Exercise and Physical Activity in the Prevention and Treatment of Atherosclerotic Cardiovascular Disease

This is a scientific statement (expert consensus document) from the American Heart Association that synthesises evidence from multiple epidemiological studies, randomised controlled trials, and meta-analyses. The writing group examined:

**Primary question:** Does regular aerobic physical activity prevent the development of coronary artery disease (CAD) and reduce symptoms in patients with established cardiovascular disease?

**Secondary questions:** What are the effects of exercise on individual cardiovascular risk factors (blood pressure, cholesterol, insulin resistance, body weight, smoking cessation)?

**Comparators:** Physically active versus sedentary individuals; exercise training versus no exercise; exercise-based cardiac rehabilitation versus usual care

**Outcome measures:** Incidence of CAD events (heart attacks, cardiac death), mortality rates, changes in blood pressure (systolic/diastolic mmHg), lipid profiles (HDL-C, LDL-C, triglycerides in mg/dL), hemoglobin A1c (%), body weight (kg), and smoking cessation rates

The statement focuses specifically on aerobic physical activity (walking, running, swimming, cycling) and does not directly evaluate resistance exercise.

The statement synthesises data from dozens of studies with vastly different populations. Key studies cited include:

**52 exercise training trials** including 4,700 subjects (meta-analysis on lipid effects)

**44 randomised controlled trials** including 2,674 participants (blood pressure meta-analysis)

**HERITAGE study:** 675 normolipidemic subjects (299 men, 376 women) who completed 5 months of supervised exercise training

**Diabetes Prevention Program:** 3,234 individuals at high risk for type 2 diabetes, followed for 2.8 years

**National Weight Control Registry:** 3,000 individuals who lost ≥10% of body weight and maintained loss for ≥1 year (average 5.5 years)

**51 randomised controlled trials** of exercise-based cardiac rehabilitation including 8,440 patients (primarily middle-aged, low-risk men post-myocardial infarction)

**Smoking cessation trial:** 281 healthy women randomised to exercise or health education

The populations studied were predominantly middle-aged adults, with some data on older adults (80+ years) in relative intensity discussions. Most cardiac rehabilitation studies included low-risk men, which limits generalisability to women and higher-risk patients.

The statement draws on multiple measurement approaches across the synthesised studies:

**Physical activity levels:** Self-reported questionnaires (occupational and leisure-time activity), treadmill performance times (exercise capacity in METs), and energy expenditure estimates (kcal/week)

**Cardiovascular fitness:** Maximal oxygen uptake (V̇O₂max), maximal heart rate, treadmill time to exhaustion

**Blood lipids:** Fasting serum HDL-C, LDL-C, and triglycerides (mg/dL)

**Blood pressure:** Resting systolic and diastolic blood pressure (mmHg), measured by standard sphygmomanometry

**Insulin resistance/glucose control:** Fasting glucose, hemoglobin A1c (Hgb A1c, %)

**Body composition:** Body weight (kg), body mass index (BMI)

**Smoking cessation:** Continuous abstinence at end of intervention and at follow-up (self-reported, biochemically verified in some studies)

**Cardiac events:** Myocardial infarction, cardiac death, all-cause mortality

Intensity of exercise was defined using metabolic equivalents (METs), where 1 MET = resting metabolic rate (~3.5 mL O₂/kg/min). Moderate intensity = 4–6 METs (40–60% of V̇O₂max); vigorous intensity = >6 METs (>60% of V̇O₂max).

**Study design:** This is a scientific statement — an expert consensus document that synthesises evidence from multiple study types including prospective epidemiological studies, randomised controlled trials, and meta-analyses. It is not a single study but rather a comprehensive review of the existing literature up to 2003.

**Key design features of the underlying studies:**

**Prospective cohort studies** (e.g., Harvard Alumni Study, multiple occupational/leisure-time activity studies): These followed large groups of people over years to decades, measuring physical activity at baseline and tracking who developed CAD. The prospective design ensures that lower activity levels preceded disease (not the reverse), which strengthens causal inference.

**Randomised controlled trials** (e.g., HERITAGE, Diabetes Prevention Program, cardiac rehabilitation trials): Participants were randomly assigned to exercise training or control conditions. Randomisation balances known and unknown confounders between groups, allowing causal conclusions about exercise effects.

**Meta-analyses:** Statistical pooling of multiple trials (e.g., 52 trials for lipids, 44 trials for blood pressure, 51 trials for cardiac rehabilitation) increases statistical power and precision of effect estimates.

**Blinding:** Most exercise studies cannot blind participants to whether they are exercising or not (you know if you're walking on a treadmill). Some outcome assessors may have been blinded, but this is inconsistently reported across the synthesised studies.

**Duration:** Exercise interventions ranged from 12 weeks (minimum for lipid meta-analysis) to 5 months (HERITAGE) to 2.8 years (Diabetes Prevention Program). Epidemiological follow-up periods spanned years to decades.

**Statistical approach:** The statement reports pooled effect sizes from meta-analyses (mean differences, risk ratios) with confidence intervals where available. The writing group assessed consistency across studies, dose-response relationships, biological plausibility, and independence from other risk factors to infer causality.

**What this design can and cannot prove:**

**Can prove:** That regular physical activity is consistently associated with lower CAD risk across diverse populations and study designs. The combination of prospective epidemiology (showing temporal sequence), randomised trials (showing causal effects on risk factors), and biological plausibility (mechanistic studies) provides strong evidence for a causal relationship.

**Cannot prove:** Precise dose-response relationships (the statement notes the blood pressure dose-response curve appears "flat"). Cannot determine optimal exercise prescription for every individual. Cannot separate effects of exercise from correlated lifestyle changes (diet, weight loss, smoking) in observational studies. Cannot prove long-term adherence effects from short-term trials.

**Major methodological weaknesses acknowledged:**

Self-reported physical activity is imprecise compared to objective fitness measures

Most cardiac rehabilitation trials studied primarily middle-aged, low-risk men

Few studies addressed exercise effects in individuals with lipid disorders or other specific subgroups

Many individual studies did not reach statistical significance (though the overall pattern was consistent)

Exercise withdrawal studies are lacking (cannot prove that stopping exercise increases risk)

**Primary outcomes — CAD prevention:**

The most physically active subjects demonstrated CAD rates approximately **half** those of the most sedentary group (50% risk reduction)

A graded relationship exists: decreasing CAD rates with increasing levels of activity (dose-response)

Studies using objective fitness measures (treadmill time) showed stronger associations than self-reported activity studies

The relationship was independent of other known atherosclerotic risk factors in many studies

**Primary outcomes — Cardiac rehabilitation (treatment of existing CAD):**

Exercise-based cardiac rehabilitation reduces mortality rates in post-myocardial infarction patients

The most comprehensive meta-analysis (51 RCTs, 8,440 patients) found significant reductions in all-cause and cardiovascular mortality (exact pooled risk ratio not provided in the abstract, but prior meta-analyses cited suggest ~20–30% relative risk reduction)

**Secondary outcomes — Lipid effects (meta-analysis of 52 trials, 4,700 subjects, ≥12 weeks):**

HDL-C increased by an average of **4.6%**

Triglycerides decreased by an average of **3.7%**

LDL-C decreased by an average of **5.0%**

**HERITAGE study (675 subjects, 5 months supervised exercise):**

Men: HDL-C increased **1.1 mg/dL (3%)** ; triglycerides decreased **5.9 mg/dL (2.7%)** ; LDL-C decreased **0.9 mg/dL (0.8%)**

Women: HDL-C increased **1.4 mg/dL (3%)** ; triglycerides decreased **0.6 mg/dL (0.6%)** ; LDL-C decreased **4.4 mg/dL (4%)**

Greater HDL-C increases may occur in individuals with baseline hypertriglyceridemia

**Secondary outcomes — Blood pressure (meta-analysis of 44 RCTs, 2,674 participants):**

Average reduction: **3.4 mmHg systolic** and **2.4 mmHg diastolic**

Normotensive subjects: **2.6/1.8 mmHg** reduction

Hypertensive subjects: **7.4/5.8 mmHg** reduction

No relationship between training frequency, session duration, or intensity and magnitude of blood pressure reduction (flat dose-response curve)

**Secondary outcomes — Insulin resistance and diabetes prevention:**

Review of 9 trials (337 type 2 diabetes patients): average reduction in Hgb A1c of **0.5% to 1.0%**

Diabetes Prevention Program (3,234 high-risk individuals, 2.8 years): Lifestyle intervention (average 4 kg weight loss, 8 MET-hr/week increase in activity) reduced onset of type 2 diabetes by **58%** compared to usual care

Lifestyle intervention was more effective than metformin (850 mg BID), which reduced diabetes onset by **31%**

**Secondary outcomes — Weight maintenance (National Weight Control Registry, 3,000 individuals):**

Average weight loss of **30 kg** maintained for average **5.5 years**

**81%** of registrants reported increased physical activity

Women expended **2,445 kcal/week** ; men expended **3,298 kcal/week** in physical activity

**Secondary outcomes — Smoking cessation (1 larger trial, 281 women, 12 weeks):**

At end of intervention: **19.4%** of exercisers achieved ≥2 months continuous abstinence vs. **10.2%** of controls

At 12-month follow-up: **11.9%** of exercisers vs. **5.4%** of controls remained continuously abstinent

**Acute effects (not dependent on long-term training):**

Serum triglycerides reduced for up to **72 hours** after vigorous exercise

HDL-C transiently increased after exercise

Systolic blood pressure reduced for up to **12 hours** after vigorous exercise

Favorable acute effects on glucose homeostasis

**In plain English:**

**CAD risk:** Being physically active roughly **halves your risk** of developing coronary artery disease. This is comparable to the risk reduction from statin therapy in primary prevention.

**Blood pressure:** The average 3.4/2.4 mmHg reduction from exercise is modest — about one-third to one-half the effect of a single blood pressure medication. However, in people with hypertension, the 7.4/5.8 mmHg reduction is clinically meaningful and may eliminate the need for medication in some mild cases.

**Cholesterol:** The 4.6% increase in HDL-C and 5% decrease in LDL-C are small compared to statins (which lower LDL-C by 30–50%) but meaningful at a population level. For an individual with LDL-C of 130 mg/dL, a 5% reduction equals about 6.5 mg/dL.

**Diabetes prevention:** A 58% reduction in diabetes onset over 2.8 years is **dramatic** — better than the leading diabetes medication (metformin, 31% reduction). This required only modest weight loss (4 kg/~9 lbs) and about 6 miles of walking per week.

**Smoking cessation:** The absolute improvement in quit rates (9.2% at 12 weeks, 6.5% at 12 months) is modest but meaningful — roughly doubling the odds of successful quitting.

**Acute effects:** A single bout of vigorous exercise lowers blood pressure for about half a day and reduces triglycerides for up to 3 days. This means the benefits of exercise are partly immediate and transient, not just cumulative.

**What the authors acknowledge:**

Most research has evaluated aerobic activity, not resistance exercise

Self-reported physical activity is imprecise compared to objective fitness measures

Few studies addressed exercise effects in individuals with lipid disorders

Exercise withdrawal studies are lacking (cannot prove that stopping exercise increases risk)

Most cardiac rehabilitation trials studied primarily middle-aged, low-risk men

The blood pressure dose-response curve appears flat (more exercise doesn't necessarily mean more blood pressure reduction)

**What a critical reader would note:**

**Publication date (2003):** This statement is over 20 years old. While the core findings remain valid, more recent research has refined dose-response relationships and identified additional mechanisms.

**No individual-level data:** As a consensus statement, it synthesises group averages. Individual responses to exercise vary enormously (some people are "non-responders" to certain exercise effects).

**Confounding in observational studies:** People who exercise also tend to eat better, smoke less, sleep better, and have higher socioeconomic status. While many studies adjusted for these factors, residual confounding is possible.

**Lack of blinding:** Exercise studies cannot blind participants, which introduces potential for placebo effects and behaviour change beyond the exercise itself.

**Heterogeneity of exercise prescriptions:** The synthesised studies used different exercise doses (frequency, intensity, time, type), making precise prescription recommendations difficult.

**Population limits:** Most data comes from middle-aged adults. Effects in young adults, elderly, women, and ethnic minorities may differ.

**No discussion of harms:** The statement does not address risks of exercise (injury, cardiac events during exercise) in any detail.

**Industry funding:** Not explicitly discussed, but some cited studies may have had funding sources that could influence results.

**For someone running their own n=1 experiment:**

### What to test (specific intervention and dose)

Test a **moderate-intensity aerobic exercise program** based on the CDC/ACSM recommendation that formed the basis of this statement:

**Type:** Brisk walking (4.8 km/h or 3 mph), jogging, cycling, swimming, or any activity using large muscle groups continuously

**Frequency:** 5–7 days per week (most, preferably all, days)

**Duration:** 30+ minutes per session (can be accumulated in shorter bouts of 10+ minutes)

**Intensity:** Moderate (4–6 METs, or 40–60% of V̇O₂max). A practical guide: you should be able to talk but not sing during the activity. Alternatively, aim for 60–75% of your maximal heart rate (rough estimate: 220 – your age, then take 60–75% of that number)

**Total weekly dose:** Aim for at least 150 minutes of moderate-intensity activity, or 75 minutes of vigorous activity, or a combination

For a more aggressive test (based on the National Weight Control Registry data), aim for:

**Energy expenditure:** 2,500–3,300 kcal/week through physical activity (roughly 25–35 miles of walking per week for a 70 kg person)

### Minimum meaningful duration

**For acute effects (blood pressure, triglycerides):** You can detect changes within **12–72 hours** of a single exercise session

**For lipid changes:** At least **12 weeks** (the minimum duration in the meta-analysis showing 4.6% HDL-C increase)

**For blood pressure changes:** **4–8 weeks** of regular exercise

**For meaningful fitness improvements:** **8–12 weeks** of consistent training

**For weight maintenance:** This is a lifelong commitment — the National Weight Control Registry participants maintained their weight loss for an average of **5.5 years**

**For diabetes prevention:** The Diabetes Prevention Program showed effects over **2.8 years**, but risk factor changes (Hgb A1c) can be seen in **3–6 months**

**Recommendation for a self-experiment:** Run your test for a minimum of **12 weeks** to see meaningful changes in lipids and blood pressure. Track acute effects (post-exercise blood pressure, next-day energy) from week 1.

### What to measure (specific