Sage Journals: Discover world-class research

Abstract

Introduction

Chronic low-grade inflammation underlies many noncommunicable diseases. Although flavonoids exhibit anti-inflammatory and antioxidant effects, population-level evidence distinguishing subclass and monomer specific associations with systemic inflammation remains limited.

Methods

We analyzed data from 12 681 U.S. adults. Dietary flavonoid intakes were assessed in relation to the neutrophil-to-lymphocyte ratio (NLR), systemic immune–inflammation index (SII), and systemic inflammation response index (SIRI). Effects were modeled per interquartile-range (IQR) increase in intake with false discovery rate (FDR) control within each outcome. Dose–response patterns were evaluated using restricted cubic splines, and heterogeneity in the cyanidin and SII association was assessed in prespecified subgroups. Mechanistic hypotheses were explored using network pharmacology and molecular docking.

Results

In fully adjusted models, higher anthocyanidin intake was associated with lower SII and SIRI but did not remain significant after FDR correction. At the monomer level, peonidin and cyanidin were inversely associated with SII (peonidin: β = −0.73, 95% CI: −1.14 to −0.32, FDR q = 0.023; cyanidin: β = −1.24, 95% CI: −2.07 to −0.41, FDR q = 0.049). Peonidin was inversely associated with SIRI (β = −0.00192, 95% CI: −0.00275 to −0.00110, FDR q = 0.007). Only cyanidin showed consistent associations across primary and spline analyses in SII. The cyanidin–SII association varied by education and by histories of stroke and chronic obstructive pulmonary disease (COPD), with stronger inverse associations among participants with at least a high-school education, without prior stroke, and with COPD. Network pharmacology identified 334 putative cyanidin–inflammation targets and highlighted hub nodes including epidermal growth factor receptor (EGFR), HSP90AA1, AKT1, SRC, PIK3R1, GRB2.

Conclusion

In this cross-sectional study, higher anthocyanidin intake, particularly cyanidin, was associated with lower SII, and in silico analyses prioritized reactive oxygen species and kinase signaling pathways as testable mechanisms. Prospective studies and experimental validation are needed to establish temporal direction and causality.

Keywords

flavonoids systemic inflammation cyanidin network pharmacology molecular docking disease prevention

Introduction

Chronic low-grade inflammation contributes to a broad range of non-communicable diseases, including cardiovascular disorders, metabolic syndrome, neurodegenerative conditions, and cancer.^1–3 Persistent immune activation coupled with oxidative stress accelerates ageing and disease progression, forming self-reinforcing cycles of cellular injury and impaired repair.^4,5 This state, often termed inflammaging, is characterized by sustained, subclinical inflammation and dysregulated redox balance that gradually undermines tissue function.^6–8 Identifying dietary exposures that may modulate this inflammation milieu is therefore an important objective in preventive and nutritional medicine.

Flavonoids are a structurally diverse family of plant-derived polyphenols found in fruits, vegetables, tea, wine, and many medicinal plants, and have been widely studied for antioxidant, anti-inflammatory, and anticancer activities.^9,10 Experimental evidence suggests that flavonoids can attenuate proinflammatory cytokine production, inhibit NF-κB and mitogen-activated protein kinase (MAPK) dependent pathways, and strengthen endogenous antioxidant defenses.^11–14 Epidemiological evidence also associates higher flavonoid consumption with improved cardiometabolic health and reduced inflammation.^15,16 However, associations vary across flavonoid subclasses and individual compounds, likely reflecting differences in chemical structure, dietary sources, metabolism, and bioavailability.

A growing body of evidence suggests that flavonoid actions in humans may be driven less by the parent compounds than by their metabolites.^17,18 After ingestion, many flavonoids undergo rapid phase II conjugation and microbiome-dependent biotransformation to smaller phenolic acids that often predominate in circulation and may mediate biological activity.¹⁹ This is particularly relevant for anthocyanins, whose metabolites have been implicated in redox and inflammation regulatory pathways.²⁰ For cyanidin-containing anthocyanins, protocatechuic acid has been proposed as a major metabolite with anti-inflammatory effects involving oxidative stress responses and NF-κB related pathways.^21,22 These observations motivate exposure assessments that distinguish flavonoids at both the subclass and monomer levels.

To quantify systemic inflammation at scale, blood count derived indices provide pragmatic markers that capture coordinated shifts across innate and adaptive immune compartments. The neutrophil-to-lymphocyte ratio (NLR), systemic immune–inflammation index (SII), and systemic inflammation response index (SIRI) integrate neutrophils and lymphocytes with platelets or monocytes.^23–25 Because redox regulation is tightly coupled to immune cell activation, these indices offer a useful framework for evaluating metabolite-informed hypotheses in population data.

Here, we used the National Health and Nutrition Examination Survey (NHANES, 2007-2008, 2009-2010, and 2017-2018) to examine associations between dietary flavonoid intake and systemic inflammation, quantified by NLR, SII, SIRI. We aimed to (i) compare total flavonoids and six subclasses across each inflammatory index, (ii) identify monomeric flavonoids most consistently associated with these indices, and (iii) generate testable mechanistic hypotheses by integrating network pharmacology and molecular docking to prioritize candidate targets and pathways.

Materials and Methods

Study Design

This cross-sectional analysis was conducted and reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement.²⁶ The NHANES is a biennial, cross-sectional program by the National Center for Health Statistics (NCHS) that evaluates the health and nutrition of the non-institutionalized U.S. population.²⁷ The NCHS Ethics Review Board approved the survey protocols, and all participants gave written informed consent following the revised Declaration of Helsinki principles. The NHANES data can be accessed publicly at https://wwwn.cdc.gov/nchs/nhanes/.

Dietary flavonoid intake data are available only in NHANES 2007-2008, 2009-2010, and 2017-2018. Therefore, we pooled participants from these cycles (initial n = 29 940) and restricted the analysis to adults aged ≥18 years (excluded n = 11 329). We then excluded participants with missing dietary flavonoid intake data (n = 1886), followed by those missing any systemic inflammation index required for the analysis (NLR, SII, or SIRI; n = 781). Finally, individuals with missing covariate data were excluded (n = 3263). After these sequential exclusions (Figure 1), the final analytic sample included 12 681 participants. Data quality was maintained by performing completeness and consistency checks without imputation. Potential selection bias was evaluated by comparing baseline characteristics of included versus excluded participants using weighted standardized mean differences and a weighted logistic regression model predicting inclusion.

Figure 1.

Flowchart of participant inclusion.

Assessment of Flavonoid Intake

Dietary flavonoid intake was estimated using data from the Flavonoid Database of the United States Department of Agriculture (USDA) Food and Nutrient Database for Dietary Studies (FNDDS).^28,29 This database provides comprehensive information on the content of 29 individual flavonoids, encompassing six major subclasses, and corresponds to each relevant NHANES dietary cycle.

Participants’ daily intake of total and specific flavonoid subclasses was estimated using two 24-h dietary recalls.³⁰ They were assisted by trained interviewers to detail all foods and beverages consumed in the previous 24 h, including meals, snacks, and drinks, with specific descriptions and quantities. The first recall was completed through an in-person interview, followed by a telephone interview approximately one week later. The average daily intake for each participant was determined by calculating the mean of the two recalls, or using the single available recall if only one was provided. To facilitate comparisons, total dietary flavonoid intake was categorized into weighted quartiles Q1–Q4.

Measurement of Systemic Inflammation

Systemic inflammation was assessed through three composite indices: the NLR, the SIRI, and the SII. These parameters offer a comprehensive assessment of inflammatory and immune status, indicating the dynamic equilibrium between innate and adaptive immune responses.^31–34 They have been increasingly recognized as reliable indicators of systemic inflammation and immune surveillance in epidemiological research.

Certified medical personnel collected venous blood samples following standardized protocols at the Mobile Examination Centers (MECs).³⁵ All procedures adhered to rigorous quality assurance standards to minimize preanalytical variability. Complete blood counts (CBCs) were obtained using automated hematology analyzers that quantified leukocyte subtypes and platelets based on optical scatter, laser detection, or electrical impedance. Neutrophils, lymphocytes, monocytes, and platelets were quantified as cells per microliter of blood. Quality control was maintained through routine calibration with standardized reference materials and ongoing performance monitoring of analytical instruments. Each assay underwent internal validation to ensure precision and reproducibility. The NHANES Laboratory and Medical Technologists Procedures Manual provides comprehensive laboratory procedures (accessible at https://wwwn.cdc.gov/nchs/nhanes/).

The indices were calculated using the formula: NLR = Neutrophil Count / Lymphocyte Count. SIRI is calculated by multiplying the neutrophil count by the monocyte count and dividing the result by the lymphocyte count. SII is calculated by multiplying the platelet count by the neutrophil count and dividing by the lymphocyte count.

Covariates

Based on prior literature and biological considerations, we prespecified a comprehensive set of covariates with established links to inflammation. Sociodemographic variables included age (years, continuous), sex (male or female), race/ethnicity (non-Hispanic White, non-Hispanic Black, other Hispanic, or other race), education (below high school, high school graduate, or above high school), poverty–income ratio (PIR, continuous), and marital status (married/living with partner, widowed/divorced/separated, or never married). Health behaviors included smoking status, categorized as never (<100 cigarets in lifetime), former (≥100 cigarets in lifetime and not currently smoking), and current (≥100 cigarets in lifetime and currently smoking). Alcohol use was defined as ever consuming ≥12 alcoholic drinks in one year. Physical activity was classified using guideline-based thresholds³⁶: active (≥150 min/week moderate activity or ≥75 min/week vigorous activity, or an equivalent combination), insufficiently active (some activity but below these thresholds), and inactive (no reported moderate or vigorous activity). Clinical and nutritional covariates included body mass index (BMI),³⁷ categorized as underweight (<18.5 kg/m²), normal weight (18.5-24.9 kg/m²), overweight (25.0-29.9 kg/m²), and obese (≥30.0 kg/m²); total energy intake (kcal/day, continuous); and glycated hemoglobin (HbA1c; %, continuous). We additionally accounted for self-reported physician diagnoses of chronic conditions that may influence inflammation milieu: asthma, congestive heart failure, coronary heart disease, myocardial infarction, stroke, cancer or malignancy, and chronic obstructive pulmonary disease (COPD).

Statistical Analysis

All analyses accounted for NHANES's complex, multistage sampling design using the survey package in R (version 4.5.1), incorporating sampling strata, primary sampling units, and Mobile Examination Center (MEC) examination weights to obtain nationally representative estimates.³⁸ To facilitate comparisons across flavonoid variables with different distributions, exposures were standardized to an interquartile-range (IQR) increment. Associations between each flavonoid and inflammatory index were evaluated using weighted linear regression: $Y_{i} = β_{0} + β_{1} Z_{i, e x p o s u r e} + \sum_{k} γ_{k} C_{i k} + ε_{i}$ , where Y_i denotes NLR, SII, or SIRI; Z_i,exposure is the IQR-standardized flavonoid class or monomer; and C_ik indicates covariates.

We constructed three models: Model 1 (M1): unadjusted; Model 2 (M2): adjusted for age and sex; Model 3 (M3): adjusted for the full covariate set. Estimates are presented as regression coefficients (β) per IQR increase in intake with 95% confidence intervals (CIs), and were visualized in forest plots. To address multiple comparisons, false discovery rate (FDR) correction was applied within each outcome for subclass and monomer level analyses, with FDR-adjusted q values reported alongside nominal P values.

Potential dose–response patterns were assessed using restricted cubic spline (RCS) models within the fully adjusted framework. Spline curves are presented as adjusted differences relative to the survey-weighted median intake (reference), with two-sided tests for both the overall association and nonlinearity. For descriptive comparisons across quartiles of total flavonoid intake, categorical variables were compared using survey-weighted χ² tests, and continuous variables using survey-weighted linear regression. All tests were two-sided, and statistical significance was defined as P < .05.

Target-Disease Intersection for Cyanidin and Inflammation

Putative cyanidin targets were retrieved from SwissTargetPrediction (https://www.swisstargetprediction.ch/) and PharmMapper (https://www.lilab-ecust.cn/pharmmapper/). SwissTargetPrediction draws on known ligand–target associations and similarity-based inference, but its coverage is shaped by existing annotations and the chemical space represented in reference datasets.³⁹ PharmMapper enables structure-based target prediction via pharmacophore mapping, although performance depends on the breadth and quality of available pharmacophore and structural models.⁴⁰

Inflammation-associated genes were retrieved from GeneCards (https://www.genecards.org/) by querying “inflammation”, which integrates evidence across multiple multi-omics databases but can include heterogeneous and publication-influenced gene–phenotype links. After mapping to reviewed human proteins in UniProt,⁴¹ overlapping targets were identified by Venn intersection to prioritize candidates for downstream network analyses.⁴² This workflow is intended to nominate candidates and is inherently hypothesis-generating, requiring experimental validation.

Protein–Protein Interaction (PPI) Network Construction and Functional Enrichment

The intersecting gene set was input into the STRING database for Homo sapiens to develop the protein-protein interaction (PPI) network.⁴³ To emphasize reliable functional links and limit spurious connectivity, we applied a high-confidence interaction threshold of 0.70 and removed isolated nodes. Network visualization and topological analysis were conducted using Cytoscape (version 3.10.3).⁴⁴ Nodes exhibiting the highest degree centrality were defined as hub targets, representing potential key regulators within the interaction network.

Functional enrichment analyses for Gene Ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were conducted via the DAVID database (https://david.ncifcrf.gov/).⁴⁵ Enrichment was considered significant for P-values less than 0.05.

Molecular Docking Validation of Hub Targets

Three-dimensional structures of the identified hub proteins were sourced from the RCSB Protein Data Bank (https://www.rcsb.org/).⁴⁶ Before docking, bound ligands and water molecules were removed, and polar hydrogens were added using AutoDockTools (v1.5.7).⁴⁷ The 3D structure of cyanidin was obtained from PubChem (https://pubchem.ncbi.nlm.nih.gov/).

Molecular docking simulations utilized AutoDock Vina (v1.2.3) with grid boxes centered on the predicted active sites of each protein.⁴⁸ Binding affinities with docking energies ≤–5.0 kcal/mol were used as a screening criterion consistent with thermodynamically favorable binding in silico for small-molecule interactions. The docking conformations and intermolecular interaction patterns were visualized and analyzed using Discovery Studio.

Results

Baseline Characteristics of Participants

Baseline characteristics across quartiles of total flavonoid intake are summarized in Table 1. The final analysis included 12 681 participants (median age: 47.00 years), representing a weighted population of 174 994 071 US adults. Median total flavonoid intake was 70.59 mg/day (IQR, 25.36-265.61). Participants in higher intake quartiles reported greater total energy intake (P < .01) and had higher socioeconomic status, reflected by the poverty–income ratio (P < .01). Sex distribution was comparable across quartiles (P = .84), whereas race/ethnicity, education level, and marital status differed by intake level (all P < .01). Obesity was less prevalent in Q4 than Q1 (36.59% vs 44.35%, P < .01). Current smoking was less common in Q4 (18.12% vs 29.75%, P < .01), accompanied by a higher proportion of never smokers (56.70% vs 46.02%). Physical inactivity declined modestly across quartiles (Q1: 25.58%, Q4: 20.42%; P < .01), while alcohol use was comparable (P = .43). HbA1c did not differ materially across intake groups (P = .49). Among inflammatory indices, SII and SIRI differed across quartiles (Q1: 494.00 vs Q4: 465.47 and Q1: 1.09 vs Q4: 1.03, respectively; both P < .01), while NLR showed only modest variation (P = .06).

Table 1.

Weighted Baseline Characteristics of Participants.

Characteristic	Overall	Q1	Q2	Q3	Q4	P-Value
Unweight n	12 681	3380	3313	3186	2802
Weight N	174 994 071	43 746 921	43 736 562	43 761 378	43 749 211
Total flavonoids, mg	70.59 (25.36, 265.61)	13.18 (7.78, 18.43)	42.61 (33.19, 55.18)	126.48 (93.25, 181.15)	541.54 (377.26, 864.20)	<.01
Total energy intake, kcal	1988.50 (1513.00, 2578.50)	1758.50 (1330.50, 2347.50)	2029.50 (1532.50, 2598.50)	2058.00 (1613.00, 2659.00)	2076.00 (1600.50, 2678.50)	<.01
HbA1c, %	5.40 (5.20, 5.80)	5.40 (5.20, 5.80)	5.50 (5.20, 5.70)	5.40 (5.20, 5.80)	5.50 (5.20, 5.80)	.49
WBC, 10³ cells/uL	6.90 (5.70, 8.40)	7.30 (6.00, 8.80)	6.90 (5.70, 8.40)	6.70 (5.60, 8.10)	6.90 (5.70, 8.20)	<.01
Neutrophil, 10³ cells/uL	4.00 (3.10, 5.20)	4.30 (3.30, 5.50)	4.00 (3.10, 5.20)	3.80 (3.00, 4.90)	4.10 (3.10, 5.10)	<.01
Lymphocyte, 10³ cells/uL	2.10 (1.70, 2.50)	2.10 (1.70, 2.60)	2.10 (1.70, 2.50)	2.00 (1.60, 2.50)	2.00 (1.60, 2.50)	<.01
Monocyte, 10³ cells/uL	0.50 (0.40, 0.70)	0.50 (0.40, 0.70)	0.50 (0.40, 0.70)	0.50 (0.40, 0.60)	0.50 (0.40, 0.70)	.02
Platelet, 10³ cells/uL	243.00 (207.00, 287.00)	250.00 (211.00, 293.00)	244.00 (205.00, 288.00)	238.00 (205.00, 280.00)	239.00 (207.00, 285.00)	<.01
NLR	1.95 (1.50, 2.58)	2.00 (1.52, 2.60)	1.94 (1.47, 2.56)	1.89 (1.47, 2.50)	1.95 (1.50, 2.62)	.06
SII	473.07 (342.00, 660.00)	494.00 (358.33, 690.00)	477.75 (340.87, 655.09)	455.00 (329.87, 630.72)	465.47 (345.10, 667.50)	<.01
SIRI	1.04 (0.72, 1.50)	1.09 (0.75, 1.53)	1.05 (0.72, 1.50)	1.00 (0.69, 1.45)	1.03 (0.75, 1.52)	<.01
Age, years	47.00 (33.00, 60.00)	42.00 (30.00, 57.00)	46.00 (33.00, 60.00)	49.00 (35.00, 61.00)	49.00 (36.00, 61.00)	<.01
Poverty income ratio	3.14 (1.58, 5.00)	2.43 (1.22, 4.25)	3.00 (1.57, 5.00)	3.48 (1.79, 5.00)	3.65 (1.91, 5.00)	<.01
Sex, (%)						.84
Female	51.31	50.8	51.65	50.73	52.05
Male	48.69	49.2	48.35	49.27	47.95
Race, (%)						<.01
Hispanic	13.15	13.99	16.06	15.61	6.94
Non-Hispanic Black	10.13	12.54	10.8	9.58	7.59
Non-Hispanic White	69.79	68.42	67.22	66.92	76.59
Other races	6.94	5.05	5.92	7.89	8.88
Education level, (%)						<.01
Above high school	59.52	47.61	58.57	66.82	65.07
Below high school	15.37	20.75	16.01	13.19	11.55
High school graduate	25.11	31.64	25.43	19.99	23.38
Marital status, (%)						<.01
Married/Living with partner	64.44	60.2	63.67	66.69	67.18
Never married	17.3	21.39	16.89	15.2	15.71
Widowed/Divorced/Separated	18.27	18.41	19.43	18.11	17.11
BMI categories, (%)						<.01
Normal	27.44	23.69	28.01	30.21	27.85
Obese	38.17	44.35	37.74	34.01	36.59
Overweight	32.86	29.92	32.81	34.38	34.34
Underweight	1.53	2.04	1.44	1.4	1.22
Smoking status, (%)						<.01
Current	19.72	29.75	18.71	12.32	18.12
Former	25.22	24.23	24.62	26.86	25.19
Never	55.05	46.02	56.67	60.82	56.7
Drinking status, (%)						.43
No	17.29	17.49	16.24	17.3	18.14
Yes	82.71	82.51	83.76	82.7	81.86
Physical activity, (%)						<.01
Active	66.08	63.17	64.9	69.38	66.87
Inactive	21.44	25.58	20.96	18.81	20.42
Less active	12.48	11.25	14.14	11.81	12.71

Continuous variables are reported as weighted median (IQR); categorical variables are shown as weighted percentages. Abbreviation: WBC, white blood cell; NLR, neutrophil-to-lymphocyte ratio; SIRI, systemic inflammation response index; SII, systemic immune-inflammation index.

Detailed subclass and compound-level intakes stratified by total flavonoid quartiles are shown in Table 2. Differences across quartiles were largely driven by flavan-3-ols, which rose from 2.96 (IQR, 0.90-5.90) mg/day in Q1 to 474.64 (IQR, 307.03-790.47) mg/day in Q4 (P < .01), alongside a steady increase in flavonols (P < .01). In contrast, some subclasses displayed less monotonic patterns across quartiles despite higher overall levels in upper intake groups. At the monomer level, the strongest gradients were observed for thearubigins, epigallocatechin 3-gallate, and epigallocatechin (all P < .01). These baseline patterns suggest that higher flavonoid intakes along with sociodemographic and lifestyle factors may be related to systemic inflammation, motivating subsequent analyses that further evaluate associations with inflammatory indices.

Table 2.

Flavonoid Intakes of Studied Participants.

Characteristic	Overall	Q1	Q2	Q3	Q4	P-Value
Total flavonoids, mg	70.59 (25.36, 265.61)	13.18 (7.78, 18.43)	42.61 (33.19, 55.18)	126.48 (93.25, 181.15)	541.54 (377.26, 864.20)	<.01
Flavonoid class
Anthocyanidins, mg	2.10 (0.06, 13.23)	0.08 (0.00, 1.10)	2.90 (0.32, 10.63)	10.49 (1.48, 40.08)	3.40 (0.24, 19.07)	<.01
Flavan-3-ols, mg	17.25 (5.13, 204.72)	2.96 (0.90, 5.90)	11.11 (5.56, 17.63)	34.64 (17.03, 113.92)	474.64 (307.03, 790.47)	<.01
Flavanones, mg	0.52 (0.04, 14.10)	0.07 (0.00, 0.34)	0.72 (0.07, 14.79)	10.00 (0.31, 40.45)	0.82 (0.10, 17.49)	<.01
Flavones, mg	0.54 (0.18, 1.16)	0.17 (0.04, 0.49)	0.49 (0.18, 1.04)	0.75 (0.31, 1.44)	0.88 (0.43, 1.64)	<.01
Flavonols, mg	13.95 (7.45, 24.95)	5.81 (3.29, 9.05)	11.40 (7.18, 17.76)	15.39 (10.78, 22.14)	32.78 (24.25, 46.00)	<.01
Isoflavones, mg	0.01 (0.00, 0.10)	0.00 (0.00, 0.04)	0.01 (0.00, 0.10)	0.01 (0.00, 0.23)	0.01 (0.00, 0.12)	<.01
Flavonoid monomer
Daidzein, mg	0.00 (0.00, 0.04)	0.00 (0.00, 0.00)	0.00 (0.00, 0.04)	0.00 (0.00, 0.12)	0.00 (0.00, 0.06)	<.01
Genistein, mg	0.00 (0.00, 0.05)	0.00 (0.00, 0.02)	0.01 (0.00, 0.05)	0.01 (0.00, 0.12)	0.01 (0.00, 0.06)	<.01
Glycitein, mg	0.00 (0.00, 0.00)	0.00 (0.00, 0.00)	0.00 (0.00, 0.00)	0.00 (0.00, 0.02)	0.00 (0.00, 0.01)	<.01
Cyanidin, mg	0.48 (0.02, 1.95)	0.04 (0.00, 0.31)	0.66 (0.10, 1.90)	1.42 (0.28, 3.50)	0.75 (0.07, 2.56)	<.01
Petunidin, mg	0.00 (0.00, 0.47)	0.00 (0.00, 0.00)	0.00 (0.00, 0.36)	0.18 (0.00, 2.18)	0.00 (0.00, 0.78)	<.01
Delphinidin, mg	0.01 (0.00, 0.81)	0.00 (0.00, 0.00)	0.02 (0.00, 0.68)	0.39 (0.00, 2.85)	0.05 (0.00, 1.22)	<.01
Malvidin, mg	0.00 (0.00, 2.13)	0.00 (0.00, 0.00)	0.00 (0.00, 1.66)	0.74 (0.00, 14.62)	0.01 (0.00, 3.29)	<.01
Pelargonidin, mg	0.00 (0.00, 0.37)	0.00 (0.00, 0.03)	0.02 (0.00, 0.46)	0.06 (0.00, 1.10)	0.01 (0.00, 0.56)	<.01
Peonidin, mg	0.07 (0.00, 0.72)	0.00 (0.00, 0.06)	0.10 (0.00, 0.66)	0.46 (0.02, 2.33)	0.14 (0.00, 1.04)	<.01
Catechin, mg	5.46 (1.89, 10.81)	1.08 (0.24, 2.60)	4.57 (2.22, 7.37)	6.78 (3.64, 11.75)	12.99 (8.85, 20.41)	<.01
Epigallocatechin, mg	0.57 (0.14, 18.25)	0.10 (0.00, 0.27)	0.29 (0.08, 0.58)	1.08 (0.40, 9.67)	44.94 (28.23, 76.92)	<.01
Epicatechin, mg	6.24 (1.58, 14.30)	0.86 (0.20, 2.48)	3.98 (1.48, 9.02)	7.72 (3.77, 13.43)	18.66 (12.53, 28.98)	<.01
Epicatechin 3-gallate, mg	0.02 (0.00, 11.53)	0.00 (0.00, 0.00)	0.00 (0.00, 0.02)	0.14 (0.02, 5.78)	29.45 (19.17, 50.92)	<.01
Epigallocatechin 3-gallate, mg	0.14 (0.00, 28.79)	0.00 (0.00, 0.00)	0.01 (0.00, 0.17)	0.41 (0.05, 14.30)	74.42 (45.26, 131.66)	<.01
Gallocatechin, mg	0.06 (0.00, 1.84)	0.00 (0.00, 0.00)	0.00 (0.00, 0.09)	0.14 (0.00, 0.83)	4.74 (3.00, 8.12)	<.01
Theaflavin, mg	0.00 (0.00, 1.30)	0.00 (0.00, 0.00)	0.00 (0.00, 0.00)	0.00 (0.00, 0.34)	4.74 (2.72, 8.15)	<.01
Theaflavin-3,3′-digallate, mg	0.00 (0.00, 1.48)	0.00 (0.00, 0.00)	0.00 (0.00, 0.00)	0.00 (0.00, 0.33)	5.23 (3.00, 9.00)	<.01
Theaflavin-3′-gallate, mg	0.00 (0.00, 1.13)	0.00 (0.00, 0.00)	0.00 (0.00, 0.00)	0.00 (0.00, 0.14)	4.51 (2.45, 7.70)	<.01
Theaflavin-3-gallate, mg	0.00 (0.00, 1.17)	0.00 (0.00, 0.00)	0.00 (0.00, 0.00)	0.00 (0.00, 0.30)	3.73 (2.17, 6.49)	<.01
Thearubigins, mg	0.00 (0.00, 98.86)	0.00 (0.00, 0.00)	0.00 (0.00, 0.00)	0.00 (0.00, 59.95)	268.52 (165.99, 453.09)	<.01
Eriodictyol, mg	0.00 (0.00, 0.08)	0.00 (0.00, 0.00)	0.00 (0.00, 0.07)	0.03 (0.00, 0.16)	0.00 (0.00, 0.12)	<.01
Hesperetin, mg	0.01 (0.00, 8.05)	0.00 (0.00, 0.00)	0.20 (0.00, 8.93)	3.85 (0.00, 28.28)	0.24 (0.00, 11.04)	<.01
Naringenin, mg	0.29 (0.02, 3.08)	0.06 (0.00, 0.22)	0.37 (0.05, 2.64)	2.86 (0.19, 8.37)	0.39 (0.07, 3.61)	<.01
Apigenin, mg	0.07 (0.01, 0.23)	0.02 (0.00, 0.06)	0.04 (0.01, 0.15)	0.08 (0.03, 0.24)	0.21 (0.10, 0.44)	<.01
Luteolin, mg	0.38 (0.12, 0.90)	0.12 (0.02, 0.39)	0.36 (0.14, 0.87)	0.54 (0.21, 1.18)	0.58 (0.26, 1.17)	<.01
Isorhamnetin, mg	0.49 (0.13, 1.12)	0.28 (0.02, 0.67)	0.57 (0.18, 1.20)	0.64 (0.20, 1.35)	0.58 (0.18, 1.35)	<.01
Kaempferol, mg	2.81 (1.03, 6.53)	0.92 (0.27, 2.13)	1.67 (0.76, 3.99)	2.79 (1.60, 4.78)	8.34 (5.74, 12.80)	<.01
Myricetin, mg	0.69 (0.28, 1.97)	0.20 (0.08, 0.39)	0.40 (0.21, 0.66)	1.02 (0.60, 1.56)	3.41 (2.23, 5.30)	<.01
Quercetin, mg	8.74 (4.71, 15.52)	3.64 (1.92, 5.84)	7.22 (4.51, 10.84)	9.76 (6.77, 14.32)	19.59 (14.00, 27.31)	<.01

Associations of Flavonoid Classes with Systemic Inflammation

To evaluate associations between flavonoid intake and systemic inflammation, we fit weighted regression models for total flavonoids and six subclasses, scaled per interquartile-range (IQR) increment. Estimates are reported from an unadjusted model (M1), an age and sex adjusted model (M2), and a fully adjusted model (M3) (Figure 2A).

Figure 2.

Associations of flavonoid classes and monomers with systemic inflammatory indices. Anthocyanidins and selected anthocyanidin monomers show the most consistent inverse associations with SII and SIRI. (A) Weighted regression coefficients (β) and 95% confidence intervals for total flavonoids and six flavonoid classes in relation to NLR, SII, and SIRI, scaled per interquartile-range (IQR) increase in intake. (B) Corresponding associations for 29 individual flavonoid monomers grouped by subclass. Points denote β estimates and horizontal bars indicate 95% confidence intervals. Results are shown for three models: M1 (unadjusted), M2 (adjusted for age and sex), and M3 (fully adjusted for covariates).

Across outcomes, the most consistent inverse pattern was observed for anthocyanidins. In the fully adjusted model, an IQR higher anthocyanidin intake was associated with lower SII (β = −2.91, 95% CI: −5.05 to −0.76, P = .010, FDR q = 0.072) and lower SIRI (β = −0.006, 95% CI: −0.011 to −0.001, P = .015, FDR q = 0.079). Flavonols were also inversely related to SIRI in M3 (β = −0.025, 95% CI: −0.046 to −0.003, P = .025, FDR q = 0.104). Although these associations were directionally consistent across the adjustment, they did not meet the conventional FDR threshold within outcome.

By contrast, evidence for associations with NLR was limited after full adjustment. Anthocyanidins were inversely associated with NLR in M2 (β = −0.008, 95% CI: −0.015 to −0.001, P = .035), but the estimate attenuated in M3 (P = .072). Several nominal associations observed in less adjusted models were similarly weaker in M3. For example, flavones were inversely related to SII in M2 (β = −4.68, 95% CI: −8.56 to −0.80, P = .019), but not after full adjustment (P = .169). Total flavonoid intake was not materially associated with NLR, SII, or SIRI in fully adjusted models. These results suggest that inflammatory variation may be more closely aligned with specific flavonoid subclasses, particularly anthocyanidins, than with total flavonoid intake.

Associations of Flavonoid Monomers with Systemic Inflammation

We next evaluated 29 flavonoid monomers in relation to NLR, SII, and SIRI (Figure 2B). In fully adjusted models, the most robust associations were observed for anthocyanidin monomers. For SII, higher intakes of peonidin and cyanidin were associated with lower values in M3 (peonidin: β = −0.73, 95% CI: −1.14 to −0.32, P = .0013, FDR q = 0.0229; cyanidin: β = −1.24, 95% CI: −2.07 to −0.41, P = .0055, FDR q = 0.0492). For SIRI, peonidin showed an FDR-significant inverse association (β = −0.00192, 95% CI: −0.00275 to −0.00110, P < .0001, FDR q = 0.00744). Cyanidin and catechin were also inversely associated with SIRI in M3 (cyanidin: β = −0.00326, 95% CI: −0.00541 to −0.00111, P = .0048, FDR q = 0.0526; catechin: β = −0.0274, 95% CI −0.0473 to −0.00745, P = .0094, FDR q = 0.0746), although these did not meet the prespecified FDR threshold.

Beyond these top associations, several monomers exhibited nominal inverse relations with SIRI after full adjustment, notably epicatechin, quercetin, and kaempferol (all P < .05), but these associations attenuated after FDR correction. For NLR, only peonidin and cyanidin were nominally associated in M3 (P < .05), but no monomer remained significant after FDR adjustment, indicating weaker monomer-specific evidence for NLR than for SII or SIRI. Overall, the monomer-level analyses corroborate the subclass findings and highlight anthocyanidin-related compounds as the most consistently associated with lower systemic inflammatory indices.

Dose–Response Relationships from Restricted Cubic Spline Analyses

To further delineate potential dose–response patterns, we fit weighted RCS models for monomers that were associated with the inflammatory indices or showed comparatively larger effect estimates in the primary regression analyses (Figure 3 and Table 3).

Figure 3.

Restricted cubic spline (RCS) analyses for dose–response relationships between key flavonoid monomers and systemic inflammation indices. Spline models show limited evidence of association with NLR but clearer inverse patterns for SII and SIRI for several monomers, with nonlinearity most evident for luteolin. RCS models relate daily intake of catechin, cyanidin, epicatechin, isorhamnetin, kaempferol, luteolin, peonidin, and quercetin to (A) NLR, (B) SII, (C) SIRI. The weighted median intake of each flavonoid served as the reference point. Fitted values are presented as differences from this reference (ie, the predicted outcome at a given exposure minus the predicted outcome at the weighted median). Shaded areas represent 95% confidence intervals around the fitted curves.

Table 3.

RCS Tests for Overall Association and Nonlinearity Between Flavonoid Monomers and Inflammatory Indices.

Monomer	NLR	NLR	SII	SII	SIRI	SIRI
Monomer	P(overall)	P(nonlinear)	P(overall)	P(nonlinear)	P(overall)	P(nonlinear)
Catechin	.4633	.3912	.4954	.5417	.1590	.2449
Cyanidin	.1707	.3331	.0434*	.2092	.0128*	.0726
Epicatechin	.2382	.1307	.5249	.3533	.2438	.1829
Isorhamnetin	.0974	.8356	.0404*	.7125	.0898	.4048
Kaempferol	.4583	.6749	.4632	.5411	.0480*	.0702
Luteolin	.0532	.0585	.0032*	.0143*	.0192*	.0180*
Peonidin	.3595	.4873	.0933	.2915	.3113	.4127
Quercetin	.1091	.1581	.3258	.4643	.0183*	.2864

P (overall) and P (nonlinear) were obtained from fully adjusted RCS models. P < .05 is indicated by an asterisk*.

For NLR (Figure 3A), RCS models provided limited evidence of dose–response relationships across the selected monomers (all P for overall association >.05), with no support for nonlinearity (all P for nonlinearity >.05). In contrast, clearer inverse patterns were observed for SII (Figure 3B) and SIRI (Figure 3C). For SII, cyanidin (P = .0434) and isorhamnetin (P = .0404) were inversely associated. Luteolin showed overall association (P = .0032) and evidence of nonlinearity (P for nonlinearity = .0143), characterized by a steeper decline at lower intakes followed by an attenuation at higher intakes, suggesting a potential threshold-like response that should be interpreted cautiously and validated in future studies. For SIRI, significant inverse associations were observed for cyanidin (P = .0128), kaempferol (P = .0480), luteolin (P = .0192), and quercetin (P = .0183); nonlinearity was again supported for luteolin (P for nonlinearity = .0180).

Subgroup Analysis of the Cyanidin–SII Association

Since cyanidin was the only monomer showing consistent associations with SII across the FDR-controlled primary regression models and the spline analyses, we next examined whether this relationship varied across prespecified population strata (Figure 4). Overall, a higher cyanidin intake (per IQR increase) was associated with a lower SII (adjusted β = −1.16, 95% CI: −1.99 to −0.33).

Figure 4.

Subgroup analysis of the association between cyanidin intake and SII. Inverse associations are broadly consistent across strata, with evidence of heterogeneity by education and by histories of stroke and COPD. Forest plot showing the association between cyanidin intake (per IQR increase) and SII across prespecified subgroups. Points indicate fully adjusted β coefficients and horizontal bars denote 95% confidence intervals.

Across most subgroups, estimates were directionally consistent, and interaction tests did not support heterogeneity by age, sex, race/ethnicity, poverty–income ratio, marital status, smoking, alcohol use, physical activity, BMI category, HbA1c, or major cardiometabolic comorbidities (all P for interaction >.05). Evidence of effect modification was observed for education (P for interaction = .04): the inverse association was strongest among high school graduates (β = −3.93, 95% CI: −6.22 to −1.64), weaker among those with education beyond high school (β = −0.90, 95% CI: −1.62 to −0.18), and not evident among those with less than a high-school education (β = 0.95, 95% CI: −1.83 to 3.73). Interaction tests were also significant for history of stroke (P for interaction = .04) and COPD (P for interaction = .04). Among participants without prior stroke, cyanidin intake was inversely associated with SII (β = −1.20, 95% CI: −2.05 to −0.35), whereas estimates in the stroke subgroup were imprecise and directionally inconsistent (β = 9.30, 95% CI: −4.07 to 22.67), potentially due to limited stability in this smaller stratum. The inverse association appeared stronger among participants with COPD (β = −5.65, 95% CI: −11.01 to −0.30) than among those without COPD (β = −1.06, 95% CI −1.82 to −0.29), although the confidence interval was wider in the COPD stratum.

Network Pharmacology Analysis of Cyanidin in Inflammation

Given the epidemiologic consistency observed for cyanidin, we used a network-pharmacology framework to generate mechanistic hypotheses linking cyanidin exposure to systemic inflammation. Putative cyanidin targets were obtained from SwissTargetPrediction and PharmMapper, and inflammation-related genes were retrieved from GeneCards. The intersection yielded 334 overlapping genes (Figure 5B), representing candidate molecular mediators linking cyanidin exposure to inflammatory regulation.

Figure 5.

Network pharmacology framework linking cyanidin to inflammation-related targets. In silico target intersection prioritizes oxidative-stress and kinase-related programs among candidate mediators. (A) Chemical structure of cyanidin. (B) Overlap between predicted cyanidin targets and inflammation-associated genes. (C) GO enrichment analysis of the shared targets. Bars indicate enrichment scores for the top GO terms across biological process (BP), cellular component (CC), and molecular function (MF) categories. (D–F) CNET plots illustrating gene–term relationships for top enriched GO terms in BP (D), CC (E), and MF (F).

GO enrichment of the 334 shared targets highlighted stress-adaptive programs centered on responses to xenobiotic and chemical stress, oxidative stress, and reactive oxygen species (ROS) metabolic processes (Figure 5C). The CNET plots (Figure 5D-F) further showed that multiple genes were mapped across these terms, suggesting convergence on a shared set of regulatory nodes. Consistent with redox-linked regulation of phosphorylation cascades, enriched terms also included positive regulation of kinase activity and related signaling functions. KEGG enrichment extended this by prioritizing interconnected kinase-centered pathways—FoxO, PI3K–Akt, Ras, and MAPK—alongside stress and survival programs such as mTOR and autophagy (Figure 6A and B). Importantly, the pathway–pathway network (Figure 6B), in which edges indicate shared genes between pathways, positioned PI3K–Akt/MAPK/Ras/FoxO as a densely connected core linking oxidative-stress programs with downstream inflammatory signaling modules. This network structure is compatible with a model in which changes in redox state may propagate through kinase cascades that coordinate cellular stress responses, survival, and inflammatory signaling.

Figure 6.

KEGG pathway enrichment of cyanidin–inflammation shared targets. Enrichment highlights a kinase-centered core (PI3K–Akt/MAPK/Ras/FoxO) linked to stress-response and immune signaling pathways. (A) Sankey plot summarizing enriched KEGG pathways for shared targets; nodes represent genes and pathways, and the bubble plot indicates gene ratio and enrichment significance. (B) Pathway–pathway network derived from enriched KEGG terms; nodes are pathways and edges indicate shared genes. Node size reflects the number of mapped genes and node color indicates enrichment significance.

Downstream of these kinase-centered modules, KEGG terms also encompassed canonical immune and cytokine pathways, including TNF, IL-17, chemokine, JAK–STAT, and T cell receptor signaling (Figure 6B). The enrichment map indicates that these immune pathways share overlapping genes with the central PI3K–Akt/MAPK/Ras/FoxO cluster, suggesting potential crosstalk through shared mediators. Additional enrichment of AGE–RAGE signaling, lipid and atherosclerosis, cellular senescence, apoptosis, and platelet activation further supports integration of metabolic stress and vascular–immune interfaces, processes that can influence circulating neutrophil, lymphocyte, and platelet dynamics. Overall, these network patterns nominate a hypothesis that cyanidin exposure may relate to systemic inflammatory burden by modulating redox homeostasis, which in turn reprograms kinase signaling and downstream immune pathways, ultimately contributing to shifts in immune-cell and platelet compartments captured by SII.

Protein–Protein Interaction Network and Identification of Hub Targets

To prioritize candidate molecular mediators among the overlapping genes, we constructed a PPI network and ranked nodes using complementary topological metrics, including degree, betweenness centrality, and closeness centrality (Figure 7A and Table 4). This analysis highlighted a set of highly connected and centrally positioned hub targets, led by epidermal growth factor receptor (EGFR) (degree = 57), HSP90AA1 (degree = 56), AKT1 (degree = 54), SRC (degree = 47), PIK3R1 (degree = 41), and GRB2 (degree = 40), with additional hubs including ESR1, HSP90AB1, ALB, and PTPN11. These hubs map to signaling modules implicated by the enrichment analyses, consistent with a network structure in which kinase-driven pathways form major interaction backbones.

Figure 7.

PPI network and molecular docking of cyanidin–inflammation hub targets. Network topology highlights central kinase-related hubs, and docking supports plausible binding of cyanidin to selected targets in silico. (A) PPI network for the shared target set, visualized in Cytoscape; node size and color reflect degree (connectivity). (B–G) Representative docking poses of cyanidin with the six prioritized hub targets: AKT1 (B), EGFR (C), GRB2 (D), HSP90AA1 (E), PIK3R1 (F), and SRC (G).

Table 4.

PPI Network Metrics of Hub Targets.

Target	ClosenessCentrality	BetweennessCentrality	Degree
EGFR	0.4228571428571429	0.09797215208016745	57
HSP90AA1	0.4192634560906516	0.10396816446976469	56
AKT1	0.43465491923641697	0.14168930310516692	54
SRC	0.38441558441558443	0.031185263675663637	47
PIK3R1	0.36363636363636365	0.022634531305767924	41
GRB2	0.3704630788485607	0.012943056473561672	40
ESR1	0.4099722991689751	0.06228662142434089	36
HSP90AB1	0.3790012804097311	0.026083203741234365	35
ALB	0.3983849259757739	0.12402455522290509	31
PTPN11	0.34579439252336447	0.0051598796717173415	30

We then performed molecular docking to assess whether cyanidin could plausibly engage the top-ranked hubs at the protein level (Figure 7B-G and Table 5). Across the six proteins prioritized for docking (AKT1, EGFR, GRB2, HSP90AA1, PIK3R1, and SRC), predicted binding energies ranged from −5.9 to −9.3 kcal/mol. The most favorable energies were observed for HSP90AA1 (−9.3 kcal/mol) and SRC (−8.8 kcal/mol), followed by EGFR (−8.4 kcal/mol) and AKT1 (−7.8 kcal/mol). Energies in this range are consistent with thermodynamically favorable binding in silico and support the plausibility of target engagement, but in vivo potency still requires further experimental confirmation. Docked poses were characterized by multiple noncovalent contacts within the predicted binding pockets (Figure 7B-G). Collectively, these in silico results nominate a hub-centered axis as a mechanistic interface through which cyanidin could influence inflammatory regulation.

Table 5.

Predicted Binding Affinities of Cyanidin with Top Six Hub Targets.

Protein (gene Symbol)	PDB ID	Vina Binding Energy (kcal/mol)
AKT1	3CQU	−7.8
EGFR	7K1I	−8.4
GRB2	6ICG	−6.1
HSP90AA1	3O0I	−9.3
PIK3R1	7CIO	−5.9
SRC	2SRC	−8.8

Discussion

In this nationally representative cross-sectional study of U.S. adults, associations between dietary flavonoid intake and blood count derived inflammatory indices differed across subclasses and individual compounds. After multivariable adjustment, total flavonoids and most subclasses showed little evidence of association with NLR, SII, or SIRI. In contrast, at the monomer level, cyanidin and peonidin were inversely associated with SII in fully adjusted models after FDR control, and peonidin was also inversely associated with SIRI. Cyanidin was the only monomer for which the inverse association with SII remained consistent in both the primary regression analyses and the restricted cubic spline models.

A plausible biochemical rationale for cyanidin begins with its structure (Figure 5A). Cyanidin contains multiple hydroxyl groups, including a catechol motif on the B ring (3′,4′-dihydroxylation), which enhances electron-donating capacity and can facilitate interactions with ROS and redox-sensitive signaling pathways.^49,50 More broadly, the number and position of hydroxyl substituents, together with the extent of conjugation, shape antioxidant behavior and anti-inflammatory potential across polyphenol classes.^51,52 In vivo, however, anthocyanidins and other flavonoids can be metabolized through phase II conjugation and microbiome-dependent biotransformation.^53,54 Circulating forms are typically glycosylated, glucuronidated, sulfated, or methylated conjugates and smaller phenolic metabolites.^55–58 Some conjugated metabolites retain signaling activity,⁵⁹ and inflammatory microenvironments may promote local deconjugation (eg, via β-glucuronidase), potentially increasing the local availability of more reactive aglycone-like species where immune cells are activated.^60,61 These processes may help explain why associations are detectable in population data even when reported intakes are modest and right-skewed, as observed for most monomers in this dataset.

RCS analyses provided further insight into dose–response shape. For several monomers, the associations were approximately linear across the observed intake range (cyanidin and isorhamnetin with SII; cyanidin, kaempferol, and quercetin with SIRI). One interpretation is that, within this range, incremental increases in intake correspond to roughly proportional increases in circulating metabolites and downstream biological exposure. For quercetin, human pharmacokinetic studies indicate relatively slow elimination, with half-lives on the order of ∼1 day,⁶² supporting the possibility of sustained exposure under habitual intake and may favor monotonic associations in observational settings. For isorhamnetin, O-methylation relative to quercetin may increase metabolic stability and cellular persistence, potentially extending the range over which intake differences remain biologically relevant.^63,64 For kaempferol, available evidence suggests that circulating exposure is dominated by glucuronidated and sulfated conjugates rather than the aglycone following phase I and phase II metabolism.⁵⁸ Under usual dietary patterns, an approximately linear spline could therefore reflect dose-proportional variation in these conjugated pools across mid-range intakes, even when absolute bioavailability is limited.⁶⁵

In contrast, luteolin exhibited a tendency toward plateauing for SII and SIRI, which may reflect a combination of biology and data distribution. Its disposition involves glucuronidation and methylation (UGTs/COMTs),⁶⁶ processes that could limit further increases in circulating bioactive pools once metabolic or transport capacity becomes rate-limiting at higher intakes. In addition, monomer intakes are right-skewed and are more susceptible to day-to-day variability and measurement error inherent to 24-h recalls. These features can visually compress the upper tail and flatten apparent dose–response curves. We also notice that peonidin showed inverse associations in the primary regression models but did not retain a concordant pattern in spline analyses. This may be due to sparse exposure with many near-zero intakes destabilizing flexible spline fits and reducing the power to detect nonlinearity or even preserve directionality. Moreover, peonidin intake is is also correlated with other anthocyanidins and shared food sources, increasing sensitivity to multivariable adjustment and to exposure misclassification from 24-h recalls. Under these conditions, a conventional linear term may capture an average contrast, while spline estimates fluctuate when data density is low in higher intake regions.

In the primary regression analyses, effect estimates often attenuated, and in some cases changed direction, after full adjustment, consistent with nonrandom patterning of flavonoid intake. In NHANES, higher flavonoid consumption tends to cluster with socioeconomic position, smoking, adiposity, and overall diet quality, factors that are independently related to systemic inflammation and can inflate associations in minimally adjusted models.⁶⁷ Age is also a major source of confounding. Ageing is accompanied by chronic low-grade inflammation and coordinated shifts in immune cell compartments that influence hematologic indices even in the absence of acute illness.⁶⁸ Adjusting for age, health behaviors, and comorbidities may therefore separate associations from correlated lifestyle and disease profiles. This interpretation is consistent with epidemiologic work showing that diet quality and inflammatory potential relate to inflammatory markers.⁶⁹ Subgroup analyses further suggest context dependance for the cyanidin and SII association. The interaction by education may reflect differences in dietary access and patterns, health literacy, and correlated behaviors that shape inflammatory status.^70,71 We also observed modification by history of stroke and COPD. The estimate among participants with prior stroke was positive but imprecise, consistent with limited stability in this smaller stratum (n = 499) and possible post-event changes in diet, medication use, or blood counts. The stronger inverse association in COPD could align with higher baseline inflammatory and oxidative-stress burden, which may widen the dynamic range of SII, although residual confounding remains possible in a cross-sectional design.⁷²

The network-pharmacology enrichment results align with a mechanistic path that is coherent with blood count derived inflammatory indices. Enriched processes centered on reactive oxygen species handling and kinase signaling (PI3K–Akt, MAPK, Ras, and FoxO), with links to AGE–RAGE and VEGF-related programs. These pathways regulate neutrophil survival, lymphocyte homeostasis, and platelet reactivity, key determinants of SII and SIRI. PI3K–Akt signaling has well-established roles in regulating neutrophil survival,⁷³ and MAPK pathways intersect with oxidative burst and inflammatory activation.⁷⁴ FoxO transcription factors integrate oxidative stress signals with apoptosis and cell-cycle control, influencing lymphocyte turnover and stress resilience.^75,76 AGE–RAGE signaling, particularly relevant in metabolic stress contexts, can amplify oxidative stress and inflammatory cascades which also promote myeloid activation and platelet responsiveness,^77,78 biologic components directly reflected in SII and SIRI. Importantly, platelets are integral to SII, and experimental work shows that cyanidin-3-glucoside can inhibit platelet activation and thrombus formation via GPVI-linked signaling,⁷⁹ supporting a biological route by which cyanidin-related exposures could relate to a platelet-weighted inflammation index. Together, these links support a model in which cyanidin associated redox modulation engages kinase networks that shape innate immune activation and platelet activity, which is then captured by composite hematologic indices.

Several limitations warrant caution in causal interpretation. First, the cross-sectional design precludes establishing temporality, and reverse causation is possible (eg, diet changes after illness). Second, flavonoid intake derived from 24-h recalls is vulnerable to measurement error and within-person variability, particularly for episodically consumed, anthocyanin-rich foods. Third, although we adjusted for a broad covariate set and applied within-outcome FDR control, residual confounding remains plausible because flavonoid intake is closely correlated with socioeconomic position and health behaviors. Fourth, the complete-case approach excluded participants with missing exposure, outcome, or covariate data. Selection related to sociodemographic characteristics (eg, age, race/ethnicity, PIR, education, and marital status) may affect generalizability and could influence effect estimates. Finally, the mechanistic analyses are in silico and are intended to prioritize hypotheses rather than demonstrate target engagement or biological effects in vivo.

Despite these constraints, the results motivate follow-up work focused on anthocyanidin composition and selected monomers, together with targeted testing of cyanidin-linked redox and kinase pathways in immune and platelet biology. Prospective cohorts with repeated dietary assessment and metabolomic profiling, complemented by experimental studies of cyanidin and its metabolites, will be needed to determine whether the observed associations reflect causal modulation of systemic inflammatory burden.

Conclusion

In this nationally representative cross-sectional analysis of US adults, higher anthocyanidins intake was associated with lower systemic inflammatory burden as indexed by SII and SIRI, while associations with NLR were weak. Cyanidin showed the most consistent inverse association with SII. Complementary in silico network pharmacology and docking analyses prioritized redox and kinase linked pathways, including PI3K–Akt/MAPK related hubs, as plausible mechanistic candidates. These findings link dietary anthocyanidins, particularly cyanidin, to variation in composite blood count derived inflammatory indices and motivate follow-up studies integrating longitudinal dietary assessment, metabolomics, and targeted experimentation to test causal mechanisms and translational relevance.

Footnotes

Abbreviations

ORCID iDs

Haomou Pu

Xiaoping Yu

Ethical Approval

The present study is a secondary analysis of publicly available data from the U.S. National Health and Nutrition Examination Survey (NHANES). NHANES protocols are reviewed and approved annually by the NCHS Ethics Review Board, and all procedures comply with the U.S. Department of Health and Human Services ethical standards. Because this research uses de-identified, publicly accessible data, no additional institutional ethical approval was required, and no approval number applies to the secondary analysis.

Statement of Informed Consent

NHANES obtains written informed consent from all participating individuals prior to data collection. Detailed consent procedures, including permission for future research use of de-identified data, are conducted by trained NCHS personnel. Because the present work uses only anonymized NHANES data released for public research purposes, no additional informed consent was required for this analysis.

Statement of Human and Animal Rights

This study involved no direct interaction with human participants, no collection of new biological specimens, and no animal experiments. All analyses were performed on de-identified NHANES datasets released by the NCHS. The study therefore adheres fully to the ethical principles outlined in the U.S. Federal Regulations for research using existing, anonymized data, and poses no risk to human or animal subjects.

Author Contributions

H.P. and H.W.: conceptualization, data curation, formal analysis, visualization, writing–original draft, writing–review and editing. J.G. and Y.C.: methodology, formal analysis, writing–review and editing. H.J.: conceptualization, writing–original draft, writing–review and editing. X.Y. and Y.Z.: supervision, project administration, resources, writing–review and editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was funded by the National Natural Science Foundation of China (No. 82073539) and the Natural Science Foundation of Sichuan Province (No. 2023NSFSC0682).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Data Availability Statement

Datasets analyzed in this study are available in the Center for Disease Control and Prevention repository; available at . Further information and requests can be directed and will be fulfilled by the corresponding author.

References

Ruiz-Núñez

Pruimboom

Dijck-Brouwer

DAJ

Muskiet

FAJ

. Lifestyle and nutritional imbalances associated with Western diseases: causes and consequences of chronic systemic low-grade inflammation in an evolutionary context. J Nutr Biochem. 2013;24(7):1183-1201. doi:10.1016/j.jnutbio.2013.02.009

Voss

. Overview of anti-inflammatory diets and their promising effects on non-communicable diseases. Br J Nutr. 2024;132(7):898-918. doi:10.1017/S0007114524001405

Phillips

Chen

L-W

Heude

, et al. Dietary inflammatory index and non-communicable disease risk: a narrative review. Nutrients. 2019;11(8):1873. doi:10.3390/nu11081873

Zhou

Bao

W-Q

Zhang

, et al. Immunometabolism and oxidative stress: roles and therapeutic strategies in cancer and aging. NPJ Aging. 2025;11(1):59. doi:10.1038/s41514-025-00250-z

Zhang

Wang

Qian

Huang

. Inflammation and aging: signaling pathways and intervention therapies. Signal Transduct Target Ther. 2023;8(1):239. doi:10.1038/s41392-023-01502-8

Higashi

. Roles of oxidative stress and inflammation in vascular endothelial dysfunction-related disease. Antioxidants. 2022;11(10):1958. doi:10.3390/antiox11101958

Baechle

Chen

Makhijani

Winer

Furman

Winer

. Chronic inflammation and the hallmarks of aging. Mol Metab. 2023;74:101755. doi:10.1016/j.molmet.2023.101755

Di Carlo

Sorrentino

. Oxidative stress and age-related tumors. Antioxidants. 2024;13(9):1109. doi:10.3390/antiox13091109

Ullah

Munir

Badshah

, et al. Important flavonoids and their role as a therapeutic agent. Molecules. 2020;25(22):5243. doi:10.3390/molecules25225243

10.

El-Seedi

Yosri

Khalifa

SAM

, et al. Exploring natural products-based cancer therapeutics derived from Egyptian flora. J Ethnopharmacol. 2021;269:113626. doi:10.1016/j.jep.2020.113626

11.

Gerritsen

Carley

Ranges

, et al. Flavonoids inhibit cytokine-induced endothelial cell adhesion protein gene expression. Am J Pathol. 1995;147(2):278-292.

12.

Endale

Park

S-C

Kim

, et al. Quercetin disrupts tyrosine-phosphorylated phosphatidylinositol 3-kinase and myeloid differentiation factor-88 association, and inhibits MAPK/AP-1 and IKK/NF-κB-induced inflammatory mediators production in RAW 264.7 cells. Immunobiology. 2013;218(12):1452-1467. doi:10.1016/j.imbio.2013.04.019

13.

Jia

Xie

, et al. (-)-Epigallocatechin-3-gallate protects against lithium-pilocarpine-induced epilepsy by inhibiting the toll-like receptor 4 (TLR4)/nuclear factor-κB (NF-κB) signaling pathway. Med Sci Monit Int Med J Exp Clin Res. 2019;25:1749-1758.

14.

Behl

Kaur

Sehgal

, et al. Flavonoids, the family of plant-derived antioxidants making inroads into novel therapeutic design against ionizing radiation-induced oxidative stress in Parkinson’s disease. Curr Neuropharmacol. 2022;20(2):324-343. doi:10.2174/1570159X19666210524152817

15.

Mazidi

Katsiki

Banach

. A higher flavonoid intake is associated with less likelihood of nonalcoholic fatty liver disease: results from a multiethnic study. J Nutr Biochem. 2019;65:66-71. doi:10.1016/j.jnutbio.2018.10.001

16.

Lucía

Sebastián

Georgina

Daniela

Raquel

. Cardiometabolic risk, biomarkers of low-grade subclinical inflammation and flavonoid intake: a cross-sectional study in Argentina. PharmaNutrition. 2022;20:100297. doi:10.1016/j.phanu.2022.100297

17.

Cassidy

Minihane

A-M

. The role of metabolism (and the microbiome) in defining the clinical efficacy of dietary flavonoids1. Am J Clin Nutr. 2017;105(1):10-22. doi:10.3945/ajcn.116.136051

18.

Baky

Elshahed

Wessjohann

Farag

. Interactions between dietary flavonoids and the gut microbiome: a comprehensive review. Br J Nutr. 2022;128(4):577-591. doi:10.1017/S0007114521003627

19.

Kiriyama

Tokumaru

Sadamoto

Kobayashi

Nochi

. Effects of phenolic acids produced from food-derived flavonoids and amino acids by the gut microbiota on health and disease. Molecules. 2024;29(21):5102. doi:10.3390/molecules29215102

20.

Hidalgo

Martin-Santamaria

Recio

, et al. Potential anti-inflammatory, anti-adhesive, anti/estrogenic, and angiotensin-converting enzyme inhibitory activities of anthocyanins and their gut metabolites. Genes Nutr. 2012;7(2):295-306. doi:10.1007/s12263-011-0263-5

21.

Vitaglione

Donnarumma

Napolitano

, et al. Protocatechuic acid is the major human metabolite of cyanidin-glucosides123. J Nutr. 2007;137(9):2043-2048. doi:10.1093/jn/137.9.2043

22.

Wang

Wei

Yan

Jin

Ling

. Protocatechuic acid, a metabolite of anthocyanins, inhibits monocyte adhesion and reduces atherosclerosis in apolipoprotein E-deficient mice. J Agric Food Chem. 2010;58(24):12722-12728. doi:10.1021/jf103427j

23.

Xia

Qing

, et al. Systemic Inflammatory Response Index (SIRI) is associated with all-cause mortality and cardiovascular mortality in population with chronic kidney disease: evidence from NHANES (2001–2018). Front Immunol. 2024;15. doi:10.3389/fimmu.2024.1338025. Epub ahead of print 15 March 2024.

24.

Zhao

, et al. The clinical value of the neutrophil-to-lymphocyte ratio, systemic immune-inflammation index, monocyte-to-lymphocyte ratio and platelet-to-lymphocyte ratio for predicting the severity of patients with autoimmune encephalitis. Front Neurol. 2025;16:1498007. doi:10.3389/fneur.2025.1498007

25.

Dadouli

Janho

Hatziefthimiou

, et al. Neutrophil-to-lymphocyte, monocyte-to-lymphocyte, platelet-to-lymphocyte ratio and systemic immune-inflammatory index in different states of bipolar disorder. Brain Sci. 2022;12(8):1034. Epub ahead of print 4 August 2022. doi:10.3390/brainsci12081034

26.

von Elm

Altman

Egger

Pocock

Gøtzsche

Vandenbroucke

. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. PLOS Med. 2007;4(10):e296. doi:10.1371/journal.pmed.0040296

27.

Zipf

Chiappa

Porter

Ostchega

Lewis

Dostal

. National health and nutrition examination survey: plan and operations, 1999-2010. Vital Health Stat Ser 1 Programs Collect Proced. 2013;56:1-37.

28.

Sebastian

Wilkinson Enns

Goldman

Steinfeldt

Martin

Moshfegh

. Flavonoid Values for USDA Survey Foods and Beverages 2007–2010. Beltsville, MD: U.S. Department of Agriculture, Agricultural Research Service, Food Surveys Research Group; 2016. http://www.ars.usda.gov/nea/bhnrc/fsrg

29.

Flavonoid Values for USDA Survey Foods and Beverages 2017–2018. U.S. Department of Agriculture, Agricultural Research Service, Food Surveys Research Group; 2022. http://www.ars.usda.gov/nea/bhnrc/fsrg

30.

NHANES Tutorials - NHANES Dietary Analyses Module. Accessed February 23, 2026. https://wwwn.cdc.gov/nchs/nhanes/tutorials/dietaryanalyses.aspx

31.

Wang

Liang

Jiang

, et al. Systemic inflammation as a mediator in the link between obesity and depression: evidence from a nationwide cohort study. BMC Psychiatry. 2025;25(1):449. doi:10.1186/s12888-025-06892-3

32.

, et al. The relationship between systemic inflammation index, systemic immune-inflammatory index, and inflammatory prognostic index and 90-day outcomes in acute ischemic stroke patients treated with intravenous thrombolysis. J Neuroinflammation. 2023;20(1):220. doi:10.1186/s12974-023-02890-y

33.

Huang

Sun

Xie

Yin

. Systemic immune-inflammation index as a versatile biomarker in autoimmune disorders: insights from rheumatoid arthritis, lupus, and spondyloarthritis. Front Immunol. 2025;16. Epub ahead of print 7 August 2025. doi:10.3389/fimmu.2025.1621209

34.

Wang

Lin

Zhang

, et al. Systemic immune-inflammatory complex index as a novel predictor of sepsis prognosis: a retrospective cohort study using MIMIC-IV. Front Med. 2025;12. Epub ahead of print 30 June 2025. doi:10.3389/fmed.2025.1608619

35.

NHANES 2017-2018 Procedure Manuals. Accessed February 23, 2026. https://wwwn.cdc.gov/nchs/nhanes/continuousnhanes/manuals.aspx?BeginYear=2017

36.

Piercy

Troiano

. Physical activity guidelines for Americans from the US Department of Health and Human Services. Circ Cardiovasc Qual Outcomes. 2018;11(11):e005263. doi:10.1161/CIRCOUTCOMES.118.005263

37.

Uzun

Fenercioglu

Senyigit

Can

. Relationship between nutritional status and systemic immune–inflammation indices across BMI categories. Nutrients. 2025;17(23):3799. doi:10.3390/nu17233799

38.

NHANES Survey Methods and Analytic Guidelines. Accessed February 23, 2026. https://wwwn.cdc.gov/nchs/nhanes/analyticguidelines.aspx#sample-design

39.

Daina

Michielin

Zoete

. Swisstargetprediction: updated data and new features for efficient prediction of protein targets of small molecules. Nucleic Acids Res. 2019;47(W1):W357-W364. doi:10.1093/nar/gkz382

40.

Wang

Shen

Wang

, et al. Pharmmapper 2017 update: a web server for potential drug target identification with a comprehensive target pharmacophore database. Nucleic Acids Res. 2017;45(W1):W356-W360. doi:10.1093/nar/gkx374

41.

The UniProt Consortium, Bateman

Martin

M-J

, et al. Uniprot: the universal protein knowledgebase in 2025. Nucleic Acids Res. 2025;53(D1):D609-D617. doi:10.1093/nar/gkae1010

42.

Oliveros

. Venny. An interactive tool for comparing lists with Venn’s diagrams; 2007. Accessed January 28, 2026. https://bioinfogp.cnb.csic.es/tools/venny/

43.

Szklarczyk

Kirsch

Koutrouli

, et al. The STRING database in 2023: protein–protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res. 2023;51(D1):D638-D646. doi:10.1093/nar/gkac1000

44.

Shannon

Markiel

Ozier

, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498-2504. doi:10.1101/gr.1239303

45.

Sherman

Hao

Qiu

, et al. DAVID: a web server for functional enrichment analysis and functional annotation of gene lists (2021 update). Nucleic Acids Res. 2022;50(W1):W216-W221. doi:10.1093/nar/gkac194

46.

Burley

Bhikadiya

, et al. RCSB Protein Data Bank (RCSB.org): delivery of experimentally-determined PDB structures alongside one million computed structure models of proteins from artificial intelligence/machine learning. Nucleic Acids Res. 2023;51(D1):D488-D508. doi:10.1093/nar/gkac1077

47.

Morris

Huey

Lindstrom

, et al. Autodock4 and AutoDockTools4: automated docking with selective receptor flexibility. J Comput Chem. 2009;30(16):2785-2791. doi:10.1002/jcc.21256

48.

Cosconati

Forli

Perryman

Harris

Goodsell

Olson

. Virtual screening with AutoDock: theory and practice. Expert Opin Drug Discov. 2010;5(6):597-607. doi:10.1517/17460441.2010.484460

49.

Olivas-Aguirre

Rodrigo-García

Martínez-Ruiz

NDR

, et al. Cyanidin-3-O-glucoside: physical-chemistry, foodomics and health effects. Molecules. 2016;21(9):1264. doi:10.3390/molecules21091264

50.

Kowalczyk

Muskała

Merecz-Sadowska

Sikora

Picot

Sitarek

. Anti-inflammatory and anticancer effects of anthocyanins in in vitro and in vivo studies. Antioxidants. 2024;13(9):1143. doi:10.3390/antiox13091143

51.

Xue

Tan

Mao

, et al. Polyphenol-based functional materials: structural insights, composite strategies, and biomedical applications. Adv Sci. 2025;12(39):e08924. doi:10.1002/advs.202508924

52.

Akter

Gul

Mumtaz

. Revisiting curcumin in cancer therapy: recent insights into molecular mechanisms, nanoformulations, and synergistic combinations. Curr Issues Mol Biol. 2025;47(9):716. Epub ahead of print 3 September 2025. doi:10.3390/cimb47090716

53.

Tian

Tan

Chen

, et al. Metabolism of anthocyanins and consequent effects on the gut microbiota. Crit Rev Food Sci Nutr. 2019;59(6):982-991. doi:10.1080/10408398.2018.1533517

54.

Xiong

H-H

Lin

S-Y

Chen

L-L

Ouyang

K-H

Wang

W-J

. The interaction between flavonoids and intestinal microbes: a review. Foods. 2023;12(2):320. doi:10.3390/foods12020320

55.

Sandoval-Ramírez

Catalán

Fernández-Castillejo

Pedret

Llauradó

Solà

. Cyanidin-3-glucoside as a possible biomarker of anthocyanin-rich berry intake in body fluids of healthy humans: a systematic review of clinical trials. Nutr Rev. 2020;78(7):597-610. doi:10.1093/nutrit/nuz083

56.

Rana

Mumtaz

. Prunin: an emerging anticancer flavonoid. Int J Mol Sci. 2025;26(6):2678. Epub ahead of print 16 March 2025. doi:10.3390/ijms26062678

57.

Spencer

JPE

Kuhnle

GGC

Williams

Rice-Evans

. Intracellular metabolism and bioactivity of quercetin and its in vivo metabolites. Biochem J. 2003;372(1):173-181. doi:10.1042/bj20021972

58.

Dabeek

Marra

. Dietary Quercetin and Kaempferol: bioavailability and potential cardiovascular-related bioactivity in humans. Nutrients. 2019;11(10):2288. Epub ahead of print 25 September 2019. doi:10.3390/nu11102288

59.

Docampo

Olubu

Wang

Pasinetti

Dixon

. Glucuronidated flavonoids in neurological protection: structural analysis and approaches for chemical and biological synthesis. J Agric Food Chem. 2017;65(35):7607-7623. doi:10.1021/acs.jafc.7b02633

60.

Bartholomé

Haenen

Hollman

, et al. Deconjugation kinetics of glucuronidated phase II flavonoid metabolites by beta-glucuronidase from neutrophils. Drug Metab Pharmacokinet. 2010;25(4):379-387. doi:10.2133/dmpk.DMPK-10-RG-002

61.

Kawai

. β-Glucuronidase activity and mitochondrial dysfunction: the sites where flavonoid glucuronides act as anti-inflammatory agents. J Clin Biochem Nutr. 2014;54(3):145-150. doi:10.3164/jcbn.14-9

62.

Hollman

van Trijp

Buysman

, et al. Relative bioavailability of the antioxidant flavonoid quercetin from various foods in man. FEBS Lett. 1997;418(1-2):152-156. doi:10.1016/S0014-5793(97)01367-7

63.

Manach

Morand

Demigné

, et al. Bioavailability of rutin and quercetin in rats. FEBS Lett. 1997;409(1):12-16. doi:10.1016/S0014-5793(97)00467-5

64.

Biswas

Kaium

Islam Tareq

, et al. The experimental significance of isorhamnetin as an effective therapeutic option for cancer: a comprehensive analysis. Biomed Pharmacother. 2024;176:116860. doi:10.1016/j.biopha.2024.116860

65.

Zabela

Sampath

Oufir

Moradi-Afrapoli

Butterweck

Hamburger

. Pharmacokinetics of dietary kaempferol and its metabolite 4-hydroxyphenylacetic acid in rats. Fitoterapia. 2016;115:189-197. doi:10.1016/j.fitote.2016.10.008

66.

Wang

Chen

Zhu

, et al. Metabolic disposition of luteolin is mediated by the interplay of UDP-glucuronosyltransferases and catechol-O-methyltransferases in rats. Drug Metab Dispos. 2017;45(3):306-315. doi:10.1124/dmd.116.073619

67.

Vieux

Maillot

Rehm

Drewnowski

. Flavonoid intakes in the US diet are linked to higher socioeconomic status and to tea consumption: analyses of NHANES 2011-16 data. J Nutr. 2020;150(8):2147-2155. doi:10.1093/jn/nxaa145

68.

Franceschi

Garagnani

Parini

Giuliani

Santoro

. Inflammaging: a new immune-metabolic viewpoint for age-related diseases. Nat Rev Endocrinol. 2018;14(10):576-590. doi:10.1038/s41574-018-0059-4

69.

Wang

Bao

Wang

Xie

. Association of dietary quality and dietary inflammatory potential with inflammatory markers: evidence from the national health and nutrition examination survey 2009-2018. Front Immunol. 2025;16:1596806. doi:10.3389/fimmu.2025.1596806

70.

Kent

Larsen-Truong

Fleming

, et al. “I always buy the purple ones … if I see them”: socioecological factors influencing anthocyanin-rich food consumption for cognitive health in older adults. Nutrients. 2023;15(5):1194. doi:10.3390/nu15051194

71.

Kuczmarski

Adams

Cotugna

, et al. Health literacy and education predict nutrient quality of diet of socioeconomically diverse, urban adults. J Epidemiol Prev Med. 2016;2(1):13000115.

72.

Yuan

Shen

Zhu

. Association between systemic immune-inflammation index and chronic obstructive pulmonary disease: a population-based study. BMC Pulm Med. 2023;23(1):295. doi:10.1186/s12890-023-02583-5

73.

Zhu

Hattori

, et al. Deactivation of phosphatidylinositol 3,4,5-trisphosphate/Akt signaling mediates neutrophil spontaneous death. Proc Natl Acad Sci U S A. 2006;103(40):14836-14841. doi:10.1073/pnas.0605722103

74.

Papaconstantinou

. The role of signaling pathways of inflammation and oxidative stress in development of senescence and aging phenotypes in cardiovascular disease. Cells. 2019;8(11):1383. doi:10.3390/cells8111383

75.

Storz

. Forkhead homeobox type O transcription factors in the responses to oxidative stress. Antioxid Redox Signal. 2011;14(4):593-605. doi:10.1089/ars.2010.3405

76.

Hedrick

Michelini

Doedens

Goldrath

Stone

. FOXO Transcription factors throughout T cell biology. Nat Rev Immunol. 2012;12(9):649-661. doi:10.1038/nri3278

77.

Senatus

Schmidt

. The AGE-RAGE axis: implications for age-associated arterial diseases. Front Genet. 2017;8:187. doi:10.3389/fgene.2017.00187

78.

Recabarren-Leiva

Burgos

Hernández

, et al. Effects of the age/rage axis in the platelet activation. Int J Biol Macromol. 2021;166:1149-1161. doi:10.1016/j.ijbiomac.2020.10.270

79.

Yao

Chen

Adili

, et al. Plant-based food cyanidin-3-glucoside modulates human platelet glycoprotein VI signaling and inhibits platelet activation and thrombus formation. J Nutr. 2017;147(10):1917-1925. doi:10.3945/jn.116.245944

The Association Between Dietary Flavonoids and Systemic Inflammation: A Cross-Sectional Analysis and Mechanistic Exploration

Abstract

Introduction

Methods

Results

Conclusion

Keywords

Introduction

Materials and Methods

Study Design

Assessment of Flavonoid Intake

Measurement of Systemic Inflammation

Covariates

Statistical Analysis

Target-Disease Intersection for Cyanidin and Inflammation

Protein–Protein Interaction (PPI) Network Construction and Functional Enrichment

Molecular Docking Validation of Hub Targets

Results

Baseline Characteristics of Participants

Associations of Flavonoid Classes with Systemic Inflammation

Associations of Flavonoid Monomers with Systemic Inflammation

Dose–Response Relationships from Restricted Cubic Spline Analyses

Subgroup Analysis of the Cyanidin–SII Association

Network Pharmacology Analysis of Cyanidin in Inflammation

Protein–Protein Interaction Network and Identification of Hub Targets

Discussion

Conclusion

Footnotes

Abbreviations

ORCID iDs

Ethical Approval

Statement of Informed Consent

Statement of Human and Animal Rights

Author Contributions

Funding

Declaration of Conflicting Interests

Data Availability Statement

References