Cold Ambient Temperature and Sudden Cardiac Arrest: Effect of Medical History on Risk Estimates

摘要

BACKGROUND AND AIMS We previously demonstrated an association between cold ambient temperature and the risk of sudden cardiac arrest in Helsinki, Finland. The current study aimed to investigate the effect of previous medical history on the observed risks. METHODS Odds Ratios (ORs) of the risk of out-of-hospital cardiac arrests (OHCAs) vs. temperature were determined by a case-crossover study of OHCAs reported to Emergency Medical systems (EMS) in Helsinki, 1998-2006. Temperature, humidity, ozone and fine particulate levels were independent variables. Exposures were expressed as 24-hr averages in the period preceding the OHCA (Lag Od), the preceding 24-hr periods (Lag 1d, Lag 2d, Lag 3d) and the average of Lag Od through Lag3d (Lag 0d3d). ORs were expressed per 1 oC decrement in ambient temperature. Cases were subsetted into disease categories based on government-determined eligibility for drug reimbursement in specific disease categories. RESULTS ORs significantly above one (p ＜ 0.05) were found for subjects with hypertension (1.04, 95% confidence interval [CI]:1.01,1.07 ) diabetes (1.04,CI:1.00,1.07), subjects with any cardiovascular disease (1.02, CI: 1.00,1.05) and for all subjects (1.03, CI: 1.02, 1.04). However these Ors were not significantly different from each other, or from subjects not in the specified disease category (p ＞ 0.05). The OR for subjects with coronary heart disease was close to one (1.01, CI: 0.98,1.04). In a subset cases where the OHCA was determined by EMS as due to myocardial infarction, the Ors for subjects with coronary heart disease and diabetes were 0.99, CI: 0.94,1.04 and 1.11, CI: 1.03,1.19 respectively. CONCLUSIONS No definitive effect of medical history on the association of OHCA risk with temperature was found. Possible reasons for the apparently lower risk for subjects with previous coronary heart disease are the effects of drug therapy or changes in the subjects' lifestyle following diagnosis of.