Abstract
Importance
Prior cross-sectional studies have suggested that very high levels of physical activity (PA) are associated with a higher prevalence of coronary artery calcium (CAC). However, less is known regarding the association between high-volume PA and progression of CAC over time.
Objective
To explore the association between PA (measured at baseline and during follow-up) and the progression of CAC over time.
Design, Setting, and Participants
This cohort study included data from 8771 apparently healthy men and women 40 years and older who had multiple preventive medicine visits at the Cooper Clinic (Dallas, Texas), with a mean (SD) follow-up time of 7.8 (4.7) years between the first and last clinic visit. Participants with reported PA and CAC measurements at each visit during 1998 to 2019 were included in the study. Data were analyzed from March 2023 to February 2024.
Exposures
PA reported at baseline and follow-up, examined continuously per 500 metabolic equivalent of task minutes per week (MET-min/wk) and categorically: less than 1500, 1500 to 2999, 3000 or more MET-min/wk.
Main Outcomes and Measures
Negative binomial regression was used to estimate the rate of mean CAC progression between visits, with potential modification by PA volume, calculated as the mean of PA at baseline and follow-up. In addition, proportional hazards regression was used to estimate hazard ratios for baseline PA as a predictor of CAC progression to 100 or more Agatston units (AU).
Results
Among 8771 participants, the mean (SD) age at baseline was 50.2 (7.3) years for men and 51.1 (7.3) years for women. The rate of mean CAC progression per year from baseline was 28.5% in men and 32.1% in women, independent of mean PA during the same time period. That is, the difference in the rate of CAC progression per year was 0.0% per 500 MET-min/wk for men and women (men: 95% CI, −0.1% to 0.1%; women: 95% CI, −0.4% to 0.5%). Moreover, baseline PA was not associated with CAC progression to a clinically meaningful threshold of 100 AU or more over the follow-up period. The hazard ratio for a baseline PA value of 3000 or more MET-min/wk vs less than 1500 MET-min/wk to cross this threshold was 0.84 (95% CI, 0.66 to 1.08) in men and 1.16 (95% CI, 0.57 to 2.35) in women.
Conclusions and Relevance
This study found that PA volume was not associated with progression of CAC in a large cohort of healthy men and women who were initially free of overt cardiovascular disease.
This cohort study examines data for men and women 40 years and older to determine the association between high-volume leisure-time aerobic physical activity and the progression of coronary artery calcium over time.
Key Points
Question
What is the association between high-volume leisure-time aerobic physical activity and the progression of coronary artery calcium (CAC) over time?
Findings
In this cohort study involving 8771 apparently healthy adults 40 years and older, high and very high volumes of physical activity during follow-up were unrelated to CAC progression. Moreover, higher baseline physical activity volumes were not associated with clinically meaningful CAC at follow-up.
Meaning
Ongoing leisure-time physical activity, even at high volumes, was unrelated to the progression of CAC, a marker of atherosclerotic cardiovascular disease.
Introduction
Habitually engaging in physical activity (PA) reduces the risk for premature death from all causes, cardiovascular disease, and cancer.1,2,3 Moreover, PA markedly lowers the risk of incurring chronic diseases such as type 2 diabetes, cardiovascular disease (eg, coronary heart disease and stroke), some types of cancer, and depression. In fact, not engaging in sufficient levels of PA (ie, not meeting the recommended guidelines) is estimated to cause 6% to 10% of major chronic diseases worldwide, such as coronary heart disease.4 To this end, the US Department of Health and Human Services recommends that adults engage in 150 minutes or more of moderate intensity aerobic PA or at least 75 minutes of vigorous-intensity activity each week (or an equivalent combination).5,6 These guidelines translate into performing at least 500 metabolic equivalents of task (MET) minutes per week (min/wk) to meet guidelines, or more than 1000 MET-min/wk to exceed guidelines.5,6,7,8,9,10,11
Meeting or exceeding the recommended PA guidelines is generally beneficial to reducing cardiometabolic risk factors,11 as well as improving adaptive physiological traits such as myocardial remodeling and increased myocardial and vascular compliance.12,13 However, some physically active individuals could incur cardiovascular events and sudden cardiac death.12,14 High volumes of aerobic PA (eg, 4-6 times the suggested guidelines) have been linked cross-sectionally to elevated subclinical coronary artery atherosclerosis.12,14,15,16 For example, in a 2017 study of 248 men, Aengevaeren et al15 found that a high volume of PA (>2000 MET-min/wk) was linked to higher CAC. Similarly, in our previous study from 2019 among 21 278 men, we found that men who engaged in very high levels of PA (≥3000 MET-min/wk) had a higher relative risk (11%) of elevated CAC (≥100 AU) than those performing less PA.14 Yet, regardless of CAC level, high volumes of PA were not associated with higher mortality risk.
It is important to note that both our prior study14 and the one by Aengevaeren et al15 assessed PA and CAC at 1 point in time (ie, cross-sectional), which prohibits determining a temporal relationship. In contrast, data are sparse on the longitudinal relationship between PA volume (including high volumes) and progression of CAC over time. Hence, the aims of the current study are 2-fold: (1) to test whether the rate of progression of mean CAC over multiple clinic visits (1998-2019) is associated with the mean level of PA during the same time period and (2) to determine whether baseline levels of PA, including high and very high volumes of exercise, are associated with CAC progression beyond a clinically meaningful threshold (CAC ≥100 AU). These aims are explored among community-dwelling men and women 40 years and older who enrolled in the Cooper Center Longitudinal Study (CCLS).
Methods
We examined the association between leisure-time PA and CAC progression among patients presenting at the Cooper Clinic (Dallas, Texas), a preventive medicine practice focused on lifestyle medicine.17 Participants in the study opted into the CCLS and provided written informed consent. The CCLS is a prospective cohort study that aims to assess the association of lifestyle behaviors with morbidity and mortality.18 Study participants are community-dwelling adults who are self-referred for preventive care or referred to the clinic by their employer. They are predominantly White and well educated and have access to medical care.14,18 The Cooper Institute institutional review board evaluates and approves the CCLS on an annual basis.
The current study considered 9665 individuals who were 40 years and older, had 2 or more clinic visits between 1998 and 2019, and had complete information on the primary study variables (eg, PA measurements and CAC scans at each visit). Participants with a personal history of myocardial infarction (n = 290), stroke (n = 175), cardiac stent (n = 168), coronary artery bypass surgery (n = 66), or body mass index (BMI) less than 18.5 (calculated as weight in kilograms divided by height in meters squared) (n = 195) were excluded.
Measures
Primary Exposure: Leisure Time Physical Activity
Leisure time PA volume was calculated from responses to survey questions pertaining to weekly frequency and mean duration in leisure-time aerobic activities during the previous 3 months.19 These activities included walking, jogging, treadmill, bicycling, swimming, tennis, basketball, soccer, fitness class, water aerobics, boot camp, elliptical, rowing, jump rope, golf (without a cart), dance, stairs, hiking, and cross-country skiing. The Compendium of Physical Activities was used to assign intensity values (in METs) to individual activities.20 Moderate and vigorous aerobic PA volume (MET-min/wk) was calculated by calculating each activity’s intensity × total minutes per week. Jogging and treadmill speed was used to refine the intensity assignment for both of these activities. If speed was not reported, jogging was entered at 8 METs, and treadmill was entered at 4.25 METs. The calculated PA volume correlated moderately with cardiorespiratory fitness (CRF) estimated using the modified Balke treadmill protocol21: Spearman correlation 0.49 for both men (n = 6168) and women (n = 1844). PA was analyzed as a continuous variable (in units of 500 MET-min/wk) or categorical variable (<1500 [referent], 1500-2999 [high], and ≥3000 [very high] MET-min/wk).14 A PA volume of 3000 MET-min/wk corresponds to approximately 5 hours per week of vigorous intensity PA, such as running at 6 miles per hour (10 minutes per mile).14,20 For analysis, baseline PA was used as well as the mean MET-min/wk at baseline and follow-up visits.
Primary Outcome: Coronary Artery Calcification
CAC was measured using electron beam tomography from 1998 to 2007 using the Imatron C-150XP or C-300 (GE Imatron).14 In 2008-2019, a 64-slice scanner (Lightspeed VCT; GE Healthcare) was used.22 Images were obtained while adhering to the standard breath-holding protocol.23 The methods for determining the Agatston calcium score have been described elsewhere.24 CAC quantification in this cohort is highly reproducible with minimal bias.25 CAC scores were analyzed as continuous variables, and time to progress to a CAC of 100 Agatston units (AU) or more was considered interval censored between clinic visits. In addition, mortality from all causes and cardiovascular disease (National Death Index Plus) was examined as a secondary outcome among a subsample of participants.
Covariates
The covariates included age (years), current smoking (no/yes), and BMI. BMI was computed based on measured weight and height using the standard formula.26 Additionally, 12-hour fasting venous blood samples were collected and assayed for fasting glucose and lipids, whereas blood pressure was measured at the clinic while adhering to an established protocol. Additional covariates included statin use (no/yes), type of CAC scanner, and time since the baseline clinic visit in the repeated measures growth model.
Statistical Analysis
Characteristics of participants at their first and last clinic visits were summarized using mean (SD) or number and percentage. Changes in characteristics between the first and last clinic visits were tested using paired t statistics for continuous variables and paired marginal homogeneity statistics for categorical variables. CAC changes and rates of change between first and last clinic visits were summarized by sex, baseline CAC, and baseline PA; trends across PA categories were tested using nonparametric Jonckheere-Terpstra statistics.
For the study’s aim 1, we used a repeated measures growth model,27 to relate CAC scores at each visit to time since baseline, and covariates on the entire sample (n = 8771). In the basic growth model, measured CAC at each visit was modeled as a negative binomial distributed count of Agatston units. The logarithm of predicted CAC at each visit was regressed on an intercept, PA at baseline, age at baseline, time since baseline, and the product of time since baseline and mean PA during the same time period, calculated as half the sum of baseline and follow-up PA. The intercept and coefficient of time since baseline were specified as random across participants. We refined the basic model by adding quadratic terms involving age and time since baseline, additional covariates at baseline, and additional covariates during follow-up, as for PA. This model adjusted for baseline age, as well as baseline and follow-up of type of scanner, current smoking, BMI, fasting glucose, total cholesterol, systolic blood pressure, and statin use. The log link connects this modeling strategy to other transformation-based progression measures.28 However, in this case, the log applies to the conditional mean of the fitted distribution instead of the CAC data, avoiding problems with CAC zeros. Details pertaining to statistical modeling of aim 1 appear in the eMethods in Supplement 1. The same model was applied to the subgroup with CAC greater than 0 at baseline (n = 3716) to explore sensitivity to participants with a baseline CAC of 0.
For aim 2, we used proportional hazards regression to examine the association between baseline PA and progression to CAC of 100 AU or more. Time to CAC progression was treated as interval censored between clinic visits.29 The analysis was applied to the subgroup of 7391 participants with CAC less than 100 AU at baseline and adjusted for age, baseline CAC, type of CAC scanner, current smoking, BMI, fasting glucose, total cholesterol, systolic blood pressure, and statin use. The same model was applied to participants with a CAC of 0 at baseline (n = 5055) to explore progression to CAC greater than 0 AU.
Follow-up for mortality extended from the last clinic visit to death or December 31, 2017, our last follow-up date for mortality. Crude incidence of all-cause and CVD mortality was calculated by dividing the corresponding number of deaths by the total follow-up time for mortality among participants whose last clinic visit was on or before December 31, 2017 (n = 6395). Trends in crude incidence across ordered categories of baseline PA within categories of baseline CAC by sex were tested in exponential survival models stratified by sex and baseline CAC. All models were fit separately to men and women. Analyses were programmed in SAS/STAT version 9.4 (SAS Institute). Data were analyzed from March 2023 to February 2024.
Results
Participants’ characteristics at baseline and the last follow-up clinic visit stratified by sex appear in Table 1. Among 8771 participants, mean (SD) ages of men and women at baseline were 50.2 (7.3) years and 51.1 (7.3) years, respectively. Mean PA in men was 1247 MET-min/wk at baseline and 1406 MET-min/wk at the last follow-up, whereas in women, it was 1206 MET-min/wk at baseline and 1283 MET-min/wk at the last follow-up. Mean CAC score increased from 95.9 to 222.1 AU in men and from 20.0 to 48.0 AU in women. The mean (SD) time between the baseline and last follow-up was 7.8 (4.8) years for men and 7.7 (4.4) years for women. In addition, the prevalence of statin use increased between baseline and last follow-up from 17% to 39% (for prevalence by sex, see Table 1). This increase was similar across categories of PA; namely, statin use increased from 18% at baseline to 40% at the last follow-up in the referent PA group, from 18% to 38% in the high PA group, and from 14% to 34% in the very high PA group. Finally, all participants had at least 2 clinic visits, with 46% having more than 2 visits, while 96% had 5 or fewer visits. One-quarter of the participants were followed up for longer than 10 years.
Table 1. Characteristics of Study Participants at First and Last Clinic Visits: CCLS 1998-2019 (N = 8771).
| Characteristic | No. (%) | |||||
|---|---|---|---|---|---|---|
| Men (n = 6661) | Women (n = 2110) | |||||
| First visit | Last visit | P value | First visit | Last visit | P value | |
| Age, mean (SD), y | 50.2 (7.3) | 58.0 (8.2) | <.001 | 51.1 (7.3) | 58.8 (8.1) | <.001 |
| Time since baseline, mean (SD), y | 0.0 (0.0) | 7.8 (4.8) | NA | 0.0 (0.0) | 7.7 (4.4) | NA |
| Current smoker | 786 (11.8) | 573 (8.6) | <.001 | 91 (4.3) | 55 (2.6) | <.001 |
| Statin use | 1345 (20.2) | 2960 (44.4) | <.001 | 187 (8.9) | 488 (23.1) | <.001 |
| PA, mean (SD), MET-min/wk | 1247 (1309) | 1406 (1556) | <.001 | 1206 (1212) | 1283 (1455) | .01 |
| PA referenta | 4638 (69.6) | 4361 (65.5) | <.001 | 1492 (70.7) | 1469 (69.6) | .13 |
| High PA, 1500-2999 MET-min/wka | 1490 (22.4) | 1640 (24.6) | 468 (22.2) | 462 (21.9) | ||
| Very high PA, ≥3000 MET-min/wka | 533 (8.0) | 660 (9.9) | 150 (7.1) | 179 (8.5) | ||
| CRF, mean (SD), METsb | 11.7 (2.1) | 11.2 (2.1) | <.001 | 9.7 (1.8) | 9.4 (1.8) | <.001 |
| BMI, mean (SD)c | 27.5 (3.8) | 27.8 (3.9) | <.001 | 24.6 (4.4) | 25.1 (4.5) | <.001 |
| Total cholesterol, mean (SD), mg/dL | 194.8 (35.9) | 180.9 (37.8) | <.001 | 200.1 (33.5) | 203.5 (36.8) | <.001 |
| Fasting glucose, mean (SD), mg/dL | 98.0 (13.9) | 99.8 (16.5) | <.001 | 92.3 (11.5) | 93.9 (11.7) | .001 |
| Blood pressure, mean (SD), mm Hgb | ||||||
| Systolic | 122.6 (12.6) | 123.2 (12.8) | <.001 | 115.8 (14.7) | 118.6 (14.7) | <.001 |
| Diastolic | 82.3 (9.1) | 79.9 (8.5) | <.001 | 77.4 (9.2) | 76.4 (8.8) | <.001 |
| CAC score | ||||||
| Mean (SD), AUd | 95.9 (301.0) | 222.1 (485.9) | <.001 | 20.0 (93.1) | 48.0 (158.1) | <.001 |
| 0 AU | 3379 (50.7) | 2063 (31.0) | <.001 | 1676 (79.4) | 1366 (64.7) | <.001 |
| 1-99 AU | 2014 (30.2) | 2196 (33.0) | 322 (15.3) | 498 (23.6) | ||
| ≥100 AU | 1268 (19.0) | 2402 (36.1) | 112 (5.3) | 246 (11.7) | ||
Abbreviations: AU, Agatston units; BMI, body mass index; CAC, coronary artery calcification; CCLS, Cooper Center Longitudinal Study; CRF, cardiorespiratory fitness; MET-min/wk, metabolic equivalent of task minutes per week; NA, not applicable; PA, physical activity.
The PA referent was <1500 MET-min/wk. For the PA referent group: among men, the mean (SD) for the first visit was 617 (462) MET-min/wk, and for the last visit, it was 650 (458) MET-min/wk; whereas among women, the mean (SD) for the first visit was 608 (456) MET-min/wk, and for the last visit, it was 598 (450) MET-min/wk. For the high PA group: among men, the mean (SD) for the first visit was 2071 (410) MET-min/wk, and for the last visit, it was 2087 (422) MET-min/wk; whereas among women, the mean (SD) for the first visit was 2122 (426) MET-min/wk, and for the last visit, it was 2086 (405) MET-min/wk. For the very high PA group: among men, the mean (SD) for the first visit was 4424 (1914) MET-min/wk, and for the last visit, it was 4710 (2601) MET-min/wk; whereas among women, the mean (SD) for the first visit was 4300 (1509) MET-min/wk, and for the last visit, it was 4833 (2179) MET-min/wk.
In men, there were 645 and 1485 missing CRF observations for the first and last visit, respectively. In women, there were 316 and 689 missing CRF observations for the first and last visit, respectively. Regarding missing observations of diastolic blood pressure: in men, there was 0 and 1 missing diastolic observation for the first and last visit, respectively; whereas in women, there was 1 and 1 missing diastolic observation for the first and last visit, respectively.
Calculated as weight in kilograms divided by height in meters squared.
Median first and last visit CAC scores were 0 and 32 AU for men and 0 and 0 AU for women, respectively.
Observed CAC changes and rates of change between participants’ first and last clinic visits by baseline CAC and PA stratified by sex are shown in Table 2. Higher levels of baseline CAC were associated with larger CAC changes (in AU) and rates of change (in AU/y) for both men and women. However, higher levels of baseline PA within CAC categories were generally not associated with CAC changes or rates of change. One exception was for men with a baseline CAC of 1 to 99 AU, where higher baseline levels of PA were associated with a lower rate of change of CAC (P = .049). Similarly, in the subgroup of participants with follow-up for mortality, higher levels of baseline PA within CAC categories were not associated with all-cause or cardiovascular disease mortality, with the exception of significantly lower all-cause mortality with higher PA in men with a baseline CAC 1 of 99 AU (P = .04) (eTables 1 and 2 in Supplement 1).
Table 2. CAC Changes Between First and Last Clinic Visits by Baseline CAC and Physical Activity Among Men and Women (n = 8771).
| Change | Baseline CAC 0 AU, PA at baseline, MET-min/wk | Baseline CAC 1-99 AU, PA at baseline, MET-min/wk | Baseline CAC ≥100 AU, PA at baseline, MET-min/wk | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0-1499 | 1500-2999 | ≥3000 | P value | 0-1499 | 1500-2999 | ≥3000 | P value | 0-1499 | 1500-2999 | ≥3000 | P value | |
| Men (n = 6661) | ||||||||||||
| No. | 2381 | 748 | 250 | 1417 | 436 | 161 | 840 | 306 | 122 | |||
| Time since baseline, mean (SD), y | 8.5 (4.9) | 8.4 (4.8) | 8.0 (4.6) | .002 | 7.4 (4.6) | 7.2 (4.6) | 6.5 (4.5) | .19 | 6.8 (4.6) | 6.8 (4.7) | 7.1 (4.7) | .50 |
| Change in CAC, mean (SD), AU | 26.5 (91.0) | 27.2 (113.0) | 17.9 (65.9) | .13 | 134.8 (224.8) | 135.1 (270.3) | 84.2 (121.8) | .20 | 392.9 (512.4) | 364.8 (467.2) | 391.0 (509.8) | .89 |
| Rate of change of CAC, mean (SD), AU/y | 2.3 (7.0) | 2.2 (7.5) | 1.5 (4.2) | .13 | 15.8 (20.3) | 14.7 (19.3) | 11.5 (13.7) | .049 | 59.8 (86.7) | 57.0 (59.9) | 51.3 (50.2) | .61 |
| Women (n = 2110) | ||||||||||||
| No. | 1184 | 373 | 119 | 220 | 76 | 26 | 88 | 19 | 5 | |||
| Time since baseline, mean (SD), y | 7.9 (4.4) | 8.4 (4.6) | 7.1 (4.0) | .97 | 7.1 (4.2) | 6.9 (4.3) | 6.6 (4.6) | .76 | 6.2 (4.3) | 6.8 (3.9) | 3.6 (2.7) | .50 |
| Change in CAC, mean (SD), AU | 8.0 (30.3) | 7.1 (35.3) | 4.7 (19.7) | .16 | 68.9 (93.7) | 83.5 (165.0) | 62.7 (83.8) | .99 | 205.4 (238.4) | 208.3 (189.5) | 254.6 (160.6) | .66 |
| Rate of change of CAC, mean (SD), AU/y | 0.7 (2.6) | 0.6 (2.4) | 0.6 (2.3) | .16 | 9.6 (11.3) | 10.5 (13.6) | 9.6 (11.7) | .55 | 39.4 (45.8) | 33.3 (27.8) | 93.6 (104.1) | .80 |
Abbreviations: Agatston units, AU; CAC, coronary artery calcification; MET-min/wk, metabolic equivalent of task minutes per week; PA, physical activity.
The mean CAC progression estimates are presented in Table 3. The rate of mean CAC progression per year from baseline was 28.5% in men and 32.1% in women, independent of mean PA during the same time period. Namely, the difference in the rate of CAC progression per year was 0.0% per 500 MET-min/wk for men and women (men: 95% CI, −0.1% to 0.1%; women: 95% CI, −0.4% to 0.5%). Furthermore, mean CAC progression estimates for men and women followed up for 10 years starting from age 50, 60, or 70 years are illustrated in Figure 1. At each starting age, parallel progression curves comparing mean PA levels of 0, 1500, or 3000 MET-min/wk showed equal relative rates of CAC progression. Similar findings were observed when comparing high (1500-2999 MET-min/wk) and very high (≥3000 MET-min/wk) PA levels to the referent (<1500 MET-min/wk) (Table 3). Consistent results were also observed when restricted to the subgroup with a baseline CAC greater than 0 AU (eTable 3 in Supplement 1). Moreover, the baseline component of the mean CAC progression model found a significant cross-sectional association between PA (continuous and categorical) and CAC among men only (Table 3). For example, at baseline, very high PA levels (vs referent) was cross-sectionally associated with 43.3% (95% CI, 15.4%-71.3%) higher CAC levels in men. Figure 1 similarly shows that cross-sectionally, individuals with higher PA have higher predicted mean CAC at any age with statistical significance among men but not women.
Table 3. Adjusted Mean CAC at Baseline and CAC Progression During Follow-Up by Physical Activity in Men and Women: CCLS 1998-2019 (N = 8771)a.
| Independent variable | Dependent variable CAC (95% CI) | |||||||
|---|---|---|---|---|---|---|---|---|
| Men (n = 6661) | Women (n = 2110) | |||||||
| % Difference at baseline | P value | % Change over time | P value | % Difference at baseline | P value | % Change over time | P value | |
| Model 1 | ||||||||
| Baseline PA, per 500 MET-min/wk | 5.5 (2.7 to 8.4) | <.001 | NA | NA | 2.8 (−4.4 to 10.1) | .45 | NA | NA |
| Time per year | NA | NA | 28.5 (27.7 to 29.3) | <.001 | NA | NA | 32.1 (29.4 to 34.7) | <.001 |
| Time × mean follow-up PA, per year per 500 MET-min/wk | NA | NA | 0.0 (−0.1 to 0.1) | .94 | N/A | N/A | 0.0 (−0.4 to 0.5) | .93 |
| Model 2 | ||||||||
| Baseline PA, high vs referent | 23.7 (5.3 to 42.0) | .01 | NA | NA | 9.1 (−32.1 to 50.3) | .66 | NA | NA |
| Baseline PA, very high vs referent | 43.3 (15.4 to 71.3) | .002 | NA | NA | 34.7 (−32.2 to 101.5) | .31 | NA | NA |
| Time per year | NA | NA | 28.4 (27.6 to 29.2) | <.001 | NA | NA | 32.1 (29.4 to 34.8) | <.001 |
| Time × mean follow-up PA per year for high vs referent PA | NA | NA | 0.5 (−0.2 to 1.2) | .16 | NA | NA | −0.1 (−3.0 to 2.8) | .95 |
| Time × mean follow-up PA per year for very high vs referent PA | NA | NA | 0.3 (−0.9 to 1.5) | .62 | NA | NA | −1.5 (−6.4 to 3.5) | .56 |
Abbreviations: CAC, coronary artery calcification; CCLS, Cooper Center Longitudinal Study; MET-min/wk, metabolic equivalent of task minutes per week; NA, not applicable; PA, physical activity.
Adjusted for baseline age, as well as baseline and follow-up type of scanner, current smoking, body mass index, fasting glucose, total cholesterol, systolic blood pressure, and statin use. The PA referent was <1500 MET-min/wk; high PA, 1500-2999 MET-min/wk; and very high PA, ≥3000 MET-min/wk.
Figure 1. Predicted Mean Coronary Artery Calcification (CAC) vs Time Starting From Age 50, 60, or 70 Years by Physical Activity.
Adjusted for baseline and follow-up type of scanner, current smoking, body mass index, glucose, total cholesterol, systolic blood pressure, and statin use. At each starting age (50, 60, or 70 years), higher physical activity was associated with higher CAC, but parallel progression curves show equal relative rates of CAC progression (% per year) for up to 10 years. MET-min/wk indicates metabolic equivalent of task minutes per week.
Finally, the association between baseline PA and CAC progression to a clinically meaningful threshold (≥100 AU) model is presented in Figure 2. Analysis revealed that very high PA (vs referent) was not associated with CAC progression to 100 AU or more in men (hazard ratio [HR], 0.84; 95% CI, 0.66-1.08) or women (HR, 1.16; 95% CI, 0.57-2.35). Similarly, high PA (vs referent) was not associated with CAC progression to 100 AU or more in men (HR, 0.99; 95% CI, 0.85-1.15) or women (HR, 0.98; 95% CI, 0.63-1.51). Similar results were found in models of progression from a baseline CAC of 0 to a CAC greater than 0 AU at follow-up (eFigure in Supplement 1).
Figure 2. Baseline Physical Activity as a Predictor of Coronary Artery Calcification (CAC) Progression to 100 Agatston Units (AU) or Greater Among Men and Women With a CAC Less Than 100 AU at Baseline (n = 7391).
The referent was less than 1500 metabolic equivalent of task minutes per week (MET-min/wk); high physical activity, 1500 to 2999 MET-min/wk; and very high physical activity, 3000 or more MET-min/wk at baseline. This multivariable model was adjusted for age, CAC, type of CAC scanner, current smoking, body mass index, fasting glucose, total cholesterol, systolic blood pressure, and statin use, all at baseline.
Discussion
The present results indicate that high and very high volumes of leisure-time aerobic PA are unrelated to CAC progression over time. Prior evidence linking high-volume exercise, such as in marathoners and triathletes, to increased CAC primarily stems from cross-sectional designs,14,15 which prohibits the study of temporal effects. While randomized clinical trials are the gold standard, our observational study provides a higher level of evidence than cross-sectional research because of its ability to determine temporality among a large sample of men and women who had repeated measurements of PA and CAC over a relatively long follow-up. These results show that exercise, even when performed at a high volume, does not appear to increase the risk for progression of CAC over time. Furthermore, the present study results show no evidence that very high volumes of PA (≥3000 MET-min/wk), even among those with elevated CAC (≥100 AU), are associated with higher mortality among a subsample with mortality surveillance data. These findings are consistent with previous research from the CCLS and others,14,30 indicating that highly active individuals with elevated CAC do not exhibit greater mortality risk. Nonetheless, there is evidence that engaging in PA could lead to adverse cardiovascular adaptions as well as acute cardiac events (related to exercise) among some individuals.31
Whereas no association was found between PA and CAC progression in the longitudinal component of our models, our baseline model found a cross-sectional association between higher PA and higher CAC in men, which is consistent with previous research.15 For example, in a study of 152 masters athletes compared with 92 controls, the athletes were more likely to have a CAC score greater than 300 AU vs controls with a comparable risk status.16 Most recently, De Bosscher et al32 evaluated 558 men in a well-balanced observational cohort, 191 of whom were lifelong endurance athletes, 191 late-onset athletes, and 176 healthy nonathletes, and found that the athletes exercised a mean 10 to 11 hours per week vs 1 hour/week in the nonathletes. Both athlete groups had more individuals with CAC 100 AU or more than the nonathlete group (23% and 16.2% vs 14.8%, respectively).
Our longitudinal findings from the CAC progression models (n = 8771) among men and women are similar to a recent study by Aengevaeren et al33 among a sample of 289 men. They found that higher exercise volume was not significantly associated with CAC progression with varying results pertaining to intensity, which was not examined in the current study. A small study by Kleiven et al34 of highly active Norwegian men (n = 61) also observed no association between PA volume and the progression of CAC. Taken together, these longitudinal results suggest that high volumes of PA do not appear to affect CAC progression. In contrast, Sung et al35 found a significant association between higher baseline PA and increased progression of CAC over time. However, this unique outcome might stem from the fact that PA was assessed with an abbreviated survey that is designed for cross-sectional PA surveillance rather than the higher-resolution estimates used here.36,37
Interestingly, there appears to be a discordance between the cross-sectional and prospective associations between high-volume PA and CAC in some though not all studies. Possible explanations include selection effects in cross-sectional sampling or an omitted factor (eg, family history of coronary heart disease) that might motivate individuals to engage in higher PA volumes. Unfortunately, neither of these explanations can be confirmed in our data. In addition, it should be noted that when coronary computed tomographic angiography (CTA) was used to quantify subclinical atherosclerosis,15,32 the plaque burden and composition varied in terms of number and morphology (calcified, noncalcified, or mixed). While the Master@Heart study showed that CAC in athletes had less favorable attributes (more mixed and noncalcified plaque),32 the Measuring Athlete’s Risk of Cardiovascular Events study showed more benign plaque composition (calcified plaque).15 Our study does not include CTA, and therefore, we were not able to evaluate the association between PA and cross-sectional and prospective changes in calcified and noncalcified plaque burden.
Strengths and Limitations
The present study has both strengths and limitations. The strengths include longitudinal data and robust analytic approach, large sample size, inclusion of both sexes, and the quality of clinical characterization. This wealth of covariates including clinical data enabled adjustment in multivariable analysis. Furthermore, comprehensive self-reported PA information was collected from participants, which is consistent with other large epidemiological datasets,38 and its correlation with CRF was equivalent or higher than other questionnaires.37,39 Nonetheless, objective activity measurement (eg, accelerometers) would be able to minimize biases, such as social desirability.40 An additional limitation pertains to the study sample, which is predominantly White and well educated, thereby limiting the generalizability to racial and ethnic minority populations.41 However, the homogeneity of the sample reduces the potential influence of unmeasured confounders, thereby increasing internal validity,14which is similar to other large cardiovascular health cohorts (eg, the Nurses’ Health Study).42,43 While our follow-up period (approximately 8 years) is comparable with other longitudinal studies (eg, 6 years in the 2023 study by Aengevaeren et al33), a longer follow-up would be advantageous to investigate exercise and CAC progression over the lifespan. Furthermore, the interpretation of CAC progression is complex in this population with PA and statin use, which both could potentially modify the plaque burden. Notably, the increase of statin use during follow-up was similar across PA groups and hence not likely a confounder. Finally, limited mortality events over the follow-up duration in this sample prohibited a comprehensive examination of this outcome in the current study.
Conclusions
This study suggests that high-volume exercise is not associated with the progression of subclinical atherosclerosis as estimated from CAC among men and women who were initially free of overt cardiovascular disease. This conclusion is based on our results showing that high and very high volumes of PA are not related to the progression of CAC (irrespective of baseline CAC levels) among this sample of community-dwelling adults. Thus, even high-volume PA likely does not accelerate CAC progression and atherosclerosis burden, though additional longitudinal research on more diverse samples is needed.
eMethods. Details on the Statistical Modeling Approach for Mean CAC Progression
eTable 1. Morality from all-causes and cardiovascular disease by baseline CAC and PA among men
eTable 2. Morality from all-causes and cardiovascular disease by baseline CAC and PA among women
eTable 3. Adjusted mean CAC at baseline and CAC progression during follow-up by physical activity in men and women
eFigure. Hazard ratios and 95% confidence intervals for baseline physical activity as a predictor of CAC progression to >0 AU among men and women with CAC=0 AU at baseline
Data sharing statement
References
- 1.Blond K, Brinkløv CF, Ried-Larsen M, Crippa A, Grøntved A. Association of high amounts of physical activity with mortality risk: a systematic review and meta-analysis. Br J Sports Med. 2020;54(20):1195-1201. doi: 10.1136/bjsports-2018-100393 [DOI] [PubMed] [Google Scholar]
- 2.Kraus WE, Powell KE, Haskell WL, et al. ; 2018 Physical Activity Guidelines Advisory Committee . Physical activity, all-cause and cardiovascular mortality, and cardiovascular disease. Med Sci Sports Exerc. 2019;51(6):1270-1281. doi: 10.1249/MSS.0000000000001939 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Li T, Wei S, Shi Y, et al. The dose-response effect of physical activity on cancer mortality: findings from 71 prospective cohort studies. Br J Sports Med. 2016;50(6):339-345. doi: 10.1136/bjsports-2015-094927 [DOI] [PubMed] [Google Scholar]
- 4.Lee IM, Shiroma EJ, Lobelo F, Puska P, Blair SN, Katzmarzyk PT; Lancet Physical Activity Series Working Group . Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet. 2012;380(9838):219-229. doi: 10.1016/S0140-6736(12)61031-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.US Department of Health and Human Services . Physical Activity Guidelines for Americans. 2nd ed. 2018. Accessed March 28, 2024. https://health.gov/sites/default/files/2019-09/Physical_Activity_Guidelines_2nd_edition.pdf
- 6.Mok A, Khaw KT, Luben R, Wareham N, Brage S. Physical activity trajectories and mortality: population based cohort study. BMJ. 2019;365:l2323. doi: 10.1136/bmj.l2323 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Piercy KL, Troiano RP, Ballard RM, et al. The physical activity guidelines for Americans. JAMA. 2018;320(19):2020-2028. doi: 10.1001/jama.2018.14854 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Paluch AE, Bajpai S, Ballin M, et al. ; Steps for Health Collaborative . Prospective association of daily steps with cardiovascular disease: a harmonized meta-analysis. Circulation. 2023;147(2):122-131. doi: 10.1161/CIRCULATIONAHA.122.061288 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Shuval K, Leonard T, Drope J, et al. Physical activity counseling in primary care: insights from public health and behavioral economics. CA Cancer J Clin. 2017;67(3):233-244. doi: 10.3322/caac.21394 [DOI] [PubMed] [Google Scholar]
- 10.Bélair MA, Kohen DE, Kingsbury M, Colman I. Relationship between leisure time physical activity, sedentary behaviour and symptoms of depression and anxiety: evidence from a population-based sample of Canadian adolescents. BMJ Open. 2018;8(10):e021119. doi: 10.1136/bmjopen-2017-021119 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Powell KE, Paluch AE, Blair SN. Physical activity for health: what kind? how much? how intense? on top of what? Annu Rev Public Health. 2011;32(1):349-365. doi: 10.1146/annurev-publhealth-031210-101151 [DOI] [PubMed] [Google Scholar]
- 12.Baggish AL, Levine BD. Coronary artery calcification among endurance athletes: “hearts of stone”. Circulation. 2017;136(2):149-151. doi: 10.1161/CIRCULATIONAHA.117.028750 [DOI] [PubMed] [Google Scholar]
- 13.Gaasch WH, Zile MR. Left ventricular structural remodeling in health and disease: with special emphasis on volume, mass, and geometry. J Am Coll Cardiol. 2011;58(17):1733-1740. doi: 10.1016/j.jacc.2011.07.022 [DOI] [PubMed] [Google Scholar]
- 14.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 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Aengevaeren VL, Mosterd A, Braber TL, et al. Relationship between lifelong exercise volume and coronary atherosclerosis in athletes. Circulation. 2017;136(2):138-148. doi: 10.1161/CIRCULATIONAHA.117.027834 [DOI] [PubMed] [Google Scholar]
- 16.Merghani A, Maestrini V, Rosmini S, et al. Prevalence of subclinical coronary artery disease in masters endurance athletes with a low atherosclerotic risk profile. Circulation. 2017;136(2):126-137. doi: 10.1161/CIRCULATIONAHA.116.026964 [DOI] [PubMed] [Google Scholar]
- 17.Shuval K, Barlow CE, Chartier KG, Gabriel KP. Cardiorespiratory fitness, alcohol, and mortality in men: the Cooper Center Longitudinal Study. Am J Prev Med. 2012;42(5):460-467. doi: 10.1016/j.amepre.2012.01.012 [DOI] [PubMed] [Google Scholar]
- 18.Shuval K, Leonard D, DeFina LF, et al. Cardiorespiratory fitness and depression symptoms among adults during the COVID-19 Pandemic: Cooper Center Longitudinal Study. Prev Med Rep. 2022;30:102065. doi: 10.1016/j.pmedr.2022.102065 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Shuval K, Chiu CY, Barlow CE, et al. Family history of chronic disease and meeting public health guidelines for physical activity: the Cooper Center Longitudinal Study. Mayo Clin Proc. 2013;88(6):588-592. doi: 10.1016/j.mayocp.2013.04.006 [DOI] [PubMed] [Google Scholar]
- 20.Ainsworth BE, Haskell WL, Herrmann SD, et al. 2011 Compendium of physical activities: a second update of codes and MET values. Med Sci Sports Exerc. 2011;43(8):1575-1581. doi: 10.1249/MSS.0b013e31821ece12 [DOI] [PubMed] [Google Scholar]
- 21.Willis BL, Morrow JR Jr, Jackson AW, Defina LF, Cooper KH. Secular change in cardiorespiratory fitness of men: Cooper Center Longitudinal Study. Med Sci Sports Exerc. 2011;43(11):2134-2139. doi: 10.1249/MSS.0b013e31821c00a7 [DOI] [PubMed] [Google Scholar]
- 22.Mao SS, Pal RS, McKay CR, et al. Comparison of coronary artery calcium scores between electron beam computed tomography and 64-multidetector computed tomographic scanner. J Comput Assist Tomogr. 2009;33(2):175-178. doi: 10.1097/RCT.0b013e31817579ee [DOI] [PubMed] [Google Scholar]
- 23.Ulzheimer S, Kalender WA. Assessment of calcium scoring performance in cardiac computed tomography. Eur Radiol. 2003;13(3):484-497. doi: 10.1007/s00330-002-1746-y [DOI] [PubMed] [Google Scholar]
- 24.Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M Jr, Detrano R. Quantification of coronary artery calcium using ultrafast computed tomography. J Am Coll Cardiol. 1990;15(4):827-832. doi: 10.1016/0735-1097(90)90282-T [DOI] [PubMed] [Google Scholar]
- 25.Cheng YJ, Church TS, Kimball TE, et al. Comparison of coronary artery calcium detected by electron beam tomography in patients with to those without symptomatic coronary heart disease. Am J Cardiol. 2003;92(5):498-503. doi: 10.1016/S0002-9149(03)00714-8 [DOI] [PubMed] [Google Scholar]
- 26.Garrow JS, Webster J. Quetelet’s index (W/H2) as a measure of fatness. Int J Obes. 1985;9(2):147-153. [PubMed] [Google Scholar]
- 27.Fitzmaurice GM, Laird NM, Ware JH. Applied Longitudinal Analysis. John Wiley & Sons; 2012. [Google Scholar]
- 28.Paixao AR, Chakravorty R, Khera A, et al. Disagreement between different definitions of coronary artery calcium progression. JACC Cardiovasc Imaging. 2015;8(6):743-744. doi: 10.1016/j.jcmg.2014.07.019 [DOI] [PubMed] [Google Scholar]
- 29.Royston P, Parmar MK. Flexible parametric proportional-hazards and proportional-odds models for censored survival data, with application to prognostic modelling and estimation of treatment effects. Stat Med. 2002;21(15):2175-2197. doi: 10.1002/sim.1203 [DOI] [PubMed] [Google Scholar]
- 30.German CA, Fanning J, Singleton MJ, et al. Physical activity, coronary artery calcium, and cardiovascular outcomes in the Multi-Ethnic Study of Atherosclerosis (MESA). Med Sci Sports Exerc. 2022;54(5):800-806. doi: 10.1249/MSS.0000000000002856 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Franklin BA, Thompson PD, Al-Zaiti SS, et al. ; American Heart Association Physical Activity Committee of the Council on Lifestyle and Cardiometabolic Health; Council on Cardiovascular and Stroke Nursing; Council on Clinical Cardiology; and Stroke Council . Exercise-related acute cardiovascular events and potential deleterious adaptations following long-term exercise training: placing the risks into perspective–an update: a scientific statement from the American Heart Association. Circulation. 2020;141(13):e705-e736. doi: 10.1161/CIR.0000000000000749 [DOI] [PubMed] [Google Scholar]
- 32.De Bosscher R, Dausin C, Claus P, et al. Lifelong endurance exercise and its relation with coronary atherosclerosis. Eur Heart J. 2023;44(26):2388-2399. doi: 10.1093/eurheartj/ehad152 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Aengevaeren VL, Mosterd A, Bakker EA, et al. Exercise volume versus intensity and the progression of coronary atherosclerosis in middle-aged and older athletes: findings from the MARC-2 Study. Circulation. 2023;147(13):993-1003. doi: 10.1161/CIRCULATIONAHA.122.061173 [DOI] [PubMed] [Google Scholar]
- 34.Kleiven Ø, Bjørkavoll-Bergseth MF, Omland T, et al. Endurance exercise training volume is not associated with progression of coronary artery calcification. Scand J Med Sci Sports. 2020;30(6):1024-1032. doi: 10.1111/sms.13643 [DOI] [PubMed] [Google Scholar]
- 35.Sung KC, Hong YS, Lee JY, et al. Physical activity and the progression of coronary artery calcification. Heart. 2021;107(21):1710-1716. doi: 10.1136/heartjnl-2021-319346 [DOI] [PubMed] [Google Scholar]
- 36.Craig CL, Marshall AL, Sjöström M, et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc. 2003;35(8):1381-1395. doi: 10.1249/01.MSS.0000078924.61453.FB [DOI] [PubMed] [Google Scholar]
- 37.Lee PH, Macfarlane DJ, Lam TH, Stewart SM. Validity of the international physical activity questionnaire short form (IPAQ-SF): a systematic review. Int J Behav Nutr Phys Act. 2011;8(1):115. doi: 10.1186/1479-5868-8-115 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Sylvia LG, Bernstein EE, Hubbard JL, Keating L, Anderson EJ. Practical guide to measuring physical activity. J Acad Nutr Diet. 2014;114(2):199-208. doi: 10.1016/j.jand.2013.09.018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Mäder U, Martin BW, Schutz Y, Marti B. Validity of four short physical activity questionnaires in middle-aged persons. Med Sci Sports Exerc. 2006;38(7):1255-1266. doi: 10.1249/01.mss.0000227310.18902.28 [DOI] [PubMed] [Google Scholar]
- 40.Brenner PS, DeLamater JD. Social desirability bias in self-reports of physical activity: is an exercise identity the culprit? Soc Indic Res. 2014;117(2):489-504. doi: 10.1007/s11205-013-0359-y [DOI] [Google Scholar]
- 41.Gafni T, Shuval K, Weinstein G, et al. Sitting time, physical activity, and cognitive impairment in midlife and older adults. J Aging Phys Act. 2022;30(3):355-363. doi: 10.1123/japa.2020-0473 [DOI] [PubMed] [Google Scholar]
- 42.Radford NB, DeFina LF, Barlow CE, et al. Progression of CAC score and risk of incident CVD. JACC Cardiovasc Imaging. 2016;9(12):1420-1429. doi: 10.1016/j.jcmg.2016.03.010 [DOI] [PubMed] [Google Scholar]
- 43.Rockhill B, Willett WC, Manson JE, et al. Physical activity and mortality: a prospective study among women. Am J Public Health. 2001;91(4):578-583. doi: 10.2105/AJPH.91.4.578 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods. Details on the Statistical Modeling Approach for Mean CAC Progression
eTable 1. Morality from all-causes and cardiovascular disease by baseline CAC and PA among men
eTable 2. Morality from all-causes and cardiovascular disease by baseline CAC and PA among women
eTable 3. Adjusted mean CAC at baseline and CAC progression during follow-up by physical activity in men and women
eFigure. Hazard ratios and 95% confidence intervals for baseline physical activity as a predictor of CAC progression to >0 AU among men and women with CAC=0 AU at baseline
Data sharing statement