gms | German Medical Science

Objective: To examine a possible association between standard meteorological variables and their changes and the occurrence of SAH.

Methods: Patients with spontaneous SAH were considered for this study. We recorded the baseline demographic data past medical history, timepoint of bleeding (day and hour of onset, clustering of SAH patients) and neurological status on admission. The meteorological variables comprised atmospheric pressure, air temperature, humid temperature, vapor pressure and the relative humidity in %. These values were measured at 3 timepoints per day. Minimum, maximum, mean and maximal differences were calculated for each day. The atmospheric pressure was additionally dichotomized for the day of bleeding with values above and below 10 hp. Univariate association between the variables was tested using Chi square tests for categorial variables, and Mann-Whitney U-test for continuous variables. In a second step a multivariate analysis was performed to find independent meteorologic predictors for the occurrence of SAH using a binary logistic regression analysis.

Results: 2347 days were evaluated from which 465 were bleeding days and 1882 were non-bleeding days, with a total of 519 patients. 48 cluster days with a total of 102 and 77 bleeding series (consecutive SAH days) with a total of 180 patients were seen. Univariate analysis showed significant differences between bleeding days and non-bleeding days for the number of change days (maximal atmospheric differences of the day >10 hp): p<0,001 (OR 4,6; 95% CI 2,9-7,5); for the maximal relative humidity: p<0,05; for the maximal difference of vapor pressure of the day 24 h before the bleeding day: p<0,006 and between cluster days and non cluster days for the number of change days p<0,001 (OR 6,7; 95% CI 2,8-15,8); for the maximal difference of temperature of the day p<0,035, and for the maximal, minimal and mean relative humidity: p<0,027, p<0,018, p<0,03 respectively. In the multivariate analyses none of these variables was retained as independent predictor for the occurrence of SAH in the model.

Conclusions: Atmospheric pressure changes of more than 10 hp per day are associated with an increased risk of SAH and with an increased probability of an accumulation of SAH patients however the multivariate analyses show that the effect of weather variables is not as strong as suggested in earlier studies.