The article “Elevated resting heart rate, physical fitness and all-cause mortality: a 16-year follow-up in the Copenhagen Male Study”, in the journal Heart (doi:10.1136/heartjnl-2012-303375) says that a low resting heart rate is predictive of longevity, even after correcting for “physical fitness, leisure-time physical activity and other cardiovascular risk factors”, at least among healthy, middle-aged Caucasian men.
This is good news for me, as I have a low resting heart rate (around 52 bpm) and fit the other criteria.
I’m a little unclear on how they did the correction, as the fitness (measured with VO2max) was very highly negatively correlated with heart rate (R=–0.34). They say they used “Cox proportional hazards regression analysis”, which seems to me to suffer the problem of all regression methods when the input variables are highly correlated. They do claim “Assumptions regarding the use of Cox proportional hazards were met by inspection of the log minus log function at the covariate mean.”
The are showing a huge difference between resting rates under 50 bpm and over 50 bpm, then small differences up to 80 bpm, and big increases for over 80 bpm and even bigger for over 90bpm. I wish that they had divided the group into equally populated classes, rather than every 10 bpm, as the 51–60bpm group has almost 36% of the subjects (1003/2978), and I’m pretty sure the low and high end of that group are different. I also wish that they had used the median group (61–70 bpm) as their baseline, rather than the unusual group having heart rates under 50 bpm. The 51–60bpm group has only a 5% lower risk than the 61–70 bpm group, while the under 50bpm group has about a 30% lower risk. It would be interesting to have analyzed the data to try to get a smooth curve of risk vs. heart rate, though that would have required more sophisticated modeling and the data may not be sufficient to produce a reasonable curve.