The central methodological innovation of the work lies in its treatment of exposure. Most earlier studies examined metals one at a time, an approach that obscures the reality that people are exposed to complex cocktails of elements simultaneously through air, food, water, and consumer products. To address this, the research team, led by Maoling Fu of Hunan University of Medicine together with colleagues at Tongji Hospital and the PLA General Hospital, deployed four complementary analytical strategies: conventional logistic regression for individual metals, weighted quantile sum regression, quantile-based g-computation, and Bayesian kernel machine regression, commonly abbreviated as BKMR. Each of the mixture methods weighs the relative contribution of every component metal to the overall effect, allowing the researchers to identify which elements drive the association and whether the combined effect is greater than the sum of its parts.

The single-metal results already revealed a consistent pattern. Blood concentrations of cadmium, lead, and manganese were each significantly associated with obesity risk, as were urinary concentrations of cadmium, lead, cobalt, cesium, antimony, and tungsten. Cadmium and lead are classic toxic heavy metals with no physiological function, while manganese is an essential trace element that becomes harmful at excessive levels. The appearance of manganese as a significant player is particularly noteworthy because the element is required for normal brain development and enzyme function, yet disturbances of manganese homeostasis have been linked in prior research to neurotoxicity and, in animal studies, to mitochondrial damage in liver tissue. A previous population-based study had also reported associations between blood manganese levels and visceral adipose tissue in adults, lending plausibility to the new pediatric findings.

When the metals were analyzed as mixtures, the blood results pointed in one clear direction. All three mixture methods, weighted quantile sum regression, quantile-based g-computation, and Bayesian kernel machine regression, converged on a significant positive association between the blood metal mixture and obesity risk, meaning that children with higher overall burdens of these circulating metals were more likely to be obese. Within that mixture, manganese emerged as the metal with the strongest contribution. This convergence across statistically distinct approaches matters because each model makes different assumptions about how mixture components interact; agreement among them substantially strengthens confidence that the observed association is not an artifact of any single modeling choice.

The urinary findings, by contrast, ran in the opposite direction. The urinary metal mixture was significantly negatively associated with obesity risk, with cadmium, lead, and cobalt showing the strongest negative contributions. At first glance this seems paradoxical: how can the same metals be positively associated with obesity when measured in blood and negatively associated when measured in urine? The authors and the broader literature suggest several possible explanations. Urinary metal concentrations reflect both exposure and excretion, and dilution effects may play a role, since obese individuals can differ in hydration status and renal function. Alternatively, the negative urinary association may reflect differences in how the body handles these metals, with obese children excreting them more efficiently, or it may indicate reverse causation, in which obesity-related physiological changes alter metal kinetics rather than metals driving obesity.

Stratified analyses added another layer of complexity. The associations between metal exposure and obesity risk were particularly pronounced among male children, suggesting that sex may modify the relationship between metal burdens and adiposity. Sex differences in body composition are well documented in pediatric populations, with boys and girls differing in visceral, subcutaneous, and total body fat distribution, and sex hormones and growth hormone axis activity are known to influence energy balance. Biological plausibility for sex-specific metal effects also exists: experimental work has shown, for example, that subacute cadmium exposure produces different metabolic changes in female rats, and cobalt treatment alters lipid metabolism in white adipose tissue in diet-induced obesity models. The new findings hint that such sex-dependent mechanisms may operate in humans as well.

The biological pathways by which heavy metals could promote obesity are an active area of investigation. Cadmium has been shown in cell studies to inhibit the differentiation of preadipocytes through the C/EBPα and PPARγ pathways, key regulators of fat cell formation, and white adipose tissue itself has been identified as a target organ for cadmium toxicity. Lead intoxication is known to disrupt endocrine function, and both cadmium and lead have been linked to obesity in adult populations in studies from Korea, China, and the United States. Manganese, meanwhile, may act through oxidative stress and through effects on the hypothalamic-pituitary-adrenal axis, a central regulator of appetite and energy balance. Oxidative stress, the imbalance between reactive oxygen species and antioxidant defenses, is a recognized consequence of redox-active metals and has been correlated with intra-abdominal fat in obese males, providing a plausible mechanistic bridge between metal exposure and fat accumulation.

The study builds on a growing body of environmental epidemiology that treats obesity not solely as a consequence of diet and physical inactivity but as a condition influenced by the chemical environment. Earlier analyses of NHANES data in adults found that cumulative exposure to heavy metal mixtures was associated with obesity and its comorbidities, and a systematic review and meta-analysis published in 2024 concluded that heavy metal exposure is associated with obesity more broadly. Pediatric evidence had been thinner, however, with some prior studies of urinary metals in children reporting mixed results. By combining a recent survey window with multiple mixture models and stratification by sex and age, the new analysis offers one of the most comprehensive pictures to date of how the metal exposome relates to childhood adiposity, and the authors note that this is the first use of multiple statistical models to confirm the combined effects of metal mixtures on childhood obesity.

Several caveats temper the conclusions. As a cross-sectional study, the analysis captures a snapshot in time and cannot establish whether metal exposure preceded the development of obesity; reverse causation remains a genuine possibility, particularly for the counterintuitive urinary findings. Measurement of metals in single blood and urine samples also provides only an estimate of longer-term exposure, and unmeasured confounders ranging from diet quality to socioeconomic circumstances could influence both metal burdens and body weight. The authors themselves are explicit that further research is needed to confirm the findings, ideally through longitudinal cohorts that track metal exposure and weight change over time in the same children.

Nevertheless, the implications are significant for public health. Childhood obesity substantially increases the risk of cardiovascular disease, type 2 diabetes, and psychosocial difficulties, and interventions that reduce environmental metal exposure, through cleaner air and water, safer consumer products, and reduced industrial emissions, are already proven public health wins for other reasons. If the associations observed here are confirmed causally, the same measures could carry an added benefit for metabolic health in children, particularly boys who appear most vulnerable. The study also highlights the value of mixture-based epidemiology: in the real world, no child is exposed to a single metal in isolation, and understanding how cadmium, lead, manganese, cobalt, and their chemical neighbors act together may be essential to grasping, and ultimately preventing, the environmental roots of the obesity epidemic.

Subject of Research: Associations between blood and urinary heavy metal exposure and childhood obesity risk in the U.S. NHANES 2015-2020 population

Article Title: Associations of blood and urinary heavy metals with obesity risk among children: NHANES 2015-2020

Article References: Fu, M., Xiang, Y., Li, X., & Wang, R. (2026). Associations of blood and urinary heavy metals with obesity risk among children: NHANES 2015-2020. Pediatric Research. https://doi.org/10.1038/s41390-026-05488-1

Image Credits: AI Generated

DOI: 10.1038/s41390-026-05488-1

Keywords: childhood obesity, heavy metals, cadmium, lead, manganese, cobalt, NHANES, metal mixtures, Bayesian kernel machine regression, environmental epidemiology, pediatric health, obesity risk