Clementine: the evidence is about hesperidin, and it disagrees with itself

Almost nothing has been tested on clementines themselves. What gets attached to them is work on hesperidin, the flavanone citrus carries, and the pooled analyses of that compound point in different directions. A 2025 meta-analysis of randomised trials of purified hesperidin, at the doses above 500 mg a day that those trials used, reported lower fasting glucose, triglycerides, total and LDL cholesterol and systolic blood pressure, with no change in insulin, HOMA-IR, HDL or diastolic pressure. Two earlier poolings found nothing. Total cholesterol moved 1.04 mg/dl in 577 people across 10 trials (95% CI -5.65 to 3.57), and fasting glucose moved 1.10 mg/dL in 318 people across 6 trials (95% CI -3.79 to 1.57), both with no heterogeneity between studies. The largest pooling, 16 trials in 845 adults, did find insulin resistance down by 0.43 HOMA-IR units (95% CI -0.82 to -0.03, p = 0.034), and then its own trial sequential analysis concluded there is not yet enough data to call that established. The reviews searched to different dates and admitted different trials, which is why this sits unresolved rather than proven in either direction. Every one of them tested purified hesperidin in a capsule. A clementine carries a small fraction of 500 mg.

The numbers, per 100 g

How much is actually in it

• The flavonoid content of clementine juice varies from batch to batch, and the batch tested in this experiment carried only hesperidin and narirutin while the batch used in the laboratory assay carried several flavonoids. — Clementine juice used in the in vitro assay compared with the juice given to the volunteers Evidence D source
  Laboratory juice contained several flavonoids; the juice actually drunk contained only hesperidin and narirutin

What your body absorbs

• Six weeks of 330 mL of orange juice a day delivered hesperidin into the blood of 37 overweight men and still changed nothing measurable in their blood vessel function. — 37 centrally overweight men in a double-blind three-period crossover, each period 6 weeks: 330 mL orange juice, a matched control drink, or a control drink enriched with 210 mg of hesperidin a day (n = 37) Evidence B source
  Plasma hesperetin metabolite exposure and 24 hour urinary excretion comparable after juice and supplement; no change in fasting or postprandial flow-mediated dilation, metabolic or anthropometric measures

What cooking and storage change

• Whole citrus fruit and citrus juice delivered their flavanones equally well: 20 people given 150 g of orange fruit or 300 g of juice absorbed and excreted the same amount once the dose was matched. — 20 adults in a randomised crossover of a single portion of orange fruit (150 g) versus juice (300 g), with plasma and urine sampling (n = 20 in the crossover, plus 109 more aged 18 to 80 for the juice arm) Evidence B source
  No difference in absorption or excretion of hesperidin or narirutin between the fruit and the juice matrix on a dose-adjusted basis

What it does to your health

• A 2025 meta-analysis of randomised trials found that hesperidin supplements above 500 mg a day for more than 12 weeks lowered fasting glucose, triglycerides, total and LDL cholesterol and systolic blood pressure. — Adults in randomised controlled trials of hesperidin supplementation, searched to August 2023, GRADE assessed (n = meta-analysis of randomised trials, count not given in the abstract) Evidence A source
  Significant reductions in fasting blood sugar, triglycerides, total cholesterol, LDL cholesterol, systolic blood pressure and TNF-alpha; nothing on insulin, HOMA-IR, QUICKI, HDL, diastolic pressure or hs-CRP
• Pooling 6 trials in 318 adults, hesperidin supplements moved fasting glucose by about 1 mg/dL and glycated haemoglobin by 0.04%, neither of which reached significance. — 318 participants across 6 randomised controlled trials of hesperidin supplementation, searched to February 2019 (n = 318 in 6 trials) Evidence A source
  Fasting glucose weighted mean difference -1.10 mg/dL (95% CI -3.79 to 1.57); HbA1c -0.04% (95% CI -0.14 to 0.04); no between-study heterogeneity

Safety, limits and interactions

• Eating 1 kg of clementines a day raised estimated midazolam clearance by 43% in a transplant patient and by 89% in a healthy volunteer, and the authors judged both changes to sit inside normal day to day variation. — Two single case experiments: one renal transplant patient taking tacrolimus who ate 1 kg of clementines a day for 4 days, and one healthy man drinking clementine juice chronically (n = 2) Evidence C source
  Estimated midazolam clearance +43% and +89%, both within the range of intra-individual variability

What people believe that the data does not support

• Citrus does not slim anyone: pooling 11 randomised trials in 512 adults, orange juice or hesperidin changed body weight by 0.31 kg and waist circumference by 0.07 cm, both within the noise. — 512 adults across 11 randomised controlled trials of orange juice or hesperidin supplementation, searched to February 2020 (n = 512 in 11 trials) Evidence A source
  Body weight -0.31 kg (95% CI -1.44 to 0.82); BMI +0.11 kg/m2; waist circumference +0.07 cm; body fat percentage -0.70%, none significant
• The European Union has authorised no health claim for hesperidin at all: the register holds four applications for it and marks every one non-authorised, including the cholesterol and heart claim. — EU register of nutrition and health claims, snapshot of 2026-09-21: 2330 claims, of which 262 authorised and 2067 non-authorised; 4 hesperidin entries, 0 authorised (n = 4 hesperidin claims, 0 authorised) Evidence E source
  Status Non-authorised for POL-HC-7258 (cholesterol and heart health), POL-HC-7257 (bone) and the two diosmin combinations

Who this works differently for

• How much citrus flavanone a person actually absorbs ranges from none at all to 59% of the dose, and older people excrete slightly less hesperetin, while sex, body mass index and the contraceptive pill made no difference. — 129 adults aged 18 to 80 who drank a single portion of orange juice, with 24 hour urinary flavanone measurement (n = 129) Evidence B source
  Urinary yield ranged from 0% to 59% of the dose; hesperetin excretion fell slightly with age (p < 0.05); no association with sex, BMI or contraceptive pill use
• In 36 older adults, a supplement combining hesperidin, diosmin and proanthocyanidins differed significantly from a control group on attention and on fall risk, with no difference between one and two doses a day. — 36 elderly patients randomised into three groups in a crossover pilot, 8 weeks on supplement, 4 weeks washout, 8 weeks on the other dose, assessed to week 20 (n = 36) Evidence B source
  Significant differences from control in cognitive function, particularly attention, and in fall risk; no difference between 1 stick and 2 sticks a day

What is still unknown

• An earlier meta-analysis of 10 randomised trials in 577 people found hesperidin changed total cholesterol by 1 mg/dl, an amount indistinguishable from zero, and moved LDL, HDL and triglycerides no further. — 577 participants across 10 randomised controlled trials of hesperidin supplementation, searched to May 2018 (n = 577 in 10 trials) Evidence A source
  Total cholesterol weighted mean difference -1.04 mg/dl (95% CI -5.65 to 3.57); systolic pressure -0.85 mmHg (95% CI -3.07 to 1.36); no heterogeneity between studies
• The largest hesperidin meta-analysis, 16 trials in 845 adults, found a statistically significant drop in insulin resistance that its own trial sequential analysis then dismissed as not yet established. — 845 participants across 16 randomised controlled trials, databases searched to July 2025 (n = 845 in 16 trials) Evidence A source
  HOMA-IR weighted mean difference -0.43 (95% CI -0.82 to -0.03, p = 0.034) and insulin sensitivity index +0.05 (95% CI 0.01 to 0.08), both overturned by trial sequential analysis

The bottom line

The USDA table here holds a single record for raw clementines, with no second sample to show how much the fruit varies. Two findings are about the fruit as a food rather than the compound. In 20 adults given either 150 g of orange fruit or 300 g of orange juice, absorption and excretion of hesperidin and narirutin were the same once the dose was matched, which was a single portion of orange rather than clementine, with juice measured fresh rather than bought. And the amount that gets in varies enormously between people. Across 129 adults aged 18 to 80 drinking one portion of orange juice, 24 hour urinary yield ran from 0 to 59 percent of the dose, with hesperetin excretion falling slightly with age (p below 0.05) and no association with sex, body mass index or use of the contraceptive pill. Urinary recovery is a proxy for absorption, and part of that spread is gut bacteria that nobody measured. The one trial at a dose a person could drink found nothing. Thirty-seven centrally overweight men took 330 mL of orange juice a day for 6 weeks in a double-blind crossover, hesperidin metabolites reached the blood as well as they did from a 210 mg supplement, and fasting and postprandial flow-mediated dilation, metabolic and body measures all stayed put, in men only, over six weeks, on a surrogate for vascular health. Citrus does not make anyone thinner. Pooling 11 randomised trials in 512 adults, orange juice or hesperidin changed body weight by 0.31 kg (95% CI -1.44 to 0.82) and waist circumference by 0.07 cm, in short trials that measured body dimensions and mostly used juice. The European Union has authorised no hesperidin claim at all. Its register holds four applications, including one for cholesterol and heart health, and marks every one non-authorised, which means the evidence did not satisfy the assessment rather than that an effect was disproved. Clementine as a fruit is absent from that register, which indexes substances. One result sits under who this works differently for, and it should be read with care: in a pilot of 36 older adults, a branded product combining hesperidin, diosmin and proanthocyanidins recorded differences from control on attention and on fall risk, with 12 people per arm, three ingredients that cannot be separated and no difference between one dose and two. On drugs and cautions, the only clementine-specific human work is small. One renal transplant patient on tacrolimus ate 1 kg of clementines a day for 4 days and his estimated midazolam clearance rose 43 percent, and one healthy man drinking clementine juice regularly rose 89 percent, changes the authors placed inside ordinary day to day variation, from two people with no control period. The juice those volunteers drank held only hesperidin and narirutin while the batch used in the laboratory assay held several flavonoids, an observation made in passing with no measured range, so what one batch does may not be what another does. We found nothing in this corpus on clementines in pregnancy, in children, or as an allergen. That silence is a gap in the evidence and not a clearance.

Sources

1. clem-r1-02 · in vitro comparison
   In contrast to the juice used for the in vitro assay comprising several flavonoids, this juice only contained hesperidin and narirutin indicating that the drug interactions potential of clementines might depend on the composition varying from batch to batch.
   Eur J Pharm Sci, 2019 · checked 2026-09-24
2. clem-r1-10 · randomised crossover trial
   Neither intervention significantly affected fasting or postprandial FMD, vascular, metabolic or anthropometric parameters, or oxylipin profiles versus CON.
   Food Funct, 2026 · checked 2026-09-24
3. clem-r1-03 · randomised crossover
   Analysis of the plasma pharmacokinetic and urinary excretion data on a dose-adjusted basis indicated no difference in absorption or excretion of either flavanone between the fruit and juice matrices.
   Br J Nutr, 2009 · checked 2026-09-24
4. clem-r1-05 · meta-analysis of RCTs
   The results of this study demonstrate that hesperidin supplementation had a significant impact on reducing FBS, TG, TC, LDL-C, SBP, and TNF-α. However, there was no significant effect observed on insulin, HOMA-IR, QUICKI, HDL-C, DBP, and hs-CRP.
   Nutr Rev, 2025 · checked 2026-09-24
5. clem-r1-07 · meta-analysis of RCTs
   The results showed that hesperidin had no significant effect on serum fasting blood glucose (weighted mean difference [WMD] = -1.10 mg/dL, 95% CI: -3.79, 1.57), plasma insulin (WMD = -0.01 μU/mL, 95% CI: -1.20, 1.19), glycated haemoglobin A1c (WMD = -0.04%, 95% CI: -0.14, 0.04), homeostasis model assessment for insulin resistance (WMD = 0.117, 95% CI: -0.06, 0.29) and quantitative insulin sensitivity check index (WMD = 0.135; 95% CI: -0.13, 0.39), with no significant between-study heterogeneity.
   Br J Clin Pharmacol, 2020 · checked 2026-09-24
6. clem-r1-01 · single case experiments
   Although an increment of 43% in the estimated midazolam clearance (eCLmet) was observed during the first experiment in a renal transplant patient on tacrolimus after 4-d consumption of clementines (1 kg/d), and an increment of +89% of eCLmet was observed during chronic consumption of clementine juice in a healthy male volunteer, these changes lie within the range of intra-individual variability.
   Eur J Pharm Sci, 2019 · checked 2026-09-24
7. clem-r1-09 · meta-analysis of RCTs
   The meta-analysis showed that orange juice and/or hesperidin had no significant effect on body weight [weighted mean difference (WMD) = -0.31 kg; 95% CI, -1.44 to 0.82, P = 0.592], body mass index [WMD = 0.11 kg/m2; 95% CI, -0.33 to 0.56, P = 0.617], fat mass [WMD = 0.14 kg; 95% CI, -0.95 to 1.22, P = 0.807], percentage body fat [WMD = -0.70%; 95% CI, -2.58 to 1.18, P = 0.463] and waist circumference [WMD = 0.07 cm; 95% CI, -1.16 to 1.30, P = 0.588].
   Clin Nutr ESPEN, 2021 · checked 2026-09-24
8. clem-r1-11 · regulatory register entry
   POL-HC-7258 Hesperidin (a component of citrus peel extract and precursor of hesperitin) (ingredient not found in the spanish food laws) Helps maintain normal blood cholesterol levels/Supports heart health (to be evaluated by EFSA). Non-authorised
   EU Register on nutrition and health claims, claim POL-HC-7258, snapshot 2026-09-21 · checked 2026-09-24
9. clem-r1-04 · randomised crossover
   In the extended urinary excretion dataset the individual variation was very large (range 0-59 % urinary yield). There was a small but significant (P < 0.05) decrease in the excretion of hesperetin (but not naringenin) with increasing age (P < 0.05), but the effects of sex, BMI and contraceptive pill use were shown not to be associated with the variation in flavanone excretion.
   Br J Nutr, 2009 · checked 2026-09-24
10. clem-r1-12 · randomised pilot trial
    Both experimental groups showed statistically significant differences from the control group in cognitive function, particularly attention, and fall risk. No domains showed statistically significant differences between the two experimental groups.
    J Am Nutr Assoc, 2026 · checked 2026-09-24
11. clem-r1-06 · meta-analysis of RCTs
    The meta-analysis revealed that hesperidin supplementation had no effect on serum total cholesterol (weighted mean difference [WMD] = -1.04 mg/dl; 95% confidence interval [CI]: -5.65, 3.57), low-density lipoprotein cholesterol (WMD = -1.96 mg/dl; 95% CI [-7.56, 3.64]), high-density lipoprotein cholesterol (WMD = 0.16 mg/dl; 95% CI [-1.94, 2.28]), and triglyceride (WMD = 0.69 mg/dl; 95% CI [-5.91, 7.30]), with no significant between-study heterogeneity.
    Phytother Res, 2019 · checked 2026-09-24
12. clem-r1-08 · meta-analysis of RCTs
    We found that hesperidin supplementation significantly reduced the homeostatic model assessment of insulin resistance (HOMA-IR) (WMD: -0.43, 95%CI: -0.82, -0.03; p = 0.034) and increased the quantitative insulin sensitivity check index (WMD: 0.05, 95%CI: 0.01, 0.08; p = 0.005)...However, subsequent trial sequential analysis revealed that hesperidin confers no statistically significant improvement in HOMA-IR.
    Front Nutr, 2025 · checked 2026-09-24