Trends in Age at First Sex: An Application of Survival Analysis Techniques to Survey Data from Africa

Basia Zaba, London School of Hygiene and Tropical Medicine
Elizabeth Pisani, Family Health International

DHS data from seven African countries conducting three or more surveys since 1985 were analyzed using survival analysis techniques, combining information on virginity status and retrospective reporting of age at first sex. Median age at first sex was classified by age group, survey year, birth cohort, residence and education, extrapolating with a gamma model where necessary. Age group comparisons generally suggested a slow secular rise in age at first sex. However, tracing birth cohorts between surveys revealed inconsistencies - median ages reported by members of a birth cohort in their teens were higher than those reported when they reached their twenties, even allowing for residence and education changes in sample composition - probably a result of young, sexually active women denying they had ever had sex. A multivariate approach allows us to separate the effects of reporting error and compositional change and estimate true changes in sexual debut over time.