Association Between Peripheral Blood Levels of Vitamin A and Autism Spectrum Disorder in Children: A Meta-Analysis
Authors:
Wang, N., Zhao, Y. and Gao, J.
Methodology
Study Design: This was a meta-analysis, combining data from multiple existing studies to provide a more robust conclusion.
Data Sources: The researchers searched PubMed, Web of Science, and Embase databases up to November 13, 2020, for relevant studies. They also reviewed reference lists to identify additional records.
Inclusion Criteria:
Studies measuring vitamin A levels in peripheral blood (serum, plasma, or whole blood) of children with ASD.
Clinical studies reporting sufficient data (e.g., vitamin A concentrations).
Exclusion Criteria:
Studies lacking sufficient data, case reports, reviews, abstracts, in vitro studies, or animal studies.
Sample Size: Five records, comprising six studies, were included, involving:
935 children with ASD
516 healthy controls
Statistical Analysis:
Hedges' g was used to measure effect size (the difference in vitamin A levels between ASD and control groups), with a 95% confidence interval (CI).
A random-effects model was applied due to significant heterogeneity among studies.
Subgroup analyses and Galbraith plots were used to explore sources of variability.
Key Findings
Overall Association:
Children with ASD had significantly lower peripheral blood levels of vitamin A compared to healthy controls.
The effect size was moderate: Hedges' g = -0.600 (95% CI: -1.153 to -0.048), with a statistically significant P-value of 0.033.
This indicates a consistent reduction in vitamin A levels among children with ASD across the included studies.
Heterogeneity:
There was high heterogeneity among the studies (I² = 94.838%, P < 0.001), suggesting variability in results that could not be fully explained by the overall analysis alone.
Subgroup Analyses:
By sample type (e.g., serum): No significant association was found in the serum subgroup (Hedges' g = -0.414, 95% CI: -1.035 to 0.208, P = 0.192), and heterogeneity remained high (I² = 95.524%, P < 0.001).
By ethnicity (e.g., Chinese population): No significant association was found in the Chinese subgroup (Hedges' g = -0.677, 95% CI: -1.148 to 0.064, P = 0.073), with persistent heterogeneity (I² = 96.170%, P < 0.001).
These subgroup results may be limited by smaller sample sizes (e.g., only three studies per subgroup).
Galbraith Plot Analysis:
Three studies were identified as major contributors to heterogeneity. When these were excluded, heterogeneity decreased (I² = 42.483%, P = 0.176), and the association remained significant (Hedges' g = -0.162, 95% CI: -0.289 to -0.035, P = 0.012).
Publication Bias:
No obvious publication bias was detected (Egger's test P = 0.269), suggesting the results were not heavily skewed by selective reporting.
Interpretation
The meta-analysis provides evidence of a significant association between lower peripheral vitamin A levels and ASD in children.
This finding supports the idea that nutritional factors, such as vitamin A, might play a role in ASD, possibly due to its importance in brain development and function.
However, the study does not establish causation—it remains unclear whether lower vitamin A levels contribute to ASD or are a consequence of the disorder (e.g., due to dietary preferences common in children with ASD).
Limitations
The authors highlighted several limitations:
Moderate Sample Size: Although the largest meta-analysis on this topic to date, the sample size (1,451 total participants) is still moderate, limiting the strength of the evidence.
Unexplained Heterogeneity: Subgroup analyses failed to fully identify the sources of variability among studies.
Confounding Factors: Factors such as age, sex, specimen collection methods, and other unmeasured variables were not accounted for due to limited data.
Lack of Longitudinal Data: The studies were cross-sectional, so changes in vitamin A levels over time could not be assessed.
Focus on Peripheral Levels: The study did not evaluate other retinoids or whether these children had clinical vitamin A deficiency.
Study Conclusion
The meta-analysis concludes that children with ASD exhibit significantly decreased peripheral vitamin A levels compared to healthy peers, strengthening the evidence of an abnormal micronutrient profile in ASD.
The authors call for further research to:
Confirm these findings with larger sample sizes.
Investigate the causal relationship between vitamin A and ASD.
Explore the clinical implications, such as whether vitamin A supplementation could benefit children with ASD.
