A large cohort study of hypothyroidism and hyperthyroidism in relation to gynecologic cancers

Background. Thyroid status may influence tumorigenesis of gynecologic cancers, yet epidemiologic studies of this relationship are limited and inconsistent. Methods. We evaluated the association of self-reported history of physician-diagnosed hypothyroidism and hyperthyroidism with medical-record confirmed endometrial (EC; all invasive adenocarcinomas) and ovarian cancer (OC; epithelial ovarian or peritoneal cancers) in Nurses' Health Study (NHS) from 1976 to 2010 and NHSII from 1989 to 2011. Cox proportional hazard models were used to estimate multivariable rate ratios (RRs) and 95% confidence intervals based on pooled cohort data. Results. We confirmed 1314 incident cases of EC and 1150 cases of OC. Neither a history of hypothyroidism nor hyperthyroidism was significantly associated with risk of EC or OC. However, having a history of hypothyroidism for 8+ years (median) was nonsignificantly inversely associated with EC (RR = 0.81; 95% CI = 0.63-1.04; P-trend with history duration = 0.11) and OC (RR = 0.87, 95% CI = 0.66-1.15; P-trend = 0.13). Having a history of hyperthyroidism for 6+ years (median) was non-significantly positively associated with EC (RR = 1.69; 95% CI = 0.86-3.30; P-trend = 0.12) but not OC (RR = 1.12; 95% CI = 0.46-2.72; P-trend = 0.95). Conclusions. A history of hypothyroidism or hyperthyroidism was not significantly associated with risk of EC or OC.