Abstract Context Hyperandrogenemia is associated with increased risks for metabolic syndrome in asymptomatic women and women with polycystic ovary syndrome. The relationship between hyperandrogenemia and metabolic syndrome in adolescent girls remains poorly defined. Objective To assess whether free testosterone concentrations predict metabolic syndrome severity (MSS) in adolescent girls. Design/Setting Using data from the U.S. National Health and Nutrition Examination Survey (2013–2016), we performed weighted regression analysis to evaluate whether free testosterone predicts MSS z-score after adjusting for body fat percentage, chronological age, and gynecological age. Similar analyses were performed using either total testosterone or sex hormone-binding globulin (SHBG) as the primary predictor variable. Participants 287 girls aged 12–19 years. Main Outcome Measure MSS z-score. Results In simple regression analyses, free testosterone and percent body fat predicted MSS z-score (R2 = 0.11 and 0.35, respectively; P 0.0001 for both), but chronological age and gynecological age did not. In the multivariable regression model (R2 = 0.41), higher free testosterone and percent body fat were independently associated with higher MSS z-score (P = 0.0081 and 0.0001, respectively), with no significant independent associations with chronological age or gynecological age. While SHBG was an independent predictor of MSS z-score (P = 0.0104) in a multivariable regression model, total testosterone was not (P = 0.77). Conclusion Higher free testosterone predicts greater MSS in a nationally representative sample of adolescent girls aged 12 to 19 years. This relationship at least partly reflects the relationship between MSS z-score and SHBG.
McCartney et al. (Tue,) studied this question.