Over the last two decades, most studies on the correlates of adolescent sexual and reproductive health in developing countries have focused on Sub-Saharan Africa and on respondents’ individual-level characteristics, according to a recent literature review.¹ Moreover, of the 222 studies included in the analysis, almost half focused on condom use (60) or sexual initiation (45). In contrast, relatively little research has examined youth in Asia, Latin America or the Middle East, looked at family or community-level characteristics, or focused on such outcomes as pregnancy or sexual coercion.

The study updated a 2005 review by the World Health Organization, which assessed studies published in 1990–2002, by merging those findings with the results of a new analysis of studies published in 2003–2010. Both rounds examined peer-reviewed studies on the positive and negative correlates of adolescent sexual and reproductive health, including those related to sexual initiation, number of sex partners, use of condoms and other contraceptives, pregnancy, early childbearing, and HIV and other STIs. Studies were eligible for inclusion if they had been conducted in a low- or middle-income country, had a sample size of at least 100 youth aged 10–24 and had used multivariate analysis. The final analytic sample consisted of 222 articles—158 from the initial review and 64 from the second. Twenty-two additional studies met the eligibility criteria but were excluded because they examined outcomes in a manner that precluded comparison with other studies (e.g., they looked at age at first sex rather than whether the respondent had ever had sex).

The majority of studies focused on Sub-Saharan Africa (133), while 47 centered on South and Southeast Asia, 37 on Latin America and the Caribbean, and five on the Middle East. The most common outcomes examined were condom use (60 studies) and sexual initiation (45 studies); pregnancy and childbearing, HIV, contraception, number of sex partners, and STIs other than HIV were each covered in 18–26 articles. Nine studies in the second round of review examined correlates of sexual coercion, a topic that had not been sufficiently studied to warrant inclusion in the original review. (Abortion was not adequately studied in either round.)

Together, the studies examined a total of 1,441 individual-level characteristics, such as demographic traits, knowledge and behaviors, in relation to adolescent sexual and reproductive health. Less commonly, studies included characteristics related to family (444), peers and partners (208), school (53) and community (41); the latter two categories were proportionally more common in the second round of the review than in the first, indicating that contextual characteristics may be increasingly viewed as important correlates.

The authors identified a number of characteristics that were significantly associated with adolescent sexual and reproductive outcomes in at least two-thirds of studies that examined them. Typically, being married, being younger, having a job, consuming alcohol, having peers or friends who had had sex, having experienced forced first sex, living in a rural area and being an orphan were positively associated with risky behaviors or negative outcomes, while having high educational attainment and having discussed reproductive health with a partner were associated with positive behaviors. Some key characteristics were linked with multiple outcomes; for instance, being relatively well-educated was associated with delayed sexual initiation, use of condoms and use of contraceptives in general.

Although adolescents’ family-level characteristics were generally understudied, one such characteristic—living with both parents—was consistently associated with reduced risks of pregnancy and childbearing. Having engaged in anal sex or commercial sex work and having had an STI in the past were associated with STI symptoms, STI diagnoses or both. The nine studies that examined sexual coercion established having been beaten by a partner, having had a reproductive tract infection and having used alcohol before sex as correlates.

The authors suggest that their findings not only help identify key correlates of risk among young people, but also indicate both met and unmet need for research. For instance, the high level of interest in Sub-Saharan Africa is likely a response to the severity of the HIV epidemic among adolescents in that region. However, increased analysis of the contexts of adolescents’ lives in other regions may be critical, especially given that "important demographic and social shifts are occurring throughout the world, including the increasing age [at] marriage, improved school enrollment, changing family structures, and transformations in technology." In addition, they emphasize the importance of filling gaps in knowledge concerning abortion and sexual coercion among young people. Finally, noting the preponderance of individual-level analysis performed in existing research, the authors point out that neighborhood context has been established as an important correlate of adolescent sexual behavior in the United States, and recommend that community-level research be extended to developing-country contexts.—H. Ball

REFERENCE

1. Mmari K and Sabherwal S, A review of risk and protective factors for adolescent sexual and reproductive health in developing countries, Journal of Adolescent Health, 2013, forthcoming.