Objectives: The aim of this study was to identify the relationship between low temperatures in winter and mortality due to cancer, cardiovascular diseases and respiratory diseases.
Study design: Case-crossover study.
Methods: A case-crossover study was performed in Cantabria (northern Spain) in the years 2004-2005; 3948 deaths were included. Odds ratios were estimated using conditional logistic regression, stratified by age, sex, and delay of exposure to low temperatures.
Results: There was an inverse dose-response relationship between temperature and mortality in the three causes of death studied; this result was consistent across genders and age groups. The higher OR for cancer mortality was seen on the first day of exposure (OR = 4.91; 95% CI: 1.65-13.07 in the whole population), and it decreased when exposure over several days in a row was considered; people aged 75 years or more were especially susceptible to cold temperatures (OR = 17.9; 95% CI: 2.38-134.8). Cardiovascular (OR = 2.63; 95% CI: 1.88-3.67) and respiratory mortality (OR = 2.72; 95% CI: 1.46-5.08) showed a weaker effect.
Conclusion: There is a striking association between the extreme cold temperatures and mortality from cancer, not previously reported, which is more remarkable in the elderly. These results could be explained by a harvesting effect in which the cold acts as a trigger of death in terminally ill patients at high risk of dying a few days or weeks later.
Copyright © 2013 The Royal Society for Public Health. Published by Elsevier Ltd. All rights reserved.