Statistics from Altmetric.com
Several studies have shown that the incidence of myocardial infarction (MI) does not follow a continuous pattern throughout the week, and that there is an increase in mortality on Mondays and possibly a decrease on Thursdays.1 The reasons for such a pattern are not well understood, and several hypothesis have been made, namely a change from leisurely weekend activities to stressful work among working patients1 or a Monday increase in the incidence of life threatening ventricular arrhythmias.2 To our knowledge, no assessment of the weekly distribution of MI mortality has been performed in France. Thus, we used the data from the French national mortality statistics and from the French MONICA (monitoring trends and determinants in coronary disease) registers to assess the weekly pattern of incidence and mortality from MI.
Mortality data for coronary heart disease (ICD codes 410–414) was obtained from the French National Mortality Statistics (INSERM SC8, Le Vésinet, France) for years 1987 to 1997. The data were provided as number of deaths according to sex, year, 10 year age group, and day of the week. The corresponding population numbers were obtained from the French National Institute of Statistics and Economical Studies (INSEE, Paris, France). MI events were also collected from the three French MONICA registers (Lille, Strasbourg, and Toulouse) for both sexes, aged 35–64 years. Statistical analysis was performed using Epi-Info (CDC, Atlanta, USA) and SAS (SAS Institute, Cary, North Carolina, USA) statistical software. National mortality data were converted to mortality rates, which were then compared by Wilcoxon test. Data from the MONICA registers was analysed by χ2. A further comparison between weekends and workdays was also performed. Significance was considered for p < 0.05.
The data from the French National Mortality Statistics contained over 17 000 coronary deaths. Average mortality rates for coronary heart disease by day of the week and for the period 1987 to 1997 are indicated in table 1. In men, significant differences were found for age groups 25–34, 35–44, and 45–54 years, higher rates occurring on Sundays and Saturdays, whereas no differences were found for older men (although with a tendency towards higher rates on Mondays) and for women (not shown). Interestingly, in men, the mortality pattern shifted from a higher mortality on weekends among young subjects to a higher mortality on Mondays among older subjects (comparison of daily number of deaths between 25–54 and 55–84 years: χ2 = 11.64, p < 0.07). Pooled data from the three French MONICA showed that, in men, MI events occurred more frequently on weekends and on Wednesdays, and that this difference was mainly caused by first time (incident) events, whereas no between-day difference was found for recurrent events. Comparison between weekends and workdays showed incident cases to be more frequent on weekends (χ2 = 7.75, p < 0.01). For women, the number of events was too low to draw any valid conclusion.
Several limitations should be acknowledged regarding data from national mortality statistics. Although they are complete and represent all deaths which occurred in France, their accuracy is lower than mortality data derived from MI registers such as MONICA. Also, they do not provide information regarding survivors of MI. Nevertheless, those misclassifications, which are inherent in national mortality statistics, would tend to reduce differences between days of the week and thus the real differences might be even greater.
The effect of age on the weekly distribution of MI mortality has seldom been studied. The higher frequency of deaths on weekends found in this study is in agreement with some studies,3 but not with others.1 Although no clear-cut explanation can be provided, it is possible that young adults engage in strenuous activity during weekends, thus increasing the risk of MI.4 This hypothesis is supported by the increase in incident (first time) MI cases on weekends than on workdays observed using MONICA data. Further, a possible explanation for the higher mortality rates on Mondays could be caused by a reporting bias, old subjects with a coronary episode on Sundays delaying their hospital admission to Mondays, thus increasing the number of events for that day. Nevertheless, since the MONICA data includes the precise timing of the MI event, such a hypothesis is unlikely.
We conclude that, for French young men, MI incidence and mortality is higher on weekends than on workdays, whereas for older subjects a trend towards higher MI mortality on Mondays is found. The reasons for such a discrepancy await further investigation