Fourteen studies were included (n was unclear but more than 1,419,833 participants). These included one RCT (n=16,608), five cohort studies (n=1,384,941) and eight population-based case control studies (n at least 13,345).
The authors stated that all studies adjusted for appropriate potential confounders in their analysis. Results were similar for fixed-effect and random-effects models. Results from only fixed-effect models were presented.
Oestrogen-only and oestrogen plus progestin hormone therapy: The risk of ovarian cancer was significantly increased among oestrogen-only users (relative risk per five years of hormone use 1.22, 95% confidence interval (CI): 1.18 to 1.27; 13 studies) and oestrogen plus progestin hormone therapy users (relative risk per five years of hormone use 1.10, 95% CI: 1.04 to 1.16; 11 studies). There was no evidence of heterogeneity or publication bias for either analysis.
Oestrogen-only versus oestrogen plus progestin hormone therapy: The risk of ovarian cancer was significantly higher among oestrogen-only compared to oestrogen plus progestin users (p=0.004; 10 studies). Results were similar for the five studies in which duration of use was similar for both types of hormone oestrogen plus progestin therapy (p=0.011).