gms | German Medical Science

Objective: To investigate the meteorological influence on the incidence of spontaneous intracerebral hemorrhage (ICH).

Method: We analyzed 295 ICH admissions between 2005 and 2013 and excluded traumatic ICH and ICH due to aneurysmal subarachnoid hemorrhage. ICH related to cavernoma or arteriovenous malformation were included as a vascular subgroup. The hourly meteorological parameters considered are: surface pressure, 2-m temperature, relative humidity and wind gusts. In addition to seasonality, we investigated for all parameters three complementary statistical measures: i) the deviation from the 10-year monthly mean; ii) the deviation from the 10-year monthly mean the change relative to the parameter's value two days before the ictus; iii) the time evolution from 5 days before to 5 days after the ICH occurrence. The statistical significance of the results is determined using a Monte Carlo simulation combined with a re-sampling technique (1000x).

Results: Although a high-resolution data set with excellent temporal and spatial resolution was used, no statistically significant and meteorological meaningful signal could be found in the overall patient population. In subgroup analyses, we saw a significant signal in the vascular subgroup (n = 80) with relative humidity below what is expected from climatology on ICH bleeding days (p=0.03) and the opposite signal higher relative humidity on bleeding days for the subgroup with cerebral amyloid angiopathy (n=47; p=0.01). Both signals could be confirmed in the independent complementary analyses ii) and iii).

Conclusions: By using high-quality meteorological data analyzed with a sophisticated and robust statistical method we found higher relative humidity on bleeding days in the vascular subgroup and lower humidity in the cerebral amyloid angiopathy group.