Alpha-Lipoic Acid (ALA)
A mitochondrial dithiol cofactor with a genuine trial series in diabetic neuropathy — 600 mg/day is the dose that works, and higher doses buy nausea, not benefit. Its glucose data are older, smaller and softer, and it carries a rare but real autoimmune hypoglycaemia risk tied to HLA type.
Reviewed & fact-checked by
BiohackingHub Research TeamEditorial Research Team · Last updated: September 9, 2026
Medical Disclaimer: The information on this page is for educational and research purposes only. It is not a substitute for professional medical advice, diagnosis, or treatment.
What Alpha-Lipoic Acid Actually Is
Lipoic acid is a short fatty acid with two sulphur atoms at one end. Bound to lysine inside the mitochondrion, it is the cofactor that lets pyruvate dehydrogenase hand glycolysis over to the citric acid cycle. Every aerobic cell makes it. Nobody is deficient in it.
The supplement is something different: free racemic lipoic acid — an equal mix of the natural R-enantiomer and the synthetic S-enantiomer — at doses several hundred times what the body carries as bound cofactor. At those doses it behaves as a dithiol antioxidant that regenerates glutathione, chelates metals and, in Germany, is a licensed prescription medicine for diabetic polyneuropathy under the name thioctic acid.
That prescription history is why ALA has a real trial series. It is also why the honest reading has to separate what it does for nerves from what it does for glucose.
The Neuropathy Trials
The series is named after acronyms, and the numbers are consistent enough to be worth tabulating.
| Trial | Design | Result |
|---|---|---|
| ALADIN (1995) | 328 patients, 3 weeks IV, 100/600/1,200 mg vs placebo | Total Symptom Score fell 63.5% on 600 mg vs 38.4% on placebo (P < 0.001); 600 mg had fewer adverse events (18.2%) than 1,200 mg (32.6%) |
| ALADIN III (1999) | 509 patients, 3 weeks IV then 6 months oral 600 mg three times daily | No symptom difference from placebo at 7 months; a modest deficit (NIS) improvement at day 19 |
| SYDNEY 2 (2006) | 181 patients, 5 weeks oral, 600/1,200/1,800 mg vs placebo | TSS fell 51%, 48% and 52% vs 32% on placebo (all P < 0.05); responder rates 62%, 50%, 56% vs 26%; nausea, vomiting and vertigo rose with dose |
| NATHAN 1 (2011) | 460 patients, 4 years oral 600 mg vs placebo | Primary composite not significant (P = 0.105); NIS (P = 0.028) and NIS-LL (P = 0.05) favoured ALA; fewer progressed |
The pattern is clear. Oral 600 mg once daily improves neuropathic symptoms within five weeks, and the meta-analysis of four intravenous trials (1,258 patients) put the symptom advantage at 24.1% relative to placebo with responder rates of 52.7% versus 36.9%. Going to 1,200 or 1,800 mg does not improve on it and roughly doubles the rate of nausea and vertigo. Over four years, 600 mg did not move the primary composite but did slow progression of clinical impairment.
That is a B: real, replicated, symptom-level, with a null long-term primary endpoint that keeps it out of A.
The Glucose Evidence
Here the profile is softer, and the assignment for this page is to say so.
The best mechanistic study is a four-week placebo-controlled pilot of 74 people with type 2 diabetes. Oral ALA at 600, 1,200 or 1,800 mg raised insulin-stimulated glucose disposal on clamp by about 27% relative to placebo (P < 0.01), with no dose–response. It was explorative, it was 1999, and it has not been repeated at scale.
The meta-analyses disagree with each other. A 2018 pooled analysis of 24 trials in people with metabolic disease reported reductions in fasting glucose (SMD −0.54), insulin, HOMA-IR (SMD −0.76) and HbA1c (SMD −1.22) — effect sizes so large they should prompt suspicion about the trials that produced them. A 2021 dose–response meta-analysis of 28 trials and 1,016 participants found ALA lowered insulin and HOMA-IR (WMD −0.48, 95% CI −0.79 to −0.16, P = 0.002) but produced no change in glucose or HbA1c.
That second result is the one to plan around: a modest improvement in a fasting insulin-derived index, with no demonstrated effect on the number a clinician actually uses. Grade C.
What ALA Has Never Shown
- A change in HbA1c in the most rigorous pooled analysis.
- Prevention of diabetes, cardiovascular events, or any hard outcome.
- A benefit at 1,200 or 1,800 mg over 600 mg, for nerves or for glucose.
- Reversal of established neuropathy. NATHAN 1 slowed progression of impairment; it did not restore nerve conduction.
The Insulin Autoimmune Syndrome Story
This is the section that belongs on every ALA label and appears on almost none.
Insulin autoimmune syndrome — Hirata disease — is spontaneous hypoglycaemia with very high insulin and anti-insulin antibodies in someone never given insulin. It is classically Japanese, tied to HLA-DRB1*04:06, and triggered by sulphhydryl drugs such as methimazole. ALA is a sulphhydryl compound, and Japanese cases on ALA were first reported in 2006.
In 2014 an Italian group described six European patients — median age 63 — who developed fasting and post-absorptive hypoglycaemia 30 to 120 days after starting ALA at 600 mg/day. Five carried HLA-DRB1*04:03, one carried *04:06. Stopping ALA reduced the episodes; all needed glucose and prednisone. Two further European cases in 2019 were both *04:03. A 2024 review of 68 genotyped cases worldwide confirmed the split: *04:06 with methimazole in East Asians, *04:03 with lipoic acid in Europeans.
This is rare. It is also underdiagnosed, because an ALA user with unexplained hypoglycaemia is usually told to eat more often. Any hypoglycaemic episode on ALA warrants stopping it and measuring insulin and insulin antibodies, not a snack.
Dosing and Form
600 mg once daily, 30 minutes before a meal. Food reduces the bioavailability of both enantiomers. That is the trialled dose, and there is no evidence that more helps.
R-ALA versus racemic. The R-enantiomer is the biologically active form, and the sodium salt of R-lipoic acid produced markedly higher and faster plasma peaks than plain R-ALA or racemic ALA in a pharmacokinetic study of 12 healthy subjects. But every efficacy trial above used racemic ALA at 600 mg, and there is no clinical trial showing that R-ALA at a lower dose matches it. If you buy R-ALA, you are buying better pharmacokinetics and no outcome data; 300 mg of R-ALA is a reasonable pharmacokinetic equivalent, not a proven one.
Who This Is Reasonable For
Someone with diabetes and symptomatic distal neuropathy — burning, stabbing, numbness — has the best case: five-week symptom relief in randomised trials at a dose that is well tolerated. That person should also have an HbA1c that is being managed, because ALA treats the symptom and not the cause.
Someone with prediabetes and no neuropathy is buying a HOMA-IR effect with no HbA1c effect. That is a C-grade proposition, honestly labelled in The Insulin Sensitivity Protocol, where ALA is the fork for people who are not candidates for inositol.
Anyone on insulin or a sulfonylurea, anyone with a history of unexplained hypoglycaemia, and anyone pregnant should leave it alone.
Related Research
Stacking Interactions
How Alpha-Lipoic Acid (ALA) interacts with other compounds
The classic mitochondrial pairing, with animal data on age-related mitochondrial decline. In humans the pairing has small trials in metabolic syndrome and none with hard outcomes
Protocols using Alpha-Lipoic Acid (ALA)
Evidence-graded stacks that include this compound
Safety Profile — Tier B
Generally safe — moderate evidence
Contraindications
- ●History of insulin autoimmune syndrome (Hirata disease) or known HLA-DRB1*04:03 / *04:06 carriage — ALA is a documented trigger
- ●Pregnancy and breastfeeding — no adequate safety data
- ●Unexplained hypoglycaemia while taking ALA — stop and get insulin and insulin-antibody levels measured
Side Effects
- ●Nausea, vomiting and vertigo — dose-dependent; increased at 1,200 and 1,800 mg/day in SYDNEY 2
- ●Serious adverse events over four years were 38.1% on 600 mg versus 28.0% on placebo in NATHAN 1, without a clear pattern
- ●Insulin autoimmune syndrome — spontaneous fasting and post-absorptive hypoglycaemia appearing 30–120 days after starting 600 mg/day, with very high insulin and anti-insulin antibodies
- ●Skin rash, uncommon