The route-(c) harm facet of the activity cluster: the transient cardiac risk of a bout of exertion,
held apart from the long-run benefit on Physical Activity Dose and Mortality. Two consensus
statements with overlapping authors, neither with a stated search method. The main one is an
AHA scientific statement (consensus design, endorsed by ACSM and AACVPR) that names no search
method — srcgrep for literature search / search strategy / PubMed / MEDLINE / systematic search
returns 0 across chunks 01-03. Every number below is a statement-grade secondary citation; the
underlying designs are named per row (case-crossover, case-control, registry, one 7-study MA).
(Franklin et al., 2020) The ACSM 2015
preparticipation consensus statement (Riebe et al.), also with no stated search method, supplies the screening
algorithm plus some relative- and absolute-risk figures.
The two documents share authors (Franklin, Thompson) and cite the same triggering primaries, so where they
agree that is one body of expert opinion, not two independent routes.
(RIEBE et al., 2015)
Bottom line
- Relative risk during a bout is large for vigorous exertion in the inactive (estimates for moderate, or pooled moderate-to-vigorous, exertion run 1.6-2.4); absolute risk is very low. The statement’s own summary: «When expressed as an absolute risk, acute events are very rare, with only 0.31 to 2.1 SCDs per 100 000 person-years.» (Franklin et al., 2020) That denominator includes non-exercisers; per bout, the Physicians’ Health Study figure is 1 SCD per 1.51 million episodes of vigorous exertion. (Franklin et al., 2020)
- The rise in relative risk is largest in one stratum: the habitually sedentary with known or occult heart disease doing unaccustomed strenuous work — «This is particularly true among habitually seden- tary individuals with known or occult CVD performing unaccustomed strenuous PA.» (Franklin et al., 2020)
- Regular activity shrinks the transient excess (the gradient below). So the harm argues for graded entry, not for staying inactive — the statement’s net: «Despite such caveats, the benefits associated with long-term exercise training outweigh the risks for the majority of the population.» The caveats it refers to are cardiac patients (where high-intensity, high-volume activity may raise mortality) and ARVC (possibly also lamin A/C mutations); the same paragraph rejects a U-shape for the general population («Longevity data argue against the U-shaped relationship.»). The long-run upper-arm question is handled on The U-Shaped Association Artifact. (Franklin et al., 2020)
- Who needs medical clearance (ACSM 2015, consensus). Inactive, with no known cardiovascular, renal disease or diabetes and no symptoms: start light-to-moderate aerobic exercise without clearance and progress gradually. Inactive with known disease, or anyone with symptoms: clearance first (symptoms during exercise mean stop, then seek clearance). Already exercising regularly with known disease and no symptoms: moderate needs no new clearance; vigorous needs clearance (one from the last 12 months suffices if nothing has changed). Basis: consensus; the panel reports a «lack of evidence that medical clearance and exercise testing (28) are effective». Decision table and caveats in the screening section below. (RIEBE et al., 2015)
Relative risk during and just after a bout
| Outcome | Population (design) | RR during/after exertion | Habitual-activity gradient |
|---|---|---|---|
| SCD | PHS men (prospective case-crossover) | ≈17-fold | see note: least active 50-74 vs regular 2-11 across PHS + Slovenia |
| SCD | Slovenia men + women (retrospective case-crossover) | 3-fold | see note |
| SCD | NHS women (prospective case-crossover) | 2.4 (1.2-4.6) | <2 h/wk 9.0 (3.3-24.3); ≥2 h/wk 1.5 (0.61-3.61) |
| MI | 7-study MA, n=5503 | 3.45 (2.33-5.13), heterogeneous | least active 4.5-107; most active 0.88-3.3 (Riebe prints 0.86-3.3, see below) |
| MI | INTERHEART, 12,461 first MIs, 52 countries (case-crossover) | 2.3 (2.0-2.7), printed as 95% CI; PAR 7.7% (see Limits for the OR 2.31, 99% CI variant) | none by baseline PA category |
| MI | Augsburg, moderate exertion ≈5 METs, within 2 h (case-crossover) | 1.6 (1.2-2.1) | — (an earlier study from the same region found no elevation with moderate exertion) |
(Franklin et al., 2020) The SCD habitual-activity
gradient is a range across two studies, not a PHS-only figure: it reads «RR=5068 to 7448» / «RR=268 to
1148» in the PDF text — reference superscripts fused onto the values, decoding to 50 (ref 68, Slovenia)
to 74 (ref 48, PHS) and 2 (ref 68) to 11 (ref 48) via the reference list (chunk 02). The MA’s frequency gradient: «In a meta-regression
analysis, the RR of MI associated with episodic PA was decreased by ≈45% for each additional time per
week a person habitually exercised.» In 2 studies the SCD excess stayed elevated even in the most
active men, so regular activity attenuates the excess without abolishing it.
(Franklin et al., 2020) Riebe’s summary of
the same MA (Dahabreh & Paulus 2011) adds the SCD arm: «Furthermore, these authors found that for every
additional exercise bout per week, there was 30% re- duction in the risk of SCD and 45% reduction in the
risk of AMI during physical activity (12).» It gives the most-active AMI range as 0.86-3.3 where Franklin
gives 0.88-3.3 for the same MA. Both describe the same quantity (the RR of exertion-related AMI in the
most habitually active groups of Dahabreh & Paulus 2011): Riebe chunk 01 prints «only 0.86–3.3»,
Franklin chunk 01 prints «(0.88–3.3)». The lower bound differs by 0.02 and the MA is not held, so the
discrepancy is recorded, not resolved; it does not change the reading (a small or absent excess in the
most active). Riebe
also gives the MA’s SCD estimate, which the table lacks: «A recent meta-analysis reported a fivefold
increased risk of SCD and 3.5-fold increased risk of AMI during vigorous- intensity physical activity (Q6
METs) (12).» (Q = ≥; no CI given.) And it sizes the residual excess in the most active: in the Onset
Study the AMI risk during vigorous work «was almost twice that at rest even for individuals who exercised
vigorously for 1 h on 5 or more days per week».
(RIEBE et al., 2015)
- Intensity is relative, not absolute. The ≥6 MET cut for vigorous activity misses the person: «the cardiac demand of any PA is determined not by the specific metabolic level but by the metabolic demand relative to the individual’s functional capacity». A load that is moderate for a fit adult can be vigorous (≥60% functional capacity) for an unfit older one. Light exertion was not associated with MI. (Franklin et al., 2020)
- INTERHEART’s null on effect modification is weaker than it looks: it stratified baseline PA by intensity category, not session frequency, the axis the MA gradient runs on (the statement’s own caveat). (Franklin et al., 2020)
Absolute risk — the denominators are not interchangeable
| Setting | Rate | Unit |
|---|---|---|
| PHS men, vigorous bouts | 1 SCD per 1.51 million | episodes |
| Middle-aged women | 1 SCD per 36.5 million | person-hours |
| Fitness facilities | 1 per 1,124,200 nonfatal; 1 per 887,526 fatal | person-hours |
| Supervised cardiac rehab, established CVD | 1 per 58,000 | patient-hours |
| Population exercise-related SCD | 0.31-2.1 | per 100,000 person-years |
| Oregon 35-65 y: sports SCA vs all other SCA | 2.1 vs 55.5 | per 100,000 person-years |
| Marathon/half (RACER, 10.9 M runners) | 0.39, labelled SCA but ≈ the death rate (see note); men 0.90, women 0.16 as given, arrest-vs-death unclear | per 100,000 participants |
| Triathlon (>9 M participants >30 y) | 1.74 | per 100,000 participants |
| Rhode Island joggers (Riebe) | 1 death per 396,000 | hours of jogging |
| YMCA sports centres (Riebe; intensity undocumented) | 1 death per 2,897,057 | person-hours |
| Marathon/half (RACER, via Riebe) | 0.20 cardiac arrests, 0.14 SCD | per 100,000 runner-hours |
(Franklin et al., 2020) The last three rows are from Riebe, which reports RACER as 1 arrest per 184,000 and 1 SCD per 256,000 runners and converts them to runner-hours. (RIEBE et al., 2015) The RACER row above carries Franklin’s label: «an SCA incidence rate of 0.39 per 100 000 participants», from «59 cases of SCA» of which «42 (71%) were fatal» among 10.9 million. 59/10.9 M is 0.54 per 100,000 (1 per 184,000) and 42/10.9 M is 0.39 (1 per ~259,000), so 0.39 is arithmetically the death rate, the same figure as Riebe’s 1 SCD per 256,000; the all-arrest rate is about 0.54. (Franklin et al., 2020) (inferred from Franklin et al., 2020; RIEBE et al., 2015) On the cardiac-rehab row the source adds «but still low». In Oregon only 63 of 1247 SCAs (5%) were sports-related. Exercise-related OHCA survival is higher than non-exercise (46.2% vs 17.2%, NL).
The rows use four incompatible denominators (episodes, person-hours, person-years, race participants) and cannot be ranked against each other without exposure-time data the statement does not give. What is comparable within one population is the Oregon row: across adults 35-65, sports-related SCA is a small slice of all sudden cardiac arrest.
Who carries the risk — strata
- Route (a), baseline risk — age. Absolute incidence rises over 35, and the greatest numbers of exertion-related SCDs fall between 40 and 64 (a count, not a peak rate); NL >35 y 3.0 vs <35 y 0.3 per 100k p-y; French men 0.4 (15-24 y) to 17.5 (55-64 y) per 100k per year, with no such rise in French women. Whether age modifies the relative risk is unsettled: a German study found a higher RR of MI in older people, the larger INTERHEART did not. (Franklin et al., 2020) The cause shifts with age: atherosclerotic CAD in middle-aged and older adults. (Franklin et al., 2020)
- Route (a) — sex. Exercise-associated SCD is 15- to 20-fold higher in men (women 0.04-0.3 vs men 0.5-5.8 per 100k p-y); at 45-55 y women’s risk is 3% (1.5%-7.5%) of men’s. Most studies find no sex difference in the RR of MI. Women are under-represented in the triggering studies (the NHS cohort had to pool moderate and vigorous exertion for lack of events). (Franklin et al., 2020)
- Route (b), effect modification on positive evidence — habitual activity. The MA meta-regression and the PHS / NHS within-study gradients all point the same way. This is observational and partly between-study, but it is a measured interaction, not a mechanistic guess. (Franklin et al., 2020) (inferred from Franklin et al., 2020) Not fully independent: the PHS (ref 48), NHS (ref 45) and MIOS (ref 41) papers share authors (Mittleman on refs 41 and 48; Albert, Manson and Chae on refs 45 and 48), and Christine Albert, senior or lead author on the PHS and NHS papers, co-wrote the statement (reference list, chunk 02). The Slovenia study is the independent arm of the SCD gradient. The statement faults INTERHEART for stratifying by intensity category rather than frequency; it raises no matching limits on the gradient studies. The wiki’s own caveats for them: recalled habitual frequency, and the chance that the inactive are also the less fit or carry occult disease.
- Route (c) — known heart disease, and the paradox. HF or post-MI with EF <40% carries the greatest risk, and supervised rehab runs at 1 event per 58,000 patient-hours. Yet most exertional events strike people without diagnosed disease: in Oregon 15-16% of sports SCAs (35-65 y; the statement gives both figures) had known heart disease vs 30% of other SCAs, and most exertional AMIs occur without prior CHD. Some but not all studies even report a higher RR of exertional AMI in people without established angina or CHD. (Franklin et al., 2020) (inferred from Franklin et al., 2020) So screening keyed to a known diagnosis cannot reach most of the people who have an exertional event.
- Competitive vs recreational. Competitive athletes ≈4.5x recreational (10-35 y), but the larger recreational pool means most young exertional SCDs occur in noncompetitive participants. (Franklin et al., 2020)
Activities and settings with higher event rates
- The statement lists activities that «may be associated» with higher event rates: racquet sports, downhill skiing, marathon and triathlon, basketball; mountaineering for certain individuals (predominantly men); snow shoveling (probably because it can produce rate-pressure products above a maximal treadmill test) and deer hunting. Mountain hiking runs at 1 SCD per 2,940,000 mountain-hours, about the same as jogging, but 1 cardiac event of any kind (including ACS) per 957,000 mountain-hours; the risk is cut >5-fold by ≥1 night acclimatizing above 1300 m. (Franklin et al., 2020)
- Endurance races: nearly half of marathon SCDs occurred in the final mile, and the male rate rose from 0.71 to 2.03 per 100k over 2000-2010; two-thirds of triathlon deaths happened during the swim and 38% of those with known experience were first-timers. The statement advises competitors at risk for cardiac events to hold pace through the race rather than sprint the final miles. (Franklin et al., 2020)
What lowers it — statement-grade practice, not trial-tested
(Franklin et al., 2020) The prudent-program advice, quoted as given: «Previously inac- tive patients with or without known CVD should also be advised to avoid unaccustomed, vigorous physical exertion and highly strenuous activities (eg, racquet sports,95 snow removal106), to recognize potential exertion-related warning signs and symptoms (eg, chest pain or pressure, lightheadedness, palpitations or arrhythmias) that require cessation of exercise and medical evaluation, and to adapt the exercise to the environment.» And for the healthy novice: «Ostensibly healthy, inactive individuals starting to exercise should begin slowly, generally with a walking program, and increase the intensity and duration of exercise as their tolerance permits.»
Cheap, reversible, low-harm, and aimed at the exact stratum where the excess sits, so per the firmness rule it gets a plain do-this, with the basis labelled (consensus practice; the triggering evidence is case-crossover, and no trial randomized progression schemes against event outcomes):
- If you are inactive, start with walking and build intensity gradually, not with a vigorous session. The basis is a between-person gradient (regular exercisers carry a smaller per-bout excess), not a measured time course: the statement gives no figure for how fast the excess falls as a habit forms, and part of the gradient may be that the least fit and those with occult disease are the inactive ones. The statement’s own rule: «novice exercisers should follow a gradual increase in exercise intensity to increase CRF without greatly increasing their cardiovascular risk by starting with vigorous exercise». (Franklin et al., 2020)
- With known heart disease, the default is supervised or home-based cardiac rehab, then walking pushed up to moderate intensity (40-59% of functional capacity, 3.0-5.9 METs), «provided they remain asymptomatic». The statement’s worry is under-dosing: patients start rehab at about 2-3 METs and up-titration is often too slow, especially once monitored supervision ends, and moderate intensity sits below the ≥6 MET vigorous level linked to triggering. (Franklin et al., 2020)
- With an inherited, structural or arrhythmic heart condition (eg Marfan syndrome, inherited cardiomyopathies), vigorous work needs specialist guidance; the statement discourages highly competitive or strenuous endurance exercise without careful consultation and encourages leisure-time low- to moderate-intensity activity. (Franklin et al., 2020)
- Warm up and cool down; scale intensity down in heat, humidity, or above ~1500 m until acclimated (the statement’s >1500 m line in chunk 02 and its 1300 m acclimatization figure in chunk 01 both cite the same altitude study, ref 101).
- Stop and get evaluated for chest/neck/jaw/arm pain or pressure, unusual breathlessness, lightheadedness, palpitations or an irregular heartbeat (Table 2 also lists ankle swelling and burning or cramping in the legs on short walks). Warning signs preceded 29% of young SCAs in Portland. (Franklin et al., 2020)
- Avoid unaccustomed strenuous one-offs — a first shovel of heavy snow, an untrained race sprint.
- Survival after an arrest depends on the people nearby. In the marathon/half-marathon registry, «The strongest predictors of survival of SCA were early initiation of bystander-administered cardiopulmonary resuscitation, use of automated defibrillators, and an underlying di- agnosis other than HCM.» And: «Regular emergency drills, bystander cardiopulmonary resuscitation, and the use of automated external defibrillators have increased survival rates in the general157 and athletic158 popula- tion.» For a gym, club or race organizer that means a defibrillator on site and staff drilled to use it. (Franklin et al., 2020)
The 2011 ACSM stand’s grades on the same precautions (same lineage, not independent). The ACSM quantity-and-quality position stand graded these items with NHLBI categories (B = RCTs, limited data; C = nonrandomized or observational; D = panel consensus): exercise is associated with an increased risk of adverse CHD events (B); warm-up, cool-down, flexibility exercise and gradual progression «may reduce the risk of CVD events and musculoskeletal injury during exercise» (C); consultation with a physician and diagnostic exercise testing «may reduce risks of exercise if medically indicated, but are not recommended on a routine basis» (C); a fitness professional for novices or people with chronic disease (D); screening for and education about warning signs (C). (Garber et al., 2011) Its one datum beyond the AHA material: the Behavior Change Consortium’s 11 activity interventions in more than 5500 middle-aged and older people used minimal or more extensive screening, «but no study reported a serious CHD event, suggesting that the former can be highly effective. Few data support the role of routine diagnostic exercise testing as an effective method for reducing the risk of exercise-related CHD events». (Garber et al., 2011) Zero events across studies that each used one screening approach (a between-study split, not randomised arms; the stand reports no within-study comparison of screening levels and no unscreened comparator) bounds the event rate low, but it cannot show that minimal screening matches extensive screening, or that screening beats none: with no events there is nothing to compare. The stand’s CVD-risk paragraph cites the AHA 2007 statement (ref 362: Thompson, Franklin et al.), which the AHA 2020 statement updates (its title ends «Placing the Risks Into Perspective–An Update»); Franklin, Barry A. co-authors the 2011 stand, the AHA 2020 statement and the ACSM 2015 algorithm, and Garber, Carol Ewing co-authors the 2011 stand and the 2015 algorithm, so agreement across the three is shared lineage. (Franklin et al., 2020) (title)
Screening before starting
- The older AHA/ACSM questionnaire would have sent 95.5% of women and 93.5% of men aged ≥40 (NHANES) to a physician, itself a barrier to starting. The statement endorses the ACSM 2015 algorithm, which drops age and risk-factor counting for four variables: current activity, known cardiovascular/metabolic/renal disease, symptoms, and desired intensity. Inactive people without disease or symptoms may start light-to-moderate exercise without medical clearance and, provided they remain asymptomatic, progress gradually in intensity. (Franklin et al., 2020) The algorithm’s primary text, decision table and stated rationale are in the subsection below. Franklin and Thompson co-authored both documents, so agreement between them is shared authorship, not independent backing.
- USPSTF 2018 recommends against resting or exercise ECG screening in low-risk adults; the companion USPSTF review of exercise-ECG screening RCTs found no improvement in health outcomes despite focusing on higher-risk populations with diabetes. The statement says exercise testing «should be considered» in high-risk middle-aged or older people, particularly inactive men with diabetes planning high-intensity endurance sport, and lists others who «might benefit»: sedentary people with multiple risk factors, elevated CAC or premature family history who plan vigorous exercise, those suspected of ignoring symptoms, and patients with critical aortic stenosis, atypical chest pain or palpitations. (Franklin et al., 2020)
- Young athletes: «There is insufficient evidence to justify routine cardiovas- cular screening to prevent sports-related acute cardiac events in young athletes.» The next sentence notes the practice is nonetheless common (history and physical, with or without ECG); AHA/ACC and ESC differ on ECG. (Franklin et al., 2020)
The ACSM 2015 preparticipation algorithm — primary text
What it is. An ACSM roundtable consensus statement. Its stated method: «After a review of the
scientific literature and extensive discussion, members of the roundtable proposed an evidence-informed
model». No search method and no evidence grading: srcgrep for literature search / search strategy /
PubMed / MEDLINE / GRADE / level of evidence / randomized / validat / sensitivity returns 0
across the whole paper (chunk 01). Its purpose is to flag people at raised risk of exercise-related SCD or
AMI, those with disease who may benefit from a medically supervised program, and those who should not
exercise until a condition is controlled. (RIEBE et al., 2015)
The decision table (Figure 2, an image read directly from p. 5; scoped to aerobic exercise). Entry
question: does the person do regular exercise, defined as «planned, structured physical activity at least
30 min at moderate intensity on at least 3 dIwkj1 for at least the last 3 months» (dIwkj1 = d/wk in the
PDF text). (RIEBE et al., 2015)
| Regular exercise? | Disease (CV, metabolic, renal) / symptoms | Medical clearance | Then |
|---|---|---|---|
| No | no disease, no symptoms | not necessary | light-to-moderate; may progress gradually to vigorous |
| No | known disease, asymptomatic | recommended | after clearance, light-to-moderate; progress as tolerated |
| No | any symptoms, regardless of disease | recommended | after clearance, light-to-moderate; progress as tolerated |
| Yes | no disease, no symptoms | not necessary | continue moderate or vigorous; progress gradually |
| Yes | known disease, asymptomatic | not for moderate; recommended before vigorous (within the last 12 months if no change in signs/symptoms) | continue moderate; after clearance, may progress gradually |
| Yes | any symptoms | stop and seek clearance | return after clearance |
The text states the first row as: «Physically inactive but otherwise healthy asymptom- atic persons may begin light- to moderate-intensity exercise without medical clearance and, in the absence of symptoms, progress gradually in intensity as recommended by current ACSM exercise prescription guidelines (2,19).» CVD = cardiac, peripheral vascular or cerebrovascular disease; metabolic = type 1 or 2 diabetes; symptoms = ischemic-type pain or discomfort (chest, neck, jaw, arms), breathlessness at rest or on mild exertion, dizziness or syncope, orthopnea or paroxysmal nocturnal dyspnea, ankle edema, palpitations, intermittent claudication, a known heart murmur, unusual fatigue with usual activities. Medical clearance = «approval from a health care professional to engage in exercise»; the clinician, not the algorithm, decides what evaluation (if any) that involves. (RIEBE et al., 2015)
What changed, and the reasons given. The old ACSM process counted CVD risk factors: asymptomatic with ≥2 risk factors meant a medical exam before vigorous work, and symptoms or known cardiovascular, pulmonary, metabolic or renal disease meant an exam plus a diagnostic exercise test before even moderate work. The 2015 process drops risk-factor counting, the low/moderate/high risk classes, and automatic referral for pulmonary disease. (RIEBE et al., 2015)
- Risk factors probably cannot pick out the rare event (argued, not measured). «First, the high prevalence of CVD risk factors among adults, combined with the extreme rarity of exercise-related SCD and AMI, suggests that the ability to predict these rare events by assessing CVD risk factors is low (42).» The argument is base-rate arithmetic (65 million US adults with hypertension and 71 million with high LDL vs a small fraction of 600,000 annual heart-disease deaths); ref 42 is a 1979 series of 18 jogging deaths whose first author, Thompson, P.D., co-wrote this paper. No sensitivity, specificity or predictive value is reported. (RIEBE et al., 2015)
- Over-referral as a barrier. «A recent study found that 95% of men and women over 40 yr of age would be advised to consult a physician before exer- cise based upon the previous risk factor-based exercise pre- participation health screening process (48).» Ref 48 (Whitfield, NHANES) is by a co-author of this paper and is the same study as Franklin’s 95.5%/93.5% figure above. Unneeded referral «may lead to a high rate of false-positive exercise test responses in some populations», with follow-up studies that were not needed. (RIEBE et al., 2015)
- Age alone is not a referral trigger. «Although increasing age is a risk factor for CVD (men, Q45 yr;
women, Q55 yr), there is no evidence that age per se is a strong predictor of exercise-associated SCD and
AMI (20), and the referral of individuals for medical clearance solely on the basis of age has been shown
to result in superfluous health care referrals (6,19,46).» (
Q= ≥ in the PDF text.) (RIEBE et al., 2015) (Franklin et al., 2020; inferred from RIEBE et al., 2015) Not joined with the age bullet under Who carries the risk: Riebe speaks of age per se (net of the disease it brings) as a predictor of exertional events, and of age-alone referral as superfluous; Franklin’s figures are crude absolute incidence, which rises steeply with age in French men (0.4 to 17.5 per 100k per year from 15-24 to 55-64 y), mostly through atherosclerotic CAD. Dropping age as a referral trigger does not make age irrelevant: for an older inactive starter, graded entry and symptom awareness matter more, not less. - Diabetes and renal disease stay in because some evidence links them to exercise-related events; pulmonary disease leaves because it does not raise exercise cardiovascular risk (the attached inactivity may), with a caution to watch for cardiac signs in smokers with COPD. (RIEBE et al., 2015) (inferred from RIEBE et al., 2015) The base-rate argument is applied unevenly: diabetes and known disease are also common relative to exertional events, and they stay in on a relative-risk signal, not on any shown predictive value. The retained branch has the same unmeasured predictive-value problem the panel used to drop risk factors.
The evidence that screening works — none, by the panel’s own account. «There is a lack of evidence that medical clearance and exercise testing (28) are effective in mitigating the risk of exercise-related cardiovascular events. There is also a lack of consensus among major organizations regarding the ex- tent of medical evaluation needed as part of the exercise preparticipation health screening process (2).» The paper sets out the spread: ACC/AHA (2013 exercise standards) recommend exercise testing before moderate or vigorous work when CVD risk is raised, acknowledging «conflicting evidence and divergent expert opinions»; PAGAC 2008 says symptomatic or diseased people who want to begin vigorous activity and have no plan yet «may wish to do so» with their provider, but mandates no contact; USPSTF 2012 recommends against routine resting or exercise ECG in low-risk asymptomatic adults and found insufficient evidence on exercise testing before starting activity. The panel calls its own recommendations «research informed when available». (RIEBE et al., 2015) (inferred from RIEBE et al., 2015) So the guidance set disagrees on the same question, while the bodies that characterize the evidence converge on it being lacking (Riebe’s panel), insufficient (USPSTF, on testing before activity) or conflicting (ACC/AHA), as Riebe reports them. The disagreement is consistent with an evidence base that does not decide the question; all three characterizations reach the page through Riebe alone, the USPSTF and ACC/AHA ones second-hand.
The load-bearing assumption under the no-clearance rule. «Less is known about the risks of SCD and AMI during light- to moderate-intensity exercise, but they are as- sumed to be substantially lower (44).» The triggering research centres on vigorous-to-near-maximal work. (RIEBE et al., 2015) Franklin’s statement cites measurements Riebe did not, both older than Riebe: von Klot 2008 (Eur Heart J, ref 71; RR 1.6 for MI, co-authored by Mittleman) and Willich 1993 (NEJM, ref 60; no elevation with moderate exertion). So Riebe’s «assumed» was partly checkable when written (type-F; shared authors, so not an independent route). (Franklin et al., 2020)
| Parameter | Riebe 2015 — quoted + locus | Franklin 2020 — quoted + locus | Same quantity? |
|---|---|---|---|
| Acute MI/SCD risk during moderate exertion vs rest | «assumed to be substantially lower» (no figure), chunk 01 | «moderate exer- tion (METs=5; RR, 1.6 [95% CI, 1.2−2.1])» in one Augsburg case-crossover study; an earlier study from the same region found no elevation; chunk 01 | partly — Franklin’s figure is MI only, within 2 h of moderate exertion, in one case-crossover study; Riebe’s assumption covers SCD and AMI with no value and no stated comparator. A gap Franklin partly fills for MI, not a disagreement |
| Light exertion | not separately addressed | «Both studies demonstrated no association between lighter levels of exertion and AMI.» chunk 01 | yes (same outcome, same comparator) |
(Franklin et al., 2020; inferred from RIEBE et al., 2015) Moderate exertion is therefore not risk-free in the one positive study: RR 1.6 for MI is about half the meta-analytic 3.5-fold AMI figure for vigorous (≥6 MET) work that Riebe cites (ref 12, Dahabreh & Paulus 2011) and well below the «approxi- mately sixfold (31) to 17-fold (1)» single-study range, which spans nonfatal AMI (sixfold, ref 31, the Onset Study) and SCD (17-fold, ref 1, Albert 2000, the PHS triggering study). This is an indicative cross-study comparison only: the hazard windows differ (within 2 h vs «during» the bout for the meta-analysis), and the 17-fold end is for sudden death, not MI. Light exertion showed no association in two studies. The no-clearance rule for light-to-moderate starts stands on that plus the rarity of events — not on any trial of screening.
Intensity is defined on two scales at once. Figure 2’s bands give relative (%HRR or VO2R), absolute
(METs) and perceived (RPE, breathing) anchors together — moderate is «40% to G60% HRR or V˙ O2R, 3 to G6
METs, 12–13 RPE, an intensity that causes noticeable increases in HR and breathing» (G = <) — without
saying which governs when they disagree. (RIEBE et al., 2015)
Franklin’s statement argues for the relative scale for cardiac load: an absolute ≥6 MET cut «does not account for the fact
that the cardiac demand of any PA is determined not by the specific metabolic level but by the metabolic
demand relative to the individual’s functional capacity».
(Franklin et al., 2020) (Franklin et al., 2020; inferred from RIEBE et al., 2015)
For an unfit older starter, an activity inside Riebe’s absolute moderate band (3 to <6 METs) can sit in
the vigorous relative band: at a peak capacity of 5 METs, a 4-MET activity is 75% of VO2 reserve
((4-1)/(5-1)), above the 60% moderate ceiling. So light-to-moderate in the first row is best read by effort — heart rate and breathing noticeably up, RPE 12-13 — not by the activity’s MET listing.
What regular exercise means here is not the activity guideline. A type-B distinction:
| Parameter | Riebe 2015 — quoted + locus | ESC 2021 activity recommendation — quoted + locus | Same quantity? |
|---|---|---|---|
| Threshold for counting as active | «planned, structured physical activity at least 30 min at moderate intensity on at least 3 dIwkj1 for at least the last 3 months» (≥90 min/wk structured exercise, sustained 3 months), chunk 01 Fig 2 | «strive for at least 150 - 300 min a week of moderate- intensity or 75 - 150 min a week of vigorous- intensity aerobic PA, or an equivalent combina- tion thereof», chunk 03 (class I, level A) | no — Riebe’s figure classifies a person as active (only planned structured exercise, sustained 3 months); the ESC range is a recommendation target for moderate or vigorous aerobic PA, with no structure or duration condition |
(European Society of Cardiology, 2021; inferred from RIEBE et al., 2015) The two thresholds sort some people into opposite camps. Someone doing three 35-minute structured moderate sessions a week (about 105 min) for the last 3 months counts as active for screening (rows 4-6) while falling short of the 150-min recommendation; someone who walks or cycles 150+ min/wk for transport but does no planned exercise counts as inactive for screening and, if free of the named diseases and of symptoms, enters at light-to-moderate before moving to vigorous. Neither cut is wrong. They plausibly answer different questions (Riebe states no rationale for its cut): the screening cut reads as asking whether the person is accustomed to sustained moderate work, the 150-min cut whether they get the long-run benefit. The practical consequence is small, since an inactive transport walker enters at a level they already exceed.
Prescription, more likely than screening, as the protection (the panel’s view). The conclusion puts the weight on how one exercises: «the hazard of exercise-related cardiovascular events may more likely be reduced by careful attention to a safe and effective exercise prescription» — a 2-3-month progressive transitional phase, warm-up and cool-down, warning-sign education, regular brisk walking for the sedentary, and counselling the inactive to avoid unaccustomed vigorous-to-near-maximal work. Resistance training is outside the algorithm: «Because there are few data regarding cardiovas- cular complications during resistance training, this risk can- not currently be determined but seems to be low (20,21,49).» (RIEBE et al., 2015)
The decision for the sedentary, asymptomatic adult with no known heart, vascular, kidney disease or diabetes: start light-to-moderate exercise (judged by effort) without a doctor’s visit, build over about 2-3 months, learn the warning symptoms, and stop and get assessed if any appear. If inactive with known disease, or with any symptom, get clearance first; if already exercising regularly with known disease and no symptoms, moderate exercise needs no new clearance and vigorous needs clearance (one from the last 12 months suffices if nothing has changed). The basis is consensus practice, not a trial: no study shows that clearance or exercise testing prevents events, and an RCT of screening against exercise-triggered death is infeasible at ~1 event per million-plus bouts (expectancy test — the absence is structural, not a null). It is cheap and reversible, and it aims at the exact stratum where the excess sits, so it earns a plain do-this. Franklin’s Oregon data above (15-16% of sports SCAs had known heart disease) point the same way: most people who have an exertional event had no diagnosis to screen on, which is why the safeguard that reaches them is graded entry plus symptom awareness rather than a clinic visit. The panel itself bounds the rule: the recommendations «are not a replace- ment for sound clinical judgment», and referral decisions «should continue to be made on an individual basis». (RIEBE et al., 2015)
Pre-exercise drugs
«Collectively, these data and more recent reports63 suggest there is insufficient evidence to recommend the routine use of either aspirin or β-blockers before strenu- ous PA or sports participation.» The next sentence leaves room for case-by-case use: «Decisions to recommend these prophylactic agents requires assessment of the in- dividual patient’s risk and a discussion with the patient of potential risks and benefits.» Aspirin did not reduce exertion-triggered MI in TIMI II or MIOS, and in RACER no plaque rupture or thrombosis was found in the 5 runners who died with CHD. The β-blocker evidence is mixed: a TIMI II analysis suggested protection, while in MIOS users still had RR 4.2 (1.8-9.9) vs 6.2 (4.7-8.2) in nonusers, not a significant difference. Cross-ref Aspirin for Primary Prevention of Cardiovascular Disease. (Franklin et al., 2020)
Cardiac patients and HIIT
HIIT beat moderate continuous training (MICT) on fitness in three meta-analyses of cardiac patients (+1.78, +1.3 and +1.4 mL/kg/min); in two of them (CHD n=594; HFrEF n=411) the edge disappeared once the arms were matched for energy expenditure, and the statement judges the benefits «comparable». On harm it says «Data on the safety of HIIT in clinical populations are scarce.» One systematic review reported an 8% adverse-event rate with HIIT in cardiometabolic disease, and in one large rehab comparison (Rognmo, n=4846) the event rate was 5.6x higher with HIIT (1 per 23,182 vs 1 per 129,456 patient-hours, absolute risk low in both; no event counts or CI given; the statement calls it «An- other study»). The statement: «additional long-term studies assessing the safety, compliance, and morbidity and mortality after HIIT are needed be- fore it can be widely adopted in patients with known or suspected CAD, especially in unsupervised, non- medical settings.» (Franklin et al., 2020) (inferred from Franklin et al., 2020) The held umbrella review (Poon 2024) cites the same Rognmo study (refs 53-57) for its claim that HIIT appears safe even in relatively high-risk populations, without reporting either of its rates, then concludes that HIIT’s safety concerns are not appreciably greater than MICT’s. Both sources call the safety data thin; Poon reads the study as reassuring, Franklin weighs the relative rate. Riebe 2015 (Franklin and Thompson co-authors) also cites Rognmo (ref 35), without its rates, in a three-reference cluster for exercise risk being «extremely low»; Franklin reports its 5.6x relative rate as a concern. (RIEBE et al., 2015) Which reading applies depends on whether one weighs the absolute or the relative figure; the cheap default for coronary patients is the cautious one (parameter table on Measuring and Raising Cardiorespiratory Fitness). The modality comparison lives on Is the Exercise Modality Doing Any Work; the supervision boundary on Supervised vs Unsupervised Exercise.
Synthesis
Transient risk vs long-run net — two different quantities. The RR rows describe the hour around a bout; the mortality curves on Physical Activity Dose and Mortality describe years of habit. The statement holds both and concludes benefit outweighs risk for most people. The magnitudes alone keep them from conflicting: the absolute acute excess is of the order of one death per million-plus vigorous bouts, and in Oregon sports-related SCA ran at 2.1 against 55.5 per 100k person-years for all other SCA. The habitual-activity gradient adds that the people who accrue the long-run benefit also carry the smallest per-bout excess. The decision consequence is a progression rule, not an avoidance rule — the relative excess is largest for unaccustomed vigorous work by the habitually inactive, so the lever is how one starts. The gradient compares people, not one person over time, so how quickly the excess falls as a habit forms is not measured. This is the cardiac counterpart of the load-progression logic on Training Load Progression and Running Injury, with a different organ and far rarer, more severe events.
Layer-1 placement. Exercise-triggered events do not reorder the hierarchy: for the inactive stratum, inactivity is the big rock and stays the big rock. They add a route-(c) constraint on how it is pulled — which is why the cheap safety rules above earn a plain do-this.
Limits
- Two shared-author consensus statements, no stated search method in either. Selection of the cited studies is unauditable; the underlying primaries are not held.
- Case-crossover designs rely on recall of exertion in the hour before the event (by survivors or proxies), and most triggering studies are in men.
- Heterogeneity: the MI MA reports substantial between-study heterogeneity; the least-active RR range (4.5-107) is too wide to carry a point estimate.
- Internal inconsistency (Franklin, two labels for what looks like one estimate): chunk 01 prints «the adjust- ed RR of AMI within 1 hour of physical exertion was estimated to be 2.3 (95% CI, 2.0−2.7), correspond- ing to a population attributable risk of 7.7% (95% CI, 6.3%−8.8%)»; chunk 02 prints «Any PA in the case period was associated with an OR of 2.31 (99% CI, 1.96–2.72), with a population attributable risk of 7.7% (99% CI, 6.3%–8.8%).» (Franklin et al., 2020) The point estimate, interval bounds and PAR (with identical PAR bounds) match to rounding, so these are probably one INTERHEART estimate printed under different labels — RR vs OR (in a case-crossover design the OR estimates the RR), 95% vs 99% interval, and heavy physical exertion vs any PA as the exposure wording. Which labels are correct needs the INTERHEART paper, which is not held. Either way the magnitude (~2.3-fold within 1 h) is the same.
- Fused superscripts: several values (the PHS and MA gradients) were decoded from PDF text where reference numbers merged with digits; the decode is consistent with the reference list but not checked against the original PDF layout.
- No unsupervised-vs-supervised event-rate comparison exists in the statement; supervised rehab has a denominator, home exercise in the same stratum does not.
- Riebe’s wording on relative risk is loose. Its introduction concedes the risk is «transiently in- creased during vigorous-intensity exercise relative to rest, especially in habitually sedentary individuals with known or occult CVD» and then says «the absolute and relative risks of a cardiovascular event during exercise are extremely low even during vigorous- intensity exercise in asymptomatic individuals (35,39,41)». The body reports an «approxi- mately sixfold (31) to 17-fold (1)» rise vs rest, not restricted to asymptomatic people, so the two are not strictly the same quantity. Read the absolute half; calling a multiple of resting risk «extremely low» on the relative scale does not match the paper’s own figures: the «extremely low» sentence is about asymptomatic people and makes no number, while the sixfold-to-17-fold figures are multiples of resting risk in mixed populations, so the two are related but not the same quantity, and Riebe never reconciles them. One of the three supporting citations, ref 35, is Rognmo 2012, a study of coronary heart disease patients in rehab, not asymptomatic people. (inferred from RIEBE et al., 2015) (RIEBE et al., 2015) (Franklin et al., 2020) (rehab setting)
- Riebe’s text and figure differ on one branch. For the active person with known disease, the text lets those cleared «within the last 12 months» continue moderate exercise without revisiting a provider; the figure box says clearance for moderate is not necessary, with no condition, and attaches the 12-month clause to vigorous work. (RIEBE et al., 2015)
- No outcome data on the 2015 algorithm — no referral rate, yield or event rate under the new rule is reported (srcgrep scope in the screening subsection). It is a consensus rule built on event rarity, the habitual-activity gradient and an assumed lower risk at light-to-moderate intensity.
Gaps
- Expectancy test: an RCT with exercise-triggered death as the outcome is infeasible at ~1 event per million-plus episodes, so the absence of trials here is structural, not a null. Case-crossover is fit-to-question for triggering.
- Does a structured progression protocol lower the excess in new exercisers? No source held tests a progression scheme against acute cardiac events — G-gap; the progression advice rests on the gradient, not on an intervention trial.
- Women’s triggering risk at vigorous intensity is under-measured (too few events to separate it).
- Does the 2015 algorithm change referrals or events? No held source reports its referral rate, missed-case rate or event outcomes — G-gap. An expectancy note: an event-outcome trial is infeasible, but an audit of referral rates and adverse events in exercise settings under the new rule is feasible, and none is held.
- Resistance training is outside the algorithm; its acute cardiac risk is «can- not currently be determined but seems to be low» on few data. (RIEBE et al., 2015)