Do carbs cause inflammation?

Not in any way the trials can find for the amount you eat. A 2026 meta-analysis of 25 randomized trials in 2,222 adults compared low-carbohydrate diets with low-fat diets over a median of 24 weeks and found no difference in CRP, IL-6 or TNF-alpha, the standard blood markers of inflammation. The type of carbohydrate matters a little more. More fiber and low glycemic index foods lowered CRP by 0.37 to 0.41 mg/L in trials, and mixed food sources that contained sugary drinks raised it.

Myth. The idea that carbohydrate as such drives inflammation does not hold up in randomized trials. Doubling the share of energy from carbohydrate, from about 26% on a low-carbohydrate diet to about 54% on a low-fat one, left the inflammatory markers where they were.

What the trials found

Cutting carbohydrate

The 2026 meta-analysis is the most direct test of the question, and its trials were mostly at low risk of bias. The standardized difference for CRP was minus 0.01 (95% CI minus 0.11 to 0.09), and IL-6 and TNF-alpha were just as flat.

Comparisons against other control diets look slightly better for cutting carbohydrate and are less secure. A 2025 meta-analysis of 60 trials in 5,511 adults found a small CRP fall with low-carbohydrate diets that did not hold up. With one trial left out, it was 0.14 mg/L (95% CI 0.28 lower to 0.00) and was no longer significant. Larger falls came in people with a BMI over 35 or a high starting CRP, which may reflect weight loss more than carbohydrate. A second 2025 meta-analysis, of 174 trials in 11,481 adults, found carbohydrate-restricted diets lowered CRP by 0.37 mg/L and TNF-alpha by 0.27 pg/mL, again mixed up with weight loss. Across 44 trials of ketogenic diets, TNF-alpha fell by 0.32 pg/mL and IL-6 by 0.27 pg/mL, with no significant effect on CRP. Nobody knows whether shifts that small matter for health.

The type of carbohydrate

In 29 trial comparisons in 1,617 people with diabetes, low glycemic index or load diets lowered CRP by 0.41 mg/L (95% CI 0.04 to 0.78), a change the authors themselves called trivial. A 2013 meta-analysis of 14 trials in obesity found a larger fall of 0.43, in units the abstract prints as mg/dl, which is probably a misprint for mg/L and which we could not check against the full text. A 2019 meta-analysis in type 2 and gestational diabetes found no CRP difference between low and higher glycemic index diets and a significant difference in IL-6, without giving its size.

Fiber does about as much. In 14 trials in adults with overweight or obesity, fiber or fiber-rich food lowered CRP by 0.37 mg/L, with an interval running from 0.74 lower to zero. The fall appeared when fiber intake was 8 g a day above the control diet.

Sugar

A 2022 meta-analysis of 64 controlled feeding trials in 4,094 adults found that fructose-containing sugars did not raise CRP when calories were matched. The source mattered. Mixed sources that contained sugary drinks raised CRP, and fruit, 100% fruit juice, sweetened soy drink and dark chocolate lowered it, on moderate to low certainty evidence over a median of 6 weeks. Which sugar it is made no difference. Across 13 intervention studies in 1,141 people, fructose and glucose diets did not differ for CRP, on low quality evidence. In a crossover trial of 24 adults drinking four sweetened drinks a day for 8 days, fructose, high-fructose corn syrup and glucose left CRP and IL-6 the same. In a 2026 crossover trial of 39 healthy adults, 66 g of sucrose a day for 14 days did not change CRP differently from drinks sweetened with saccharin or stevia. Both trials were short and small. Sugar has its own page, sugar and inflammation.

Why the observational studies say more

A 2014 systematic review set the two kinds of evidence side by side for adults without inflammatory disease. For fiber, 13 of 16 observational studies found a link with lower inflammatory markers, against 1 of 11 trials. For whole grain the count was 6 of 7 against 1 of 7. The reviewers concluded that the observational benefit may reflect confounding. People who eat more fiber and whole grain differ in many other ways.

Unsettled. A 2025 meta-analysis of 45 observational studies linked high added sugar intake with 66% higher odds of Crohn's disease (odds ratio 1.66, 95% CI 1.21 to 2.29) and 59% higher odds of ulcerative colitis. Most of those studies were retrospective and asked people with the disease to recall their diet, which is prone to recall bias. The link is unsettled for that reason, and it gives no absolute risk.

For the rest of what is known about carbohydrates, see the carbohydrate page.

Who should be careful

Not for everyone. Cutting carbohydrate hard in the hope of lowering inflammation has a price in blood lipids. In the meta-analysis of 174 trials, carbohydrate-restricted diets raised LDL cholesterol by 4.81 mg/dL (95% CI 2.58 to 7.05), ketogenic diets raised LDL and total cholesterol the most, and lean mass fell. The authors advise monitoring.

That is the one caution in this evidence. Our search found no evidence on carbohydrates and inflammation in pregnancy, in children or alongside medicines. The 2019 glycemic index analysis included women with gestational diabetes, but its abstract reports the CRP result for type 2 diabetes only.

What expert bodies say

We found no agency that addresses carbohydrates and inflammation directly. The closest is a 2024 evidence map by the US Department of Veterans Affairs on anti-inflammatory diets. It says some evidence suggests that simple carbohydrates, like saturated fat, may promote inflammation, and that the mechanisms are not understood. It offers no pooled effect of its own, and its wording is "may promote". The WHO and European sugar opinions we checked are about dental caries and body weight, and they do not address inflammation.

How we searched

What would change this answer

• Trials that hold body weight steady while changing carbohydrate. The larger CRP falls came with weight loss, and the current trials cannot pull the two apart.
• Trials with outcomes people feel, such as flares of inflammatory disease, in place of blood markers whose shifts of a few tenths of a mg/L have no known clinical meaning.
• Prospective cohorts of added sugar and inflammatory bowel disease that record diet before diagnosis and report absolute risk.

Sources

1. ev-crbi-01 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 2222
   No significant differences were found between LC and LF diets for interleukin-6 (SMD: -0.01; 95% CI: -0.12 to 0.09), C-reactive protein (SMD: -0.01; 95% CI: -0.11 to 0.09), or tumor necrosis factor-alpha (SMD: -0.03; 95% CI: -0.18 to 0.11), nor for adiponectin, leptin, or resistin.
   
   Who: adults in RCTs of low-carbohydrate (about 26% of energy) versus low-fat (about 54% carbohydrate) diets; median 24 weeks
   Effect: CRP SMD -0.01 (95% CI -0.11 to 0.09); IL-6 SMD -0.01 (-0.12 to 0.09); TNF-alpha SMD -0.03 (-0.18 to 0.11); risk of bias mostly low
   Certainty: The most direct answer: doubling the carbohydrate share does not change inflammation markers compared with a low-fat diet.
   Nutr Res, 2026 · checked 2026-10-07 · we read the abstract
2. ev-crbi-02 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 5511
   Sensitivity analysis by omitting each study one at a time showed the meta‐analysis result was influenced by Abbaspour Rad et al. (Rad et al. 2023) study, which changed the significance of the results (MD = −0.14; 95% CI: −0.28 to 0.00; p = 0.052) while the heterogeneity remained unchanged (Cochrane's Q‐test, p = 0.002, I 2 = 37.2%).
   
   Who: adults in RCTs of low-carbohydrate versus control diets; larger falls with BMI over 35 or baseline CRP over 4.5 mg/L
   Effect: CRP with one trial omitted: MD -0.14 mg/L (95% CI -0.28 to 0.00; p = 0.052), not significant
   Certainty: Fragile overall effect; a larger effect in people with obesity may reflect weight loss, not the carbohydrates themselves.
   Food Science & Nutrition, 2025, Effect of Low-Carbohydrate Diets on C-Reactive Protein Level in Adults: A Systematic Review and Meta-Analysis of Randomized Controlled Trials · checked 2026-10-07 · we read the fulltext
3. ev-crbi-03 · Meta-analysis or systematic review · meta-analysis of randomized trials · n = 11481
   For cardiovascular health, compared with higher-carbohydrate diets, CRDs significantly reduced blood pressure: SBP (–2.05 mmHg) and DBP (–1.26 mmHg); lipid profiles: TG (–15.11 mg/dL), LDL–HDL ratio (–0.18), TG-HDL ratio (–0.32), VLDL (–3.34 mg/dL), and ApoB-ApoA1 ratio (–0.03); and other endothelial and inflammatory markers: E-selectin (–2.23 ng/mL), CRP (–0.37 mg/L), and TNF-α (–0.27 pg/mL).
   
   Who: adults in RCTs comparing diets with 45% or less of energy from carbohydrate against higher-carbohydrate diets, 27 countries
   Effect: CRP -0.37 mg/L; TNF-alpha -0.27 pg/mL; systolic BP -2.05 mmHg; triglycerides -15.11 mg/dL
   Certainty: Effects are mixed with weight loss; larger effect in women and people with overweight.
   Am J Clin Nutr, 2025 · checked 2026-10-07 · we read the fulltext
4. ev-crbi-04 · Meta-analysis or systematic review · meta-analysis of randomized trials · n = 11481
   Moderate-carbohydrate diets offered balanced benefits, whereas ketogenic diets produced greater weight loss but greater increases in LDL and total cholesterol.
   
   Who: adults in RCTs of carbohydrate-restricted diets, including ketogenic diets
   Effect: LDL +4.81 mg/dL (95% CI 2.58 to 7.05); total cholesterol +4.32 mg/dL; larger rises on ketogenic diets; lean mass loss
   Certainty: Safety: sharp carbohydrate restriction for an "anti-inflammatory" effect has a cost for lipids; the authors advise monitoring.
   Am J Clin Nutr, 2025 · checked 2026-10-07 · we read the abstract
5. ev-crbi-05 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 44 trials
   The pooled findings showed that a KD has an effect on lowering TNF-α (WMD: -0.32 pg/mL; 95% CI: -0.55, -0.09; P = 0.007) and IL-6 (WMD: -0.27 pg/mL; 95% CI: -0.52, -0.02; P = 0.036) compared with control groups.
   
   Who: participants in RCTs of ketogenic diets versus control diets
   Effect: TNF-alpha WMD -0.32 pg/mL (-0.55 to -0.09); IL-6 WMD -0.27 pg/mL (-0.52 to -0.02); CRP, IL-8, IL-10 not significant
   Certainty: Small cytokine shifts; clinical significance unknown.
   Nutr Rev, 2025 · checked 2026-10-07 · we read the abstract
6. ev-crbi-06 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 14 trials
   Weighted mean differences in change of C-reactive protein [WMD: -0.43 mg/dl, (95% CI -0.78 to -0.09), p = 0.01], and fasting insulin [WMD: -5.16 pmol/L, (95% CI -8.45 to -1.88), p = 0.002] were significantly more pronounced in benefit of low GI/GL diets.
   
   Who: RCTs of low versus high GI/GL diets in obesity management; effect held after excluding type 2 diabetes
   Effect: CRP WMD -0.43 mg/dl (95% CI -0.78 to -0.09); fasting insulin -5.16 pmol/L; fat-free mass -2.3 lb (-1.04 kg)
   Certainty: CRP units in the abstract are mg/dL, as printed; probable typo (mg/L).
   Nutr Metab Cardiovasc Dis, 2013 · checked 2026-10-07 · we read the abstract
7. ev-crbi-07 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 1617
   trivial reductions in fasting blood glucose (−0.36 mmol/L (−0.42 to −0.19), P<0.001; substantial heterogeneity; I2=54%, P<0.001) and CRP (−0.41 mg/L (−0.78 to −0.04), P=0.03; no substantial heterogeneity, I2=24%, P=0.26).
   
   Who: adults with type 1 or type 2 diabetes, mostly middle-aged and overweight, on drugs or insulin; trials of 3 weeks or more
   Effect: CRP MD -0.41 mg/L (95% CI -0.78 to -0.04); HbA1c -0.31%; moderate certainty for most secondary outcomes
   Certainty: Review for EASD guidelines; effect on CRP is small.
   BMJ, 2021 · checked 2026-10-07 · we read the fulltext
8. ev-crbi-08 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 10 trials
   The results of the meta-analysis did not show significant differences (p > 0.05) between low GI and higher GI diets with respect to adiponectin and C-reactive proteins in patients with type 2 diabetes, but a significant difference (p < 0.001) was observed between the two groups in relation to interleukin-6.
   
   Who: patients with type 2 diabetes and women with gestational diabetes; 10 RCTs reviewed, 9 pooled
   Effect: CRP and adiponectin p > 0.05; IL-6 p < 0.001 in favor of low GI
   Certainty: The abstract does not give effect sizes for inflammation, only p.
   Nutrients, 2019 · checked 2026-10-07 · we read the abstract
9. ev-crbi-09 · Meta-analysis or systematic review · meta-analysis of RCTs · n = 14 trials
   Meta-analysis including 14 RCTs showed that intervention with dietary fiber or fiber-rich food, compared with control, produced a slight, but significant reduction of 0.37 mg/L (95% CI -0.74, 0) in circulating CRP level among this population.
   
   Who: overweight and obese adults; reduction seen when fiber intake was 8 g/day higher than control
   Effect: CRP -0.37 mg/L (95% CI -0.74 to 0)
   Certainty: The upper CI bound is zero; the effect is borderline significant.
   Int J Food Sci Nutr, 2015 · checked 2026-10-07 · we read the abstract
10. ev-crbi-10 · Meta-analysis or systematic review · systematic review and meta-analysis of controlled feeding trials · n = 4094
    The evidence provides good indication that mixed sources that contain SSBs increase CRP, while most other food sources have no effect with some sources (fruit, 100% fruit juice, sweetened soy beverage or dark chocolate) showing decreases, which may be dependent on energy control.
    
    Who: predominantly healthy, mixed-weight or overweight adults; 12 food sources of fructose-containing sugars; median 6 weeks
    Effect: total fructose-containing sugars: no CRP effect in substitution trials; mixed sources with SSBs increased CRP; fruit, 100% juice, soy drink decreased it; moderate-to-low certainty
    Certainty: A conclusion about the source, not about sugar as a molecule; short duration.
    Nutrients, 2022 · checked 2026-10-07 · we read the abstract
11. ev-crbi-11 · Meta-analysis or systematic review · systematic review and meta-analysis of controlled intervention studies · n = 1141
    The comparator of fructose interventions vs glucose control groups showed no differences for CRP (MD: −0.03 mg/L 95% CI −0.52 to 0.46, I2 = 44%)
    
    Who: adults or adolescents in controlled intervention studies of at least 2 weeks
    Effect: fructose vs glucose CRP MD -0.03 mg/L (95% CI -0.52 to 0.46); HFCS vs sucrose MD 0.21 mg/L (-0.11 to 0.53); low quality
    Certainty: Compares sugars with each other, not sugar versus its absence.
    Nutrients, 2018, Effect of Dietary Sugar Intake on Biomarkers of Subclinical Inflammation: A Systematic Review and Meta-Analysis of Intervention Studies · checked 2026-10-07 · we read the fulltext
12. ev-crbi-12 · Randomized controlled trial(s) · randomized double-blind crossover trial · n = 24
    Fasting plasma concentrations of C-reactive protein and IL-6 did not differ significantly at the end of the 3 diet periods.
    
    Who: normal-weight to obese adults without fructose malabsorption; drinks supplied 25% of estimated energy needs
    Effect: no significant difference in fasting CRP or IL-6 between the three sugars
    Certainty: Short (8 days), no sugar-free control.
    Am J Clin Nutr, 2016 · checked 2026-10-07 · we read the abstract
13. ev-crbi-13 · Randomized controlled trial(s) · randomized double-blind crossover trial · n = 39
    Triglycerides were higher after sucrose than saccharin (p < 0.05), while no differences were seen for cholesterol, apolipoproteins, or CRP.
    
    Who: healthy normal-weight adults; sucrose 66 g/day versus saccharin or steviol glycosides 220 mg/day for 14 days
    Effect: no differences for cholesterol, apolipoproteins or CRP; triglycerides higher after sucrose than saccharin
    Certainty: Small sample, short, healthy people.
    Nutrients, 2026 · checked 2026-10-07 · we read the abstract
14. ev-crbi-14 · Observational data · systematic review of observational and intervention studies · n = 60 studies
    Benefits of higher fiber and whole-grain intakes suggested by observational studies may reflect confounding.
    
    Who: adults without inflammatory disease; 60 studies of GI/GL, fiber and whole grain
    Effect: fiber: 13 of 16 observational studies inverse, 1 of 11 trials significant; whole grain: 6 of 7 vs 1 of 7
    Certainty: Explains why observational "anti-inflammatory" data on carbohydrates should not be read as causal.
    Am J Clin Nutr, 2014 · checked 2026-10-07 · we read the abstract
15. ev-crbi-15 · Observational data · systematic review and meta-analysis of observational studies · n = 523730
    Pooled analysis showed that added sugar intake was associated with increased risk of Crohn's disease (OR 1.66; 95% Cl 1.21-2.29; n = 523,730; 14 studies) and ulcerative colitis (OR 1.59; 95% CI 1.25-2.02; n = 787,228; 18 studies).
    
    Who: general population; 11 prospective and 34 retrospective studies
    Effect: Crohn's disease OR 1.66 (1.21 to 2.29); ulcerative colitis OR 1.59 (1.25 to 2.02); soda/sweetened drinks OR 1.58 and 1.72
    Certainty: Mostly retrospective studies, prone to recall bias; level C.
    United European Gastroenterol J, 2025 · checked 2026-10-07 · we read the abstract
16. ev-crbi-16 · Position of an expert body · agency evidence map (VA Evidence Synthesis Program) · n = —
    Despite the lack of understanding of the mechanisms by which inflammation might increase the risk for or the progression of chronic diseases, some evidence suggests that some nutrients, eg, simple carbohydrates and saturated fat, may promote inflammation.
    
    Who: adults; evidence map of anti-inflammatory diets for chronic conditions
    Effect: statement of possibility, no pooled effect
    Certainty: Position without its own quantitative analysis; "may promote", not "demonstrated".
    US Department of Veterans Affairs, Evidence Synthesis Program, Beyond Diabetes, Obesity, and Cardiovascular Disease: An Evidence Map of Anti-Inflammatory Diet, 2024 · checked 2026-10-07 · we read the section