Citation

Brandl EJ, Lett TA, Bakanidze G, Heinz A, Bermpohl F, Schouler-Ocak M. Int. J. Biometeorol. 2018; 62(5): 843-850.

Affiliation

Psychiatric University Hospital of Charité at St. Hedwig-Hospital, Große Hamburger Str. 5-11, 10115, Berlin, Germany.

Copyright

(Copyright © 2018, International Society of Biometeorology, Publisher Holtzbrinck Springer Nature Publishing Group)

DOI

10.1007/s00484-017-1485-z

PMID

29204686

Abstract

The specific impact of weather factors on psychiatric disorders has been investigated only in few studies with inconsistent results. We hypothesized that meteorological conditions influence the number of cases presenting in a psychiatric emergency room as a measure of mental health conditions. We analyzed the number of patients consulting the emergency room (ER) of a psychiatric hospital in Berlin, Germany, between January 1, 2008, and December 31, 2014. A total of N = 22,672 cases were treated in the ER over the study period. Meteorological data were obtained from a publicly available data base. Due to collinearity among the meteorological variables, we performed a principal component (PC) analysis. Association of PCs with the daily number of patients was analyzed with autoregressive integrated moving average model. Delayed effects were investigated using Granger causal modeling. Daily number of patients in the ER was significantly higher in spring and summer compared to fall and winter (p < 0.001). Three PCs explained 76.8% percent of the variance with PC1 loading mostly on temperature, PC2 on cloudiness and low pressure, and PC3 on windiness. PC1 and PC2 showed strong association with number of patients in the emergency room (p < 0.010) indicating higher patient numbers on warmer and on cloudy days. Further, PC1, PC2, and PC3 predicted the number of patients presenting in the emergency room for up to 7 days (p < 0.050). A secondary analysis revealed that the effect of temperature on number of patients was mostly due to lower patient numbers on cold days. Although replication of our findings is required, our results suggest that weather influences the number of psychiatric patients consulting the emergency room. In particular, our data indicate lower patient numbers during very cold temperatures.

Language: en

Keywords

Autoregressive integrated moving average model; Climate; Factor analysis; Granger causal modeling; Psychiatric emergencies