Zinc-to-copper imbalance in the urinary trace element profile of children with autism spectrum disorder

Abstract Autism spectrum disorder (ASD) is associated with excitatory-inhibitory imbalance, oxidative stress, immune dysregulation, and mitochondrial dysfunction, all of which depend on tightly regulated trace-element homeostasis. Comprehensive, quality-controlled urinary trace-element profiling in well-matched ASD cohorts remains limited. Thirty trace elements were measured by inductively coupled plasma mass spectrometry (ICP-MS) in first morning urine from 241 children screened for the study. Elements were tiered by the proportion of values below the limit of detection (LOD): 17 elements with < 20% censored values formed the primary panel, 2 (copper and tin) with 20–50% censoring formed the secondary panel, and 11 elements with > = 50% censoring were reported descriptively only. The primary analysis used 52 age- (± 1 year) and sex-matched ASD-control pairs aged 4–14 years, constructed by maximum-cardinality bipartite matching. Group comparisons used Mann–Whitney U and Wilcoxon signed-rank tests with Benjamini-Hochberg false discovery rate (FDR) correction; effect sizes were expressed as rank-biserial correlations. Pre-specified secondary analyses were limited to the Zn: Cu, Se: Hg and Ca: Mg ratios. Exploratory analyses included sex-stratified comparisons, Spearman element-element correlations, unsupervised clustering, permutation testing and Bayesian estimation; sensitivity analyses included creatinine-normalised and covariate-adjusted analyses. No individual element differed between ASD and controls after FDR correction (smallest adjusted p = 0.64); barium showed the largest single-element reduction (− 33.0%, raw p = 0.071). Total toxic and essential element sums and their ratio did not differ between groups ( p = 0.68, 0.15 and 0.25). The Zn: Cu ratio was 31.8% lower in ASD (median 36.4 vs. 53.4; Mann–Whitney p = 0.033; paired Wilcoxon signed-rank p = 0.021; permutation p = 0.019; rank-biserial r = − 0.24, 95% CI − 0.44 to − 0.03), whereas zinc (− 16.3%, p = 0.25) and copper (+ 11.1%, p = 0.30) did not differ individually. Adjusted for sex and age across all 241 children, the ratio was 24.2% lower in ASD (95% CI − 37.4 to − 8.2%, p = 0.005). A nominally significant reduction in the Zn: Cu ratio was confined to boys (− 28.3%, p = 0.017), whereas the reduction in girls was not significant (− 20.1%, p = 0.51); however, the sex-by-group interaction was non-significant ( p = 0.19), and sex-specificity was not established. The ratio finding did not survive FDR correction across the three a priori ratios ( p = 0.098), a Bayes factor of 0.91 indicated no decisive evidence for either hypothesis, and the difference attenuated when the 32% of copper values below the LOD were excluded ( p = 0.20). Unsupervised clustering did not separate ASD from control children. The study used spot morning urine rather than 24-hour collections, the matched cohort of 52 pairs is powered only for medium effects, approximately one third of copper values fell below the analytical limit of detection, and no independent replication cohort was available. Within this urinary panel, ASD was associated with a modest, selective reduction in the Zn: Cu ratio rather than a generalised elevation or depletion of trace elements. Because urinary excretion need not reflect tissue burden, these data neither support nor exclude the tissue-level bioaccumulation of mercury or lead reported elsewhere in the ASD literature. The reduced Zn: Cu ratio is best regarded as a hypothesis-generating signal that requires independent replication before any biomarker application can be considered.