Keto in MS Lowers Leptin, but Nerve-Injury Marker Stays Flat

A ketogenic diet can move metabolic blood markers in multiple sclerosis. At three months, pooled evidence found that leptin fell by 2.63 ng/mL, with a 95% confidence interval of −3.03 to −2.24. That is a precise, statistically significant result. The harder question is whether it changes the course of MS. So far, the best available blood marker of ongoing nerve-cell injury has stayed flat.

That split matters. Lower leptin and higher adiponectin may reflect meaningful metabolic changes, especially for someone with MS who also has obesity, insulin resistance, or metabolic syndrome. But neither marker is a validated substitute for fewer relapses, slower disability progression, less MRI activity, or better day-to-day function. A diet can improve a lab panel and still fail to protect the central nervous system.

Ketosis changes adipokines consistently

A 2025 systematic review and meta-analysis of ketogenic-diet studies in people with MS reported lower leptin at both measured time points. The pooled mean difference was −2.63 ng/mL at three months (95% CI, −3.03 to −2.24; p < 0.00001) and −2.18 ng/mL at six months (95% CI, −2.92 to −1.43).

Adiponectin moved in the opposite direction: the pooled point estimate increased by 1.78 mcg/mL at three months and 1.65 mcg/mL at six months. Those estimates fit the usual biology of adiponectin, an adipose-derived hormone often associated with a more favorable metabolic state. But the available report does not provide confidence intervals for these adiponectin estimates, so readers cannot judge their precision as fully as they can for leptin.

These are real signals, and they persist for at least six months in the available pooled data. Still, calling them MS disease modification would skip several steps. Leptin and adiponectin sit at the intersection of fat mass, calorie intake, insulin sensitivity, and immune signaling. They are interesting biomarkers; they are not clinical endpoints.

The neuroprotection test came back negative

The key counterweight is neurofilament light chain, or NfL. Damaged axons release this protein, and serum NfL is one of the best-validated blood biomarkers of neuroaxonal injury in MS. If a ketogenic diet reduced ongoing nerve injury over the studied periods, NfL would be a sensible place to look.

It did not show a statistically significant change. In the 2025 pooled analysis, the mean difference was −0.10, with a 95% confidence interval from −0.61 to 0.40 and p > 0.05. Because that interval crosses zero, the studies cannot distinguish a small reduction from no effect or a small increase in NfL.

The most direct contributing study was a 2023 six-month study of 65 people with relapsing MS on a biomarker-verified ketogenic diet. It measured NfL as a safety and efficacy endpoint. The null result does not prove that keto can never affect neuroaxonal injury; the study duration and sample size leave room for effects too small to detect. It does mean that current human evidence does not show the neuroprotective effect often implied by keto-for-MS claims.

Lower leptin is a metabolic finding. Lower NfL would be evidence that the diet reduced a marker of nerve injury. Current studies show the first result, not the second.

Why the leptin result may not belong to ketosis alone

The largest unresolved issue is calories. Ketogenic diets commonly lead to lower energy intake and weight loss. Leptin responds quickly to calorie deficit and lower fat mass, while adiponectin can rise with weight loss. No MS ketogenic-diet trial in this evidence base used an isocaloric calorie-restricted comparison diet or a weight-stable ketogenic-diet arm.

That omission makes the mechanism uncertain. A lower leptin level could result from ketosis, calorie restriction, weight loss, or some combination. A 2024 meta-analysis of randomized trials across metabolic conditions found similar ketogenic-diet effects on leptin and adiponectin outside MS. That consistency supports the reality of the metabolic response, but it weakens the claim that the response reflects MS-specific immune control.

This is more than a technical footnote. If moderate calorie restriction or another dietary pattern can produce comparable metabolic improvements with less dietary burden, ketosis itself may not be the active ingredient. The existing studies cannot answer that question.

Do people with MS actually feel better?

The evidence for patient-important outcomes is thinner than the biomarker evidence. A 2022 systematic review of nine randomized trials examined ketogenic diets and quality of life across several chronic diseases, including MS. The studies differed too much for a pooled analysis. The MS-specific randomized trial included in that review found no significant improvement in mental or physical health quality of life. Its available reporting does not allow a detailed appraisal of its sample size, duration, ketosis verification, or quality-of-life instrument.

A 2023 network meta-analysis of dietary interventions for fatigue and quality of life in MS also found too much variation among studies to make a strong ketogenic-diet-specific conclusion. It did not establish keto as superior for fatigue or quality of life.

There is one more encouraging, but narrower, signal. A 2024 prospective study enrolled 45 people with relapsing MS in a six-month ketogenic-diet intervention. Thirty-nine completed sleep assessments at baseline and six months. Self-reported daytime sleepiness improved by a mean 1.90 points on the Epworth Sleepiness Scale. Sleep is a legitimate quality-of-life target in MS, but this was not evidence that keto slowed disability or reduced relapses.

The often-cited 2022 Phase II study in relapsing MS adds useful information about feasibility, tolerability, and biomarker-confirmed ketosis. It was not described as a randomized controlled trial, so it cannot establish that keto caused any reported clinical change. A later 2025 systematic review likewise supports keto as a feasible adjunctive strategy, not a replacement for disease-modifying treatment.

What competing diets tell us

Keto does not have a monopoly on plausible dietary benefits in MS. A 2024 cross-sectional study in 227 community-dwelling older adults with MS linked greater Mediterranean-diet adherence with lower disability, better quality-of-life measures, and favorable serum biomarkers. Cross-sectional data cannot prove that the diet caused those differences; people with less disability may also find a Mediterranean pattern easier to follow. Still, the study is a reminder that an association with better outcomes does not require ketosis.

A randomized trial of a modified Mediterranean diet versus a traditional Iranian diet in relapsing MS also reported fatigue and disability outcomes, setting a more useful bar for dietary research: patient outcomes, not only molecules in blood. Keto needs well-controlled trials that meet that bar.

What to do with this evidence

For a person with MS, a ketogenic diet may be reasonable to consider for metabolic goals, provided it fits medical care, medications, nutritional needs, and the practical reality of eating this way for months or years. The strongest evidence supports changes in leptin and adiponectin, not protection against axonal injury or proven improvement in disability.

• Do not use keto in place of disease-modifying therapy.
• If you try it, treat it as a monitored experiment rather than a cure: track fatigue, sleep, symptom burden, weight, and metabolic measures alongside usual MS care.
• Blood beta-hydroxybutyrate is a better way to verify ketosis than urine strips alone.
• Plan for gastrointestinal adaptation, micronutrient monitoring, and the social burden of a restrictive diet.
• Ask whether a less restrictive calorie-controlled or Mediterranean-style pattern could meet your metabolic goals more sustainably.

The current verdict is restrained but useful: keto changes peripheral metabolism in MS, and it may help some people with metabolic health or sleep. It has not yet shown that it reduces the nerve injury that drives irreversible disability.