Adolescent sexual behavior: Estimates and trends from four nationally representative surveys

Context: Accurate information about trends over time in adolescent sexual b ehavior is essential to understand changes in adolescent pregnancy and sexu ally transmitted diseases and to monitor the progress of health promotion a ctivities in the United States Methods: Estimates from the National Survey of Family Growth (NSFG), the Na tional Survey of Adolescent Males (NSAM), the Youth Risk Behavior Survey (Y RBS) and the National Longitudinal Study of Adolescent Health (Add Health) were compared. While methodologies and populations varied by survey, adoles cents aged 15-17 who attend high school were a common subpopulation among a ll four. For each survey the prevalence of sexual intercourse, contraceptiv e use and multiple sexual partners was measured in this population. Results: Trend comparisons fell into four categories. First some similar si gnificant trends were found across surveys. The proportion of all males and of white males who reported ever having had sexual intercourse decreased s ignificantly, while condom use rose significantly among males in both the N SAM and the YRBS. For such behaviors as ever having had sexual intercourse (among Hispanic males and black females), using the pill and using the cond om (among all females) and having four or more lifetime sexual partners (am ong white males), a significant trend was found in one survey while a simil ar but nonsignificant trend was found in another. Several trend comparisons were not significant in any survey. Finally having had intercourse in the past three months (among all males and all females), having had two or more partners in the past three months (for males) and having had four or more lifetime sexual partners (among white females and all males) showed a signi ficant trend in one survey but lacked a parallel nonsignificant trend in an other. Prevalence estimates in 1995 differed significantly in at least one comparison of surveys for all behaviors except having four or more lifetime sexual partners (both genders) and having two or more recent sexual partne rs (females). Gender differences within the YRBS and between the NSFG and t he NSAM generally were consistent. Conclusions: Trends over time and gender differences were similar across su rveys, underscoring their value for tracking adolescent sexual behaviors. D ifferences in prevalence estimates across surveys probably result from diff erences in question wording, diverse interview settings and modes of data c ollection, and varying statistical power. These findings suggest a need to increase our understanding of how methodologies influence survey response i n research on adolescents.