Whether from grade school health class, decades of television PSAs, or my own oft-repeated refrains on the topic, anyone reading this will be well aware of the conventional wisdom that exercise reduces risk of heart disease. So when a study this year reported that longer durations of physical activity were associated with increased coronary artery calcification (CAC, a marker of atherosclerotic plaques),1 you can imagine the flurry of questions and concerns that followed. But as is so often the case, these findings are less than they might initially seem, so let’s set the record straight.
What they did
To investigate how the intensity and duration of physical activity (PA) impact the relationship between activity levels and CAC, authors Pavlovic et al. evaluated data from 23,383 generally healthy men (mean age: 51.7±8.3 years) from the Cooper Center Longitudinal Study cohort. Participants had undergone a CAC scan and completed a thorough questionnaire on PA between 1998 to 2019. The questionnaire was designed to assess exercise habits over the previous three months, and using these results, the investigators categorized participants based on their average PA intensity (calculated as total volume divided by total duration and binned as follows: 1 MET; 3 to 5.9 METs; 6 to 8.9 METs; or 9 to 12 METs) as well as their weekly duration of activity (0 hours/week; >0 to <2 hours/week; 2 to 5 hours/week; or ≥5 hours/week). They then analyzed these data to characterize their association with CAC after adjustment for age, smoking status, BMI, circulating glucose and cholesterol levels, and resting systolic blood pressure.
What they found
As one might expect, results on PA intensity demonstrated that CAC levels (as measured in Agatston units, or AU) was inversely correlated with average PA intensity, meaning that higher exercise intensity was associated with less artery calcification. Specifically, for every one-MET increase in average PA intensity, mean CAC decreased by 3.1% (95% CI: -4.6% to -1.6%).
More surprisingly, however, the researchers observed a direct correlation between PA duration and CAC. For each additional hour of weekly exercise duration, mean CAC score increased by 2.4% (95% CI: 1.1% to 3.7%). Further, in comparing higher (≥5 hours/week) versus lower (<5 hours/week) durations as a categorical variable, the authors report that average CAC in the higher duration group were 19.9% (95% CI: 10.1% to 29.7%) greater than averages among the lower duration group.
Pavlovic et al. note that their findings with respect to exercise intensity and duration were independent of each other(Pinteraction=0.246). In all, they conclude that both low intensity and long duration of physical activity are independently associated with elevated CAC, leading to public concerns that exercising too much might in fact be detrimental to cardiovascular health.
Another case of misleading correlations
So do these results mean that everyone should cut down on their exercise time and only focus on short bursts of high intensity training? Absolutely not. Concerns over “too much” exercise as a result of this study are yet another example of inappropriate causal inference based on strictly correlational data, and once again, closer inspection reveals plenty of alternative explanations for the apparent association.
A correlation between weekly PA duration and CAC score does not tell us anything about whether increased PA duration in any way might cause increased artery calcification, and conclusions about causality are even less valid when relying exclusively on a cross-sectional snapshot, as was the case with this investigation. Indeed, a recent longitudinal study by the same research group suggests that such a causal link does not exist, as no differences in CAC progression were observed across different volumes of PA over a mean follow-up period of 7.8 years.2 In that paper, the authors themselves admit that various biases or confounding factors could lead to a discordance between cross-sectional and longitudinal data. It’s even conceivable that a reverse causality relationship might be at play, as those who learn they have a high CAC score might be more inclined to engage in more exercise in an attempt to mitigate their risk of cardiovascular events.
Additionally, Pavlovic et al. report the association between exercise duration and CAC as a dose-dependent relationship (i.e., CAC increases incrementally with every incremental increase in PA duration), but this dose-dependency falls apart when we look at the group of participants recording zero hours of PA per week. Mean CAC score in this group (226.9±625.9) was much higher than it was for any other category of weekly PA duration (the next highest being 199.7±566.6, among the ≥5 hours/week group), and the 0 hours/week group also included a higher percentage of individuals with CAC ≥100 than any other group.
All of this isn’t even to mention the limited representation of the study cohort (mostly white and entirely male) or the fact that the analyses relied on self-reported PA data, a notoriously unreliable means of collecting information. Biases in reporting, particularly with respect to health-related behaviors like exercise, can significantly influence results from participant questionnaires – and thus, influence the results of the studies relying on these questionnaires.
A question of risk and benefit?
Still, Pavlovic et al. are not the first to report a seemingly paradoxical relationship between exercise volume and CAC. Some who are reading this may recall an interview I had with Dr. Ethan Weiss, in which we discussed evidence that endurance athletes with high cardiorespiratory fitness often tend to have an unexpectedly high degree of coronary artery calcification. As he explained, such a link might be plausible given the increased shear stress on vasculature associated with extreme exercise, but even at this upper end of activity levels, evidence to date suggests that increased CAC in these individuals is not associated with increased risk of cardiovascular or all-cause mortality (again, work done in part by the research group behind the Pavlovic et al. study).3
Ultimately, CAC is only valuable as far as it is an indicator of one’s risk of cardiovascular events or mortality – a low CAC score is not an end in itself, but rather a means of ensuring low CV risk. Thus, even if large volumes of exercise do promote CAC progression (again, not clearly established), if this effect doesn’t correspond to an increase in CV risk, it’s possible that the CAC is particularly stable – in which case, the reduction in CV risk associated with more exercise (for instance, by reducing blood pressure and maintaining metabolic health) almost certainly outweigh any increase in risk associated with a higher CAC score.
Exercise is still good bet for heart health
In all, the results of the present study provide very little meaningful information with regard to exercise intensity or duration and their relationship to CAC, let alone to cardiovascular mortality. While we have focused our critiques on the aspect of the study that has received the most attention and concern – i.e., the results regarding exercise duration – it’s worth noting that they similarly apply to results regarding exercise intensity. (Faulty methodology is faulty methodology, whether results happen to align with existing knowledge or not.)
Meanwhile, exercise has many well-established benefits that reduce risk of cardiovascular disease, such as by improving metabolic health, helping to maintain a healthy body weight and body composition, and reducing blood pressure. So on the balance, we have very questionable risks against very clear and substantial rewards. The choice seems obvious to me…
For a list of all previous weekly emails, click here.
References
- Pavlovic A, DeFina LF, Leonard D, et al. Coronary artery calcification and high-volume physical activity: role of lower intensity vs. longer duration of exercise. Eur J Prev Cardiol. 2024;31(12):1526-1534. doi:10.1093/eurjpc/zwae150
- Shuval K, Leonard D, DeFina LF, et al. Physical activity and progression of coronary artery calcification in men and women. JAMA Cardiol. 2024;9(7):659-666. doi:10.1001/jamacardio.2024.0759
- DeFina LF, Radford NB, Barlow CE, et al. Association of all-cause and cardiovascular mortality with high levels of physical activity and concurrent coronary artery calcification. JAMA Cardiol. 2019;4(2):174-181. doi:10.1001/jamacardio.2018.4628
I found this article fascinating and very helpful. Having been told I had CAC when I had a thoracic CT some years ago I was slightly alarmed at the time. A life long non smoker and a keen (if not particularly talented) endurance athlete (best place top ten finisher in ITU World Triathlon age group championships) and still a keen cyclist and weight trainer, I wondered if I was pushing myself too hard aged 71. This analysis by Peter was excellent and has dispelled some of my concerns. Time to get out for a run in the Florida sunshine!
Interesting as always but as this study (and most)only used male participants, this information doesn’t apply to me.
Thank you for taking time to help us. Excellent article. Integrating HIIT into workout at 58 vs the long workouts is making a difference. Good to better understand the conflicting results.
How many women were part of the cohort?
What were their results?
That begs the question- if a person has a calcium score due to heavy exercise (5 Ironmans so far) and a normal lipid panel, should they be on statins? I have had two doctors say yes but that makes no sense.
Peter is a big believer in low LDL numbers which statins help with. Perhaps you could discuss with your cardiologist.
Anecdotal comment. I’m a 77 year-old male. I’ve been a distance mountain runner/hiker for decades including 15 Pikes Peak Ascent races and several Imogene Pass races. Other stuff includes a 36 mile Doc Holliday race and summiting 14,000 ft peaks over 85 times. No history of family heart disease. Rarely eat red meat. Last year my calcium score was 1346, but the plaque is evenly distributed with no blockages or significant narrowing.
My CAC score, about 8 years ago (I am 73 now) was 1800 (way off the chart: I suppose it is even higher now). My G.P. was very concerned. But my cardiologist did not think it was concerning, as no blockage of any sort was indicated. In my 30s and 40s I was a competitive runner (2:47 marathon and 24-hr 100 mile) and obviously trained very hard in that period. I have kept it up now, mostly with nordic skiing (50 km race every March). So this article and the cited study seem to confirm my cardiologist’s view.
Aside: in 2013 I started having atrial fibrillation. An ablation in 2018 cured it completely! This too might be related to many years of intense training.
Your experience mirrors mine including the AFib. I have a pacemaker now for bradycardia.
Thank you for sharing your research and knowledge!
As one of those individuals with a long history of high-intensity, high-duration exercise (40 years running marathons, racing bikes, triathlon, etc.) and a very high CAC score (800+ Agatston units) I can only take partial comfort in these conclusions. Sure, the plaques I’ve acquired may be more stable and less threatening than those who don’t excersie much … but I wish I didn’t have them at all. I take statins (and have never noticed side effects) and try to follow other healthy lifestyle choices, but I dislike wondering if I’d be better off limiting my exercise habit to brisk walking rather than the many things I find more inspiring and satisfying.
Isn’t it also likely that the type of CAC in exercisers is the more stable type, vs the softer less stable type?
I particularly agree with this sentence :
“ It’s even conceivable that a reverse causality relationship might be at play, as those who learn they have a high CAC score might be more inclined to engage in more exercise in an attempt to mitigate their risk of cardiovascular events. ”
Indeed, it seems to me that many people who consider themselves at “high risk” of CAVD , tend to either (1) avoid any PA at all or (2) gravitate towards frequent and low intensity PA. As you said, it would be much more beneficial to perform a longitudinal study where difference between initial and final CAC is observed as a function of PA duration.
Well hopefully more people will get fit less fat less stress
Thank you Peter
Still wonder: what is best, moderate or intense exercise?
Thank you. Would love to hear more exploration on this topic of exercise, CAC, and CV risk. This “MARC-2” study on exercise intensity vs volume in middle-aged men suggests that too much high intensity training leads to higher CAC scores. I don’t think the findings change Dr Attia’s take homes, but perhaps this aspect can be folded into a future CV topic for more robustness. https://pubmed.ncbi.nlm.nih.gov/28450347/
I’m an active professional ,age 74. Never was a smoker nor a drinker. I have been an athlete
all my life , used to play Badminton, Tennis and Cricket 7 days a week during my college days and a NCAA player in Badminton in my 20’s. Have been captaining USTA Tennis Teams over 30 years Plus . My cholesterol levels was never more than 200, Triglycerides less than 60, LDL has been below normal and HDL has been above normal with out any meds. Diagnosed to have Type II AODM with in last 10 years, no Hypertension. My calcium score was 1100, followed by Stress Test, absolutely normal, CT Angio showed 90% blockage of LAD. Never had SOB, Chest discomfort, Cardiolite Stress test was normal, followed by Catherization with 2″ long 95% blockage of LAD requiring 2 stents. Probably partly Genetics and dietary life style
As a retired cardiovascular radiologist I have observed that the coronary arteries are larger in marathoners and probably in others that train for long periods. Thus the increased calcification and presumable plaques do not narrow the lumen enough to be clinically significant.
Got a study today in my feed on Vigorous Interval Training reducing immune response. Now reading this … seems a convergence of interest on questioning exercise while supporting it. Just reminds me of the hypervitaminosis case of a 89 year old man taking vitamin D. EIGHTY-NINE YEARS OLD. Anyway they fail to mention he was essentially taking a bottle of pills per day, while reporting vitamin D levels low in the populace.
PS Also don’t go out in the sun due to melanoma…
As noted, there is no direct evidence that high intensity exercise with or without factoring in duration, is causal to CAC, especially when based on cross-sectional data with weaknesses beyond the design.
It is also important to note that there is even less evidence that there is no causal association. By “less evidence” I mean there is more weak evidence of an association that there is weak evidence of no association. Moreover, the theoretical evidence that rapid high-pressure pulses of blood at artery branches (“hammering”) can damage arteries, perhaps setting up inflammatory conditions.
The answers to these questions will likely come from the development of a concsnsus among different forms of research including observational studies and enhanced studies of high-pressure blood flow at arterial junctions. I’m agnostic on the outcome but will not be surprised to learn that there are direct causal pathways.
Love it…Curious to know what David Goggins’ Levels are cause that’s the standard bearer for Ultra-Exercise, IMO.