Different effects of continuous oestrogen-progestin and transdermal oestrogen with cyclic progestin regimens on...

Seventy-five postmenopausal women were randomly allocated to receive either continuous oral 17 beta-oestradiol 2 mg day-1 and norethisterone acetate 1 mg day-1 (E2/NETA) or transdermal treatment consisting of 28-day cycles with patches delivering 17 beta-oestradiol 50 micrograms day-1 combined with oral cyclic medroxyprogesterone acetate 10 mg day-1, on days 17-28 (E2/MPA). At baseline, the plasma lipid and lipoprotein concentrations, composition and concentrations of low-density lipoprotein (LDL) subclasses (LDL1, LDL2 and LDL3) isolated by density-gradient ultracentrifugation were similar in the two groups. The post-heparin plasma hepatic lipase activity (HL) correlated inversely with the percentage of total LDL found in LDL1 (buoyant LDL) and directly with the percentage of LDL found in LDL3 (dense LDL). After 12 months of hormone replacement therapy (HRT), the total and LDL-cholesterol concentration of the E2/ NETA group decreased by 14% and 17% respectively (P < 0.001), while in the E2/MPA group these parameters remained unchanged. The lowering of LDL-cholesterol in the E2/NETA group was a consequence of a significant reduction of the large, buoyant LDL particles (LDL1) from 103 mg dL-1 to 60 mg dl-1 (P < 0.001) and of a decrease of cholesterol content of LDL particles in the major LDL subclass, LDL2. In the E2/MPA group, the concentration of LDL1 decreased, but less than in the oral group. In both groups, a significant increase in the concentration of the LDL3 subclass was observed, indicating an overall shift to denser LDL particles. After 12 months, the post-heparin plasma HL activity decreased only in the E2/NETA group (by 12%). The inverse correlation between post-heparin plasma HL activity and LDL1 persisted in both groups, but the direct correlation between HL and LDL3 vanished in the E2/NETA group and subsided in the E2/MPA group. Our results indicate that HRT has multiple effects on LDL subclasses and suggest that these changes cannot be explained by changes in HL activity.