Eating More Antioxidants Cuts Heart Attack and Stroke Risk, But Only Up to a Point

Eating more fruits, vegetables, and whole grains packs your diet with antioxidants, and a new large-scale study suggests that loading up on them is associated with meaningfully lower risk of two of the world's deadliest cardiovascular conditions. But the same research, drawing on data from over 66,000 UK Biobank participants followed for more than 12 years, finds that the relationship is not linear. Past a certain threshold, extra antioxidant intake stops helping, and may even edge slightly in the wrong direction.

CDAI Score Links Six Antioxidant Nutrients to Cardiovascular Risk

The study, published in Frontiers in Nutrition, tracked incident ischemic heart disease (IHD) and stroke using a metric called the Composite Dietary Antioxidant Index, or CDAI. The CDAI aggregates intake of six antioxidant micronutrients: carotene, selenium, zinc, and vitamins A, C, and E, all calculated from standardized 24-hour dietary recalls. The researchers then mapped CDAI scores against cardiovascular outcomes over a mean follow-up of 12.1 years for IHD and 12.4 years for stroke, after adjusting for age, sex, ethnicity, socioeconomic status, BMI, smoking, physical activity, and a range of clinical conditions.

The headline results are striking, if hedged. Below an IHD threshold value of roughly -0.30 on the CDAI scale, each one-unit increase in the index was associated with an 11% reduction in IHD risk. Below the corresponding stroke threshold of -0.29, a one-unit rise corresponded to an 18% lower stroke risk. Kaplan-Meier survival curves confirmed the pattern: cumulative IHD incidence was highest among those in the lowest CDAI quartile and lowest among those in the third quartile, with the highest-intake group showing an intermediate trajectory. For stroke, the lowest incidence appeared in the second quartile, not the highest.

Why Ischemic Heart Disease and Stroke Remain a Massive Global Problem

The stakes for getting diet right here are substantial. Globally, ischemic heart disease affected an estimated 254 million people in 2021, with nearly 9 million deaths attributed to IHD that year alone. Stroke compounds the burden further, and both conditions share the same underlying biology: oxidative stress and systemic inflammation that damage artery walls and drive atherosclerosis. Research on dietary antioxidants and heart disease prevention has long proposed that oxidized LDL accumulation, which antioxidants can slow, is a central mechanism in coronary artery disease. The appeal of a dietary lever is obvious: unlike statins or blood pressure medications, food is accessible, modifiable, and free of side effects at normal dietary doses.

The difficulty is that individual antioxidant nutrients, studied in isolation, have yielded inconsistent results in clinical trials. The CDAI approach tries to get around this by treating them as a package rather than singling out, say, vitamin E or beta-carotene alone, on the theory that synergistic effects between nutrients matter more than any single compound.

How the UK Biobank Study Was Structured

The researchers started with the full UK Biobank cohort, which recruited roughly 500,000 adults aged 40 to 69 between 2006 and 2010. After excluding anyone with pre-existing IHD or stroke, missing dietary data, or incomplete covariate information, 66,382 participants remained. Dietary data came from the Oxford WebQ, a web-based 24-hour dietary recall tool validated in UK Biobank participants across more than 211,000 people. Incident cases of IHD and stroke were identified by linking to hospital inpatient records and death registries using ICD-10 codes.

Statistical analysis used Cox proportional hazards models, both for CDAI treated as a continuous variable and divided into quartiles. Restricted cubic spline functions were used to test for nonlinear dose-response relationships, and two-piecewise linear regression pinned down the specific inflection points where the risk curve changed slope. Subgroup analyses across gender, education, ethnicity, and clinical factors showed consistent results throughout.

Participants with higher CDAI scores tended to have healthier lifestyle profiles overall: they were less likely to be current smokers and more likely to be physically active, which the authors adjusted for but which cannot be fully disentangled from dietary effects in an observational study.

The Ceiling Effect: Why More Antioxidants Is Not Always Better

The most practically important finding may be what happens above those threshold values. Once CDAI climbs past roughly -0.30 for IHD and -0.29 for stroke, the risk-reduction benefit plateaus and, in the IHD curve, nudges slightly upward. The authors interpret this as consistent with known biology: antioxidants at high concentrations can disrupt the body's redox balance and blunt adaptive hormetic responses, the beneficial low-level stress signals that help cells maintain their own defenses. Dietary antioxidants regulate endothelial function partly through nitric oxide pathways, and those pathways appear to have an optimal operating range rather than a simple more-is-better dose curve.

This fits with a body of evidence from high-dose supplementation trials showing that megadoses of isolated antioxidants like beta-carotene or vitamin E do not reduce cardiovascular risk and may, in some populations, increase it. The CDAI-based analysis suggests the signal is real even for dietary, not supplemental, antioxidants: the transition from insufficient to adequate intake matters far more than pushing intake ever higher.

What This Means for Diet Recommendations

The practical message from this study is targeted rather than sweeping. The authors argue the priority should be improving antioxidant intake among people who are currently low, specifically those below the identified CDAI thresholds, through increased consumption of fruits, vegetables, and whole grains. That is a more modest and arguably more useful conclusion than a blanket call to eat more of everything antioxidant-related.

Key limitations temper the findings. Self-reported dietary recalls are vulnerable to recall bias and misreporting. The CDAI covers only six nutrients and does not account for supplements or the full complexity of diet. And as a prospective observational study, it can identify associations but not establish causation. The UK Biobank's "healthy volunteer" bias also means the cohort skews toward healthier adults, which may not reflect higher-risk populations where the findings matter most.

Even so, for a study built on a cohort this size, tracked over more than a decade, with consistent results across demographic and clinical subgroups, the nonlinear dose-response relationship is a finding worth taking seriously. The inflection points suggest there is a meaningful difference between eating too few antioxidants and eating enough, and that the gap between those two states carries real cardiovascular consequence.

Evidence rating

EMERGING — 78/100

Caveats:

• Observational design cannot establish causation — residual confounding from unmeasured lifestyle factors cannot be ruled out.
• Dietary intake was based on self-reported 24-hour recalls, which are prone to recall bias and measurement error.
• The CDAI does not capture all dietary antioxidants and excludes supplement intake, potentially underestimating total antioxidant exposure.
• The UK Biobank 'healthy volunteer' cohort skews healthier than the general population, limiting generalizability.
• Funding source was not disclosed by the authors.

Large prospective cohort study (n=66,382, UK Biobank) with approximately 12 years of follow-up using Cox proportional hazards models and restricted cubic splines. Demonstrates an association between a composite dietary antioxidant index and reduced risk of ischemic heart disease and stroke below identified threshold values, but as an observational design it cannot establish causation and residual confounding remains possible. Findings have not yet been independently replicated using the same composite index in an external cohort.