Seasonal variation in onset of myocardial infarction - A 7-year single-center study in Italy

Like many other serious acute cardiovascular and cerebrovascular events, acute myocardial infarction ( AMI) shows seasonal variation, being most frequent in the winter. We sought to investigate whether age, gender, and hypertension influence this pattern. We studied 4014 ( 2259 male and 1755 female) consecutive patients with AMI presenting to St. Anna Hospital of Ferrara, Italy between January 1998 and December 2004. Some 1131 (28.2%) of the AMI occurred in persons <65 yrs of age, and 2883 (71.8%) in those >= 65 yrs of age. AMI was over-represented in males (82% in the <65 yr group vs. 56.6% in the >= 65 yr group (chi(2) = 13.99; p< 0.001). Hypertension had been previously documented in 964 (24%) of the cases. There were 691 (17.2%) fatal case outcomes; fatal outcomes were significantly higher among the 3054 normotensive ( n = 614 or 20.1%) than the 964 hypertensive cases (n = 77 or 8%; chi(2) = 74.94, p< 0.001). AMIs were most frequent in the winter ( n = 1076 or 26.8% of all the events) and least in the summer ( n = 924 or 23.0% of all the events; chi(2) = 12.36, p = 0.007). The greatest number of AMIs occurred in December (n = 379 or 9.44%), and the lowest number in September (n = 293 or 7.3%; chi(2) = 11.1, p = 0.001). Inferential chronobiological (Cosinor) analysis identified a significant annual pattern in AMI in those >= 65 yrs of age, with a peak between December and February - January for the total sample ( p< 0.005), January for the sample of males ( p = 0.014), February for fatal infarctions ( p = 0.017), and December for non-fatal infarctions (p = 0.006). No such temporal variations were detected in any of these categories in those <65 yrs of age. The annual pattern in AMI was also verified by Cosinor analysis in the following hypertensive subgroups: hypertensive males ( n = 552: January, p = 0.014), non-fatal infarctions in hypertensive patients (n = 887: January, p = 0.018), and elderly normotensives ( n = 1556: November, p = 0.007).