If you have an autistic child, your supplement shelf probably grew one bottle at a time. A fish oil a pediatrician suggested, a magnesium from a parent group, then a multivitamin, a “calm” blend, a sleep chewable, maybe a probiotic after a rough patch. Each one looked reasonable on its own.
The trouble is that supplements don't act one at a time. A child takes them together, and the same nutrient often sits in three different bottles. That's where most of the real risk lives: not in a single label, but in the total.
We ran into this while building our own protocols, and it forced a rule that's the most useful thing a parent can borrow. One of each. One product per active ingredient, whether that's the fish oil or the B6 hiding inside a magnesium powder, and for a short list of ingredients, none at all in a home stack. Here's the reasoning, so you can run the same check on any shelf, ours included.
The risk hides in the overlap, not the label
Vitamin B6 is the clearest example. It sits in about twenty of the products we sell, and most parents wouldn't guess it, because a lot of it rides inside magnesium powders, where the eye lands on “magnesium,” not the B vitamins underneath. B6 has a real ceiling. Too much of it over time damages sensory nerves, a syndrome first pinned down in people taking megadoses (Schaumburg 1983, PMID 6308447). Australia's medicines regulator dropped its warning threshold to 10 mg a day in 2022, because people were reaching it unknowingly through their magnesium.
So two products that each look fine can push a child past a limit neither one crosses alone. Our cap is 25 mg of B6 a day for an adult and 10 mg for a child, counted across the whole shelf, and any product over 10 mg should carry a nerve-symptom caution. The method under the cap is the part you can reuse anywhere: write down every product, list its actives, and add up each active across the stack. You're checking the totals, not the bottles.
The nutrients you count across the whole shelf
The same counting applies to every nutrient with an upper limit. Magnesium is capped at 350 mg a day from supplements for anyone nine and older, and lower for younger children; someone with kidney disease shouldn't add magnesium at all, because failing kidneys can't clear it (Vermeulen 2023, PMID 36771254). One magnesium product covers it, taken in the evening, where it settles the gut and fits a wind-down.
Fat-soluble vitamins are stricter, because the body stores the excess instead of flushing it. One vitamin D source, under the age limit, which is 4,000 IU a day for an adult and 2,500 for a one-to-three-year-old. Preformed vitamin A worries us most in small children: a single drop of a concentrated liquid A can deliver more than double a toddler's daily limit, and vitamin A overload has driven blood calcium dangerously high in a child with autism (O'Neal 2020, PMID 32658863; Hathcock 1990, PMID 2197848). That's why our A and D liquids say “one drop, not one dropper.” A full dropper is a stack of drops, and infants have landed in hospital with high blood calcium and kidney deposits after a caregiver squeezed out a dropperful of vitamin D (Gérard 2022, PMID 35178365; Pizzini 2024, PMID 39629100).
Iron gets its own rule for a blunter reason. A bottle of it is one of the most common causes of fatal poisoning in a young child who swallows the lot at once (Bateman 2018, PMID 29587543). So one iron source, stored where a child can't physically reach it. Iron also competes with other minerals, so take it away from calcium, zinc and coffee, with a little vitamin C.
Fish oil has no hard ceiling, but the FDA flags about three grams a day as the upper end, and higher doses thin the blood, so you don't want two oils stacked or one next to other blood-thinning ingredients. One oil, and remember it's a fish allergen.
A few ingredients are one-only for what they do, not how much
Some ingredients earn the one-only rule through interaction, not arithmetic. 5-HTP is the sharpest case. It's a serotonin precursor, and stacked with an antidepressant, a migraine triptan, or a few other medicines it can drive serotonin too high, a reaction that can turn dangerous (Mikkelsen 2023, PMID 37309284). Plenty of families in our community have a child or a parent on an SSRI, so this isn't hypothetical. We use one 5-HTP product at most, keep it out of every child's stack, and expect any 5-HTP label to say plainly: don't combine it with an antidepressant without your prescriber. Watch for it hiding in sleep and calm blends, where some labels carry no medication warning at all.
Ashwagandha is an adults-only ingredient, for two separate reasons. The first is the liver: the NIH's LiverTox database rates it a likely cause of clinically apparent liver injury, and causality-checked case reports back that up (Björnsson 2020, PMID 31991029). The second is the thyroid, which it can tip into overactivity (van der Hooft 2005, PMID 16355578). One ashwagandha product at a time, none for anyone pregnant, none with existing liver or thyroid trouble, and stop it for jaundice, dark urine or unusual fatigue.
Probiotics surprise people here, because they're the friendliest thing on the shelf. For most children they're low-drama, and the rule is simply one live-culture product at a time. The exception matters, though. A child who is severely immunocompromised, has a central line, or was born very premature should skip live cultures entirely: in those situations the organisms have caused bloodstream infections, including an infant death that prompted FDA action in 2023 (Thye 2026, PMID 42074240).
The products that don't belong in a home stack at all
A handful of ingredients don't belong in a build-it-yourself stack at all. They can sit on a practitioner's protocol, with monitoring, but not on a shelf you manage on your own.
- Lithium, even the low-dose “orotate” form sold as a supplement. It has a narrow margin between a working dose and a harmful one, affects the thyroid and kidneys, and interacts with common blood-pressure and pain medicines. Not something to self-manage, and not for children.
- Huperzine A. It works like a class of Alzheimer's drugs, blocking the enzyme that clears acetylcholine, and the trial evidence for it is thin and low in quality (Li 2008, PMID 18425924). There's no established children's dose.
- Chelators such as EDTA. The FDA has never approved an over-the-counter chelation product for anything, and children have died from chelation, among them a five-year-old autistic boy (Baxter and Krenzelok 2008, PMID 18949650; Brown 2006, PMID 16882789). This one isn't a grey area.
- Binders like activated charcoal and zeolite. They grab drugs and nutrients without discriminating, so they can quietly pull a medication below its working dose and cancel out the rest of what you give.
- Antimicrobial herbs such as wormwood, black walnut and berberine. There's little to no pediatric safety data behind them, and berberine can displace bilirubin in a newborn, a known risk to the developing brain (Chan 1993, PMID 8513024).
When you take it changes how much lands
Timing isn't housekeeping; it decides how much of a dose gets absorbed, and in one case it's a safety rule outright. Digestive enzymes go at the start of a meal. Fat-soluble vitamins and fish oil want a meal with some fat in it. Magnesium fits the evening. Iron goes away from calcium, zinc and coffee, which all blunt it. And if a clinician ever has you use a binder, it goes two hours clear of everything else, so it can't soak up the medicine and nutrients you meant to give.
How this shapes our protocol
Our Daily Base is where the one-of-each rule shows up most plainly. For adults it's three products: Omega-3 Mini, the single fish oil; Vitamin D3 Micellized; and Magnesium Night. The vitamin D is deliberately K-free, because vitamin K interferes with blood thinners and the fish oil already nudges bleeding risk. Magnesium Night carries zero B6, so it can be the stack's magnesium without touching the B6 allowance, and it's built for the evening. For children the Daily Base is Spectrum Formula + Iron Chewable, the child's one iron source, which needs child-resistant storage, plus Omega-3 Mini, both under a practitioner's direction.
It's called a base because the Calm, Stress and Mood protocols build on top of it, not beside it. They share the same fish oil, vitamin D and magnesium, so a family running two of them still takes one of each, not two. Add a protocol and you count actives per person per day, not per order, dropping whatever's already covered. See the Daily Base protocol for the exact products and doses.
Two honest limits. Most of what's known about these ingredients, benefits and harms alike, comes from adults and from single-ingredient studies, not from trials in autistic children, and no stack like ours has been tested as a whole. Where a product gives less than a studied dose, like the 3 mg of iron in the child multivitamin, well short of what it takes to correct low iron under a doctor, we say so on the product page rather than let the label imply more. The point isn't to sell you more bottles. It's to help you take fewer, on purpose.
Sources
- Schaumburg H, et al. Sensory neuropathy from pyridoxine abuse: a new megavitamin syndrome. NEJM. 1983. PMID 6308447.
- Vermeulen EA, et al. Magnesium administration in chronic kidney disease. Nutrients. 2023. PMID 36771254.
- O'Neal S, et al. Hypercalcemia from hypervitaminosis A in a child with autism. J Pediatr Endocrinol Metab. 2020. PMID 32658863.
- Hathcock JN, et al. Evaluation of vitamin A toxicity. Am J Clin Nutr. 1990. PMID 2197848.
- Gérard AO, et al. Severe hypercalcemia following vitamin D intoxication in an infant. Front Pediatr. 2022. PMID 35178365.
- Pizzini C, et al. Nephrocalcinosis in an infant due to a vitamin D food-supplement overdose. Front Pediatr. 2024. PMID 39629100.
- Bateman DN, et al. Iron overdose epidemiology, clinical features and concentration-effect relationships. Clin Toxicol (Phila). 2018. PMID 29587543.
- Mikkelsen N, et al. Serotonin syndrome: a focused review. Basic Clin Pharmacol Toxicol. 2023. PMID 37309284.
- Björnsson HK, et al. Ashwagandha-induced liver injury: a case series. Liver Int. 2020. PMID 31991029.
- van der Hooft CS, et al. Thyrotoxicosis following the use of ashwagandha. Ned Tijdschr Geneeskd. 2005. PMID 16355578.
- Thye AY, et al. A review of probiotic interventions for necrotising enterocolitis and sepsis in preterm infants. Int J Mol Sci. 2026. PMID 42074240.
- Li J, et al. Huperzine A for Alzheimer's disease. Cochrane Database Syst Rev. 2008. PMID 18425924.
- Baxter AJ, Krenzelok EP. Pediatric fatality secondary to EDTA chelation. Clin Toxicol (Phila). 2008. PMID 18949650.
- Brown MJ, et al. Deaths resulting from hypocalcemia after administration of edetate disodium. Pediatrics. 2006. PMID 16882789.
- Chan E. Displacement of bilirubin from albumin by berberine. Biol Neonate. 1993. PMID 8513024.