A population-based study of the associations of stroke occurrence with weather parameters in Siberia, Russia (1982–92)


Professor Valery Feigin MD, PhD, Clinical
Trials Research Unit, University of Auckland, Private Bag 92019, Auckland, New Zealand (fax: +64 93731710;
e-mail: v.feigin@ctru.auckland.ac.nz).


Previous studies have established a seasonal variation in stroke occurrence, but none have assessed the influence of inclement weather conditions on stroke incidence in a general population of Russia. We performed a stroke population-based study in the Oktiabrsky District of Novosibirsk, Siberia, Russia. Included in the analysis were 1929 patients with their first occurrence of ischemic stroke (IS), 215 patients with their first occurrence of intracerebral hemorrhage (ICH) and 64 patients with their first occurrence of subarachnoid hemorrhage (SAH): all patients were aged between 25 and 74 years. The cumulative daily occurrence of total strokes and stroke subtypes was evaluated in relation to aggregated daily mean values of ambient temperature, relative humidity and air pressure by means of Poisson regression analysis to estimate the rate ratio (RR) with corresponding confidence interval (CI) and to identify the weather parameters of most importance. In a multivariate analysis, with adjustment for the effects of season, solar and geomagnetic activity, and age of the patients, low ambient temperature (RR 1.32; 95% CI 1.05–1.66) and mean value of air pressure (RR 0.986; 95% CI 0.972–0.999) were important predictors of IS occurrence, while mild ambient temperature (RR 1.52; 95% CI 1.04–2.22) was an important predictor of ICH occurrence. No relationship between SAH occurrence and any one of the weather parameters studied was revealed. There was no interaction between any meteorological variables that was statistically significant. Inclement weather conditions are associated with the occurrence of IS and ICH in Siberia, Russia. Among the meteorological parameters studied, low ambient temperature and mean air pressure are the most important predictors of IS occurrence, whereas the occurrence of ICH is associated with mild ambient temperature. There is no association between any one of the weather parameters studied and the occurrence of SAH.