Does magnesium reduce inflammation?

Not in any way you could count on. The largest review, a 2026 meta-analysis drawing on 78 randomized trials of magnesium supplements, not all of which measured inflammation, found interleukin-6 fell by 1.10 pg/mL (95% CI 0.27 to 1.93) on moderate certainty evidence, but only at doses of 300 mg a day or more and in organic forms, and the authors say the clinical meaning is uncertain. In the same review, C-reactive protein, the standard blood marker of inflammation, did not change. A 2021 review of 18 trials in 927 adults found no effect on CRP, IL-6 or TNF-alpha at all.

Unsettled. The trial reviews point in different directions. One 2022 analysis finds a small fall in CRP, two others find none, and the IL-6 result in 2026 is not repeated in the 2021 or 2022 reviews. Where a fall does appear, it is in people whose CRP was already raised or who had metabolic syndrome, and with large differences between the trials pooled.

What the trials found

C-reactive protein

The 2026 meta-analysis put the pooled change in CRP at 0.15 mg/L, with an interval from -0.01 to 0.30 that touches zero. In an exploratory subgroup, doses under 300 mg a day were linked to a rise in CRP. The 2021 review of 18 trials gave a change of -0.49 mg/L, with a wide interval from -1.72 to 0.75, and found no relation with dose or duration. Its full text was not available to us, so its heterogeneity is known only from the abstract.

The 2022 meta-analysis of 15 placebo-controlled trials in 737 people did find a fall, a standardized difference of -0.356 (95% CI -0.659 to -0.054). Its participants were mostly people with diabetes, prediabetes or obesity, followed for a median of 12 weeks, most often on 250 mg a day of magnesium oxide. Heterogeneity was high, at 74.8%, and the result is in standardized units, so it does not translate into mg/L. The same review found no change in IL-6 or TNF-alpha.

Who the fall shows up in

A 2017 meta-analysis of 11 trials found no effect overall. In the subgroup whose CRP was above 3 mg/L at the start, CRP fell by 1.12 mg/L (95% CI 0.18 to 2.05). That is a subgroup analysis. A 2025 meta-analysis of 8 trials in 444 people with metabolic syndrome, on magnesium for at least 2 months, found a standardized fall of -0.327 (95% CI -0.602 to -0.053), with p = 0.048, and the effect disappeared in most subgroups.

Another 2025 meta-analysis, of 9 trials in 509 adults with overweight or obesity, found hs-CRP fell when magnesium was taken together with vitamin D or E (mean difference -1.19, 95% CI -1.95 to -0.42, with no unit stated in the abstract). With magnesium and vitamin E the fall was not significant. Because every arm was a combination, the share that belongs to magnesium cannot be separated.

Single trials

In 49 people with stable COPD, 300 mg a day of magnesium citrate for 6 months gave lower CRP than placebo (beta -3.2, 95% CI -6.0 to -0.4), with no change in lung function. The trial did not reach its planned size and was registered retrospectively. In 86 people with prediabetes, 250 mg a day of magnesium oxide for 12 weeks did not change CRP against placebo. That dose is below the 300 mg at which the 2026 review saw its IL-6 effect.

Diet, outside trials

Observational data link eating the least magnesium with 49% higher odds of a CRP of 3 mg/L or more than eating the most (OR 1.49, 95% CI 1.18 to 1.89), pooled from 3 cross-sectional studies. These figures are odds and should not be read as risk, and we have no absolute rate to set them against. Magnesium in food comes with vegetables, whole grains and nuts, so this pattern cannot show that magnesium itself is the cause.

Who should be careful

The dose where the 2026 review saw an IL-6 effect, 300 mg a day or more, sits close to the upper limit the US National Institutes of Health sets for supplemental magnesium: 350 mg a day for adults, and 65 to 350 mg for children and teens depending on age. Magnesium in food does not count toward it.

Not for everyone. High doses of magnesium from supplements or medicines often cause diarrhea, sometimes with nausea and cramping, according to NIH. Magnesium oxide, the form most often used in the inflammation trials, is among the forms most often reported. If your kidneys work poorly, the risk of magnesium toxicity rises because the excess is no longer cleared, and supplements are a question for your doctor. Magnesium can also interact with oral bisphosphonates, tetracyclines and quinolone antibiotics.

The 2026 signal that doses under 300 mg a day went with a rise in CRP came from an exploratory subgroup. It is a flag, and it is not evidence of harm.

We have no cards on magnesium supplements for inflammation in pregnancy or in children, beyond the upper limits above.

What expert bodies say

We found no position from a US or European body on magnesium and inflammation. The NIH magnesium fact sheet has no section on it, and the EU register of health claims holds no authorized magnesium claim on inflammation. The NIH positions used on this page are about safety: the upper limit, gut side effects, kidney function and drug interactions.

How we searched

What would change this answer

• A large preregistered trial in adults with raised hs-CRP, above 3 mg/L, or with low magnesium status, at 300 mg a day or more of an organic magnesium form for at least 12 weeks, reporting both CRP and IL-6. The one registered trial we found, of magnesium lactate, is listed as completed with no results posted.
• Trials that report outcomes people feel or that change disease, rather than blood markers alone.
• Agreement between the meta-analyses. Today one finds an IL-6 effect without a CRP effect, another the reverse, and a third neither.

The rest of the nutrient is in the magnesium guide, and the sleep question has its own evidence page.

Sources

1. ev-mginf-01 · Meta-analysis or systematic review · systematic review and dose-response meta-analysis of RCTs · n = 78 trials (all outcomes)
   The meta-analysis demonstrated that Mg supplementation significantly reduced serum IL-6 levels (WMD: −1.10 pg/mL, 95% CI: −1.93, −0.27) in comparison with the control group.
   
   Who: adults in 78 RCTs of magnesium supplementation versus control reporting cardiometabolic risk factors, a subset reporting inflammatory biomarkers
   Effect: IL-6 WMD -1.10 pg/mL (95% CI -1.93 to -0.27); GRADE moderate certainty
   Certainty: Strong heterogeneity; the IL-6 reduction held only at a dose >= 300 mg/day and organic forms; the authors say clinical significance is uncertain.
   Nutrients, 2026 · checked 2026-10-07 · we read the fulltext
2. ev-mginf-02 · Meta-analysis or systematic review · systematic review and dose-response meta-analysis of RCTs · n = 78 trials (all outcomes)
   Subgroup analyses indicated that Mg supplementation was associated with a significant increase in CRP levels at Mg doses < 300 mg/day.
   
   Who: adults in RCTs of magnesium supplementation versus control reporting CRP
   Effect: CRP WMD 0.15 mg/L (95% CI -0.01 to 0.30); significant CRP increase in the subgroup with doses below 300 mg/day
   Certainty: Subgroups are an exploratory analysis; GRADE for CRP is moderate. A CRP rise at small doses is a signal, not demonstrated harm.
   Nutrients, 2026 · checked 2026-10-07 · we read the fulltext
3. ev-mginf-03 · Meta-analysis or systematic review · systematic review and dose-response meta-analysis of RCTs · n = 927
   Our results indicate that the supplementation with magnesium had no statistically significant effect on serum concentrations of CRP (WMD, - 0.49; 95% CI, - 1.72 to 0.75 mg/L; P = 0.44)
   
   Who: adults in 18 RCTs of oral magnesium supplementation
   Effect: CRP WMD -0.49 mg/L (95% CI -1.72 to 0.75); IL-6 WMD -0.03 pg/mL (-0.40 to 0.33); TNF-alpha WMD 0.12 pg/mL (-0.08 to 0.31); no dose or duration relation
   Certainty: Search up to December 2020; no full text in PMC, so heterogeneity and quality are from the abstract only.
   Biol Trace Elem Res, 2022 · checked 2026-10-07 · we read the abstract
4. ev-mginf-04 · Meta-analysis or systematic review · systematic review and meta-analysis of placebo-controlled RCTs · n = 737
   Among 737 participants in 15 RCTs, compared to placebo, Mg supplementation significantly decreased serum CRP (SMD = −0.356; 95% CI: −0.659 to −0.054; p = 0.02) (Figure 2)
   
   Who: 737 participants (mean age 47, 62.5% female, mean BMI 29) in 15 RCTs, mostly with diabetes, prediabetes or obesity; median follow-up 12 weeks, most often 250 mg/day magnesium oxide
   Effect: CRP SMD -0.356 (95% CI -0.659 to -0.054); I2 74.8%
   Certainty: Effect in standardized units; heterogeneity was explained by the share of women across groups. Overall risk of bias is low.
   Nutrients, 2022 · checked 2026-10-07 · we read the fulltext
5. ev-mginf-05 · Meta-analysis or systematic review · systematic review and meta-analysis of placebo-controlled RCTs · n = at least 3 trials per marker
   On the contrary, in the meta-analyses performed with at least three studies, Mg supplementation did not affect the serum levels of IL-6, total antioxidant capacity, glutathione (GSH), tumor necrosis factor alpha
   
   Who: participants in RCTs of oral magnesium versus placebo reporting IL-6 and TNF-alpha
   Effect: no significant effect on IL-6, TNF-alpha, total antioxidant capacity or glutathione
   Certainty: Other markers (fibrinogen, IL-1) fell only in 1-2 RCTs, without pooling.
   Nutrients, 2022 · checked 2026-10-07 · we read the fulltext
6. ev-mginf-06 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 11 trials
   a significant reduction of CRP values was observed in the latter subgroup (WMD: -1.12 mg/L, 95% CI: -2.05, -0.18, p=0.019)
   
   Who: participants in 11 RCTs of oral magnesium supplementation, stratified by baseline plasma CRP above or at most 3 mg/L
   Effect: overall CRP WMD -0.11 mg/L (95% CI -0.75 to 0.52); baseline CRP > 3 mg/L: WMD -1.12 mg/L (-2.05 to -0.18)
   Certainty: Subgroup analysis; the abstract conclusion writes "mg/dL", the results "mg/L".
   Curr Pharm Des, 2017 · checked 2026-10-07 · we read the abstract
7. ev-mginf-07 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 444
   Compared with placebo, magnesium intake notably lowered C-reactive protein (CRP) levels in serum (SMD = -0.327; 95% CI: -0.602 to -0.053; p = 0.048).
   
   Who: patients with metabolic syndrome in 8 RCTs of oral magnesium versus placebo lasting at least 2 months
   Effect: CRP SMD -0.327 (95% CI -0.602 to -0.053); clearer in women and at 12 to 16 weeks
   Certainty: Borderline significance (p = 0.048); in most subgroups the effect disappeared.
   Front Nutr, 2025 · checked 2026-10-07 · we read the abstract
8. ev-mginf-08 · Meta-analysis or systematic review · systematic review and meta-analysis of RCTs · n = 509
   Co-supplementation of magnesium and vitamin D/E lowered levels of serum hypersensitivity C-reactive protein (hs-CRP) (MD: -1.19, 95%CI: -1.95, -0.42, p = 0.002).
   
   Who: obese or overweight participants in 9 RCTs of magnesium co-supplemented with vitamin D or vitamin E
   Effect: hs-CRP MD -1.19 (95% CI -1.95 to -0.42); with vitamin D: MD -0.66 (-1.17 to -0.14); with vitamin E: not significant
   Certainty: Combination with vitamin D or E, so the effect of magnesium cannot be isolated. MD units not given in the abstract (No. 6: mg/L removed from the claim).
   Front Nutr, 2025 · checked 2026-10-07 · we read the abstract
9. ev-mginf-09 · Randomized controlled trial(s) · double-blind randomized placebo-controlled trial · n = 49
   Linear mixed models revealed significantly lower CRP values in the intervention group than in the placebo group at the 6 month follow-up (β = - 3.2, 95% CI - 6.0, - 0.4, p = 0.03).
   
   Who: stable-phase COPD patients given 300 mg/day magnesium citrate or placebo
   Effect: CRP beta -3.2 (95% CI -6.0 to -0.4) versus placebo at 6 months; no change in lung function or physical performance
   Certainty: The planned sample size was not reached; registration is retrospective.
   Aging Clin Exp Res, 2022 · checked 2026-10-07 · we read the abstract
10. ev-mginf-10 · Randomized controlled trial(s) · double-blind randomized placebo-controlled trial · n = 86
    The mean changes of HOMA-IR index, total cholesterol, LDL-cholesterol, triglyceride, uric acid and C-reactive protein levels as well as anthropometric indices and blood pressure in supplemented and placebo groups did not differ significantly.
    
    Who: people with prediabetes given magnesium oxide 250 mg/day or placebo for 12 weeks
    Effect: no significant difference in CRP change versus placebo
    Certainty: The dose is below 300 mg, at which the 2026 MA showed an effect on IL-6.
    Sci Rep, 2022 · checked 2026-10-07 · we read the abstract
11. ev-mginf-11 · Observational data · meta-analysis of cross-sectional studies · n = 32918
    The pooled OR (95% CI) of having CRP ≥ 3 mg/l was 1.49 (1.18-1.89) on comparing the lowest to the highest group of Mg intake from three studies with the data available.
    
    Who: general population in cross-sectional studies (7 studies, 32,918 participants; OR pooled from 3)
    Effect: OR 1.49 (95% CI 1.18 to 1.89) for CRP >= 3 mg/L, lowest vs highest dietary magnesium intake
    Certainty: Cross-sectional data: dietary Mg comes together with vegetables, whole grains, and nuts, so causality is not established.
    Eur J Clin Nutr, 2014 · checked 2026-10-07 · we read the abstract
12. ev-mginf-12 · Position of an expert body · agency fact sheet · n = —
    The Tolerable Upper Intake Level for supplemental magnesium is 350 mg for adults, and it ranges from 65 to 350 mg for children and adolescents, depending on age.
    
    Who: Adults; children and adolescents 65 to 350 mg depending on age
    Effect: UL for supplemental magnesium 350 mg/day (adults); food magnesium not counted
    Certainty: The IL-6 signal in the 2026 meta-analysis is at doses >= 300 mg, i.e., close to the upper limit.
    NIH Office of Dietary Supplements, Magnesium fact sheet for health professionals · checked 2026-10-07 · we read the section
13. ev-mginf-13 · Position of an expert body · agency fact sheet · n = —
    However, high doses of magnesium from dietary supplements or medications often result in diarrhea that can be accompanied by nausea and abdominal cramping
    
    Who: people taking high-dose magnesium supplements or magnesium-containing medicines
    Effect: diarrhea, nausea, abdominal cramping; oxide, carbonate, chloride and gluconate most often reported
    Certainty: Magnesium oxide is the most common form in RCTs on inflammation.
    NIH Office of Dietary Supplements, Magnesium fact sheet for health professionals · checked 2026-10-07 · we read the section
14. ev-mginf-14 · Position of an expert body · agency fact sheet · n = —
    The risk of magnesium toxicity increases with impaired renal function or kidney failure because the ability to remove excess magnesium is reduced or lost
    
    Who: People with impaired renal function or kidney failure
    Effect: higher risk of magnesium toxicity
    Certainty: Chronic inflammation often accompanies CKD, and it is this group for whom supplements without a doctor are not suitable.
    NIH Office of Dietary Supplements, Magnesium fact sheet for health professionals · checked 2026-10-07 · we read the section
15. ev-mginf-15 · Position of an expert body · agency fact sheet · n = —
    Magnesium may interact with certain medications, such as oral bisphosphonates, tetracyclines, and quinolone antibiotics.
    
    Who: people taking oral bisphosphonates, tetracyclines or quinolone antibiotics
    Effect: drug interactions
    Certainty: Spread the intake over time, per the ODS fact sheet.
    NIH Office of Dietary Supplements, Magnesium fact sheet for health professionals · checked 2026-10-07 · we read the section