Table 4.

Summary results of the secondary outcomes for the 28 women who completed the trial (per-protocol analysis set)

		Placebo (adjusted mean)	MLE4901 (adjusted mean)	Adjusted estimate of percentage point difference between treatment means	p value comparing adjusted treatment means
Hot flush severity		5·70 (5·09 to 6·38)	3·27 (2·92 to 3·66)	−41% (−32 to −49)	<0·0001
Hot flush bother		5·56 (4·96 to 6·22)	2·92 (2·61 to 3·27)	−45% (−36 to −53)	<0·0001
Hot flush interference		26·48 (20·02 to 35·03)	7·94 (5·76 to 10·95)	−58% (−40 to −76)	<0·0001
MENQOL domain score
	Vasomotor	3·98 (3·38 to 4·69)	2·05 (1·74 to 2·42)	−45% (−33 to −58)	<0·0001
	Psychosocial	2·58 (2·30 to 2·90)	2·18 (1·94 to 2·45)	−15% (−5 to −25)	0·0083
	Physical	2·93 (2·63 to 3·27)	2·42 (2·17 to 2·69)	−19% (−9 to −28)	0·0002
	Sexual	2·15 (1·84 to 2·51)	1·98 (1·68 to 2·30)	−8% (5 to −23)	0·24
Number of flushes detected by sweat monitor per 24 h		26·91 (23·16 to 31·27)	16·22 (13·99 to 18·80)	−43% (−30 to −55)	<0·0001

All outcomes were compared during the final week of the 4 week treatment period with MLE4901 and exact-match placebo to allow adequate time for an effect to be observed. For all endpoints, a generalised linear model with gamma error structure was used. A standard crossover analysis was implemented, with period, administration sequence, and treatment as fixed effects and subject as a random effect. All possible baseline covariates were initially included in exploratory analyses but the only covariates that were significant, in each case, were the baseline values of the endpoint being analysed. All other demographic covariates were therefore excluded from the final model. From each model, adjusted (least squares) means and differences between means were estimated, together with associated 95% CIs. p values refer to the comparison of the mean values of the two treatments (placebo and MLE4901). MENQOL=Menopause-Specific Quality of Life questionnaire.