Feeding a vegan child: what the evidence actually requires

Since 1980, doctors have published at least thirty case reports of infants who developed severe, sometimes permanent neurological damage from vitamin B12 deficiency, and in nearly every case the route was the same: an exclusively breastfed baby whose vegan mother had not supplemented, or had stopped once the baby arrived. That is the sharpest possible version of the risk this guide is about, and it sits right next to the fact that the Academy of Nutrition and Dietetics and the British Dietetic Association both state that a well-planned vegan diet, including one followed from birth, is nutritionally adequate and appropriate at every stage of life. Both things are true. The professional consensus is not that veganism is safe for children full stop, it is that it is safe when specific, known gaps are closed, and unsafe in the exact way that closing them is skipped.

That consensus is not unanimous, and a parent deserves to know where the disagreement actually lives. The German Nutrition Society, the DGE, spent years explicitly advising against vegan diets for infants, children, pregnant and breastfeeding women, citing insufficient data. In June 2024 it revised that position: rather than endorsing vegan childhoods, it moved to neither recommending for nor against a vegan diet in these groups, still citing limited long-term data and calling for qualified nutritional counselling wherever one is chosen. The Academy of Nutrition and Dietetics and the BDA looked at similar evidence and reached the more permissive conclusion. Neither side is citing a fabricated number to get there. They differ on how much weight to put on the absence of long-term outcome trials in vegan infants, which genuinely does not exist at the scale it does for formula-fed or omnivorous children.

A two-year-old's stomach holds roughly a cup. An adult vegan diet built around beans, whole grains and vegetables is bulky and high in fibre relative to its calories, which is exactly the property that makes it good for an adult's heart and gut. In a small child it can mean filling up before taking in enough energy to grow, which is the single most common way a vegan child's diet fails in practice, and it is a mechanical problem rather than a planning failure of the parent's values. The German VeChi Diet Study, led by Weder and colleagues in 2019, measured this directly: vegan toddlers aged one to three ate 21.8 grams of fibre per 1,000 kilocalories against 12.2 grams for omnivorous toddlers on the same rough calorie intake, nearly double the fibre density for the same energy.

The same study is also the honest complication in this guide, because the fear that vegan toddlers simply do not grow as well turned out not to hold up in that cohort: total energy intake did not differ significantly between vegan, vegetarian and omnivorous toddlers, at roughly 970 to 990 kilocalories a day across all three groups, and weight-for-height, height-for-age and weight-for-age z-scores showed no significant differences either. A small share of vegan toddlers, 3.6%, were classified as stunted against 2.4% of vegetarians and none of the omnivores in that sample, a gap too small on those numbers to call decisive but large enough that a paediatrician should be watching for it rather than assuming it away.

The practical fix is energy density: nut and seed butters, avocado, added oils, and full-fat plant milks rather than the low-fat versions, stirred into a child's food rather than relied on as separate snacks. None of this matters much for an adult eating three or four meals from a large stomach; for a toddler eating six small ones, it is close to the whole game.

The recommended intake rises with age rather than staying fixed: 0.5 micrograms a day as an adequate intake for infants seven to twelve months, 0.9 micrograms as a full RDA at one to three years, and 1.2 micrograms at four to eight years, according to the National Academies' dietary reference intake tables. Symptoms of infant deficiency, when they appear, are not subtle: vomiting, lethargy, failure to thrive, low muscle tone and a stalling or reversal of developmental milestones, typically between two and twelve months of age. Treatment with B12 corrects the blood picture within days, but some case reports describe neurological effects that did not fully resolve, which is why this is framed as a mechanism, not a scare: the deficiency is easy to prevent and hard to fully undo.

Iron from plants is non-heme iron, absorbed at a lower and more variable rate than the heme iron in meat, and its uptake depends heavily on what else is on the plate: vitamin C markedly increases non-heme absorption in the same meal, while the tannins in tea and the phytates in some grains suppress it. The practical rule for a vegan child's plate is to pair an iron source, lentils, tofu, fortified cereal, with a vitamin C source, citrus, peppers, strawberries, at the same sitting rather than relying on the iron content of the day's food in isolation. The iron evidence page covers the absorption science and population-level deficiency risk in full.

Iodine cuts the other way from most nutrients on this list, because a vegan child can end up either short or in excess, and both are thyroid problems. A Czech cross-sectional study found 41.9% of vegan children below adequate urinary iodine status against 19.6% of omnivorous children, but the fix parents reach for first, seaweed, is not simple: iodine content across seaweed species ranges from about 16 to nearly 3,000 micrograms per gram, a spread wide enough that unmonitored seaweed use can just as easily push a child into iodine excess as fix a deficiency. Iodised salt used in cooking and a monitored supplement are the more predictable route; the iodine evidence page has the dosing detail.

Calcium and vitamin D are handled together in practice because a young child gets most of both from milk in a mixed diet, and a vegan child needs both replaced deliberately: a calcium-fortified plant milk or juice at a dose that matches dairy milk's roughly 300 milligrams per cup, plus greens, tofu set with calcium sulfate, and fortified foods, alongside a vitamin D supplement that most paediatric guidance now recommends for breastfed infants of any diet. The calcium evidence page and the site's vitamin D coverage set out the numbers by age. Omega-3 DHA is the remaining gap: a vegan diet supplies the precursor alpha-linolenic acid from flax and walnuts, but the body's conversion of that precursor into DHA is inefficient, and the omega-3 evidence page makes the case that an algae-derived DHA supplement, the same source fish get theirs from, is the more direct route for pregnancy, infancy and early childhood specifically, where DHA supports nervous system development at a pace that later childhood does not repeat.

Breast milk from a vegan mother with adequate B12, iodine and DHA status is nutritionally equivalent to any other breast milk, and breastfeeding is not itself a compromise. The failure mode described above, maternal depletion feeding through to the infant, is specifically about supplementation continuing through lactation, not about the choice to breastfeed. Formula is the sharper constraint: soy-based formula is the only widely available infant formula that is not derived from cow's milk, and it is not vegan in every jurisdiction's regulatory sense because some formulations include vitamin D3 sourced from lanolin, so a parent who wants full certainty needs to check the label rather than assume. Plant milks, oat, almond, rice, pea, are not a substitute for formula or breast milk in the first year at all: they are too low in protein, fat and several micronutrients to support an infant, a point paediatric bodies make regardless of whether the family is vegan.

Weaning follows the same logic as any baby's, offered a few weeks earlier or later depending on the child, but the calorie-density point from the second section starts mattering at exactly this stage: a vegan baby's first foods should include mashed avocado, nut or seed butter thinned with water, and full-fat cooking rather than the pureed vegetables alone that a well-meaning parent might reach for first. Toddlerhood is where B12, iron and calcium tracking becomes routine rather than occasional, and school age is where the main new problem is not nutrient content but logistics, covered next.

A school lunch that travels well and stays vegan without special handling tends to be built from things that are naturally shelf-stable at room temperature for a few hours: hummus and crackers, a bean or lentil salad with an acidic dressing that does not spoil quickly, roasted chickpeas, nut-butter sandwiches where nut policies allow them, and fruit. The energy-density point still applies at five and six years old, so a lunch that is all raw vegetables and fruit will often come home half eaten, not from pickiness but because it did not carry enough calories to get through an active school day.

Fussy eating in a vegan toddler is the same phenomenon as fussy eating in any toddler, a normal developmental phase tied to neophobia rather than to the diet, and the fix is the same too: repeated low-pressure exposure to a food, sometimes a dozen or more tries, rather than substitution or bribery. Birthday parties and other people's houses are a social problem more than a nutritional one, and the approach that keeps a child included rather than singled out is to send a labelled version of whatever is being served, a vegan cupcake alongside the birthday cake, rather than asking a host to rebuild their menu.

A vegan child's growth should be tracked on the same percentile charts as any child's, and more consistently than a family might otherwise bother with, because the failure modes above, low B12, low iron, insufficient calories, show up first as a flattening growth curve before they show up as anything else. A parent who does not intend to track growth and bloodwork with a paediatrician or registered dietitian, particularly through infancy and toddlerhood, is taking on a real risk that this guide cannot talk them out of. The most useful thing to bring to that appointment is not a defence of the diet but a written account of what the child actually eats over a few typical days, since most paediatricians who raise concerns are working from the general population's average vegan diet, not the specific one in front of them, and a concrete food record moves the conversation from ideology to arithmetic fast.

The pregnancy and childhood evidence page covers the clinical trial and cohort evidence behind these positions in depth, and the supplement guide covers the adult version of this decision; neither substitutes for a clinician who has actually examined the child in question, which is the entire point of this section.