Content uploaded by Laura Katus
Author content
All content in this area was uploaded by Laura Katus on Oct 20, 2023
Content may be subject to copyright.
STUDY PROTOCOL
The Brain Imaging for Global Health (BRIGHT) Project:
Longitudinal cohort study protocol [version 1; peer review:
awaiting peer review]
Sarah Lloyd-Fox 1,2, Sam McCann3, Bosiljka Milosavljevic 1,2, Laura Katus4,
Anna Blasi 5, Chiara Bulgarelli2,5, Maria Crespo-Llado6, Giulia Ghillia 3,
Tijan Fadera7, Ebrima Mbye7, Luke Mason3, Fabakary Njai7, Omar Njie7,
Marta Perapoch-Amado 8, Maria Rozhko1, Fatima Sosseh7, Mariama Saidykhan7,
Ebou Touray7, Sophie Moore3,7, Clare Elwell5, The BRIGHT Project team
1Psychology, University of Cambridge, Cambridge, England, UK
2Psychological Sciences, Birkbeck University of London, London, England, UK
3Women's and Children's Health, Kings College London, London, UK
4School of Human Sciences, University of Greenwich, London, England, UK
5Medical Physics and Biomedical Engineering, University College London, London, England, UK
6Institute of Lifecourse and Medical Sciences, University of Liverpool, Liverpool, England, UK
7The Medical Research Council Unit The Gambia at the London School of Hygiene and Tropical Medicine, Banjul, The Gambia
8Psychology, University of East London, London, England, UK
First published: 18 Oct 2023, 7:126
https://doi.org/10.12688/gatesopenres.14795.1
Latest published: 18 Oct 2023, 7:126
https://doi.org/10.12688/gatesopenres.14795.1
v1
Abstract
There is a scarcity of prospective longitudinal research targeted at
early postnatal life which maps developmental pathways of early-
stage processing and brain specialisation in the context of early
adversity. Follow up from infancy into the one-five year age range is
key, as it constitutes a critical gap between infant and early childhood
studies. Availability of portable neuroimaging (functional near infrared
spectroscopy (fNIRS) and electroencephalography (EEG)) has enabled
access to rural settings increasing the diversity of our sampling and
broadening developmental research to include previously
underrepresented ethnic-racial and geographical groups in low- and
middle- income countries (LMICs). The primary objective of the Brain
Imaging for Global Health (BRIGHT) project was to establish brain
function - using longitudinal data from mother - for-age reference
curves infant dyads living in the UK and rural Gambia and investigate
the association between context-associated moderators and
developmental trajectories across the first two years of life in The
Gambia. In total, 265 participating families were seen during
Open Peer Review
Approval Status AWAITING PEER REVIEW
Any reports and responses or comments on the
article can be found at the end of the article.
Gates Open Research
Page 1 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
pregnancy, at 7-14 days, 1-, 5-, 8-, 12-, 18- and 24-months post-
partum. An additional visit is now underway at 3–5 years to assess
pre-school outcomes. The majority of our Gambian cohort live in
poverty, but while resource-poor in many factors they commonly
experience a rich and beneficial family and caregiving context with
multigenerational care and a close-knit supportive community.
Understanding the impact of different factors at play in such an
environment (i.e., detrimental undernutrition versus beneficial
multigenerational family support) will (i) improve the
representativeness of models of general cognitive developmental
pathways from birth, (ii) identify causal pathways of altered
trajectories associated with early adversity at both individual and
group level, and (iii) identify the context-associated moderators (i.e.
social context) that protect development despite the presence of
poverty-associated challenges. This will in turn contribute to the
development of targeted interventions.
Keywords
Gambia, UK, infancy, development, undernutrition, longitudinal,
neuroimaging, global health, fNIRS, EEG
Gates Open Research
Page 2 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Corresponding author: Sarah Lloyd-Fox (sl868@cam.ac.uk)
Author roles: Lloyd-Fox S: Conceptualization, Formal Analysis, Funding Acquisition, Investigation, Methodology, Project Administration,
Resources, Supervision, Validation, Visualization, Writing – Original Draft Preparation, Writing – Review & Editing; McCann S: Data
Curation, Formal Analysis, Investigation, Project Administration, Software, Supervision, Visualization, Writing – Original Draft
Preparation, Writing – Review & Editing; Milosavljevic B: Conceptualization, Data Curation, Formal Analysis, Funding Acquisition,
Investigation, Methodology, Project Administration, Software, Supervision, Validation, Visualization, Writing – Original Draft Preparation,
Writing – Review & Editing; Katus L: Data Curation, Formal Analysis, Investigation, Methodology, Software, Supervision, Visualization,
Writing – Original Draft Preparation, Writing – Review & Editing; Blasi A: Data Curation, Formal Analysis, Investigation, Methodology,
Project Administration, Software, Supervision, Validation, Writing – Review & Editing; Bulgarelli C: Data Curation, Formal Analysis,
Software, Supervision, Validation, Writing – Review & Editing; Crespo-Llado M: Data Curation, Formal Analysis, Investigation,
Methodology, Project Administration, Supervision, Validation, Writing – Review & Editing; Ghillia G: Data Curation, Formal Analysis,
Investigation, Writing – Review & Editing; Fadera T: Investigation, Writing – Review & Editing; Mbye E: Investigation, Project
Administration, Supervision, Writing – Review & Editing; Mason L: Data Curation, Formal Analysis, Methodology, Software, Supervision;
Njai F: Investigation, Writing – Review & Editing; Njie O: Investigation, Software; Perapoch-Amado M: Data Curation, Investigation,
Project Administration, Writing – Review & Editing; Rozhko M: Data Curation, Investigation, Methodology, Project Administration,
Writing – Review & Editing; Sosseh F: Investigation; Saidykhan M: Formal Analysis, Investigation, Writing – Review & Editing; Touray E:
Data Curation, Investigation, Project Administration, Supervision, Writing – Review & Editing; Moore S: Conceptualization, Funding
Acquisition, Methodology, Project Administration, Resources, Supervision, Writing – Review & Editing; Elwell C: Conceptualization,
Funding Acquisition, Methodology, Project Administration, Resources, Supervision, Writing – Review & Editing;
Competing interests: No competing interests were disclosed.
Grant information: The BRIGHT Study is funded by the Bill and Melinda Gates Foundation (OPP1127625 and core funding MC‐A760‐
5QX00 to the International Nutrition Group by the Medical Research Council UK and the UK Department for International Development
(DfID) under the MRC/DfID COncordant agreement. S.L.F is supported by a UKRI Future Leaders Fellowship (MR/S018425/1). S.E.M is
supported by a Wellcome Trust Senior Research Fellowship (220225/Z/20/Z). L.K. was further supported by an ESRC Postdoctoral
Fellowship, grant number ES/T008644/1 and a BBSRC grant, G114265. B.M. is supported by an Economic and Social Research Council
(ESRC) Secondary Data Analysis Initiative Grant G110102 (ES/V016601/1). C.B. was further supported by a Leverhulme Trust Early Career
Fellowship (ECF-2021-174).
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Copyright: © 2023 Lloyd-Fox S et al. This is an open access article distributed under the terms of the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
How to cite this article: Lloyd-Fox S, McCann S, Milosavljevic B et al. The Brain Imaging for Global Health (BRIGHT) Project:
Longitudinal cohort study protocol [version 1; peer review: awaiting peer review] Gates Open Research 2023, 7:126
https://doi.org/10.12688/gatesopenres.14795.1
First published: 18 Oct 2023, 7:126 https://doi.org/10.12688/gatesopenres.14795.1
Gates Open Research
Page 3 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Introduction
Background
The first 1000 days of life, which describes the developmental
period between conception and two years of age, is character-
ised by prodigious physiological, psychological and physical
change. As such, this period represents a critical window for
brain development, during which plasticity to environmen-
tal factors is greatest. According to UNICEF, 19.5% of the
world’s children live in poverty, the majority of whom reside
within sub-Saharan Africa (51.7%) and South Asia (35.7%).
Infants and children growing up in poverty may be exposed to a
range of biological and/or psychosocial risk factors both pre-
and postnatally. Such risk factors include lower parental income
and educational level, parental mental health issues, reduced
access to recreational and educational activities (particularly
in rural communities), undernutrition, food insecurity, envi-
ronmental hazards and poor sanitation (Giovanelli et al., 2016;
Jensen et al., 2017; Smith et al., 2015; Worku et al., 2018).
While resource poor in many factors, there can also be many
beneficial context-associated moderators, for example some
communities who grow up in poverty also experience a rich
and beneficial family and caregiving context with multigen-
erational social support and a close-knit supportive community.
The impact of environment on neurocognitive development is
therefore dynamic and multi-faceted, affecting biological, social,
and behavioural developmental processes. For example, under-
nourished infants may seek, and consequently receive, less
stimulation from caregivers. This lack of social stimulation is,
in turn, linked to changes in brain function which are likely to
precede changes in behaviour (East et al., 2017). Despite these
multifaceted links between different factors, evidence examin-
ing the impact of poverty-related risk and developmental out-
comes oftentimes focusses on only narrow subsets of relevant
factors, which leaves open questions regarding the interplay
across domains. As noted by Nobel Laureate Esther Duflo
in 2019, ‘Our goal is to make sure that the fight against
poverty is based on scientific evidence. It starts from the idea
that often the poor are reduced to caricatures and often even
people who try to help them do not actually understand what are
the deep roots of the problems’ (Cho, 2019). Here, we describe
the rationale behind and cohort characteristics of the Brain
Imaging for Global Health (BRIGHT) project, which works
with family cohorts in the UK and The Gambia. We first present
relevant literature that informed the design of the BRIGHT
project, before providing details on the study protocol and
characteristics of the two cohorts.
Childhood poverty has been associated with lower perform-
ance on language, memory and cognitive control tasks (Farah
et al., 2006). At a global level, this is reflected by one third
of pre-school-aged children in low-and-middle-income coun-
tries (LMICs) failing to reach age-appropriate milestones in
cognitive and/or socio-emotional development (McCoy et al.,
2016). Within the first years of life, one major poverty-
associated risk factor that poses a considerable risk to early
child development is stunting (low length/height for age
against an international reference), which occurs as the result
of chronic growth failure and affects one in five children under
five years of age globally (Development Initiatives, 2018).
Furthermore, the interplay between the impact of undernutri-
tion and compensatory factors (i.e., positive parenting prac-
tices), thought to scaffold early child development, is complex.
A recent study in rural Cambodia looking at the joint role of
parenting and nutritional status – in relation to inequities in
family wealth – found that, while more stimulating and sup-
portive parenting practices were associated with improved
developmental outcomes in three to five year olds, this was
strongest for non-stunted children (Berkes et al., 2019). We
therefore urgently need to further our understanding of brain and
cognitive development during early childhood in the context of
poverty-associated risk factors. This is especially relevant as
compromised development of a core set of age-appropriate
skills in childhood has a significant impact on subsequent
academic achievement, mental health and economic status - and
consequently the potential to lead full and productive lives and
support future generations (Alderman et al., 2014; Hackman
& Farah, 2009; Martorell et al., 2010; Victora et al., 2008). The
United Nations Sustainable Development Goals have conse-
quently identified the reduction of poor cognitive development
during childhood in LMICs as a key priority for global health
research and interventions (UN, 2015). All LMICs fall within
what are also known as Majority World countries, where 85%
of the world’s population live (Alam, 2008). However, over the
past 15-30 years, only 3 - 17% of published child development
journal articles (Moriguchi, 2022; Nielsen et al., 2017) are from
Majority World countries and only 5% of child development
interventions (Draper et al., 2023), meaning that the world’s
child population are under-represented in our theoretically
driven understanding of development (Draper et al., 2022). In
the following section we review some of the existing litera-
ture on links between poverty and brain development as well as
describe how a small number of studies are beginning to address
the under-representation of child development research in LMIC
(Majority World) contexts.
Over the last decade several large-scale studies from the United
States have shown links between poverty and brain develop-
ment in childhood (Barch et al., 2016; Hair et al., 2015; Luby
et al., 2013; Noble et al., 2015); for example, children of
parents with high school education had roughly 3% less
cortical volume than those with university level education, and
those with parental incomes below $25,000 had 6% less than
those making over $150,000 (Noble et al., 2015). These find-
ings have led to models relating specific components of
socioeconomic status (SES), stress and brain structural and
functional development (Farah, 2017; Hackman & Farah,
2009; Noble et al., 2012). However, the majority of these
studies rely heavily on correlational analyses in later child-
hood rather than studying development at an early age to fully
understand the mechanistic processes driving these differ-
ences (for a review of studies that have looked at general
differences in brain volume and poverty in infancy see Hurt &
Betancourt, 2016). While the development of new methodol-
ogy has increased our understanding of brain and cognitive
development in infants and young children over the last decade,
this research has been largely restricted to financially mobile
Page 4 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
participants within high-income countries (Henrich et al., 2010).
Studies that examine the impact of extreme poverty (defined
by the World Bank as living on less than $1.9 per household
member per day) on brain development are extremely scarce.
Furthermore, it remains unclear whether poverty-associated
risk factors influence brain development more severely
during sensitive periods of development (i.e. fetal life/early
infancy/early childhood), or whether their impact builds over
time depending on the chronicity, pervasiveness, severity and/
number of incidences (Berens et al., 2017; Jensen et al., 2017).
Furthering our understanding of the developmental impact of
early environmental adversity in both the short- and long-term
is thus of high priority, particularly during the understudied
period of the first 1000 days of life.
A further consideration is that the majority of child develop-
ment research conducted in LMICs has been limited to the
use of behavioural assessments of cognitive development to
measure the effect of exposure to early adversity. Such meas-
ures are often undertaken later in childhood rather than at the
time that vulnerability to exposure is most critical i.e. during
prenatal and early postnatal life (Sabanathan et al., 2015). Fur-
thermore, it is important to note that developmental trajectories
of perceptual, motor and language domains have different
timescales and cascading effects on one another. Therefore,
they may have different key periods of sensitivity to insults
(for an example of a key sensitive period for visual cortex see
Hensch, 2005). Thus, it is imperative that neurocognitive
development is studied from as close to birth as possible –
ideally with a longitudinal framework to track age-related
changes – taking contemporaneous measurements of brain
function and behaviour in parallel with measurements of expo-
sure to environmental challenges. Furthermore, such research
could offer new pathways for the provision of widely appli-
cable, objective paradigms and methods that can assess early
brain development in hard-to-reach populations (Isaacs,
2013). This would be synergistic with current large scale glo-
bal health initiatives to optimise behavioural measures of early
cognitive development (Murray-Kolb et al., 2014; Richter et al.,
2019) such as the Global Scales of Early Development (GSED)
(Cavallera et al., 2023). Furthermore the introduction of objec-
tive brain imaging paradigms to global health research could
address some of the current challenges associated with behav-
ioural measures of child development (Isaacs et al., 2008;
Perkins et al., 2017). The optimisation of tools for measuring
neurocognitive development will in turn support the develop-
ment of early intervention strategies from the first days and
months of life offering the potential for large lifetime cost
savings (i.e. “1001 Critical Days” cross party manifesto, UK; Sure
Start (Cattan et al., 2021)).
Over the last five to ten years, several research collaborations
have been established to bring new neurocognitive technol-
ogy (i.e. eye-tracking and neuroimaging tools) to the field of
infant and early child brain and cognitive development research
in LMICs: these include within The Gambia (Begus et al.,
2016; Katus et al., 2019; Katus et al., 2020; Katus et al., 2023;
Katus et al., 2022b; Lloyd-Fox et al., 2014; Lloyd-Fox
et al., 2017; Lloyd-Fox et al., 2019); Cote D’Ivoire
(Jasinska & Guei, 2018); Guinea Bissau (Roberts et al.,
2017); Malawi (Forssman et al., 2017; Pyykkö et al., 2019;
Pyykkö et al., 2020); Bangladesh (Jensen et al., 2019; Perdue
et al., 2019; Turesky et al., 2019; Xie et al., 2018); India
(Wijeakumar et al., 2019); South Africa (Wedderburn
et al., 2020) and Brazil (Alarcão et al., 2021). These
global infant and child brain development studies cover age
ranges from 0 – 65 months of age, reporting on areas as wide
ranging as social information processing (Lloyd-Fox et al.,
2017; Perdue et al., 2019; Xie et al., 2019), early brain ana-
tomical and connectivity development (Collins-Jones et al.,
2021; Fishell et al., 2020; Turesky et al., 2020; Turesky et al.,
2019), the development of brain networks associated with
visual working memory (Wijeakumar et al., 2019) and the
development of attentional/neural markers of habituation and
novelty detection (Katus et al., 2020; Katus et al., 2023; Katus
et al., 2022b; Lloyd-Fox et al., 2019). Finally, a recent step
change in research has been to begin to use measures of brain
and cognitive development to understand the impact of interven-
tions within global health studies; as evidenced by the recent
work by Alarcão and colleagues in Brazil to measure the efficacy
of a home-visiting program for adolescent mothers for enhanc-
ing early infant brain development and behaviour (Alarcão
et al., 2021).
In this paper we describe the BRIGHT project, a follow-on from
a pilot study, which ran from 2012-2014, in which we demon-
strated the feasibility of combined neuroimaging, behavioural
assessments and growth measures in longitudinal and cross-
sectional studies from birth to 24 months of age in rural Gambia
(Begus et al., 2016; Lloyd-Fox et al., 2017; Lloyd-Fox et al.,
2014; Papademetriou et al., 2013). Importantly we showed that
fNIRS, can be easily implemented in rural contexts such as in
The Gambia and used from the first weeks of life to provide
quantitative and objective markers of neurocognitive func-
tion. We identified testing paradigms that elicit reliable brain
responses that can be used to chart development as a func-
tion of age, and which aligned with findings in age-matched
groups of infants from studies conducted in the UK. As part
of our pilot study, we also successfully performed a qual-
ity control assessment of the adaptation and administration of
a behavioural assessment, the Mullen Scales of Early Learning
(MSEL), for use in rural Gambia (Milosavljevic et al., 2019).
The BRIGHT project has been established to extend this pilot
phase to a larger longitudinal observation cohort study of infant
and early child development from birth to two years of age, with
an additional follow-up at pre-school age (3-5 years).
The BRIGHT project design overview
The BRIGHT project (Phase I), which ran from 2015 – 2020,
established two prospective cohorts of families in the UK and
The Gambia using a longitudinal multi-methods approach.
Families were recruited during pregnancy and, following
delivery, longitudinal measures of infant brain and cognitive
development were conducted from 0-24 months of age across
10 data collection phases: antenatal – recruitment and 32-36
weeks’ gestation, and postnatal – 1-3, 7-14 days, 1, 5, 8, 12, 18
and 24 months of age. Additional data on diet and health were
collected continuously across this time period. The project
Page 5 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
implemented brain imaging measures (fNIRS and electroen-
cephalography [EEG]), neurocognitive behavioural measures
(utilising eye-tracking methods), population-specific cogni-
tive developmental measures (MSEL and the Communicative
Development Inventory [CDI]), family-caregiving assessments
(Family Care Indicators [FCI]), caregiver-infant interaction
videos and questionnaires) and home environment measures
(Language Environment Analysis [LENA]) alongside regular
collection of biological, socioeconomic, parental health and
nutritional measures at both sites. This data collection frame-
work was implemented to allow the modelling of longitudinal
changes in brain function, cognitive development, and growth
within the rural Gambian population. Further, the collection
of parallel behavioural and environmental data was designed
to enable the identification of critical developmental
moderators, mediators, and markers of risk and resilience.
The purpose of the BRIGHT project is to firstly establish
longitudinal trajectories across populations, and secondly to
provide a framework for in-depth investigations of inter-
individual differences within the Gambian cohort. Given that
neuroimaging data provided the backbone of this project, it
was essential that a UK cohort was also established to meas-
ure the longitudinal developmental trajectories of the different
fNIRS and EEG paradigms across different populations as
several of these had not been studied across this longitudinal
time span before in any population internationally. This was
chosen to broadly match the context of previously acquired
developmental neuroimaging data, given that to date the vast
majority of research of this kind has been undertaken in high
income countries (HICs).
The target cohort sizes (The Gambia n=200, UK n=60,)
were based on previous infant fNIRS and EEG studies con-
ducted in the UK, which indicated that sample sizes from 20
(moderate effect size) to 42 (small effect size) were sufficient
to determine regions of significant cortical brain activation in
response to stimuli. The Gambian cohort was designed to be
larger to allow within-cohort sub-group comparisons and
individual differences analyses; for example, grouping by
growth trajectories (mild, medium and severe markers of
undernutrition) on the assumption that approximately 25–30%
of the cohort would be stunted (z-score of length-for-age
< 2 standard deviations below the WHO reference) by two years
of age (Nabwera et al., 2017).
BRIGHT Kids (BRIGHT project phase II)
Previous research highlights a marked impact of exposure to
early adversity and neurobehavioral outcomes at preschool age.
With this in mind, in 2023 we conducted a follow-up assess-
ment at preschool age in the Gambian cohort of the BRIGHT
project at 3–5 years. This cross-sectional follow up will allow
us to examine additional questions, regarding the long-term
stability of our infant neural markers to predict long-term
outcomes.
Objectives
The primary objectives of the BRIGHT project are to:
(1) develop brain and neurocognitive function-for-age
curves from birth to 24 months of age using prospective
longitudinal datasets from the UK and The Gambia.
These reference curves will be used to enable
age-adjusted group comparisons of differences in
average trajectories, group-wise differences in vari-
ability, and for characterizing the range of individual
developmental trajectories within each cohort.
(2) establish the association between context-associated
moderators, including poverty-associated risk factors
(i.e., undernutrition and consequent growth faltering),
and developmental trajectories across the first two
years of life in The Gambia.
(3) establish the association between context-associated
moderators and developmental trajectories across the
first two years of life in The Gambia and pre-school
outcomes at three to five years.
Secondary aims of interest are to:
1. Assess whether infants with similar trajectories of
growth have the potential to reach the same developmental
milestones within the first 24 months of life.
2. Establish whether neuroimaging markers of brain func-
tion are more robust indicators of development within
individual infants across age, as compared to behavioural
measures.
3. Assess the capability of fNIRS, EEG and eye-
tracking methods to deliver specific and early biomarkers
of altered developmental pathways.
Here we describe the formation of a common BRIGHT study
protocol across the two sites (The Gambia and the UK) and,
where appropriate, site-specific additional measures, par-
ticularly those focused on family context, nutrition, diet, and
biological samples are presented for The Gambia. We briefly
outline how we recruited participants at each site; selected and
implemented experimental neurocognitive and behavioural
measures at each site; how we standardised lab practices across
sites to ensure comparability; and how we developed analyti-
cal pipelines for the different datasets. We also describe the
demographic and socioeconomic distribution of our cohorts.
Methods
The Gambian site and population
The Gambia is situated on the West coast of Africa, bordering
Senegal. The majority (60%) of the roughly 2.4 million
inhabitants of The Gambia live in the coastal regions
surrounding the capital, Banjul, while the remainder of the
population live rurally, often supporting themselves through
subsistence farming (Hennig et al., 2017) living in extended,
multi-generational households (Brotherton et al., 2021; Kea,
2013; Sear & Mace, 2009). The Gambia is one of the lowest
ranking countries with regard to gross national income, years
of schooling, and life expectancy, with over half of adults
never having received formal education (Hennig et al.,
2017). School attendance has risen rapidly over the last
decades thanks to the introduction of free universal education,
Page 6 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
and 97% of children now attend to primary level relative to
66.7% in the early 1970s (The Gambia Government National
Education Statistics, 2018; CEICdata.com). Preschool education
has also become increasingly available across the timeframe of
the BRIGHT project (Blimpo et al., 2019). Childcare is viewed
as a shared responsibility among family members, with grand-
mothers and older sisters having the biggest role in supporting
parents (Brotherton et al., 2021; Sear & Mace, 2009). Islam is the
predominant religion and raising children in accordance
with religious and community values is of high importance
(Sosseh et al., 2023). Marriages are commonly polygamous
with over half of married women living with one or two
co-wives (Hennig et al., 2017), though gendered hierarchies and
intra-household relations are dynamic and subject to change
(Kea, 2013; Kea, 2020). Furthermore, over the past decades infant
and child mortality has decreased, birth spacing has increased,
and overall family size has reduced (Nabwera et al., 2017).
The Gambian arm of the BRIGHT project was hosted at a rural
site of the Medical Research Council The Gambia Unit at the
London School of Hygiene and Tropical Medicine (MRCG@
LSHTM; www.mrc.gm). The UK Medical Research Coun-
cil (MRC) has a long-standing research partnership with The
Gambia, established in the late 1940’s. Currently, research
conducted within MRCG@LSHTM is focused on three
broad themes centred around major public health priorities,
specifically Vaccines and Immunity, Disease Control and
Elimination, and Nutrition and Planetary Health; the latter of
which the BRIGHT project is situated within.
The BRIGHT project was undertaken at the Keneba Field Sta-
tion of MRCG@LSHTM, situated in the rural West Kiang
region, 145 km inland from the capital. Seasonality has an
impact on nutrient availability for the population living here
as weather patterns alternate between four months of heavy
rainfall (July-October) and eight months of extreme dryness -
directly affecting the availability of key nutrients (Moore et al.,
1997). In 2015 at the onset of the project, the Keneba Field Sta-
tion was relatively isolated, accessed via unmade roads and
required to independently maintain all facilities necessary for
research and clinical care (e.g., generator powered electricity,
bore hole water supply, satellite communication). However,
over the course of the study, the country and local region
have been witnesses to several changes. At the local level
several infrastructure improvements have been made, includ-
ing the road being tarmacked, and therefore allowing greater
access to urban resources, and the field station and local
community now have nationally sourced electrical power.
BRIGHT project participants were drawn from Keneba and
surrounding villages within a 20km radius of the field station.
All women of reproductive age (18–45 years) who were reported
to be pregnant within the West Kiang Demographic Surveillance
System between June 2016 and March 2018, spoke Mandinka
as their primary language, and were expected to reside in
West Kiang for the duration of the project were invited to
participate (see Figure 1 for recruitment pathway in The
Gambia). Further eligibility criteria for study participation
pertaining to the pregnant women included: carrying a singleton
pregnancy, < 36 weeks’ gestation on presentation to the first
antenatal study visit and being medically fit to participate, as
determined by the study midwife. The project was designed
to recruit participants so that deliveries were spread evenly
throughout the recruitment period, aiming for around 10–15
deliveries per month. This was to ensure that workload was
achievable and as consistent as possible, ensuring timely
scheduling of follow-up visits. For this reason, an additional
exclusion criterion of ‘gestational age incompatible with study
requirements’ was introduced. From an ultrasound scan at the
first antenatal study visit, gestational age was measured, and
expected delivery date calculated. If a participant was due to
deliver in a month that was already at full capacity, they were
excluded at this point. Postnatally, mother-infant dyads were
excluded from the project if the infant was diagnosed with a
developmental disability e.g., Down’s Syndrome or cerebral
palsy. Participants were free to withdraw from participation
at any point in the study.
The UK site and population
In the UK, participants were recruited from the city of
Cambridge and surrounding villages. Demographically,
the population in Cambridgeshire is representative of that
across the UK with regard to ethnicity, employment rates and
family structure (Cambridge County Council, 2011). The area
however differs from the rest of the UK with regard to levels of
education within the population, with twice as many inhabit-
ants holding a higher education degree (Cambridge County
Council, 2011). The research involved in the UK arm of the
BRIGHT project was conducted at dedicated facilities either
within the Evelyne Perinatal Imaging Unit at the Rosie Hospital,
Cambridge University Hospitals NHS Foundation Trust or
within the Department of Psychology, University of Cambridge.
Once per week during the recruitment phase, families who
attended an antenatal clinic at the Rosie Maternity Unit at
Cambridge University Hospitals between June 2016 and
January 2017, with a healthy singleton pregnancy less than 36
weeks gestational age, were approached and given information
about the project (see Figure 2 for UK recruitment pathway).
Ethical considerations
Protocols were approved by the relevant committee at each
site. In The Gambia, ethical approval was given by the joint
Gambia Government - MRC Ethics Committee (SCC 1351)
and the Scientific Coordinating Committee at the MRC Unit
The Gambia. Additional approval was granted for the BRIGHT
Kids follow up (Project reference 22737). Informed consent
was obtained in writing, or via thumbprint if individuals were
unable to write, from all parents/carers prior to participation.
In the UK, the study was approved by the National Research
Ethics Service East of England Committee, NHS Health
Research Authority (REC reference 13/EE/0200), and informed
written consent was obtained from parents of infants to par-
ticipate. The project is guided by a consistent set of principles
which ensured that the infants’ and child’s wellbeing is always
prioritised. Infants/children are always with their caregiver.
The protocols were designed to be engaging and interesting to
the infants and children, and the setup comfortable. Caregivers
were made aware that the study can be interrupted, rescheduled,
Page 7 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Figure 1. Summary of study visits: The Gambia. EPDS, Edinburgh Postnatal Depression Scale; PRAS, Pregnancy-related Anxiety Scale;
PSS, Perceived Stress Scale; PANAS, Positive and Negative Aect Scale; NBAS, Neonatal Behavioural Assessment Scale; SES, Socioeconomic
Status; MIPH, Maternal and Infant Physical Health; fNIRS, functional near infrared spectroscopy; EEG/ERP, electroencephalography/event
related potentials; SOC, Social versus Non-Social Response; HaND, Habituation and Novelty Detection; FC, Functional Connectivity Networks;
WM, Working Memory; DI, Deferred Imitation task; PCI, Parent-Child Interaction; MSEL, Mullen Scales of Early Learning; LENA, Language
Environment Analysis; FCI, Family Care Indicators; CDI, Communicative Development Inventory; * indicates assessments undertaken in the
family’s home at the later time points.
Page 8 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Figure 2. Summary of study visits: The UK. Abbreviations: EPDS, Edinburgh Postnatal Depression Scale; PRAS, Pregnancy-
related Anxiety Scale; PSS, Perceived Stress Scale; PANAS, Positive and Negative Aect Scale; NBAS, Neonatal Behavioural Assessment
Scale; SES, Socioeconomic Status; MIPH, Maternal and Infant Physical Health; fNIRS, functional near infrared spectroscopy; EEG/ERP,
electroencephalography/event related potentials; SOC, Social versus Non-Social Response; HaND, Habituation and Novelty Detection; FC,
Functional Connectivity Networks; WM, Working Memory; DI, Deferred Imitation task; PCI, Parent-Child Interaction; MSEL, Mullen Scales of
Early Learning; LENA, Language Environment Analysis; FCI, Family Care Indicators; CDI, Communicative Development Inventory; * indicates
assessments undertaken in the family’s home at the later time points.
Page 9 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
or stopped at any time if the infant/child became fussy or
tired, or, if the caregiver expressed a wish to end the study.
Each research team includes researchers fluent in Mand-
inka and/or English, as relevant. Data protection and con-
fidentiality shape our approach to data sharing within the
BRIGHT research team and externally, and is detailed in the
Standardisation of protocol across sites section.
Study protocol
The full study protocol is shown in Table 1 and outlined
below. The draft protocol was developed through the integra-
tion of expertise from our international multi-disciplinary
(psychology, neuroscience, medical physics and bioengineer-
ing, maternal and infant health and nutrition, global health)
research leadership team. Following this, all measures were
Table 1. Summary of BRIGHT protocol. Note on abbreviations: * The UK only; ▵ The Gambia only; ^ Recorded
every two weeks from 2 weeks to 24 months of age; L,W, H, HC,MUAC, KHL (L – length, W – weight, H – height,
HC – head circumference, MUAC – mid to upper arm circumference, KHL – knee to heel length).
Study Measure Time point: Antenatal Birth 7-14
d
1
mo
5
mo
8
mo
12
mo
18
mo
24
mo
Neuroimaging measures
fNIRS: Social/Non-social x x x x x x
fNIRS: Habituation and Novelty Detection x x x x x x
fNIRS: Functional connectivity x x x x x x
fNIRS: Working memory x x x x
fNIRS*/Behavioural: Deferred imitation x x
EEG: Auditory Oddball x x x
Behavioural/Neurocognitive measures
Neonatal Behavioural Assessment Scale (NBAS) x
Eye-tracking: Cognitive control x x x x x
Eye-tracking: Habituation x x x
Eye-tracking: Gap/Overlap x x x x x
Eye-tracking: Non-social contingency x x x
Eye-tracking: Face popout x x x x x
Eye-tracking: Dynamic scenes x x x x x
Eye-tracking: Word-picture-matching x x
Mullen Scales of Early Learning (MSEL) x x x x x
Parent-Child Interaction x x x x x x
LENA language assessment in home x x x
LENA in PCI x ▵ x ▵ x ▵ x ▵ x x
Tablet-based Cognitive Assessment*x * x *
Questionnaires/Interviews – Infant/Child
Communication Development Inventory (CDI) x x x
Adapted Oxford Sleep Diary x x x x x x
Food Frequency Q (FFQ) ▵ ^x x x x x x x x
Infant feeding online Q (IFQ) * ^ x x
Food diary (Intake24UK) * x x x x
Early Childhood Development Index ^
Page 10 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
reviewed during several multi-site web-based meetings to
identify the necessary adaptations and translations required
to ensure each paradigm and assessment was appropriate for
the population and culture of the cohort. Furthermore, some
field testing and adaptation had already been undertaken in
previous pilot phases within our research group (Lloyd-Fox
et al., 2014; Milosavljevic et al., 2019).
During the protocol development phase, for paradigms that
used images, videos, or audio that included people, actors
representative of the ethnicity and language of the participants
were used. For paradigms that included toys or objects (either real
or in image form), appropriate representatives of the contextual
environment of each cohort were identified. For The Gambia
only, when appropriate, questionnaires were translated and
Study Measure Time point: Antenatal Birth
7-14
d
1
mo
5
mo
8
mo
12
mo
18
mo
24
mo
Questionnaires/Interviews – Family
Edinburgh Postnatal Depression Scale (EDPS)
– Maternal x x x x x
Edinburgh Postnatal Depression Scale (EDPS)
– Paternal *x x x x x
Pregnancy Related Anxiety form (PRAS)
– Maternal x
Pregnancy Related Anxiety form (PRAS)
– Paternal*x
Pregnancy Specic Anxiety (PSA) – Maternal x
Pregnancy Specic Anxiety (PSA) – Paternal*x
Positive Negative Aect Schedule (PANAS)
– Maternal x x x x
Positive Negative Aect Schedule (PANAS)
– Paternal*x x x x
Perceived Stress Scale (PSS) – Maternal x x x x
Perceived Stress Scale (PSS) – Paternal*x x x x
Socioeconomic Status (SES) x * x ▵ x * x
Demographic and Family Information x * x ▵ x * x x *
Family details (from DSS – Gambia or antenatal
call UK) x x * x *
Family Caregiving Questionnaire (FCQ) x x x
Family Care Indicators (FCI) ▵x x x
Clinical measures / Medical details
Healthy Pregnancy Questionnaire *x
Antenatal Medical form *x
Delivery Information and Baby check ▵x
Anthropometric measures – infant
(L,W,HC,MUAC, KHL)
x
(L,W,HC)
x
(L,W,HC) x x x x x x
Anthropometric measures – mother (W, H)
Maternal blood sample ▵x
Maternal urine sample ▵x
Maternal breast milk sample ▵x
Infant blood sample ▵x x x x x x
Infant urine sample ▵x x x
Page 11 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
administered in Mandinka (e.g., all mental health question-
naires, and the Mullen Scales of Early Learning [MSEL]). A
full adaptation process with forward and backward translation
by the authors and the BRIGHT Project team in Keneba was
undertaken for all questionnaires and assessments (for an
example see Milosavljevic et al., 2019). As Mandinka is not a
written language and literacy rates among caregivers were
low, the questionnaires were converted into interview ver-
sions and conducted by trained field assistants. Of note, where
translations were undertaken for standardised measures that
were not open-source and free to use (i.e., Mullen Scales of
Early Learning [MSEL]) we purchased the equivalent number
of copies of the original forms from the publisher that would
be required to administer the measure at each age point.
Neurocognitive measures were selected, where possible, on the
basis of test re-test reliability and previous evidence indicat-
ing that they showed robust data quality (i.e., fNIRS (Blasi
et al., 2014); Eye-tracking (Jones et al., 2019); EEG
(Dzhelyova et al., 2019; Räikkönen et al., 2003)). Within each
battery of measures, we selected a combination of well tested
and robust paradigms, and, when necessary to allow us to
target particular cognitive domains or informative metrics,
paradigms with novel designs were developed by the
BRIGHT research group (i.e., fNIRS tasks to assess working
memory, delayed imitation, habituation, repetition suppres-
sion and novelty detection). Tasks were administered in a
pre-determined order across the study visit where possible
(i.e., anthropometrics were taken at the end of the session to opti-
mise infants’ attention and energy for experimental tasks), and
also within a testing modality (i.e., in the fNIRS session infants
viewed paradigms in a set order according to the stimulus pres-
entation scripting framework). On occasion, when infants tired
before completing the full session, families were invited to
return on a separate day to complete the tasks, but where pos-
sible infants were encouraged to continue after a nap and/or
feed within the same visit. We found that a second visit was
required more often in The Gambian cohort than in the UK.
Details of the session and completion of tasks were recorded in a
Session Log Form at each visit.
Neuroimaging measures
Electroencephalography (EEG)
Electroencephalography (EEG) has a long-standing tradition
in neurodevelopmental research. It provides a direct meas-
ure of infants’ neural responses to stimuli without requiring
them to overtly respond or to follow task instructions. Through
the use of innovative, wireless EEG hardware, it is now
possible to implement EEG tasks in remote rural contexts and in
the absence of standardised lab settings (Katus et al., 2019). The
EEG task implemented in the BRIGHT project assessed audi-
tory habituation and novelty detection, at 1, 5 and 18 months
of infant age (for a description of the full protocol see Katus
et al., 2020). Due to the nature of the sounds (pure tones, bursts
of white noise etc.), no adaptations had to be undertaken allow-
ing for identical protocols at both project sites. Infants were
presented with auditory stimuli for approximately 15 minutes
while asleep (at 1 month) or awake (at 5 and 18 months): dur-
ing the latter an experimenter quietly entertained the infant
with bubbles or silent toys to maintain calm attention during
the task.
Functional near infrared spectroscopy (fNIRS)
Functional near infrared spectroscopy (fNIRS) is a relatively
recent addition to the battery of neuroimaging measures avail-
able to neurodevelopmental research. fNIRS measures the
haemodynamic response to the neural activation measured by
EEG. It has become the technique of choice for many studies
given its ease of use with infants and young children, improved
spatial resolution (relative to EEG) and low cost (relative to
MRI) (Gervain et al., 2023; Lloyd-Fox et al., 2010). In addi-
tion, fNIRS is relatively portable, opening a pathway for
implementation in the remote and/or out-of-lab settings often
associated with global health research contexts (Blasi et al.,
2019; Katus et al., 2019).
The fNIRS paradigms implemented in the BRIGHT project
assessed a range of cognitive functions and domains, namely
social cognition (Lloyd-Fox et al., 2017), habituation and
novelty detection (Lloyd-Fox et al., 2019), working memory,
deferred imitation and functional connectivity. Paradigms
were included, at age-appropriate time points, across the 1,
5, 8, 12, 18 and 24 months, as well as in BRIGHT Kids at
3-5 years of age. Paradigms contained auditory and/or vis-
ual stimuli and were presented while infants were asleep at
1 month of age and while infants were awake and alert at all
other time points. Audio and visual stimuli were adapted with
site-relevant content (see the Preliminary Results section and
Katus et al., 2019). The full fNIRS battery lasted 24 min for the
shortest sessions at 1 and 5 months to up to 35 min for the long-
est at 8 and 12 months (where we included a live behavioural
Deferred Imitation task). When accounting for preparation
time such as settling the infant, taking head measurements,
capping, and photographing headgear, the total assessment time
was approximately 45 minutes. This multi-domain battery
was designed to interrogate whether global health risk factors
impact on development to result in global/cross-domain differ-
ences in brain activity or localized/domain-specific differences
or altered function.
Behavioural/neurocognitive assessments
Neonatal Behavioural Assessment Scale (NBAS)
The NBAS is a structured clinical assessment of infant neurol-
ogy and behaviour, which can be performed within the first
few days of life. The NBAS is regarded as the most compre-
hensive examination of newborn behaviour available. It has
been used across multiple cultures and in different LMICs
(Zambia: Brazelton et al., 1976, Chile: Ayala et al., 2021, Mex-
ico: Soler-Limón et al., 2019, Kenya: Super & Harkness, 2020).
The NBAS is a standard protocol which requires an initial
training period culminating in assessment and certification
as described in The Neonatal Behavioral Assessment Scale
Manual (Brazelton & Nugent, 1995). Prior to the BRIGHT
Project, we conducted a qualitative pilot study to assess the
cultural acceptability and feasibility of using the NBAS within the
rural low resource settings of families living in the West Kiang
region of The Gambia. To this end, fifteen infants were assessed
Page 12 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
with the NBAS, and their families’ feedback showed that the
NBAS was acceptable to parents in this population (Bartram,
2018). In line with feedback from parents from other countries
and populations, a few aspects of the assessment (specifically,
shining a light over closed eyes while sleeping, covering eyes
with cloth while awake and undressing the infant) were ques-
tioned or perceived negatively. While no items were altered
or removed from the assessment, the more controversial items
were introduced with special care during all study visits. In both
the UK and The Gambia, the NBAS was performed between
7–14 days after birth during a pre-arranged home visit. Admin-
istration times for the NBAS ranged from 20 to 45 minutes,
depending on the infants’ state of alertness.
Eye-tracking neurocognitive battery
Eye-tracking is a non-invasive and well-tolerated measure
in infant neurodevelopmental research (Jones et al., 2019).
The tasks in our eye-tracking battery were selected to pro-
vide broad coverage of several key domains of neurocognitive
functioning, including working memory (Elsabbagh et al.,
2009; Elsabbagh et al., 2013a; Johnson, 1995; Scerif et al.,
2005), visual attention (Kaldy et al., 2011), habituation
(Webb et al., 2010), reversal learning (Wass, 2015), social
versus non-social visual preference (Elsabbagh et al., 2013b)
and language learning (Fernald et al., 2008). Table 1 displays
the specific tasks used per age point. Many of the tasks,
and all of the fixation stimuli that preceded a task, were
gaze-contingent, that is they rely on the child’s gaze to pro-
ceed through the battery. The duration of the battery therefore
varied slightly between participants but averaged 20 minutes.
The majority of the tasks focused on the use of visual stimuli
(accompanied by simple alerting sounds) with no language
modification required for use in The Gambia. Therefore, the
eye-tracking battery was run with an existing stimulus pack-
age used in other longitudinal large scale cohort studies (devel-
oped by researchers at the Centre for Brain and Cognitive
Development, Birkbeck, University of London) to aid future
comparative analyses. An additional task (word-picture
matching task) was developed specifically for the BRIGHT
project to assess language comprehension of participants at
the ages of 18 and 24 months. The task measured processing
efficiency (speed and accuracy) in terms of infants’ ability
to direct their gaze to one of two visual stimuli to match
to a spoken target noun. This task was adapted from the
behavioural Looking-while-Listening eye movement meth-
odology used by Fernald and colleagues (2008), which had
been recently adapted for a research project conducted in
neighbouring Senegal (Weber et al., 2017). During the design
phase of the paradigm, common items were photographed and
audio recordings of sentences relating to the photographed
items were made in the appropriate language at each site.
A series of pilot studies were then completed in both the UK
and The Gambia to identify population- and age- appropriate
word-picture stimulus pairs.
Mullen Scales of Early Learning (MSEL)
The MSEL measures cognitive ability and motor development
using five scales: Gross Motor, Visual Reception, Fine Motor,
Expressive Language, and Receptive Language. In both the UK
and The Gambia, the MSEL was performed at visits to the research
lab conducted at 5, 8, 12, 18 and 24 months of age, as well as at
3–5 years of age in The Gambia. During each visit, the MSEL
was conducted using the standardized protocol appropri-
ate for the age of the participant, as detailed in the Mullen
Scales of Early Learning Manual and the Item Administration
Book (Mullen, 1995) and the MSEL training DVD. Dur-
ing a pilot phase to optimize and adapt the MSEL for use in
The Gambia, n = 171 infants were tested across the age ranges
described above (Milosavljevic et al., 2019). The MSEL is
broadly similar to the Bayley Scales of Infant and Toddler
Development, both of which have been adapted for use
across multiple countries, but originate in the U.S.A. and are
normed to this population. A further restriction of this meas-
ure is that it is not open source and must be purchased, both
the toolkit, manual and the assessment forms, i.e. a fee per
participant is required to be paid.
Parent-child interaction videos
We assessed parent-child interaction styles in both the UK and
The Gambia at 1, 5, 8, 12, 18 and 24 months of age. The parent
and child engaged in a video recorded, 10-minute free-play
session, which consisted of five minutes of play without
toys and five minutes with a standardised collection of toys
provided by the research team. The parent and child were
seated on a mat in front of a mirror, to ensure that both of their
faces were visible on the recording. At the younger ages, the
infants were placed on a baby mat facing the parent, but, as
they became more mobile, they were allowed to move around
the room. The parent was instructed to play with their child as
they normally would at home. For the younger time points the
parent-infant dyad were left alone in the room to encourage
a more relaxed environment, however, at later time points the
researcher remained in the room to be able to move the cam-
era around as the child became more mobile. These videos
can be coded to assess multiple aspects of parental and child
behaviour and engagement. Parental interactive characteristics
have been shown to associate with child neural and cognitive
development across a range of cultural contexts (i.e. Bozicevic
et al., 2016; Sethna et al., 2017).
Language Environment Analysis system (LENA)
The Language Environment Analysis system (LENA) provides
automated counts of the linguistic environment. A digital lan-
guage processor (DLP) is worn by the participant in the front
pocket of a specially designed vest and able to record the audio
environment within 1- to 3- metres. In BRIGHT, this was used
to record the acoustic environment of the participants during
a typical day in the home. Our standard protocol assessed the
auditory home environment using LENA at the 12, 18 and
24 months of age. The recordings took place during two con-
secutive days, with seven hours of recording per day. During
these recordings, parents were asked to complete a Family
Caregiving Questionnaire or Interview. In the UK this
was a logbook describing the main activities performed by the
toddler, the locations in which the recording had taken place,
who was around during the recording as well as technical
details regarding the usage of the device. In The Gambia,
this was conducted as an interview by a field assistant at the
end of each recording day.
Page 13 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Via the LENA software, we extracted, 1) Adult word counts
(AWC) defined as the number of adult words the key child
hears - these adult words may be or may not be directed to the
key child; 2) Key child vocalization counts (CVC), defined as
words or prelinguistic babbling produced by the key child (cry-
ing and laughing sounds are not included in this category);
3) Conversational turn taking (CTT), identified as those
instances when the key child and an adult speak one at a
time in alternating turns. The reliability of these estimates
has been shown in multiple languages; including English
(Gilkerson & Richards, 2008; Xu et al., 2009), French (Canault
et al., 2016), Dutch (Busch et al., 2018), Spanish (Weisleder
& Fernald, 2013), Vietnamese (Ganek & Eriks-Brophy,
2018) and Korean (Pae et al., 2016). In line with previous
research, field assistants, with Mandinka as their first language,
transcribed a subset of recordings to determine reliability
estimates for the Mandinka language. In addition to the home
visits, within The Gambia a sub-sample (N = 40) of infants
were followed more intensively from 1- 24 months of age
with LENA recording included during the PCI sessions where
the context is restricted to one parent and the target child.
Tablet-basedased cognitive assessment
The Babyscreen software application V1.83 (Hello Games,
Guilford, UK) was used to measure key domains of neurode-
velopment, including selective attention, working memory,
and general learning ability. The task consists of 18 items and
provides two performance variables: number of items accu-
rately completed and speed of item completion (Twomey
et al., 2018). The task was administered on an Apple iPad
(6th generation, 9.7” screen). Given the young age of the
BRIGHT participants and varying exposure to touch screen
technology, two free play tasks were administered at the
start of the testing session to familiarise the participants
with the tablet, these involved drawing on the screen and
moving shapes around.
In the UK, the task was administered to participants at 18
and 24 months of age (Macrae et al., 2022). In The Gambia,
several challenges arose, and the task was removed from the
protocol: whereas pilots of the task were well-received by
infants and parents, the density of assessment during the study
visit meant that infants often were too fatigued to complete
this task. Additionally, participants were more reluctant to touch
or play with the tablet in the lab setting, even after encourage-
ment from examiners and mothers. Attempts were made to
complete the task at separate home visits. However, this had
disadvantages as it increased the burden on the testing team
and introduced wider variance in testing conditions. Therefore,
the task was excluded from the main BRIGHT protocol in The
Gambia.
Questionnaires/interviews – infant/child
Communicative development inventory (CDI)
The McArthur-Bates CDI (Fenson et al., 2007) was used to
assess language development at the 12, 18 and 24 month time
points. In the UK, the full English version was used, which
consists of a vocabulary checklist that asks parents to report
how many words their child can understand and how many they
can understand and say. The questionnaire also asks about
the child’s use of grammar and gestures.
An adaptation of the CDI was developed for use in the Mandinka
language, following guidelines outlined by the MacArthur-Bates
CDI Advisory Board. To construct an inventory of words
for use in the Mandinka adaptation, a list of 200–250 words
was compiled; these were taken from the standard CDI, the
Malawian CDI and the Senegalese CDI. Mandinka-speaking
field staff translated these words into Mandinka and
suggested alternatives if words were not applicable in the
West Kiang district (i.e., baby buggy/stroller) or more affected
by seasonality (i.e. some food items). The inventory probed
specific categories from the MacArthur-Bates CDI, such as
animals, food and drink, and clothing. It also asked whether
the child had started to combine words and use more com-
plex sentences. Subsequently, this inventory underwent pilot
testing in two phases, with a total of 60 mothers of children
aged 24 to 48 months. During the first phase, 30 women
were interviewed, and the list was revised, removing words
that were not frequently endorsed and adding new words that
had been suggested by the pilot participants themselves.
Subsequently, a second phase of pilot testing was conducted,
where a further 30 women were interviewed. From these inter-
views, the list was reduced further by (1) eliminating all
words which less than 10% of mothers said their child knew;
(2) selecting 54 words of moderate difficulty (known by 40
to 70 % of the children); (3) selecting 18 easy words known
by 70 – 100% of the children; and (4) selecting 18 more
difficult words for which 10-40% of the children knew. This
adaptation received full approval by the CDI committee as
an official adaptation into the Mandinka language.
Parental report sleep diary (adapted from the brief infant sleep
questionnaire)
A daily sleep log was administered over three consecutive days
in the week prior to each lab visit and averaged over the moni-
tored period at 1, 5, 8, 12, 18 and 24 months of age. This diary was
adapted from the Brief Infant Sleep Questionnaire (Sadeh, 2004)
and methods for parental reporting (Sadeh, 1996) into a three-day
diary differentially for each site. In the UK the diary was sent to
caregivers by post, and completed by caregivers, with an option
to fill out an online version if preferred. The diary asked for a
record of all periods of sleep (time and location) over a 72-hour
period including information on anything that made the day/night
unusual relative to their regular routine (i.e., illness/activity).
In The Gambia the questionnaire was adapted in several ways
following advice from the local ethics committees and through
the formation of a consultation group comprised of local
research staff who live and work in the West Kiang region of
The Gambia and who had young children (i.e., to provide guid-
ance on the range of locations parents might use for daytime
naps). Firstly, the local population in West Kiang do not adhere
to strict observance of equinoctial hours, nor do they necessarily
possess a time piece in each home. Therefore, the diary was
adapted so that caregivers could answer questions about
sleep based around the sections of the day that are divided
by prayer calls from the local mosques (i.e., morning, after-
noon, evening, last prayer time). The prayer time calls differed
Page 14 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
slightly by season, which will be accounted for in analyses.
Therefore, while fragmentation of sleep, location and number of
daytime naps could be accurately recorded, length of sleep was
restricted to an approximation. As with other questionnaires,
the sleep diary was administered as an interview at the lab
visit with caregivers asked to recall the previous three days
and nights. While it would have been more accurate to
interview caregivers day by day, the research team did not have
the capacity for this many home visits. Families were, how-
ever, reminded to attend to their child’s sleeping patterns for
three days prior to the study visit when they were notified
about their visit date the week before the scheduled research
visit.
Questionnaires/interviews – Family
Family Care Indicators (FCI)
In The Gambia, families were asked to complete the FCI
questionnaire when the infants were 12, 18 and 24 months
of age. The development of this set of indicators was initi-
ated by UNICEF to provide measures of family care practises
and resources with globally relevant application (Kariger et al.,
2012). The items measure the support provided by caregivers
for a stimulating environment for infants to learn from, and the
caregiving resources available within the home. For example,
caregivers were asked about number of books and play items
in the home, who was engaging with the child at home and
how many different types of stimulating activities the child
was encouraged to do. The questionnaire was developed by
an international panel of experts who reviewed existing sur-
veys used in low- and high-income countries (i.e., Home
Observation for Measure of the Environment Inventory;
Caldwell & Bradley, 1984) and field-tested new candidate
questionnaire items across populations in five low-income
countries before finalising this set of indicators for use in
global health and epidemiological studies.
Parental mental health
Parental mental health was assessed using a range of
questionnaires, starting at the antenatal visit and followed up
until the 24-month time point. In the UK, the original English
versions of the questionnaires were given to both parents (where
applicable) to complete in their own time. In The Gambia,
questionnaires were translated into Mandinka and admin-
istered in interview format (see below). Since mothers
always accompanied infants to visits and fathers were often
working away from home, we could only collect data on
maternal mental health.
The Edinburgh Postnatal Depression Scale (EPDS) (Cox et al.,
1987) was administered at the antenatal, 1, 5, 12 and 24
month visits. This is a 10-item self-report questionnaire that
asks participants to rate how frequently they have experi-
enced a range of depressive symptoms in the last seven days.
Items are scored on scale of 0-3 (“No, not at all” to “Yes,
most of the time”) and possible scores range from 0 to 40. A
cut-off of 10 is considered to indicate elevated levels of
depression. The EPDS is a validated tool used to screen for
postnatal depression and has previously been used to assess
maternal mental health in The Gambia, as well as other LMICs
(Coleman et al., 2006; Nabwera et al., 2018).
The Positive and Negative Affect Schedule (PANAS) (Watson
et al., 1988) was administered at the 1, 5, 12 and 24 month vis-
its. This 20-item self-report questionnaire asks participants
to rate how frequently they have experienced a range of posi-
tive and negative emotions in the past few hours. Items are
scored on a range of 1-5 (“Very slightly or not at all” to
“Extremely”). There are six items that correspond to the Posi-
tive Affect (PA) and the Negative Affect (NA) scales, which
are summed to compute scores for each scale, with a possible
maximum of 30 for each scale.
The Percieved Stress Scale (PSS) (Cohen et al., 1983) is a
10-item self-report questionnaire that asks participants to rate
how often they have experienced a series of stress-related feel-
ings in the last month. Items are scored on a scale of 0–4
(“Never” to “Very often”), with a possible total score of 40.
The PSS has been shown to have robust psychometric prop-
erties across diverse low- and middle-income settings (Katus
et al., 2022a) and across different modes of assessment
(Murray et al., 2023). The PSS was administered at the antenatal, 1,
5, 12 and 24 month visits.
The Pregnancy Related Anxiety Scale (PRAS) (Rini et al.,
1999) and the Pregnancy Specific Anxiety scale (Roesch
et al., 2004) were used as measures of anxiety related to preg-
nancy at the antenatal visit only. The PRAS is a 10-item
scale that asks respondents to rate how frequently they have
experienced a range of concerns related to their pregnancy in
the last few months. The scale is rated on a scale of 1-4 (“Not
at all” to “Very Much” or “Never” to “A lot of the time”).
The total score is computed by summing scores on all items,
with a maximum score of 40 possible for the scale. The PSA
asks participants to rate how often they have felt a range of
emotions in the last week. Scores range from 1-5 (“Never”
– “Always”). A total score is generated by summing the
scores on four items that are specific to anxiety (anxious, con-
cerned, afraid, panicky). Total scores can range from 4-20. The
paternal versions of these questionnaires ask fathers to rate
their feelings in reference to their partner’s pregnancy.
Over a period of eight months (2015-2016) each questionnaire
was adapted following World Health Organization guidelines
(World Health Organization, 2013), the questionnaire devel-
opers, and procedures described in other studies using these
measures in LMICs (Hanlon et al., 2008; Kohrt et al., 2016;
Nabwera et al., 2018; Tesfaye et al., 2010; Weobong et al.,
2015). The adaptation protocol was the same for all measures
and involved a core team of five researchers, as well as an addi-
tional nine staff members who supported this intermittently
where needed. an initial translation from English to Mandinka
by a panel of three Gambian research staff, the local PI (MD),
who were all native speakers, and two researchers from the
UK who were experienced in mental health data collection.
Following best practice guidelines outlined by (Peña, 2007),
we attempted to align the translated items as closely as
Page 15 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
possible to the original English, while taking into account cul-
tural equivalence. This involved replacing English idioms and
medical terminology with phrases that would be comprehensible
in Mandinka, and changing the structure of statements into ques-
tions (Kohrt et al., 2016). The translation process involved
several rounds of back translation by Mandinka-speaking
staff who were blind to the original questionnaires and the
translation process, as well as several meetings with local
clinical staff (i.e. midwives) and visits to families to discuss
the language of the questions with mothers and pregnant women.
Where discrepancies were noted between the original English
and the back-translation, or regional variations in wording iden-
tified, the panel made necessary adjustments. The translated
questionnaires were pilot tested with N=12 volunteers from
West Kiang, to assess their understanding of the measures,
corrections were made where issues with comprehension
emerged.
While every attempt was made to ensure equivalence
between the original English and the Mandinka translations,
one item on the EPDS (item 10) that asked about suicidal
ideation/behaviours was changed because of the highly sensi-
tive nature of the question in this culture and local population.
(Nabwera et al., 2018) noted that, due to the highly commu-
nal way of life in this community, the desire to be isolated from
others was seen as a sign that the individual may be suffer-
ing from a mental health problem. Therefore, this item was
changed to ask participants whether they wanted to be isolated
or alone. Participants who scored above clinical cut-off
(a score of 10) on the EPDS were given the opportunity to be
referred to the MRC clinic for support. Furthermore, some
of the English words used, in the PANAS in particular, were
not differentiable in Mandinka, and so the number of items
were reduced in The Gambian version relative to the one
administered in the UK.
To simplify administration and reduce recall of response
options, mothers were first asked whether they had experi-
enced the issue described in the question (yes/no) and, only
if they responded with a yes, would the interviewer elaborate
with the frequency options (Hanlon et al., 2008). To help
mothers remember the time period that each questionnaire was
referring to, they were administered in order of timeframe,
from shortest to longest, and the timeframe was reiterated with
each question. Finally, to reduce the impacts of stigma, moth-
ers were reminded that all participants were being asked the
same questions.
Field staff responsible for administering the interviews received
extensive training on understanding the conceptual framework
of each measure. Subsequently, they were trained in admin-
istration using vignettes and role play scenarios, practicing
administration and managing different types of potential
responses.
Socioeconomic status (SES), demographic and family
information
Families were asked to complete a questionnaire (UK) or inter-
view (The Gambia) regarding their family demographics
and socioeconomic circumstances. These were conducted
as a series of questionnaires/interviews spanning from the
first antenatal visit across the postnatal sessions, tailored to
ask questions relevant to each time point, and with reduced
time burden at each session, given that questions became an
update on whether circumstances had changed. At both sites
information was gathered on biological parents, biological
grandparents (ethnicity, date, and place of birth) and any other
applicable caregivers of the key participating infant/child. The
caregiver information gathered included ethnicity, age, car-
egiving role, employment status, highest level of education,
languages spoken. Wider family information was collected
including parity of parents, size, and composition of house-
hold. Finally, housing information was collected pertaining to
indication of wealth, i.e., number of bedrooms (UK), wall/floor
materials, access to water, durable assets (The Gambia). In
The Gambia this information was gathered through a combi-
nation of observation by the field assistant during home visits
and reports from the participant during interview.
Pregnancy, birth and family health information
In the UK and The Gambia mothers were asked to complete
a questionnaire (UK) or interview (The Gambia) regarding
their pregnancy, birth, and family medical history. Informa-
tion gathered included: (i) antenatal information on maternal
obstetric and medical history; (ii) fetal ultrasound information
including gestational age and anthropometric foetal measure-
ments (The Gambia only); (iii) birth and delivery information
including neonatal anthropometric measurements (The Gambia
only) and a maternal health check.
Growth and diet measures
Anthropometric measures
Anthropometric measurements were made by research assistants
in the UK and by field assistants in The Gambia. In both sites,
measurements were taken in triplicate, following standard
protocols and all staff underwent training. In the Gambian
sample, maternal height and weight were measured in late
pregnancy and infant length, weight and head circumference
were measured at birth. In addition, infant length, weight, head
circumference, mid-upper-arm-circumference (MUAC) and
knee-heel length were measured at both sites at 7-14 days, 1,
5, 8, 12 ,18 and 24 months of age. At birth, infant length was
measured using a flexible length mat, and a fixed length board
(SECA 417) was used thereafter. Infant weight was meas-
ured using a calibrated electronic baby scale (SECA 336), with
a precision of 10g. Mid-upper-arm circumference and head
circumference were both measured using a SECA 201 head and
body measuring tape, precise to 1mm. Knee-heel length was
measured using a calliper, also precise to 1mm.
Dietary data
In The Gambia, infant feeding data was collected every
two weeks from birth to 24 months of age. The question-
naire was administered verbally by a field assistant at the
participant’s home or by telephone if a home visit was not
possible. The mother was asked to report on the infant’s diet
in the two weeks prior to the questionnaire. Details included
whether the infant received breastmilk feeds, and/or other
Page 16 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
liquids, semi-solid or solid foods. The questionnaire included
common examples of local weaning foods, as well as free text
space for additional items. The mother was also asked to report
the frequency (never, once, more than once, most days) at
which the infant received each food or drink.
In the UK, similar feeding questionnaires were completed
monthly by parents online, from 1-7 months of age. In addi-
tion, parents completed a detailed food diary reporting all food
and drink consumed by their infant, for four consecutive days
prior to each study visit from 8 months onwards (8, 12, 18,
24 months). This data was then coded on a food composition
database (DINO; (Fitt et al., 2015)). Mothers also completed
online 24-hour dietary recall questionnaires reporting on their
own diets, in pregnancy and at 6 and 12 months postnatally,
using the Intake24 UK platform.
Biological samples
In The Gambia, to investigate nutritional factors in more detail,
samples of breast milk, blood and urine were collected. Mater-
nal venous blood and urine were collected in late pregnancy
(34-36 weeks’ gestation) and breast milk was collected from the
mother at the five-month visit. Infant urine was collected at 5, 12
and 24 months of age. Samples were stored at -70°C for subse-
quent analysis. In addition, infant blood samples were collected
at all infant visits, alternating between a 0.5mL capillary sample
(at 1, 8 and18 months) and a 3mL venous sample (5, 12 and 24
months). On each sample, a full blood count was run using
a Medonic analyser and the remaining sample was centri-
fuged. Plasma and cell pellets were separated and stored
at -70°C for subsequent nutritional analysis and DNA
extraction, respectively.
Standardisation of protocol across sites
All measures were collected using Standardised Operating
Procedures (available upon request). Longitudinal infant and
toddler testing requires standardisation of the (i) equipment
(ii) environment in which the measures are administered,
(iii) experimental protocol, and (iv) behaviour of research-
ers during administration of measures. The site in The Gambia
had not previously undertaken research of this kind until the
pilot phase of the BRIGHT project (Lloyd-Fox et al., 2014;
Milosavljevic et al, 2019). Therefore, all equipment and test-
ing materials had to be purchased prior to the start of the study
sessions. To reduce site differences due to hardware, an iden-
tical set of equipment was purchased for both sites. While it
is challenging outside of a research lab context to replicate
the environment that the testing is undertaken in, where
possible, we replicated the UK room setups at the Gambian
site. In the UK (both at the hospital and university sites)
testing rooms for visits from 1 – 24 months of age were
sound proofed and windowless with temperature and lighting
control. In The Gambia, the rooms were air conditioned to
control the temperature, and the neurocognitive testing (fNIRS,
EEG, eye-tracking) room was windowless with some light
control. However, none of the rooms were sound-proofed.
Therefore, environmental noise was more inconsistent
across data collection within the Gambian sample as external
sounds could sometimes be heard within the testing rooms.
At both sites testing at 7 – 14 days of age was done at the
family’s home, therefore environmental noise differed between
the cohorts (for example family size was generally larger
in The Gambia and houses often had open windows and
doors– see Table 3). To address this, researchers at both sites
optimised data quality where possible by discussing the needs
of each measure with the family who were present during data
collection (i.e., they discussed with the family that there would
be times when they needed the room to be quiet during a
measure of attention to sound or light, or when they might
need their help in eliciting a smile from their baby)).
Broadly, the protocol was identical across sites. As described
above, some measures were site-specific either because a
measure was found to be unsuccessful at one site (i.e., the
tablet task) or because the measure is only relevant at one site
(i.e., FCI). For the neurocognitive testing we adopted the use of
the TaskEngine framework, which was developed for a separate
multi-site neurocognitive study, Eurosibs (Jones et al., 2019),
to optimise data quality and standardisation of acquisition.
This framework allows the presentation of the paradigms to
be identical across sites and produces identical data outputs
for cross-site quality control reports and analysis.
As described above, the order of testing was kept consistent
across sites, however we have adopted the practice of stand-
ardization with flexibility to be responsive to the needs of
the individual infant/toddler. Before the onset of the project,
project coordinators for both sites were trained in the UK for a
period of two months. Following this, training continued in The
Gambia for one – two months (depending on the measure) and
the first study sessions were conducted under supervision. The
BRIGHT project team were committed to building long-term
capacity for neurodevelopmental research at MRC Keneba and
therefore across the duration of the project, trained and sup-
ported local researchers to conduct and co-ordinate all aspects of
the research in The Gambia. To facilitate harmonization across
sites, staff were trained in the practical detail of data collec-
tion and administration as well as researcher responsiveness to
infant behaviour.
Working in partnership with local researchers during their train-
ing phase, we discussed the observed infant responses and
behaviours elicited by each paradigm and used this experi-
ence to devise a standardized strategy for responding to the
needs of the infant and caregiver. This strategy was agreed
at both sites and included in the Standard Operating Proce-
dures (SOPs). These steps took into account the infant’s state of
alertness and fussiness, as well as the caregivers needs,
to ensure optimal data collection and participant comfort.
Significant fussiness typically leads to inadequate data qual-
ity across both the neurocognitive and behavioral meas-
ures. Fussiness was defined as excessive motion (i.e., infant
wriggling on lap, or toddler walking away from task), inat-
tention to the task (i.e., looking away) or negative affect (i.e.,
crying, negative vocalizations). Furthermore, particularly in The
Gambia, a further category of inattentiveness existed as infants,
Page 17 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Table 2. Participant characteristics (age/sex) and retention rate of cohort at each time point. Note: If infant became
tired or fussy before session was complete a call back was arranged for a second visit to complete testing; DOV = date of visit; N =
sample size; SD = standard deviation of mean.
The Gambia (N = 214)
Sex (female/male) 103/111
Maternal age at birth
Mean, SD (min-max) 29.76 (6.61), 18.2 – 44.7
Timepoint 7-14 days 1 months 5 months 8 months 12 months 18 months 24 months
Enrolled at DOV (N) 205 204 200 193 192 191 185
Attended visit (N) 157 185 198 188 188 177 161
Mean age in days (SD) 12.3 (3.87) 36.0 (5.64) 159.9 (10.14) 247.2 (11.39) 372.7 (14.05) 558.6 (17.13) 745.0 (27.94)
Range in days 5 - 44 29 - 65 148 - 208 211 - 314 353 - 428 533-641 722-896
% attended visit 76.5 90.7 99.0 97.4 97.9 92.7 87.0
% attended two visits
per age point* n/a 3.2 22.2 23.9 25.9 37.9 16.8
UK (N = 62)
Sex (female/male) 31/31
Maternal age at birth
Mean, SD (min-max) 32.96 (2.93), 28.4 – 40.8
Timepoint 7-14 days 1 months 5 months 8 months 12 months 18 months 24 months
Enrolled at DOV (N) 62 61 60 60 60 57 50
Attended visit (N) 58 60 58 57 59 55 50
Mean age in days (SD) 12.2 (3.33) 33.2 (5.53) 155.8 (6.54) 251.7 (9.89) 375.7 (12.51) 557.1 (15.02) 736.9 (15.79)
Range (days) 7 – 23 22 – 56 144 – 184 235 – 279 353 – 411 536 – 603 700 - 784
% attended visit 93.6 98.4 96.7 95 98.3 96.5 100
% attended two visits
per age point* n/a n/a n/a n/a n/a 27.3 20
on occasion, become drowsy or fell asleep during the task, likely
an effect of the climate. We agreed upon a hierarchy of responses,
tailored to the task, to ensure maximum participation. For
example, during screen-based tasks (i.e., eye-tracking, fNIRS) to
address possible boredom, when the infant began to fidget and
look away, “attention-grabbers” (non-social short sounds) were
employed to re-orient the participant to the screen. These were
automatically recorded by the task presentation framework.
If this was ineffective the researcher moved through the
following strategies (also applicable to other measures); par-
ents were asked to speak reassuringly and hold hands, the
infant was given something to hold or at older age points
infants/toddlers are offered a snack (i.e. rice cake/rusk), a
short break was offered before re-engaging with the task, or if
none of these strategies were successful families were given
a longer break for a nap or feed. If the parent was happy to
resume the session continued. If a participant was unable to
complete the full session within a day, families were asked if
they were happy to return on a subsequent visit. All “manual”
strategies of engagement and breaks for naps and food are
recorded in the Session Log Form.
Data analysis plan
Quality control
Frequent refresher training sessions and quality control
have been implemented across the duration of the BRIGHT
project. At the Gambian site, due to the number of partici-
pants, time frame of testing and number of age points measured,
testing load was very high at the peak of the project – with
up to four infants tested per day, seven days per week, with
staff on rotating schedules - for over two years. To main-
tain high data quality, research staff were trained to be highly
specialised in a subset of measures rather than every measure.
A researcher might oversee the neurocognitive tasks (EEG,
fNIRS, eye-tracking), behavioural measures (MSEL, NBAS),
mental health parent interviews or anthropometric measures.
Page 18 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Table 3. Demographic characteristics and socioeconomic
status, The Gambia. This socioeconomic information is derived
from that reported or observed at the 7-14-day home visit. n=172;
IQR = inter quartile ratio; NA = not applicable.
Mothers (%) Fathers (%)
Education
No formal education 59.4 55.0
Some primary education 12.9 5.9
Complete primary education 3.5 4.7
Some secondary education 18.2 7.1
Complete Secondary education 5.9 27.2
Household characteristics
Number of children, Median (IQR)
min-max 5(4), 1-10 6 (7), 1-23
Number of wives NA 1 (1), 0-4
Assets
Household (median (IQR),
min-max)
Number of people in
household 11 (8), 3-36
Diet
Meals per week containing meat 1 (1), 0-5
Meals per week containing sh 6 (1), 0-7
Meals per week containing sh
or meat 7 (0), 2-12
Housing Attribute Households (%)
Primary Water Source
Open Public Well 2.25
Protected Public Well 7.87
Public Tap 87.6
Piped water in compound 2.25
Cooking Fuel
Firewood 96.1
Charcoal 3.95
Toilet Facilities
Pit Latrine 96.07
Improved Pit latrine 1.69
Flush Toilet 2.25
Flooring
Earth, Sand, Mud 17.4
Cement 70.2
Vinyl 0.56
Tiles 5.62
Carpet 6.18
For example, for fNIRS studies, to minimize data loss,
designated researchers were trained to monitor the system’s
performance, detect potential problems with the acquisition
and instructed to implement basic repairs (Blasi et al., 2019).
Designated researchers with expertise in a particular measure
were also supervised by a senior team member with special-
ist knowledge of the measure or paradigm. The majority
of the senior team members were located in the UK. From
the start of the project, web-based multi-site meetings were
routinely held. For the duration of the project, these consisted
of fortnightly group meetings (which include senior team
members and all designated researchers at the sites of data
collection) and fortnightly quality control meetings (which
included at minimum one senior team members and designated
researchers). At these meetings recruitment, data quality,
testing practices, outputs and other issues were discussed. Fur-
thermore, during the quality control meetings training sessions
were also conducted, including researcher responsiveness,
as and when the need arose. For example, during the life
course of the project, training in the MSEL had to be adminis-
tered as each new age point was reached to ensure the data was
being collected in a standardised way across sites. During the
BRIGHT-Kids phase of the project, training on all meas-
ures was complete, therefore, team meetings were reduced to
fortnightly quality control meetings until completion of data
collection.
After data collection, each data set required extensive quality
control assessments.
For example, EEG and NIRS data needed to be assessed for
artefact from motion and infant inattentiveness, and segments
of data removed as appropriate. Session level data from the
“Session Log Form” was also reviewed so that contextual vari-
ables such as infant state (e.g., fussiness), interference from
Housing Attribute Households (%)
Roong
Corrugate 100
Walls
Mud 12.9
Earth Bricks 3.93
Cement/ Burnt Bricks 82.0
White Lime 1.12
Household Assets
Electricity 2.81
Television 16.3
Refrigerator 5.06
Bicycle 73.0
Motorbike 14.0
Other vehicle 7.87
Page 19 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
researcher/caregiver (e.g. talking during task), experimenter
error (e.g. video not recorded), technical issues (e.g. computer
program crashed, headgear misaligned for NIRS/EEG), or
position of participant (e.g. infant facing wrong way, or par-
ent out of view during interaction video) could be taken into
account. Data was marked for validity at each stage of this
process to monitor data attrition during the stages of the
processing stream. Data quality metrics were extracted for all
paradigms at each site and age point. Specific guidance for
conducting EEG and fNIRS studies in global health con-
texts to maintain high quality control measures can be found
in two methods papers concerning the BRIGHT project (Blasi
et al., 2019; Katus et al., 2019)
Data storage and handling
While it would be preferable for data pre-processing and analysis
to be conducted at the site of data acquisition, such skills
are highly specialised and usually attributed to post gradu-
ate education in neuroimaging and such expertise among local
researchers in The Gambia is, as of 2023, very limited. This
is likely because developmental neuroscience is a very new
field of research in The Gambia, and as such, opportunities to
learn the required analytic skills have previously been lack-
ing. This has been identified as an important focus for capac-
ity building in the long term. However, to ensure timely data
quality control during the BRIGHT project, much of the data
analysis was conducted within the UK and a reliable data trans-
fer method had to be established to enable this. A dual protocol
was designed to ensure the integrity of the data transfer.
As a standard procedure, initially, the data was stored and
backed-up locally on site in a separate location to the research
data acquisition computers. Personal data (i.e., contact details,
DOB) was stored securely at each site within locked cabinets as
well as within a password encrypted electronic database iso-
lated from the research data. Data from the anthropometric
measures, parent-report questionnaires, and behavioural assess-
ments (i.e., NBAS, MSEL) was pseudonymised and housed
in password-protected encrypted databases locally. All data
obtained using paper forms was double entered on the local
databases to ensure reliability across staff. Experimental data
(i.e., eye-tracking, neuroimaging measures, parent-infant
videos) was pseudonymised and stored on password encrypted
storage hard drives at each local site. Following this, a Secured
File Transfer Protocol (SFTP) server functioned as a bridge
between sites. This system allowed the transfer of data in both
directions to account for (i) planning of protocol updates,
(ii) software and stimuli transfer from the development site
(London, UK) to cohort sites (Cambridge, UK and Keneba, The
Gambia), (iii) data transfer from the cohort site and (iv) feed-
back on pilot data and quality control checks (see Blasi et al.,
2019 for further detail). A second full copy of all research data
was transferred to a secure password encrypted server in one
of the participating UK centres (at the Centre for Brain and
Cognitive Development, Birkbeck, University of London).
One consistent challenge was the unreliable internet and/or
poor bandwidth, which lead to slow transfer times for large
files. The fNIRS tasks of the BRIGHT project, for exam-
ple, involved acquiring over 160 files, equating to over 3GB
of data, in total across all time points for each participant
(Blasi et al., 2019). Furthermore, during the life course of the
project over 25,000 data files were collected for the infant/
child measures alone (this excluded datasets for anthropo-
metric data, biological data and family questionnaire data).
Therefore, when the rate of data transfer became restricted, we
prioritized the transfer of specific data types based on the data
quality control checks and analysis pipelines required for each.
Data transfers, quality control checks and inventories were car-
ried out at regular intervals. Access was fully audited, and to
ensure data security, access is governed by a management team.
A web-based interface enables internal BRIGHT researchers to
access the database using personalized login details to search,
filter, and download data. Data access is overseen by the data
management team in the UK and The Gambia, and access is
granted for internally pre-registered projects (see Data.
Data analysis and statistical plan
Overall, we aim to identify developmental brain and neuro-
cognitive function-for-age curves across the first two years of
life, establish which context associated moderators (i.e., under
nutrition, family income, caregiving/family support) impact
significantly on infant development and how these associate
with pre-school outcomes at three to five years. Further to the
plans outlined for each aim below, tests of normality and
sensitivity analyses (comparing observed values and imputed
missing values) will be conducted. Non-linear tests of
significance and interpolation approaches will be applied where
appropriate.
To address Aim 1, brain and neurocognitive function-for-age
curves from birth to 24 months of age will be generated across
1, 5, 8, 12, 18 and 24 months of age. These will be derived from
the fNIRS, EEG, MSEL and neurocognitive eye-tracking bat-
teries of tasks. These reference curves will be used to enable
age-adjusted group comparisons of differences in average trajec-
tories, group-wise differences in variability, and for character-
izing the range of individual developmental trajectories within
each cohort. To generate appropriate metrics for each data-
set we will explore which type of derivative measures and the
level of complexity required to meaningfully capture brain
and cognitive change across this developmental window. For
the neuroimaging data we will explore these derivative meas-
ures (i.e., localisation versus globalisation of brain response,
latency of response, profiles of change in haemoglobin) by
running time varying parameter models to explore variation
across and within these values over our varying age points.
Following this, several approaches will be undertaken,
including longitudinal growth modelling to explore relation-
ships between measures, and regularisation methods such as
Page 20 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
latent function connectivity modelling, to explore whether
we can derive a common measure across the different brain
function tasks which predict rate of change in other measures
such as MSEL and the neurocognitive battery.
To address Aim 2, a combination of quantitative analyses will be
undertaken to establish the association between context-associ-
ated moderators (i.e., undernutrition in mother and infant and
consequent growth faltering, parental mental health, SES factors)
and developmental trajectories across the first two years of
life in The Gambia. To generate appropriate metrics for each
dataset we will explore which type of derivative measures
most meaningfully capture measures of poverty (i.e. maternal
iron status versus infant iron levels, physical growth at birth,
versus change in growth measures over time, SES measures
of income, household size, household assets). Structural equa-
tion modelling (SEM) and hypothesis-driven regressions
will explore how these latent pre- and post-natal variables
associate with latent outcomes of infant development and
regression analyses conducted to understand the directional
relationships between outcome variables.
To address Aim 3, outputs from Aims 1 and 2 will be nomi-
nated using lasso regression coefficients in relation to longitu-
dinal brain and cognitive development trajectories (i.e. across
language, motor, sensory, attentional correlates, brain connec-
tivity, ) across 0–24 months of life using SEM, applying full
information maximum likelihood to account for missingness
and to identify developmental-hypothesis driven clusters to
explore how context-associated moderators (including risk fac-
tors of poverty) of altered infant development, and the impact
of these in turn on pre-school outcome measures of early
child development.
Preliminary results
Recruitment and retention of participants: The Gambia
Figure 3 illustrates the recruitment and retention of par-
ticipants in The Gambian cohort. In total 280 families were
recruited and consented into this project. A total of 58 families
were excluded from the study prior to delivery, as outlined in
Figure 3, leaving a total of 222 families enrolled at delivery.
Eight infants were stillborn, leaving a total postnatal cohort of
214 mother-infant dyads. A further seven infants were lost to
neonatal death during the first two weeks of life and two families
chose to withdraw from the study prior to the first postna-
tal home visit at 7–14 days of infant age, leaving 205 mother
infant dyads within the study. At the 24-month time point,
185 families remained enrolled within the study.
Recruitment and retention of participants, UK
Figure 4 illustrates the initial recruitment process in the UK,
and the retention of participants throughout the study. In total,
62 families were recruited and consented into this study, under-
took the antenatal assessment and took part in their first post-
natal home visit when the infants were 7–14 days of age.
Participating families live either in the centre of Cambridge
(N = 22) or in surrounding urban or rural communities within
a 20-mile radius (N = 40).
Participant attendance to study visits
This data was derived from the first data collection point and
includes all available data for N = 214 (The Gambia) and N
= 62 (UK). At each age point a preferred testing window was
established; +/- 1 week at 7-14 days, +/- 2 weeks at 1 – 8
months, +/- 1 month at 12 – 24 months. To minimise data loss,
we allowed data collection up to + 1 month at 5 – 8 months,
and + 2 months from 12 months upwards. This became
necessary on occasion, particularly if both parents returned to
work or when families had travelled outside of the region and
were unavailable at the time of test. The total number of par-
ticipants who attended each visit is given in Table 2. Due to the
COVID-19 pandemic, testing of the full 24-month time point
was suspended in The Gambia during their first country-wide
lockdown. Following a prolonged period of suspension of
research, a decision was made to end data collection for the
BRIGHT project, therefore 20 participants could not be invited
for their final time point at 24 months of age. For those families
enrolled at each time point (i.e., excluding those who had
been withdrawn for health reasons or because they moved
away) we have experienced a high retention rate for the major-
ity of the completed time points (> 90% of enrolled cohort
attended visits). In The Gambia the exception has been at the
7 – 14 days visit, where the proportion of families attend-
ing the visit fell to 76.6%. This occurred as a result of moth-
ers who gave birth away from their village and would therefore
stay with other family members during the first weeks of their
infant’s life. Consequently, it was sometimes challenging to
identify that these women had given birth and arrange their first
visits. Furthermore, during late 2016 – early 2017 the country
experienced political unrest during the general elections, which
impacted on our ability to schedule infants at 7–14 days
and 1 month of age for their visit. In the UK, retention rates
have been high in the first time points, for example 91.9%
of participants were seen at all five of their visits during
their first year of life. While the majority of those families
were still enrolled in the study at 18 and 24 months of age and
attended the visit, we experienced higher rates of self-withdrawal
in the UK as the testing burden on some participating families
became too high (the most common reasons for withdrawal
were either that both parents had returned to work or that
the mother was expecting/had given birth to a further child).
Demographic characteristics and socioeconomic status
(SES), The Gambia.
A summary of the family demographic and SES distribu-
tion of The Gambian cohort is given in Table 3. Families live in
multigenerational households with up to 36 members per
compound. Polygamy was common within the cohort with
38.8% of fathers having more than one wife. Consequently,
while mothers had on average 4.4 children, including the
infant enrolled in the study, fathers had on average 6.9, with a
range of 1 – 23 children attributed to a single father.
For the generation of parents within our cohort, formal school-
ing was readily available when they themselves were children,
therefore, on average, mothers and fathers within the study
had completed three and four years of schooling respectively.
Page 21 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Figure 3. Recruitment and retention of participants, The Gambia.
Page 22 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Figure 4. Recruitment and retention of participants, UK.
Page 23 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Over five times more fathers (26.6%) than mothers (5.62%)
had completed high school (grade 12). Subsistence farming
was the most common profession among both mothers
(64.4%) and fathers (43.4%). In terms of livestock, fathers
owned more livestock (sheep, goats, donkeys, or cows) than
mothers. Goats were the most commonly owned livestock
among both mothers and fathers, with 45.5% mothers and
46.6% fathers owning at least one goat. Cows are the most valu-
able livestock, and ownership was heavily skewed in favour
of the fathers with 29.7% of fathers compared to 6.7% moth-
ers owning at least one cow. Around one third (34%) of farming
fathers sold a proportion of their produce, however the
mothers’ farm produce was largely consumed by the household,
with only a small minority (6%) of mothers selling any of their
harvest. Durable asset ownership was low, while 73.3% of
households reported owning a bicycle, only 14% owned a motor-
bike, 7.9% owned a vehicle, 16.3% a television and 5.1% a
refrigerator.
Within the cohort, the mother was reported to be the pri-
mary caregiver for all infants at 7–14 days of age, and for
all but one child, where the primary caregiver was reported
as the aunt, at 18 months of age. As can be seen in Figure 5, at
7-14 days of age, the most common secondary caregiver
was the grandmother (41.1%), whereas by 18 months almost
half (45.0%) of all participants reported an elder sibling to
be the secondary caregiver. Fathers were reported as second-
ary or tertiary caregivers in 28.2% of families at 7–14 days,
and in 32.2% of families when infants were 18 months old.
All primary caregivers reported that their first language was
Mandinka. In addition, 16.7% of primary caregivers reported
that they spoke a second language and 3.5% of primary car-
egivers spoke three languages. Mandinka was also the most
common first language among secondary and tertiary caregiv-
ers, the only exceptions being three participants for whom
their tertiary (n=2) or secondary and tertiary (n=1) caregivers’
first language was an alternative local language. In addition,
25% of secondary caregivers and 13% of tertiary caregivers
spoke a second language and a small minority spoke a
third language (2.3% and 4.1% of secondary and tertiary
caregivers respectively) (Figure 6). Other than Mandinka, lan-
guages spoken included local languages such as Fula, Wolof
and Jola, as well as English and Arabic.
Demographic characteristics and socioeconomic status
(SES): UK
A summary of the family demographic and SES distribution of
the UK cohort is given in Table 4.
As can be seen in Figure 7, at 8 months of age, caregiving in
the UK cohort was primarily provided by the mother and/or
father. Other reported caregivers at this age included grand-
parents and a childminder. Similarly, at 18 months of age,
caregiving was primarily provided by the mother or jointly by
the mother and father, with the exception of one family who
reported the grandmother as the primary caregiver. Nursery
workers, childminders and grandparents were also frequently
reported as additional caregivers at this age (Figure 8).
In the UK we followed a recruitment strategy that encom-
passed natural population variance within the region of recruit-
ment. A large proportion of the population living in the city of
Cambridge and surrounding areas is multi-lingual. Consequently,
a significant proportion of our recruited infants were exposed
to multiple languages (Figure 8). As highlighted in Figure 9,
exposure to languages both within, and outside of, the home
was recorded antenatally and then at 8, 18 and 24 months of
age. Within the home, the proportion of households exposed
to English-only varied from 55 – 72.9% across age points. A
further 13.6 – 27.5% of households had one additional lan-
guage, 10 – 15.4% had two additional languages and 0 – 7.5%
of infants were not exposed to English within the home at all.
The languages that the infant was exposed to outside of the
home differed with 75.6 – 85.7% of families reporting that
infants were exposed only to the English language outside of
home across 8 – 24 months of age. A further 11.4 – 14.7%
reported that their infants heard English and one additional
language outside of the home, one further family reported
that their infants heard two other languages in addition to
Figure 5. Additional caregivers at 7–14 days and 18 months of age, The Gambia. The Figures display the make-up of secondary
and tertiary caregivers at 7–14 days and 18 months of age in The Gambian cohort. Secondary caregiver; a person who looks after the
infant a substantial proportion of the time. Tertiary caregiver; a person who looks after the infant ‘sometimes’. A. n=170, B. n=165, C.
n=171, D. n=168.
Page 24 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Table 4. Demographic characteristics and socioeconomic status, UK.
Note that this summary is compiled from the 18 month visit; SD = standard
deviation of the mean.
Mother
(%)
Father
(%)
Education
Secondary 3.2 6.5
Tertiary 3.2 4.8
Undergraduate 33.9 33.9
Postgraduate 54.8 48.4
Family Ethnicity
Caucasian 87.1 83.9
Asian 3.2 6.5
Black 1.6 1.6
Mixed/Other 3.2 3.2
Parental SOC Classication
Higher managerial or professional 69.8 76.8
Intermediate 22.6 19.6
Routine/semi-routine 3.8 3.6
Unemployed > 6 months 3.8 0
Household (mean (SD), min-max)
No of children (incl. key child) 1.19 (0.4), 0-3)
Annual Household Income Percent of Households (%)
£20,000 - £29,999 1.6
£30,000 - £39,999 1.6
£40,000 - £59,999 25.8
£60,000 - £79,999 37.1
£80,000 - £99,999 21.0
£100,000 - £149,999 6.5
> £149,999 1.6
Do not wish to answer 3.2
Figure 6. Number of languages spoken by caregivers, The Gambia. The Figures show the proportion of primary (n=172), secondary
(n=171) and tertiary (n=167) caregivers at 18 months of age, who spoke one, two or three languages.
Page 25 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Figure 8. Number of languages heard in the home by infants at 18 months.
Figure 9. Percent of English heard by UK infants across age points. Note: While the majority of families completed this questionnaire
antenatally (95.2%), the proportion of families who provided this information at 8 – 24 months dropped signicantly to 62.9 – 74.2%
complete datasets from the full cohort. This was due either to families withdrawing from the study at 18 – 24 months or missing data within
the demographic questionnaire.
Figure 7. Caregivers reported at 8 and 18 months of age, UK. The Figure shows the primary caregivers and other reported caregivers
at 8 and 18 months of age for the UK cohort. 8 months of age n=46, 18 months of age n=48.
Page 26 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
English, and one family reported that their infant did not hear
English outside of their home at all. At this stage we wish to
avoid premature consideration of cultural and linguistic fac-
tors that could affect the interpretation of our results within
the UK, given that this was not a primary focus of our project.
However, this factor will be taken into account during our
planned analyses.
BRIGHT Kids: Implementation, challenges, and
successes
The protocol for BRIGHT Kids is outlined below and summarised
in Table 5.
Aims of BRIGHT Kids
While a growing body of research has begun to identify early
markers of risk and resilience, the consequences of adver-
sity often do not become fully manifest until later in child-
hood. Preschool age is a period where children experience rapid
progress in a number of developmental domains (e.g., language,
executive functions), as well as an increase in external demands
from caregivers. Consequently, this is a period where delays
in development start to become outwardly observable and
to interfere with everyday functioning. Thus, to fully under-
stand the consequences of early exposure to context-associated
moderators, it was necessary to continue tracking the
development of children in The Gambian cohort beyond
infancy and into the preschool period.
The preschool age follow-up (hereafter “BRIGHT Kids”)
has several key aims: (i) to assess a broader, age-appropriate
set of developmental outcomes; (ii) to further elucidate
developmental trajectories of the measures that had been admin-
istered since infancy; and (iii) to examine associations between
biomarkers of risk in infancy and outcomes at preschool age.
To examine these aims at the BRIGHT Kids follow-up, we
implemented a combination of existing and new measures,
described below. Prior to inviting the BRIGHT sample
for the follow-up, we recruited a separate pilot cohort of 24
participants to evaluate the implementation of the new
measures into the protocol.
Participant retention, sample size and age distribution
In 2021, all participants that had attended the 24-month visit
were invited to take part in the BRIGHT Kids follow-up.
Of these, 181 provided informed consent for participation.
Two families were traveling and could not be reached and
two declined to participate. A further three families with-
drew from BRIGHT Kids after consenting and six remained
in the study but did not attend the scheduled visits. Finally,
one child was withdrawn from the study due to evidence of
developmental delay (which interfered with their ability to
complete any of the assessments). This left a final sample of
171 (49.7% female) participants that took part in BRIGHT
Kids.
Table 5. Protocol for BRIGHT-Kids. Summary of abbreviations; fNIRS, functional near infrared
spectroscopy; EEG/ERP, electroencephalography/event related potentials; L – length, W – weight,
H – height, HC – head circumference, MUAC – mid to upper arm circumference, KHL – knee to heel
length. * Indicates a new measure introduced for BRIGHT-Kids.
Measure
Neuroimaging measures Questionnaires/Interviews – Infant/Child
fNIRS: Social/Non-social Food Insecurity Q (FFQ) *
fNIRS: Habituation and Novelty Detection Dietary Diversity Q (IFQ) *
fNIRS: Functional connectivity Early Childhood Development Index *
EEG: Auditory Oddball Questionnaires/Interviews – Family
Behavioural/Neurocognitive measures Generalised Anxiety Disorder – Maternal *
Eye-tracking: Cognitive control Patient Health Questionnaire – Maternal *
Eye-tracking: Gap/Overlap Perceived Stress Scale (PSS) – Maternal
Eye-tracking: Non-social contingency Impact of COVID-19 *
Eye-tracking: Face popout Family Care Indicators (FCI)
Eye-tracking: Dynamic scenes Clinical measures/ Medical details
Eye-tracking: Word-picture-matching Anthropometric measures – infant (L,W,HC,MUAC, KHL)
Mullen Scales of Early Learning (MSEL)
Tablet-based Early Years toolbox *
Test of Gross Motor Development *
Page 27 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
As a result of over recruitment in the early stages of the
BRIGHT project and interruptions to testing caused by a politi-
cal crisis in December 2016-January 2017 (which resulted in
the short-term closure of MRC Keneba), a recruitment pause
was needed in the summer of 2017. This recruitment pause
gave the number of infants born into the study per month a
bimodal distribution. Furthermore, the initial BRIGHT-Kids
phase of the BRIGHT project was conducted over a com-
pressed time window relative to the earlier longitudinal time
points of the BRIGHT project, due to funding and pandemic
related restrictions during 2021-2022. Children were assessed
starting with the oldest (first recruited) and ending with the
youngest (last recruited). Therefore, the ages of children
at follow-up reflected the same bimodal distribution as the
number of infants born per month. The age range of partici-
pants in Phase I of BRIGHT-Kids was 45–63 months (M=52.8,
SD=5.06); see Figure 10.
With the support of additional funding, due to this bimodal dis-
tribution of ages, a Phase II data collection of BRIGHT-Kids,
has now been established to collect data during 2023. This
additional follow-up is designed to assess the younger half
of the age distribution again when they are age matched with
the older participants from Phase I. The same BRIGHT-Kids
protocol is being utilised to allow for us to assess poten-
tial outcome and exposure effects by collecting data from all
participants when they reach 4 – 5 years.
Protocol for neuroimaging, ET, and MSEL
A number of paradigms that had previously been used in the
BRIGHT protocol were retained in BRIGHT Kids. These
included fNIRS tasks (social cognition, habituation and novelty
detection, functional connectivity), EEG, ET tasks and the
MSEL. The aforementioned assessments were either developed
using stimuli that was suitable for broader age ranges through-
out early childhood (fNIRS, EEG, ET) or had specific items
suitable for preschool-aged children (MSEL). Using consistent
paradigms enabled further tracking of developmental tra-
jectories, which is informative because it tells us whether a
particular biomarker is relevant at only one developmental/age
period, or if it continually predicts outcomes throughout
early childhood (Loth et al., 2017). While the neuroimag-
ing paradigms were suitable for the age group tested, several
practical considerations were required to ensure suitability of
the headgear.
fNIRS headgear. Data acquired at this visit were collected
using a headgear based on an EasyCap, rather than the cus-
tom-built headband that were used at prior visits. These caps are
made of soft fabric that covers the whole head. The optodes
were held in place with purpose-made holders for a secure
fit on the head and the design was changed to have a pointy,
rather than flat, end to better penetrate hair in older children.
Both of these changes allowed for better data acquisition,
particularly considering the variety of hair types at this
age.
Practical considerations. One challenge of undertaking
fNIRS and EEG with older children is that they have more
hair than infants, and children (girls in particular) often have
their hair tied or in braids (Katus et al., 2019). Several steps
were taken to maintain data quality and account for a vari-
ety of hair types. If participants had large braids, mothers were
asked to undo these before the study visit. When fitting the
fNIRS or EEG caps, research staff brushed aside stray hair
to ensure that the optodes/electrodes made contact with the
scalp. To acknowledge this additional burden on participating
families, and to thank them for their time, we hired an assist-
ant, who was local in West Kiang, to act as a “hairdresser”
for the participants. At the end of the assessments, she washed
out the EEG gel from the children’s hair and re-braided the hair
of every child (with caregiver consent).
Figure 10. Distribution of participant age (in days) in the BRIGHT Kids follow up.
Page 28 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Protocol for additional behavioural measures
Gross motor skills
The Test of Gross Motor Development 3rd Edition TGMD-3
(Ulrich, 2019) was used to assess gross motor (GM)
skills. The MSEL does not provide age-appropriate
test items, nor does it describe developmental GM milestones,
for children older than 39 months (Mullen, 1995). The TGMD-3
is suitable for children aged 3–11 years and assesses GM skills
across two domains: locomotor and ball skills. The locomo-
tor subscale tests skills that require coordinated and fluid
body movements (e.g., running, skipping, hopping). The ball
skills subscale evaluates proficiency in throwing, catching and
striking movements. Age-normed scores can be computed
for both scales, which are then combined to form the Gross
Motor composite score.
Tablet-based executive functioning assessments
Tablet based assessments from the Early Years Toolbox
(Howard & Melhuish, 2017) were used to assess executive
functions (EFs) across three domains: working memory
(“Mr Ant” task), Inhibitory Control (the Go/No-Go task) and
Cognitive Flexibility (Card Sorting task). These tasks were pre-
viously used among preschool aged children (3–5 years) in
South Africa, who demonstrated good EF abilities and, on some
scales, outperformed their Australian counterparts (Howard
et al., 2020). Task instructions were presented as an audio
playing from the tablet. These had been translated into Man-
dinka and recorded by field staff and embedded into the app.
While the tablet task was challenging at the earlier time
point, we found this tablet task to be successful in The Gambia
with our cohort at 3 – 5 years (Milosavljevic et al., 2023) .
Parent-report measures
We continued to use the Family Care Indicators (FCI, described
above) to assess play materials, enrichment activities and car-
egiving. In addition to this, a number of new questionnaires
were implemented to assess child development, caregiver
wellbeing and the home environment.
Child development
The Early Childhood Development Index (ECDI2030) is a
caregiver report questionnaire that assesses the achievement of
fundamental developmental milestones among children aged
24–59 months. The measure was developed by UNICEF
with the aim of creating a tool that could be both nationally
representative and also provide internationally comparable
data on child development (for full details, see the UNICEF
ECDI resources). The questionnaire consists of 20 items that
assess a range of skills related to a learning, health, and
psychosocial wellbeing. Since a Mandinka version was not avail-
able, we translated the tool following the customisation and
translation guidelines outlined by UNICEF.
Maternal mental health and caregiving
Maternal wellbeing has important implications for child well-
being and development beyond infancy and into school age
(Bennett et al., 2016; Kingston & Tough, 2014). Therefore,
we continued to assess maternal mental health at the BRIGHT
Kids visits. The Perceived Stress Scale (PSS, described above)
continued to be used, as this assessment is not specific to preg-
nancy and was, thus, deemed suitable for this time point. In
addition to this, two new assessments of depression and anxi-
ety that were designed to assess these mental health constructs
more generally and were not specific to pregnancy, were intro-
duced. The Generalised Anxiety Disorder-7 (Spitzer et al.,
2006) and the Patient Health Questionnaire-9 (PHQ-9; Kroenke
et al., 2001) were implemented to assess anxiety and depres-
sion, respectively. These measures were translated and adapted
following the same method as the previously used mental health
assessments. For the adaptation of these two measures, we col-
laborated with the PRECISE-DYAD study, who were assessing
maternal health in a different area of The Gambia (for further
details see PRECISE-DYAD website) to create a Mandinka
version of these measures that could be used in multiple
sites and studies. As with the previous measures, these were
administered as interviews. The two measures are described
in more detail below.
The GAD-7 is a seven-item self-report questionnaire that asks
respondents to report how often they experienced a range of
anxiety-related symptoms in the last two weeks. Responses
are rated on a Likert scale ranging from 0–3 (not at all-
nearly every day). The scale generates a total score, which
can range from 0–21. There is also one item assessing the
severity of symptoms, where respondents are asked to report
how much the symptoms that they reported add difficulty to
their daily lives on a scale of 0–3 (not at all to very difficult).
The GAD-7 has previously been used in Kenya (Nyongesa
et al., 2020), where it was reported to have good psychomet-
ric properties and the same factor structure as the original
English. For our translations, all of the items were deemed to
be appropriate for the local setting by the field staff working on
the translations.
The PHQ-9 is a nine-item self-report questionnaire that asks
respondents to rate how often they experienced a range of
depression-related symptoms in the last two weeks. Responses
are rated on a Likert scale ranging from 0–3 (not at all-nearly
every day). Scores can range from 0–27 and different ranges
indicate varying severity of depression (0–4 – none; 5–9 – mild;
10–14 – moderate; 15–19 – moderately severe; and 20–27 –
severe). Similar to the GAD-7, the PHQ-9 has one item assessing
severity of symptoms, where respondents are asked to rate
how much difficulty they experience in their daily lives as a
result of reported symptoms on a scale of 0–3 (not at all to very
difficult). Mothers who report elevated levels of depression
were offered an opportunity for referral to the on-site clinic,
as was done at previous visits. The field team found most of
the items on this scale suitable for the local population. How-
ever, as with the EPDS, there were concerns about asking
participants to report on suicidal thoughts. Thus, as for the
EPDS, this item was changed to asking participants how often
they wanted to be alone or isolated.
Impact of COVID-19
The BRIGHT Kids follow up was conducted during the COVID-
19 pandemic and, thus, it was vital to better understand how
the pandemic impacted on each family’s wellbeing, financial
Page 29 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
situation and access to healthcare. The questionnaire was
adapted from a measure that was being used in a UK study at
the time to assess the impact of the pandemic on families with
young children (Aydin et al., 2022). Items that were deemed
by field staff to be irrelevant or unsuitable for the West Kiang
community were removed. The final measure consisted of a
series of questions that asked the caregiver to report on how
their living and financial situation had changed due to the pan-
demic, whether they had family members who got sick with
COVID-19, their ability to access healthcare, and their general
feelings and concerns related to the situation.
Nutrition and child growth
Measures of child physical size, namely height, weight, and
head circumference, continued to be measured at this visit
(see above for details). In addition to this, measures of family
food insecurity and the child’s dietary diversity were also
implemented.
The Dietary Diversity Questionnaire (Swindale & Bilinsky,
2006)) is a caregiver report questionnaire that asks respond-
ents to indicate what the child eats on a typical day, from 12
food groups. There is also an option for “miscellaneous”. The
Food Insecurity questionnaire asks two sets of questions. The
first comprises of a single item that asks respondents if anybody
in the family has missed a meal due to food shortage in the last
seven days. Subsequently, there are seven items asking respond-
ents to indicate any food shortages in the household over
the previous month.
Discussion
Good scientic practice
We recognise the importance of maximising outputs from the
data collected in the BRIGHT project, both by serving the
participants and communities that have agreed to partake in
this research and the wider scientific community by provid-
ing access to the collected data for further analysis. Access
to any data collected during or generated by the BRIGHT
project is fully audited, and to ensure data security, is overseen
by the data management team in the UK and The Gambia.
While data sharing is critically important to maximising the ben-
efit of research, we must also consider the need to protect the
confidentiality of this sensitive group (particularly the infants
within the mother-infant dyads, who as minors do not con-
sent for themselves). Furthermore, to generate maximum value
from this dataset we must link data points together (i.e. NIRS/
EEG data with outcome data or contextual factor data). Due to
the nature of the data being collected (i.e. collected from a spe-
cific geographical location, longitudinal dataset of several
datapoints) the majority of the data cannot be fully de-identified
under the guidance included in the European General Data Pro-
tection Regulation (GDPR). Furthermore, some BRIGHT meas-
ures include photo, audio and video material, and therefore
is inherently identifiable and requires even stricter governance.
In line with other EU based studies of this nature (i.e. EUROSIBS,
(Jones et al., 2019)), currently, access to the data sets for the
wider scientific community is governed by the data manage-
ment team to ensure that users comply with all relevant data
protection laws and have appropriate ethical permissions.
Access operates via a project approval form, which is reviewed
by a committee consisting of representatives from each
cohort site, and implements relevant data sharing agreements.
Collaborations are encouraged, and projects are evaluated
primarily on their consistency with the ethical principles and
aims of the project that the families signed up to when par-
taking in this study. All planned analyses (both internal to the
BRIGHT team and external) are pre-specified either on an
internal database monitored by our management committee or
via web-based pre-registration platforms. These procedures
continue to be evaluated annually and updated to optimise
the BRIGHT Project’s value to the scientific community and
public priorities.
Dissemination of key findings has, and will continue, to
take place, through presentation of findings by the BRIGHT
research team at national and international conferences, inter-
nal reports and peer reviewed publications hosted on our
website, and through direct interaction with key stakeholders
and public engagement events.
Lessons learned
During the life course of the BRIGHT Project (and the pilot
phases that preceded this) we have encountered many chal-
lenges and opportunities that provide us with important guid-
ance for future work. Firstly, the importance for harmonisation of
goals, sense of purpose and motivation across the research team
and study sites should not be underestimated. Only through our
dedicated research team, which includes the 46 authors of this
paper, and countless more research midwives, clinical staff,
lab assistants, village community assistants, drivers, mechan-
ics, (even a hairdresser for the children’s hair) and research
centre staff have we managed to succeed in achieving what
we have thus far in the BRIGHT Project. In particular in The
Gambia, staff coordination was critical while we managed the
study workload across the time course of the study with the
research team working on rolling interleaved working hours so
that the participants could be tested seven days a week during
the peak period of longitudinal timepoints. Furthermore, a
strength of our study came in the establishment of rigorous
SOPs – adapted for each site, age point and measure obtained
– and overseen by the quality control and data handling proce-
dures and regular meetings outlined above. The pilot phase of
our project enabled us to identify areas where sites differed and
discuss with the research team at each site how to adapt best to
the contextual factors that were relevant. An important area
was to operationalise our management of infant behaviour to
ensure successful collection of neuroimaging and behavioural
infant measures. We broadly harmonised these approaches
and used a series of sequential “attention getters” within task,
and calming strategies (such as a snack, break or comfort-
ing social strategies) to maintain successful data collection,
with levels of attrition being broadly similar across sites and
measures. However, there were some differences noted across
sites. In The Gambia for example, the climate required us to
more carefully monitor environmental conditions, a signifi-
cant rise in temperature could lead the infant to become drowsy
Page 30 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
and fall asleep during the session at a more frequent rate at
some age points relative to the UK. In the UK, in contrast, we
found that age related changes in the behaviour of individuals
led to some losing attention more frequently and moving around
or leaving their parent’s lap at older ages relative to younger.
While in The Gambia this age-related behavioural shift was
less evident. It is possible that some of these site differ-
ences exist due to fluctuations in the testing practice of the
research team, or differences in caregiving approaches or infant
motivational behaviours that led to differences in perform-
ance during the sessions, or differences in other contextual
factors of the sample at each site (for example the proportion
of infants and families having a stressful or long journey to
the research site). It is hard to tease apart these differences
systematically, but our approach to data analysis will include
provision for these contextual factors. It should also be noted
that heterogeneity can be a strength, and individual variance
within sites as well as across sites can be used to further
our understanding of infant and early childhood develop-
ment. As noted by the EUROSIBS consortium, “multisite
studies need to consider balancing the resources put into
standardisation with the desirability of testing whether metrics
are robust across natural variation between sites” (Jones et al.,
2019).
Building on this, a strength of our project comes in our approach
to incorporating contextual information in the design, devel-
opment and adaptation of measures to different contexts.
We were faced with a need to adapt measures to be broadly
applicable across different cultural and linguistic contexts,
while also ensuring we designed site specific measures when
appropriate. While many of these measures had been used
extensively in the UK, and other HICs, not all of these had
been used across all the age points of the BRIGHT protocol,
so piloting work focused on ensuring that the questionnaires
and paradigms were relevant across the developmental time
frame that we are studying at both sites. Furthermore, in the
Gambia, our approach was to ensure Gambian research staff
co-produced the necessary adaptation of measures and assessed
the appropriateness of tools in a context that they had not been
used previously to take advantage of the local expertise at each
site. Furthermore, staff in The Gambia are also involved in data
processing and quality control to ensure that, when possible, our
training strategies incorporate not just data collection knowl-
edge transfer but also steps towards transferring knowledge on
data analysis and dissemination. The latter is certainly an area
for future development, as this area of research (developmental
psychology/neuroscience) is relatively new to the country and
so capacity building is a key priority for our group. While we
strive to be sensitive to cultural context, a limitation of some
of the measures used in this project is that they are drawn from
a pool of existing developmental measures of cognitive out-
comes which have largely been derived from tasks developed for
high-income settings. While we went to lengths to adapt these
measures, we cannot fully eliminate possibility of cultural
bias. Equally, these measures likely overlook some skills that
are relevant in a rural, farming community as many measures
have been designed in urban contexts, therefore we cannot
be certain that we are measuring all of the crucial predictors
of SES (Hermida et al., 2019). For example, some studies
have defined preschool attendance and access to services as a
better predictor of cognitive outcomes than income. In focuss-
ing on poverty initially, we also overlook many of the strengths
of this community and culture, which are highly relevant.
While we strive to understand potential protective factors in
this environment, we may not have fully captured the complex
family dynamics that exist, and will focus on furthering our
understanding in future work.
A final strength of the BRIGHT project comes from the
diversity of measures incorporated in the cohort design and the
density of time points. The spacing of study visits and the range
of cognitive, brain and contextual measures employed means
that we are able to more accurately pinpoint when certain skills
emerge and when they start to become relevant for later
outcomes. Likewise, we are also able to investigate whether
a particular marker is relevant throughout childhood or only
during a specific age/time in development.
Future eorts
The ultimate objective of the next phase of the BRIGHT project
is the identification and validation of a marker, or “fingerprint”
combination of brain function markers, that predict the contri-
bution of exposure phenotypes (i.e., undernutrition / caregiving
context) to the substantive variation in developmental out-
comes seen in infants born into a low-income setting, such
as rural Gambia. Innovative modelling and analytic approaches
will be applied to allow us to address the following aims: (1) To
employ feature extraction and time varying parameter modelling
to identify the most reliable “neural fingerprinting” biomarkers
(from fNIRS and EEG) of longitudinal developmental cortical
specialization; (2) To create a profile of metrics across develop-
mental outcomes in later childhood (executive function, language,
general cognitive development, adaptive skills) at the group
level and identify latent classes that sub-group individuals into
developmental profiles of pre-academic skills; (3) To
expand our predictive models of longitudinal fNIRS and EEG
trajectories of early developing brain networks (0–2 years) to
determine which age points and “fingerprint” of biomarkers are
the most consistent and reliable predictors of developmental out-
comes of language, pre-academic skills and functional brain
specialization in later childhood (3–5 years); (4) To model expo-
sure phenotypes and establish health, social and environmental
risk and resilience determinants of developmental trajectories
from birth to preschool which associate with early biomarkers
of development, to identify primary targets of intervention.
Conclusions
The BRIGHT project is a comprehensive, and multi-method
study of development, during the first two years of life in
the UK and the first five years in The Gambia. The combina-
tion of neuroimaging, behavioural, parent-report measures,
and biological samples provides a unique opportunity to study
a variety of context associated moderators, which may, or
may not be related to poverty, as well as the mechanistic proc-
esses underlying associations between markers and outcomes.
Page 31 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
On one hand, we hope that our work will be an asset to global
health research, where the study of neurocognitive develop-
ment in early infancy, particularly with the use of neuroimag-
ing tools, is still emergent. On the other hand, this work has
broader value for developmental research in general. Due to
the logistical and financial constraints of longitudinal research,
our project is among the few to assess early development,
across a large number of study visits, and that incorporates
such a variety of methods. Thus, the generated results will
enable us to identify critical windows for developmental vul-
nerability and act as rationale to guide future interventions
which aim to protect and enrich the developing brain within
context associated risk contexts. We propose that our project
provides a roadmap for other researchers interested in con-
ducting studies of neurocognitive development in LMICs
with similar contextual factors.
Data availability
Underlying data
No data are associated with this article.
Extended data
We recognise the importance of maximising outputs from the
data collected in the BRIGHT project, both by serving the par-
ticipants and communities that have agreed to partake in this
research and the wider scientific community by providing access
to the collected data for further analysis. Access to any data
collected during or generated by the BRIGHT project is fully
audited, and to ensure data security, is overseen by the data
management team in the UK and The Gambia. While data shar-
ing is critically important to maximising the benefit of research,
we must also consider the need to protect the confidentiality of
this sensitive group (particularly the infants within the mother-
infant dyads, who as minors do not consent for themselves).
Furthermore, to generate maximum value from this dataset we
must link data points together (i.e. NIRS/EEG data with out-
come data or contextual factor data). Due to the nature of the
data being collected (i.e. collected from a specific
geographical location, longitudinal dataset of several datapoints)
the majority of the data cannot be fully de-identified under the
guidance included in the European General Data Protection
Regulation (GDPR).
The data used to support this study are stored in the Brain
Imaging for Global Health Data Repository. The conditions
of our ethics approval do not allow public archiving of pseu-
donymised study data. The data cannot be fully anonymized
due to the nature of combined sources of information, such
as neuroimaging, sociodemographic, geographic and health
measures, making it possible to attribute data to specific
individuals, and hence, falling under personal information, the
release of which would not be compliant with GDPR guide-
lines unless additional participant consent forms are completed.
Our data sharing procedures were created in consultation with
stakeholders and external consultation (Begum-Ali et al.,
2023).
Collaborations are encouraged, and projects are evaluated pri-
marily on their consistency with the ethical principles and aims
of the project that the families signed up to when partaking in
this study. All planned analyses (both internal to the BRIGHT
team and external) are pre-specified either on an internal
database monitored by our management committee or via
web-based pre-registration platforms. These procedures con-
tinue to be evaluated annually and updated to optimise the
BRIGHT Project’s value to the scientific community and pub-
lic priorities. To access the data, interested readers should
contact the BRIGHT coordinator on the Contact page of our
website. Access will be granted to named individuals follow-
ing ethical procedures governing the reuse of sensitive data.
Specifically, requestors must pre-register their proposal, and
clearly explain the purpose of the analysis so as to ensure that
the purpose and nature of the research is consistent with that
to which participating families originally consented. Addi-
tionally, requestors must complete and sign a data sharing
agreement to ensure data is stored securely. Approved projects
would need to adhere to the BRIGHT project’s policies
on Ethics, Data Sharing, Authorship and Publication.
Legal copyright restrictions do not permit us to publicly archive
the full set of behavioural tests and task paradigms used in
this experiment. Readers seeking access to these tests are
advised to contact the lead author or the reference list.
No part of the study procedures or analysis plans was
preregistered prior to the research being conducted.
Contributor Role Role Denition Authors
Conceptualization Ideas; formulation or evolution of overarching research goals and
aims.
Main Authors: SLF, BM, SEM, CEE
BRIGHT Study Team: TA, MDH, AP
Data Curation Management activities to annotate (produce metadata), scrub data
and maintain research data (including software code, where it is
necessary for interpreting the data itself) for initial use and later
reuse.
Main Authors: BM, SMC, LK, AB, MCL,
GG, LM, MP, CB, EM, ET
BRIGHT Study Team:
Formal Analysis Application of statistical, mathematical, computational, or other
formal techniques to analyze or synthesize study data.
Main Authors: SLF, BM, SMC, LK, AB,
MCL, GG, LM, CB, MS, ON
BRIGHT Study Team:
Page 32 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Contributor Role Role Denition Authors
Funding Acquisition Acquisition of the nancial support for the project leading to this
publication.
Main Authors: SLF, BM, LK, SEM, CEE
BRIGHT Study Team: TA, MDH, AP
Investigation Conducting a research and investigation process, specically
performing the experiments, or data/evidence collection.
Main Authors: SMC, BM, LK, SLF, AB,
MCL, GG, TF, EM, FN, ON, MP, MR, FS,
MS, ET
BRIGHT Study Team: LA, ABa, CBT, SB,
DC, MC, YD, NH, ED, SDa, SD, AJ, SJ, BJ,
MJ, OK, KK, JL, LS, LSt
Methodology Development or design of methodology; creation of models. Main Authors: SLF, BM, LK, AB, MCL,
LM, SEM, MR
BRIGHT Study Team: CBT, MDH, JL
Project Administration Management and coordination responsibility for the research
activity planning and execution.
Main Authors: SLF, BM, SMC, MCL, EM,
MP, MR, ET, SEM, CEE
BRIGHT Study Team: LA, TA, SB, NH,
SD, PN
Resources Provision of study materials, reagents, materials, patients,
laboratory samples, animals, instrumentation, computing resources,
or other analysis tools.
Main Authors: SLF, CEE, SM,
BRIGHT Study Team: TA, EC, AF
Software Programming, software development; designing computer
programs; implementation of the computer code and supporting
algorithms; testing of existing code components.
Main Authors: AB, LM, CB, BM, SMC,
LK
BRIGHT Study Team:
Supervision Oversight and leadership responsibility for the research activity
planning and execution, including mentorship external to the core
team.
Main Authors: SLF, BM, SMC, LK, AB,
CB, MCL, EM, LM, MR, SEM, CEE
BRIGHT Study Team: LA, CBT, SB,
MDH, SD
Validation Verication, whether as a part of the activity or separate, of the
overall replication/reproducibility of results/experiments and other
research outputs.
Main Authors: SLF, AB, BM, MCL, CB,
SCB,
BRIGHT Study Team:
Visualization Preparation, creation and/or presentation of the published work,
specically visualization/data presentation.
Main Authors: SLF, BM, SMC, LK
BRIGHT Study Team:
Writing – Original Draft
Preparation
Creation and/or presentation of the published work, specically
writing the initial draft (including substantive translation).
Main Authors: SLF, BM, SMC, LK
BRIGHT Study Team:
Writing – Review and
Editing
Preparation, creation and/or presentation of the published work by
those from the original research group, specically critical review,
commentary or revision – including pre- or post-publication stages.
Main Authors: SLF, BM, SMC, LK, AB,
CB, MCL, GG, TJ, EM, LM, FN, ON, MP,
MR, FS, MS, ET, SEM, CEE
BRIGHT Study Team: LA, TA, ABarjo,
CBT, SB, DC, MC, EC, YD, MDH, AF, NH,
ED, SD, AJ, SJ, BJ, MJ, OK, KK, AP, PN, LS,
LStein
Author contributions
Main authors: S. Lloyd-Fox*, S. McCann*, B. Milosavljevic*,
L. Katus*, A. Blasi, C. Bulgarelli, M. Crespo-Llado, G.
Ghillia, T. Fadera, E. Mbye, L. Mason, F. Njai, O. Njie,
M. Perapoch-Amado, M. Rozhko, F. Sosseh, M. Saidykhan,
E. Touray, S.E. Moore^, C.E. Elwell^ and the BRIGHT Study Team
BRIGHT Study Team: L. Acolatse, T. Austin, A. Barjo,
S.C. Bartram, S. Budge, D. Camara, M. Ceesay, E. Comma,
Y. Dampha, S. Darboe, M. de Haan, A. Faal, N. Hayes, E. Drammeh,
S. Drammeh, A. Janneh, S. Jarju, B. Jobarteh, M. Joof, O. Kambi,
K. Kora, J. Larrieta, A. Prentice, P. Nshe, L. Sanyang,
L. Steiner.
Page 33 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
References
Alam S: Majority World: Challenging the West’s Rhetoric of Democracy.
Amerasia Journal. 2008; 34(1): 87–98.
Publisher Full Text
Alarcão FSP, Shephard E, Fatori D, et al.: Promoting mother-infant
relationships and underlying neural correlates: Results from a randomized
controlled trial of a home-visiting program for adolescent mothers in
Brazil. Dev Sci. 2021; 24(6): e13113.
PubMed Abstract | Publisher Full Text | Free Full Text
Alderman H, Behrman JR, Grantham-McGregor S, et al.: Economic perspectives
on integrating early child stimulation with nutritional interventions. Ann N
Y Acad Sci. 2014; 1308(1): 129–138.
PubMed Abstract | Publisher Full Text
Ayala A, Christensson K, Christensson E, et al.: Newborn infants who received
skin-to-skin contact with fathers after Caesarean sections showed stable
physiological patterns. Acta Paediatr. 2021; 110(5): 1461–1467.
PubMed Abstract | Publisher Full Text | Free Full Text
Aydin E, Weiss SM, Glasgow KA, et al.: COVID-19 in the context of pregnancy,
infancy and parenting (CoCoPIP) study: Protocol for a longitudinal study of
parental mental health, social interactions, physical growth and cognitive
development of infants during the pandemic. BMJ Open. 2022; 12(6):
e053800.
PubMed Abstract | Publisher Full Text | Free Full Text
Barch D, Pagliaccio D, Belden A, et al.: Eect of Hippocampal and Amygdala
Connectivity on the Relationship Between Preschool Poverty and School-
Age Depression. Am J Psychiatry. 2016; 173(6): 625–634.
PubMed Abstract | Publisher Full Text | Free Full Text
Bartram SC: Early infant social behaviour and maternal perinatal
depression in Kiang West, The Gambia. [Phd, University of Warwick], 2018.
Reference Source
Begum-Ali JB, Holman R, Goodwin A, et al.: Parent attitudes towards data
sharing in developmental science. PsyArXiv, 2023.
Publisher Full Text
Begus K, Lloyd-Fox S, Halliday D, et al.: Using fNIRS to Study Working Memory
of Infants in Rural Africa. Adv Exp Med Biol. 2016; 876: 273–279.
PubMed Abstract | Publisher Full Text
Bennett IM, Schott W, Krutikova S, et al.: Maternal mental health, and
child growth and development, in four low-income and middle-income
countries. J Epidemiol Community Health. 2016; 70(2): 168–173.
PubMed Abstract | Publisher Full Text | Free Full Text
Berens AE, Jensen SKG, Nelson CA 3rd: Biological embedding of childhood
adversity: From physiological mechanisms to clinical implications. BMC
Med. 2017; 15(1): 135.
PubMed Abstract | Publisher Full Text | Free Full Text
Berkes J, Raikes A, Bouguen A, et al.: Joint roles of parenting and nutritional
status for child development: Evidence from rural Cambodia. Dev Sci. 2019;
22(5): e12874.
PubMed Abstract | Publisher Full Text
Blasi A, Lloyd-Fox S, Johnson MH, et al.: Test-retest reliability of functional
near infrared spectroscopy in infants. Neurophotonics. 2014; 1(2): 025005.
PubMed Abstract | Publisher Full Text | Free Full Text
Blasi A, Lloyd-Fox S, Katus L, et al.: FNIRS for Tracking Brain Development in
the Context of Global Health Projects. Photonics. 2019; 6(3): 89.
PubMed Abstract | Publisher Full Text | Free Full Text
Blimpo M, Carneiro P, Jervis P, et al.: Improving Access and Quality in Early
Childhood Development Programs: Experimental Evidence from the
Gambia. SSRN Electronic Journal. 2019.
Publisher Full Text
Bozicevic L, De Pascalis L, Schuitmaker N, et al.: Longitudinal Association
between Child Emotion Regulation and Aggression, and the Role of
Parenting: A Comparison of Three Cultures. Psychopathology. 2016; 49(4):
228–235.
PubMed Abstract | Publisher Full Text | Free Full Text
Brazelton TB, Koslowski B, Tronick E: Neonatal behavior among urban
Zambians and Americans. J Am Acad Child Psychiatry. 1976; 15(1): 97.
PubMed Abstract | Publisher Full Text
Brazelton TB, Nugent JK: Neonatal behavioral assessment scale. Cambridge
University Press, 1995.
Brotherton H, Daly M, Johm P, et al.: “We All Join Hands”: Perceptions of the
Kangaroo Method Among Female Relatives of Newborns in The Gambia.
Qual Health Res. 2021; 31(4): 665–676.
PubMed Abstract | Publisher Full Text | Free Full Text
Busch T, Sangen A, Vanpoucke F, et al.: Correlation and agreement between
Language ENvironment Analysis (lenaTM) and manual transcription
for Dutch natural language recordings. Behav Res Methods. 2018; 50(5):
1921–1932.
PubMed Abstract | Publisher Full Text
Caldwell BM, Bradley RH: Home Observation for Measurement of the
Environment (HOME)-revised edition. University of Arkansas, Little Rock,
1984.
Reference Source
Cambridge County Council: Cambridge City annual demographic and socio-
economic report. 2011.
Reference Source
Canault M, Le Normand MT, Foudil S, et al.: Reliability of the Language
ENvironment Analysis system (LENATM) in European French. Behav Res
Methods. 2016; 48(3): 1109–1124.
PubMed Abstract | Publisher Full Text
Cattan S, Conti G, Farquharson C, et al.: The Health Eects of Universal Early
Childhood Interventions: Evidence from Sure Start.2021; 82.
Reference Source
Cavallera V, Lancaster G, Gladstone M, et al.: Protocol for validation of the
Global Scales for Early Development (GSED) for children under 3 years of
age in seven countries. BMJ Open. 2023; 13(1): e062562.
PubMed Abstract | Publisher Full Text | Free Full Text
CEIC data: CEICdata.Com. 2018.
Reference Source
Cho A: Economics Nobel honors trio taking an experimental approach to
ghting poverty. Science. 2019.
Publisher Full Text
Cohen S, Kamarck T, Mermelstein R: A GLOBAL MEASURE OF PERCEIVED
STRESS. J Health Soc Behav. 1983; 24(4): 385–96.
PubMed Abstract
Coleman R, Morison L, Paine K, et al.: Women’s reproductive health and
depression: a community survey in the Gambia, West Africa. Soc Psychiatry
Psychiatr Epidemiol. 2006; 41(9): 720–727.
PubMed Abstract | Publisher Full Text
Collins-Jones LH, Cooper RJ, Bulgarelli C, et al.: Longitudinal infant fNIRS
channel-space analyses are robust to variability parameters at the
group-level: An image reconstruction investigation. NeuroImage. 2021;
237: 118068.
PubMed Abstract | Publisher Full Text | Free Full Text
Cox JL, Holden JM, Sagovsky R: Detection of Postnatal Depression.
Development of the 10-item Edinburgh Postnatal Depression Scale. Br J
Psychiatry. 1987; 150(6): 782–786.
PubMed Abstract | Publisher Full Text
Development Initiatives: 2018 Global Nutrition Report: Shining a light to
spur action on nutrition. Bristol. 2018.
Reference Source
Draper CE, Barnett LM, Cook CJ, et al.: Publishing child development research
from around the world: An unfair playing eld resulting in most of the
world’s child population under-represented in research. Infant Child Dev. 2022.
Publisher Full Text
Draper CE, Klingberg S, Wrottesley SV, et al.: Interventions to promote
development in the next 1000 days: A mapping review. Child Care Health Dev.
2023; 49(4): 617–629.
PubMed Abstract | Publisher Full Text
Dzhelyova M, Jacques C, Dormal G, et al.: High test-retest reliability of a
neural index of rapid automatic discrimination of unfamiliar individual
faces. Vis Cogn. 2019; 27(2): 127–141.
Publisher Full Text
Acknowledgements
We are indebted to the families and communities that wel-
come us into their homes, and attend visits to our research sites
to generate the data for the BRIGHT Project. Furthermore
we would like to make special mention to Dr Momodou
Darboe, who was the lead Co-Investigator in Keneba at
LSHTM @ MRCG, and contributed to the application for the
grants that have funded this work, and to supporting the team
at this local site. He died in 2021 before he was able to see
the conclusion of this research project.
Page 34 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
East P, Lozo B, Blanco E, et al.: Infant Iron Deciency, Child Aect, and
Maternal Unresponsiveness: Testing the Long-Term Eects of Functional
Isolation. Dev Psychol. 2017; 53(12): 2233–2244.
PubMed Abstract | Publisher Full Text | Free Full Text
Elsabbagh M, Fernandes J, Webb SJ, et al.: Disengagement of visual attention
in infancy is associated with emerging autism in toddlerhood. Biol
Psychiatry. 2013a; 74(3): 189–194.
PubMed Abstract | Publisher Full Text | Free Full Text
Elsabbagh M, Gliga T, Pickles A, et al.: The development of face orienting
mechanisms in infants at-risk for autism. Behav Brain Res. 2013b; 251:
147–154.
PubMed Abstract | Publisher Full Text | Free Full Text
Elsabbagh M, Volein A, Holmboe K, et al.: Visual orienting in the early
broader autism phenotype: Disengagement and facilitation. J Child Psychol
Psychiatry. 2009; 50(5): 637–642.
PubMed Abstract | Publisher Full Text | Free Full Text
Farah MJ: The Neuroscience of Socioeconomic Status: Correlates, Causes,
and Consequences. Neuron. 2017; 96(1): 56–71.
PubMed Abstract | Publisher Full Text
Farah MJ, Shera DM, Savage JH, et al.: Childhood poverty: Specic
associations with neurocognitive development. Brain Res. 2006; 1110(1):
166–174.
PubMed Abstract | Publisher Full Text
Fenson L, Marchman VA, Thal DJ, et al.: MacArthur-Bates Communicative
Development Inventories: User’s guide and technical manual (2nd ed.).
Brookes, 2007.
Reference Source
Fernald A, Zangl R, Portillo AL, et al.: Looking while listening: Using eye
movements to monitor spoken language comprehension by infants and
young children. In: Developmental Psycholinguistics: On-line Methods in Children’s
Language Processing. John Benjamins Publishing Company, 2008; 97–135.
Publisher Full Text
Fishell AK, Arbeláez AM, Valdés CP, et al.: Portable, eld-based neuroimaging
using high-density diuse optical tomography. NeuroImage. 2020; 215:
116541.
PubMed Abstract | Publisher Full Text
Fitt E, Cole D, Ziauddeen N, et al.: DINO (Diet In Nutrients Out) - an integrated
dietary assessment system. Public Health Nutr. 2015; 18(2): 234–241.
PubMed Abstract | Publisher Full Text | Free Full Text
Forssman L, Ashorn P, Ashorn U, et al.: Eye-tracking-based assessment of
cognitive function in low-resource settings. Arch Dis Child. 2017; 102(4):
301–302.
PubMed Abstract | Publisher Full Text | Free Full Text
Ganek HV, Eriks-Brophy A: A Concise Protocol for the Validation of Language
ENvironment Analysis (LENA) Conversational Turn Counts in Vietnamese.
Comm Disord Q. 2018; 39(2): 371–380.
Publisher Full Text
Gervain J, Minagawa Y, Emberson L, et al.: Using functional near-infrared
spectroscopy to study the early developing brain: Future directions and
new challenges. Neurophotonics. 2023; 10(2): 023519.
PubMed Abstract | Publisher Full Text | Free Full Text
Gilkerson J, Richards JA: The LENA Natural Language Study. (LENA Technical
Report LTR-02-2). LENA Foundation, 2008; 1–26.
Reference Source
Giovanelli A, Reynolds AJ, Mondi CF, et al.: Adverse Childhood Experiences and
Adult Well-Being in a Low-income, Urban Cohort. Pediatrics. 2016; 137(4):
e20154016.
PubMed Abstract | Publisher Full Text | Free Full Text
Hackman DA, Farah MJ: Socioeconomic status and the developing brain.
Trends Cogn Sci. 2009; 13(2): 65–73.
PubMed Abstract | Publisher Full Text | Free Full Text
Hair NL, Hanson JL, Wolfe BL, et al.: Association of Child Poverty, Brain
Development, and Academic Achievement. JAMA Pediatr. 2015; 169(9):
822–829.
PubMed Abstract | Publisher Full Text | Free Full Text
Hanlon C, Medhin G, Alem A, et al.: Detecting perinatal common mental
disorders in Ethiopia: Validation of the self-reporting questionnaire and
Edinburgh Postnatal Depression Scale. J Aect Disord. 2008; 108(3): 251–262.
PubMed Abstract | Publisher Full Text
Hennig BJ, Unger SA, Dondeh BL, et al.: Cohort Prole: The Kiang West
Longitudinal Population Study (KWLPS)-a platform for integrated research
and health care provision in rural Gambia. Int J Epidemiol. 2017; 46(2): e13.
PubMed Abstract | Publisher Full Text | Free Full Text
Henrich J, Heine SJ, Norenzayan A: Beyond WEIRD: Towards a Broad-based
Behavioral Science. Behav Brain Sci. 2010; 33(2–3): 111–135.
Publisher Full Text
Hensch TK: Critical period mechanisms in developing visual cortex. Curr Top
Dev Biol. 2005; 69: 215–237.
PubMed Abstract | Publisher Full Text
Hermida MJ, Shalom DE, Segretin MS, et al.: Risks for Child Cognitive
Development in Rural Contexts. Front Psychol. 2019; 9: 2735.
PubMed Abstract | Publisher Full Text | Free Full Text
Howard SJ, Cook CJ, Everts L, et al.: Challenging socioeconomic status: A
cross-cultural comparison of early executive function. Dev Sci. 2020; 23(1):
e12854.
PubMed Abstract | Publisher Full Text
Howard SJ, Melhuish E: An Early Years Toolbox for Assessing Early Executive
Function, Language, Self-Regulation, and Social Development: Validity,
Reliability, and Preliminary Norms. J Psychoeduc Assess. 2017; 35(3): 255–275.
PubMed Abstract | Publisher Full Text | Free Full Text
Hurt H, Betancourt LM: Eect of socioeconomic status disparity on child
language and neural outcome: How early is early? Pediatr Res. 2016; 79(1–2):
148–158.
PubMed Abstract | Publisher Full Text
Isaacs EB: Neuroimaging, a new tool for investigating the eects of early
diet on cognitive and brain development. Front Hum Neurosci. 2013; 7: 445.
PubMed Abstract | Publisher Full Text | Free Full Text
Isaacs E, Oates J, ILSI Europe a.i.s.b.l: Nutrition and cognition: Assessing
cognitive abilities in children and young people. Eur J Nutr. 2008; 47 Suppl
3(3): 4–24.
PubMed Abstract | Publisher Full Text
Jasińska KK, Guei S: Neuroimaging Field Methods Using Functional Near
Infrared Spectroscopy (NIRS) Neuroimaging to Study Global Child
Development: Rural Sub-Saharan Africa. J Vis Exp. 2018; (132): 57165.
PubMed Abstract | Publisher Full Text | Free Full Text
Jensen SKG, Berens AE, Nelson CA: Eects of poverty on interacting
biological systems underlying child development. Lancet Child Adolesc
Health. 2017; 1(3): 225–239.
PubMed Abstract | Publisher Full Text
Jensen SKG, Tofail F, Haque R, et al.: Child development in the context of
biological and psychosocial hazards among poor families in Bangladesh.
PLoS One. 2019; 14(5): e0215304.
PubMed Abstract | Publisher Full Text | Free Full Text
Johnson MH: The inhibition of automatic saccades in early infancy. Dev
Psychobiol. 1995; 28(5): 281–291.
PubMed Abstract | Publisher Full Text
Jones EJH, Mason L, Begum Ali J, et al.: Eurosibs: Towards robust
measurement of infant neurocognitive predictors of autism across Europe.
Infant Behav Dev. 2019; 57: 101316.
PubMed Abstract | Publisher Full Text | Free Full Text
Kaldy Z, Kraper C, Carter AS, et al.: Toddlers with Autism Spectrum Disorder
are more successful at visual search than typically developing toddlers.
Dev Sci. 2011; 14(5): 980–988.
PubMed Abstract | Publisher Full Text | Free Full Text
Kariger P, Frongillo EA, Engle P, et al.: Indicators of Family Care for
Development for Use in Multicountry Surveys. J Health Popul Nutr. 2012;
30(4): 472–486.
PubMed Abstract | Publisher Full Text | Free Full Text
Katus L, Blasi A, McCann S, et al.: Longitudinal fNIRS and EEG metrics of
habituation and novelty detection are correlated in 1-18-month-old
infants. Neuroimage. 2023; 274: 120153.
PubMed Abstract | Publisher Full Text | Free Full Text
Katus L, Foley S, Murray AL, et al.: Perceived stress during the prenatal
period: Assessing measurement invariance of the Perceived Stress Scale
(PSS-10) across cultures and birth parity. Arch Womens Ment Health. 2022a;
25(3): 633–640.
PubMed Abstract | Publisher Full Text | Free Full Text
Katus L, Hayes NJ, Mason L, et al.: Implementing neuroimaging and eye
tracking methods to assess neurocognitive development of young infants
in low- and middle-income countries [version 2; peer review: 2 approved].
Gates Open Res. 2019; 3: 1113.
PubMed Abstract | Publisher Full Text | Free Full Text
Katus L, Mason L, Milosavljevic B, et al.: ERP markers are associated with
neurodevelopmental outcomes in 1-5 month old infants in rural Africa and
the UK. Neuroimage. 2020; 210: 116591.
PubMed Abstract | Publisher Full Text | Free Full Text
Katus L, Milosavljevic B, Rozhko M, et al.: Neural Marker of Habituation at
5 Months of Age Associated with Deferred Imitation Performance at 12
Months: A Longitudinal Study in the UK and The Gambia. Children (Basel).
2022b; 9(7): 988.
PubMed Abstract | Publisher Full Text | Free Full Text
Kea P: ‘The complexity of an enduring relationship’: Gender, generation,
and the moral economy of the Gambian Mandinka household. J R Anthropol
Inst. 2013; 19(1): 102–119.
Publisher Full Text
Kea P: Gambian educational migration, care and the persistence of the
domestic moral economy. Child Geogr. 2020; 18(6): 642–653.
Publisher Full Text
Kingston D, Tough S: Prenatal and Postnatal Maternal Mental Health and
School-Age Child Development: A Systematic Review. Matern Child Health J.
2014; 18(7): 1728–1741.
PubMed Abstract | Publisher Full Text
Kohrt BA, Luitel NP, Acharya P, et al.: Detection of depression in low resource
settings: Validation of the Patient Health Questionnaire (PHQ-9) and
cultural concepts of distress in Nepal. BMC Psychiatry. 2016; 16(1): 58.
PubMed Abstract | Publisher Full Text | Free Full Text
Page 35 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Kroenke K, Spitzer RL, Williams JBW: The PHQ-9: Validity of a brief depression
severity measure. J Gen Intern Med. 2001; 16(9): 606–613.
PubMed Abstract | Publisher Full Text | Free Full Text
Lloyd-Fox S, Begus K, Halliday D, et al.: Cortical specialisation to social stimuli
from the rst days to the second year of life: A rural Gambian cohort. Dev
Cogn Neurosci. 2017; 25: 92–104.
PubMed Abstract | Publisher Full Text | Free Full Text
Lloyd-Fox S, Blasi A, Elwell CE: Illuminating the developing brain: The past,
present and future of functional near infrared spectroscopy. Neurosci
Biobehav Rev. 2010; 34(3): 269–284.
PubMed Abstract | Publisher Full Text
Lloyd‐Fox S, Blasi A, McCann S, et al.: Habituation and novelty detection
fNIRS brain responses in 5- and 8-month-old infants: The Gambia and UK.
Dev Sci. 2019; 22(5): e12817.
PubMed Abstract | Publisher Full Text | Free Full Text
Lloyd-Fox S, Papademetriou M, Darboe MK, et al.: Functional near infrared
spectroscopy (fNIRS) to assess cognitive function in infants in rural Africa.
Sci Rep. 2014; 4: 4740.
PubMed Abstract | Publisher Full Text | Free Full Text
Loth E, Charman T, Mason L, et al.: The EU-AIMS Longitudinal European
Autism Project (LEAP): Design and methodologies to identify and validate
stratication biomarkers for autism spectrum disorders. Mol Autism. 2017;
8(1): 24.
PubMed Abstract | Publisher Full Text | Free Full Text
Luby J, Belden A, Botteron K, et al.: The Eects of Poverty on Childhood Brain
Development: The Mediating Eect of Caregiving and Stressful Life Events.
JAMA Pediatr. 2013; 167(12): 1135–1142.
PubMed Abstract | Publisher Full Text | Free Full Text
Macrae E, Milosavljevic B, Mason L, et al.: Executive functions in infancy:
Measurement using a novel tablet task and exploration of longitudinal
attentional and cognitive predictors. [Preprint]. PsyArXiv. 2022.
Publisher Full Text
Martorell R, Horta BL, Adair LS, et al.: Weight Gain in the First Two Years of
Life Is an Important Predictor of Schooling Outcomes in Pooled Analyses
from Five Birth Cohorts from Low- and Middle-Income Countries. J Nutr.
2010; 140(2): 348–354.
PubMed Abstract | Publisher Full Text | Free Full Text
McCoy DC, Peet ED, Ezzati M, et al.: Early Childhood Developmental Status
in Low- and Middle-Income Countries: National, Regional, and Global
Prevalence Estimates Using Predictive Modeling. PLoS Med. 2016; 13(6):
e1002034.
PubMed Abstract | Publisher Full Text | Free Full Text
Milosavljevic B, Cook CJ, Fadera T, et al.: Executive functioning skills and their
environmental predictors among pre‐school aged children in South Africa
and The Gambia. Dev Sci. 2023; e13407.
PubMed Abstract | Publisher Full Text
Milosavljevic B, Vellekoop P, Maris H, et al.: Adaptation of the Mullen Scales
of Early Learning for use among infants aged 5- to 24-months in rural
Gambia. Dev Sci. 2019; 22(5): e12808.
PubMed Abstract | Publisher Full Text | Free Full Text
Moore SE, Cole TJ, Poskitt EME, et al.: Season of birth predicts mortality in
rural Gambia. Nature. 1997; 388(6641): 434.
PubMed Abstract | Publisher Full Text
Moriguchi Y: Beyond bias to Western participants, authors, and editors in
developmental science. Infant and Child Development. 2022; 31(1).
Publisher Full Text
Mullen EM: Mullen scales of early learning (AGS ed.). Circle Pines, MN:
American Guidance Service Inc. 1995.
Murray AL, Xiao Z, Zhu X, et al.: Psychometric evaluation of an adapted
version of the perceived stress scale for ecological momentary assessment
research. Stress Health. 2023.
PubMed Abstract | Publisher Full Text
Murray-Kolb LE, Rasmussen ZA, Scharf RJ, et al.: The MAL-ED Cohort Study:
Methods and Lessons Learned When Assessing Early Child Development
and Caregiving Mediators in Infants and Young Children in 8 Low- and
Middle-Income Countries. Clin Infect Dis. 2014; 59(suppl_ 4): S261–S272.
PubMed Abstract | Publisher Full Text | Free Full Text
Nabwera HM, Fulford AJ, Moore SE, et al.: Growth faltering in rural Gambian
children after four decades of interventions: A retrospective cohort study.
Lancet Glob Health. 2017; 5(2): e208–e216.
PubMed Abstract | Publisher Full Text | Free Full Text
Nabwera HM, Moore SE, Mwangome MK, et al.: The inuence of maternal
psychosocial circumstances and physical environment on the risk of
severe wasting in rural Gambian infants: A mixed methods approach. BMC
Public Health. 2018; 18: 109.
PubMed Abstract | Publisher Full Text | Free Full Text
Nielsen M, Haun D, Kärtner J, et al.: The persistent sampling bias in
developmental psychology: A call to action. J Exp Child Psychol. 2017; 162:
31–38.
PubMed Abstract | Publisher Full Text
Noble KG, Houston SM, Brito NH, et al.: Family income, parental education
and brain structure in children and adolescents. Nat Neurosci. 2015; 18(5):
773–778.
PubMed Abstract | Publisher Full Text | Free Full Text
Noble KG, Houston SM, Kan E, et al.: Neural correlates of socioeconomic
status in the developing human brain. Dev Sci. 2012; 15(4): 516–527.
PubMed Abstract | Publisher Full Text | Free Full Text
Nyongesa MK, Mwangi P, Koot HM, et al.: The reliability, validity and factorial
structure of the Swahili version of the 7-item generalized anxiety disorder
scale (GAD-7) among adults living with HIV from Kili, Kenya. Ann Gen
Psychiatry. 2020; 19(1): 62.
PubMed Abstract | Publisher Full Text | Free Full Text
Pae S, Yoon H, Seol A, et al.: Eects of feedback on parent-child language
with infants and toddlers in Korea. First Language. 2016; 36(6): 549–569.
Publisher Full Text
Papademetriou MD, Richards J, Correia T, et al.: Cortical Mapping of 3D
Optical Topography in Infants. In: S. Van Huel, G. Naulaers, A. Caicedo, D.
F. Bruley, and D. K. Harrison (Eds.): Oxygen Transport to Tissue XXXV. 2013; 789:
455–461.
Publisher Full Text
Peña ED: Lost in translation: Methodological considerations in cross-
cultural research. Child Dev. 2007; 78(4): 1255–1264.
PubMed Abstract | Publisher Full Text
Perdue KL, Jensen SKG, Kumar S, et al.: Using functional near-infrared
spectroscopy to assess social information processing in poor urban
Bangladeshi infants and toddlers. Dev Sci. 2019; 22(5): e12839.
PubMed Abstract | Publisher Full Text | Free Full Text
Perkins JM, Kim R, Krishna A, et al.: Understanding the association between
stunting and child development in low- and middle-income countries:
Next steps for research and intervention. Soc Sci Med. 2017; 193: 101–109.
PubMed Abstract | Publisher Full Text
Pyykkö J, Ashorn U, Chilora E, et al.: Associations between individual
variations in visual attention at 9 months and behavioral competencies at
18 months in rural Malawi. PLoS One. 2020; 15(10): e0239613.
PubMed Abstract | Publisher Full Text | Free Full Text
Pyykkö J, Forssman L, Maleta K, et al.: Early development of visual attention
in infants in rural Malawi. Dev Sci. 2019; 22(5): e12761.
PubMed Abstract | Publisher Full Text
Räikkönen K, Birkás E, Horváth J, et al.: Test-retest reliability of auditory
ERP components in healthy 6-year-old children. Neuroreport. 2003; 14(16):
2121–2125.
PubMed Abstract | Publisher Full Text
Richter L, Black M, Britto P, et al.: Early childhood development: An
imperative for action and measurement at scale. BMJ Global Health. 2019;
4(Suppl 4): e001302.
PubMed Abstract | Publisher Full Text | Free Full Text
Rini CK, Dunkel-Schetter C, Wadhwa PD, et al.: Psychological adaptation and
birth outcomes: The role of personal resources, stress, and sociocultural
context in pregnancy. Health Psychol. 1999; 18(4): 333–345.
PubMed Abstract | Publisher Full Text
Roberts SB, Franceschini MA, Krauss A, et al.: A Pilot Randomized Controlled
Trial of a New Supplementary Food Designed to Enhance Cognitive
Performance during Prevention and Treatment of Malnutrition in
Childhood. Curr Dev Nutr. 2017; 1(11): e000885.
PubMed Abstract | Publisher Full Text | Free Full Text
Roesch SC, Schetter CD, Woo G, et al.: Modeling the types and timing of
stress in pregnancy. ANXIETY STRESS COPIN. 2004; 17(1): 87–102.
Publisher Full Text
Sabanathan S, Wills B, Gladstone M: Child development assessment tools
in low-income and middle-income countries: How can we use them more
appropriately? Arch Dis Child 2015; 100(5): 482–8.
PubMed Abstract | Publisher Full Text | Free Full Text
Sadeh A: Evaluating Night Wakings in Sleep-Disturbed Infants: A
Methodological Study of Parental Reports and Actigraphy. Sleep. 1996;
19(10): 757–762.
PubMed Abstract | Publisher Full Text
Sadeh A: A Brief Screening Questionnaire for Infant Sleep Problems:
Validation and Findings for an Internet Sample. Pediatrics. 2004; 113(6):
e570–e577.
PubMed Abstract | Publisher Full Text
Scerif G, Karmilo-Smith A, Campos R, et al.: To look or not to look? Typical
and atypical development of oculomotor control. J Cogn Neurosci. 2005;
17(4): 591–604.
PubMed Abstract | Publisher Full Text
Sear R, Mace R: Family Matters Kin, Demography and Child Health in a
Rural Gambian Population. In: G. Bentley and R. Mace (Eds.), Substitute
Parents. Berghahn Books, 2009; 50–76.
Publisher Full Text
Sethna V, Pote I, Wang S, et al.: Mother-infant interactions and regional
brain volumes in infancy: An MRI study. Brain Struct Funct. 2017; 222(5):
2379–2388.
PubMed Abstract | Publisher Full Text | Free Full Text
Smith T, Noble M, Noble S, et al.: The English Indices of Deprivation 2015:
Page 36 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
Technical report. Department for Communities and Local Government, 2015.
Reference Source
Soler-Limón KM, Rivera-González R, Velázquez-Moctezuma J, et al.:
Neurobehavior of mexican newborns in conditions of low perinatal risk.
Acta Pediátrica de México. 2019; 40(3): 113–121.
Reference Source
Sosseh SAL, Barrow A, Lu ZJ: Cultural beliefs, attitudes and perceptions
of lactating mothers on exclusive breastfeeding in The Gambia: An
ethnographic study. BMC Womens Health. 2023; 23(1): 18.
PubMed Abstract | Publisher Full Text | Free Full Text
Spitzer RL, Kroenke K, Williams JBW, et al.: A Brief Measure for Assessing
Generalized Anxiety Disorder: The GAD-7. Arch Intern Med. 2006; 166(10):
1092–7.
PubMed Abstract | Publisher Full Text
Super CM, Harkness S: Culture and the perceived organization of newborn
behavior: A comparative study in Kenya and the United States. New Dir
Child Adolesc Dev. 2020; 2020(172): 11–24.
PubMed Abstract | Publisher Full Text
Swindale A, Bilinsky P: Household Dietary Diversity Score (HDDS) for
Measurement of Household Food Access: Indicator Guide. FHI 360/FANTA,
2006.
Reference Source
Tesfaye M, Hanlon C, Wondimagegn D, et al.: Detecting postnatal common
mental disorders in Addis Ababa, Ethiopia: Validation of the Edinburgh
Postnatal Depression Scale and Kessler Scales. J Aect Disord. 2010; 122(1–2):
102–108.
PubMed Abstract | Publisher Full Text
Turesky TK, Jensen SKG, Yu X, et al.: The relationship between biological and
psychosocial risk factors and resting-state functional connectivity in 2-
month-old Bangladeshi infants: A feasibility and pilot study. Dev Sci. 2019;
22(5): e12841.
PubMed Abstract | Publisher Full Text | Free Full Text
Turesky T, Xie W, Kumar S, et al.: Relating anthropometric indicators to brain
structure in 2-month-old Bangladeshi infants growing up in poverty: A
pilot study. Neuroimage. 2020; 210: 116540.
PubMed Abstract | Publisher Full Text | Free Full Text
Twomey DM, Wrigley C, Ahearne C, et al.: Feasibility of using touch screen
technology for early cognitive assessment in children. Arch Dis Child. 2018;
103(9): 853–858.
PubMed Abstract | Publisher Full Text
Ulrich DA: Test of Gross Motor Development. (3rd ed.). Pro-Ed, 2019.
Reference Source
Victora CG, Adair L, Fall C, et al.: Maternal and child undernutrition:
Consequences for adult health and human capital. Lancet. 2008; 371(9609):
340–357.
PubMed Abstract | Publisher Full Text | Free Full Text
Wass SV: Applying cognitive training to target executive functions during
early development. Child Neuropsychol. 2015; 21(2): 150–166.
PubMed Abstract | Publisher Full Text | Free Full Text
Watson D, Clark LA, Tellegen A: Development and validation of brief
measures of positive and negative aect: The PANAS scales. J Pers Soc
Psychol. 1988; 54(6): 1063–1070.
PubMed Abstract | Publisher Full Text
Webb SJ, Jones EJH, Merkle K, et al.: Toddlers with elevated autism symptoms
show slowed habituation to faces. Child Neuropsychol. 2010; 16(3):
255–278.
PubMed Abstract | Publisher Full Text | Free Full Text
Weber A, Fernald A, Diop Y: When Cultural Norms Discourage Talking to
Babies: Eectiveness of a Parenting Program in Rural Senegal. Child Dev.
2017; 88(5): 1513–1526.
PubMed Abstract | Publisher Full Text
Wedderburn CJ, Subramoney S, Yeung S, et al.: Neuroimaging young children
and associations with neurocognitive development in a South African
birth cohort study. NeuroImage. 2020; 219: 116846.
PubMed Abstract | Publisher Full Text | Free Full Text
Weisleder A, Fernald A: Talking to Children Matters: Early Language
Experience Strengthens Processing and Builds Vocabulary. Psychol Sci. 2013;
24(11): 2143–2152.
PubMed Abstract | Publisher Full Text | Free Full Text
Weobong B, Ten Asbroek AH, Soremekun S, et al.: Determinants of postnatal
depression in rural ghana: Findings from the don population based cohort
study. Depress Anxiety. 2015; 32(2): 108–119.
PubMed Abstract | Publisher Full Text
Wijeakumar S, Kumar A, Reyes LMD, et al.: Early adversity in rural India
impacts the brain networks underlying visual working memory. Dev Sci.
2019; 22(5): e12822.
PubMed Abstract | Publisher Full Text | Free Full Text
Worku BN, Abessa TG, Wondafrash Kibebew M, et al.: Eects of home-based
play-assisted stimulation on developmental performances of children
living in extreme poverty: A randomized single-blind controlled trial. BMC
Pediatr. 2018; 18(1): 29.
PubMed Abstract | Publisher Full Text | Free Full Text
World Health Organization: Process of translation and adaptation of
instruments. 2013.
Reference Source
Xie W, Jensen S, Wade M, et al.: Child Growth Predicts Brain Functional
Connectivity and Future Cognitive Outcomes in Urban Bangladeshi
Children Exposed to Early Adversities. bioRxiv. 2018.
Publisher Full Text
Xie W, Kumar S, Kakon SH, et al.: Chronic inammation is associated with
neural responses to faces in bangladeshi children. NeuroImage. 2019; 202:
116110.
PubMed Abstract | Publisher Full Text | Free Full Text
Xu D, Richards JA, Gilkerson J, et al.: Automatic Childhood Autism Detection
by Vocalization Decomposition with Phone-like Units. Proceedings of the 2Nd
Workshop on Child, Computer and Interaction. 2009; 5: 1–7.
Publisher Full Text
Page 37 of 37
Gates Open Research 2023, 7:126 Last updated: 18 OCT 2023
