Toddler Diets and Toxic Metals: What a Swedish Study Found

A study of national dietary survey data, published on September 25, 2026, reports that toddlers consuming plant-heavy diets rich in meat substitutes, plant-based milk alternatives, nuts, and seeds show higher urinary levels of cadmium compared to peers eating traditional diets. Researchers from Karolinska Institutet analyzed biomarkers across hundreds of Swedish children, observing that while overall toxic metal concentrations remained low, plant-focused patterns noticeably shifted exposure markers in the youngest age groups.

Key takeaways:

• A September 25, 2026 Karolinska Institutet study found 1.5-year-old toddlers eating plant-based alternatives, nuts, and seeds had elevated urinary cadmium.
• Swedish market basket data cited by the authors put mean cadmium at 11 ug/kg in meat alternatives and 1.9 ug/kg in milk alternatives, versus 1.3 ug/kg in traditional meat and 0.05 ug/kg in lean dairy.
• The cadmium disparity attenuated by age 4 as diets diversified, though high-diversity diets introduced higher markers for arsenobetaine and methylmercury from seafood.
• Toxic metal levels were generally low, no health outcomes were measured, and the paper does not advise parents to avoid plant-based milks, nuts or seeds.

What happened

On September 25, 2026, researchers published an open-access study in the Journal of Exposure Science & Environmental Epidemiology titled "Identifying dietary sources and patterns of exposure to toxic metals and other trace elements in Swedish toddlers and young children" (DOI: 10.1038/s41370-026-00973-5). Led by corresponding author Maria Kippler at the Institute of Environmental Medicine at Karolinska Institutet, along with co-first authors M. R. Islam and M. Trask, the paper evaluated toxic metals and trace elements across young children nationwide.

The project analyzed data collected between 2021 and 2023 from Riksmaten Young Children, a national dietary survey conducted by the Swedish Food Agency with a register-based random sample drawn by Statistics Sweden. Funding came from the Swedish Environmental Protection Agency, the Swedish Research Council Formas, and Karolinska Institutet. The researchers declared no competing interests.

Dietary intake was captured using a two-day web-based food diary converted to habitual daily intake across 14 food groups. The research team then measured toxic metals (arsenic, cadmium, lead, mercury) and essential elements using inductively coupled plasma mass spectrometry (ICP-MS) in blood and urine samples from 305 children aged 1.5 years and 288 children aged 4 years.

What changed

Rather than looking only at individual food items in isolation, the study grouped diets using principal component analysis (PCA) and k-means clustering. At age 1.5 years, urinary cadmium correlated positively with intake of plant-based milk alternatives as well as nuts and seeds. Furthermore, higher adherence to the plant-based alternatives PCA pattern showed a statistically significant link with higher urinary and blood cadmium (p-trend 0.018).

When dividing the 1.5-year-olds into distinct diet clusters, the plant-based alternatives cluster (n = 43) showed higher urinary cadmium compared to the high-diversity cluster (n = 116) and the western-style cluster (n = 146). However, this cluster did not show higher blood cadmium levels.

By age 4, the pattern changed. The plant-based cluster (n = 35) and the western-style cluster (n = 112) showed similar urinary cadmium levels, both remaining higher than the high-diversity cluster (n = 141), with no differences in blood cadmium across the three groups. The study authors highlighted that the link between plant-based alternatives and cadmium was less obvious in 4-year-olds, who consumed more varied overall diets.

Who is affected

The population examined consists of Swedish toddlers aged 18 months and young children aged 4 years. The findings specifically concern young children whose early feeding regimens rely heavily on plant-based milk alternatives, meat substitutes, and frequent portions of nuts and seeds.

The study cited market basket data from the Swedish Food Agency (2022) showing substantial differences in background cadmium levels across food categories: plant-based meat alternatives averaged 11 ug/kg and milk alternatives averaged 1.9 ug/kg, compared to 1.3 ug/kg in conventional meat products and just 0.05 ug/kg in lean dairy. The authors suggest that soy, peas and other legumes take up cadmium from soil, though the study did not test this.

Dietary trade-offs also surfaced in other dietary patterns. Children who consumed canned tuna showed slightly higher blood mercury (median 1.12 vs 0.91 ug/L at 1.5 years; 1.22 vs 0.91 ug/L at 4 years), and those who ate game meat had slightly elevated blood lead (median 7.61 vs 6.35 ug/L at 1.5 years, p = 0.001). Understanding chemical exposures across childhood diets requires monitoring multiple categories, much like recent investigations into how Burritos, Cheese, Fish Tied to PFAS in Kids' Blood: Study revealed overlapping environmental exposures.

What it means for shoppers

For parents and caregivers, this study does not recommend eliminating plant-based foods, legumes, nuts, or seeds. The authors did not assess clinical health outcomes such as kidney function or bone mineral density, and metal concentrations across the study population remained generally low.

Diet variety cut both ways in the data. The high-diversity cluster, rich in fruits, vegetables and seafood, had the lowest urinary cadmium but higher arsenobetaine and methylmercury exposure, alongside better selenium status. Median blood selenium was 103 ug/L at both ages, and the authors wrote that around half of the children appear to have suboptimal selenium status.

Caregivers reviewing ingredient lists should also pay attention to how plant-based alternatives are formulated. For example, texturizers and thickeners like those discussed in Is Carrageenan Safe? Uses, Safety, and Regulatory Status were mentioned by the researchers, who suggested that some additives such as carrageenan can accumulate cadmium. The study did not measure cadmium in any specific product or brand, but an ingredient list does show what a plant-based drink is made from.

The background

The European Food Safety Authority (EFSA) established a tolerable weekly intake (TWI) for cadmium of 2.5 micrograms per kilogram of body weight on March 20, 2009. EFSA noted at the time that vegetarians, children, and populations in contaminated areas could experience dietary exposure up to twice the TWI, identifying cereals, vegetables, nuts, pulses, and potatoes as primary contributors. EFSA says cadmium is primarily toxic to the kidney.

In the United States, the Food and Drug Administration addresses heavy metal contamination in infant and toddler foods through its Closer to Zero initiative. FDA issued final lead action levels for processed baby foods on January 6, 2025. On the program page reviewed (current as of January 2025), cadmium was listed in the "Propose" phase with draft guidance targeted for 2025.

The authors of the Karolinska Institutet study proposed several potential mechanisms for why plant-based clusters showed higher cadmium biomarkers. Beyond the higher baseline cadmium content of legumes, the authors suggested that meat avoidance and subsequent lower iron stores could upregulate cadmium absorption in the gut. Phytate present in nuts and seeds could also influence mineral bioavailability, though these explanations were not tested in the study.

What to watch next

The next US marker to watch is FDA's cadmium guidance under Closer to Zero; the program page reviewed does not confirm whether that draft has been issued.

There are clear limits to what this single publication establishes. The investigation was cross-sectional, which precludes establishing temporality or drawing causal conclusions. In addition, the 1.5-year-old plant-based cluster was small, with only 43 children in the diet group and 65 toddlers providing urine samples across the entire 1.5-year cohort, which reduced statistical power and overall generalizability. The study analyzed broad food categories rather than specific commercial brands or milk types (such as oat versus almond versus soy), did not calculate whether individual children exceeded EFSA's tolerable intake, and took place entirely in Sweden.