Distribution of Body Hair in Young Australian Women and Associations With Serum Androgen Concentrations

Context: An important element of the diagnosis of polycystic ovary syndrome is hyperandrogenism. Objective: To determine the distribution of modified Ferriman-Gallwey (mF-G) scores, as a measure of facial and body hair growth, and associations between the mF-G scores and serum androgen concentrations, including 11-oxygenated androgens. Design: Cross-sectional study of non-health-care-seeking women, aged 18 to 39 years, recruited from the eastern states of Australia from November 2016 to July 2017. Participants and measurements: Participants provided an mF-G self-assessment that corresponded to their appearance when not using treatment for excess hair. Androgens were measured in 710 women by liquid chromatography and tandem mass spectrometry. Results: The distribution of the mF-G scores was right-skewed. The median (range) mF-G score of all participants (73.1% Caucasian) was 5 (0-36). The mF-G scores were negatively associated with age (r(s) = 0.124; P < 0.0001) and positively associated with body mass index (BMI) (r(s) = 0.073; P < 0.0001). Only androstenedione remained significantly associated with mF-G scores when controlling for age and BMI. Cluster analysis identified 2 groups with mF-G score of < 10 and >= 10. Repeating the cluster analysis using the combined vector of mF-G score and androstenedione returned a similar cluster structure, and again separated the 2 groups at a mF-G score < 10 versus >= 10. Conclusions: A self-assessed mF-G score >= 10 is indicative of excess body hair. Androstenedione, as well as testosterone, should be measured when hyperandrogenism is being evaluated. The lack of association between mF-G scores and the 11-oxygenated androgens highlights the need for a better understanding of these steroids.