Dr, S. Moharty et al have presented a meta-analysis on the relationship between exercise level and risk of atrial fibrillation (AF). In what is the largest data analysis to date on this topic they found that women's risk of AF is reduced at all levels of exercise while in men low and moderate levels of exercise seem protective risk of AF increases at the highest levels.
The term coined by Jim Manzi,( ref 5) "high causal density", seems appropriate in regard to the etiology of AF. Both inactivity and allegedly " too much exercise" both increase the risk of AF as do aging,obesity, alcohol , hypertension , perhaps pericardial fat deposition. and apparently being too tall as well as a number of other putative risk factors.The trick is to figure out what is "too much". Moharty's paper suggests there may be no "too much" for women ,at least they did not demonstrate a threshold, a claim which I believe has not been made before.
Reviewing 22 studies (665,750 subjects) they found:
1) In men moderate exercise was protective ,OR .72 , while vigorous exercise increased AF risk with an OR of 3.30 (1.97--4.63). When the authors excluded some small case control studies that had very wide confidence intervals the estimated risk level was reduced but not eliminated to OR of 2.45 ( 1.1-3.8)
2 )In women moderate exercise was protective with an OR of 0.91 while vigorous exercise seemed even better with an OR of 0.72. So for them more was better without discernable limit?
Previous studies were conflicting with some showing in men the so called J or U shaped curve while others claimed greater physical activity was associated with a lower risk of AF. i.e. a continuing downsloping risk curve as exercise levels increase. Moharty's study indicates that in women the curve is continuously down sloping and supported the existence of a U-shaped curve in men.
Intensity levels of exercise in most (all?) of the studies is at best a "coarse grain" indication. Some of the exercise levels were based on the subjects' declaration of their exercise level. In Anderson's study of cross country skier exercise level was defined as finishing one race versus those who finished 5 or more. In some trials AF was self reported, in others AF was physician confirmed. Often exercise level reflects duration of exercise per week with no consideration of level of intensity of exercise and some times quantified by number of endurance events completed usually without consideration of intensity, i.e completion times.
The effect of men's age on risk of AF at various exercise levels is also a consideration. At least it seemed to be in Aizer's paper. This was an analysis of AF in men in a post hoc analysis of the randomized aspirin study as part of the Physicians health study. (Ref 1) The only increased AF risk was shown in men less than age fifty at the highest self reported exercise level.Suggested non-causal explanations for this include survivor effect and the increased risk of so called lone-AF in otherwise healthy runners typically occuring in a younger or middle aged man
Another study presented data that suggested increased risk of AF in younger but not older men. This study by Nikola Drca from Sweden was a long term followup of 44,000 men with retrospective estimates of their exercise levels. Those 30 year olds who exercised more than 5 hours per week has an increased RR (1.19) but that was not the case for fifty year olds. Again, could this be a survivor effect or a high number of so-called "lone parasympathetic AF" in the younger runners? Here we have the relative risk of less than 2, a topic I have ranted about before. An age differential effect was not mentioned in Moharty's paper.
Being the pattern seeking story telling creatures that people are ( ref 4) there are bound to be speculations as to why/how this purported gender difference could occur. Here is one such such speculation found in a 2011 article by Wilhelm ( ref 3). The authors compared male and female non elite runners and found that the men has subtle changes in diastolic function, slightly large left atria ,changes in heart rate variability and higher BP during exercise, a constellation of findings arguably making male runners, holding training levels constant, more at risk for AF and in fact 4 of the 70 men studied did have episodes of paroxysmal AF.
How to determine where moderate ( and allegedly therefore benefit) ends and excessive ( and allegedly therefore harm) begins either with group data or for an individual is the problem, one that I suspect is not solvable.
1. Aizer,a et al. Atrial fibrillation is association with different levels of exercise at different ages in men. Heart 2014;100, 1037-1042.
2.Drca, N Atrial fibrillation is associated with different levels of physical activity at different ages in men. Heart, 2014 ,100 (13), p 1037
3. Wilhelm, M et al.Gender Differences of atrial and ventricular remodeling and autonomic tone in non elite athletes.American Coll cardio nov15,2011, vol 108, pg 1489
4."Humans are pattern seeking story telling creatures and are quite adept at telling stories about patterns whether they are true or not.: Usually attributed to Michael Sherma.The phrase became a theme for Ed Lemer, UCLA economist, in his Book "Macroeconomic Patterns and Stories.
5."Uncontrolled:The surprising payoff of trial-and-error" Jim Manzi ,2012 Basic Books