Background and Goal:
Sleep problems were significantly increased on simvastatin (
simva
) (but not pravastatin) vs placebo in the UCSD Statin Study. Sleep problems on simva predicted glucose rise. Weight gain has also been reported as a statin side effect. We sought to capitalize on existing data to assess whether sleep problems on simva related to weight gain in men.
Method:
442 men without known diabetes or CVD were randomized to simva 20mg or placebo for 6 mon. One hundred eighty and 186 completed single-item self-rating of change in sleep problems vs baseline (
Δslpprob
). Weight (lb) was measured at baseline and 6 mon. Missing 6 mon values were imputed.
Analyses:
A. Regressions stratified by treatment assessed prediction of weight change by Δslpprob, adjusted for baseline weight. B. Regressions assessed prediction of weight change by the interaction term of simva (vs placebo) x Δslpprob, adjusted for the components of the interaction and baseline weight. Since age-related muscle loss may complicate weight change in elderly; and young adults have low vulnerability to metabolic problems, analyses were repeated excluding these groups.
Results:
A. Increased sleep problems on simva predicted weight
gain
(significant), but on placebo predicted weight
loss
(nonsignificant). B. The Δslpprob x simva interaction term significantly predicted weight gain. When that was parceled out, simva, outside of the sleep relationship, negatively predicted weight change. Exclusion of young adults and elderly strengthened significance of findings (Table).
Discussion:
Sleep problems, which differentially arise on simva, differentially predict weight gain on simva. This expands the metabolic effects to which sleep problems on simva may contribute and might possibly favor mediation by sleep apnea (a reported complication of simva). Once the sleep problem effect is considered, simva use predicted weight
loss
. The relative contribution of fat vs muscle loss (vs other) requires exploration.