Mitochondria are the structures inside cells that turn nutrients from food into ATP, the energy a cell can spend. The brain, the muscles, the heart and the gut all draw on a steady supply, which is why problems with energy metabolism can show up in so many different places at once. In autism research, mitochondria have had a long and sometimes overstated run. This article sticks to what the studies measured.
Two terms that get mixed up
Primary mitochondrial disease is a medical diagnosis caused by genetic changes that affect how mitochondria work; it can involve one organ or several. Mitochondrial dysfunction is a broader description of abnormal mitochondrial function, which can occur with a primary disease or secondarily alongside another illness or cellular stress (Rose 2018, PMID 30039193). The two are related, but they are not interchangeable. Dysfunction does not mean an inherited disease, and it does not mean a permanent energy deficit. The Mitochondrial Medicine Society's consensus statements, which set the diagnostic and care standards most clinicians work from, treat the distinction as fundamental (Parikh 2015, PMID 25503498; Parikh 2017, PMID 28749475).
How common is it in autism? The 2012 meta-analysis
The most quoted number comes from a systematic review and meta-analysis of the literature up to 2010 (Rossignol 2012, PMID 21263444). Across the included studies, the prevalence of mitochondrial disease in autistic children was 5.0% (95% confidence interval 3.2 to 6.9%), against roughly 0.01% in the general population. Abnormal biomarker values were far more common than disease itself, and the average values of lactate, pyruvate, carnitine and ubiquinone differed between autistic children and controls, with some markers tracking autism severity.
The same review pooled 18 publications describing 112 children who had both autism and a mitochondrial disease diagnosis. In that group, developmental regression (52%), seizures (41%), motor delay (51%), gastrointestinal abnormalities (74%) and elevated lactate (78%) were all more frequent than in autistic children without the diagnosis. That is a description of a diagnosed subgroup, not a checklist for every autistic child, and the authors said so: the data suggest a spectrum of mitochondrial dysfunction of differing severity.
What the biomarker studies add, twelve years on
A 2024 systematic review and meta-analysis covered 204 articles on biomarkers of mitochondrial dysfunction in autism (Frye 2024, PMID 38703861). It found a raised prevalence of elevated lactate (17%), pyruvate (41%), alanine (15%) and creatine kinase (9%) in autistic groups, and moderate to large group-level differences in mean pyruvate, the lactate-to-pyruvate ratio, ATP and creatine kinase. Thirteen controlled studies reported mitochondrial DNA deletions or variations, and the pooled data showed differences in mitochondrial DNA copy number. One line of work described a subgroup with unusually high mitochondrial respiration, which came with increased sensitivity to physiological stress and with neurodevelopmental regression.
Two cautions sit beside those findings. A group-level difference is not a test for an individual child, and the traditional markers are widely reported as abnormal in autism yet remain non-specific (Rose 2018, PMID 30039193). Lactate rises with crying, exertion and a difficult blood draw. A normal lactate does not exclude a mitochondrial problem, and an abnormal one does not diagnose it.
Signs described in the clinical literature
The 2018 review lists the clinical features that have been described in autistic children with mitochondrial dysfunction: unusual developmental regression, especially when triggered by an inflammatory event such as an infection or fever; gastrointestinal symptoms; seizures; motor delays; fatigue and lethargy (Rose 2018, PMID 30039193). None of these is specific to mitochondria. Poor sleep, anemia, pain and many other conditions produce similar pictures, which is one more reason the first step is a medical evaluation rather than a supplement.
Some patterns need urgent care, not a plan: a sudden loss of skills, severe new weakness, breathing difficulty, difficulty waking, a first or prolonged seizure, or vomiting with an inability to keep fluids down.
How it is diagnosed, and why there is no single test
There is no routine test that confirms or rules out mitochondrial disease. A clinician reviews symptoms, growth, family history and the examination, then chooses selected blood or urine tests and, when indicated, genetic testing. A 2025 review of diagnostic practice describes most of the available tests as non-specific or shared with other conditions, and makes the case for a genetics-first pathway now that molecular testing is faster and cheaper (Bhai 2025, PMID 40177789). The consensus statements set out the same logic: diagnosis rests on clinical and molecular data together, supplemented when needed by biochemical and tissue evidence (Parikh 2015, PMID 25503498).
The care standards are plainer than the biology. Enough food and fluids, attention to a restricted diet, investigation of persistent sleep problems, activity matched to the child's ability, and a dietitian when intake is narrow or several supplements overlap (Parikh 2017, PMID 28749475; Sue 2022, PMID 34505344). Those basics apply to every child, with or without a diagnosis.
What the supplement trials show
One randomized trial addresses this directly. Thirty autistic children were randomized to liquid L-carnitine at 50 mg per kilogram of body weight per day or to placebo for three months (Geier 2011, PMID 21629200). The carnitine group improved on the Childhood Autism Rating Scale (a change of 2.03 points, 95% confidence interval 0.31 to 3.7), on a clinical global impression score and on the parent-rated ATEC, and rises in serum free carnitine correlated with gains in hand muscle strength and cognitive scores. Adherence was above 85% and the treatment was well tolerated. The authors themselves called for further studies. Fourteen years later the trial has not been repeated at scale, so it stands as one small positive result, not an established treatment.
The numbers matter for reading our own labels. For a 25 kg child the trial dose was 1,250 mg of L-carnitine a day. The MITOboost capsules carry 200 mg of acetyl-L-carnitine per four-capsule serving and the MITOboost drink 500 mg per scoop, alongside coenzyme Q10, N-acetylcysteine and alpha-lipoic acid. Our protocol is not a replica of the trial, and we do not present it as one.
Combinations of coenzyme Q10, carnitine, B vitamins and alpha-lipoic acid, the so-called mitochondrial cocktail, are common clinical practice. The Mitochondrial Medicine Society's recommendations for them are consensus-based, which is the society's own description of the state of the evidence (Parikh 2015, PMID 25503498). Correcting a measured deficiency is a different act from using a broad combination, and small trials of single ingredients do not establish a need in every child.
How this shapes our protocol
Energy Support is built from the nutrients normal energy metabolism uses: coenzyme Q10, acetyl-L-carnitine, N-acetylcysteine and alpha-lipoic acid, with magnesium for normal muscle and nerve function. It comes in two routes, a strawberry drink for anyone who does not swallow capsules and a capsule formula for those who do, with one magnesium beside either. B vitamins and trace minerals are additions for an identified need, never automatic, because the drink already carries both. For children every amount is set by the practitioner; the label servings are adult servings.
What we do not claim: that the protocol treats autism or mitochondrial disease, that a research marker is a diagnosis, or that a supplement replaces the evaluation a tired or weak child needs. We grade the evidence behind this protocol C: biomarker studies and clinical consensus, plus one small randomized trial of a single ingredient at a dose our products do not match. Judge it the way the care standards suggest: stamina, recovery after activity, sleep, appetite and bowel comfort, written down week by week, and reviewed with the clinician at eight to twelve weeks.
Parent guide
Download the Energy support parent guide (PDF, 11 pages)
Free, no sign-up. A printable version of the Energy Support protocol for parents: what to notice, how each product fits, and space to write the plan your clinician sets. The protocol itself, with every dose and caution, is at Energy Support.
Sources
- Rossignol DA, Frye RE. Molecular Psychiatry, 2012. PMID 21263444.
- Frye RE, et al. Neurobiology of Disease, 2024. PMID 38703861.
- Rose S, et al. Molecular Diagnosis & Therapy, 2018. PMID 30039193.
- Geier DA, et al. Medical Science Monitor, 2011. PMID 21629200.
- Parikh S, et al. Genetics in Medicine, 2015. PMID 25503498.
- Parikh S, et al. Genetics in Medicine, 2017. PMID 28749475.
- Sue CM, et al. Internal Medicine Journal, 2022. PMID 34505344.
- Bhai S, et al. Muscle & Nerve, 2025. PMID 40177789.