Does Lead Have a Connection to Autism? A Systematic Review and Meta-Analysis

Abstract

Environmental pollutants, particularly toxic trace metals with neurotoxic potential, have been related to the genesis of autism. One of these metals that stands out, in particular, is lead (Pb). We conducted an in-depth systematic review and meta-analysis of peer-reviewed studies on Pb levels in biological materials retrieved from autistic children (cases) and neurotypical children (controls) in this work. A systematic review was conducted after the careful selection of published studies according to established criteria to gain a broad insight into the higher or lower levels of Pb in the biological materials of cases and controls, and the findings were then strengthened by a meta-analysis. The meta-analysis included 17 studies (hair), 13 studies (whole blood), and 8 studies (urine). The overall number of controls/cases was 869/915 (hair), 670/755 (whole blood), and 344/373 (urine). This meta-analysis showed significantly higher Pb levels in all three types of biological material in cases than in controls, suggesting a higher body Pb burden in autistic children. Thus, environmental Pb exposure could be related to the genesis of autism. Since no level of Pb can be considered safe, the data from this study undoubtedly point to the importance of regularly monitoring Pb levels in autistic children.

Keywords: autism; biological material; lead; meta-analysis; systematic review.

Figure 1

PRISMA flow diagram of literature search and study identification, screening, eligibility, inclusion, and exclusion. Combined data for hair, whole blood, and urine.

Figure 2

Forest plot for random-effects meta-analysis. Differences in Pb levels in hair samples between controls and cases are shown. The size of each square is proportional to the weight of the study. Based on the observed studies, the diamond symbol indicates the pooled overall effect size. Abbreviation: CI, confidence interval.

Figure 3

Funnel plot for assessment of publication bias in in observational studies comparing hair Pb levels of controls and cases. The figure shows the studies’ effect sizes (Hedges’s g) relative to their precision (inverse of SE). Circles represent observed studies. Based on the observed studies, the diamond symbol indicates the pooled overall effect size. Publication Bias—Egger’s regression test: t₁₇ = 3.972, p = 0.001. Begg and Mazumdar rank correlation—Kendall’s τ = −0.389, p = 0.029.

Figure 4

Forest plot for random-effects meta-analysis. Differences in Pb levels in whole blood samples between controls and cases are shown. The size of each square is proportional to the weight of the study. Based on the observed studies, the diamond symbol indicates the pooled overall effect size. Abbreviation: CI, confidence interval.

Figure 5

Funnel plot for assessment of publication bias in observational studies comparing whole blood Pb levels of controls and cases. The figure shows the studies’ effect sizes (Hedges’s g) relative to their precision (inverse of SE). Circles represent observed studies. Based on the observed studies, the diamond symbol indicates the pooled overall effect size. Publication Bias—Egger’s regression test: t₁₃ = 3.040, p = 0.011. Begg and Mazumdar rank correlation: Kendall’s τ = −0.461, p = 0.033.

Figure 6

Forest plot for random-effects meta-analysis. Differences in Pb levels in urine samples between controls and cases are shown. The size of each square is proportional to the weight of the study. Based on the observed studies, the diamond symbol indicates the pooled overall effect size. Abbreviation: CI, confidence interval.

Figure 7

Funnel plot for assessment of publication bias in observational studies comparing urine Pb levels of controls and cases. The figure shows the studies’ effect sizes (Hedges’s g) relative to their precision (inverse of SE). Circles represent observed studies. Based on the observed studies, the diamond symbol indicates the pooled overall effect size. Publication Bias—Egger’s regression test: t₈ = 4.225, p = 0.0056. Begg and Mazumdar Rank Correlation—Kendall’s τ = −0.857, p = 0.003.