Investigating changes in mortality attributable to heat and cold in Stockholm, Sweden.

Projections of temperature-related mortality rely upon exposure-response relationships using recent data. Analyzing long historical data and trends may extend knowledge of past and present impacts that may provide additional insight and improve future scenarios. We collected daily mean temperatures and daily all-cause mortality for the period 1901-2013 for Stockholm County, Sweden, and calculated the total attributable fraction of mortality due to non-optimal temperatures and quantified the contribution of cold and heat. Total mortality attributable to non-optimal temperatures varied between periods and cold consistently had a larger impact on mortality than heat. Cold-related attributable fraction (AF) remained stable over time whereas heat-related AF decreased. AF on cold days remained stable over time, which may indicate that mortality during colder months may not decline as temperatures increase in the future. More research is needed to enhance estimates of burdens related to cold and heat in the future.