“…Our study was somewhat smaller than but comparable in design to a previous report (3), but a bigger sample would have allowed analysis for specific hyperplasia histologies, follow-up biopsies, and influence of tobacco and alcohol use. We also did not evaluate endometrial pattern or echogenicity, but the impact of these is controversial (4,5). Although most of our sample was hyperandrogenic and anovulatory, we did not specifically analyze different PCOS phenotypes (6), which may have affected the results (7).…”