Analyzing the Association Between Lunar Phases and Mortality in Sepsis

Sepsis remains a common cause of morbidity and mortality worldwide. Despite decades of research, a complete understanding of the factors predisposing individuals to poor outcomes remains uncertain. The influence of lunar phase on human health has become a popular topic of public discourse but remains scientifically unclear. We sought to test the hypothesis whether lunar phase influences outcomes in patients with sepsis using 2 contemporary datasets.

First, we conducted a retrospective cohort study of adult patients diagnosed with sepsis (using “Sepsis 3” definitions) from a previously curated cohort of sepsis patients using patient encounters at 2 hospitals from January 2021 thru Apil 2023. To evaluate whether moon phase at hospital admission influenced in-hospital mortality, we used a multivariable logistic regression with the primary outcome of in-hospital mortality and primary exposure of moon phase, which was calculated using moon illumination percentage. The regression was adjusted for covariates, including age group, sex, race, ethnicity, socioeconomic indices, Charlson Comorbidity Index, and initial sequential organ failure assessment CSOFA) score. Then, we conducted a secondary analysis using Epic Cosmos, a dataset including over 300 million patient records from over 1,848 hospitals and 42,100 clinics, to evaluate unadjusted 90-day mortality across lunar phases. For all analyses, a p value < 0.05 was considered significant.

We did not find any statistically significant association between moon phase and in-hospital mortality. Compared to the reference (first quarter), the ORs for mortality were 1.39 (95% CI: 0.31–6.23, p = 0.721) for full moon, 1.62 (95% CI: 0.48–5.50, p = 0.518) for waning gibbous, 1.63 (95% CI: 0.48–5.55, p = 0.518) for waxing gibbous, 2.08 (95% CI: 0.62–7.00, p = 0.322) for waning crescent, 1.93 (95% CI: 0.57–6.51, p = 0.375) for waxing crescent, 2.14 (95% CI: 0.48–9.54, p = 0.404) for last quarter, and 2.15 (95% CI: 0.62–7.46, p = 0.31) for new moon. From Epic Cosmos, we also observed that mortality rates remained stable across all lunar phases. In our institutional dataset, we also found that age and initial SOFA were both significantly associated with mortality.

Our study found no association between moon phase and in-hospital mortality among adult patients diagnosed with sepsis at our institution.