Approach

Our analysis is based on a review of published literature (with a focus on multicountry studies within sub-Saharan Africa), as well as the results of our synthesis of national representative Demographic and Health Surveys (DHS) data for 33 countries in sub-Saharan Africa with a most recent survey since 2010 (data extracted from StatCompiler) and national HIV surveys.6 7 The surveys were on average conducted in 2014. We computed the median ages at first sex, marriage and birth based on the recall of these events by women and men aged 25–29 years (or 30–34 years if the event median was 25 years or higher).

In order to obtain further insights into more recent trends, we also analysed primary DHS data for respondents 15–24 years in a life table analysis of reproductive events in 10 large-population countries with a survey around 2004, 2010 and 2015. The 10 countries—Ethiopia, Ghana, Kenya, Malawi, Mozambique, Nigeria, Tanzania, Uganda, Zambia and Zimbabwe—are part of the Countdown to 2030 for women’s, children’s and adolescents’ health (http://countdown2030.org/) regional initiative for data analysis in sub-Saharan Africa and were selected because they have populations over 15 million, together accounting for 53% of the total population in sub-Saharan Africa, and conducted three surveys in the relevant periods. The descriptions of the surveys, indicators and methods are shown in online supplementary file 1.

Key dimensions of inequality: who is left behind?

Gender: girls continue to bear the brunt

ASRH inequalities in sub-Saharan Africa for women are rooted within social, cultural and economic spheres that influence their future prospects, decision-making power and autonomy.8 Such gender inequalities are critical for ASRH throughout adolescence, even at ages 10–14 years,9 and influence an individuals’ ability to realise his/her sexual and reproductive health rights, especially girls.

Sexual initiation, which occurs at similar age range in sub-Saharan Africa as many other parts of the world,10 marks the beginning of exposure to a range of sexual and reproductive health risks such as early pregnancy and HIV infection. In the most recent DHS in 33 countries, the median ages at first sex were 17.4 and 18.4 years for girls and boys, respectively (figure 1 and table 1). There was widespread variation between countries, ranging from 16.0 to 21.5 years for females (Mozambique and Rwanda, respectively) and from 16.2 to 24.1 years for males (Angola and Senegal, respectively). The gender gap was much larger in West and Central Africa (2.0 years) than in Eastern and Southern Africa (0.3 years), mainly due to a younger age at first sex among girls in West and Central Africa. Our trend analysis for 10 countries showed a reduction of the gender gap for age at first sex during 2004–2015. The median age at first sex for girls increased from 17.4 to 17.7 years but a small decrease was observed for boys, reducing the gender gap to 0.2 years in the most recent surveys (table 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Median age at first sex, age at first marriage and age at first birth in 33 countries in sub-Saharan Africa (14 countries in Eastern and Southern and 19 countries in West and Central Africa), most recent DHS 2010–2018. Dash denotes median value of countries in subregion. Dot denotes country median age at first sex, marriage or birth for both men and women aged 25–29 years (or 30–34 years for men only if the median age at first marriage was 25 years or higher). AGO, Angola; BDI, Burundi; COM, Comoros; ETH, Ethiopia; GMB, Gambia; GAB, Gabon; GHA, Ghana; LSO, Lesotho; MWI, Malawi; MOZ, Mozambique; NER, Niger; RWA, Rwanda; SEN, Senegal; TCD, Chad.

View this table:

• View inline
• View popup

Table 1

Key indicators of adolescent sexual and reproductive health (country medians), 33 countries with most recent DHS since 2010*, sub-Saharan Africa by subregion

View this table:

• View inline
• View popup

Table 2

Trends in age at first sex, age at first marriage and age at first birth in years, median value from 10 countries, 2004–2015

Marriage patterns in sub-Saharan Africa are characterised by substantial age differences between the partners and high prevalence of child marriage among girls (before age 18).11 Child marriage, which is an SDG indicator, is not only a violation of a young person’s rights with potentially far-reaching consequences for ASRH and educational opportunities.12 While child marriage does occur among boys, in most countries it is far less common than among girls.13 In our analysis of the 33 countries the median ages at first marriage among women 25–29 and men 25–34 years were 19.4 and 24.9 years, respectively (figure 1 and table 1). The median male–female age at marriage gaps were 4.9 and 6.4 years in Eastern and Southern and West and Central Africa, respectively. Several studies have reported an increase in age at first marriage and decline in child marriage during the last decades.14 15 The reductions in child marriage, however, were not uniform, with several countries showing no change over time such as Chad and Niger where over 60% of all girls were married before the age of 18 years.14 In our analysis of 10 countries with recent data, age at first marriage increased by 0.7 years for both women and men during 2004–2015 (table 2).

The incidence and prevalence of HIV and other STIs among adolescents remains a key public health challenge in much of sub-Saharan Africa.16 17 There is a clear gender divide, with adolescent girls at much higher risk for HIV transmission than boys.16 18 19 Sexual transmission plays a key role, as paediatric infections (mother-to-child transmission during the pregnancy, delivery or post partum) are unlikely to cause such gender disparities. In addition to biological factors,20 the age difference between sexual partners is an important driver of the epidemic among adolescent girls and young women, because older male partners are more likely to be HIV infected than adolescent boys. Condom use may also be due to limited sexual negotiation power and skills of girls.21 In the most recent surveys in 30 countries with HIV testing, the country median HIV prevalence rate among girls 15–19 years was 1.2%, two times higher than for boys of the same age. There are also marked country and regional differences with HIV prevalence at 15–19 years Eastern and Southern Africa four times higher than in West and Central Africa (table 1).

Condom use for HIV/STI prevention is underpinned by complex gender and cultural dynamics within sexual relationships and is driven by the ability for both partners to openly communicate and negotiate protection options.22 In 32 countries with data from a recent survey in sub-Saharan Africa, nearly 38% of sexually active adolescent girls reported condom use at last sex with a non-marital non-cohabiting partner, compared with 51% of adolescent boys (table 1). Condom use among both adolescent boys and girls was considerably higher in Eastern and Southern Africa than in West and Central Africa, perhaps because of greater condom promotion efforts due to the more severe HIV epidemics.5

More education: later sex, marriage and childbearing

Educational attainment is strongly related to sexual and reproductive behaviours. Higher levels of school attainment are associated with later initiation of sex, marriage and childbirth (figure 2). Girls with less education (none or primary) initiated sex 2.2 years earlier, were married 4.4 years earlier and had their first child 2.5 years earlier than girls with secondary or higher education, based on 33 countries with recent data. Differences in sexual initiation were non-existent for boys, but more educated young men married 3.6 years later than those who were less educated. The gender gaps in age at first sex and marriage by educational status did not reduce much in the 10 countries with surveys around 2004 and 2015 (table 2). Yet, evidence shows that educational changes are an important driver of the overall increases in age at first sex and marriage.15 Decomposition analysis showed that this is primarily due to relatively more adolescent girls and boys reaching higher levels of education where later sex and marriage are more common, rather than changes within the educational groups.23

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Median age at first sex, age at first marriage and age at first birth for girls and boys, by level of education, urban–rural residence and household wealth, 33 countries in sub-Saharan Africa (14 countries in Eastern and Southern and 19 countries in West and Central Africa), most recent DHS 2010–2018. Data extracted from STATcompiler are available and reported by wealth quintile. DHS, Demographic and Health Survey.

Even though there has been an increase in contraceptive use among young single women in sub-Saharan Africa during the last two decades,5 family planning coverage (defined as demand satisfied for family planning with modern methods) among adolescent single girls was still only 38% in 32 recent surveys, with considerably higher levels in Eastern and Southern Africa compared with West and Central Africa. For married adolescents, family planning coverage was even lower (table 1). Attending school is associated with greater family planning coverage, but our analysis of trends in nine countries (no recent data for Mozambique) showed that the gap between those with lower and those with higher education reduced in the last decade.

In the initial years of the HIV epidemic, education was positively associated with HIV prevalence.24 Over time, education has become less important as HIV risk factor,25 presumably due to greater use of preventive measures (such as sexual partner reduction and condom use) by more educated women and men. A study of trends among people 15–24 years in seven African countries, however, did not find a consistent pattern or change in HIV prevalence by education during 2004–2011.24 Our analysis shows that the median HIV prevalence in 30 countries with surveys since 2010 was about the same among both sexes with secondary education or higher compared with those with less education in both subregions of sub-Saharan Africa (figure 3). In Eastern and Southern Africa, HIV prevalence was higher among more educated young women.­

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

HIV prevalence (%) among girls/women and boys/men 15–24 years, by residence, education level and wealth quintile (poorest vs richest), 28 countries in sub-Saharan Africa (12 countries in Eastern and Southern and 16 countries in West and Central Africa), most recent surveys 2010–2018. Online supplementary file: study populations and data sources, definitions, online supplementary tables 1–8 and online supplementary figures 1–3. Wealth quintile is used as a stratified for consistency and comparability as HIV prevalence data extracted from reports and population-based HIV impact assessemnt (PHIA) are available and reported by wealth quintile.

Urban–rural gap remains wide

A rapidly increasing proportion of adolescents are living in urban settings. According to data from the UN Population Division,26 the proportion of adolescents 10–19 years who were living in urban areas in sub-Saharan Africa reached at 38% in 2015 (28% in Eastern and Southern Africa and 48% in West and Central Africa), an increase from nearly 34% in 2005. Adolescents living in urban settings, especially those in low-income settings in metropolitan areas, are more likely to be exposed to greater sexual mixing and risks of HIV than their rural counterparts. At the same time, urban settings may offer better access to contraceptives including condoms, safe abortion and different social norms regarding marriage and fertility. Several studies have shown that adolescents living in urban areas have lower rates of adolescent pregnancy and early marriage, as well as far greater contraceptive use.27–29 Most analyses describe urban–rural differences as a dichotomous stratification, but recent analysis suggests that the urban poor often have equally poor indicators to their rural counterparts.30 Aggregated analysis of DHS data from 33 countries showed that rural girls started sex 1.3 years earlier and married 2.7 years earlier than urban girls (figure 2). There was no difference for boys in age at first sex, but rural men married 3.3 years earlier than urban men. Rural girls had their first birth almost 2 years earlier than urban girls. The urban–rural gaps in ASRH indicators have not reduced during the past decade, according to our in-depth analysis of 10 countries (table 2). In fact, while median ages at first sex, marriage and birth were increasing, the gaps between the urban and rural adolescents increased during 2004–2015 for both sexes. First births to women under the age of 20 years are becoming increasingly concentrated within rural populations.31

The coverage of family planning with modern methods differs greatly between urban and rural adolescents, but data from our 10 country analysis suggests that rural contraceptive use is increasingly faster than among urban populations. Family planning coverage among single sexually active girls 15–24 years increased from 34% to 49% during 2004–2015, reducing the gap with urban girls to almost none (51% in 2004 and 53% in 2015) (table 3). Among married adolescents and young women, there were large increases in family planning coverage during 2004–2015 but the absolute urban–rural gap remained the same (online supplementary figures 1–3).

In Eastern and Southern Africa, HIV prevalence is higher in urban than in rural areas, especially among women 15–24 years (medians 6.3% and 4.5%, respectively) (figure 3). In West and Central Africa, urban women also had a higher prevalence than their rural counterparts (medians 1.5% and 0.9%, respectively). It has also been shown that the urban poor have a greater risk of HIV than other urban residents.32 Condom use at last sex with a non-marital non-cohabiting partner was 49% and 61% among urban adolescent girls and boys, respectively, which was about 1.5 times higher than among rural adolescent girls.

Wealth-related inequalities

While there are relatively few multicountry studies examining socioeconomic inequalities in ASRH in sub-Saharan Africa, there is clear evidence of a wealth gradient for a number of indicators such as early marriage, age at first birth and contraceptive usage indicating that poorer adolescents face significant barriers in realising their reproductive health and rights.33 The data from the DHS surveys from 33 countries provide ample evidence of large disparities in ASRH by wealth quintile (figure 2). Compared with adolescent girls from the richest wealth quintile, adolescent girls from the poorest wealth quintile had their first sex 2.2 years earlier, married 4.6 years earlier and gave birth 4.1 years earlier (table 2). Boys in both wealth quintiles started sex at the same age, but those in the poorest households married 4.3 years earlier than those in the richest households. HIV prevalence was higher among both sexes in the richest quintiles among 15–24 years in both regions and most countries. Overall, the wealth-related gaps are even larger than for education and place of residence.

For the analyses of trends over time in the 10 countries, we computed wealth tertiles from an asset index score34 instead of the more conventional quintiles to reduce sampling errors. The results show that gaps for first sex, marriage and childbearing between the poorest and richest tertiles increased except for marriage among males (table 2). The median age at first birth has increased by 0.8 and 1.3 years among those from urban and wealthiest families, respectively, during the period 2004–2015.