Hormone therapy (HT) has been the subject of considerable controversy. Some prospective trials have shown decreased cardiovascular mortality in users of HT compared with nonusers. Others, notably the Women's Health Initiative (WHI) trial, showed an increase in cardiovascular events but no change in cardiovascular mortality. Some of the discrepancies in trial results might be because of differences in the ages of participants. Salpeter and associates conducted a meta-analysis to discover how the age at which women begin HT affects mortality.

The authors performed a systematic review of randomized controlled trials of HT lasting six months or more, with at least one death reported. Outcome measures were total deaths, deaths resulting from specific causes, and all-cause mortality. An arbitrary age cutoff of 60 years was chosen to determine effects of HT on younger and older patients.

A total of 30 trials were analyzed. Of 14,147 patients treated with HT, 518 died. Of 12,561 placebo participants, 501 died. The total mortality odds ratio (OR) for women taking HT was 0.98. The ORs for cardiovascular deaths and cancer deaths were 1.10 and 1.03, respectively. Excluding cancer and cardiovascular deaths, HT was associated with a 33 percent reduction in mortality. The younger age group was represented in 17 trials. The OR for total mortality in this group was 0.61. For cardiovascular and cancer mortality, the ORs were 0.68 and 0.69, respectively. In the older group, represented by 13 trials, the OR was 1.03 for total mortality, 1.11 for cardiovascular mortality, and 1.07 for cancer mortality. Thus, HT was associated with a significantly greater reduction (39 percent) in mortality in younger users than in older users.

The authors hypothesize that the discrepancy between observational study findings and randomized trials may be caused by this age differential. They cite the findings of the Nurses' Health Study, which showed a reduction in cardiac events and cardiovascular and total mortality, and the WHI study, in which an increase in cardiovascular events was evident. Both trials showed similar increases in breast cancer, stroke, and pulmonary embolism among HT users, as well as decreases in colon cancer and hip fracture. The WHI trial showed a nonsignificant decrease in cardiac events in those women who started HT shortly after menopause, whereas in those who started later there was an increase. The authors conclude that given these findings, and in light of their meta-analysis results, there may still be a role for HT use in younger women, beginning shortly after menopause.