Lifestyle Lowers Heart Risk Across Genetic Profiles, With Limits

A high cardiovascular polygenic risk score does not cancel the value of exercise, diet, or avoiding cigarettes. Across European, Norwegian, Korean, and multiethnic U.S. cohorts, favorable lifestyle factors tracked with lower cardiovascular risk in every genetic-risk group studied. The more interesting question—whether people with high genetic risk benefit more than everyone else—remains unsettled.

The distinction matters. "Lifestyle can offset genetic risk" sounds like a settled gene-by-environment finding. The studies support a narrower, more useful conclusion: genetic risk and lifestyle both predict cardiovascular health, and healthier behavior remains worthwhile regardless of a person's score. Several studies report statistically significant interactions, especially for blood pressure and triglycerides. Others, including work from the same UK Biobank population, find little formal evidence that genes change the size of lifestyle benefit.

Fitness helped across genetic-risk groups, but the interaction missed conventional significance

The most direct evidence for hard cardiovascular events comes from a 2026 UK Biobank prospective cohort study by Collings et al. The researchers followed 69,447 participants for a median of 12 years and recorded 5,448 incident atherosclerotic cardiovascular disease events. They estimated cardiorespiratory fitness with a non-exercise prediction model, then examined fitness alongside a cardiovascular polygenic risk score.

Each one-standard-deviation increase in estimated fitness was associated with an 8% lower hazard of atherosclerotic cardiovascular disease overall (hazard ratio 0.92, 95% confidence interval 0.89–0.95). The association appeared in each genetic-risk stratum: HR 0.87 (95% CI 0.80–0.94) in the low-risk group, 0.94 (95% CI 0.90–0.98) in the intermediate group, and 0.91 (95% CI 0.86–0.97) in the high-risk group.

Those estimates support fitness as a favorable marker across the score range. They do not establish that fitness works differently according to genetic risk. The study's multiplicative interaction test was borderline (P=.06), and its additive interaction test was null (P=.79). In other words, the stratum-specific associations pointed in the same helpful direction, but the investigators could not confirm a conventional statistical interaction.

The absolute-risk results complicate the usual precision-medicine story further. Over 10 years, the difference between low and high fitness was about 1.6 percentage points in the low-genetic-risk group, equivalent to a number needed to treat of roughly 63, versus 1.2 points and an NNT of roughly 83 in the high-risk group. If those estimates reflect a causal fitness intervention, they suggest a slightly larger absolute event difference at lower genetic risk—not preferential benefit for people with high scores.

Replication spans populations, but not one shared cardiovascular endpoint

A 2024 prospective study by Iribarren et al. adds a multiethnic U.S. perspective. It followed 60,568 Kaiser Permanente Northern California members through December 2022 for incident coronary heart disease. The genetic measure was CARDIO inCode, a 12-SNP panel rather than a genome-wide score. Smoking, Mediterranean-diet adherence, and physical activity each had independent associations with coronary heart disease.

Within the highest genetic-risk quintile, participants with all three unfavorable lifestyle factors had about twice the coronary heart disease risk of those with all three favorable factors. That is a large relative contrast, roughly consistent with a 50% lower relative risk in the favorable-lifestyle comparison. But the cohort consisted of insured, relatively health-engaged members, and a 12-SNP panel captures far less genetic architecture than genome-wide scores. Its effect sizes should not be treated as interchangeable with studies that use different scores or outcomes.

The clearest cross-ethnic replication of formal interaction comes from a 2025 analysis by Hur et al. involving 100,543 UK Biobank participants of European ancestry and 7,211 Korean Genome and Epidemiology Study participants. The endpoint was hypertriglyceridemia, a cardiovascular risk factor rather than myocardial infarction, stroke, or cardiovascular death. The investigators used a three-variant panel involving LPL, APOA5, and GCKR.

High triglyceride genetic-risk combinations were associated with 2.2-fold higher risk in the UK cohort and 2.6-fold higher risk in the Korean cohort. Plant-based diet (P=.0008), alcohol intake (P=.0022), and smoking (P<.001) showed replicated PRS-by-lifestyle interactions across the two populations. The relevant biology is plausible because these loci participate in chylomicron and VLDL remodeling. Plausible biology, however, does not turn a triglyceride result into proof that the same interaction changes clinical cardiovascular-event risk.

Norwegian blood-pressure data produced the strongest interaction statistics

In the 2024 HUNT Study analysis by Øvretveit et al., 49,234 Norwegian participants underwent a broad screen of 250 polygenic-score-by-environment combinations for blood pressure traits. The researchers applied Bonferroni correction, which raises the bar for chance findings. Estimated cardiorespiratory fitness was the strongest modifier of systolic blood pressure polygenic risk (P=4.26×10⁻¹¹). Fruit and berry intake also interacted with systolic (P=2.56×10⁻⁷) and diastolic blood pressure risk (P=5.02×10⁻⁴), while vegetable intake did so for systolic (P=1.63×10⁻⁶) and diastolic pressure (P=1.31×10⁻³).

These are compelling statistical signals, but the endpoint remains blood pressure, not clinical cardiovascular disease. The fitness estimate also came from a non-exercise prediction model rather than measured VO₂max. Exposure misclassification often weakens associations, yet prediction error can also make a biological interpretation less certain.

The contradictory UK Biobank result is part of the story

A 2022 UK Biobank study by Livingstone et al. examined 76,958 White British participants and found that an unhealthy lifestyle score and polygenic risk independently predicted cardiovascular disease and mortality. It reported limited evidence of formal gene-by-environment interactions.

This does not erase the fitness findings from the later UK Biobank analysis. The studies used different lifestyle definitions, covariate strategies, and outcomes. It does show that interaction detection depends on analytical choices and is not yet a stable, model-independent biological fact. A favorable lifestyle can coexist with high genetic risk; the evidence does not yet justify telling people that a high score amplifies the payoff from lifestyle change.

Current evidence supports universal lifestyle action across genetic-risk tiers. It does not support using a high cardiovascular polygenic risk score to promise a larger lifestyle benefit.

What the evidence can—and cannot—tell you

All five main cohorts are observational. They can adjust for measured differences between participants, but they cannot fully remove confounding from socioeconomic conditions, prior illness, health care access, or the tendency for genetic and lifestyle factors to cluster. No Mendelian-randomization study in this evidence base tests whether lifestyle causally modifies polygenic-risk-stratified cardiovascular-event risk.

A 2025 observational and Mendelian-randomization study by Wang et al. does support a causal role for physical activity in cardiometabolic multimorbidity among older adults with hypertension. It did not test polygenic-risk interaction, so it strengthens the case for activity without settling the gene-by-lifestyle question.

Likewise, a 2024 study by Carbonneau et al. found that epigenetic age acceleration mediated the association between the American Heart Association's Life's Essential 8 lifestyle score and cardiovascular disease and mortality. That offers a possible route through which accumulated lifestyle exposures influence vascular risk. It is not a PRS-stratified interaction study and cannot confirm that epigenetic aging explains these findings.

What to do with a high genetic-risk score

Prioritize cardiorespiratory fitness, dietary quality, smoking abstinence, and moderate alcohol intake. Fitness has the most consistent signal across the clinical-event and blood-pressure studies. Mediterranean-style and plant-forward dietary patterns, along with smoking avoidance, also align with favorable results in the multiethnic coronary heart disease and cross-ethnic triglyceride cohorts.

Use a polygenic score as risk information, not as an excuse to abandon ordinary prevention or buy an individualized intervention package. The studies agree on the practical point: no tested genetic-risk group showed that healthier behavior was neutral or harmful. They disagree on whether high genetic risk creates a special, larger response. For now, the honest prescription is universal: improve the behaviors with established cardiovascular value, then manage blood pressure, lipids, and other clinical risks with a clinician who can put a genetic score in context.