In the interesting article by Naska et al,1 siesta in healthy individuals was associated with reduced coronary mortality and the association was more pronounced in working men. The authors used the end point of coronary mortality as opposed to all-cause mortality and argue that coronary mortality is an outcome likely to be related to a stress factor. While this may be true, the very adjudication of coronary mortality is fraught with subjectivity. The cause of death usually is determined using death certificates (as was also done in this study), and we know that the information is biased and often inaccurate. Death is usually a complex process (in most cases), and to clearly identify a cardiac cause in patients with multiple comorbidities may not be easy. Although cardiac disease may be severe, it still is not always the primary cause of death. In a study of 384 death certificates in a university hospital, the death certificates were filled incorrectly in 59% of cases,2 presumably because practicing physicians confused the “cause of death” with the “mechanism of death.” Cardiac arrest is the most common mechanism of death in all cases (other than brain death). Studies have shown that the diagnosis of death is proved wrong in about 29% to 42% of cases after necropsy.3,4 This inaccuracy rate will be much higher given the low necropsy rates in general practice. We therefore agree with the observation by Lauer et al5 that using an end point of coronary death may introduce some “softness” to the “hard” end point of all-cause mortality.
Bangalore S, Sawhney S, Messerli FH. Siesta, All-Cause Mortality, and Cardiovascular Mortality: Is there a “Siesta” at Adjudicating Cardiovascular Mortality?. Arch Intern Med. 2007;167(19):2143. doi:10.1001/archinte.167.19.2143-a