Cardiovascular disease (CVD) ranks among the leading causes of death and disability worldwide (1), with its pathological basis primarily rooted in atherosclerosis. Atherosclerosis is not merely a simple lipid deposition but an active pathological process triggered by endothelial dysfunction and driven by the combined forces of oxidative stress and chronic inflammation (2). Under the influence of risk factors such as hypertension, smoking, and diabetes, vascular endothelial cells endure persistent oxidative stress (3). This leads to reduced nitric oxide (NO) bioavailability and activates pro-inflammatory signaling pathways, thereby promoting monocyte migration, foam cell formation, and the initial establishment of atherosclerotic plaques (4).

In recent years, plant-based dietary patterns have garnered significant attention for their potential role in cardiovascular health (5). Flavonoids, a class of polyphenolic compounds widely present in plant foods such as fruits, vegetables, tea, red wine, and dark chocolate, are considered key bioactive components contributing to these potential benefits (6). Based on chemical structure, flavonoids can be classified into six major subclasses: flavonones, flavones, flavan-3-ols (including proanthocyanidins), flavonols, anthocyanins, and isoflavones (7). Significant differences exist among these subclasses in terms of bioavailability, dietary sources, and biological activity (8).

Multiple epidemiological studies and randomized controlled trials indicate that flavonoid intake is associated with a lower CVD risk. For example, Grassi et al. (2) systematically reviewed the association of tea flavonoids with endothelial function, noting that flavan-3-ols may improve flow-mediated vasodilation by enhancing NO synthesis or inhibiting its degradation, potentially delaying the progression of atherosclerosis. Furthermore, Bondonno et al. (9) observed in the Multi-Ethnic Study of Atherosclerosis that higher flavonoid intake was significantly associated with lower rates of abnormal ankle-brachial index and carotid plaque development, suggesting that flavonoid intake may be inversely associated with atherosclerosis in distinct vascular regions.

Although existing evidence supports the potential of flavonoids in cardiovascular risk reduction, their associations with different stages of atherosclerosis—such as initial onset, progression, and clinical events—remain incompletely understood. Furthermore, the heterogeneous associations among flavonoid subclasses, their differential relations in specific populations, and their link with clinical endpoints require further systematic evaluation.

Therefore, this study, based on the large-scale prospective UK Biobank cohort, aims to systematically investigate the association between total flavonoid and its subclass intake and the occurrence, progression, and all-cause mortality of post-atherosclerosis cardiovascular disease. It will also conduct in-depth analyses of potential differences across various population subgroups to provide scientific evidence to inform dietary guidelines and strategies for cardiovascular health promotion.

This study utilized a cohort of 207,571 adults without baseline atherosclerosis from the UK Biobank, employing a three-level stepwise confounder-adjusted Cox proportional hazards model alongside subgroup analyses to further explore the independent associations of flavonoid subclasses. It systematically assessed the association between total dietary flavonoids and their subclasses with three key outcomes: the risk of atherosclerosis onset, the risk of post-atherosclerosis cardiovascular disease onset, and all-cause mortality. Research findings indicate that total dietary flavonoid intake exhibits a significant inverse association with the risk of developing atherosclerosis. According to the fully adjusted model (Model 3), participants in the highest quartile (Q4, ≥203.26 mg/day) of total dietary flavonoid intake exhibited a significantly lower risk of atherosclerosis (HR = 0.781, 95% confidence interval: 0.634–0.961, p = 0.020) was significantly lower than that in the lowest quartile (Q1, ≤58.61 mg/day). Subgroup analyses further confirmed that this inverse association remained consistent across different genders, BMI strata, and among individuals with hypertension or diabetes—particularly pronounced in adults aged 55–69 years. In contrast, after adjusting for dietary factors, the association between total flavonoid intake and the risk of cardiovascular disease following atherosclerosis was no longer statistically significant. Furthermore, no significant association was observed between total dietary flavonoid intake and all-cause mortality. The inverse association of flavonoids with atherosclerosis exhibited significant subclass specificity. Flavan-3-ols emerged as the core effect subgroup (Q4 vs. Q1: HR = 0.660, 95% CI: 0.539–0.807, p < 0.001), whereas anthocyanins (Q4 vs. Q1: HR = 0.816, 95% CI: 0.685–0.971, p = 0.022) and flavonols (Q4 vs. Q1: HR = 0.733, 95% CI: 0.608–0.884, p = 0.002) also exhibited independent inverse associations. Conversely, the effects of flavones and flavanones were no longer significant after multivariate adjustment, indicating fundamental differences in the associations with health outcomes of distinct flavonoid subclasses.

CVD is the leading cause of death and disability worldwide, with its pathological basis primarily rooted in atherosclerosis (10). Atherosclerosis is not merely a simple lipid deposition but an active pathological process initiated by endothelial dysfunction and driven by the combined forces of oxidative stress and chronic inflammation (11). Under the influence of risk factors such as hypertension, smoking, and diabetes, vascular endothelial cells endure persistent oxidative stress. This reduces the bioavailability of NO and activates pro-inflammatory signaling pathways (e.g., NF-κβ) (12). Endothelial cells within the body enter a pro-inflammatory state, increasing vascular permeability. Lipids infiltrate and oxidize, while monocytes chemotactically migrate, adhere, and differentiate into macrophages. These phagocytose oxidized lipids to form foam cells, ultimately establishing the rudiments of atherosclerotic plaques (13, 14). Unstable atherosclerotic plaques may rupture, leading to thrombus formation and blood flow interruption, which can result in serious cardiovascular events such as myocardial infarction and stroke. Previous reports indicate that atherosclerosis is a primary cause of mortality and morbidity (15). Therefore, inhibiting endothelial dysfunction, oxidative stress, and chronic inflammation at their source is a key strategy for potentially delaying the onset and progression of atherosclerosis (16), thereby potentially reducing the overall burden of CVD.

Against this backdrop, this study may hold some significance for public health. The risk of developing atherosclerosis gradually decreases with increasing total dietary flavonoid intake. After adjusting for covariates, the highest intake group showed a significant 21.9% lower risk compared to the lowest intake group (HR = 0.781, 95% CI: 0.634–0.961, p = 0.020), confirming the independent association between flavonoid intake and atherosclerosis development. Furthermore, flavan-3-ols, the predominant subclass accounting for 61.4% of total flavonoid intake, emerged as the primary contributor to this association. The high-intake group exhibited a 34.0% lower risk compared to the lowest group (HR = 0.660, 95% CI: 0.539–0.807, p < 0.001); Anthocyanins (HR = 0.816, 95% CI: 0.685–0.971, p = 0.022) and flavonols (HR = 0.733, 95% CI: 0.608–0.884, p = 0.002) also demonstrated independent inverse associations. Furthermore, subgroup analyses confirmed that the inverse association of total dietary flavonoids with atherosclerosis remained consistent across different genders, ages, BMIs, smoking statuses, and hypertension/diabetes statuses. This suggests that increasing flavonoid intake may not be limited to specific low-risk populations but represents a broadly applicable dietary strategy for risk reduction. Its benefits remain consistent across diverse demographic groups, particularly offering a practical dietary intervention direction for individuals with coexisting risk factors such as hypertension and diabetes.

The observed associations of flavonoids are deeply rooted in their bioactivity against the core mechanisms of atherosclerosis. The negative correlation observed in this study can be rationally explained at the molecular and cellular levels. First, improving endothelial function is a key component of flavonoid action. Randomized controlled trials have confirmed that flavan-3-ols, in particular, significantly promote NO release and restore vasodilatory function (12). In a study by Dalgaard et al. (17), flavonoid intake was found to be negatively correlated with atherosclerotic CVD, with the strongest association observed in peripheral atherosclerotic disease. Our study corroborates this finding, identifying Flavan-3-ols as the core subclass driving the inverse associations (Q4 vs. Q1: HR = 0.660). This aligns with Flavan-3-ols being the most abundant dietary flavonoid in humans and the subclass with the strongest evidence for endothelial protection. Second, flavonoids are potent antioxidants and anti-inflammatory agents. They neutralize reactive oxygen species (ROS) and mitigate vascular wall inflammation by upregulating the Nrf2 pathway and suppressing the NF-κβ pathway (18, 19). We found that both anthocyanins and flavanols exhibit independent inverse associations. Previous studies indicate that anthocyanin metabolites reduce IL-6 and vascular cell adhesion molecule-1 (VCAM-1) in human vascular endothelial cells stimulated by oxidized low-density lipoprotein cholesterol or cluster of differentiation 40 (CD40) ligand (20). Furthermore, higher anthocyanin and flavanol intake correlates with reduced systemic inflammation levels. For example, a study by Cassidy et al. (21) showed that intakes of flavonols and anthocyanins were significantly inversely associated with total inflammatory score (IS): flavonols intake in the highest quintile was 42% lower than that in the lowest quintile (−0.72, P for trend = 0.01), while anthocyanin intake in the highest quintile showed a 73% reduction compared to the lowest quintile (−1.48, P for trend < 0.001), with anthocyanins exhibiting superior anti-inflammatory potential.

The associations effects of flavonoids appear to be disease stage-specific. We found that the association between flavonoid intake and CVD risk after atherosclerosis weakens after adjusting for dietary factors, while showing no significant association with all-cause mortality. This result suggests that the primary “target” of flavonoids may lie in the initiation and early progression stages of atherosclerosis. During this phase, flavonoids may help delay or mitigate the formation of initial lesions by improving endothelial function and suppressing oxidative stress and inflammation (22, 23). However, once the disease progresses to advanced stages, where complex necrotic cores and calcification form within plaques, stability becomes more dependent on mechanical structure and acute inflammatory responses (24, 25). At this point, dietary flavonoid intervention alone may struggle to reverse established severe lesions or prevent acute complications. Similarly, Parmenter et al. (26) found that a high-flavonoid diet showed no association with mortality risk in CVD populations but was associated with a lower CVD incidence risk, consistent with our findings. Furthermore, this study revealed the “heterogeneity” within the flavonoid family through subclass analysis. While total flavonoids demonstrated inverse associations, these were primarily driven by core subclasses including flavan-3-ols, anthocyanins, and flavanones. The effects of flavone and flavonol, however, ceased to be significant after adjusting for multiple factors. This may be attributed to their low dietary prevalence, differences in bioavailability, or inherently weaker biological activity. This finding may suggest that in future dietary guidance or nutritional intervention studies, “flavonoids” should not be treated as a homogeneous group. Instead, greater emphasis should be placed on those subclasses with clearly demonstrated strong associations and their primary dietary sources.

This study has several limitations. First, dietary intake was assessed using the Oxford WebQ online 24-h dietary recall questionnaire, which may not accurately reflect long-term dietary habits, particularly for infrequently consumed flavonoid-rich foods. Additionally, the temporal discrepancy between baseline covariate collection and dietary assessment may introduce exposure misclassification, although we enhanced accuracy by restricting analyses to participants completing ≥2 assessments with adequate energy intake. Second, flavonoid intake estimates carry uncertainty: Food grouping (e.g., combining different “berries”) necessitates the use of averages, and the questionnaire struggles to fully capture the impact of cooking processing on flavonoid content. Furthermore, the Oxford WebQ food database does not perfectly match the USDA database, and these factors may collectively underestimate the true association between flavonoids and health outcomes to some extent. Although Oxford WebQ demonstrated reasonable validity against objective biomarkers, recall bias, over- or under-reporting cannot be entirely ruled out (27). Third, regarding outcome ascertainment, cases of atherosclerosis were identified primarily through linkage to hospital inpatient records and death registries using ICD-10 codes (I70). While this ensures high specificity for clinically significant disease, it may miss early-stage, subclinical, or less severe cases managed exclusively in primary care settings. Consequently, the incidence rates reported in our study may underestimate the true burden of atherosclerosis in the population. Finally, as an observational study, although we adjusted for a wide range of potential confounders including socioeconomic status and lifestyle factors (e.g., smoking, BMI), residual confounding cannot be fully excluded. It is possible that high flavonoid intake serves as a marker for a generally healthier lifestyle or unmeasured behaviors that contribute to the observed cardiovascular benefits.

This study suggests that higher dietary flavonoid intake, particularly from foods rich in flavan-3-ols, anthocyanins, and flavonols, is associated with a significantly lower risk of incident atherosclerosis. This inverse association was particularly pronounced in middle-aged and older adults aged 55 to 69. Moreover, incorporating common flavonoid-rich foods—such as apples, berries, and tea—into daily diets may represent a valuable strategy to support cardiovascular health and potentially mitigate the risk of atherosclerosis.