The short-term effects of ambient cold temperature on mortality have been well documented in the literature worldwide. However, less is known about which subpopulations are more vulnerable to death related to extreme cold. We aimed to examine the personal characteristics and underlying causes of death that modified the association between extreme cold and mortality in a case-only approach. Individual information of 197,680 deaths of natural causes, daily temperature, and air pollution concentrations in cool season (November-April) during 2002-2011 in Hong Kong were collected. Extreme cold was defined as those days with preceding week with a daily maximum temperature at or less than the 1st percentile of its distribution. Logistic regression models were used to estimate the effects of modification, further controlling for age, seasonal pattern, and air pollution. Sensitivity analyses were conducted by using the 5th percentile as cutoff point to define the extreme cold. Subjects with age of 85 and older were more vulnerable to extreme cold, with an odds ratio (OR) of 1.33 (95 % confidence interval (CI), 1.22-1.45). The greater risk of extreme cold-related mortality was observed for total cardiorespiratory diseases and several specific causes including hypertensive diseases, stroke, congestive heart failure, chronic obstructive pulmonary disease (COPD), and pneumonia. Hypertensive diseases exhibited the greatest vulnerability to extreme cold exposure, with an OR of 1.37 (95 % CI, 1.13-1.65). Sensitivity analyses showed the robustness of these effect modifications. This evidence on which subpopulations are vulnerable to the adverse effects of extreme cold is important to inform public health measures to minimize those effects.
Keywords: Case-only study; Cause of death; Effect modifiers; Extreme cold; Mortality.