You know the feeling. The child is finally asleep, the house has gone quiet, and instead of winding down you're lying there with your jaw tight and tomorrow's appointments running on a loop. Parents of autistic children carry more of this than most. A meta-analysis that pooled study after study found their parenting stress runs markedly higher than in parents of neurotypical children or children with other disabilities (Hayes and Watson 2013, PMID 22790429).
The children feel it too, in their own way. Anxiety travels closely with autism: across studies of more than two thousand autistic young people, close to 40% met criteria for at least one anxiety disorder (van Steensel 2011, PMID 21735077). Their stress biology also tends to look different, with a less predictable daily cortisol rhythm and bigger cortisol spikes to new or unpredictable situations (Corbett 2006, PMID 16005570; Corbett 2008, PMID 18592041).
So the question parents keep bringing us is a fair one. Is there anything in a bottle that helps with everyday stress, for the child or for the parent, and does the research actually back it up? Here's what the trials found, where they stop, and the two safety lines that shape what we'll put anywhere near a child.
Magnesium has the plainest evidence, and it's still modest
Start with the least glamorous mineral. In a randomized trial of 264 stressed adults with low blood magnesium, taking 300 mg of magnesium a day cut perceived-stress scores by about 42% over eight weeks (Pouteau 2018, PMID 30562392). That same trial pitted magnesium plus vitamin B6 against magnesium on its own, and adding B6 didn't help the average person; only the most severely stressed subgroup got a small extra benefit. The magnesium was doing the work.
That fits the biology. Stress and magnesium tug on each other, because stress drives magnesium out of the body and running low on magnesium sharpens the stress response, a loop a 2020 review lays out in detail (Pickering 2020, PMID 33260549). The honest caveat is who was studied: adults, and mostly adults who started out low. No one has run this trial in autistic children.
B vitamins help a little, and B6 is where we get careful
B vitamins are cofactors the body spends faster when it's under load, which is the rationale for topping them up. The trials are real, the effects small. A 90-day course of a high-dose B-complex lowered personal strain and low mood in working adults under stress (Stough 2011, PMID 21905094), and a meta-analysis of vitamin and mineral supplementation found a modest drop in perceived stress across healthy samples (Long and Benton 2013, PMID 23362497).
The catch is one of those very B vitamins. B6 turns from helper to hazard when it accumulates: too much pyridoxine over time injures sensory nerves, the numbness and pins-and-needles of a peripheral neuropathy (Schaumburg 1983, PMID 6308447). B6 also hides in a lot of products at once, magnesium blends included, so a family can stack its way to a high daily total without realising. That's why we hold total B6 across the whole plan to a low ceiling, well under the amounts tied to nerve damage, and print the neuropathy caution on any product that carries a meaningful dose. With this one vitamin, more isn't better; past a point it's the direction of harm.
Ashwagandha and Relora move cortisol, but read the fine print
Ashwagandha has the loudest stress research and the biggest asterisks. A meta-analysis of 12 randomized trials in about a thousand adults found large drops in stress and anxiety scores against placebo (Akhgarjand 2022, PMID 36017529). Two of those trials measured the stress hormone directly: 60 days of an ashwagandha root extract lowered serum cortisol by roughly 28% (Chandrasekhar 2012, PMID 23439798), and a separate trial saw morning cortisol drop about 23% with anxiety ratings falling alongside it (Lopresti 2019, PMID 31517876). Relora, a magnolia-and-phellodendron bark blend, has a smaller footprint: one small, industry-linked trial in stressed adults lowered salivary cortisol about 18% over four weeks (Talbott 2013, PMID 23924268).
Now the asterisks. The meta-analysis rated its own certainty low, with the individual trials scattering widely. Every one of these studies was run in adults, most between 25 and 48; there's no ashwagandha stress trial in children at all. And a lower cortisol reading in a study is not the same thing as a calmer child or a steadier parent at six in the evening. Cortisol is a marker, not the feeling. The symptom effects that do show up are short-term, and several of the trials were funded by the companies selling the extract. One last bit of precision: those trials used particular branded extracts (KSM-66 and Shoden) that aren't identical to the Sensoril form many products use, so what the evidence supports is the ingredient at a studied dose, not one exact extract.
Why we won't say "adrenal fatigue"
You'll see plenty of stress supplements sold on the idea that long-term stress "burns out" the adrenal glands into a state called adrenal fatigue, which the pills then restore. The concept doesn't hold up. A systematic review combed 58 studies for evidence that the condition is real and concluded it isn't, and that the tests used to diagnose it don't hold up either (Cadegiani and Kater 2016, PMID 27557747). We don't use the phrase, and we don't build a protocol on glandular "adrenal" extracts, because the premise under them is missing. The stress physiology worth acting on is the cortisol and nervous-system biology above, not a worn-out gland.
Glycine at night, and how long to give any of this
For the evening, magnesium isn't the only lever. Glycine is an amino acid that doubles as a mild wind-down aid: 3 grams before bed improved next-day fatigue and reaction time in adults whose sleep had been cut short (Bannai 2012, PMID 22529837). It carries no hormone, which is why we like it as a gentle bedtime option. (Melatonin is the better-studied choice for sleep onset in autistic children, but that's a separate topic with its own safety rules, and it has its own article.)
Whatever you try, give it a fair, time-boxed run and then judge it honestly. Magnesium and the adaptogens work over weeks, not days: the ashwagandha cortisol effect built across roughly 60 days in the trials, magnesium's was measured at eight weeks, and the B-complex trial ran 90. Glycine you can read the next morning. Set a review date, decide in advance on the one or two things that would count as a change (calmer evenings, an easier bedtime), and if nothing has moved by that date, stop rather than piling on more.
Keep the ranking straight, too. These are adjuncts. When a child's everyday stress tips over into a diagnosable anxiety disorder, the first-line treatment for anxiety in autistic young people is cognitive behavioural therapy, not a capsule (Wood 2020, PMID 31755906).
How this shapes our protocol
Our Stress and Resilience protocol is built to match the evidence above and its limits. For adults it pairs Calming Cortisol, which carries the ashwagandha, Relora and L-theanine from those trials, with Magnesium Night for the evening and a small Omega-3 Mini as a foundation. Spectrum B Complex and Benfotiamine is an optional add-on, kept as the plan's single source of vitamin B6 at 20 mg, with the neuropathy caution on it. Glycine Powder is offered as the evening alternative to Magnesium Night, for anyone who wants the wind-down without a mineral load.
For children the plan is narrower and practitioner-directed, because most of this research sits in adults and ashwagandha is an adults-only ingredient (more on that below). The child set leans on L-theanine through the GABA CoFactors Spray Tonic, plus glycine and the Spectrum Formula and Iron multivitamin. L-theanine is the one calming ingredient with a trial in children: 400 mg of it improved sleep in boys with ADHD (Lyon 2011, PMID 22214254), and 200 mg a day eased stress and sleep ratings in adults (Hidese 2019, PMID 31623400). The honest limit is dose. The child spray gives about 30 mg a serving, well under those trial amounts, so we describe it as supportive rather than as a treatment.
Two safety lines sit over all of it. In a small number of people ashwagandha has injured the liver (Björnsson 2020, PMID 31991029) and pushed thyroid hormones up (van der Hooft 2005, PMID 16355578; Hayashi 2024, PMID 38559552), so Calming Cortisol is for adults only and stays away from anyone with liver or thyroid disease, anyone on thyroid medication, and pregnancy. The B6 ceiling above is a fixed rule, not a preference.
And the candid part: nearly every trial here tested a single ingredient in adults, not a blend, and no one has studied our combination as a whole. Some of our doses match the studies (magnesium at 300 mg, glycine at 3 grams, ashwagandha's daily total inside the 240 to 600 mg trial range) and some fall short on purpose (the Relora in Calming Cortisol is 500 mg a day against about 750 mg in its one trial, and the child theanine spray is a fraction of the studied dose). We'd rather tell you that than imply the bottle does more than the research shows.
Sources
- Hayes SA, Watson SL. 2013. J Autism Dev Disord. PMID 22790429.
- van Steensel FJA, et al. 2011. Clin Child Fam Psychol Rev. PMID 21735077.
- Corbett BA, et al. 2006. Psychoneuroendocrinology. PMID 16005570.
- Corbett BA, et al. 2008. J Psychiatry Neurosci. PMID 18592041.
- Pouteau E, et al. 2018. PLoS One. PMID 30562392.
- Pickering G, et al. 2020. Nutrients. PMID 33260549.
- Stough C, et al. 2011. Hum Psychopharmacol. PMID 21905094.
- Long SJ, Benton D. 2013. Psychosom Med. PMID 23362497.
- Schaumburg H, et al. 1983. N Engl J Med. PMID 6308447.
- Akhgarjand C, et al. 2022. Phytother Res. PMID 36017529.
- Chandrasekhar K, et al. 2012. Indian J Psychol Med. PMID 23439798.
- Lopresti AL, et al. 2019. Medicine (Baltimore). PMID 31517876.
- Talbott SM, et al. 2013. J Int Soc Sports Nutr. PMID 23924268.
- Cadegiani FA, Kater CE. 2016. BMC Endocr Disord. PMID 27557747.
- Bannai M, et al. 2012. Front Neurol. PMID 22529837.
- Wood, 2020. PMID 31755906.
- Lyon MR, et al. 2011. Altern Med Rev. PMID 22214254.
- Hidese S, et al. 2019. Nutrients. PMID 31623400.
- Björnsson HK, et al. 2020. Liver Int. PMID 31991029.
- van der Hooft CS, et al. 2005. Ned Tijdschr Geneeskd. PMID 16355578.
- Hayashi M, et al. 2024. Cureus. PMID 38559552.