Abstract

The recent decline in stroke mortality, and its seasonal variation, have not been satisfactorily explained through any single factor. Nevertheless, several causes might operate through a single mechanism, namely salt loss variation. The increased use of diuretics could explain the trend, and physiological salt loss variations might explain the cycle. The associations between mortality and meteorological variables were therefore examined. The examination was negative in that temperature correlations were equally strong in winter and in summer, with no support for the hypothesis that temperature-dependent salt loss was a contributing cause. It was found in addition, unexpectedly, that stroke mortality showed strong correlations with atmospheric pollution levels, both in winter and in summer. These correlations were strengthened, rather than dissipated, by standardisation for season and for temperature. The pattern for stroke mortality differed, in these respects, from acute myocardial infarction. The pollution correlations of hypertension deaths were similar to those from stroke, and they were jointly more powerful than correlations with deaths from bronchitis.