“Formula milk feeding causes autism”

A belief circulates in some parenting communities — sometimes framed as a concern, sometimes as a firm claim — that feeding a baby formula instead of breastfeeding raises the risk of autism, or that something in infant formula causes it. This causes significant distress for parents who couldn't or didn't breastfeed. Is it true?

• Formula does not cause autism. No study has established a causal relationship between infant formula feeding and autism spectrum disorder. Autism is rooted in genetics and prenatal brain development — processes that are complete or well underway before a baby is born and before any feeding decision is made.

• Some studies find an association between less breastfeeding and higher autism rates — but the direction is likely reversed. Meta-analyses have found that children with ASD tend to have shorter breastfeeding durations. A 2023 meta-analysis found that not breastfeeding was associated with roughly 1.8 times the odds of autism diagnosis. A 2024 Autism Research study found that both exclusive and partial breastfeeding were associated with lower odds of ASD diagnosis. These numbers sound striking, but they do not mean formula caused the autism.

• The most important confound: autistic infants are harder to breastfeed. This is the key piece most reporting misses. Feeding difficulties are among the earliest observable signs of autism — appearing in infancy, months before any diagnosis. Autistic babies frequently experience dysregulated sucking, weak or inconsistent latch, sensory overload during feeding, and oral motor difficulties. These challenges make breastfeeding harder and shorter, not the other way around. A 2025 preprint and multiple review studies explicitly identify this reverse causation as the most plausible explanation for the observed association.

• Large population studies find no link with formula specifically. A large Korean population-based study of over 150,000 children found no significant association between soy-based infant formula and autism, epilepsy, or ADHD. A foodfacts.org analysis of the broader evidence base concluded the causal claim is unsupported.

• Soy formula — a specific concern. A handful of smaller studies have explored whether soy formula in particular may be associated with seizure risk in autistic children (due to phytoestrogens and other compounds). Evidence is conflicting and preliminary — one PLOS One retrospective study found a possible association with seizures in children with autism and Fragile X syndrome; a large Korean cohort found no such signal. This remains an open question for a very specific subgroup, not a finding that generalises to autism causation.

• Socioeconomic confounding matters. Breastfeeding rates are strongly influenced by socioeconomic factors — income, education, workplace policies, support. Many of these same factors correlate with autism diagnosis rates. Studies that do not fully control for these variables will find associations that reflect social patterns, not feeding biology.

The claim that formula causes autism is a myth. The association between formula feeding and autism exists in the data, but it flows in the opposite direction: autistic neurology makes breastfeeding difficult, so autistic children are more likely to end up on formula — not because formula made them autistic, but because their already-developing autistic nervous system made breastfeeding harder from the start.

This matters enormously for parents. Formula feeding is a safe, nutritionally complete choice. Parents who formula-fed their children bear no responsibility for their child's autism — the neurodevelopmental trajectory was set before the first feed, in genetics and prenatal development.

• Formula is safe. There is no evidence that switching to or continuing formula raises autism risk
• If your child is struggling to breastfeed, feeding difficulties in early infancy can occasionally be an early sign worth discussing with your paediatrician — not because formula is dangerous, but because early support for feeding challenges (from a lactation consultant or OT) can sometimes help identify children who may benefit from early developmental monitoring
• If you feel guilt about not breastfeeding, the research does not support it — autism was not caused by feeding choices

• Not breastfeeding and risk of ASD: a meta-analysis — PMC 2023
• Breastfeeding patterns associated with later ASD diagnosis — Autism Research / Wiley 2024
• Feeding-related early signs of ASD: a narrative review — PMC 2024
• The interplay between breastfeeding and autism spectrum — medRxiv preprint 2025
• Infant formula and autism: why this claim is unsupported — FoodFacts.org
• Neurodevelopmental outcomes in infants fed soy formula — ScienceDirect
• Association of breastfeeding with ASD risk: systematic review and meta-analysis — ScienceDirect