Colon cancer risk and different HRT formulations: a case-control study
Centre for Epidemiology & Health Research Berlin, Invalidenstr. 115, 10115 Berlin, Germany
BMC Cancer 2007, 7:76 doi:10.1186/1471-2407-7-76Published: 8 May 2007
Most studies have found no increased risk of colon cancer associated with hormone replacement therapy (HRT), or even a decreased risk. But information about the effects of different HRT preparations is lacking.
A case-control study was performed within Germany in collaboration with regional cancer registries and tumor centers. Up to 5 controls were matched to each case of colon cancer. Conditional logistic regression analysis was applied to estimate crude and adjusted odds ratios (OR) and 95% confidence intervals (95% CI). Stratified analyses were performed to get an impression of the risk associated with different estrogens and progestins.
A total of 354 cases of colon cancer were compared with 1422 matched controls. The adjusted overall risk estimate for colon cancer (ColC) associated with ever-use of HRT was 0.97 (0.71 – 1.32). No clinically relevant trends for ColC risk were observed with increasing duration of HRT use, or increasing time since first or last HRT use in aggregate.
Whereas the overall risk estimates were stable, the numbers in many of the sub-analyses of HRT preparation groups (estrogens and progestins) were too small for conclusions. Nevertheless, if the ColC risk estimates are taken at face value, most seemed to be reduced compared with never-use of HRT, but did not vary much across HRT formulation subgroups. In particular, no substantial difference in ColC risk was observed between HRT-containing conjugated equine estrogens (CEE) or medroxyprogesterone acetate (MPA) and other formulations more common in Europe.
Ever-use of HRT was not associated with an increased risk of colon cancer. In contrary, most risk estimates pointed non-significantly toward a lower ColC risk in HRT ever user. They did not vary markedly among different HRT formulations (estrogens, progestins). However, the small numbers and the overlapping nature of the subgroups suggest cautious interpretation.