Extended Data Fig. 3: Sensitivity analysis of heat-related risks of non-fatal myocardial infarction (MI) at a maximum lag of 3 days stratified by medication intake in Augsburg, Germany, during the warm seasons from 2001 to 2014.
From: Triggering of myocardial infarction by heat exposure is modified by medication intake
(a) Effect modification by six medication types with more than 10% of intake frequency among the MI survivors (n = 2,124). The odds ratio (OR) is for the 95th percentile of temperature (95th; 24.2 °C) relative to the minimum MI risk temperature (MMIT; 7.5 °C). Data are presented as point estimates of OR ± 1.96 standard error (that is, 95% confidence intervals). Two-sided P-value (0.060 and 0.028 for antiplatelet and beta-receptor blockers intake, respectively; multiple comparisons were not adjusted) indicates the statistical significance of the difference in risk estimates between users and non-users of a specific medication, based on the z score calculated using the coefficients and standard errors for users and non-users. (b) Exposure-response curves for the temperature-MI relationship stratified by antiplatelet or beta-receptor blockers intake. Solid lines represent the point estimates of OR; shaded areas represent the 95% confidence intervals. Solid vertical lines are the minimum MI temperatures (MMIT; 7.5 °C) and dashed vertical lines are the 95th percentile of temperature (24.2 °C).
