In Zimbabwe's rural Manicaland Province, having premarital sex before age 18 is positively associated with men's and women's lifetime number of partners, number of recent partners and number of current sexual relationships, but also with consistent condom use.1 Early first sex is also associated with an increased risk of HIV infection for women; this relationship is significant when the number of years of sexual activity is taken into account, but not when the number of lifetime partners is controlled for. Males' age at first sex is not associated with their risk of acquiring HIV infection. These are among the details about timing of sexual debut, its determinants and its consequences provided by results of a survey conducted in 1998–2000.

The survey was carried out among a population-based sample of men aged 17–54 and women aged 15–44. Researchers conducted face-to-face interviews with participants, who wrote their responses to sensitive questions on cards that they dropped into a locked box. Free HIV counseling, as well as treatment for other sexually transmitted infections, was offered in each of the 12 communities involved. Analyses were based on data from 4,138 men and 4,948 women who provided consistent information about their age and their age at first sex.

Eighty-three percent of men and 81% of women in the sample were sexually experienced. The median age at first sex was 19 years for men and 18 for women. Results of chi-square tests for trends suggest that the median for men declined over the 30 years preceding the survey; the median for women was stable for 20 years but appears to have increased during the last 10 years before the survey.

In one set of analyses, the researchers calculated hazard ratios to estimate associations between sexual experience among those younger than 25 and a wide range of socio-economic characteristics and AIDS knowledge. At the multivariate level, few significant associations emerged. Among young men, students and those who were unemployed were less likely than manual laborers to have had sex (hazard ratios, 0.6 and 0.9, respectively), members of traditional churches were more likely than Anglicans to be sexually experienced (1.4), and the likelihood of sexual experience was greater among minority tribes than among the Man-yika, the predominant group (1.3). Among young women, students, those in skilled occupations and those who listened to the radio daily had a reduced likelihood of being sexually experienced (0.2–0.8), as did 15–19-year-olds who had completed their primary education (0.7). Members of minority tribes and women who had previously lived in an urban area or in a different rural area had an increased likelihood of being sexually experienced (1.3–1.5).

The next set of analyses used linear and logistic regression to determine the association between sexual debut before age 18 and selected outcomes related to sexual behavior: numbers of lifetime partners (with and without adjustment for number of years of sexual activity), partners in the past month, sexual encounters in the past two weeks and current relationships; consistent condom use in the past two weeks; and divorce. For both males and females who had first had sex before age 18 outside marriage, the coefficients were positive and significant for all outcomes except divorce (which was nonsignificant). These results indicate that individuals who had early premarital sex subsequently engaged in more sexual activity than others, but also that they were more likely to protect themselves by using condoms. Fewer results were significant for men and women whose first intercourse had occurred before age 18 within marriage, and the magnitude of the associations was smaller.

Finally, logistic regression was used to measure associations between age at first sex and the likelihood of acquiring HIV infection. In multivariate models controlling for the number of years of sexual activity, women who had initiated intercourse at ages 12–16 were significantly more likely than those whose sexual debut had been at age 21 or later to become HIV-infected (odds ratio, 1.6). Age at first intercourse was not a significant predictor of HIV risk for males in this model or for either gender in a model that controlled for the lifetime number of partners.

Because the survey was cross-sectional, the researchers note, the data do not imply "that early sexual debut leads to a spiral of decline and riskier sexual behavior." The investigators conclude that delaying sexual debut would have only a limited effect on lifetime HIV risk at the population level because "the casual nature of…sexual activity rather than…early sexual debut appears to be what leads to…infection" among individuals in Manicaland who begin their sexual lives at a young age.—D. Hollander


1. Hallett TB et al., Age at first sex and HIV infection in rural Zimbabwe, Studies in Family Planning, 2007, 38(1):1–10.