Age-of-consent requirements and adolescent HIV testing in low-and middle-income countries: multinational insights from 51 population-based surveys

Abstract

Background:

Pervasive social and structural barriers—including national policies—inhibit HIV testing uptake among priority populations, including adolescents. We assessed the relationship between age-of-consent policies for HIV testing and adolescent HIV testing coverage in 51 low- and middle-income countries.

Methods:

We pooled data from household surveys (2010–2020) and calculated the weighted country-level prevalence of lifetime HIV testing separately for adolescent girls and boys (ages 15–19). We then abstracted age-of-consent requirements for HIV testing across countries. Using multivariable linear regression, we estimated the average difference in national HIV testing coverage estimates for adolescent girls and boys by age-of-consent restrictions for HIV testing.

Results:

National HIV testing coverage estimates ranged from 0.7% to 72.5% among girls (median: 18.0%) and 0% to 73.2% among boys (median: 7.5%) in Pakistan and Lesotho, respectively. In adjusted models, HIV testing coverage in countries requiring parental consent for individuals <18 years was, on average, 9.4 percentage-points (pp) lower (95% confidence interval [95%CI] −17.9pp to −0.9pp) among girls and 9.3pp lower (95%CI: −17.3pp to −1.2pp) among boys, relative to countries with less restrictive policies (age-of-consent: ≤16 years). Compared to countries with less restrictive (age-of-consent: ≤14 years) policies, HIV testing prevalence was significantly lower among girls (β −10.5pp, 95%CI: −19.7pp to −1.3pp) and boys (β −10.5pp, 95%CI −19.2pp to −1.8pp) in countries with more restrictive (age-of-consent: 18 years) parental consent requirements.

Conclusions:

Age-of-consent policies are persistent obstacles to adolescent HIV testing. Repealing parental consent requirements for HIV testing is needed to expand coverage and accelerate progress towards global HIV treatment and prevention targets.

Background

Expanded access to testing remains imperative to close gaps in both the HIV prevention and treatment cascades. Global estimates suggest that knowledge of HIV serostatus impedes achievement of the UNAIDS 95–95–95 goals for HIV epidemic control.¹ A pooled multi-country study estimated that 31% of African adults have never tested for HIV, and 46% of Africans living with HIV are unaware of their HIV status.² Recent population-level estimates from high-burden settings show that these HIV testing coverage gaps persist, despite HIV testing service expansion and investments in novel testing modalities like self-testing.^3,4 The fraction of adolescents and young people (aged 15–24 years), in particular, who report ever testing for HIV ranges from 29% to 80% in 12 high-burden sub-Saharan African countries—falling well-below testing coverage thresholds required for HIV epidemic control.^5,6 Adolescents and young adults are a priority population in ending the AIDS pandemic, accounting for 29% of new HIV infections in 2019.⁷ Despite substantial reductions in AIDS mortality since the scale-up of antiretroviral therapy globally, HIV-related morbidities remain a leading cause of death among adolescents in sub-Saharan Africa.⁸ Identifying and remedying salient barriers to adolescent HIV testing is, therefore, an urgent public health priority.

Social and structural factors are widely acknowledged as forces inhibiting HIV testing uptake among adolescents globally. A systematic review found that among included studies, negative/judgmental attitudes from healthcare providers; HIV stigma and sanctioning of sexual behavior; suboptimal HIV knowledge and low perceived HIV risk; caregiver involvement; and facility inaccessibility (i.e. distance, inconvenient hours, lack of youth-friendly services) emerged as key barriers to adolescent HIV testing.⁹ Legal age-of-consent for HIV testing is another critical, albeit understudied, dimension of adolescent HIV testing uptake. Qualitative studies have explored how misunderstandings of age-of-consent requirements or inconsistencies in their implementation constrain HIV testing accessibility for adolescents.^10–13 To our knowledge, only one study has quantitatively examined the relationship between parental consent policies with adolescent HIV testing, finding that less restrictive age-of-consent requirements were associated with higher rates of adolescent HIV testing.¹⁴ However, this study was restricted to sub-Saharan African countries—the results of which may not be generalized to other low- and middle-income countries with different HIV epidemic profiles.

Furthermore, the COVID-19 pandemic has resulted in unprecedented disruptions to HIV prevention and treatment services, interrupting delivery of community-based HIV testing platforms in many settings.^15–17 These service disruptions are likely to have a disproportionate impact on adolescents, who are more likely to engage with HIV testing services when delivered in community settings.^18,19 Considering interruptions to vital HIV testing services precipitated by the COVID-19 pandemic, addressing persistent barriers to adolescent HIV testing, including age-of-consent requirements, is of paramount importance. To that end, we assessed the relationship between age-of-consent policies for HIV testing and adolescent HIV testing prevalence in 51 low-and middle-income countries.

Methods

Design

Between 2010 and 2020, we identified nationally representative household surveys implemented in low- and middle-income countries that estimated HIV testing coverage among adolescents. These surveys included Demographic and Health Surveys (DHS), AIDS Indicator Surveys (AIS), and Population-Based HIV Impact Assessments (PHIA). Across surveys, two-stage sampling proportional to population size was used to identify households in standard enumeration areas, from which individuals aged 15 years and older were recruited for survey completion. In countries where multiple surveys were conducted over the observation period, we abstracted data from the most recent survey (Table 1).

Table 1.

National age-of-consent requirements for HIV testing, adult HIV prevalence, and survey used in analysis (N = 51).

Country	Age of consent for HIV testing	Adult HIV prevalence	Survey (Year)
Afghanistan	None	0.1	2015 DHS
Cameroon	None	3.2	2018 DHS
Comoros	None	0.1	2012 DHS
Myanmar	None	0.1	2015–16 DHS
Eswatini	Age 12	28.0	2016–17 PHIA
Rwanda	Age 12	2.7	2019–20 DHS
Armenia	Age 14	0.2	2015–16 DHS
Benin	Age 14	1.0	2017–18 DHS
Cambodia	Age 14	0.7	2014 DHS
Colombia	Age 14	0.4	2015 DHS
Congo	Age 14	3.2	2011–12 DHS
Gabon	Age 14	3.8	2012 DHS
Gambia	Age 14	1.8	2019–20 DHS
Lesotho	Age 14	23.2	2016–17 PHIA
Liberia	Age 14	1.1	2019–20 DHS
Malawi	Age 14	9.4	2015–16 DHS
Mozambique	Age 14	11.4	2015 AIS
Namibia	Age 14	12.9	2017 PHIA
Senegal	Age 14	0.4	2017 DHS
South Africa	Age 14	17.1	2016 DHS
Togo	Age 14	2.5	2013–14 DHS
Uganda	Age 14	6.1	2016–17 PHIA
Burundi	Age 16	1.2	2016–17 DHS
Cote d’Ivoire	Age 16	2.5	2017–18 PHIA
Dominican Republic	Age 16	1.0	2013 DHS
Guinea	Age 16	1.4	2018 DHS
Haiti	Age 16	1.9	2016–17 DHS
Nepal	Age 16	0.1	2016 DHS
Niger	Age 16	0.3	2012 DHS
Tanzania	Age 16	5.2	2016–17 PHIA
Zambia	Age 16	11.8	2018 DHS
Zimbabwe	Age 16	12.6	2020 PHIA
Albania	Age 18	0.1	2017–18 DHS
Angola	Age 18	1.8	2015–16 DHS
Burkina Faso	Age 18	1.0	2010 DHS
Chad	Age 18	1.3	2014–15 DHS
Democratic Republic of the Congo	Age 18	0.9	2013–14 DHS
Ethiopia	Age 18	1.0	2017–18 PHIA
Ghana	Age 18	1.8	2014 DHS
Guatemala	Age 18	0.3	2014–15 DHS
Honduras	Age 18	0.4	2011–12 DHS
India	Age 18	0.3	2015–16 DHS
Kenya	Age 18	5.1	2014 DHS
Kyrgyz Republic	Age 18	0.1	2012 DHS
Maldives	Age 18	0.1	2016–17 DHS
Mali	Age 18	1.0	2018 DHS
Nigeria	Age 18	1.3	2013 DHS
Pakistan	Age 18	0.1	2017–18 DHS
Papua New Guinea	Age 18	0.8	2016–18 DHS
Sierra Leone	Age 18	1.5	2019 DHS
Timor-Leste	Age 18	0.1	2016 DHS

AIS: AIDS Indicator Survey; DHS: Demographic and Health Survey; PHIA: Population-Based HIV Impact Assessment.

For each country, we abstracted the weighted percentage of lifetime HIV testing (i.e. ever tested for HIV by self-report) among adolescent girls and boys (ages 15–19). We also abstracted the presence or absence of age-of-consent requirements for HIV testing across countries from national legal frameworks reported to the Joint United Nations Programme on HIV/AIDS (UNAIDS).²⁰ For each country, we determined whether parental consent requirements were present and the age-of-consent for HIV testing stipulated in these policies (none, age 12, age 14, age 16, or age 18).

Other country-level covariates included HIV epidemic type (generalized/mixed or concentrated, defined using a threshold of 1% adult HIV prevalence)²¹; World Bank income level (low, low-middle, or upper-middle)²²; per-capita health expenditure (in USD$ per 100,000 population)²³; and UNAIDS region (Asia-Pacific, East and Southern Africa, Eastern Europe and Central Asia, Latin America, or West and Central Africa). We abstracted country-level estimates from the most recent survey year.

Statistical analysis

We first calculated descriptive sample statistics for each country. Because our outcome measure captured adolescent HIV testing prevalence (only in the aggregate), we collapsed parental consent policies into a single reference group (“age 14 or younger”) for countries without age-of-consent requirements or stipulating age 14 or age 12 as the age-of-consent for HIV testing. To estimate associations of parental consent requirements with adolescent HIV testing uptake, we used multivariable linear regression to calculate average (mean) differences in national HIV testing coverage for adolescent girls and boys, separately, by age-of-consent requirements. We modeled age-of-consent requirements in two ways: dichotomously, using a cutpoint of 18 years (age-of-consent ≤16 years versus 18 years) and 16 years (age-of-consent ≤14 years versus ≥16 years), respectively; and categorically (age-of-consent ≤14 years, 16 years, versus 18 years). Across statistical models, our adjustment set included epidemic type, income level, health expenditure, and region—all of which were theorized to confound the relationship between age-of-consent policies and adolescent HIV testing uptake. We subsequently stratified regression models by region to identify heterogeneities in the effect of parental consent laws on adolescent HIV testing. We report regression coefficients with 95% confidence intervals (95%CI) and p-values. We managed and analyzed data in Stata/IC 15.1 (StataCorp LLC, College Station, Texas).

Results

National adolescent HIV testing coverage estimates varied widely, ranging from 0.7% (Pakistan) to 72.5% (Lesotho) among 15–19-year-old girls and 0% (Pakistan) to 73.2% (Lesotho) among 15–19-year-old boys. Median national HIV testing estimates were 18.0% among girls and 7.5% among boys. HIV testing coverage was greater in countries with less restrictive age-of-consent policies (age 14 or younger) (median 23.6% among girls and 11.8% among boys) compared to countries with stipulating age 16 (median 23.3% among girls and 11.2% among boys) or age 18 (median 10.0% among girls and 3.9% among boys) as the age-of-consent for HIV testing (Figure 1).

Figure 1.

HIV testing coverage by age-of-consent requirements for HIV testing.

Countries where the age-of-consent for HIV testing was 18 years had, on average, 9.4 percentage-point (pp) lower HIV testing coverage among girls (95%CI: −17.9 to −0.9; p=0.032) and 9.3pp lower coverage among boys (95%CI: −17.3 to −1.2, p=0.026), relative to countries where the age-of-consent was 16 years or younger (Table 2). Countries where the age-of-consent for HIV testing was at least 16 years were associated with 8.2pp lower HIV testing coverage among boys (95%CI: −6.0 to −0.5, p=0.038) relative to countries where the age-of-consent was 14 years or younger (association was not significant for girls).

Table 2.

Associations of age-of-consent policies with HIV testing coverage for adolescent girls and boys, by age-of-consent policy (N = 51).

	Adolescent girls			Adolescent boys
	β	95%CI	p-value	β	95%CI	p-value
Model 1a
≤ 16 years	1.00	—	—	1.00	—	—
18 years	−9.4	−17.9 to −0.9	0.032	−9.3	−17.3 to −1.2	0.026
Model 2b
≤ 14 years	1.00	—	—	1.00	—	—
≥ 16 years	−8.0	−16.2 to 0.23	0.056	−8.2	−16.0 to −0.5	0.038
Model 3c
≤ 14 years	1.00	—	—	1.00	—	—
16 years	−3.6	−14.6 to 7.3	0.507	−4.3	−14.6 to 6.1	0.414
18 years	−10.5	−19.7 to −1.3	0.026	−10.5	−19.2 to −1.8	0.019

^aNotes: Model one compares HIV testing prevalence in countries with an age-of-consent requirement of 18 years to countries with age-of-consent requirements age 16 years or younger (referent).

^bModel two compares HIV testing prevalence in countries with age-of-consent requirements of 16 years or older to countries with age-of-consent requirements age 14 years or younger (referent).

^cModel three compares HIV testing prevalence in countries with an age-of-consent requirement of 18 years and 16 years, respectively, to countries with age-of-consent requirements age 14 years or younger (referent). All models adjusted for epidemic type (generalized/mixed or concentrated), World Bank income level (low, low-middle, or upper-middle), per-capita health expenditure, and UNAIDS region (Asia-Pacific, East and Southern Africa, Eastern Europe and Central Asia, Latin America, or West and Central Africa).

Bolded values represent statistically significant beta coefficients at the p < 0.05 level.

When modeling policy as a categorical variable, countries with more restrictive (age-of-consent: 18 years) HIV testing policies were associated with 10.5pp lower HIV testing coverage among girls (95%CI: −19.7 to −1.3, p=0.026) and 10.5pp lower HIV testing coverage among boys (95%CI: −19.2 to −1.8, p=0.019) relative to countries with the least restrictive policies (age-of-consent: 14 years or younger). There was no significant effect for girls or boys comparing countries where the age-of-consent for HIV testing was 18 years and 16 years.

Table 3 presents unadjusted estimates for the effect of age-of-consent policies on adolescent HIV testing, stratified by region. Due to small sample sizes, no significant associations between age-of-consent policies and adolescent HIV testing were observed, except among adolescent boys in West and Central African countries (β −5.5pp, 95%CI: −10.6 to −0.5, p=0.033). Nevertheless, the magnitude of effect estimates varied by region (girls: −1.9pp in Asia-Pacific to −10.8pp in East and Southern Africa; boys: −2.8pp in Asia-Pacific to −13.6pp in East and Southern Africa).

Table 3.

Region-specific associations of age-of-consent policies (18 years ≤16 years) with HIV testing coverage for adolescent girls and boys, by age-of-consent policy (N = 51).

Region	N	Adolescent girls			Adolescent boys
		β	95%CI	p-value	β	95%CI	p-value
Asia-Pacific	9	−1.9	−9.7 to 5.9	0.583	−2.8	−7.1 to 1.4	0.157
East and Southern Africa	15	−10.8	−35.1 to 13.3	0.349	−13.6	−39.2 to 12.0	0.273
Latin America and the Caribbean	9	−5.5	−21.1 to 10.1	0.346	−5.1	−12.1 to 1.9	0.101
West and Central Africa	15	−7.4	−16.1 to 1.3	0.091	−5.5	−10.6 to −0.5	0.033

Notes: Only unadjusted effect estimates are reported due to small sample sizes in each region. Eastern Europe and Central Asia were excluded from stratified analysis due to only three observations. Across stratified models, countries with age-of-consent policies for adolescents ≤16 years served as the referent group. Bolded values represent statistically significant beta coefficients at the p<0.05 level.

Discussion

In this study using national data from 51 low- and middle-income countries, we found that more restrictive age-of-consent policies for HIV testing were associated with lower levels of national HIV testing coverage among adolescent girls and boys. The magnitude of these effects varied by region, albeit non-significantly—indicating that the implementation of these age-of-consent policies likely have heterogeneous impacts on adolescent HIV testing globally. These findings suggest that age-of-consent requirements may present obstacles to HIV testing among adolescents, and interventions are urgently needed to remove legal barriers to HIV services.

It is unsurprising that, overall, rates of adolescent HIV testing were higher among adolescent girls than boys. Especially in countries with generalized HIV epidemics, HIV testing has been well-integrated into antenatal care, which represents a key HIV testing platform for women.²⁴ This may partially help explain discrepancies in HIV testing between adolescent girls and boys, although adolescent pregnancy rates have fallen drastically in the last two decades in many settings.²⁵ Although not a focus of this study, emancipation exemptions to age-of-consent policies in the context of adolescent marriage and/or pregnancy may reinforce the disproportionate HIV risks and HIV service-seeking responsibilities that women and adolescent girls shoulder globally. Gendered expectations and norms surrounding healthcare-seeking, including HIV testing, could also explain observed heterogeneities in HIV testing coverage among adolescent girls and boys.^26–28

Individual-level interventions such as creating demand for HIV testing among adolescents and their caregivers, reducing HIV-related stigma, and expansion of youth-friendly HIV testing services could mitigate the harmful consequences of these age-of-consent requirements.⁹ Unlike a prior study assessing associations between age-of-consent policies and HIV testing among adolescents in sub-Saharan Africa,¹⁴ we found similar effects of age-of-consent policies among both boys and girls, though boys had lower average rates of HIV testing across policy environments. Interventions that focus on increasing HIV testing for adolescent boys—including HIV self-testing, male-specific programming, and social and behavior change communication—should, therefore, be prioritized across policy contexts.^29–31 Ultimately, in order to meaningfully expand coverage and ensure equity of HIV testing for adolescents, health systems interventions, specifically repealing parental consent policies, should be considered alongside other efforts to increase HIV testing among adolescents.

Findings from this study should be viewed in light of various limitations. First, due to multicollinearity with HIV epidemic type and adolescent HIV testing, lifetime adult HIV testing prevalence could not be included in multivariable analysis without destabilizing statistical models, particularly given the small sample size. Second, information on age-of-consent policies for HIV testing were agnostic to emancipation exemptions for adolescent girls in the context of marriage, pregnancy, and antenatal care. Future policy analyses should identify documented exemptions to HIV testing age-of-consent policies for pregnant and/or married adolescent girls. Third, this analysis used cross-sectional, nationally aggregated data, so we were unable to examine factors associated with HIV testing behavior at the individual level or secular trends in HIV testing over time. Fourth, we examined the association between HIV testing and presence of age-of-consent policies; this does not account for differential design, implementation, or enforcement of policies within or between countries, which likely impact HIV testing uptake.³² Fifth, data abstracted from household surveys were collected by self-report and may, therefore, be susceptible to response biases. Sixth and finally, countries self-reported age-of-consent requirements to UNAIDS, which could induce selection bias if countries with less restrictive HIV testing requirements were more likely to communicate age-of-consent policy information to multilateral bodies

Conclusions

Age-of-consent requirements are persistent obstacles to adolescent HIV testing, with more restrictive policies associated with decreased HIV testing coverage among adolescent girls and boys. In addition to individual-level interventions targeted at increasing testing uptake in these groups, repealing parental consent requirements to HIV testing is needed to expand and accelerate progress towards global HIV treatment and prevention targets.