We estimated the overall prevalence of sexual intercourse within the last 12 months at 16.4% and 9.7% among male and female, respectively, in-school adolescents in Zambia. Our finding that 9.4% of adolescents of age less than 15 years had sexual intercourse in the previous 12 months to the survey indicate that adolescents in Zambia have their first sexual intercourse at a young age. This is in support of the findings in the Demograhic and Health Survey of 2001–2002 that the median age at first sexual intercourse was 16.8 and 17.8 years for females and males, respectively [1]. In Zimbabwe, the median age at first sexual intercourse has been reported as 19 years for males and 18 years for females [13]. These estimates are much lower than those that have been reported for the United States between 1991 and 2005 where adolescents in school Grades 9 to 12 had prevalence of sexual intercourse within the last 3 months exceeding 33% [14]. While we can make comparisons of prevalence between different settings, it is also important to recognize that differences may actually arise from differences in study questions. In the Zambia GSHS, students were asked: During the past 12 months, have you had sexual intercourse? Sexual intercourse was not specifically defined. In a study by Thurman et al in South Africa, study participants were asked whether they had ever engaged in sex, specifically described as full penile-vaginal penetration [15]. In the United States [13], the duration of recall was 3 months compared to 12 months in the GSHS. There may therefore be limitation in comparing studies that have a specific definition or definitions of sexual intercourse to those that did not.

The constellation of potentially harmful behaviours within an individual is cause for concern. Our study shows that adolescents who engage in sexual intercourse are also likely to have smoked dagga, used alcohol, and had been drunk. Ohene et al have reported that among Caribbean youth, being sexually active was a good marker of also being involved in other problem behaviours [16].

In Zambia, the school health program includes the following topics for Grades 7–9: love sex and abstinence; reproductive health; sexually transmitted infections; understanding HIV and AIDS, for Grade 7, working together safely; children's rights, growing up; and understanding gender, for Grade 8; and sexual feeling and behaviour; about pregnancy; sexually transmitted infections; understanding HIV and AIDS; and coping well with life, for Grade 9. This curriculum may require to be changed with these other harmful practices in mind. Dealing with sexual behaviour alone may not lead to healthy adolescent behaviours. On the other hand, it is possible that preventing a single harmful behaviour may also result in gains in the prevention of other associated behaviors.

It is of particular interest that adolescents who reported having close friends were more likely to have engaged in sex than those without close friends. It is possible that those adolescents who have friends may use this pool of friends as source of sexual partners. It is also plausible that having had sex may make one popular and be attracted to new friends. In a study of US adolescents, Brady and Halpern-Felsher [17] reported that some adolescents experienced popularity in the immediate period following sexual intercourse.

In our study, males were more likely to have reported sexual intercourse than females (16.4% versus 9.7%). Van Duynhoven et al have reported that females were more likely to report more reliable sexual histories than males [18]. It is not possible to determine whether such disparity in reporting may have occurred in our study. However, the fact that we identified male gender, substance use (illicit drugs and alcohol use) and increasing age as positively associated with history of sexual intercourse calls for consideration of these factors in efforts to prevent sexual intercourse in adolescents or at least identify those that may require targeting for safer sex, such as condom use promotion, education on illicit drugs and prohibition of alcohol use.

We found that adolescents who reported having parental supervision were less likely to engage in sexual intercourse in bivariate analysis but quite the opposite association in multivariate analysis. Bingham and Crockett [19] have demonstrated parental support and supervision to be protective against experience with sexual intercourse. In other studies, adolescents who have parental support are more likely to use condoms and other modern family planning methods if they engage in sexual intercourse [20,21]. We cannot determine with certainty why adolescents who report sexual intercourse were also likely to report parental supervision. It is quite possible that adolescents who engage in problem behaviours, as a consequence may attract increased parental supervision.