OBJECTIVE One of the mechanisms currently proposed to explain the cardioprotective effect of hormone replacement therapy (HRT) is the antioxidative property of estrogens. Considering that HRT involves the administration of an estrogen, usually combined with a progestin and sometimes with an androgen, we investigated the following in vitro: (1) the effect of estradiol, progesterone, and testosterone on the oxidation of low density lipoprotein; (2) the possible pro-oxidative effect of progesterone and testosterone on native low density lipoprotein; and (3) the possible modification of the antioxidant effect of estradiol on low density lipoprotein induced by progestins. DESIGN Low density lipoprotein was isolated from blood samples obtained from 20 untreated postmenopausal women and divided in multiple aliquots, each containing 0.5 mg LDL protein. In Protocol 1 (n = 10) different doses of estradiol, progesterone, and testosterone ranging from 0 to 26 micrograms/ml were tested inducing oxidation with 15 microM copper sulfate. In Protocol 2 (n = 6) we studied the rate of oxidation of low density lipoprotein incubated with progesterone or testosterone without any oxidative induction. In Protocol 3 (n = 10) we studied the concomitant effect of 15 microM estradiol with four separate progestins (progesterone, medroxyprogesterone acetate, norethindrone, and norgestrel) in different doses (0, 5, 15, and 50 microM). After incubation for 4 h at 37 degrees C, malonaldehyde was measured as a marker of low density lipoprotein oxidation. The results were expressed in mean +/- SD. RESULTS Protocol 1: Estradiol induced a dose-dependent decrease in malonaldehyde generation, from a baseline of 61.8 +/- 30.2 nmol/mg protein to 11.6 +/- 7.1 nmol/mg protein at the highest dose of estradiol tested (p < 0.0001). Progesterone or testosterone did not modify malonaldehyde generation. Protocol 2: Progesterone and testosterone did not show pro-oxidative action. Protocol 3: Estradiol 15 microM alone induced a 35% decrease in malonaldehyde generation, from a baseline of 75.4 +/- 25.4 to 49.3 +/- 18.8 nmol/mg protein (p < 0.0001). Norgestrel and norethindrone did not modify the antioxidant effect of estradiol (p > 0.05). Progesterone and medroxyprogesterone acetate induced a further reduction of malonaldehyde concentration to 37.2 +/- 20.8 and 38.6 +/- 18.2 nmol/mg protein, only at the highest dose tested (p < 0.02 and p < 0.01, respectively). CONCLUSIONS Our results demonstrate that, in contrast with the potent antioxidant effect of estradiol, progesterone and testosterone did not show any pro- or antioxidant effect on low density lipoprotein in vitro. Furthermore, progestins did not counteract the antioxidant effect of estradiol in vitro.