Are heavy metals in baby food a real risk, and how can you reduce exposure?

Arsenic, lead, cadmium and mercury occur naturally in soil and water and also come from human activity, so crops take them up. Babies are more vulnerable than adults because they eat more food for their body weight and their brains are still developing¹⁰.

What testing shows. In 2021 a US House subcommittee published internal test results from four of the seven baby-food makers it asked. They included finished foods with up to 180 ppb inorganic arsenic and 641 ppb lead, and several companies tested only ingredients, not finished products⁹. Independent testing of 564 products found lead in 37% and cadmium in 57%; organic certification made no difference, and rice-containing products were higher⁷.

What the rules say. The FDA's lead action levels for packaged foods for children under two, finalized in January 2025, are 10 ppb for most foods and 20 ppb for single-ingredient root vegetables and dry infant cereals⁶. Infant rice cereal has had a 100 ppb inorganic arsenic action level since 2020^9,2. Action levels for cadmium and inorganic arsenic in other baby foods are a 2026 FDA priority¹². These are nonbinding, and the FDA recommends a varied, nutrient-dense diet rather than avoiding particular foods⁶.

Is it harming children? Not shown for baby food itself. The best evidence comes from higher-exposure settings: across 1,333 children, IQ fell by about 3.9 points as blood lead rose from 2.4 to 10 µg/dL⁸, and a meta-analysis linked arsenic to lower IQ but found little data on cadmium¹¹.

How to lower exposure

• Vary grains and foods. Among infants who ate no seafood, those who ate rice cereal had more than three times the urinary arsenic of those who ate no rice¹. Oat, barley and multigrain cereals are easy swaps, and the FDA advises variety to limit exposure to any one contaminant².
• Cook rice in extra water. Draining it removes about 40% to 60% of inorganic arsenic, though it also washes out added iron and folate³; parboiling first and finishing with fresh water removed 54% to 73%⁴. See rice safety and reducing exposure.
• Look at lead sources at home. Common sources include soil, water and paint in older homes⁵. See lead in tap water and cadmium in food.
• Ask about blood lead tests. Routine testing is how food-related lead exposure was found in 2023⁵.

The House investigation

In November 2019 the subcommittee asked seven manufacturers for internal test data. Four responded and three refused⁹. The report cited one company's finished foods testing up to 180 ppb inorganic arsenic (over 25% of its tested products exceeded 100 ppb), another's finished foods up to 129 ppb, a finished product with 641 ppb lead, and cadmium above 5 ppb in 75% of one brand's carrots, up to 87 ppb. One company's own analysis showed finished foods held 28% to 93% more inorganic arsenic than its ingredient testing predicted⁹. The limits matter: the data were company-supplied, mixed ingredients with finished foods, and reported maxima rather than typical levels. Comparing them with bottled-water limits shows scale but is not a food standard. Bair's review notes that the House reports suggest subpar testing, lenient or absent standards and limited oversight let these contaminants persist¹⁰.

Independent testing

Gardener and colleagues tested 564 US baby foods across formula, cereals, meals, juices, pouches and snacks⁷. Lead was detected in 37% (maximum 183.6 µg/kg; 75th percentile 5.6) and cadmium in 57% (maximum 103.9 µg/kg; 75th percentile 9.5). Against FDA lead guidelines, 1% of solid foods exceeded the limit at two servings; against California's stricter Proposition 65 limits, 26% did. Certified-organic products were not lower, and rice products were higher for both metals. It was a convenience sample, so it shows what exists rather than market averages.

Rice and infant arsenic exposure

Karagas followed 759 infants in the New Hampshire Birth Cohort Study. About 80% met rice cereal during the first year, and 32.6% of 12-month-olds ate rice snacks. Among 12-month-olds who ate no seafood, geometric mean urinary arsenic was 9.53 µg/L with rice cereal, 4.97 with rice snacks and 2.85 with no rice (all P < .01). Infant rice snacks contained 36 to 568 ng/g arsenic, 5 to 201 ng/g of it inorganic¹. Urine is a snapshot, and no health outcomes were measured.

Does it affect development?

Lanphear pooled seven cohorts (1,333 children). Using a log-linear model, a rise in blood lead from 2.4 to 30 µg/dL went with a 6.9-point IQ decrement (95% CI 4.2 to 9.4), and 3.9 points (95% CI 2.4 to 5.3) between 2.4 and 10 µg/dL. The slope was steeper at lower levels (p = 0.015)⁸. Blood lead reflects every source, not food alone. Rodríguez-Barranco's review covered 41 studies (18 arsenic, 6 cadmium, 17 manganese): 13 of 18 arsenic studies found neurodevelopmental harm, a meta-analysis tied a 50% rise in urinary arsenic to a 0.4-point lower IQ at ages 5 to 15, and only two cadmium studies found any association¹¹. These studies measured exposure from all sources, not baby food specifically, and we found no study linking typical baby-food intake to a developmental outcome.

Cooking and lead events

Gray's lab study found excess-water cooking cut inorganic arsenic by about 40% in polished, 60% in parboiled and 50% in brown rice, but removed 50% to 70% of added iron, folate, thiamin and niacin from enriched polished and parboiled rice; rinsing alone barely helped³. Menon found parboil-and-absorb removed 73% (white) and 54% (brown) of inorganic arsenic without significant zinc loss⁴. Neither tested infants.

Troeschel reported 566 lead cases across 44 states after contaminated cinnamon in applesauce pouches; 55% were in children under 2, with a median peak blood lead of 7.2 µg/dL (range 3.5 to 39.3)⁵. That was adulteration, not typical baby-food composition, but routine testing caught it.

Bottom line (confidence 3 of 5)

Metals in baby food are real and measurable, and exposure-reduction steps work. Whether typical levels harm development is unproven, so cut the avoidable exposures and keep feeding a varied diet.