Seroprevalence of Human Herpesvirus 8 Infection in Individuals From Health Care Centers in Mozambique: Potential for Endemic and Epidemic Kaposi's Sarcoma

Abstract

Human herpesvirus 8 (HHV-8) infection is common in sub-Saharan Africa, but its prevalence in Mozambique is unknown. The seroprevalence of HHV-8 in a cohort of individuals seen at public health centers in Northern (n = 208), Central (n = 226), or Southern (n = 318) Mozambique was examined. All individuals were interviewed to obtain socioeconomic, demographic and clinical data and were tested for serum anti-HHV-8 antibodies using an immunofluorescence assay. The overall frequency of HHV-8 antibodies was 21.4% and, in spite of the diversity of epidemiological characteristics of the tested individuals, did not differ significantly among regions: 18.7%, 24.3% and 21.4% in the North, Center, and South, respectively (chi(2), 2.37; P = 0.305). The variables that were associated significantly with the presence of HHV-8 antibodies were gender, age, level of education, number of siblings and HIV serostatus, but these differed across the regions. In the North, although tested individuals lived under poor socioeconomic conditions, no association between HHV-8 infection and household variables was detected, with the exception of the number of siblings (P = 0.042). In the Central region, HHV-8 infection was associated with gender (P = 0.010), the number of household members (P = 0.031), and the place of attendance (P = 0.021). In the South, HHV-8 infection was associated with the number of siblings (P = 0.023) and HIV status (P = 0.002). The overall prevalence of HHV-8 seropositivity increased with age. These results demonstrate that Mozambique is another country in Africa with endemic HHV-8 infection, and, because of the AIDS epidemic, continued access to antiretroviral treatment is necessary to avert an outbreak of AIDS-Kaposi's sarcoma.

Full text

Journal of Medical Virology 82:1216–1223 (2010)
Seroprevalence of Human Herpesvirus 8 Infection
in Individuals From Health Care Centers in
Mozambique: Potential for Endemic and
Epidemic Kaposi’s Sarcoma
Adele Caterino-de-Araujo,
1,2
* Rolanda Carmen Rafael Manuel,
3
Rosana Del Bianco,
4
Elizabeth Santos-Fortuna,
1
Mariana Cavalheiro Magri,
1,2
Joyce Matie Kinoshita Silva,
1
and Rui Bastos
3
1
Immunology Department, Instituto Adolfo Lutz, Secretary of Health of Sa
˜o Paulo, Sa
˜o Paulo, SP, Brazil
2
Faculdade de Cie
ˆncias Farmace
ˆuticas, USP, Sa
˜o Paulo, SP, Brazil
3
Dermatology Unit, Hospital Central de Maputo, University Eduardo Mondlane, Maputo, Mozambique
4
AIDS Reference Center of Sa
˜o Paulo, Secretary of Health of Sa
˜o Paulo, Sa
˜o Paulo, SP, Brazil
Human herpesvirus 8 (HHV-8) infection is com-
mon in sub-Saharan Africa, but its prevalence in
Mozambique is unknown. The seroprevalence
of HHV-8 in a cohort of individuals seen at
public health centers in Northern (n ¼208),
Central (n ¼226), or Southern (n ¼318) Mozam-
bique was examined. All individuals were inter-
viewed to obtain socioeconomic, demographic
and clinical data and were tested for serum anti-
HHV-8 antibodies using an immunofluorescence
assay. The overall frequency of HHV-8 antibodies
was 21.4% and, in spite of the diversity of
epidemiological characteristics of the tested
individuals, did not differ significantly among
regions: 18.7%, 24.3% and 21.4% in the North,
Center, and South, respectively (w
2
, 2.37;
P¼0.305). The variables that were associated
significantly with the presence of HHV-8 anti-
bodies were gender, age, level of education,
number of siblings and HIV serostatus, but
these differed across the regions. In the
North, although tested individuals lived under
poor socioeconomic conditions, no association
between HHV-8 infection and household varia-
bles was detected, with the exception of the
number of siblings (P¼0.042). In the Central
region, HHV-8 infection was associated with
gender (P¼0.010), the number of household
members (P¼0.031), and the place of attendance
(P¼0.021). In the South, HHV-8 infection was
associated with the number of siblings (P¼0.023)
and HIV status (P¼0.002). The overall prevalence
of HHV-8 seropositivity increased with age.
These results demonstrate that Mozambique is
another country in Africa with endemic HHV-8
infection, and, because of the AIDS epidemic,
continued access to antiretroviral treatment is
necessary to avert an outbreak of AIDS-Kaposi’s
sarcoma. J. Med. Virol. 82:1216–1223,
2010. ß2010 Wiley-Liss, Inc.
KEY WORDS: epidemiology; human herpes-
virus 8 (HHV-8); HIV; risk fac-
tors; serology
INTRODUCTION
Human herpesvirus 8 (HHV-8, also known as Kaposi’s
sarcoma [KS]-associated herpesvirus) is the etiologic
agent of KS [Moore and Chang, 1995]. The global
seroprevalence of HHV-8 is uneven and depends on
the geographic region and the analyzed population. In
the general population, a low HHV-8 seroprevalence
(<3–5%) has been reported in Central, Northern, and
Western Europe, Asia, and North America [Hengee
et al., 2002; Edelman, 2005]. An intermediate HHV-
8 prevalence (5–35%) has been reported in Mediterra-
nean, Eastern European, and Caribbean countries, as
Study conducted in the Immunology Department, Instituto
Adolfo Lutz, Sa
˜o Paulo, SP, Brazil.
Grant sponsor: Ministe
´rio da Cie
ˆncia e Tecnologia/Conselho
Nacional de Desenvolvimento Cientı
´fico e Tecnolo
´gico (MCT/
CNPq), Brasil, (PROAFRICA); Grant numbers: 490452/2007-8,
304372/2006-4 (Fellowship to ACA); Grant sponsor: Instituto
Adolfo Lutz (CTC) 55/06; Grant sponsor: International Health
Alliance (Mozambique; partial support).
*Correspondence to: Adele Caterino-de-Araujo, Immunology
Department, Instituto Adolfo Lutz, Av. Dr. Arnaldo, 355, 118
andar, 01246-902 Sa
˜o Paulo, SP, Brazil.
E-mail: caterino@usp.br; caterino@ial.sp.gov.br
Accepted 4 February 2010
DOI 10.1002/jmv.21789
Published online in Wiley InterScience
(www.interscience.wiley.com)
ß2010 WILEY-LISS, INC.

well as in Brazil [Hengee et al., 2002; Edelman, 2005;
Nascimento et al., 2008]. Finally, a high HHV-8
prevalence (>50%) has been reported in Central and
Southern Africa and in some regions of South America
[Hengee et al., 2002; Dedicoat and Newton, 2003; Cunha
et al., 2005; Edelman, 2005; Mohanna et al., 2007]. In
HHV-8 endemic populations, such as those in Africa
and Amazonia, no evidence exists demonstrating a
difference in the prevalence of HHV-8 between men
and women [Dedicoat and Newton, 2003; Cunha et al.,
2005; Sissolak and Mayaud, 2005; Mohanna et al.,
2007]; however, in endemic regions KS is more common
in adult men than in women [Dedicoat and Newton,
2003; Sissolak and Mayaud, 2005].
After the onset of the AIDS pandemic, an explosion in
the incidence of KS was observed worldwide, and this
tumor became more frequent in several African coun-
tries and in women and children [Sissolak and Mayaud,
2005; Szajerka and Jablecki, 2007]. The impact of HIV
on HHV-8 infection is not completely understood, but is
likely due, at least in part, to the immunosuppressive
effect of HIV. Other mechanisms may play a role, such as
HIV Tat protein-enhanced entry of HHV-8 into endo-
thelial cells and reactivation of HHV-8 from a latent
state [Caterino-de-Araujo, 2007]. Notably, antiretrovi-
ral (ARV) therapy has contributed to a reduction in the
number of AIDS-KS cases, probably in part by effects on
immune reconstitution and the clearance of HIV and
consequently of HHV-8 [Sissolak and Mayaud, 2005;
Caterino-de-Araujo, 2007; Szajerka and Jablecki, 2007].
To date, no large and well-documented study of HHV-
8 seroprevalence has been conducted in Mozambique.
One survey, which examined a cohort of 161 HIV-
positive patients seen at the Drug Resource Enhance-
ment against AIDS and Malnutrition (DREAM) pro-
gram of the Community of Sant’Egidio from Maputo
and Beira cities and 166 HIV-negative patients from
Hospital Central de Maputo, Mozambique, detected an
overall HHV-8 seroprevalence of 51.1% (52.7% among
HIV-positive patients and 49.6% among HIV-negative
patients) [Ceffa et al., 2007]. However, different results
were obtained in a preliminary study of HHV-8 infection
in Hospital Central de Maputo, where great variation
in the HHV-8 antibody frequency was observed depend-
ing on the analyzed group and epidemiological varia-
bles. Among 31 individuals from the staff, an HHV-8
seropositivity rate of 9.7% was observed, and among
67 outpatients, an HHV-8 seropositivity rate of 16.4%
was observed. In these outpatients, HHV-8 seropositiv-
ity could be correlated with having only a primary school
education and HIV infection. Among 34 hospitalized
patients from the Dermatology Unit, 47.1% were HHV-
8-seropositive and associations between HHV-8 seropo-
sitivity and gender, native to Central region, prior
blood transfusions, and KS were observed [Caterino-de-
Araujo et al., 2009]. To confirm and extend these
preliminary data, the present study was undertaken
with the following objectives: (i) to evaluate the
seroprevalence of HHV-8 infection in individuals
attending public health care centers in Northern,
Central and Southern Mozambique, regions where the
HIV infection rate varies considerably; (ii) to identify
any association between HHV-8 infection and socio-
economic, demographic or clinical variables; and (iii) to
speculate on HHV-8 and HIV infections in endemic and
epidemic KS outbreaks when preventive transmission
measurements may not be introduced and ARV therapy
may not be available universally or could be interrupted.
MATERIALS AND METHODS
Study Population
From August to September 2008, two trained doctors
(ACA and RCRM) interviewed 752 individuals who were
seen at public health care centers in Mozambique
using a standard questionnaire about socioeconomic
and demographic characteristics, including gender, age,
education, marital status, number of siblings, number
of household members, occupation, place of residence,
household conditions (electricity and water), and place
of attendance. Other variables such as HIV status, prior
blood transfusion, history of sexually transmitted dis-
eases (STD) or Kaposi’s sarcoma, sexual lifestyle and the
STD history of the their sexual partner(s) were recorded,
as well as the actual participant clinical complaint or
status. Two health care centers, one from an urban
area and one from a periurban or rural area in each of
three provinces in Mozambique (Nampula, Sofala, and
Maputo) located in Northern, Central, and Southern
regions, respectively, were selected for study. The
number of individuals interviewed in each province
was 208 in Nampula, 226 in Sofala, and 318 in Maputo.
The mean age of the 752 participants was 29.9 years,
varying from 2 months to 79 years. In Northern, Central
and Southern Mozambique, the mean age was 24.5 years
(2 months to 79 years), 31.6 years (1–72 years), and
32.3 years (6 months to 77 years), respectively. Table I
presents the major characteristics of the individuals
separated by region for comparative purposes.
The study was approved by the Ministry of Health of
Mozambique and the Research and Ethics Committees
of the participating institutions, and written informed
consent was obtained before blood collection from all
study participants or parents (in case of children).
Specimen Collection and Serologic Tests
Blood samples were collected on filter paper by
finger prick, dried at room temperature, stored under
refrigeration and sent on dry ice to Instituto Adolfo Lutz,
Sa
˜o Paulo, Brazil, for analysis. HHV-8 antibodies were
assessed in eluted blood using ‘‘in house’’ immunofluo-
rescence assays (IFA-LANA and IFA-Lytic) as previ-
ously described [Silva et al., 2007]. The sensitivity and
specificity of the assays were confirmed in patients with
AIDS-KS and healthy persons from Sa
˜o Paulo. Briefly,
BCBL-1 cells, a cell line latently infected with HHV-8,
were left unstimulated (latent infection) or stimulated
with tetradecanoyl phorbol ester acetate (TPA) to induce
lytic replication and were used to prepare slides for the
J. Med. Virol. DOI 10.1002/jmv
Seroprevalence of HHV-8 Infection in Mozambique 1217

IFA-LANA or IFA-Lytic assays, respectively. Samples
were eluted from 6 mm diameter dried blood punches in
200 ml phosphate-buffered saline (PBS, dilution 1:40),
and then diluted 1:50, 1:100, or 1:200 in PBS/skim milk.
Reference positive serum was run in each assay. All
slides were scored by two independent observers
using fluorescence microscopy. A sample was scored as
positive in the IFA-LANA assay when clear, punctate
nuclear staining was observed in greater than 90% of the
cells and as positive in the IFA-Lytic assay when strong
membrane and cytoplasmic staining was detected in 5–
10% of cells. Samples were deemed seropositive for
HHV-8 when at least one IFA (LANA and/or Lytic) at a
1:100 dilution was positive. All positive samples were
titrated to an endpoint using twofold serial dilution for
further analysis.
HIV infection was assessed in the same blood sample
using an enzyme-linked immunosorbent assay which
detects both HIV-1 and HIV-2 infections (Vironostika
HIV Uni-Form II plus O, BioMe
´rieux, The Netherlands).
This assay was conducted only for comparative purposes
and was not confirmed by Western blot analysis.
J. Med. Virol. DOI 10.1002/jmv
TABLE I. Characteristics of the Study Population From Northern, Central, and Southern Mozambique
Characteristic
Northern Province
of Nampula
Central Province
of Sofala
Southern Province
of Maputo
Mozambique all
provinces
n%n%n% n %
Gender
Male 83 39.9 86 38.1 94 29.6 263 35
Female 125 60.1 140 61.9 224 70.4 489 65
Age (years)
10 48 23.3 8 3.9 20 6.3 76 10.2
11–20 30 14.6 26 11.5 28 8.9 84 11.2
21–30 60 29.1 85 37.6 100 31.6 245 32.8
31–40 40 19.4 57 25.2 97 30.7 194 25.9
41–50 17 8.3 35 15.4 48 15.2 100 13.4
50 11 5.3 15 6.6 23 7.3 49 6.5
Education
None 67 32.5 36 15.9 27 8.7 130 17.5
Primary 71 34.5 86 38.1 108 34.7 265 35.7
Secondary 50 24.3 82 36.3 125 40.2 257 34.6
Senior secondary 11 5.3 16 7.1 29 9.3 56 7.5
Higher 7 3.4 6 2.6 22 7.1 35 4.7
Marital status
Single 77 37.4 86 38.1 128 40.3 291 38.9
Married 100 48.5 104 46 151 47.5 355 47.3
Separated/divorced 21 10.2 7 3.1 22 6.9 50 6.6
Widowed 8 3.9 29 12.8 17 5.3 54 7.2
Number of siblings
0 76 36.9 64 28.3 93 29.4 233 31.2
1 25 12.1 33 14.6 57 18.1 115 15.4
2–4 65 31.6 91 40.3 130 41.1 286 38.2
>4 40 19.4 38 16.8 36 11.4 114 15.2
Number of household members
1–4 71 43.6 80 37.6 114 38.4 265 39.4
5–8 82 50.3 97 45.5 150 50.5 329 48.9
>8 10 6.1 36 16.9 33 11.1 79 11.7
Occupation
Household 33 21.0 72 35.7 59 20 164 24.4
Laborer/professional 49 31.2 98 44.5 186 63.3 333 49.6
Students 33 21.0 36 16.4 45 15.3 114 17
Peasant/farmer 42 26.8 14 6.4 4 1.4 60 8.9
Place of residence
Urban 84 41.2 212 93.8 214 68.2 510 68.5
Rural 120 58.5 14 6.2 100 31.8 234 31.5
Household electricity
Yes 46 23 123 54.4 213 67.8 382 51.6
Household water
Yes 26 13 87 38.5 181 57.6 294 39.7
Place of attendance
Urban Health Center 69 33.2 101 44.7 185 58.2 355 47.2
Periurban/Rural Health
Center
139 66.8 125 55.3 133 41.8 397 52.8
HIV
Yes 67 32.2 148 65.5 140 44 355 47.2
Blood transfusion
Yes 16 7.7 25 11.1 35 11.1 76 10.1
1218 Caterino-de-Araujo et al.

Statistical Analysis
Differences in the seroprevalence of HHV-8 were
evaluated using the Chi-square (w
2
) test or Fisher’s exact
test, as appropriate. A Pvalue less than 0.05 was
considered to be significant. Logistic regression was
used to calculate the odds ratios (OR) and 95%
confidence intervals (CI).
RESULTS
Table I presents the characteristics of the studied
population, with more females than males who sought
health care services. In Northern Mozambique, more
individuals from the rural site and children under the
age of 10 years participated in the study. While the
North region had the least affluent household conditions
(lack of water and electricity), these conditions were
also reported in the Center and South, albeit at a lower
frequency. In addition, the peasant/farmer occupation
was more frequent in the North. The highest number of
individuals who resided in an urban area was observed
in Central Mozambique (Table I).
The major complaints of participants also varied
according to the region. In the North, children and
adolescents often sought health care services because
of parasitic infections and dermatologic diseases/
manifestations such as scabies, lice, pityriasis, impetigo
and herpes, while in the Central region, diseases related
to HIV/AIDS were the most frequent complaints
of the participants (e.g., fever, weight loss, cough,
tuberculosis, herpes zoster, syphilis, piodermitis,
seborrheic dermatitis, and diarrhea). Similar to Central
Mozambique, the same complaints were observed in the
South with the addition of sexually transmitted diseases
(e.g., condiloma acuminatus and syphilis) and related
manifestations such as ulcers, burbles, genital warts,
and leucorrhoea. Of interest, 20.7%, 37.6%, and 22.3% of
the studied population from Northern, Central and
Southern regions, respectively, sought health care
services because they were HIV positive or they were
on ARV therapy.
In spite of the socioeconomic, demographic and clinical
differences among the groups, no differences in the HHV-
8-seropositivity rates were detected. The overall seropre-
valence of HHV-8 was 21.4% and did not differ signifi-
cantly among the regions: 18.7% in Northern, 24.3% in
Central and 21.4% in Southern Mozambique (w
2
, 2.37;
P¼0.305). Epidemiological data describing the frequency
of HHV-8 antibody detection and the risk of infection
are presented in Table II. The variables that were
associated significantly with HHV-8 antibodies were
gender (P¼0.024), age (P¼0.031), level of education
(P¼0.007), number of siblings (P¼0.001), and HIV
serostatus (P¼0.002), but they differed across the
regions. In the North, no statistical association was
observed between HHV-8 infection and all of the examined
variables (all P>0.05), except for the number of siblings
(P¼0.042). In the Central region, HHV-8 infection was
more frequently detected in men (P¼0.010), even though
the majority of participants were female; infection was also
correlated with the number of household members
(P¼0.031) and attending an urban health care center
(P¼0.021). In contrast, in the South, HHV-8 infection was
associated with the number of siblings (P¼0.023) and HIV
status (P¼0.002) (Table II).
An overall linear increase in the frequency of HHV-8
antibodies was observed with age in individuals from
Mozambique (Fig. 1); however, when each region was
analyzed separately, there was no significant difference
(Table II). This age pattern contrasted with the age
pattern detected for HIV (Fig. 1).
When the number of HHV-8-seropositive participants
allowed for other statistical analyses, significant HHV-8
associations were detected with gender (OR 1.55, CI 0.96–
2.5), high number of household members (OR 2.65, CI
0.81–8.87) and attending an urban health care center (OR
2.06, CI 1.06–3.99) in Central Mozambique. In the North
and South regions HHV-8 was associated with education
level (OR 1.55, CI 0.58–4.19, and OR 1.53, CI 0.81–2.92,
respectively); an association of HHV-8 and HIV serostatus
was detected only in the South (OR 2.30, CI 1.29–4.16).
Importantly, the results obtained with HIV serology
presented different results than the results obtained
using the interview. In the North, 20.7% of individuals
reported being HIV positive; however, when tested in
the laboratory, 32.2% of individuals were HIV positive.
Similar results were seen in both Central (37.5%
reported in the interview vs. 65.5% based on lab
serology) and Southern (22.3% reported in the interview
vs. 44% based on lab serology) regions. Table III presents
the seroprevalence of HHV-8 infection among individ-
uals from the three regions according to HIV serostatus.
Eight participants from the South had KS, five of
whom were HIV positive. Among the KS patients, six
(75%) tested positive for HHV-8 antibodies.
Previous blood transfusion was associated with HHV-
8 infection in participants from the Southern region only
when a multivariate analysis was conducted (OR 1.56,
CI 0.66–3.65).
All of the participants reported heterosexual behavior
and no illicit drug use, except for two participants from
the Central region, who reported drug use. Although
some participants reported having more than two sexual
partners during the last 6 months, no statistical analysis
could be performed with these variables. In addition, a
history of STDs was not associated with HHV-8 infection
(all P>0.05), but when participants were asked about
this variable for their sexual partners, the HHV-8
seroprevalence increased from 42.9% to 50% in parti-
cipants from the North, 10.5% to 31.6% in participants
from the Center and 24% to 30.8% in participants from
the South. Again, the low number of HHV-8-seropositive
participants who knew their partner’s history of STDs,
HIV status, or KS status did not allow for statistical
analysis of these results (data not shown).
DISCUSSION
Few data are available on the seroprevalence of
and risk factors associated with HHV-8 infection in
J. Med. Virol. DOI 10.1002/jmv
Seroprevalence of HHV-8 Infection in Mozambique 1219

Mozambique. Only two preliminary studies have been
conducted, and these studies present different results
[Ceffa et al., 2007; Caterino-de-Araujo et al., 2009]. The
reported heterogeneity in HHV-8 seroprevalence could
be an artifact resulting from interlaboratory variability
in the assays or may be due to variability in the age or
characteristics of the analyzed population. In fact, all of
these variables could account for the discordant results
that have been reported. For instance, one study used an
ELISA that is available commercially to detect HHV-8
J. Med. Virol. DOI 10.1002/jmv
TABLE II. Seroprevalence of HHV-8-Specific Antibodies in Individuals Attending Public Health Centers in Northern, Central,
and Southern Mozambique According to Epidemiological Data
Epidemiological data
Northern Central Southern Total
HHV-8
seropositive
n/total (%)
a
P
HHV-8
seropositive
n/total (%)
a
P
HHV-8
seropositive
n/total (%)
a
P
HHV-8
seropositive
n/total (%)
a
P
Gender
Male 15/83 (18.1) 0.952 29/86 (33.7) 0.010 25/94 (26.6) 0.142 69/263 (26.2) 0.024
Female 23/125 (18.4) 26/140 (18.6) 43/224 (19.2) 92/489 (18.8)
Age (years)
10 4/48 (8.3) 0.090 2/8 (25) 0.749 1/20 (5) 0.111 7/76 (9.2) 0.031
11–20 9/30 (30) 6/26 (23.1) 3/28 (10.7) 18/84 (21.4)
21–30 9/60 (15) 19/85 (22.4) 19/100 (19) 47/245 (19.2)
31–40 11/40 (27.5) 15/57 (26.3) 24/97 (24.7) 50/194 (25.8)
41–50 4/17 (23.5) 7/35 (20) 12/48 (25) 23/100 (23)
50 1/11 (9.1) 6/15 (40) 8/23 (34.8) 15/49 (30.6)
Education
None 7/67 (10.4) 0.066 4/36 (11.1) 0.144 4/27 (18.5) 0.175 15/130 (11.5) 0.007
Primary 18/71 (25.4) 24/86 (29.3) 31/108 (28.7) 73/265 (27.5)
Secondary 9/50 (18) 21/82 (25.6) 26/125 (20.8) 56/257 (21.8)
Senior secondary 4/11 (36.4) 3/16 (18.8) 3/29 (10.3) 10/56 (17.9)
Higher —3/6 (50) 3/22 (13.6) 6/35 (17.1)
Marital status
Single 13/77 (16.9) 0.944 20/86 (23.3) 0.985 25/128 (19.5) 0.812 58/291 (19.9) 0.776
Married 19/100 (19) 26/104 (25) 33/151 (21.9) 78/355 (22)
Separated/divorced 4/21 (19) 2/7 (28.6) 5/22 (22.7) 11/50 (22)
Widowed 2/8 (25) 7/29 (24.1) 5/17 (29.4) 14/54 (25.9)
Number of siblings
0 9/76 (11.8) 0.042 13/64 (20.3) 0.763 12/93 (12.9) 0.023 34/233 (14.6) 0.001
1 7/25 (28) 9/33 (27.3) 15/57 (26.3) 31/115 (27)
2–4 9/65 (13.8) 22/91 (24.2) 27/130 (20.8) 58/286 (20.3)
>4 13/40 (32.5) 11/38 (28.9) 13/36 (36.1) 37/114 (32.5)
Number of household members
1–4 12/71 (16.9) 0.629 19/80 (23.8) 0.031 21/114 (19.3) 0.499 52/265 (19.6) 0.127
5–8 19/82 (23.2) 19/97 (19.6) 38/150 (25.3) 76/329 (23.1)
>8 2/10 (20) 15/36 (41.7) 7/33 (21.2) 24/79 (30.4)
Occupation
Household 6/33 (18.2) 0.080 14/72 (19.4) 0.300 12/59 (20.3) —32/164 (19.5) 0.542
Laborer/professional 8/49 (16.3) 24/98 (24.5) 46/186 (24.7) 78/333 (23.4)
Students 12/33 (36.4) 13/36 (36.1) 5/45 (11.1) 30/114 (26.3)
Peasant/farmer 6/42 (14.3) 4/14 (28.6) 2/4 (50) 12/60 (20)
Place of residence
Urban 16/84 (19) 0.897 53/212 (25) 0.560 42/214 (19.6) 0.279 111/510 (21.8) 0.799
Rural 22/120 (18.3) 2/14 (14.3) 25/100 (25) 49/234 (20.9)
Household electricity
Yes 8/46 (17.4) 0.751 31/123 (25.2) 0.740 44/213 (20.7) 0.669 83/382 (21.7) 0.942
No 30/154 (19.5) 24/103 (23.3) 23/101 (22.7) 77/358 (21.5)
Household water
Yes 3/26 (11.5) 0.440 24/87 (27.6) 0.368 34/181 (18.8) 0.198 61/294 (20.7) 0.639
No 35/174 (20.1) 31/139 (22.3) 33/133 (24.8) 99/446 (22.2)
Place of attendance
Urban Health Center 14/79 (20.3) 0.595 32/101 (31.7) 0.021 39/185 (21.1) 0.877 85/355 (23.9) 0.109
Periurban/Rural Health
Center
24/139 (17.3) 23/125 (18.4) 29/133 (21.8) 76/397 (19.1)
HIV
Yes 14/67 (20.9) 0.499 38/148 (25.7) 0.518 41/140 (29.3) 0.002 93/355 (26.2) 0.002
No 24/141 (17) 17/78 (21.8) 27/178 (15.2) 68/397 (17.1)
Blood transfusion
Yes 4/16 (25) 0.698 7/25 (28) 0.651 10/35 (28.6) 0.263 21/76 (27.6) 0.159
No 34/192 (17.7) 48/201 (23.9) 57/280 (20.4) 139/673 (20.7)
a
HHV-8 seropositive indicates latent (IFA-LANA) and/or lytic-(IFA-Lytic) seropositivity; P0.05 was considered significant.
1220 Caterino-de-Araujo et al.

antibodies [Ceffa et al., 2007]; the other used an ‘‘in
house’’ immunofluorescence assay [Caterino-de-Araujo
et al., 2009]. In addition, differences in the charac-
teristics of the participants, including age and gender
and socioeconomic, demographic and clinical variables
were observed between the studies.
The results obtained in the present study showed a
high diversity in the socioeconomic, demographic and
clinical variables of the participants. Nonetheless, the
socioeconomic and demographic characteristics of each
region did not alter the overall HHV-8 seroprevalence
that is similar to the rate observed in endemic
populations from Mediterranean and some African
countries [Edelman, 2005; Mbulaiteye et al., 2005;
Nascimento et al., 2008]. In addition, the linear increase
in the HHV-8 seroprevalence with age is similar to the
age pattern observed in countries with endemic HHV-8
infection [DeSantis et al., 2002; Plancoulaine et al.,
2004; Cunha et al., 2005; Mbulaiteye et al., 2005], and
contrasts with the epidemic age pattern detected with
HIV infection.
In South Africa, where KS is endemic and has risen at
least threefold (and continues to rise) after the HIV
epidemic, the increase in the HHV-8 seroprevalence
with age is similar in men and in women [Sitas and
Newton, 2000]. In some villages from rural Tanzania,
the HHV-8 seroprevalence has been reported to be as
high as 89% in adults over 45 years old [Mbulaiteye
et al., 2003]. In Cameroon, a seroprevalence of 62% was
detected among individuals between 15 and 40 years of
age [Plancoulaine et al., 2004]. The age-related increase
in the HHV-8 seroprevalence has also been observed
among hospitalized patients and local volunteers in
Malawi [DeSantis et al., 2002]. In these HHV-8 endemic
populations, transmission occurs within families, from
mother and other relatives to children via nonsexual
routes and between spouses via sexual routes. Based on
the results obtained in this study, the same routes of
HHV-8 transmission could be occurring in Mozambique
while for HIV, sexual intercourse seems to be the most
important route of virus transmission.
Although similar frequencies of HHV-8 antibodies
were detected among males and females from endemic
regions, no explanation for the higher HHV-8 seropre-
valence among males in Central Mozambique was
obtained. In addition, the association between HHV-8
infection and a high number of household members
in the Center and a high number of siblings in the
North mirrored the characteristics of these populations
and is most likely not a true associated risk factor.
However, one cannot exclude the frequent and close
contact among HHV-8-infected individuals, which may
facilitate HHV-8 transmission and acquisition.
In Uganda an association between HHV-8 infection
and both low socioeconomic status and the lack of
tap water has been described, and the investigators
attributed the limited access to water and poorer
hygiene as risk factors for HHV-8 infection [Mbulaiteye
et al., 2005]. In the present study, despite the parti-
cipants (mostly from Northern Mozambique) reporting
not having water and electricity at home, no correlation
between these variables and HHV-8 seropositivity was
detected. In contrast, having only a primary school
education was associated with HHV-8 infection; a
similar association has been described elsewhere in
Africa [Sitas et al., 1999; Sitas and Newton, 2000;
Klaskala et al., 2005]. Regarding other epidemiological
variables, in South African black cancer patients,
an association between HHV-8 infection and a low
socioeconomic status, such as pension/disability, non-
active or domestic occupation, has been observed
among females [Wojcicki et al., 2004]; however, these
J. Med. Virol. DOI 10.1002/jmv
TABLE III. Seroprevalence of Anti-HHV-8 Antibodies in HIV-Positive and -Negative
Individuals From Northern, Central, and Southern Mozambique
HIV
seronegative
a
HIV
seropositive
a
w
2
Pn%n%
HHV-8 seronegative (n)
a
Northern (170) 117 68.8 53 31.2 48.19 <0.001
Central (171) 61 35.7 110 64.3 28.08 <0.001
Southern (250) 151 60.4 99 39.6 21.63 <0.001
HHV-8 seropositive (n)
a
Northern (38) 24 63.2 14 36.8 5.26 0.022
Central (55) 17 30.9 38 69.1 16.04 <0.001
Southern (68) 27 39.7 41 60.3 5.76 0.016
a
According to the Materials and Methods Section; P0.05 was considered significant.
Fig. 1. HHV-8 and HIV seroprevalence in Mozambique according to
age group.
Seroprevalence of HHV-8 Infection in Mozambique 1221

associations were not detected in Mozambique. Inter-
estingly, in Uganda, where HHV-8 is relatively
common, the prevalence of HHV-8 antibodies increased
with age, but showed little association with nearly 50
other epidemiological variables [Newton et al., 2003].
Among antenatal women in Zambia, of 154 variables
evaluated by logistic regression analyses, only three risk
factors were associated with HHV-8 seropositivity:
diagnosis of genital warts, HIV-1 co-infection and
having only a primary school education [Klaskala
et al., 2005].
Interestingly, in the present study, a history of STDs
was not associated with HHV-8 infection. Nonetheless,
when participants were asked about this variable for
their sexual partners, an increase in the HHV-8-
seropositive rate was observed. However, participant
omission or lack of knowledge regarding this informa-
tion, which may have affected the analysis, could not be
excluded.
Concerning previous blood transfusions, when logistic
regression analyses were conducted, an association with
HHV-8 seropositivity was detected in Mozambique.
The same association has been documented in hospi-
talized patients with cancer in Uganda [Newton et al.,
2003]. Of note, malaria, which occurs in a great
number of individuals several times throughout life in
Mozambique, was the major cause of blood transfusion
in study participants.
Sexual lifestyle, including homosexual contacts, a
high number of sexual partners and oral and/or anal
intercourse, as well as drug addition, were not asso-
ciated with HHV-8 infection in Mozambique. Again,
participant omission of these behaviors could not be
excluded.
Finally, regarding HIV, it is important to highlight
the difference in the data obtained through the inter-
view and the serologic results. Despite some omission of
participants, one could speculate that approximately
36%, 43%, and 49% of HIV-infected individuals seen in
health care centers in Northern, Central, and Southern
Mozambique, respectively, do not know their correct
HIV status. Therefore, these patients were not on ARV
therapy and could be transmitting HIV; this scenario
is concerning. The strong association between HIV
infection and KS in Africa and worldwide has been
described extensively [Hengee et al., 2002; Edelman,
2005; Sissolak and Mayaud, 2005; Caterino-de-
Araujo, 2007], as has the extraordinary potential of
ARV therapy in preventing KS-AIDS in resource-
limited settings [Cassol et al., 2005; Flanigan et al.,
2005].
Of interest, the overall estimated prevalence of HIV in
Mozambique was 9%, 18%, and 21%, respectively in the
North, Central, and South regions, and 50,000 HIV-1
infected patients in Mozambique were on ARV treat-
ment (29% of eligible individuals) [Ministe
´rio da Sau
´de,
2007]. In 2006, the Ministry of Health of Mozambique
estimated the prevalence of HIV to be 31.1%, 34.8% and
31.8% for Nampula, Beira and Maputo cities, respec-
tively [Ministe
´rio da Sau
´de, 2006].
Therefore, the results obtained in the present study
are important to demonstrate more effectively to the
government of Mozambique the need for preventive
transmission measurements and universal access to
HIV diagnostic tests and treatment, which in turn could
diminish the impact of an outbreak of endemic and
epidemic KS. Detailed analyses of the titers and the
importance of specific antibodies against latent and lytic
HHV-8 antigens are currently in progress, and future
research should focus on the effect of HHV-8 infection on
adults and children in order to avoid and properly treat
this disabling and stigmatizing disease in Mozambique.
ACKNOWLEDGMENTS
We are grateful to the study participants, the staff at
the health care centers and the nurses who collected
blood samples (Ligia Leonardo and Gilda Langa from
Hospital Central de Maputo). They are also grateful to
the Directors and Heads of CS Chamanculo (cidade de
Maputo) and CS Ndlavela (provı
´ncia de Maputo), CS
Ponta-Ge
ˆa and CS Manga Nhaconjo (cidade de Beira)
and CS 25 de Setembro and CS Anchilo (cidade de
Nampula) for allowing the study to be conducted and to
the Ministry of Health of Mozambique for allowing the
study to be reported. The authors are also indebted to
Instituto Adolfo Lutz and Centro de Refere
ˆncia e
Treinamento em AIDS of Sa
˜o Paulo, Brazil for allowing
the study to be conducted and the participation of the
staff, particularly Ana Ke
´sia de Souza Lima. The
authors also thank the Ministe
´rio da Sau
´de of Brazil
for providing the serologic HIV kits.
REFERENCES
Cassol E, Page T, Mosam A, Friedland G, Jack C, Lalloo U, Kopetka J,
Patterson B, Esterhuizen T, Coovadia HM. 2005. Therapeutic
response of HIV-1 subtype C in African patients coinfected with
either Mycobacterium tuberculosis or Human herpesvirus-8.
J Infect Dis 191:324–332.
Caterino-de-Araujo A. 2007. The fight against Kaposi’s sarcoma in
AIDS—Lessons from Brazil. South Afr J HIV Med 26:53 –54.
Caterino-de-Araujo A, Manuel RCR, Del Bianco R, Santos-Fortuna E,
Magri MC, Silva JMK, Bastos R. 2009. First survey for detecting the
presence of human herpesvirus 8 infection (HHV-8) in Maputo,
Mozambique. Braz J Infect Dis 13:200 –202.
Ceffa S, Buonomo E, Altan AM, Erba F, Germano P, Guidotti G, Liotta
G, Magnano San Lio M, Scarcella P, Palombi L, Marazzi MC. 2007.
Seroprevalence of HHV-8 in a cohort of HIV-negative and HIV-
positive patients in Mozambique. Ann Ig 19:519–523.
Cunha AMG, Caterino-de-Araujo A, Costa SCB, Santos-Fortuna E,
Boa-Sorte NCA, Gonc¸alves MS, Costa FF, Galva
˜o-Castro B. 2005.
Increasing seroprevalence of Human herpesvirus 8 (HHV-8) with
age confirms HHV-8 endemicity in Amazon Amerindians from
Brazil. J Gen Virol 86:2433–2437.
Dedicoat M, Newton R. 2003. Review of the distribution of Kaposi’s
sarcoma-associated herpesvirus (KSHV) in Africa in relation to the
incidence of Kaposi’s sarcoma. Br J Cancer 88:1–3.
DeSantis SM, Pau CP, Archibald LK, Nwanyanwu OC, Kazembe PN,
Dobbie H, Jarvis WR, Jason J. 2002. Demographic and immune
correlates of human herpesvirus 8 seropositivity in Malawi, Africa.
Int J Infect Dis 6:266–271.
Edelman DC. 2005. Human herpesvirus 8: A novel human pathogen.
Virol J 2:1–91.
Flanigan TP, Campbell T, Harwell J, Kumarasamy N. 2005. The
extraordinary hope of antiretroviral therapy in South Africa (even
for patients with tuberculosis or Kaposi sarcoma!). J Infect Dis
191:321–323.
J. Med. Virol. DOI 10.1002/jmv
1222 Caterino-de-Araujo et al.

Hengee UR, Ruzicka T, Tyring SK, Seeber S. 2002. Update on Kaposi’s
sarcoma and other HHV-8 associated diseases. Part. 1: Epidemio-
logy, environmental predispositions, and therapy. Lancet Infect Dis
2:281–292.
Klaskala W, Brayfield BP, Kankasa C, Bhat G, West JT, Mitchell CD,
Wood C. 2005. Epidemiological characteristics of human herps-
virus-8 infectio in a large population of antenatal women in Zambia.
J Med Virol 75:93–100.
Mbulaiteye SM, Pfeifer RM, Whitby D, Brubaker GR, Shao J, Biggar
RJ. 2003. Human herpesvirus 8 infection within families in rural
Tanzania. J Infect Dis 187:1780–1785.
Mbulaiteye SM, Biggar RJ, Pfeifer RM, Bakaki PM, Gamache C, Owor
AM, Katongole-Mbidde E, Ndugwa CM, Goedert JJ, Whitby D,
Engels EA. 2005. Water, socioeconomic factors, and human
herpesvirus 8 infection in Ugandan children and their mothers.
J Acquir Immune Defic Syndr 38:474–479.
Ministe
´rio da Sau
´de. 2006. Repu
´blica de Moc¸ambique. Relato
´rio Anual
2006. PNC ITS/HIV/SIDA –Aconselhamento e Testagem em
Sau
´de.
Ministe
´rio da Sau
´de. 2007. Repu
´blica de Moc¸ambique. Boletim
DNAM –Direcc¸a
˜o Nacional de Assiste
ˆncia Me
´dica, N
o
2–
Vol. III. Julho 2007.
Mohanna S, Portillo JA, Carriquiry G, Vidal J, Ferrufino JC, Sanchez J,
Bravo F, Gotuzzo E. 2007. Human herpesvirus-8 in Peruvian blood
donors: A population with hyperendemic disease? Clin Infect Dis
44:558–561.
Moore PS, Chang Y. 1995. Detection of herpesvirus-like DNA
sequences in Kaposi’s sarcoma in patients with and without HIV
infection. N Engl J Med 332:1181–1185.
Nascimento MC, Souza VA, Sumita LM, Freire W, Weiss H, Sabino EC,
Franceschi S, Panutti CS, Mayaud P. 2008. Prevalence of, and risk
factors for Kaposi’s sarcoma-associated herpesvirus infection
among blood donors in Brazil: A multi-center serosurvey. J Med
Virol 80:1202–1210.
Newton R, Ziegler J, Bourboulia D, Casabonne D, Beral V, Mbidde E,
Carpenter L, Reeves G, Parkin DM, Wabinga H, Mbulaiteye S, Jaffe
H, Weiss R, Boshoff C, Uganda Kaposi’s Sarcoma Study Group.
2003. The sero-epidemiology of Kaposi’s sarcoma-associated
herpesvirus (KSHV/HHV-8) in adults with cancer in Uganda. Int
J Cancer 103:226–232.
Plancoulaine S, Abel L, Tre
´goue
¨t D, Duprez R, van Beveren M,
Tortevoye P, Froment A, Gessain A. 2004. Respective roles
of serological status and blood specific antihuman
herpesvirus 8 antibody level in human herpesvirus 8 intrafamilial
transmission in a highly endemic area. Cancer Res 64:8782 –
8787.
Silva JMK, Magri MC, Jacob F, Santos-Fortuna E, Caterino-de-Araujo
A. 2007. Otimizac¸a
˜o e avaliac¸a
˜o do ensaio de imunofluoresce
ˆncia
indireta para pesquisa de anticorpos anti-herpesvı
´rus humano 8
(HHV-8) usando sangue colhido em papel de filtro. Rev Inst Adolfo
Lutz 6:68–72.
Sissolak G, Mayaud P. 2005. AIDS-related Kaposi’s sarcoma:
Epidemiological, diagnostic, treatment and control aspects in sub-
Saharan Africa. Trop Med Int Health 10:981 –992.
Sitas F, Newton R. 2000. Kaposi’s sarcoma in South Africa. J Natl
Cancer Inst Monogr 28:1–4.
Sitas F, Carrara H, Beral V, Newton R, Reeves G, Bull D,
Jentsch U, Pacella-Norman R, Bourboulia D, Whitby D, Boshoff
C, Weiss R. 1999. Antibodies against human herpesvirus 8 in black
south African patients with cancer. N Engl J Med 340:1863 –
1871.
Szajerka T, Jablecki J. 2007. Kaposi’s sarcoma revisited. Aids Rev
9:230–236.
Wojcicki JM, Newton R, Urban MI, Stein L, Hale M, Patel M, Ruff P,
Sur R, Bourboulia D, Sitas F. 2004. Low socioeconomic status and
risk for infection with human herpesvirus 8 among HIV-1 negative,
South African black cancer patients. Epidemiol Infect 132:1191 –
1197.
J. Med. Virol. DOI 10.1002/jmv
Seroprevalence of HHV-8 Infection in Mozambique 1223