, isolation and instances of violence4. In Ghana,
Congo and Tanzania, where participants stated that there was no stigma as dementia-like
symptoms were accepted as a “normal part of old age”3, 5, 10, symptoms were associated with a
much-changed role. In Ghana and Congo, PLWD were routinely referred to as childlike3, 5, whilst in
Nigeria, symptoms (forgetfulness) were associated with a loss of authority12.
Impact
“Since he is sick, everything is done by me, he cannot do anything.
That is why there are changes in my life, and I feel like a person who
is also sick”
Wife of PLWD, South Africa6
The strongest theme related to the impact of the condition was the effect that the need for
supervision had upon the activities of the caregiver, limiting their ability to do other things as they
used to, namely, income generating activities (in Ghana and South Africa)3, 6, 10 as well as social
activities6 (in South Africa). Participants in Ghana and South Africa who presented caregiving as a
collective responsibility shared among family members3, 4 sounded more positive about the burden
of caregiving as the impact upon any one individual was felt to be minimised. In South Africa,
the positive impact of the work of caregiving was highlighted by one family caregiver6 and one
residential care-worker11 . However, residential care-workers highlighted the difficulties associated
with aggressive behaviour, passivity and racism from residents. Costs of care - transportation,

ALZHEIMER’S DISEASE INTERNATIONAL46
food, medication - were mentioned as an additional burden that was often difficult to manage in
Ghana, Congo and South Africa3, 5, 6, particularly in the context of other economic burdens within
the family3. In Ghana, although most older people had health insurance and this was perceived to
be helpful, it was perceived to cover only a small proportion of the costs associated with chronic
illness3.
Responses
Care arrangements
“Each one has a role to play. During the day, I am not there. My dad’s
niece actually comes in to help my wife look after him. And as far as
his medication is concerned - that is where my sister comes in; she
will take care of the medical side of things. I am on the financial side
of everything”
Male caregiver of PLWD, South Africa4
Care, mainly delivered by families, was shaped by a need to show gratitude for the caregiving
provided to the current carers by PLWD earlier in their lives- findings from South Africa indicate
a typology of reciprocity4 and a notion of collectivity and responsibilities associated with that
kinship6. Results from Ghana, South Africa and Tanzania suggest that women tend to be the
hands-on caregivers for PLWD3, 4, 10, with men often playing a role in decision-making about
care3, particularly related to finances4. There was evidence that caregiving was often a collective
responsibility with different components assigned to different family members in Ghana3,
commonly children of the PLWD but in Tanzania also involving grand-children10. Two studies from
South Africa provided insight into the role of paid care (respite and residential) from the perspective
of family caregivers and residential facility staff. Hired help was perceived as a difficult but
necessary component of the caregiving strategy for PLWD, in order to give family members a break
so that they were able to regain energy to continue in their role4. Residential staff outlined strategies
for managing often challenging symptoms of PLWD: these centred around communicating in
such a way as to encourage trust and respect - using touch, verbal and non-verbal ways to
communicate11. In another study carried out in South Africa, nurses reported the positive impact
of serving patients traditional foods which was felt to trigger reminiscence and joy, resulting in a
calming effect7. Approaches were felt to be rendered ineffective when PLWD were over-medicated,
resulting in passivity, which consequently care workers found difficult to manage11.
Help-seeking
“Yes I will bring them to you [allopathic services] and I see the
collaboration will be good… because you will treat them, sometimes
they get treated in hospital and sometimes they come to us and we
treat them. So it won’t be a problem because you can’t do everything
by yourself!”
Traditional healer, Tanzania8
The most salient overall theme regarding help-seeking was one of pluralism: in Ghana and
Tanzania, participants used both biomedical, traditional and faith healing to treat the problems of
older people3, 10. This was often experimental, with participants trying out one type of treatment,
monitoring the response and then switching to a different one as necessary3, 10. However, there
was some evidence that explanatory models influenced patterns of help-seeking/attitudes to
treatment. In Tanzania, traditional healers believed that they could cure cases where the cause was
supernatural but not if the only cause was ageing8. Older participants shared this view in Congo5.
Some older people and their family members shared this view that modern medicine could not
help problems that were deemed to be “beyond human control”10. Similarly, in seeking help from
biomedical services in Ghana and Tanzania, participants sought help from biomedical services
for problems other than dementia-like symptoms, aiming to improve the overall comfort of older
people, rather than cure problems entirely3, 10. In Tanzania, faith healers and prayers (their modus
operandi) were perceived to be helpful but not curative and were seen as a “safe” treatment8, 10, as
compared to traditional healers whom participants in Tanzania and South Africa perceived to be

47DEMENTIA IN SUB-SAHARAN AFRICA
potentially dangerous5, 8. In Tanzania, faith and traditional healers were positive regarding potential
collaboration with biomedical services in the management of PLWD, and requested ongoing
involvement in case management of patients they referred, in addition to training and education
about dementia8.
Unmet needs
Emotional support
“Emotionally it’s a strain and there’s no way you can do it alone. You
need someone. My advice to anyone would be to use whatever help
you can get. Don’t push anyone away, the more hands the better”
Family caregiver of PLWD, South Africa4
The need for emotional support for caregivers was a strong theme in two South African studies4, 6.
This took the form of having someone to talk to and good communication between family members
involved in the care of the older person4. A lack of emotional support was perceived to manifest as
feelings of depression and strain4. This was felt to be essential to enable the caregiver to continue
to deliver the work of caregiving. Costs of transportation to clinics, namely, hiring transport, were
perceived as burdensome6 .
Knowledge
“Doctors should educate us more because we don’t know much
about this problem”
Family caregiver of PLWD, Tanzania10
Education, in order to improve knowledge and understanding of dementia, among caregivers and
local communities was deemed to be essential: both to counteract beliefs related to witchcraft
which in South Africa, were perceived to put older people living with dementia at risk9 and in South
Africa and Tanzania were education was felt to be necessary to improve the level of support for
caregivers in their role4, 10. Non-Governmental Organisations (NGOs) (Dementia South Africa and
Alzheimer’s South Africa) were identified as useful sources for information4. In fact, the potential
for NGOs and civil society to take the lead in terms of educating communities and caregivers was
recognised by participants in South Africa and Tanzania9, 10 .
Economic support and services
“I wish they could give us soup. She can no longer swallow even
nutritional porridge… She can’t swallow at all... and wound care
medicines, because when taking her to the clinic one needs to hire
transport and pay for all those things”
Family caregiver of PLWD, South Africa6
In South Africa and Tanzania, people living with dementia and their caregivers were dependent
upon families and communities in to support the costs of care4, 6, 10. Costs of transportation to
clinics were felt to be particularly problematic6. One nurse participant in South Africa suggested
that government provision of local, township-based geriatric facilities would not only provide a
much-needed service for PLWD but could also play an important role in improving community
understandings of the condition9. In Tanzania, it was suggested that churches and NGOs “should
actively be involved in this problem like in other social problems” and that one way in which they
might do this could be through provision of day centres.
3.4 Discussion
Our systematic review reveals a small but useful evidence base which offers important insights
into the everyday reality of living with dementia in SSA. Experiences of living with dementia and
the responses of families, health services and communities take place in the context of cultural
understandings of dementia and beliefs about the causality of the condition. The biomedical label

ALZHEIMER’S DISEASE INTERNATIONAL48
of dementia was absent from participant narratives in all but studies carried out in residential care
facilities in South Africa. As has been found elsewhere, although not perceived to be an inevitable
component of the ageing process, dementia was generally seen as a part of the overarching
condition of growing old which afflicted some but not all older people13-15. This perceived normality
helped to differentiate dementia from “madness” that is perceived to affect younger people,
despite the overlap of some symptoms. Our findings suggest that belief in witchcraft or other
sorcery as a cause of dementia was a minority view, albeit one that was associated with negative
consequences: stigma and discrimination, violence and threat of violence. It is important to
note that this finding emerged alongside a strong appetite for information and knowledge about
dementia among caregivers, care-workers, healers and community, in recognition of the fact that
a lack of understanding of the condition fuelled fear and discrimination and perpetuated negative
beliefs to those affected by the disease.
Although most participants reported that PLWD were not subjected to stigma or discrimination,
there is evidence to suggest that their symptoms were perceived as a regression to a childlike
state, which inevitably led to infantilisation, loss of authority and discontinuity of role. In HIC, both
maintaining a sense of continuity and participating in life have been identified as important factors
in determining quality of life of PLWD16. The voices of people living with dementia were largely
absent from findings, even when study authors reported that they were included as interviewees. In
HIC where there are large numbers of older people and dementia is recognised as a public health
priority, the importance of knowledge derived from first-hand experience of dementia is recognised
as a key contributor to comprehensive understanding of the disease and its impact16.
As has been found elsewhere, in relation to help-seeking for mental disorders, explanatory models
also influenced patterns of help-seeking and attitudes to treatment of dementia17. PLWD and their
caregivers as well as healers, generally believed that if the cause of symptoms was supernatural,
traditional or faith healing was likely to be more efficacious than biomedical treatments. That
said; families were also experimental and pluralistic in their approach to treatment of any kind,
monitoring responses and switching or stopping a particular approach if it was perceived to be
having no positive effect. There was consensus that if the cause was ageing, the condition was
incurable. Families and healers alike focussed upon addressing remediable co-morbidities and
providing holistic treatments to improve the comfort and wellbeing of the older person.
Despite the fact that rapid demographic ageing means across the region means that a failure to
address the health of older people will inhibit attainment of development goals, health systems in
SSA remain focussed upon services for infectious diseases, child health and adults of reproductive
age and are not equipped to deal with chronic health problems associated with older age18. Healers
outside the biomedical system are arguably further ahead of biomedical health services in providing
holistic, community-based care which meets the needs of older people. It is therefore unsurprising,
that although unclear, the proportion of PLWD in SSA who seek help from non-biomedical healers
is likely to be significant. Mushi et al. found that 41% of participants in the qualitative study had
visited Christian faith healers and 19% had visited traditional healers10. Hindley et al. found that
traditional and faith healers were positive about the idea of collaboration with biomedical services
in the treatment and management of problems associated with older age8. Indeed, the potential
advantages of collaboration between different models of healing have been recognised but not yet
fully implemented in the context of increasing access mental healthcare in low and middle income
country settings19. Likely benefits include: cultural acceptability, accessibility, and a person-centred
approach19.
Consistent with demographic ageing and changing epidemiological profile, quantitative evidence
points to a growing prevalence of functional impairment among older people in SSA and a growing
need for support from others to carry out essential everyday tasks20. Caregiving is managed
by families with no involvement or support from government sources. The sustainability of this
approach is highly questionable. It is notable that many of the symptoms reported by caregivers
as the most difficult to manage at home (such as aggression and incontinence) are largely absent
from narratives identified in our review. This may be an artefact of the focus of topic guides used
in the studies included in this review, which tend not to focus on obtaining detailed symptomology
of the condition. In any case, these symptoms are common features of mid-stage dementia,
which will only become more common as SSA populations rapidly age. It is of note that the only
evidence related to caregiving outside of the household comes from South Africa. The call for more
training about dementia and narratives relating to the challenges of managing difficult behaviour
associated with dementia are illustrative of the challenges accompanying the necessary provision

49DEMENTIA IN SUB-SAHARAN AFRICA
of long-term care around the world. In some middle income countries, societal changes (such as
greater participation in the labour market, urbanisation, higher levels of divorce, later marriage), are
leading the established role of female family members as unpaid caregivers to become contested21.
Findings illustrate the additional financial pressure that caring for a PLWD can bring to often already
stretched household budgets. A lack of formal governmental provision to support the economic
impacts of older age and disability means that the family are the sole source of social security for
older people22. It remains to be seen to what extent traditional notions of filial duty and reciprocity
will hold fast in the face of societal change and “modernisation”22, 23.
3.5 Recommendations
Provision of culturally sensitive education about dementia
Findings suggest that there is a desire across different stakeholder groups to gain a better
understanding of dementia, in order to both counteract stigma and to enhance management of the
condition. Although associations between beliefs in the supernatural as the cause of dementia and
stigma should be addressed, it is important to do this using a culturally sensitive approach. In the
absence of effective biomedical treatments for dementia, the medicalisation of dementia should
be approached with caution. There is little evidence that replacing existing local understandings
of dementia with Western diagnostic categories will lead to improvements in outcomes for PLWD
and their families. Rather, a culturally sensitive but common-sense approach is required. Utilisation
of existing conceptualisations of dementia as a normal part of old age in combination with feasible
practical advice for caregivers and PLWD regarding the management of symptoms would seem
to be the best pathway to provision of education that offers the best opportunity of being effective
whilst remaining culturally acceptable and supportive of the norm of familial reciprocity. Some
research participants suggested that NGOs, civil society and religious institutions might be best-
placed to deliver this kind of training within community and healthcare settings. This possibility
should be explored.
Advocacy to strengthen the voices of people living with dementia
The design of policies and services which successfully meet the needs of PLWD and their carers
will be impossible without the input derived from lived experience of the disease. Evidence from
research carried out in Ethiopia suggested that societal stigma severely limited the ability of
people living with severe mental illness to participate in research related to the design of mental
health services24. Therefore, it is important that governments, non-governmental organisations
and researchers make efforts to address this disparity and find innovative ways to ensure that the
voices of PLWD are heard. Lessons in how to achieve this be may be learnt from elsewhere: both
from other successful global advocacy movements for disadvantaged and stigmatised groups (e.g.
people living with HIV) as well as from the empowerment of PLWD in HIC settings.
Development and scale-up of integrated health services for older people
Although policy frameworks directed at ageing populations are beginning to emerge, these have
yet to effect change in policy and practice25. Design and implementation of services designed to
meet the needs of older people in SSA lags behind that of mental health, itself a marginalised area
of public health in the region. The urgent need to build capacity within primary care to manage
chronic conditions and co-morbidities associated with old age has however been recognised by
WHO. ICOPE is a WHO-led programme to develop evidence-based guidelines for the prevention
and management of dependence by non-specialist healthcare workers, focussing on problems
(such as nutrition, mobility, falls, cognition, mood and behaviour, sensory impairment and
incontinence) rather than specific conditions26, 27. The potential for integration of healers from
outside the biomedical system into the implementation and scale-up of ICOPE should be explored.
Working with existing pluralist approaches to healthcare for older people is likely to enhance
acceptability and healthcare workers may have much to learn from traditional and faith healers
experiences of managing the health problems of older patients.
Development of policies related to long-term care and economic security of older
people
Evidence from elsewhere suggests that in the face of changing societal norms and ageing
populations, family systems of social security struggle to cope with increased numbers of older
people living with chronic health problems, leading to indebtedness, family strain and conflict and
household rationing of healthcare and food21. The generalisability of these findings to the SSA

ALZHEIMER’S DISEASE INTERNATIONAL50 51DEMENTIA IN SUB-SAHARAN AFRICA
Table 3.1
Description of selected studies
Authors/Country Participants Research questions/aims Themes identified by authors
1Adebiyi et al 2016
Nigeria12
Community members (n=36): older people,
middle-aged, youth
To investigate the occurrence of implied/
enacted stigma
Implied stigma
Enacted stigma
2Agyeman et al 2017
Ghana3
Caregivers and PLWD (n=28) To explore sociocultural beliefs,
understandings, perceptions and
behaviours related to dementia
To explore beliefs about causality and
stigma and help-seeking behaviours
To explore care arraignments and the
impacts of caregiving
Symptoms
Understandings of cognitive symptoms
Help-seeking
Course of cognitive illness
Care arrangements
Economic impacts of care
Stigma
3Deist and Greef 2017
South Africa4
Children of older PLWD who are caregivers (n=21) To identify, explore and describe resilience
factors that are present in families caring
for a PWD
Social support
Financial stability
Religion and spirituality
Optimism
Positive communication patterns
Family closeness
Acceptance
Managing the illness
4Faure-Delage et al 2012
Republic of Congo5
PLWD (n=27); their family members (n=31);
cognitively impaired older people (n=90); their
family members (n=92); hospital workers (n=33)
To describe the sociocultural
representations of dementia in Brazzaville
Knowledge about the phenomenon called ‘dementia’
Recognition of dementia
Patterns of distress
Perceived causes
Help-seeking
5Gurayah et al 2015
South Africa6
Caregivers of PLWD (n=5) To explore experiences of being a carer for
a PLWD
Views and responsibilities of the caregiver
Impact of caregiving
Services to assist the caregiver
6Hanssen and Kuven 2015
South Africa7
Nurses in geriatric facilities (n=4); family members
of PLWD (n=1)
What does it mean to institutionalised
patients with dementia to be served
dishes they recognise from childhood and
youth
Belonging and joy
Improved appetite
Food and reminiscence
7Hindley et al 2016
Tanzania8
Traditional healers (n=11)/faith healers (n=10)
PLWD (n=18) (and their carers, n=17)
How do traditional and faith healers
understand and manage people with
dementia?
Why do PLWD present to traditional and
faith healers
Conceptualisation of dementia by healers: normal part of ageing process
People with dementia and carer reasons for seeking help and experiences of treatment: prayers,
plants and witchcraf t
Collaboration with allopathic healthcare services
8Mkhonto et al 2017
South Africa9
Nurses in residential care facilities (n=7)
Family member of PLWD (n=1)
To explore and describe the link between
culture and dementia care, focussed on
the belief in dementia as witchcraft and
PLWD as witches
Belief in witchcraft causing fear of PLWD
Need of knowledge and education
9Mushi et al 2014
Tanzania10
People living with dementia (PLWD) (n=21)
Carers (n=16)
What are the perceived causes and
symptoms of dementia?
What are the experiences of PLWD and
their carers and what meaning is attached
to these?
What are the health seeking behaviours of
PLWD and their carers?
Knowledge and perception of dementia
Symptoms of dementia
Stigma against dementia
Health seeking behaviour
Source of social support
Views on how to address the problem
10 Van Wyk et al 2016
South Africa11
Staff in residential care facilities (n=17) Explore unmet needs of staff working with
residents with dementia in long-term care
facilities
Explore which behaviours of PLWD staff
found challenging and how they managed
these
Hear from staff how they learn to deal
with distressing/challenging behaviour
Attitudes toward PLWD
Distressed behaviour
Management techniques
Training and education
Managing distressing behaviour more effectively

ALZHEIMER’S DISEASE INTERNATIONAL52
region requires further research23. With the numbers of older people living with dementia projected
to rise in SSA more rapidly than elsewhere and an almost total lack of governmental support for
older people, there is an urgent demand for governments to begin to develop and implement
policies related to both economic security of older people and long-term care for chronic
conditions associated with older age.
3.6 Strengths and limitations of our approach
We believe this is the first synthesis of qualitative research evidence on dementia from the SSA
region. This is timely. As quantitative evidence relating to the prevalence, incidence and overall
burden of disease advances, it is equally important to develop the knowledge base around lived
experience of the disease. Both of these will be important in terms of shaping the next steps of the
research and policy agenda. There are a number of limitations to our approach. Chosen search
terms were pragmatic, chosen to limit the number of results in order to identify highly relevant
records. However, results will only include studies using the category of “dementia”. It is possible
that we have missed studies which have samples of people with dementia-like symptoms who
were not identified as such. Some authors may have used the broad term of cognitive decline or
cognitive impairment as synonymous with dementia, but we restricted our search to dementia. We
note that, while unlikely, the modern American DSM 5 recommending the non-use of “dementia”
where possible (but instead using major neurocognitive disorder) may have prompted some
researchers in SSA after 2013 to have studied but not used the term “dementia”. We searched
titles and abstracts published in English only. It is possible that we may have missed some relevant
articles published in other languages. We included articles published in peer reviewed journals only.
Although this acted as a mechanism for quality control, it means that we have excluded all grey
literature. It is plausible that reasonable quality, relevant evidence exists in the form of reports and
student dissertations.
Given the heterogeneity of cultures, languages and beliefs within the region, selection of SSA as
an area of interest is potentially misleading. The extent to which findings from particular settings
are generalizable to other countries, cultures and communities within the region is debatable. This
is perhaps most striking when considering the likely relevance of findings from South Africa, an
upper middle income country with an epidemiological and sociodemographic profile unlike that
found elsewhere in SSA, to experiences of people living in communities elsewhere in the region.
For example, South African studies included a minority of white participants, whose experiences,
due to historical, political, social and economic conditions are highly unlikely to be representative
of the majority of South Africans affected by dementia, let alone people living elsewhere in SSA.
That said, there are some common threads which support the relevance of SSA as a region in
the context of dementia and ageing research and advocacy18. These include: rapid demographic
ageing, under-resourced and over-stretched health systems, and a lack of prioritisation of the
health and social security of older people on public health and policy agendas.

53DEMENTIA IN SUB-SAHARAN AFRICA
References
1. Thomas, J. and A. Harden, Methods for the thematic
synthesis of qualitative research in systematic reviews.
Bmc Medical Research Methodology, 2008. 8.
2. University of Umeå. OpenCode 4.0. 2013 20/06/2017];
Available from: http://www.phmed.umu.se/english/
units/epidemiology/research/open-code.
3. Agyeman N, G.M., Nyame S, Tawiah C, Owusu-Agyei
S, Prince M, Mayston R,, “When someone be comes
old then every par t of the body too becomes old”:
experie nces of living with dementia in Kintampo, rural
Ghana (in press). Transcultural Psychiatry, 2017.
4. Deist, M. and A.P. Greeff, Living with a parent with
dementia: A family resilience study. Dementia: The
International Journal of Social Re search and Practice,
2017. 16(1): p. 126-141.
5. Faure-Delage, A ., et al., Socio-Cultural Pe rceptions
and Representations of Dementia in Brazzaville,
Republic of Congo: The EDAC Survey. Dementia and
Geriatric Cognitive Disorders Extra, 2012. 2(1): p. 84-
96.
6. Gurayah, T., Caregiving for people with dementia in
a rural context in South Africa. South African Family
Practice, 2015. 57(3).
7. Hanssen, I. and B.M. Kuven, Moments of joy and
delight: the meaning of traditional food in dementia
care. Journal of clinical nursing, 2016. 25(5-6): p. 866 -
874 .
8. Hindley, G., et al., The role of traditional and faith
healer s in the treatment of dementia in Tanzania
and the potential for collaboration with allopathic
healthcare ser vices. Age & Ageing, 2016. 04: p. 04.
9. Mkhonto, F. and I. Hanssen, When people with
dementia are perceived as witches. Consequences for
patients and nurse education in South Afric a. Journal
of Clinical Nursing, 2017. 30: p. 30.
10. Mushi, D., et al., Social representation and practices
related to dementia in Hai District of Tanzania. BMC
public health, 2014. 14: p. 260.
11. van Wyk, A ., J. Manthorpe, and C. Clark, The
behaviours that dementia care home staff in South
Africa find challe nging: An exploratory study.
Dementia, 2016. 12: p. 12.
12. Adebiyi, A.O., et al., Enacted and implied stigma
for dementia in a community in south-west Nigeria.
Psychogeriatrics, 2016. 16(4): p. 268-273.
13. World Health Organisation, A .s.D.I., Dementia: a public
health priorit y. 2012: Geneva.
14. Connell, C.M., et al., Black and white adult family
members’ attitudes toward a dementia diagnosis. J Am
Geriatr Soc, 2009. 57(9): p. 1562-8.
15. Ineichen, B., The epidemiology of dementia in Africa: a
review. Soc Sci Med, 2000. 50(11): p. 1673-7.
16. Gorska, S., K. Forsy th, and D. Maciver, Living With
Dementia: A Meta-synthesis of Qualitative Research on
the Lived Experience. Gerontologist, 2017.
17. Bhikha, A.G., et al., A systematic review of explanator y
models of illness for psychosis in developing countries.
Int Rev Psychiatry, 2012. 24(5): p. 450-62.
18. Ab oderin, I.A. a nd J.R. Beard, Older people’s health
in sub-Saharan Africa . Lancet, 2015. 385(9968): p. e9-
e11.
19. Gureje, O., et al., The role of global traditional and
complementary systems of medicine in treating mental
health problems. Lancet Psychiatry, 2015. 2(2): p. 168-
17 7.
20. Ab oderin, I. and J. Hof fman, Research Debate on
‘Older Carers and Work’ in Sub -Saharan Africa?
Current Gaps and Future Frame s. J Cross Cult
Ge r o n tol, 2 017.
21. Mayston R, L.-S.P., Gallardo S, Wang H, Huang Y,
Montes de Oca V, Ezeah P, Guerra M, Sosa AL, Liu
Z, Uwakwe R , Guerchet M, Prince M,, A journey
without maps- understanding the costs of caring for
dependent older people in Nigeria, China, Mexico and
Peru (in press). PLoS ONE, 2017.
22. Heslop A, G.M., Chronic poverty and older people in
the developing world, in Chronic Povert y Research
Centre Working Paper. 2002.
23. Ab oderin, I. and J. Hof fman, Families, Intergenerational
Bonds, and Aging in Sub-Saharan Africa. Can J Aging,
2015. 34(3): p. 282-9.
24. Mayston, R., et al., Participator y planning of a primary
care ser vice for people with severe mental disorde rs in
rural Ethiopia. Health Policy Plan, 2016. 31(3): p. 367-
76.
25. World Health Organisation, Multisectoral ac tion for a
life course approach to healthy ageing: draft global
strategy and plan of action on ageing and health. 2016:
Geneva.
26. World Health Organisation, Integrate d care for older
people. 2017 09/06/2017]; Available from: http://www.
who.int/ageing/health-systems/integrated-care/en/.
27. A, T.J., et al., Identifying common impairments in frail
and dependent older people: validation of the COPE
assessment for non-specialised health worker s in low
resource primary health care settings. BMC Geriatr,
2015. 15: p. 123.

ALZHEIMER’S DISEASE INTERNATIONAL54
CHAPTER 4
Social protection and the health
of older people in sub-Saharan
Africa
4.1 Introduction
Conceptually, social protection refers to any intervention that seeks to reduce individuals’
vulnerability to adverse outcomes as a consequence of specific risks1. Among policy-makers,
this is usually limited to two broad areas of intervention: regular cash payments and insurance
against potential healthcare costs. Across international development networks increased attention
has been given to social protection over the past 15 years, particularly cash transfers for mothers
and pensions for older people2. This chapter examines the development of social protection in
sub-Saharan Africa and assesses effects on the health status of older people. It is not possible to
do justice to the diversity of regional experience, so the chapter mainly focusses on Ghana and
South Africa. Country selection reflects data availability and the presence of social protection
schemes of particular interest. The chapter starts by briefly reviewing the development of pension
provision across the region, followed by a similar review of health insurance. The final part includes
discussion and analysis of the effects of these interventions on health outcomes, followed by some
policy recommendations.
4.2 Pensions for older people
As in other regions, it is helpful to distinguish between non-contributory “social” pensions, and
separate contributory schemes providing retirement benefits for specific sets of workers. It is
possible to establish a broad categorisation of pension provision in sub-Saharan Africa. A number
of countries have relatively extensive and generous social pension schemes, usually combined with
separate contributory arrangements. These countries are mainly located in southern Africa and
include South Africa, Botswana, Namibia and Mauritius. In the rest of the region, the coverage of
pension schemes and the value of benefits are considerably more limited, although all countries
have generous provision for the military, police and civil service. The coverage of contributory

55DEMENTIA IN SUB-SAHARAN AFRICA
schemes is limited by the small number of people employed in the formal sector of the labour force
and by low wages, which reduce capacity for making monthly contributions. The coverage of social
pension schemes is limited by state fiscal capacity and competing priorities in public expenditure.
A number of countries have developed pilot social pension schemes, usually with the support of
external funding. In many cases, such as Ghana, these provide only a small number of low value
benefits. In other cases, such as Kenya, more progress has been made to scale-up social pension
provision, although benefits remain much lower than schemes in southern Africa.
Across sub-Saharan Africa the focus of academic research and policy debate is almost entirely on
social pensions. Despite this, far larger amounts of public funds are usually devoted to contributory
schemes for relatively privileged social groups. For example, in Uganda 0.4 percent of GDP is
spent on civil service pensions, but these cover only 2.5 per cent of the older population3. In
some countries in the region civil servants are permitted to retire at 45 or even younger, and in
some cases pensions are worth the full amount of final salaries3. This neglected aspect of social
protection in Africa is highly inequitable and crowds out public expenditure for more vulnerable
groups of older people.
4.2.1 South Africa: an example of extensive pension provision
In South Africa non-contributory pensions (known as old age grants) are well-established, dating
back to the Apartheid era. Old age grants are notionally means-tested and available for men aged
65 and over and women aged 60 and over. It is estimated that around 80 per cent of eligible older
people received an old age grant4. The basic grant value is about US$110 a month (US$112 for
people aged 70 and over). In 2017, 3.3 million old age grants are being paid out. The scheme
is funded through general taxation, and costs around 1.5 per cent of GDP. Additionally, the
government pays out around 1.1 million non-contributory disability benefits to adults aged under
60, including chronic conditions such as Alzheimer’s disease.
In recognition of the costs of paid care at home, the Department of Social Development offers
a “Grant-in-Aid” benefit to older people who “require regular care”. Eligibility requires a medical
report or assessment report (less than three months old) that shows the applicant needs full-time
care. In theory, medical eligibility for the grant should be reviewed every year. In 2012, 68,000
benefits worth around £17 a month were being paid out, at a total cost of £14 million. In its
Strategic Plan for 2012-15, the Department observed that there was a need to generate greater
awareness of these grants5. Progress seems to have been made: in 2016, the number of benefits
being awarded had risen to 152,000. However, the geographical distribution of the benefits remains
uneven. In 2015, 18 per cent of South Africans aged 70 or over lived in the province of Kwazulu
Natal, but the province accounted for 33 per cent of Grant-in-Aid benefits.
South Africa’s current system of non-contributory cash transfers has been a focus of international
interest for both researchers and policy-makers working in ageing and development. The scheme
has been put forward as an example of “best practice” for other developing countries to emulate6.
It has been claimed that the social pension does not just benefit older people, but is pooled across
entire households7. Indeed, in settings of high unemployment, it may represent the only reliable
source of household income. It is sometimes claimed that household dependence on the old age
grant increases the authority of older people and hence the potential willingness of family members
to care for frail older relatives8. However, a separate issue is whether older people share their
pension income willingly or are coerced or bullied into doing so9. Frail older people, including those
with cognitive impairments, may be particularly vulnerable to financial exploitation.
South Africa also has several large contributory pension schemes, although these receive little
attention from academics or policy-makers outside the country. These cover people with lifetime
employment in the formal sector of the labour market, mainly focussing on more affluent groups.
For example, in 2000 over 60 per cent of white men and 40 per cent of white women aged over
60 received an employer-provided private pension; for black Africans coverage was under 10 per
cent10. By 2004 contributory pension funds were paying out around 1.1 million benefits: roughly
half the number of old age grants11. South Africa has a separate set of pension provisions for civil
servants, providing around 230,000 retirement benefits a year12.

ALZHEIMER’S DISEASE INTERNATIONAL56
4.2.2 Ghana: a more limited pension provision
Ghana can be taken as broadly representative of most other countries outside southern Africa,
where coverage mainly consists of contributory schemes for relatively affluent occupations and
where social pensions are largely non-existent.
Ghana’s main contributory fund, the Social Security and National Insurance Trust (SSNIT), provides
retirement pensions mainly for civil servants and formal sector workers. Coverage of the labour
force is estimated at around 10 per cent and in 2001 the scheme paid out 50,000 pensions,
representing less than 5 per cent of Ghana’s population aged 60 and over. Some efforts have been
made to extend protection to informal sector workers with a dedicated voluntary scheme, but
uptake has been low due to high rates of cash poverty. Despite high administration costs, SSNIT
pension claims typically take around four months to process. In 2008 the government approved
a bill to combine different contributory funds into a single three pillar system, but this makes no
specific provision for non-contributory social pensions.
The Livelihood Empowerment Against Poverty Program (LEAP) has been extended across Ghana
since March, 2008 and reaches over 35,000 households13. LEAP benefits are worth between US$8
and US$15 per month, depending on the number of qualifying individuals in the household. The
scheme targets a range of different vulnerable population groups, of which one is people aged 65
and over. Specific data on the number of older people receiving LEAP payments are unavailable,
but since there were 934,000 people aged 65 and over in 2015, only a small minority of them are
likely to be covered by the scheme. As such, the large majority of older Ghanaians receive no
pension: either from the contributory system or from LEAP.
4.2.3 New universal social pension schemes
Since the 1990s, often with the support of overseas agencies, a number of sub-Saharan African
countries established pilot social pension programmes. However, typically these schemes have
been small-scale, means-tested and pay small pensions. For example, in 2011 Uganda’s Senior
Citizens Grant paid a monthly pension of US$7 to around 4 per cent of its population aged 60 and
over. In some cases, these pilot schemes were discontinued after international funding stopped.
One sub-Saharan African country conformed to neither the more embracing southern African
model nor the minimalist model found in elsewhere. In 2004 Lesotho established a universal
pension for all people aged 70 and over worth around US$37 a month. More recently, a number
of other sub-Saharan African countries have established similar schemes. For example, in 2016
Zanzibar established a universal pension worth US$9 a month for people aged 70 and over, with
Kenya proposing to set up a similar scheme from 2018.
4.3 Health insurance: does Ghana show the way forward?
In sub-Saharan Africa, only a small number of health insurance schemes have attained more than
30 per cent coverage of their populations. In most countries, low rates of enrolment and high rates
of drop out mean coverage struggles to exceed 10 per cent. Coverage is highly concentrated
on more affluent social groups. In Nigeria, for example, only 4 per cent of the population, mainly
federal government workers, are covered by public health insurance schemes14. In contrast to its
generous pension coverage, South Africa does not yet have a universal health insurance system,
with coverage concentrated among civil servants and formal sector workers. The South African
ministry of health does not aspire to achieve universal coverage until 2026 at the soonest15.
Ghana represents an exceptional case, with a national health insurance scheme that has received
much international attention and is sometimes framed as a potential model for other resource-
constrained countries. Around 66 per cent of the population are enrolled in the scheme16.
Membership requires payment of an annual registration fee, ranging from around US$3 to US$30,
depending on the socio-economic profile of the registration district. People over the age of 69 with
contributory pensions and/or enrolled in LEAP are exempt from making payments. This explains
why rates of enrolment are higher among over 70s (63 per cent in 2011) than for adults aged 18
to 49 (32 per cent)17. In theory, the scheme offers a minimum package that covers 95 per cent of
diseases reported in health facilities in Ghana and requires no co-payment. In practice, most NCDs,
such as most forms of cancer, are not covered18. More generally, questions have been raised about
the capacity of the national health insurance scheme to deliver improved health outcomes without
substantial upgrading of Ghana’s limited health service infrastructure19.

57DEMENTIA IN SUB-SAHARAN AFRICA
4.4 The effects of social protection on the health status of
older people
Increasing attention is being given to the potential of pensions and other cash transfers to improve
health outcomes of older people. A recent study focussed on Ethiopia, Mozambique, Tanzania and
Zimbabwe observed a:
“positive impact of cash transfers on access to health services, but this impact is limited by the fact
that many vulnerable older people are excluded, and benefit levels are low”.20
It also noted:
“Cash transfers cannot compensate for major weaknesses in health systems”20.
Likewise, a more general review of cash transfers and health in low and middle-income countries
concludes:
“There are good reasons to be somewhat sceptical about the promise of these programmes…”21
The availability of and access to health services (including preventive services) across most of sub-
Saharan Africa is extremely limited, particularly for NCDs and for poorer groups living in rural areas.
As such, the capacity of social pensions to have significant impacts on the health status of older
people will remain very limited unless substantial improvements are made to the suitability, quality
and accessibility of these services. In other words, cash transfers should not be viewed as a stand-
alone intervention.
The challenges of translating pension coverage into improved health for older people are evident in
South Africa. Here, despite the generosity of the old age grant and other pension schemes, older
people experience one of the highest rates of uncontrolled hypertension reported worldwide22.
Key contributing factors include low levels of health awareness among older people and the
generally poor quality of health services, especially in rural areas. The most important risk factor for
hypertension (and several other common NCDs) is obesity. A national survey conducted in 2008
found that 72 per cent of people aged 50 or over were either overweight or obese22. Consequently,
addressing high rates of disease among older people in South Africa will require a combined
strategy of awareness-raising, health promotion and investment in key health services.
It might be expected that health insurance would have a more direct effect on enhancing health
outcomes for older people. Just as with pensions, however, any effect is contingent on the
generosity of the scheme, health awareness and the availability of suitable services. Simply being
enrolled in health insurance or utilising health services should not in itself be considered positive
outcomes, unless it can be shown that these lead to improved health status. This is particularly
true where the quality of services is uneven, as is the case in most of sub-Saharan Africa. Despite
the many studies of Ghana’s national health insurance scheme, robust evidence of enhanced
health outcomes for older people is relatively scant. It has been shown that rates of awareness of
hypertension are higher among older people with insurance (31 per cent) than those without (18
per cent)23. Likewise, rates of hypertension control are higher for those with insurance (6 per cent)
than those without (4 per cent). Thus, while insurance does improve outcomes to some degree, 94
per cent of insured older people with hypertension were still not effectively managing the condition.
Robust data for other NCDs, such as diabetes and heart disease, are not available, but patterns are
likely to be similar.
4.5 Conclusion
Overall, the effects of social protection on the health status of older people in sub-Saharan Africa
remain limited by a range of different factors. Most critically, there is an urgent need to bolster the
“supply-side” of health service infrastructure. There is also a need to re-orientate services away
from a near-exclusive focus on infectious disease and mother and child health towards NCDs and
conditions associated with later life20, 24. Social pensions have considerable potential to enhance
the wellbeing of older people, especially in those countries where nutritional intake is widely
inadequate. The real benefits of health insurance for older people will remain marginal and notional,
unless entitlement to services is translated into genuine access.
Consequently, the most effective approach to enhance the well-being of older people in much of
sub-Saharan Africa would be to combine a general reorientation and upgrading of basic health
services with extended provision of cash transfers. It is often argued that limited fiscal resources

ALZHEIMER’S DISEASE INTERNATIONAL58
represent an insurmountable barrier to such interventions. However, this chapter shows that
many countries in the region already devote considerable resources to contributory pension and
health schemes, which only include small, more affluent sectors of their population. With specific
reference to dementia, there may be particular scope to enhance health worker knowledge of
the condition and of simple interventions, such as cognitive stimulation. This should come to be
seen as a core component of sub-Sahara African primary health care systems, rather than an
unaffordable luxury.
References
1. Hozmann, R . and S. Jorgensen, Social protection as
social risk management: conceptual underpinnings for
the social protection sector strategy pape r. . Journal of
International Development., 1999. 11(7): p. 1005-1027.
2. Devereux, S. and P. White, Social protection in Africa:
evidence, politics and rights. . Poverty and Public
Policy, 2010. 2(3): p. 53-77.
3. Schwartz, A. and M. Abels, Issues for civil service
pension reform in sub-Sa haran Africa. , W.B. Group,
Editor. 2016.
4. Woolard, I., K . Hart tgen, and S. Klasen, The evolution
and impac t of social security in South Africa.,
S.A.L .a.D.R. Unit, Editor. 2010, University of Cape
Tow n.
5. Depar tment of Social Development, Strategic Plan
2012-2015. . 2012, Government of South Africa, .
6. HelpAge International, Age a nd security. How social
pensions can deliver effective aid to poor older people
and their families. 2004, HelpAge International: London
7. Lloyd-Sherlock, P., et al., Pensions, poverty and
wellbeing in later life: comparative research from South
Africa and Brazil. Journal of Aging Studies 2012. 26(3):
p. 243 -252.
8. Ambler, K., Bargaining with Grandma: The Impact of
the South African Pension on Household Decision
Making. . 2011, University of Michigan. .
9. Lloyd-Sherlock, P., P. Penhale, and P. Ayiga, Financial
abuse of older people in low and middle-income
countries: the case of South Africa. 2017.
10. Lam, D., M. Leibbrandt, a nd V. Ranchhod, Labor force
withdrawal of the elderly in South Africa, in Aging in
sub-Saharan Africa: recommendations for furthering
research. , B.M. Cohen, J.;, Editor. 2006, National
Research Council. : Washington D.C.
11. Van den Heever, A., Pension reform and old age
grants in South Africa., in Mimeo, . 2007, University of
Pretoria.
12. Hendricks, F., The private affairs of public pensions in
South Africa: debt, development and corporatisation.
2009, UNRISD: Geneva.
13. Handa, S., et al., Livelihood empowerment against
povert y programme impact evaluation report. 2013,
Carolina Population Center, University of North
Carolina.: Chapel Hill, NC.
14. Onoka, C., et al., Promoting univers al financial
protection: constraints and enabling factors in scaling-
up coverage with social health insurance in Nigeria.
Health Re search Policy and Systems, 2013.
15. Ataguba, J.E., C. Day, and D. McInty re, Monitoring
and evaluating progress towards Universal Health
Coverage in South Africa. PLoS Med, 2014. 11(9): p.
e100168 6.
16. Jehu-Appiah, C., et al., Household perceptions and
their implications for enrolment in the National Health
Insurance Scheme in Ghana. Health Policy and
Planning, 2012. 27(3): p. 222-233.
17. Duku, S., C. van Dullemen, and C. Fenenga, Does
Health Insurance Premium Exemption Policy for Older
People Increase Access to Health Care? Evidence from
Ghana. . Journal of Aging and Social Policy., 2015.
27(4): p. 331-347.
18. Bosu, W., A comprehensive review of the policy and
programmatic response to chronic non-communicable
disease in Ghana. . Ghana Medical Journal 2012. 46(2
Suppl): p. 69-78.
19. Witter, S. and B. Ga rshong, Something old or
something new? Social health insurance in Ghana.
Bmc International Health and Human Rights, 20 09. 9.
20. HelpAge International, Cash transfers and older
people’s access to healthcare: A multi-countr y study in
Ethiopia, Mozambique, Tanzania and Zimbabwe. 2017.
21. Fer nald, L.C., Promise, and risks, of conditional cash
transfer programmes. Lancet, 2013. 382(9886): p. 7-9.
22. Lloyd-Sherlock, P., et al., Hypertension among
older adults in low- and middle-income countries:
prevalence, awareness and control. Int J Epidemiol,
2014. 43(1): p. 116-28.
23. Lloyd-Sherlock, P., et al., Social protection and
the health of older people in developing countries:
establishing the health effects of pensions and health
insurance. Inter national Social Security Review, 2012.
65(4): p. 51-68.
24. World Health Organization, Wor ld report on age ing and
health. 2015, World Health Organization,: Geneva.

59DEMENTIA IN SUB-SAHARAN AFRICA
Personal story
Mauritius – Mrs C
Mrs C is 85 years old, and lives at home with her unmarried son and
daughter. Around eight years ago, they began to notice that their
mother was becoming forgetful, but she was still very socially active
– visiting friends, and taking the bus daily into the city of Port Louis
to visit a friend’s shop. Mrs C’s forgetfulness amplified, as she started forgetting
to pay bills and to take her antihypertensive medications, and misplacing small
things. However, she still enjoyed her independence and was going out alone
undeterred.
Things changed three years ago when the family built a new house, and Mrs C
became disorientated by the change and had difficulty with daily activities. Her
children did not want her to take the bus alone, but she resisted and this became
difficult for them to handle. On one occasion, she decided to go to hospital on
her own and did not come back until late. When asked where she had gone, she
said she was unwell. Her family contacted a doctor and she was diagnosed with
dementia.
Since her diagnosis, Mrs C is not on any medications, but has attended the
Alzheimer Centre organised by Alzheimer Association Mauritius. She started
attending twice a week, and spending the rest of the week with a neighbour, but
has since started attending daily during the week. In the evening she stays at
home with the supervision of her daughter, and at the weekend is supported by
her children, who repeat the activities they have learned with her at the Centre.
Mrs C is active and still likes to go to malls and to the sea side. She sometimes
repeats herself, but her children are tolerant and patient with her. Thankfully, she
has good familial support and she receives social support from the Centre. She
is benefitting a lot and is now calmer, having received a lot of love and human
warmth.

ALZHEIMER’S DISEASE INTERNATIONAL60
CHAPTER 5
Conclusions and
recommendations
5.1 Summary
5.1.1 Prevalence and numbers affected, costs, incidence, and mortality
We estimate now that 2.13 million people were living with dementia in sub-Saharan Africa in 2015,
with numbers projected to nearly double every 20 years, increasing to 3.48 million by 2030 and
7.62 million by 2050. These estimates are slightly higher than those for the 2015 World Alzheimer
Report1. The increases are accounted for by two factors. First, the 2015 UN population estimates
include larger numbers of older people than the 2012 estimates which had been used previously.
Second, the evidence base regarding the prevalence of dementia has slightly expanded, including
data from three additional studies. This resulted in a slight increase in age-standardised prevalence
of dementia in SSA than had been estimated previously, but it does not mean that prevalence has
increased since.
These estimates are now closer to the estimates for other high income or low and middle income
regions, and suggest that previous estimates may have underestimated the scale of the dementia
epidemic in the region.
Based on these latest estimates of numbers of people living with dementia in SSA, the total
costs of dementia are now estimated at US$ 6.2 billion in 2015. The upward revision of the costs
compared to the previous estimates is only explained by the larger numbers of people with
dementia estimated in this report. Distributions of the costs were not different from previous
estimates, with two-thirds to three-quarters of the total costs and cost per person attributed to
informal care costs. Contribution of formal social care to overall costs remains low in the region.
However, one important limitation is the absence of relevant published research on resource use
and costs of dementia in the region. Our cost estimates are therefore mainly relying on imputations.
From our review and meta-analysis of incidence rates in SSA, we estimate that there were over
367,000 new cases of dementia in 2015, most of them arising in Eastern and Western SSA.
Although this number is lower than our previous estimate, for the first time we have been able to

61DEMENTIA IN SUB-SAHARAN AFRICA
provide an estimate relying on evidence of dementia incidence from SSA countries only, rather than
relying on evidence from all LMIC.
We reviewed for the first time the evidence on mortality risk for people with dementia living in SSA.
Pooling the results from four studies carried out in Western and Central Africa, we estimated that
people with dementia had a two-fold increased mortality risk compared to people not affected by
this condition. This risk is similar to the hazards reported in other low and middle income countries.
Furthermore, the evidence was indicating a higher mortality risk for people with more severe
dementia compared to the ones with lower severity.
5.1.2 Risk factors
In the course of this review, we have examined the evidence existing on factors associated with
dementia in sub-Saharan Africa. The evidence covers non-modifiable (age, sex, genetics) as well
as modifiable risk factors (education, vascular disease risk factors, early-life, lifestyle, psychosocial
and psychological risk factors).
While both older age and female sex show strong associations with higher dementia prevalence
and incidence as it is the case more globally, the evidence on genetic risk factors (mainly focused
on the APOE gene) remains limited and contradictory. When investigated, the APOE ε4 allele, which
is the allele with the greatest known effect on the risk of developing dementia, does not seem to be
systematically associated with dementia despite a high frequency within sub-Saharan populations.
The possibility of gene-environment or gene-gene interactions that would explain those negative or
weak associations with dementia needs to be explored further.
Most of the evidence regarding modifiable risk factors is provided by cross-sectional rather than
longitudinal cohort studies, which would demonstrate aetiological pathways more clearly. The
strongest evidence on modifiable risk factors is probably the association of vascular disease
risk factors (hypertension, hypercholesterolemia and peripheral artery disease) with dementia.
Evidence on early-life, psychological, psychosocial and lifestyle risk factors remains very tentative.
Considering the scale of the HIV/AIDS epidemic in sub-Saharan Africa and the ageing of the
population, including of those treated by antiretroviral therapy, HIV infection seems a relevant
potential risk factor to investigate in the region. Our review on HIV-associated dementia (HAND)
revealed that the evidence in older adults is completely lacking, either because older people were
excluded from most of the studies on HAND or because population-based studies on dementia
among older people did not screen for HIV or HAND. This is an important knowledge gap that
needs to be addressed in future studies.
5.1.3 The experience of people living with dementia
Understanding beliefs, perceptions and experiences relating to illness, care, help-seeking and
treatment are critical to understanding the nature of the problem, and are therefore an essential
complement to quantitative understanding of the scale of the epidemic of dementia in SSA. A small
but useful evidence base was revealed in our systematic review, offering valuable insights into how
people live with dementia in this region.
The biomedical label of dementia was absent from narratives in almost all studies. The only
exception was found in residential care settings in South Africa – in other studies dementia was
generally seen as a part of the ageing process, which affected some but not all older people.
Beliefs in witchcraft or other sorcery were found to be a minority view, although one that was
associated with stigma and discrimination, violence and threat of violence. It was recognised that
the lack of awareness and understanding of dementia could contribute to fear and discrimination,
and negative beliefs relating to those affected by the condition. On the other hand, perceptions
of symptoms shown by people living with dementia as a regression to a childlike state were not
generally linked to stigma or discrimination. However, despite their contribution to the research, the
voices of people with dementia were largely absent from reported narratives and findings.
There was a general belief that if the cause of symptoms was supernatural, traditional or faith
healing would be more efficient than biomedical treatments. But families were also pluralistic
in their approaches to treatment. The condition was always recognised as incurable if the
acknowledged cause was ageing. The proportion of people with dementia who seek help from non-
biomedical healers in SSA was significant.

ALZHEIMER’S DISEASE INTERNATIONAL62
For now, the only evidence related to caregiving outside of the household comes from South Africa.
When care was delivered by families, women tended to be the hands-on carers whilst caring was
also a collective responsibility in some settings. Education was felt to be necessary to improve the
level of support for carers in their role. The potential for non-governmental associations (including
Alzheimer associations) and civil society to lead education initiatives towards communities and
carers was recognised.
Yet the evidence identified covered only five countries, with all sub-regions other than Southern
SSA represented by only one or two studies. Given the heterogeneity and diversity of cultures,
languages and beliefs within SSA, the generalisability of those results to countries or settings other
than those where they originated is debatable.
5.1.4 Social protection for older people
With the rapidly growing number of older people living in SSA, who are more likely to develop
chronic diseases such as dementia, increasing needs in support and health and social services
are forecasted. Social protection, which seeks to reduce individuals’ vulnerability, is therefore a
necessity for older people across SSA. This is even more true for people living with dementia.
In that perspective, we examined the development of social protection programs in SSA, mainly
based on cash transfers and health insurance, and assessed their effects on the health of older
people by looking in particular at two countries for which data was available (South Africa and
Ghana).
A number of countries, mainly in Southern SSA, have social pension schemes mostly combined
with contributory arrangements. Some countries have developed pilot social pension schemes
with the support of external funding while others have dedicated larger amounts of public funds to
develop contributory schemes for relatively privileged social groups (civil servants). Overall social
protection in SSA is highly inequitable and most often excludes the most vulnerable groups of older
people.
In SSA, only a small number of health insurance schemes have reached more than 30% coverage
of their populations. Coverage is usually highly concentrated on more affluent social groups.
However, Ghana’s national health insurance scheme, free for people aged over 69 years, is quite
unique and has been presented as a potential model for other resource-constrained countries.
For now, the capacity of this scheme to improve health outcomes in the context of limited health
services is debatable.
The availability of and access to health services across most of SSA is very limited, including
for older people. There are significant challenges of translating pension coverage and health
insurance into improved health and social outcomes. For now, the effects of social protection on
the health status of older people in SSA remain limited by several factors, including weak health
infrastructures and services still focussed on infectious diseases and child and maternal health.
Social pensions have potential to enhance the wellbeing of older people. Health insurance for older
people will improve older people’s health only if entitlement to services ensures access to services
that meet their needs.
5.2 Global Action Plan on Dementia
In May 2017, the World Health Organization’s Global Plan of Action on Dementia was unanimously
adopted at the 70th session of the World Health Assembly in Geneva. The plan follows ten years of
advocacy by ADI for a global response to the growing impact of dementia worldwide, and includes
targets for the advancement of dementia awareness, risk reduction, diagnosis, care and treatment,
support for care partners and research (see Box 2). The Global Plan of Action on the Public Health
Response to Dementia 2017-20252 commits Member States to take practical steps by 2025 to
encourage progress in those different domains.

63DEMENTIA IN SUB-SAHARAN AFRICA
Box 2
The global plan on dementia: actions and targets
Action 1. Dementia as Public Health Policy
Target 75% of countries will have developed or updated national policies, strategies,
plans or frameworks for dementia, either stand-alone or integrated into other
policies, by 2025.
Action 2. Dementia Awareness and Friendliness
Target 100% of countries will have at least one functioning public-awareness campaign
on dementia to foster a dementia-inclusive society by 2025. 50% of countries will
have at least one dementia-friendly initiative to foster a dementia-inclusive society
by 2025.
Action 3. Dementia Risk Reduction
Target The relevant global targets defined in the Global Action Plan for prevention and
control of non-communicable disease (NCDs) 2013-2020 and any future revisions
are achieved for risk reduction and reported. (The risk factors for dementia are
the same as other NCDs such as cardiovascular diseases, cancers, chronic
respiratory diseases and diabetes. The following measures can reduce the risk
of cognitive decline and dementia: increase physical activity, prevent and reduce
obesity, promote balanced and healthy diets, stop smoking and harmful use of
alcohol, promote social engagement, cognitive stimulating activities and learning,
and prevent and manage diabetes and hypertension – especially in midlife – and
depression.)
Action 4. Diagnosis, Treatment, and Care
Target In at least 50% of countries, as a minimum, 50% of the estimated number of
people with dementia are diagnosed by 2025.
Action 5. Support for Dementia Care
Target 75% of countries provide support and training programs for carers and families of
people with dementia.
Action 6. Information systems for Dementia
Target 50% routinely collect a core set of dementia indicators through their national
health and social information systems every two years by 2025.
Action 7. Dementia Research and Innovation
Target The output of global research on dementia doubles between 2017 and 2025.
From Rees G., Australian Journal of Dementia Care 2017; 6(4):7-9.

ALZHEIMER’S DISEASE INTERNATIONAL64
5.2.1 What does the plan mean for sub-Saharan Africa?
While it is estimated that 58% of all people with dementia live in low and middle income countries1,
we have estimated in this report that 2.13 million people with dementia were living in sub-Saharan
Africa, a number that is expected to almost double every 20 years. Increasing and emerging needs
will therefore arise in this region as more dementia awareness might be induced by the increasing
numbers of people affected.
Health and care systems will be increasingly challenged. Specialist services, already under-
resourced, will not have the capacity to deal with the levels of demand, which will likely increase
faster than the development of the specialist workforce3. Primary care services, potentially better
suited to facilitate care coordination for people with complex multimorbidities (including dementia),
are currently not designed and not trained to assume such a responsibility. As highlighted
previously, the focus on infectious diseases and maternal and child health rather than non-
communicable and chronic diseases is still very strong in SSA. Besides the weakness of healthcare
services, the erosion of customary family care systems and old-age support (reputed to be an
asset for African populations) will certainly continue given the context of social, demographic
and economic changes within the region. The implications of these considerable changes and
increasing needs must be considered urgently by countries and the African Union.
For now, no dementia plans or strategies have been established by countries in SSA. The issue of
dementia is often within the remit of Ministries of Health and Social Affairs; sometimes included in
national plans for Non-Communicable Diseases or Mental Health. The World Health Organization
(WHO) has a role in monitoring health trends, disseminating information, and providing leadership,
policy guidance and technical assistance to governments. After almost a decade of efforts and
energy deployed towards the dementia field, with the inclusion of dementia in the 10 ‘priority
neurological and mental disorders’ for the WHO Mental Health Gap Action Plan (mhGAP)4, the
WHO/ADI joint report ‘Dementia: a public health priority’5 and the conference on Global Action
Against Dementia, the Global Action Plan on Dementia2 has now been adopted. This plan includes
specific actions for Member States, international, regional and national level partners. Indicators
and targets, which can be used to evaluate levels of implementation, progress and impact, have
been set. For the first time, governments can now be held accountable on the advancements of
dementia awareness, risk reduction, diagnosis, care and treatment, support for care partners and
research in their countries.
5.2.2 Awareness
Raising awareness is the foundation of the public health approach to addressing the dementia
epidemic. The core message is that dementia is a disease causing disability and not an inevitable
consequence of ageing5. Awareness-raising and understanding are essential to counter the fatalism
and stigma that can often be associated with dementia, and may also help family members and
carers to access support and cope better with their caregiving role. Awareness often needs to be
raised at different levels: the broader community, people with dementia and their families, and
those who provide health and social services to them.
Dementia awareness and friendliness are one of the seven action areas listed in the WHO Global
Plan. It is believed that increasing public awareness, acceptance and understanding of dementia
and making the social environment dementia friendly will enable people with dementia to
participate in the community and develop their autonomy2. We have seen in the past that ‘Dementia
Friendly Communities’ were popular among the governments that had developed policies and
plans. Dementia Friends initiatives are aiming at transforming the way the nation thinks, acts and
talks about dementia. It seeks a change in attitudes and behaviour towards dementia, for people
with dementia and their carers to be treated with respect and dignity, and for communities to be
able to support people with dementia so they can ‘live well with dementia’. One of the proposed
actions for the WHO Secretariat is to build upon the WHO Global Network of Age-friendly Cities
and CommunitiesA
1 to integrate and link dementia friendly initiatives, learning from previous
experiences of what works in different contexts. The current limitation is that the network does not
include any members (city or community) in the whole African continent, where experiences could
be more efficiently shared.
A http://www.who.int/ageing/projects/age_friendly_cities_network /en/

65DEMENTIA IN SUB-SAHARAN AFRICA
However, Dementia Friends initiatives have already been introduced in the sub-Saharan region. A
Dementia Friends programme was launched in Nigeria in early 2016 by Dementia Nigeria with the
support and guidance of the Alzheimer’s Society (UK), using outline and materials adapted from the
UK programmeB
2. Within a few months, over 21 Dementia Friends Champions (trained volunteers
who encourage others to learn a little bit about dementia) were spread in 19 of the 36 Nigerian
states. Information sessions were held in schools, hospitals, places of worship and other places
within villages and towns. Over 80,000 people in Nigeria have now become Dementia Friends and
positive changes within communities are described by Dementia Nigeria. The strength of their
programme was to get individual village chiefs and elders involved, who would then cascade the
learning to the lower chiefs and then to individual households within their communities. Having key
respected members of the community engaged in the programme helped to give the Dementia
Friends messages some weight. The role of associations in dementia education and awareness is
often key and has been recognised. Currently, only seven Alzheimer’s associations are members
of the ADI network, from countries spread over the sub-Saharan region. The Nigerian experience
regarding the implementation of the Dementia Friends initiative is a good opportunity for all those
associations and beyond to learn how this challenging task can be successfully approached.
5.2.3 Access to services
With the new United Nations Sustainable Development Goals, all UN Member States have agreed
to try to achieve Universal Health Coverage by 2030. This includes financial risk protection,
access to quality essential healthcare services and access to safe, effective, quality and affordable
essential medicines and vaccines for all. A challenge will have to be faced by many low and middle
income countries, especially in SSA where the availability of and access to health services can be
very limited, particularly for NCDs, populations from rural areas and for older people.
Problems of access to services are numerous, including limited help-seeking due to low awareness
and financial barriers (especially in case of a continuous need for care, when there is limited or no
reimbursement for health and social care, when people have insufficient outcomes to cover out-of-
pocket payments)1.
For countries in SSA, like many low and middle income countries, one of the main issues is the lack
of coverage of services. Specialists caring for people with dementia (neurologists, psychiatrists,
geriatricians, psychologists, occupational and physical therapists, specialist nurses) are too few (if
available at all) to provide services to a significant number of people with dementia, and they are
mostly based in the main urban areas. This problem is not exclusive to dementia and has been
identified for other non-communicable diseases or mental health domains1. However, the increase
in demand for services related to dementia will be particularly acute and will add strain on already
under-resourced specialist services.
‘Task-shifting’ or ‘task-sharing’ approaches have been presented as potential solutions in this
context, whereby much of the delivery of care is contributed by non-specialist primary care and
community services, which would be supported and trained by more experienced specialists.
Rationale for a move towards a more task-shifted model of care for dementia can be based on
two arguments3. First, mobilising the non-specialist primary care workforce can relieve resource
constraints and therefore facilitate scaling up of the model. This is particularly relevant for low
and middle countries including SSA. Secondly, this approach will likely help to reach productive
efficiency, when more healthcare services are provided at a given quality and cost or when the
same healthcare services are provided at a given quality but lower cost3. The cost of care per
person with dementia is then assumed to decrease. When considering the task-shifted model of
dementia care pathways for South Africa in the World Alzheimer Report 2016, the main challenges
identified remained the competing claims to limited resources, aggravated by inequities in access
to care, stigma, and lack of professional training and awareness3.
Although the task-shifted approach for dementia care is attractive, such a transition will not
be achieved overnight and assumes a complex process, including changes of roles and
responsibilities, additional resources and new models for the delivery of care. Non-specialists
will need to be trained to take on new and unfamiliar tasks, with the support and supervision of
specialists. Specialists will therefore become more focused on service delivery management rather
than delivery of frontline care (reserved for complex cases), and will need to ensure the coverage of
services and the quality of care provided.
B https://blog.alzheimers.org.uk/campaigns/dementia-friends-nigeria-tackling-a-global-health-epidemic/

ALZHEIMER’S DISEASE INTERNATIONAL66
For sub-Saharan African countries, like other low and middle income countries, the implementation
of task-shifted approaches will most likely rely on the WHO mhGAP guideline for dementia
diagnosis and management4. In summary, dementia healthcare would be provided by community
healthcare workers providing outreach in the community (hence boosting detection), a case
manager and the primary care physician, screening for those presenting with subjective memory
impairment. Only a small proportion of patients (5%) would be referred to the specialist because of
unusual presentations or young onset for example. In complement of the mhGAP guidelines, the
assessment and management of physical comorbidity to optimise physical health can be provided
by the community healthcare worker, through the WHO Integrated Care for Older People (ICOPE)
programme6. Post-diagnostic support and care would be provided by community healthcare
workers and primary care physicians.
The readiness of primary care services to assume such roles will vary among countries, depending
on existing levels of population coverage, workforce resources and competencies, and capacity to
provide continuous care for chronic diseases.
In the Global Action Plan, the WHO secretariat proposed to develop and implement guidelines,
tools and materials such as model training covering core competencies relating to dementia for
health and social care workers, as well as to provide guidance on strengthening the implementation
of the dementia component of the mhGAP programme to enhance capabilities of existing human
resources and train more staff, and the ability to provide quality care through primary care2.
5.2.4 Prevention
One of the key areas of the Global Action Plan is dementia risk reduction. While the strongest
evidence for possible causal associations with dementia identified in the 2014 Word Alzheimer
Report reviews were those with low education in early life, hypertension in midlife, and smoking
and diabetes across the life course7, the evidence from studies in SSA is not consistent. More
research needs to be dedicated to identifying potential risk factors in SSA, some of which might be
specific to the region, before we can even assess the likely impact of dementia risk reduction on
the incidence of dementia.
Considering that cardiovascular health is becoming increasingly compromised in many SSA
countries, actions to promote good health need to be supported without waiting for more robust
research to be carried out. Health promotion activities, publicising risk associations to achieve
change through aggregated modification of lifestyles and behaviours, have been effective in the
past improving cardiovascular health8. The focus on control of detection and treatment of diabetes
and hypertension, reduction of levels of obesity, smoking cessation, increased physical activity and
better education are public health priorities in many countries worldwide. This certainly needs to
be reinforced in SSA countries where non-communicable diseases are still competing for attention
with communicable diseases and where levels of education can still be low, especially in rural
areas. Yet, tackling NCD prevention and management in SSA countries, including the assessment
of risk factors, identification and active management of high risk status, early disease detection and
long-term follow-up with regular monitoring and promotion of adherence to treatment, means there
is a need for primary health care to be substantially strengthened9.
The WHO recommends linking dementia with other programmes, policies and campaigns on non-
communicable disease risk reduction, and raising awareness of the links between dementia and
other non-communicable diseases. The message that dementia, alongside heart disease, stroke
and cancer, may be prevented through increased adoption and implementation of public health
strategies needs to be heard by the public as well as policymakers.
The potential for dementia risk reduction has been modelled multiple times, mostly based on
data from high-income countries10-12 where age-specific dementia incidence has been reported
to decline13-16. However, even with urgent action to address this problem in LMIC and SSA, it is
unlikely that trends in dementia incidence and prevalence will follow the same patterns, at least
short- to medium-term.
5.2.5 Rights
People with dementia and their carers have the same human rights as every other citizen. It is
widely recognized that people with dementia are frequently denied their basic rights in both the
community and care homes. The lack of ethical, social and legal protection of people with dementia
in sub-Saharan countries is evident. People with dementia are not always involved in decision-

67DEMENTIA IN SUB-SAHARAN AFRICA
making processes and their wishes and preferences for care are not often respected. As mentioned
in this report, the voices of those affected by dementia are often forgotten or not sought, when they
might have a lot to tell us about their experience of living with dementia and how they would wish
their rights to be respected. It is essential that rights are recognized, respected and protected to
empower people with dementia, those who support them and the community as a whole.
The Global Action Plan for Dementia is grounded on cross-cutting principles, including respect
for the Human Rights of people with dementia and universal health and social care coverage for
dementia2. More recently, the World Health Organization’s newly elected director-general,
Dr Tedros Adhanom Ghebreyesus, reiterated that universal health coverage would be a priority
during his mandate, which would address the issue of health as a human right.
Within that framework, the basic principle of equity (the absence of avoidable or remediable
differences among groups of people, whether those groups are defined socially, economically,
demographically, or geographically) is therefore important. It means that all people affected by
dementia should be acknowledged as having equal status and value, and should be accorded
equal access to diagnosis, treatment, care and support, regardless of age, sex, socioeconomic
status, ethnicity or country of residence. With the perspective of a disease-course modifying
treatment for dementia by 2025, one of the impressive commitments of the G7 Global Action
on Dementia, it is essential that issues of equity are addressed. The WHO Global Action Plan
indicates that:
‘designing and implementing health programmes for universal health coverage must include
financial risk protection and ensuring equitable access to a broad range of promotive, preventive,
diagnostic and care services (including palliative, rehabilitative and social support) for all people with
dementia and their carers’2
As highlighted in the World Alzheimer Report 2015 and the present report, this objective will take
time and commitment to achieve in sub-Saharan Africa.
5.2.6 Research prioritisation
Dementia research is globally underfunded with respect to the burden of the disease and the
societal economic cost. While ADI and several international and national organisations have called
for nation states to dedicate 1% of their respective societal economic cost to dementia research,
investment in dementia research remains uneven globally.
The WHO Global Action Plan on Dementia proposed actions regarding dementia research and
innovation, stating that implementation of research aligned with identified priorities and social
and technological innovations can increase the likelihood of effective progress towards better
prevention, diagnosis, treatment and care for people with dementia. Actions for Member states
include:
• the development, implementation and monitoring of national research agendas on
prevention, diagnosis, treatment and care which would fill the gaps in evidence to support
policy and practice;
• the increase in investment in dementia research and innovative health technologies; and
• supporting the development of technological innovations that respond to the physical,
psychological and social needs of people with dementia, their carers, or people at risk of
developing dementia2.
Although this Action Plan will support the development of dementia research globally and guide
nations to define adequate research agendas, the results of this report show the gap in evidence
for almost all domains of the dementia field in the sub-Saharan region. Estimating the prevalence
and incidence of the disease in this region is still a task that needs to be prioritised, and new
studies need to be ideally designed to be able to monitor any secular trends in incidence and
mortality when feasible. Those studies should as much as possible be conducted in nationally
representative samples, and could potentially be nested within national surveys of health and
ageing. An alternative would be to build upon the existing networks of Health and Demographic
Surveillance Sites across sub-Saharan Africa to develop dementia studies, as has already been
done on ageing and health with the SAGE and HAALSI studies.

ALZHEIMER’S DISEASE INTERNATIONAL68
Future studies should also consider investigating a diverse range of areas using multiple
methodologies to help filling gaps of evidence highlighted in this report on:
• the experience of people with dementia and their carers, the understanding of the condition
and how it can lead to stigma, discrimination and help-seeking behaviours, that would
support adequate education, counteract stigma and enhance the management of dementia;
• the use of services, the description of care systems and care arrangements, especially
informal care, which would allow a better estimation of the costs of dementia;
• the potential effect of genetic and modifiable risk factors of dementia and how they relate
to geographic variations and changes in dementia prevalence and incidence, which would
support the relevant implementation of programmes, policies and campaigns on non-
communicable disease and dementia risk reduction;
• the impact of the HIV/AIDS epidemic on both the dementia epidemiology and the changes in
social and economic context for older people in SSA;
• the impact of social protection and health insurance on the access to services for people
with dementia and their carers.
Despite the striking gaps in evidence for the dementia field in SSA, we must acknowledge the
overall challenge brought by the epidemiologic and demographic transitions in a region where most
health systems are not equipped to provide the comprehensive care needed to manage complex
health states and where there may already be difficulties dealing with individual conditions,
especially among older people. The World Health Organization in its report on Ageing and Health17
recently recognised that meeting the needs of ageing populations would require significant
changes in the way health systems are structured and health care is delivered. Many existing
services were designed to cure acute conditions or symptoms, often resulting in disconnected and
fragmented disease management as well as a lack of coordination across care providers, settings
and time.
Moving away from disease-based curative models towards the provision of older-person-centred
and integrated care is one way health systems could be transformed17, which would likely benefit
older populations in SSA. The development of generalizable models of community healthcare for
frail and dependent older people, who are likely to develop dementia, is therefore fundamental.
The WHO has initiated the development of evidence-based guidelines on integrated care for older
people (ICOPE) with a focus on less resourced settings. Targeted at non-specialist health workers,
they will guidehome-based interventions for older people that can prevent, reverse or slow
declines in intrinsic capacity. Those guidelines, to be published by the WHO by the end of 2017,
will cover issues such as malnutrition, mobility loss, urinary incontinence, falls, hearing and visual
impairments, depression, and cognitive, behavioural and psychological problems6.
5.3 Concluding remarks and call to action
This report tries as far as possible to give a faithful snapshot of the situation today. The lack of
available data continues to make it difficult to create a baseline which future publications of this
type can use to ascertain whether there have been changes and, more crucially, any progress.
Lack of data will not deter ADI and its focused, energetic and committed members from working
to find solutions to a very real and tangible problem. But the solutions for Africa can only come
from Africa. Alzheimer associations in Ghana, Kenya, Namibia, Nigeria, South Africa, Zambia and
Zimbabwe – our members at the date of publication – cannot be alone in their fight, we need more
to join them. They are forging ahead, creating hope and activity were there was little, mobilizing
goodwill and resources to find solutions, and trying to sensitise their governments.
Governments in the region have a real opportunity right now – thanks to the excellence of these
grassroots organisations – to join forces with them and by adopting and implementing the seven
key areas of the WHO Global Action Plan and the recommendations in this report, show the rest of
Africa the way. At the same time, we need the African Union to listen to the important truths about
the scale and cost of dementia today and in the future. Now is the time to protect those in society
who are most vulnerable.

69DEMENTIA IN SUB-SAHARAN AFRICA
References
1. Prince, M., et al., World Alzheimer Repor t 2015 - The
Global Impact of dementia: an analysis of prevalence,
incidence, costs and trends. 2015: London.
2. World Health Organization, Draft global action plan on
the public health re sponse to dementia. 2016, World
Health Organization: Geneva.
3. Prince, M., et al., World Alzheimer Repor t 2016 -
Improving healthcare for people living with dementia.
Coverage, quality, costs now and in the future. 2016,
Alzheimer’s Disease International: London. p. 131.
4. World Health Organization, mhGAP Intervention Guide
- Version 2.0 for mental, neurological and substance
use disorders in non-specialized health settings. 2016,
World Health Organization: Geneva.
5. World Health Organization, Dementia: a public health
priority. 2012, World Health Organization: Geneva. p.
112.
6. World Health Organization. WHO Integrated Care for
Older People (ICOPE) 10/08/2017]; Available from:
http://www.who.int/ageing/health-systems/icope/en/
7. Prince, M., et al., World Alzheimer Repor t 2014 -
Dementia Risk Reduction: An analysis of prote ctive
and modifiable risk factors. 2014: London.
8. Vartiainen, E., et al., Thirt y-five-year trends in
cardiovascular risk factors in Finland. Int J Epidemiol,
2010. 39(2): p. 504-18.
9. Beaglehole, R., et al., Improving the prevention and
management of chronic disease in low-income and
middle-income countries: a priority for primary health
care. Lancet, 2008. 372(9642): p. 940-9.
10. Barnes, D.E. and K. Yaffe, The projected effect of risk
factor reduction on Alzheimer’s disease prevalence.
Lancet Neurol, 2011. 10(9): p. 819-28.
11. Livingston, G., et al., Dementia prevention,
interve ntion, and care. Lancet, 2017.
12. Norton, S., et al., Potentia l for primary prevention of
Alzheimer’s disease: an analysis of population-based
data. Lancet Neurol, 2014. 13(8): p. 788-94.
13. Gao, S., et al., Dementia incidence declined in African-
Americans but not in Yoruba. Alzheimers Dement,
2016. 12(3): p. 244-51.
14. Matthews, F.E., et al., A two decade dementia
incidence comparison from the Cognitive Function
and Ageing Studies I and II. Nat Commun, 2016. 7: p.
113 98.
15. Qiu, C., et al., Twenty-year change s in dementia
occurrence suggest decreasing incidence in central
Stockholm, Sweden. Neurology, 2013. 80(20): p. 1888-
94.
16. Satizabal, C.L., et al., Incidence of Dementia over
Three De cades in the Framingham Hea rt Study. N Engl
J Med, 2016. 374(6): p. 523-32.
17. World Health Organization, World report on ageing and
health. 2015, World Health Organization: Geneva.
Appendix A: Global Burden of Disease (GBD) Regions
Table A.1
GBD regions and countries
GBD Region Countries Relationship to WHO
regions used for ADI/
Lancet estimates
Sub-Saharan
Africa, Central
Angola, Central African Republic, Congo, Democratic
Republic of the Congo, Equatorial Guinea, Gabon
A mixture of AFRO D and AFRO E
Sub-Saharan
Africa, East
Burundi, Comoros, Djibouti, Eritrea, Ethiopia, Kenya,
Madagascar, Malawi, Mauritius, Mozambique, Rwanda,
Seychelles, Somalia, South Sudan, Sudan, Uganda,
United Republic of Tanzania, Zambia
AFRO E except for Comoros (A FRO D)
and Somalia and Sudan (EMRO D)
Sub-Saharan
Africa, Southern
Botswana, Lesotho, Namibia, South Africa, Swaziland,
Zimbabwe
AFRO E
Sub-Saharan
Africa, West
Benin, Burkina Faso, Cameroon, Cape Verde, Chad,
Cote d’Ivoire, Gambia, Ghana, Guinea, Guinea-Bissau,
Liberia, Mali, Mauritania, Niger, Nigeria, Sao Tome and
Principe, Senegal, Sierra Leone, Togo
AFRO D

ALZHEIMER’S DISEASE INTERNATIONAL70
Personal story
Zimbabwe – Mrs Z
Mrs Z is 94 years old, and lives on the farm where she and her
husband raised their eight children. After retirement, they both
stayed at the farm, and Mrs Z remained active, as Chairwoman
at the local church, and as both Secretary and Treasurer of the
Women Fellowship. Her grandchildren came to the farm, and enjoyed spending
time with them.
Her husband and four of her children sadly passed away. After the death of her
youngest daughter in 2013, she began to wake up during the night and start
praying continuously, mentioning the name of her dead daughter. Her children
attributed this to the mourning process. By 2014, they started to notice changes
in her behaviour: she would forget things and wanted to stay at her maternal farm
although her in-laws were no longer there. Mrs Z started leaving the family home
to go to neighbouring houses without informing anyone where she was going.
Her children became worried for her safety as she was travelling 25km to reach
the Women Fellowship, and the farm is surrounded by tall grass, which is home
to pythons. She was increasingly spending time alone in her room, and packing
clothes to visit relatives who had passed away.
In December 2016, Mrs Z was admitted to hospital and was diagnosed with
congested cardiac failure. When her daughter described to the physician the
changes in Mrs Z’s behaviour, he initially said it was just old age. However, after
speaking to her further, he explained that she was going through the early stages
of dementia. A month later, Mrs Z started becoming more confused and had
trouble sleeping.
At this point the family got in contact with Zimbabwe Alzheimer’s and Related
Disorders Association (ZARDA), who have been a great help to Mrs Z and her
family. They have done much to raise awareness and train carers in communities
across Zimbabwe – a country in which there is still a pervasive stigma
surrounding dementia.
Her daughter continues to care for her, and has support from a maid. She
ensures her mother’s safety by locking the gates, but she freely walks around
the garden, and enjoys getting some fresh air and exercise. Her daughter does
some basic occupational therapy by giving her tasks to exercise her brain. The
local community, particularly members of the church, neighbours, and extended
family, remain supportive of Mrs Z, and often visit her. Having seen ZARDA’s work
and interventions, Mrs Z’s daughter says: “It is my hope and wish to research this
condition in different cultures and tribes. A big thank you to ZARDA”.

About ADI
Alzheimer’s Disease International (ADI) is the international federation
of Alzheimer associations throughout the world. Each of our 90
members is a non-profit Alzheimer association supporting people
with dementia and their families. ADI’s mission is to strengthen and
support Alzheimer associations, to raise awareness about dementia
worldwide, to make dementia a global health priority, to empower
people with dementia and their care partners, and to increase
investment in dementia research.
What we do
• Support the development and activities of our member associations
around the world.
• Encourage the creation of new Alzheimer associations in countries where
there is no organisation.
• Bring Alzheimer organisations together to share and learn from each other.
• Raise public and political awareness of dementia.
• Stimulate research into the prevalence and impact of Alzheimer’s disease
and dementia around the world.
• Represent people with dementia and families on international platforms at
the UN and WHO.
Key activities
• Raising global awareness through World Alzheimer’s Month™ (September
every year).
• Providing Alzheimer associations with training in running a non-profit
organisation through our Alzheimer University programme.
• Hosting an international conference where staff and volunteers from
Alzheimer associations meet each other as well as medical and care
professionals, researchers, people with dementia and their carers.
• Disseminating reliable and accurate information through our website and
publications.
• Supporting the 10/66 Dementia Research Group’s work on the prevalence
and impact of dementia in developing countries.
• Supporting global advocacy by providing facts and figures about
dementia, and monitoring as well as influencing dementia policies.
ADI is based in London and is registered as a non-profit
organisation in the USA. ADI was founded in 1984, has been in
official relations with the World Health Organization since 1996
and has had consultative status with the United Nations since
2012. ADI is partnered with Dementia Alliance International (DAI),
a collaboration of individuals diagnosed with dementia providing
a unified voice of strength, advocacy and support in the fight for
individual autonomy for people with dementia.
You can find out more about ADI at www.alz.co.uk/adi

Alzheimer’s Disease International:
The International Federation
of Alzheimer’s Disease and
Related Disorders Societies, Inc.
is incorporated in Illinois, USA,
and is a 501(c)(3) not-for-profit
organization
Alzheimer’s Disease International
64 Great Suffolk Street
London SE1 0BL
UK
Tel: +44 20 79810880
www.alz.co.uk