Twenty-eight studies (n=5,755) in total were included: 10 placebo-controlled RCTs, 7 RCTs, 4 cross-sectional studies, 2 prospective studies, 2 partly randomised cohort studies, an open-label study and 2 longitudinal studies (one randomised and one prospective).
Ten RCTs (of which eight were placebo-controlled) were double-blinded. Drop-out rates within individual treatment and control groups ranged from 0 to 50%, where reported. Most of the study groups had drop-out rates below 20%, but 3 studies (one placebo-controlled RCT of muscle strength, one placebo-controlled RCT, and one partly randomised cohort of body composition) had one or more treatment groups with a drop-out rate of 40% or more.
Effects on muscle strength (6 RCTs and 5 observational studies).
Two double-blind, placebo-controlled RCTs, both of which included an exercise component, and one other RCT reported statistically significant positive effects of HRT on muscle strength (isometric back extensor muscle strength; knee extension and vertical jumping height). However, the 2 RCTs of exercise had high drop-out rates in the exercise groups, which confounded the data. In comparison, one placebo-controlled RCT and one other RCT found no effects of HRT. One cross-sectional study also reported positive effects for HRT, but another cross-sectional study and 2 prospective studies reported no effect of HRT on muscle strength. Further details of the studies were reported in the review.
One double-blind, placebo-controlled RCT reported a significant improvement in handgrip strength for tibolone in comparison with placebo in postmenopausal women, in addition to a positive effect on body mass index-adjusted isometric knee extension strength. One cross-sectional study also reported non significant positive effects on mean knee extensor strength for tibolone and oestrogen, in comparison with no HRT.
Effects on body composition (11 RCTs, one open-label controlled trial and 5 observational studies).
Five studies (one placebo-controlled RCT, one RCT, one prospective longitudinal study and 2 partly randomised cohort studies) reported that conventional HRT had a significant positive effect on body composition. However, 5 RCTs (3 double-blind, placebo-controlled RCTs, one placebo-controlled RCT and one RCT) failed to find any effect of HRT on body composition. One double-blind, placebo-controlled RCT, one RCT, one open-label controlled study and one randomised longitudinal study reported significant positive effects for tibolone on body composition. One cross-sectional study reported a negative effect for HRT. Further details were reported in the review.