Incidence rates of endometrial hyperplasia, endometrial cancer and hysterectomy from 1980 to 2003 within a large prepaid health plan

Obesity strongly increases the risk of endometrial cancer and is projected to increase current and future endometrial cancer incidence. In order to fully understand endometrial cancer incidence, one should also examine both hysterectomy, which eliminates future risk of endometrial cancer, and endometrial hyperplasia (EH), a precursor that prompts treatment (including hysterectomy). Hysterectomy and EH are more common than endometrial cancer, but data on simultaneous temporal trends of EH, hysterectomy and endometrial cancer are lacking. We used linked pathology, tumor registry, surgery and administrative datasets at the Kaiser Permanente Northwest Health Plan to calculate age-adjusted and age-specific rates, 1980-2003, of EH only (N = 5,990), EH plus hysterectomy (N = 904), hysterectomy without a diagnosis of EH or cancer (N = 14,926) and endometrial cancer (N = 1,208). Joinpoint regression identified inflection points and quantified annual percentage changes (APCs). The EH APCs were -5.3% (95% confidence interval [CI] = -7.4% to -3.2%) for 1980-1990, -12.9% (95% CI = -15.6% to -10.1%) for 1990-1999 and 2.4% (95% CI = -6.6% to 12.2%) for 1999-2003. The EH-plus-hysterectomy APCs were -8.6% (95% CI = -10.6% to -6.5%) for 1980-2000 and 24.5% (95% CI = -16.5% to 85.7%) for 2000-2003. Hysterectomy rates did not significantly change over time. The endometrial cancer APCs were -6.5% (95% CI = -10.3% to -2.6%) for 1980-1988 and 1.4% (95% CI = -0.2% to 3.0%) for 1988-2003. Hysterectomy rates were unchanged, but increased endometrial cancer incidence after 1988 and the reversal, in 1999, of the longstanding decline in EH incidence could reflect the influence of obesity on endometrial neoplasia.