Ambient Temperature Effect on Acute Myocardial Infarction by Risk Factors: Daily Data From 2000 to 2017, Taiwan

BACKGROUND: A U-shaped relationship between temperature and acute myocardial infarction (AMI) was observed, but the risk factors were rarely included. OBJECTIVES: The authors sought to examine AMI’s cold and heat exposure after considering their risk groups. METHODS: Daily data on ambient temperature, newly diagnosed AMI, and 6 known risk factors of AMI for the Taiwan population from 2000 to 2017 were created by linking 3 Taiwan national databases. Hierarchical clustering analysis was performed. Poisson regression was performed on the AMI rate with the clusters along with the daily minimum temperature in cold months (November-March) and the daily maximum temperature in hot months (April-October). RESULTS: There were 319,737 patients with new-onset AMI over 109.13 billion person-days, corresponding to the incidence rate of 107.02 per 100,000 person-years (95% CI: 106.64-107.39 person-years). Hierarchical clustering analysis identified 3 distinct clusters (1: age <50 years, 2: age ≥50 years without hypertension, and 3: mainly age ≥50 years with hypertension) with AMI incidence rates of 16.04, 105.13, and 388.17 per 100,000 person-years, respectively. Poisson regression revealed that below 15 °C, cluster 3 had the highest risk of AMI per 1°C reduce in temperature (slope = 1.011) compared with clusters 1 (slope = 0.974) and 2 (slope = 1.009). However, above the 32 °C thresholds, cluster 1 had the highest risk of AMI per 1 °C increase in temperature (slope = 1.036) compared with clusters 2 (slope = 1.02) and 3 (slope = 1.025). Cross validation showed a good fit for the model. CONCLUSIONS: People ≥50 years of age with hypertension are more susceptible to cold-related AMI. However, heat-related AMI is more prominent in individuals <50 years of age.