The evidence does not justify throwing every antioxidant, vitamin and mitochondrial supplement into a bottle. It does justify a much more interesting question: can carefully selected, nonredundant interventions attack several components of ALS biology at the same time?
By James Lyons-Weiler, PhD
Amyotrophic lateral sclerosis has been remarkably resistant to the idea that there should be one molecular target, one drug and one decisive therapeutic mechanism. That resistance may be telling us something important about the disease.
Motor-neuron degeneration in ALS occurs in a biological environment in which mitochondrial dysfunction, abnormal energy metabolism, oxidative injury, neuroinflammation, impaired proteostasis, glutamate-related excitotoxicity, neuromuscular-junction failure, skeletal-muscle abnormalities, altered one-carbon metabolism and, increasingly, disturbances of the intestinal microbiome have all been implicated. Which processes initiate disease, which drive progression and which are secondary consequences remain intensely studied questions. What is increasingly difficult to defend, however, is the assumption that every potentially useful intervention should be judged as though ALS were a single-pathway disorder.
That point matters when we look at nutritional and supplement-based interventions. The conventional question has usually been: Does supplement X treat ALS? For most supplements the answer, based on available evidence, is no—or at least not demonstrably so. Creatine is a good example. Its mitochondrial rationale is excellent, it performed well enough in animal models to justify human trials, and three randomized trials involving 386 participants ultimately failed to demonstrate significant effects on survival, ALSFRS-R decline or respiratory deterioration (Pastula et al., 2012). Vitamin D is another. Deficiency should be corrected for ordinary medical reasons, but controlled ALS data have not shown that raising vitamin D levels meaningfully slows motor deterioration (Trojsi et al., 2020).
Those failures are important. But they do not prove that metabolic, inflammatory or oxidative biology is irrelevant to ALS. They show that correcting one piece of a complex system may be insufficient.
That distinction is the basis for a different research question.
Instead of asking which supplement is the treatment for ALS, we should be asking which small set of interventions has enough evidence to justify combination testing, which mechanisms they cover, which ones are redundant, and which biological measurements could tell us whether they are doing what we expect them to do.
A 2025 review from Richard Bedlack and colleagues arrived at a similar conclusion after more than fifteen years of ALSUntangled evaluations. They identified eight alternative or off-label treatments having both plausible ALS mechanisms and at least some human clinical signal. Among the supplement-like interventions were acetyl-L-carnitine, L-serine, methylcobalamin and nicotinamide riboside/pterostilbene. The authors explicitly raised the possibility that targeting several mechanisms simultaneously may be more productive than repeatedly attacking one mechanism at a time (Bedlack et al., 2025).
That observation deserves to be pushed further.
Fifteen Candidates Worth Putting on the Research Table
The list below is not a proposed treatment regimen. It is a candidate library for rational combination research. Some entries have randomized human ALS data. Others have only animal evidence. They should not be treated as evidentially equivalent.
The Human Signals Are More Interesting Than They First Appear
Acetyl-L-carnitine deserves particular attention because it has already crossed the animal-to-human boundary. In a randomized, double-blind trial of 82 people with ALS receiving riluzole, 3 grams per day of ALCAR or placebo was administered for 48 weeks. The prespecified self-sufficiency outcome favored ALCAR, and several secondary measures also generated encouraging signals (Beghi et al., 2013).
That study was not definitive, but the signal was sufficient to justify going back rather than simply consigning ALCAR to the supplement graveyard. That is now happening. The ALCALS trial, NCT06126315, is recruiting 246 participants in a Phase II/III randomized, blinded, placebo-controlled study testing 1.5 and 3 grams per day. Its planned measurements are unusually informative for a combination-therapy discussion: ALSFRS-R, FVC, survival and biomarkers including PGC-1α, neurofilament light, creatine kinase, 4-hydroxynonenal and several muscle-related markers. As of September 2026, that trial remains recruiting.
Nicotinamide riboside plus pterostilbene is another human signal worth watching. A 32-person randomized, double-blind pilot study reported favorable changes in ALSFRS-R, pulmonary function and muscular strength over four months (de la Rubia et al., 2019). Small trials can produce large-looking effects by chance, and this one requires replication. But replication is exactly what is being attempted. The multicenter NO-ALS trial, NCT04562831, enrolled 380 participants and is evaluating whether high-dose NR/pterostilbene slows ALSFRS-R decline. The trial is active but no longer recruiting, with completion listed for October 31, 2026; published efficacy results are not yet available.
The biochemical rationale is substantial. NR supplies a precursor for NAD+, which sits near the center of cellular energy metabolism, mitochondrial function and several signaling systems. Pterostilbene has been investigated in relation to sirtuin and oxidative-stress biology. But this creates our first lesson in combination design: NR/PT and NMN should probably be viewed as competing approaches to the same NAD+ problem, not automatically as two ingredients to stack together.
Methylcobalamin presents an even more important distinction. Ordinary B12 supplementation and ultra-high-dose injectable methylcobalamin should not be conflated. In the JETALS Phase III trial, patients with early ALS and moderate progression receiving 50 mg intramuscular methylcobalamin twice weekly had significantly less ALSFRS-R decline during the 16-week randomized treatment period (Oki et al., 2022). Japan subsequently approved high-dose mecobalamin, marketed as Rozebalamin, for slowing functional deterioration in ALS in September 2024.
This is no longer appropriately described as someone taking a vitamin pill. At those doses and by that route, methylcobalamin is a pharmacologic ALS intervention.
But it also points directly toward a metabolic axis that might matter more broadly: homocysteine and one-carbon metabolism.
A 2026 report involving 120 people with sporadic ALS found that oral B12 and folate supplementation increased B12 and folate concentrations and lowered homocysteine. Higher B12 concentrations were associated with slower early progression, although supplementation was not associated with a significant survival difference. That does not establish oral B12 plus folate as a disease-modifying treatment. It does suggest that a combination study could reasonably measure B12, folate and homocysteine rather than treating vitamin administration as a biologically unmeasured act.
Folate belongs on the research list for that reason. Folinic acid specifically has not been established as an ALS therapy by the evidence reviewed here. The testable target is the one-carbon/homocysteine pathway; the exact intervention remains a separate question.
L-serine occupies another interesting niche. A randomized dose-ranging Phase I study enrolled 20 people with ALS and tested four twice-daily dose levels for six months. L-serine was generally tolerated, and comparison with historical controls generated an exploratory suggestion that functional decline might be slower at the highest exposure (Levine et al., 2017). The evidentiary limitations are obvious: this was principally a safety study, it was tiny, and the efficacy comparison was historical rather than a contemporary placebo group. But L-serine is attractive because its hypothesized effects do not simply duplicate mitochondrial supplementation. It has been investigated in relation to ER stress, protein handling, neuroinflammation and BMAA biology (Bedlack et al., 2025).
Curcumin has more human ALS evidence than many people realize, although the studies remain small. One randomized trial using a curcumin preparation reported relative stability of ALSFRS-R alongside changes in oxidative-stress markers (Chico et al., 2018). A separate 54-person randomized, double-blind study of nanocurcumin added to riluzole observed one death or ventilation-dependency event among 27 treated participants versus six among 27 placebo recipients over twelve months, producing a significant difference in survival curves, but no significant differences in the other clinical outcomes (Ahmadi et al., 2018).
Those studies are not sufficient to declare curcumin effective. They are sufficient to say that curcumin has progressed beyond pure biochemical speculation.
The Preclinical Group Is Where Combination Research Becomes Especially Interesting
Spermidine may be the most intriguing recent addition. In 2026, Fiorucci and colleagues studied spermidine in the SOD1-G93A model and found that treatment partly restored expression of more than 4,000 genes in skeletal muscle, including PGC-1α and mitochondrial genes, enhanced mitochondrial bioenergetics and delayed muscle weakness.
That result deserves attention because ALS is not merely a story of dying motor neurons. Disruption of the neuromuscular junction and skeletal-muscle metabolism can precede terminal neuronal loss. A supplement that influences autophagy, mitochondrial regulation and muscle biology could therefore occupy a different therapeutic niche from a compound aimed principally at neuronal oxidative injury.
NMN approaches another part of the same metabolic landscape. In SOD1-G93A mice, dietary NMN modestly extended lifespan and delayed motor dysfunction while improving neuromuscular-junction physiology, mitochondrial structure and innervation and reducing reactive astrogliosis (Lundt et al., 2024). Again, this is animal evidence. But the NMJ result is particularly relevant because motor-unit failure in ALS is not restricted to neuronal cell bodies.
The glutathione story offers one of the most useful lessons in the entire literature. A cystine-rich undenatured whey preparation increased glutathione availability in SOD1-G93A mice, delayed disease onset and slowed loss of grip strength. It did not extend survival on its own. When investigators combined it with riluzole, however, the animals retained the onset benefit associated with glutathione support and the survival benefit associated with riluzole.
That is not proof of human efficacy. It is something conceptually important: an actual experiment in which two interventions with different effects were combined rather than merely discussed.
The intestinal microbiome provides another nonredundant target. In TDP-43 mutant mice, Zhang and colleagues found impaired intestinal barrier function, altered microbial composition, elevated inflammatory cytokines and increased TDP-43 pathology. Treatment with either butyrate or the multistrain probiotic VSL#3 improved rotarod performance and intestinal motility, strengthened barrier markers, lowered inflammatory mediators and reduced TDP-43 aggregation in experimental tissues (Zhang et al., 2024).
The observation is provocative because TDP-43 pathology is central to most ALS. It connects microbial metabolites not merely with gastrointestinal symptoms but with inflammatory signaling, barrier integrity and protein aggregation.
Galactooligosaccharides offer an older version of the same systems-level idea. In SOD1-G93A mice, GOS and GOS-containing prebiotic yogurt delayed disease onset and prolonged lifespan while increasing folate and B12, lowering homocysteine, improving muscle mitochondrial activity and reducing glial activation (Song et al., 2013).
The important implication is not that someone with ALS should immediately take butyrate, VSL#3 and GOS simultaneously. Those three candidates overlap substantially. They define a microbiome/barrier/metabolite therapeutic module from which an intervention could be selected and tested.
Alpha-lipoic acid belongs in a similar intermediate category. Its mitochondrial cofactor function, antioxidant behavior and anti-inflammatory effects are biologically credible, and ALS animal studies have been encouraging. But ALSUntangled’s 2025 review found no published controlled clinical efficacy trial in people with ALS. The human reports either involved combinations or uncontrolled observations that prevent isolation of an ALA-specific effect.
Rosmarinic acid and anthocyanins are further back in the translational queue but remain interesting. Rosmarinic acid substantially delayed symptoms and improved survival in a SOD1-G93A mouse experiment while reducing oxidative injury and neuronal loss (Seo et al., 2015). A strawberry-derived anthocyanin preparation delayed disease onset, extended survival, preserved grip strength and neuromuscular junctions and reduced astrogliosis in another SOD1-G93A experiment (Winter et al., 2018).
These should not be mistaken for human ALS treatments. They are candidate probes into biological processes that could eventually earn that status.
Combination Therapy Does Not Mean “Take Everything”
This is where supplement culture and rational therapeutics part company.
If ALS is biologically heterogeneous and multimechanistic, combination therapy makes sense. But indiscriminate stacking does not follow from that premise. In fact, the fifteen-candidate list immediately shows why a kitchen-sink approach is scientifically weak.
NR/PT and NMN both target NAD+ biology. A rational trial would normally choose between them before testing them together.
Butyrate, probiotics and GOS all occupy overlapping microbiome territory. Using all three at once would make it impossible to determine which manipulation mattered.
Curcumin, alpha-lipoic acid, rosmarinic acid and anthocyanins all touch oxidative and inflammatory biology to differing degrees. Four compounds nominally attacking oxidative stress do not necessarily constitute four independent treatments.
B12 and folate belong to a common one-carbon/homocysteine system and make more sense when their biochemical effect is actually measured.
This leaves a much more defensible combination architecture. One intervention might address cellular energy and mitochondrial function. Another could address protein handling/autophagy. A third could address the gut-barrier/inflammatory axis. A fourth might address one-carbon metabolism or glutathione, preferably according to measured biochemical abnormalities. A fifth, if justified, could focus directly on skeletal muscle or neuromuscular-junction preservation.
That approach turns supplements into a hypothesis-driven therapeutic experiment.
It also forces us to specify what success would look like.
If an NAD+ intervention is included, NAD-related metabolites should be measured where feasible. If B12 and folate are included, B12, folate and homocysteine should be followed. If glutathione support is included, glutathione/redox measures should be considered. If the microbiome is being deliberately manipulated, stool metagenomics, short-chain fatty acids or appropriate barrier and inflammatory markers become relevant. ALSFRS-R, respiratory function and survival remain essential clinical outcomes, but biomarkers can answer a question that a functional score alone cannot: Did the intervention actually move the biological pathway it was selected to move?
That is how rational combination therapy becomes falsifiable.
Negative Evidence Matters Just as Much
A combination strategy becomes worthless if every plausible compound automatically earns a place in the cocktail.
Creatine should teach caution. Its mechanistic case was strong. Human trials were negative.
Vitamin D deficiency should be corrected, but supplementation has not demonstrated convincing ALS disease modification.
Zinc has legitimate SOD1 biology, but the clinical evidence is extremely thin, and high-dose zinc can itself become problematic; ALSUntangled concluded in 2025 that the evidence did not support recommending zinc to slow ALS progression.
Omega-3 fatty acids illustrate an even more important trap. Epidemiologic studies of dietary omega-3 intake and ALS risk have produced interesting associations, yet high-dose EPA actually shortened survival in the SOD1-G93A mouse model and increased a toxic lipid-peroxidation product. Anti-inflammatory is not a synonym for beneficial in ALS.
TUDCA offers perhaps the best recent reminder not to confuse plausible mechanisms and early signals with established clinical efficacy. After encouraging Phase II work, the 334-participant European Phase III TUDCA-ALS trial failed its primary ALSFRS-R endpoint and showed no significant benefit on secondary outcomes including survival.
Negative trials are not inconveniences to be explained away. They are information.
They tell us which hypotheses need revision.
The Research Question Now
The most interesting finding from this review is therefore not that fifteen supplements might help ALS.
It is that the candidates distribute themselves across a limited number of biologically distinct modules, and a handful have advanced far enough to justify carefully designed combination research.
At the human-evidence end of the spectrum, acetyl-L-carnitine, NR/pterostilbene, high-dose methylcobalamin, curcumin formulations and L-serine deserve particular attention, although their evidence is far from equivalent. ALCAR is now being retested in a 246-person Phase II/III trial. NR/PT is being evaluated in the 380-person NO-ALS study. Ultra-high-dose methylcobalamin already has a positive Phase III study and Japanese regulatory approval in early ALS.
At the translational end, spermidine, NMN, glutathione support and microbiome-directed interventions are especially interesting because they extend combination design into muscle bioenergetics, NMJ preservation, autophagy, redox control, barrier integrity and TDP-43 biology.
The mistake would be to turn those observations into another supplement recipe.
The opportunity is to turn them into a trial.
A particularly informative study would not ask whether fifteen compounds taken simultaneously work. It would select several nonredundant modules, document target engagement with biomarkers, measure ALSFRS-R and respiratory trajectories longitudinally, and use a design capable of removing ineffective components while preserving promising combinations. Factorial, platform or response-adaptive approaches could substantially outperform the traditional sequence of testing one plausible intervention after another.
ALS is heterogeneous enough that biomarker-defined subgroups may matter as well. A person with elevated homocysteine does not necessarily present the same therapeutic target landscape as a person with normal one-carbon metabolism. A person with substantial metabolic derangement or rapid weight loss may not present the same landscape as someone metabolically stable. Genetic ALS further complicates the picture.
Combination medicine is already routine in diseases where biology refuses to cooperate with single-target thinking. Oncology, HIV and tuberculosis did not advance by insisting that every useful component first prove that it could defeat the entire disease alone.
ALS may eventually demand the same intellectual transition.
The evidence reviewed here does not establish a fifteen-supplement treatment for ALS. It establishes something more useful: a finite set of testable interventions, several distinct biological modules, enough human signals to justify serious investigation, and enough failures to tell us that the next generation of studies must be designed more intelligently than the last.
References
Ahmadi M, Agah E, Nafissi S, et al. Safety and efficacy of nanocurcumin as add-on therapy to riluzole in patients with amyotrophic lateral sclerosis: a pilot randomized clinical trial. Neurotherapeutics. 2018.
Bedlack RS, et al. The scientific and therapeutic rationale for off-label treatments in amyotrophic lateral sclerosis. Annals of Neurology. 2025.
Beghi E, Pupillo E, Bonito V, et al. Randomized double-blind placebo-controlled trial of acetyl-L-carnitine for ALS. Amyotrophic Lateral Sclerosis and Frontotemporal Degeneration. 2013;14:397–405.
Chico L, Ienco EC, Bisordi C, et al. Amyotrophic lateral sclerosis and oxidative stress: a double-blind therapeutic trial after curcumin supplementation. CNS Neurol Disord Drug Targets. 2018;17:767–779.
de la Rubia JE, Drehmer E, Platero JL, et al. Efficacy and tolerability of EH301 for amyotrophic lateral sclerosis: a randomized, double-blind, placebo-controlled human pilot study. Amyotroph Lateral Scler Frontotemporal Degener. 2019;20:115–122.
Fiorucci C, Rossi MN, Di Santo R, et al. Transcriptomic analysis reveals the beneficial effects of spermidine in an ALS mouse model. Biomolecules. 2026;16:566.
Giacobbe A, Hiana J, Wang O, et al. ALSUntangled #79: alpha-lipoic acid. Amyotroph Lateral Scler Frontotemporal Degener. 2026;27:233–237.
Hu N, Ding J, Tian H, et al. Impacts of oral supplementation of vitamin B12 and plasma levels of homocysteine on progression and survival in a Chinese ALS cohort. Neurological Research. 2026;48:957–965.
Levine TD, Miller RG, Bradley WG, et al. Phase I clinical trial of safety of L-serine for ALS patients. Amyotroph Lateral Scler Frontotemporal Degener. 2017.
Lundt S, Zhang N, Polo-Parada L, Wang X, Ding S. Dietary NMN supplementation enhances motor and NMJ function in ALS. Experimental Neurology. 2024;374:114698.
Oki R, Izumi Y, Fujita K, et al. Efficacy and safety of ultrahigh-dose methylcobalamin in early-stage amyotrophic lateral sclerosis: a randomized clinical trial. JAMA Neurology. 2022;79:575–583.
Pastula DM, Moore DH, Bedlack RS. Creatine for amyotrophic lateral sclerosis/motor neuron disease. Cochrane Database of Systematic Reviews. 2012.
Ross EK, Winter AN, Wilkins HM, et al. A cystine-rich whey supplement delays disease onset and prevents spinal cord glutathione depletion in the hSOD1-G93A mouse model of ALS. Antioxidants. 2014;3:843–865.
Seo JS, Choi J, Leem YH, Han PL. Rosmarinic acid alleviates neurological symptoms in the G93A-SOD1 transgenic mouse model of ALS. Experimental Neurobiology. 2015;24:341–350.
Song L, Gao Y, Zhang X, Le W. Galactooligosaccharide improves animal survival and alleviates motor-neuron death in the SOD1-G93A mouse model of ALS. Neuroscience. 2013;246:281–290.
Trojsi F, Siciliano M, Passaniti C, et al. Vitamin D supplementation has no effects on progression of motor dysfunction in amyotrophic lateral sclerosis. European Journal of Clinical Nutrition. 2020;74:167–175.
Winter AN, Ross EK, Wilkins HM, et al. An anthocyanin-enriched extract from strawberries delays disease onset and extends survival in the hSOD1-G93A mouse model of amyotrophic lateral sclerosis. Nutritional Neuroscience. 2018;21:414–426.
Zhang Y, Xia Y, Sun J. Probiotics and microbial metabolites maintain barrier and neuromuscular functions and clean protein aggregation to delay disease progression in TDP43 mutation mice. Gut Microbes. 2024;16:2363880.
Editorial note: This review is a research-prioritization discussion, not a clinical treatment protocol. Evidence levels differ substantially across candidates, and combination safety and efficacy have not been established for the multi-agent strategies discussed here.
IPAK-EDU is grateful to Popular Rationalism as this piece was originally published there and is included in this news feed with mutual agreement. Read More
Leave a Reply