Objective: To identify reporting biases and to determine the influence of inconsistent reporting on observed trends in the timing of age at first sex and age at marriage.
Methods: Longitudinal data from three rounds of a population-based cohort in eastern Zimbabwe were analysed. Reports of age at first sex and age at marriage from 6837 individuals attending multiple rounds were classified according to consistency. Survival analysis was used to identify trends in the timing of first sex and marriage.
Results: In this population, women initiate sex and enter marriage at younger ages than men but spend much less time between first sex and marriage. Among those surveyed between 1998 and 2005, median ages at first sex and first marriage were 18.5 years and 21.4 years for men and 18.2 years and 18.5 years, respectively, for women aged 15–54 years. High levels of reports of both age at first sex and age at marriage among those attending multiple surveys were found to be unreliable. Excluding reports identified as unreliable from these analyses did not alter the observed trends in either age at first sex or age at marriage. Tracing birth cohorts as they aged revealed reporting biases, particularly among the youngest cohorts. Comparisons by birth cohorts, which span a period of >40 years, indicate that median age at first sex has remained constant over time for women but has declined gradually for men.
Conclusions: Although many reports of age at first sex and age at marriage were found to be unreliable, inclusion of such reports did not result in artificial generation or suppression of trends.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Contributors: SG and GPG designed the research, PM, ZM and CN performed the research, IC and TH analysed the data. All authors contributed to preparing the manuscript.
Funding: The Wellcome Trust and the Joint United Nations Programme on HIV/AIDS supported the Manicaland HIV/STD Prevention Project. IC received PhD funding from the Division of Epidemiology, Public Health & Primary Care, Imperial College London.
Competing interests: None.
Ethics approval: Written informed consent was given by every participant at each survey. Ethical approval for the study was granted by the Medical Research Council of Zimbabwe (number 02187) and St Mary’s, London local research ethics committee (03.66 R&D 03/SB/004E).