RESEARCH ARTICLE
Predictors of HIV Testing among Youth in
Sub-Saharan Africa: A Cross-SectionalStudy
Ibitola O. Asaolu
1
, Jayleen K. Gunn
2
, Katherine E. Center
3
, Mary P. Koss
1
, Juliet
I. Iwelunmor
4
, John E. Ehiri
1
*
1 Department of Health Promotion Sciences, Mel and Enid Zuckerman College of Public Health, University
of Arizona, 1295 N Martin Avenue, Tucson, Arizona, 85724, United States of America, 2 Department of
Epidemiology and Biostatistics, Mel and Enid Zuckerman College of Public Health, University of Arizona,
1295 N Martin Avenue, Tucson, Arizona, 85724, United States of America, 3 Department of Obstetrics and
Gynecology, University of Arizona, 1501 N Campbell Ave, Tucson, Arizona, 85724, United States of
America, 4 Department of Kinesiology and Community Health, University of Illinois at Urbana Champaign,
1206 S. 4th Street, Champaign, Illinois, 61820, United States of America
* jehiri@email.arizona.edu
Abstract
Introduction
In spite of a high prevalence of HIV infection among adolescents and young adults in sub-
Saharan Africa, uptake of HIV testing and counseling among youth in the region remains
sub-optimal. The objective of this study was to assess factors that influence uptake of HIV
testing and counseling among youth aged 15–24 years in sub-Saharan Africa.
Methods
This study used the Demographic and Health Survey (DHS) data from countries that repre-
sent four geographic regions of sub-Saharan Africa: Congo (Brazzaville), representing cen-
tral Africa (DHS 2011–2012); Mozambique, representing southern Africa (DHS 2011);
Nigeria, representing western Africa (DHS 2013); and Uganda, representing eastern Africa
(DHS 2011). Analyses were restricted to 23,367 male and female respondents aged 15–24
years with complete data on the variables of interest. Chi-square tests and logistic regres-
sion models were used to assess predictors of HIV testing. Statistical significance was set
at p< 0.01.
Results
The analysis revealed that a majority of the respondents were female (78.1%) and aged
20-24-years (60.7%). Only a limited proportion of respondents (36.5%) had ever tested for
HIV and even fewer (25.7%) demonstrated comprehensive knowledge of HIV/AIDS. There
was a significant association between HIV testing and respondents’ gender, age, age at
sexual debut, and comprehensive knowledge of HIV in the pooled sample. Older youth
(adjusted OR (aOR) = 2.19; 99% CI = 1.99–2.40) and those with comprehensive knowl-
edge of HIV (aOR = 1.98; 1.76–2.22) had significantly higher odds of ever being tested for
PLOS ONE | DOI:10.1371/journal.pone.0164052 October 5, 2016 1 / 12
a11111
OPEN ACCESS
Citation: Asaolu IO, Gunn JK, Center KE, Koss MP,
Iwelunmor JI, Ehiri JE (2016) Predictors of HIV
Testing among Youth in Sub-Saharan Africa: A
Cross-Sectional Study. PLoS ONE 11(10):
e0164052. doi:10.1371/journal.pone.0164052
Editor: Claire Thorne, UCL Institute of Child Health,
University College London, UNITED KINGDOM
Received: April 12, 2016
Accepted: September 19, 2016
Published: October 5, 2016
Copyright: © 2016 Asaolu 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
author and source are credited.
Data Availability Statement: The data may be
found as follows: Congo (Brazzaville, DHS 2011-
2012): http://dhsprogram.com/what-we-do/survey/
survey-display-388.cfm; Mozambique (DHS 2011):
http://dhsprogram.com/what-we-do/survey/
survey-display-362.cfm; Nigeria (DHS 2013):
http://dhsprogram.com/what-we-do/survey/
survey-display-438.cfm; Uganda (DHS 2011):
http://dhsprogram.com/what-we-do/survey/
survey-display-399.cfm.
Funding: The authors received no financial support
(external or internal) for this study.
HIV than younger respondents and those with limited HIV/AIDS knowledge respectively.
Furthermore, men had lower odds of HIV testing than women (aOR = 0.32; 0.28–0.37).
Conclusions
Reaching youth in sub-Saharan Africa for HIV testing continues to be a challenge. Public
health programs that seek to increase HIV counseling and testing among youth should pay
particular attention to efforts that target high-risk subpopulations of youth. The results fur-
ther suggest that these initiatives would be strengthened by including strategies to increase
HIV comprehensive knowledge.
Introduction
Globally, significant progress has been made in efforts to address the burden of the Human
Immuno-Deficiency Virus (HIV) and Acquired Immune-DeficiencySyndrome (AIDS), lead-
ing to a 38% reduction in new infections since 2001 [1]. In spite of this remarkable progress,
HIV/AIDS remains a major global health problem [1]. According to the United Nations Joint
Programme on HIV/AIDS (UNAIDS), 2.1 million people were newly infected with HIV in
2013, and an estimated 35 million people were living with the virus [2]. Specifically, there has
been a decreasing trend in incidence of HIV infection among adolescents in many countries.
After over three decades since the first case of HIV was discovered, there is evidence that over-
all, knowledge about the virus infection is increasing, leading to a trend towards less risky sex-
ual health behaviors, and thus reduction in new infections [3]. In spite of this, HIV remains
particularly devastating for adolescents and young adults aged 15–24 years who account for
approximately 50% of all new HIV infections and 33% of persons living with HIV/AIDS world-
wide [4]. Globally, there is an estimated 1.2 billion adolescents (10–19 year-olds), constituting
18% of the world’s population [5]. Available evidenceshows that about 2.2 million of these
(60% of them, female) are living with HIV [5], and many are unaware of their infection [6].
Adolescence and emerging adulthood constitute a period of transition, growth, exploration,
and opportunities. During this phase of life, young people develop an increased interest in sex,
with concomitant risks for sexually transmitted infections (STIs), including HIV. Furthermore,
adolescenceand emerging adulthood is known as a stage of development that is characterized
by a high sense of invulnerability [7], which is often associated with increased risk taking [8, 9].
A study by Anderson and colleagues found an inverse relationship betweenperceived risk of
HIV and sexual debut; youth with high levels of perceivedrisk of HIV had lower odds of having
sexual intercourse compared to those with low perceived risk of HIV [9].
Sub-Saharan Africa is the region where persons aged 10–19 years constitute the largest pro-
portion of the population [5]. Youth in this region are particularlyat a heightened risk for HIV
infection with 2.2% of young women and 1.1% of young men living with HIV as of 2013 [1].
According to available evidence,only 10% of young men and 15% of young women aged 15–
24 years in sub-Saharan Africa knew their HIV status in 2013 [1]. Barriers to HIV testing
include lack of awareness of available services,low perception of personal risk, fear of negative
consequencesassociatedwith a positive test result (including stigma), concerns about confi-
dentiality, financial burden of testing, and lack of HIV/AIDS knowledge[10–12]. These barri-
ers have contributed to generations of individuals whose lives, dreams, and potential
contribution to economic development have been undermined by a largely preventable disease.
Young women in sub-Saharan Africa are particularly vulnerable to HIV infection;of all young
Predictors of HIV Testing among Youth in Sub-Saharan Africa: A Cross-Sectional Study
PLOS ONE | DOI:10.1371/journal.pone.0164052 October 5, 2016 2 / 12
Competing Interests: The authors have declared
that no competing interests exist.
women aged 15–24 living with HIV globally, 80% of them reside in sub-Saharan Africa [1, 9–
12]. In many parts of sub-Saharan Africa, the HIV infection risk for adolescent girls is 2–6
folds higher than that of their male counterparts by the time they are aged 20–24 years [1]. The
higher prevalence of HIV infection among young women is associated with lower levels of
accurate HIV knowledge, age-disparate sexual relationships, gender-based violence, poverty,
certain cultural practices, and early sexual debut [10–14].
Although it is recognized that this population has different needs from those of children
and adults, there are few global initiatives that are specifically devoted to the promotion of HIV
counseling, testing, and linkage to care among adolescents and young adults. The problem of
limited initiatives is of major global concern because HIV counseling and testing is an impor-
tant entry point for most HIV/AIDS care and prevention interventions [14]. Given the wider
availability of antiretroviral therapy, efforts to increase uptake of HIV counseling and testing
among youth have the potential to lead to early diagnosis, prompt linkage to care, improved
treatment outcomes, and reduction in new infections [4].
It is important to understand and address the challenges that young people in sub-Saharan
Africa face in accessing HIV counseling and testing, with the hope that such an understanding
might help inform future research, practice, and policy regarding interventions that address
young people’s unique barriers to HIV counseling and testing. The objective of this study was
to assess factors that influenceuptake of HIV testing among youth aged 15–24 years in sub-
Saharan Africa.
Methods
Survey
This study used the Demographic and Health Survey (DHS) data from four countries that rep-
resent unique regions of sub-Saharan Africa: Congo (Brazzaville, DHS 2011–2012), represent-
ing Central Africa; Mozambique (DHS 2011), representing Southern Africa; Nigeria (DHS
2013), representing Western Africa; and Uganda (DHS 2011), representing Eastern Africa. The
DHS is a nationally representative survey that uses a multistage and stratified design to collect
information on population health, HIV/AIDS, malaria, and nutrition within each country [15].
The individual women’s and men’s data of the DHS dataset were used in this analysis. Only
youth, defined by the United Nations “as individuals between the ages of 15 and 24 years” [16]
were included in the analysis.
Measures
The variable assessing comprehensive knowledge of HIV was created using UNICEF’s defini-
tion of comprehensive HIV knowledge [17]. Participants were classified as having comprehen-
sive HIV knowledge if they correctly:1) identified two methods of preventing sexual
transmission of HIV; 2) acknowledgedthat a healthy looking person can have HIV; and 3)
rejected two common misconceptions about HIV transmission, i.e., HIV can be transmitted
through mosquito bites or by sharing food with an HIV infected person. The variable on hav-
ing a history of sexually transmitted infection (STI) was classified as ‘yes’ if a respondent
answered ‘yes’ to any of the following items: had any STI in last 12 months; had genital sore/
ulcer in last 12 months; or had abnormal or bad smelling genital discharge in last 12 months
[18]. History of STI was categorized as ‘no’ if the respondent answered ‘no’ to these questions.
Urban versus rural residence (considered as a system-level predictor of HIV testing) was used
as a proxy for access to health care facilities.
Predictors of HIV Testing among Youth in Sub-Saharan Africa: A Cross-Sectional Study
PLOS ONE | DOI:10.1371/journal.pone.0164052 October 5, 2016 3 / 12
Statistical Analysis
Due to oversampling of certain populations, country-specificindividual weights were used in
all analyses. Using weights makes the data more representative of the study population on a
national level [15]. Chi-square tests and logistic regression models were used to assess predic-
tors of HIV/AIDS testing. Statistical significancewas reduced from p < 0.05 to p < 0.01 to con-
trol for type-1 error emanating from the large sample size of the DHS data. All data analyses
were conducted on SAS 9.4 (Cary, North Carolina) using the ‘PROC SURVEY’ command. The
Demography and Health Surveys were approved by the Institutional ReviewBoard of ORC
Macro. In addition, this secondary data-analysis was reviewedand approved as exempt by the
Institutional Review Board of the University of Arizona. Participants' information were anon-
ymized and de-identifiedprior to analysis.
Results
Pooled Sample
The analytic sample consisted of 23, 367 respondents including the following: 4,482 (19.2%)
Congolese;5,301 (22.7%) Mozambicans; 10,942 (46.8%) Nigerians; and 2,642 (11.3%) Ugan-
dans. The demographic characteristicsof respondents are presented in Table 1. The total sam-
ple consisted predominantly of females (78.2%) and youth aged 20 to 24 (60.7%). About half
(50.2%) of the respondents had at least a secondary school education, and slightly more than
half (58.1%) lived in rural areas. About 3 out of 5 respondents (62%) had no sexual partner
other than their current partner or spouse, within the past year and did not report symptoms
suggestiveof an STI (88.1%) in the 12 months preceding the survey. Few (25.7%) demonstrated
comprehensive knowledge of HIV/AIDS, and only a limited proportion (36.5%) had ever been
tested for HIV.
Table 2 presents results of the relationship between individual and system-level factors and
uptake of HIV testing and counseling among youth in the study countries. HIV testing was sig-
nificantly associated with respondents’ gender (p< 0.0001), age (p< 0.0001), age at sexual
debut (p<0.0001), comprehensive knowledge of HIV (p<0.0001), number of sexual partners
in the past year (p<0.0001), and history of STIs (p<0.0001). Male youth had a lower propor-
tion (23.1%) of HIV testing than female youth (40.3%). Similarly, fewer youth (26.5%) aged 15
to 19 years tested for HIV than those aged 20–24 years (40.3%). More youth with comprehen-
sive knowledgeof HIV (50.3%) tested for HIV compared to those without comprehensive
knowledgeof the virus (31.8%). Among youth who reported a history of STI, 49.3% had tested
for HIV compared to 34.8% of youth with no history of STI.
Results from adjusted logistic regression models are presented in Table 3. Associations
between respondents’ characteristicsand HIV testing were similar across all countries and
within the pooled sample. Positive associations were evident betweenHIV testing and the fol-
lowing variables: age, age at sexual debut, and comprehensive knowledge of HIV. Youth aged
20–24 years had higher odds of HIV testing than youth aged 15–19 years (adjusted OR
(aOR) = 2.19; 99% (1.99–2.40)). Compared to youth with age at sexual debut of less than 15
years, those who were older at sexual debut (15–18 years, or 19 years and above) had higher
odds of HIV testing (aOR = 1.22; 1.07–1.39 and aOR = 1.47; 1.24–1.74 respectively).Also,
those with comprehensive knowledge of HIV had higher odds of being tested for HIV com-
pared to youth without comprehensive HIV knowledge (aOR = 1.98;1.76–2.22). Male youth
had lower odds of being tested than female youth (aOR = 0.32; 0.28–0.37).
Predictors of HIV Testing among Youth in Sub-Saharan Africa: A Cross-Sectional Study
PLOS ONE | DOI:10.1371/journal.pone.0164052 October 5, 2016 4 / 12
Table 1. Demographic and health characteristics of participants.
CONGO MOZAMBIQUE NIGERIA UGANDA TOTAL
n
a
%
b
n
a
%
b
n
a
%
b
n
a
%
b
n
a
%
b
Sex of Respondents
Male 1228 28.3 1184 22.2 2256 19.6 521 18.7 5189 21.8
Female 3254 71.7 4117 77.8 8686 80.4 2121 81.3 18178 78.2
Age of Respondents
15 to 19 years 2148 43.1 2449 46.1 3788 34.6 1000 38.0 9385 39.3
20–24 years 2334 58.9 2852 53.9 7154 65.4 1642 62.0 13982 60.7
Age at Sexual Debut
Less than 15 years 1502 27.5 1492 28.4 2433 23.6 544 21.6 5971 22.3
15–18 years 2821 67.3 3509 65.5 6274 56.5 1724 65.1 14328 61.6
19 years and above 159 5.2 300 6.1 2235 19.9 374 13.3 3068 13.1
Area of Residence
Urban 1642 71.5 2438 38.4 3731 34.9 950 25.2 8761 41.9
Rural 2840 28.5 2863 61.6 7211 65.1 1692 74.8 14606 58.1
Education
None 226 3.2 689 15.7 3527 36.0 151 3.8 4593 21.4
Primary 1262 20.7 2617 52.3 1439 12.9 1420 57.1 6738 28.4
Secondary or higher 2994 76.1 1995 31.0 5976 51.1 1071 39.1 12036 50.2
Wealth Index
Poorest 1483 14.0 649 17.8 1893 19.4 443 15.3 4468 17.5
Poorer 1240 20.3 750 17.7 2347 22.0 424 17.6 4761 20.2
Middle 696 22.4 890 17.8 2429 20.7 431 19.1 4446 20.2
Richer 592 22.6 1150 20.7 2403 20.1 483 19.4 4628 20.7
Richest 471 20.7 1862 26.0 1870 17.8 861 28.6 5064 21.4
Comprehensive Knowledge of HIV
c
No 3466 73.7 3799 74.4 8398 76.5 1781 66.4 17444 74.3
Yes 1016 26.3 1502 25.6 2544 23.5 861 23.5 5923 25.7
Number of Sexual Partners (excluding spouse) in the past 12 months
None 1893 40.9 2971 60.2 7226 69.8 1825 70.6 13915 62.0
1 or more 2589 59.1 2330 39.8 3716 30.2 817 29.4 9452 38.0
Sexually Transmitted Infection in Past 12 months
No 3745 81.0 4830 91.5 10140 92.6 2024 74.8 20739 88.1
Yes 737 19.0 471 8.5 802 7.4 618 25.2 2628 11.9
Ever-Tested for HIV
No 3228 68.6 2631 54.7 8184 75.3 600 22.5 14643 63.5
Yes 1254 31.4 2670 45.3 2758 24.7 2042 77.5 8724 36.5
Received Results of HIV Test
d
No 120 6.6 201 8.4 370 13.7 107 5.1 798 9.0
Yes 1134 93.4 2469 91.6 2388 86.2 1935 94.9 7926 91.0
TOTAL (N) 4482 5301 10942 2642 23367
Notes:
“a”
= Unweighted frequency of respondents
“b”
= Weighted percent
“c”
= Comprehensive knowledge assesses respondents’ understanding of HIV transmission and prevention.
“d”
= Adds up to the total number of people who tested for HIV
doi:10.1371/journal.pone.0164052.t001
Predictors of HIV Testing among Youth in Sub-Saharan Africa: A Cross-Sectional Study
PLOS ONE | DOI:10.1371/journal.pone.0164052 October 5, 2016 5 / 12