To determine whether the mechanisms of C-reactive protein production differ depending on the presence or absence of a progestin in the regimen.

We examined data from the Postmenopausal Estrogen Progestin Intervention (PEPI) study, a 5-group (3 different combined estrogen-progestin regimens, conjugated equine estrogen-only, and placebo) randomized clinical trial. This substudy included 221 postmenopausal women assigned to active treatment groups who took at least 80% of pills and had stored plasma specimens available to assess 12-month changes in estrone, sex hormone binding globulin, interleukin (IL)-6, and C-reactive protein levels.

All treatments resulted in increases in estrone, sex hormone binding globulin, and C-reactive protein at 12 months compared with baseline values. In all progestin-containing groups, 12-month change in IL-6 was positively correlated with 12-month change in C-reactive protein (r between 0.34 and 0.65, each P <.010). By contrast, in the conjugated equine estrogen-only group, 12-month change in IL-6 was negatively correlated with 12-month change in C-reactive protein (r value −0.31, P = .055). In adjusted models predicting 12-month C-reactive protein change, an interaction between change in IL-6 and treatment group was highly significant (P=.0008-P <.0001) for each of the progestin-containing groups compared with the conjugated equine estrogen-only group. In the conjugated equine estrogen-only group, the change in C-reactive protein per unit increase in IL-6 was −0.88, whereas in the progestin-containing groups it ranged from 1.46 to 2.85 (P <.0001 for each comparison with conjugated equine estrogen-only).

Progestins in combination with conjugated equine estrogen potentiate the IL-6 (inflammatory)-mediated stimulation of C-reactive protein. These findings support the hypothesis that progestins plus estrogen, not estrogen alone, generate C-reactive protein through an inflammatory mechanism.