Increased suicide risk among younger women in winter during full moon in northern Europe. An artifact or a novel finding?

Abstract

Available evidence suggests that there is no effect of moon phases on suicidal behavior. However, a Finnish study recently reported elevated suicide rates during full-moon, but only among premenopausal women and only in winter. This could not be replicated in an Austrian study and stirred a discussion about whether the Finnish finding was false-positive or if there are unaccounted moderator variables differing between Finland and Austria. The goal of the present study was to provide another replication with data from Sweden, which is geographically more comparable to Finland than Austria. We also investigated the discussed moderator variables latitude and nightly artificial brightness. There were 48,537 suicides available for analysis. The fraction of suicides during the full-moon quarter in winter did not differ significantly from the expected 25% among premenopausal women (23.3%) and in the full sample (24.7%). The incidence risk ratios for full moon quarter in Poisson regression models were 0.96 (95% CI: 0.90–1.02) for premenopausal women and 1.01 (95% CI: 0.99–1.04) for the full sample. According to Bayes-factor analysis, the evidence supports the null-hypothesis (no association) over the alternative hypothesis (some association). We found similar results when we split the data by latitude and artificial nightly brightness, respectively. In line with the Austrian study, there was no increase of suicides in Sweden among premenopausal women in winter during full-moon. The results from the Finnish study are likely false positive, perhaps resulting from problematic but common research and publication practices, which we discuss.

STROBE and pre-registration of observational studies

“…The Finnish study that reported elevated suicide rates during a full moon only among premenopausal women and only in winter, which was debunked by Plöderl et al,2 questions the integrity and the professionalism of our system. Hypotheses were not pre-specified and results only identified in subgroups, not to mention the lack of rationale.…”