Assessing the causal association between celiac disease and autism spectrum disorder: A two-sample Mendelian randomization approach.
Genetic data show celiac disease does not cause autism, so skip routine celiac screens unless GI signs show up.
01Research in Context
What this study did
The team used a genetics trick called Mendelian randomization. They asked: do genes that raise celiac risk also raise autism risk?
If the answer were yes, celiac disease might cause autism. They pulled large genetic datasets and ran the numbers.
What they found
The genes linked to celiac disease did not predict autism. The study found no causal path between the two conditions.
How this fits with other research
Higgins et al. (2021) looked at maternal gluten antibodies in Swedish mothers. They also saw no autism link, backing up the null result.
Katz et al. (2003) warned that gluten-free diets can leave autistic kids low on amino acids. Abiodun’s finding supports skipping the diet unless true celiac signs appear.
Bravo-Muñoz et al. (2025) found small autism risk bumps from maternal diabetes or PCOS. Those risks are separate from celiac and remain worth watching.
Why it matters
You can stop ordering celiac blood tests for every autistic client. Save the test for kids with stomach pain, poor growth, or other clear GI symptoms. This frees time and money for evidence-based interventions.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Remove celiac labs from your standard autism intake checklist and teach staff to spot red-flag GI symptoms instead.
02At a glance
03Original abstract
The association between celiac disease (CD) and autism spectrum disorder (ASD) remains inconclusive. Reports from different observational studies have become controversial, necessitating exploration of the causal relationship between CD and ASD. To assess true causality, this study used a two-sample Mendelian randomization (MR) analysis to determine the causal association between CD and ASD. Summary-level data from a genome-wide association study (GWAS) of the European population were used to select instrument variables (IVs) at genome-wide significance (p < 5 × 10-8). The strength of IVs was also evaluated with F-statistics. The inverse variance weighted method (IVW) was the primary MR analysis, supported by other MR tests such as the weighted median method and weighted mode. The presence of horizontal pleiotropy was tested with MR-Egger and MR-PRESSO while other sensitivity analyses such as heterogeneity, leave-one-out analysis, and scatterplot were used to assess the validity of our MR results. Our study did not show an association between CD and ASD (OR, 0.994; 95% CI, 0.935-1.057; p = 0.859). There was also no evidence of horizontal pleiotropy (MR-Egger intercept = 0.015; p-value = 0.223) and heterogeneity (Q = 14.029; p-value = 0.051). These results were also complemented by the leave-one-out analyses, forest plot, and scatter plot, which showed that none of the SNPs influenced the result. The result of this study shows that CD is not causally associated with ASD.
Autism research : official journal of the International Society for Autism Research, 2025 · doi:10.1186/s13052-023-01484-x