Objective: This study determined whether two different formulations of hormone therapy (HT): oral conjugated equine estrogens (o-CEE; 0.45 mg/d, n = 209), transdermal 17β-estradiol (t-E2; 50 μg/d, n = 201) plus cyclic progesterone (Prometrium, 200 mg) or placebo (PBO, n = 243) affected sleep domains in participants of the Kronos Early Estrogen Prevention Study.
Methods: Participants completed the Pittsburgh Sleep Quality Index at baseline and during the intervention at 6, 18, 36, and 48 months. Global sleep quality and individual sleep domain scores were compared between treatments using analysis of covariance, and correlated with vasomotor symptom (VMS) scores using Spearman correlation coefficients.
Results: Global Pittsburgh Sleep Quality Index scores (mean 6.3; 24% with score >8) were similar across groups at baseline and were reduced (improved sleep quality) by both HT (average change -1.27 [o-CEE] and -1.32 [t-E2]) when compared with PBO (-0.60; P = 0.001 [o-CEE vs PBO] and P = 0.002 [t-E2 vs PBO]). Domain scores for sleep satisfaction and latency improved with both HT. The domain score for sleep disturbances improved more with t-E2 than o-CEE or PBO. Global sleep scores significantly correlated with VMS severity (rs = 0.170, P < 0.001 for hot flashes; rs = 0.177, P < 0.001 for night sweats). Change in scores for all domains except sleep latency and sleep efficiency correlated with change in severity of VMS.
Conclusions: Poor sleep quality is common in recently menopausal women. Sleep quality improved with both HT formulations. The relationship of VMS with domains of sleep suggests that assessing severity of symptoms and domains of sleep may help direct therapy to improve sleep for postmenopausal women.