Hair Calcium Levels Are Linked to Gut Microbiome Shifts and Behavior in Autistic Children

A preliminary study published in the journal Metabolites has found that hair calcium levels in children with autism spectrum disorder (ASD) correlate with measurable differences in their gut microbial communities and with behavioral severity scores across multiple domains. The findings, from a team led by Li J and colleagues, are observational and cannot establish cause and effect, but they add a new layer to the growing body of work exploring how nutritional factors might interact with gut microbiota dysbiosis and neurodevelopment in children with ASD.

Hair Calcium Scores Tied to More Severe ASD Behavioral Symptoms

The research team measured calcium in scalp hair using inductively coupled plasma mass spectrometry (ICP-MS), a technique that captures longer-term mineral exposure rather than day-to-day fluctuations in the blood. Among 183 children with confirmed ASD diagnoses, higher hair calcium was positively associated with worse scores on the Childhood Autism Rating Scale (CARS) and on the Autism Treatment Evaluation Checklist (ATEC), particularly in the Speech/Language/Communication subdomain (ATEC-1) and the Sensory/Cognitive Awareness subdomain (ATEC-3). One specific ATEC-3 item, "draws or scribbles," remained significantly associated with hair calcium even after Benjamini-Hochberg false discovery rate correction for multiple testing.

The authors are careful to note that hair calcium reflects peripheral mineral deposition, not blood calcium or overall systemic calcium status. In other words, a high hair calcium reading does not mean a child has too much calcium in their diet or bloodstream; it is a proxy measure with its own interpretation limits. None of the participants had received calcium supplementation, and children who had taken antibiotics, probiotics, or prebiotics within four weeks of sample collection were excluded.

63 Gut Microbial Species Differ Between High- and Low-Calcium Groups

To examine the microbiome angle, the team divided participants into low-calcium (lowest quartile) and high-calcium (highest quartile) groups and performed shotgun metagenomic sequencing on fecal samples using the Illumina HiSeq X Ten platform. Overall gut microbial diversity, measured by Shannon index, species richness, and Bray-Curtis beta diversity, did not differ significantly between the two groups. What did differ was the relative abundance of specific species: 63 microbial species showed significantly different abundances between the groups, with 36 enriched in the high-calcium group and 27 in the low-calcium group.

The high-calcium group was enriched in species related to Clostridium, Eisenbergiella, Enterococcus, Blautia, and Faecalicatena. The low-calcium group showed higher abundances of species related to Mediterraneibacter, Enterobacter, Bifidobacterium, Parasutterella, and certain other Clostridium taxa. The same ANCOM-BC2 differential abundance analysis also identified 22 MetaCyc metabolic pathways that differed between groups: 19 were enriched in the high-calcium group, including pathways for aromatic compound degradation, nucleotide degradation, amino acid metabolism, and fatty acid biosynthesis. The low-calcium group was enriched in three pathways, including pyruvate fermentation to butanoate and histidine biosynthesis, both of which involve metabolites with known roles in gut physiology and, potentially, neural signaling.

Specific species also associated independently with behavioral scores. Collinsella aerofaciens was positively associated with ATEC-1 (Speech/Language/Communication) and ABC scores. Alistipes onderdonkii and Bacteroides nordii were negatively associated with selected ATEC subdomains or total scores. Metabolic pathways including TCA cycle IV and pyruvate fermentation to propanoate I were linked to behavioral measures such as ATEC-2, ATEC-4, and CARS.

A Cross-Sectional Design That Cannot Prove Causality

The study's core methodological limit is its cross-sectional, single-cohort design. Because all measurements were taken at one point in time, there is no way to determine whether calcium-associated microbial differences preceded, followed, or simply co-occurred with differences in behavioral severity. The authors adjusted for a wide range of covariates, including age, sex, gastrointestinal symptoms, picky eating, sleep disturbance, hyperactivity, allergy history, and mood variability, but residual confounding remains possible.

A meaningful sex imbalance between the two calcium groups adds another caution: the low-calcium group contained 36 males out of 46 participants, compared to only 20 males out of 46 in the high-calcium group (p = 0.001). Although sex was included as a covariate in the analyses, this imbalance warrants careful interpretation. The study also lacked comprehensive dietary intake data, which would be needed to understand why hair calcium levels differed in the first place. Shotgun metagenomics measures what metabolic pathways the microbial community is genetically capable of running, not what metabolites it is actually producing; direct metabolomic measurements would be needed to confirm the functional implications. As researchers studying the gut microbiome and autism have repeatedly observed, inconsistent gut microbiome results across autism studies reflect both the disorder's heterogeneity and the methodological variability across cohorts.

Why the Calcium-Microbiome Connection in ASD Deserves Further Study

Calcium is not usually the first nutrient that comes to mind when thinking about gut bacteria, but the mineral plays roles in neuronal excitability, synaptic transmission, vitamin D metabolism, immune regulation, and metabolic homeostasis, all systems that are relevant to both gut microbial ecology and neurodevelopment. Children with ASD frequently show altered dietary patterns and gastrointestinal symptoms, which could plausibly influence mineral absorption and deposition patterns. The Sankey diagram the authors use to visualize their findings draws multi-level associations connecting microbial taxa, metabolic pathways, and behavioral severity scores, illustrating how domain-specific patterns, where one bacterium links to communication scores and another to sensory scores, are consistent with the clinical heterogeneity that makes ASD so difficult to study.

Fermentation-derived metabolites such as propionate and butyrate, which featured in the low-calcium-enriched pathways, have been implicated in host metabolism and nervous system function in prior animal and human research. Biogenic amines and nucleotide pathways, also represented in the differential pathway analysis, may affect neurotransmitter precursors and cellular energy balance. Whether these pathway-level differences translate to meaningful metabolite differences in the blood or brain of the children studied remains an open question.

What Researchers Need to Do Next

The authors call for independent cohort replication, longitudinal designs that can establish temporal ordering, and metabolomic studies that would move from measuring functional genetic potential in the microbiome to measuring actual metabolite concentrations. Broader age ranges and typically developing comparison groups would strengthen any follow-up work. Microbiota-based interventions for autism, including fecal microbiota transplant, probiotics, and dietary modification, are already being explored in clinical research; understanding how host mineral status shapes the microbial environment could eventually inform how such interventions are designed or personalized. For now, this study is best read as hypothesis-generating: it identifies a set of calcium-associated microbial and metabolic features worth tracking in larger, better-controlled studies rather than a therapeutic target ready for clinical translation.

Evidence rating

INVESTIGATIONAL — 80/100

Caveats:

• Cross-sectional design: this study shows associations only and cannot establish whether calcium, gut microbiota, or behavior causally influence each other.
• Hair calcium reflects peripheral mineral deposition, not blood calcium or systemic nutritional status — results should not be interpreted as evidence about dietary calcium intake.
• Findings are from a single cohort of Chinese children with ASD and have not been independently replicated.
• A significant sex imbalance existed between the low- and high-calcium groups; despite statistical adjustment, residual confounding cannot be ruled out.
• Metagenomic sequencing measures functional potential, not actual metabolite production — pathway findings require metabolomic validation.
• Funding source was not disclosed by the authors.

Cross-sectional cohort study (n=183 children with ASD) using shotgun metagenomic sequencing and hair calcium measured by ICP-MS. The design establishes associations among hair calcium, gut microbial species, metabolic pathways, and behavioral severity scores, but cannot determine temporal ordering or causality. Findings are from a single cohort and have not yet been independently replicated.

References

1. Jialin Li, Xinjie Xu, Huinuo Wang, Rui Gao, Bing Li, Xin You (2026). Relationship Between Calcium and Gut Microbial Composition and Metabolic Pathways in Children with Autism. Metabolites. https://doi.org/10.3390/metabo16060405
2. Federico Anaclerio, Maria Minelli, Ivana Antonucci, Valentina Gatta, Liborio Stuppia (2024). Microbiota and Autism: A Review on Oral and Gut Microbiome Analysis Through 16S rRNA Sequencing. Biomedicines. https://doi.org/10.3390/biomedicines12122686
3. Jennifer Fouquier, Nancy Moreno Huizar, Jody Donnelly, et al. (2021). The Gut Microbiome in Autism: Study-Site Effects and Longitudinal Analysis of Behavior Change. mSystems. https://doi.org/10.1128/msystems.00848-20
4. Evelyn Takyi, Khemlal Nirmalkar, James Adams, Rosa Krajmalnik-Brown (2025). Interventions targeting the gut microbiota and their possible effect on gastrointestinal and neurobehavioral symptoms in autism spectrum disorder. Gut Microbes. https://doi.org/10.1080/19490976.2025.2499580