Exploring the influence of moon phases and weather on mortality in a palliative care unit over a ten year period

Despite a common belief among public and sometimes medical professionals that moon phases affect mortality rates, studies do not confirm generally higher mortality rates. However, periods of extreme heat or cold have been shown to cause fluctuations in mortality, especially in vulnerable patients. This study aimed to examine whether mortality rates in a 10-bed palliative care unit were influenced by moon phases or weather conditions (temperature, barometric pressure, humidity). Data were gathered from records of the Wuerzburg University Hospital palliative care unit, the Wuerzburg weather station, and the NASA lunar calendar. Daily death rates were calculated over a 10-year period (2010-2019). Bivariate analyses and linear multivariate regression analyses were used to explore potential relationships between moon phases, weather, and mortality. Linear associations, a prerequisite for regression analysis, were ensured using Box-Tidwell tests. Over 3652 days, 3120 patients were treated, with 1508 (48.4%) deaths recorded. The average daily death rate was 0.41 (SD = 0.65), with a maximum of five deaths per day. No deaths occurred on 2411 days (66%). The average death rates were similar across the moon phases: 0.40 (SD = 0.71) for the full moon, 0.42 (SD = 0.66) for the new moon, 0.40 (SD = 0.62) for the waxing moon, and 0.43 (SD = 0.66) for the waning moon, with no significant association (ANOVA: F(3, 3648) = 0.51, p = 0.67). Weather variations were not significantly associated with death rates. The multivariate linear regression analysis confirmed that no combined effects, e.g., of different weather aspects, were found (F(6, 3643) = 0.41, p = 0.87). Our findings align with those of previous studies, which revealed no associations between the moon phases and mortality rates. Unlike general mortality trends, temperature did not significantly impact death rates in the palliative care unit, possibly because of controlled environmental factors. Limitations included a low incidence of extreme weather, which may have reduced the statistical power. Trial registration This study did not involve any health care interventions for human participants and, therefore, did not require registration.