In many developing countries, considerable proportions of women are married during adolescence, yet few studies have compared characteristics and outcomes of girls married in early, middle and late adolescence. In a study conducted by Annabel Erulkar using data from a population-based survey in seven of Ethiopia’s nine major regions, 17% of 20–24-year-old women had married before age 15, 30% at ages 15–17 and 26% at ages 18–19; the remainder either had married after age 19 or had not yet married [see article]. Compared with young women who had married at age 18 or 19, those married before age 15 were less likely to have known about their marriage beforehand and more likely to have experienced forced first marital sex. Young women who had never attended school had higher odds of having been married before age 15 than of not having been married by age 20. Educational attainment was positively associated with foreknowledge and wantedness of marriage, and with high levels of marital discussions about fertility and reproductive health issues. The author recommends that programs focus on delaying the earliest marriages, and on girls who have never attended or have left school.

Social desirability bias is an intrinsic problem in studies that rely on self-reported sexual behavior. The extent of bias, and the ways in which it manifests itself, often differ between males and females. To test how data collection methods affect gender differences in reporting, Christine Kelly and colleagues compared reports of sexual behavior from two groups of never-married males and females aged 16–18 [see article]. Data were collected through face-to-face interviews in one group and through audio computer-assisted self-interviewing (audio-CASI) in the other. In adjusted regression models, males interviewed via audio-CASI were less likely to report ever having had a girlfriend than those interviewed face-to-face, but were more likely to report having had sex with a relative or teacher. For females, reports of ever having had a boyfriend did not differ between interview modes. A small proportion of females reported ever having had sex with a relative or teacher via audio-CASI, but none did so face-to-face. The authors note that audio-CASI increases privacy and may therefore reduce the influence of societal expectations on reporting, but it also removes the opportunity for interviewers to probe respondents or clarify questions.

Abundant data exist on the sexual and reproductive behaviors of middle and older adolescents and young adults in Sub-Saharan Africa, but relatively little is known about those of young adolescents. Using data drawn from the first two waves of the Transition to Adulthood study, Milly Marston and colleagues identified variables that predicted sexual debut between waves among 12–16-year-olds living in two informal settlements in Nairobi, Kenya [see article]. Overall, 92 (53 males and 39 females) of the 1,754 respondents who had never had intercourse at Wave 1 reported having had sex at Wave 2. For both males and females, sexual debut was associated with having permanently dropped out of school, having never attended school and having severe family dysfunction. Lack of parental supervision predicted sexual debut only among males, and low aspirations did so only among females. Males and females who had medium or low self-esteem were less likely than those with high self-esteem to experience sexual debut between survey waves. According to the authors, the findings suggest that policies and programs encouraging aspirations for education and employment are likely to reduce early sexual debut among young women, and that those encouraging participation in groups that promote positive social behaviors are likely to do so for young men.

Although most maternal deaths are preventable, maternal mortality remains high in many developing countries. Tracking progress toward the Millennium Development Goal of a 75% reduction in maternal mortality between 1990 and 2015, Sarah Zureick-Brown and colleagues estimated that the annual number of maternal deaths worldwide declined by 34% between 1990 and 2008 (from 546,000 to 358,000), as did the maternal mortality ratio for the world as a whole (from 400 to 260 maternal deaths per 100,000 live births) [see article]. The majority of the global burden of maternal deaths shifted from Asia to Sub-Saharan Africa during the study period, driven by differential trends in fertility, the HIV/AIDS epidemic and access to reproductive health care.

Also in This Issue

In Ghana, an estimated two-thirds of abortions are obtained illegally under unsafe conditions and an estimated 15–30% of maternal deaths result from abortion complications. In a Special Report, Vanessa Dalton and colleagues describe the establishment of the International Family Planning Fellowship Program in Ghana in 2008 to address unsafe abortion by providing physicians with advanced training in abortion and family planning [see article]. The authors note that this postgraduate training program could serve as a model and resource for other countries in their efforts to improve women’s health. —The Editors