Hesperidin cuts CRP by 0.43 mg/L in overweight adults — but only supplements can reach the effective dose

A meta-analysis of 16 randomized controlled trials found that supplementing with hesperidin — a flavonoid concentrated in citrus peel and orange juice — reduced C-reactive protein (CRP) by −0.43 mg/L (95% CI −0.68 to −0.17, p = 0.001) in overweight or obese adults compared with placebo. 1 The authors rated this finding as moderate-certainty evidence under the GRADE framework — the most reliable signal in the analysis.

Two other inflammatory markers moved in the same direction: tumor necrosis factor-alpha (TNF-α) fell by −2.84 pg/mL (95% CI −4.81 to −0.86, p = 0.005) and the endothelial adhesion marker VCAM-1 dropped by −27.16 ng/mL (95% CI −50.13 to −4.19, p = 0.020). 1 Both, however, carried very-low GRADE certainty — the confidence intervals are wide, and the true effect could differ substantially from the point estimates. IL-6, total antioxidant capacity (TAC), and ICAM-1 showed no significant effect.

The CRP reduction is modest by clinical standards: a 0.43 mg/L drop sits well below the 2–3 mg/L threshold that separates low from elevated cardiovascular risk on standard hsCRP panels. But among the 16 trials pooled, it was consistent enough to clear a moderate-certainty threshold, and it emerged specifically in a population where baseline inflammation tends to run higher than in normal-weight adults.

The CRP signal here is new. A 2023 meta-analysis by Khorasanian et al., which pooled 13 RCTs across general adult populations, found no significant CRP reduction: WMD −0.01 mg/L (95% CI −0.05 to 0.03), with substantial heterogeneity (I² = 66.9%). 2 Zhu et al. pooled three more trials (16 vs. 13) and restricted inclusion to overweight and obese participants. That restriction likely explains the divergence: people with excess body fat carry chronically elevated adipose-derived cytokines, giving anti-inflammatory interventions more room to move the CRP needle than in metabolically healthier populations.

What did hold across both analyses was the TNF-α signal. Khorasanian et al. also found a significant TNF-α reduction (WMD −2.74 pg/mL, I² = 82.1%), and Zhu 2026 replicated the direction. 2 The high heterogeneity in both analyses means this effect is not uniform across trials — some trials show strong TNF-α responses, others do not. Until subgroup data from the full Zhu 2026 manuscript become available, it's not possible to identify which dosing or population characteristics drive the variance.

Where the 2026 paper breaks new ground: it adds VCAM-1 as an outcome variable. VCAM-1 (vascular cell adhesion molecule-1) is an endothelial surface protein whose elevation reflects early vascular inflammation and is associated with atherosclerosis progression. A −27.16 ng/mL reduction is biologically plausible given hesperidin's known capacity to inhibit NF-κB signaling, but very-low GRADE certainty means it should be treated as a preliminary signal rather than a reliable clinical estimate.

Zhu et al. searched PubMed, Web of Science, Embase, Scopus, and the Cochrane Library from inception through April 4, 2026, with two authors independently screening results. 1 The final dataset included 16 RCTs with 845 participants, all overweight or obese based on BMI criteria (exact thresholds not specified in the abstract). Interventions covered hesperidin in its various forms — purified hesperidin capsules, glucosyl-hesperidin, and hesperidin-containing citrus preparations — compared against placebo or control.

Quality assessment followed the Cochrane Risk of Bias 2 (RoB 2) tool for individual trial quality; evidence certainty was graded using the GRADE approach. Continuous outcomes were pooled as mean differences (MD) with 95% confidence intervals, except TAC, which was pooled as standardized mean differences (Hedges' g) because the assay methods differed across studies.

What the abstract does not report — and what will be critical to a full appraisal — includes per-study sample sizes and durations, dose ranges across included trials, I² values for each outcome (heterogeneity across the 16 studies), and any subgroup analyses by dose, formulation, or BMI category. The authors state explicitly: "Marked between-study differences in intervention matrix, design structure, and reporting methods preclude dose- or matrix-specific recommendations at present." 1 That caveat means the pooled effect estimate should be read as an average signal across heterogeneous trials, not a precise prescription.

Hesperidin is the primary flavanone in sweet oranges and tangerines. It concentrates in the white pith (albedo) and the peel — the parts most commonly discarded — with juice containing considerably less than the intact fruit.

Based on data from the Phenol-Explorer polyphenol composition database, orange juice hesperidin content varies substantially by processing: 3

The clinical trials included in meta-analyses of hesperidin have used doses ranging from approximately 292 to 1,000 mg/day of purified hesperidin, with effective doses generally above 500 mg/day based on the Khorasanian 2023 dose-response analysis. 2 At 500 mg, a person drinking orange juice from concentrate would need to consume approximately 950 ml (just over four 8-oz glasses) to approach that threshold — roughly 400–440 calories from juice alone, with the accompanying sugar load. Fresh-squeezed juice would require closer to 1,900 ml for the same hesperidin dose.

Whole citrus fruit delivers more than juice because hesperidin resides primarily in the peel and pith. Dried tangerine peel (chenpi in traditional Chinese medicine) is among the most concentrated dietary sources. But even eating whole oranges regularly falls short of what trials used unless consumption is substantial and consistent.

The practical implication is direct: food-first strategies cannot reliably replicate the hesperidin doses studied in RCTs. Supplementation — purified hesperidin capsules or the enhanced-bioavailability glucosyl-hesperidin form — is the only feasible route to the dose range where the CRP-lowering signal appears. This does not mean food sources are without value; they deliver hesperidin alongside other citrus flavonoids (narirutin, eriocitrin, naringenin) whose combined effects are not captured in single-compound supplementation trials. But the dose math does not favor food as the primary vehicle for a therapeutic anti-inflammatory effect.

Accepted, not yet published. The full text, supplementary materials, individual forest plots, and complete methods section are unavailable. COI declarations and funding sources for the paper itself have not been disclosed. The caveat about "marked between-study differences" in the abstract suggests heterogeneity that the pooled point estimates alone do not capture.

Population specificity. The entire analysis was conducted in overweight or obese adults. The CRP-lowering finding should not be generalized to normal-weight individuals, where the 2023 broader-population meta-analysis found no effect. 2

Intervention heterogeneity. Trials used different forms of hesperidin (pure compound, citrus extract, combined formulations), different doses, and presumably different durations. Dose- and formulation-specific conclusions cannot be drawn from the abstract-level data available now.

Effect size in clinical context. A −0.43 mg/L CRP reduction is statistically significant but clinically modest. Standard risk stratification thresholds for hsCRP (low risk: <1 mg/L; intermediate: 1–3 mg/L; high: >3 mg/L) suggest the absolute CRP shifts from hesperidin supplementation are unlikely to move most patients from one risk category to another on their own. The effect is more consistent with modest adjunctive anti-inflammatory support than with a primary therapeutic intervention.

GRADE downgrades for secondary outcomes. Both TNF-α and VCAM-1 achieved statistical significance but were rated very-low certainty — meaning the evidence base for these outcomes is sufficiently limited that the true effect could be "substantially different from the estimate of effect." 1 These findings are directionally interesting, not clinically established.

Safety. Hesperidin is generally well-tolerated in the trial literature, with no serious adverse events reported across prior meta-analyses. The Norwegian Scientific Committee for Food Safety (VKM) found no identified exposure-related adverse effects at supplement doses used in trials. Health Canada has noted a potential bleeding risk with diosmin-plus-hesperidin combination products, though whether that applies to hesperidin alone is unclear. Standard cautions apply for pregnancy, breastfeeding, and individuals with bleeding disorders.

The Zhu 2026 meta-analysis adds a moderately reliable signal to a previously null finding: hesperidin supplementation reduces CRP in overweight and obese adults, and it does so with consistent direction across 16 RCTs. The GRADE upgrade from the prior null result is meaningful — but the information gaps from the still-unpublished full text, the modest absolute effect size, and the explicit authorial caution about formulation heterogeneity all argue for measured clinical use rather than confident recommendation.

For health-conscious adults with excess body weight and elevated hsCRP: hesperidin supplementation at 500 mg/day or higher is the only route likely to approach the doses used in clinical trials. Standard citrus intake — even two to three glasses of orange juice daily — delivers a fraction of that threshold and comes with substantial added-sugar costs. If you or a patient are considering hesperidin as part of a dietary inflammation-reduction strategy, a glucosyl-hesperidin formulation (which shows improved absorption versus standard hesperidin) taken with meals is the most evidence-adjacent option available. This is adjunctive to, not a substitute for, established anti-inflammatory dietary patterns (weight loss, reduced ultra-processed food, increased omega-3 intake) where the effect sizes and evidence certainty are larger.

For registered dietitians: the most useful takeaway from this paper may be the population specificity rather than the compound itself. The CRP signal emerged in overweight/obese adults and was absent in general-population meta-analyses — consistent with the hypothesis that adiposity-driven baseline inflammation is the condition where citrus flavonoids find traction. If counseling a client with BMI > 25 and elevated hsCRP who asks about anti-inflammatory supplements, hesperidin at 500–1,000 mg/day is now backed by 16 RCTs with a moderate-certainty pooled effect, making it a more defensible discussion than many widely marketed options. The authors themselves call for "further high-quality randomized trials with longer follow-up, standardized biomarker assessment, and better characterization of formulations" — 1 which means the evidence base is real but still evolving. When the full text and supplementary materials are published, the dose-subgroup data will clarify whether 300 mg/day is sufficient or whether the effect is concentrated above 600–1,000 mg/day.

One concrete action today: For overweight or obese adults with hsCRP above 1 mg/L who want to use hesperidin as an adjunctive supplement, the current best estimate from 16 RCTs is: 500–1,000 mg/day of hesperidin (or glucosyl-hesperidin for better bioavailability), taken consistently for at least six to eight weeks before re-checking CRP. Eating more whole citrus fruit — particularly the white pith and segments rather than only juice — adds modest additional hesperidin alongside fiber and other flavonoids that may contribute independently to inflammation reduction. Both approaches are low-risk, but only supplementation is likely to reach the dose range where the meta-analytic signal lives.