Seventy-six RCTs were included in the review: 71 parallel designs and five crossover trials. Follow-up ranged from four weeks to two years.
Lipoprotein(a): Studies showed a neutral or slightly lowering effect of non-oral hormone therapy on lipoprotein(a) compared with baseline, but not compared with control groups. This was not significantly different from the decrease caused by oral hormone therapy.
Homocysteine: Reported changes in homocysteine were non-significant and did not differ from the lack of effect seen with oral hormone therapy.
Inflammation: All except two studies of non-oral hormone therapy reported non-significant changes or no changes in C-reactive protein, one showed a significant increase and one a significant decrease. All studies that assessed oral therapy reported significant increases from baseline, which were significantly different to controls in most cases. Changes in cell adhesion molecules appeared to be less pronounced with non-oral therapy than with oral therapy.
Markers of endothelial dysfunction: Data on this outcome were limited, but it appeared that changes in markers of endothelial dysfunction were not significant. Endothelin-1 did not show differences between oral and non-oral therapy; there were larger decreases in asymmetric dimethylarginine with oral therapy.
Anticoagulant proteins: Changes in antithrombotic proteins were small and mostly non-significant. Where significant changes were found these were usually in relation to baseline rather than to control groups; differences between oral and non-oral groups were not considered to be significant overall.
Resistance to activated protein C: Significant increases in resistance to activated protein C were found, although these were considered to be of smaller magnitude than those seen with oral therapy.
Markers of coagulation: A significant decrease in factor VII coagulation activity was found during non-oral hormone therapy, while none of the other markers for coagulation showed an overall significant change. There were no significant differences between oral and non-oral groups with the possible exception of prothrombin fragment 1+2, where there may be fewer changes in the non-oral groups.
Markers of fibrinolysis: The decreases in total plasminogen activator and total plasminogen activator-inhibitor were absent or were smaller in patients who were given non-oral hormone therapy than those seen in trials of oral hormone therapy.