Incidence of stroke and season of the year: evidence of an association.

Evidence of seasonal variation in the incidence of stroke is inconsistent. This may be a likely consequence of one or more methodological shortcomings of the studies investigating this issue, including inappropriate analytic models, insufficient length of time, small sample size, and a regional (vs. national) focus. The authors' objective was to ascertain whether an association exists between season of the year and the incidence of stroke by using a methodological approach designed to overcome these limitations. The authors used a longitudinal study design involving 72,779 veterans hospitalized for stroke at any Veterans Affairs hospital nationally during the years 1986-1995. These data were analyzed by using time series methods. There was clear evidence of a seasonal occurrence for stroke in general. This seasonal effect was found for ischemic stroke, but not for hemorrhagic stroke. The peak occurrence was in mid-May. Neither the region (i.e., climate) nor the race of the patient substantially modified the seasonal trend. An explanation for this pattern remains to be determined.