Strategies for combining and scaling sexual maturity indicators: the Heartfelt Study

Primary objective: To identify the best variable or combination of sexual m aturity variables to use in investigations of adolescent health, and explor e possible re-scaling of five-stage sexual maturity stages into 'dummy vari ables' ('advanced'/'delayed'). Research design: Cross-sectional. Subjects: Tri-ethnic sample of 384 US adolescents 11-16 years of age (Heart felt Study). Methods: Sexual maturity variables are genital, breast and pubic hair stage s obtained by clinical examination by nurse practitioners, testes size, men arche and age at menarche. Results: Principal factor analysis of sexual maturity variables, height and age, revealed high commonality among the variables, as no more than a sing le factor appeared in 27 of 28 factor analyses (by sex and age groups). Gen ital singe (boys) and breast stage (girls) were most highly related to the first principal factor independent of age and ethnicity. ANOVA of 'key vari ables' indirectly reflecting endocrine function suggested that some stages could be combined. Conclusions: A single variable should suffice to describe sexual maturity a t the studied ages, or variables could be combined in a (unstandardized) se xual maturity index (SMI): SMI (Girls) = 0.998 x BREAST + 0.900 x PUBIC HAIR + 1.815 x YEARS SINCE MEN ARCHE, SMI (Boys)= 1.104 x GENITAL + 1.328 x PUBIC HAIR + 5.997 x TESTES SI ZE (mL), There is modest support for re-scaling a sexual maturity variable into 'del ayed' and 'advanced' groups if needed, thus: (stages 1, 2, 3, vs 4, 5).