Alpha-Lipoic Acid for Diabetic Neuropathy: Why Meta-Analyses Clash

Oral alpha-lipoic acid (ALA) has not shown clinically meaningful relief of diabetic peripheral neuropathy symptoms at six months in the highest-rigor review available. That conclusion conflicts with four meta-analyses that report large improvements in symptom scores. The disagreement is less mysterious than it looks: the reviews pooled different kinds of treatment protocols and asked different clinical questions.

The central distinction is duration and route. Short courses of intravenous ALA can produce rapid changes in symptom scores. Whether a person taking 600 mg of oral ALA each day gets durable, noticeable relief is a separate question. When a 2024 Cochrane review limited its analysis to randomized trials lasting at least six months, the apparent benefit largely disappeared.

The six-month result falls below a noticeable benefit

The Cochrane review included three randomized controlled trials, totaling 816 participants, that met its minimum six-month treatment requirement. For the Total Symptom Score (TSS) at six months, the pooled estimate from 330 participants was a mean difference of −0.16 points with ALA versus placebo (95% confidence interval −0.83 to +0.51). The review rated this evidence moderate certainty.

That estimate matters clinically because the review used a minimal clinically important difference of 0.97 TSS points. Even the most favorable end of the confidence interval, −0.83, remained below that threshold. The result therefore does not say that ALA produced a small but meaningful improvement that happened to miss statistical significance. It says the available six-month evidence is compatible with little or no patient-noticeable symptom benefit.

For neurological impairment measured with the Neuropathy Impairment Score of the Lower Limbs (NIS-LL), the estimate was −1.02 points (95% CI −2.93 to +0.89; 245 participants). This result was low-certainty evidence: the interval crossed zero and was imprecise enough to leave some uncertainty. All three long-duration trials had high overall risk of bias because of attrition. None reported a validated quality-of-life measure, so the trials cannot tell us whether treatment changed sleep, daily function, or pain interference.

Why the positive meta-analyses look so different

Several other reviews found statistically significant improvements in subjective neuropathy symptoms. A 2024 network meta-analysis by Prado et al. pooled 11 trials of oral and intravenous ALA. Its direct comparison for ALA 600 mg per day versus placebo found a standardized mean difference in composite TSS of −1.05 (95% CI −2.07 to −0.04; p=0.04). Its network estimate was −1.68 (95% CI −2.8 to −0.6).

Those are large-looking numbers, but the direct analysis had I²=98.18%. In practical terms, almost all observed variation came from differences among trials rather than ordinary sampling noise. The studies varied in administration route, treatment duration, baseline severity, attrition, and outcome assessment. Combining a short intravenous loading protocol with months of oral dosing creates one numerical average, but it does not create one biologically or clinically uniform treatment.

The issue is especially clear in the ALADIN and SYDNEY trial series. These studies used intravenous loading over roughly three to five weeks and reported acute reductions in TSS. They were eligible for positive reviews that mixed routes and durations, but they were structurally excluded from Cochrane's chronic-management question. That exclusion was not a dismissal of the intravenous signal; it was a decision that a few weeks of clinic-administered IV treatment cannot establish sustained benefit from long-term oral supplementation.

A 2023 review by Orellana-Donoso et al. illustrates the problem. Its six trials, involving 1,077 participants, had a mean follow-up of only 31 days, ranging from 1 to 84 days. One short-course IV cohort reported TSS changes of −5.72 ± 1.53 with ALA and −1.83 ± 1.97 with placebo (p<0.001). That is an acute IV result, not evidence that an oral supplement maintains the same effect over six months or longer. The review did not publish a route-stratified subgroup analysis.

Subjective symptom improvements did not track nerve measures

A 2026 systematic review by Vergara Salinas et al. included 15 studies and quantitatively synthesized 12. It reported statistically significant symptom-subscale findings: paresthesia had an SMD of −1.04 (95% CI −1.24 to −0.84; p<0.00001), numbness had an SMD of −0.23 (95% CI −0.44 to −0.01; p=0.04), and the Hamburg Pain Adjective List had an SMD of −1.00 (95% CI −1.15 to −0.85; p<0.00001).

Yet the same review found no significant changes in sural sensory nerve action potential, peroneal motor nerve conduction velocity, HbA1c, or nitric oxide. Those measures do not settle whether a patient feels better; pain and tingling are patient-important outcomes. They do, however, weaken claims that ALA regenerates nerves, corrects diabetic neuropathy, or modifies the disease process. A broader 2024 review of disease-modifying therapies for diabetic peripheral neuropathy provides similar context: ALA has not demonstrated significant objective nerve-function changes as a standalone disease-modifying treatment.

The gap between subjective and objective endpoints has several possible explanations. Placebo response can affect symptom scales. Attrition can distort results when participants who do poorly are more likely to leave a study. An acute effect on symptoms could also fail to translate into nerve repair. The available trials cannot cleanly separate those explanations, so the responsible conclusion is narrower than either "ALA works" or "ALA does nothing."

The oral-only review narrows one problem, not all of them

A 2023 meta-analysis by Hsieh et al. excluded intravenous ALA and focused on oral treatment. That makes it more relevant to supplement use than reviews that blend IV and oral protocols. But it did not require the six-month minimum used by Cochrane, so it does not fully answer the long-term question. Its oral-only design reduces route confounding; it does not eliminate the duration difference that separates the positive literature from the Cochrane analysis.

This is why averaging all published estimates would mislead. The positive reviews and the Cochrane review are not simply five votes on the same proposition. Some estimate short-term symptom changes after mixed IV and oral treatment. Cochrane estimates sustained outcomes in longer trials. The latter question is the one most relevant to someone buying oral ALA for chronic diabetic neuropathy.

A combination trial raises a different clinical question

ALA may still have a role as part of pain treatment rather than as a standalone supplement. In a multicenter, double-blind, non-inferiority trial, Zárate et al. studied adults with painful diabetic peripheral neuropathy who did not achieve a 50% pain reduction during pregabalin titration. Participants received either pregabalin 150 mg twice daily or a fixed combination of pregabalin 80 mg plus thioctic acid 400 mg twice daily for 12 weeks. The combination produced non-inferior pain control with a more favorable adverse-effect profile than high-dose pregabalin alone.

That finding does not validate oral ALA monotherapy. It tests whether ALA can help lower the pregabalin dose while preserving pain control. The result needs replication, but it is a reasonable subject for a clinician-patient discussion when pregabalin side effects limit treatment.

What to do with ALA now

ALA appears reasonably well tolerated in the long-duration trials: Cochrane found no significant increase in adverse events versus placebo. A separate randomized trial of 210 normoglycemic adults with idiopathic pain also evaluated oral ALA at 400 or 800 mg per day for two months, but its non-diabetic population cannot establish benefit for diabetic neuropathy.

• Short-term IV ALA has an acute symptom signal, but its durability after the loading period is unknown.
• For oral ALA 600 mg per day, the best available six-month evidence does not support clinically meaningful symptom relief or disease modification.
• ALA plus lower-dose pregabalin is a distinct, emerging treatment strategy rather than proof that ALA alone works.

For chronic diabetic peripheral neuropathy, oral ALA is better viewed as an optional, generally tolerated supplement with uncertain practical value—not a proven treatment for nerve repair or reliable symptom control.