Intellectual impairment in school-age children exposed to manganese from drinking water

Abstract

Background: Manganese is an essential nutrient, but in excess it can be a potent neurotoxicant. Despite the common occurrence of manganese in groundwater, the risks associated with this source of exposure are largely unknown.

Objectives: Our first aim was to assess the relations between exposure to manganese from drinking water and children's intelligence quotient (IQ). Second, we examined the relations between manganese exposures from water consumption and from the diet with children's hair manganese concentration.

Methods: This cross-sectional study included 362 children 6-13 years of age living in communities supplied by groundwater. Manganese concentration was measured in home tap water (MnW) and children's hair (MnH). We estimated manganese intake from water ingestion and the diet using a food frequency questionnaire and assessed IQ with the Wechsler Abbreviated Scale of Intelligence.

Results: The median MnW in children's home tap water was 34 µg/L (range, 1-2,700 µg/L). MnH increased with manganese intake from water consumption, but not with dietary manganese intake. Higher MnW and MnH were significantly associated with lower IQ scores. A 10-fold increase in MnW was associated with a decrease of 2.4 IQ points (95% confidence interval: -3.9 to -0.9; p < 0.01), adjusting for maternal intelligence, family income, and other potential confounders. There was a 6.2-point difference in IQ between children in the lowest and highest MnW quintiles. MnW was more strongly associated with Performance IQ than Verbal IQ.

Conclusions: The findings of this cross-sectional study suggest that exposure to manganese at levels common in groundwater is associated with intellectual impairment in children.

Figure 1

Distribution of MnH by quintiles of (A) estimated manganese intake from water consumption (n = 302), and (B) estimated manganese intake from the diet (n = 288). (Central bar: 50th percentile; lower and upper bounds of the rectangle: 25th and 75th percentiles; lower and upper tails: 5th and 95th percentiles. Observations outside the 95% CIs are not shown.)

Figure 2

Mean Full Scale IQ (± SE), with respect to manganese exposure indicators, adjusted for covariates in model B. (A) IQ is plotted by median of tap water manganese concentration (μg/L) quintiles. The medians and ranges of MnW are as follows: 1st quintile (lowest), 1 (0–2); 2nd, 6 (3–19); 3rd, 34 (20–66); 4th, 112 (67–153); and 5th (highest), 216 (154–2,700). (B) IQ is plotted by median of manganese intake from water consumption (μg/kg/month) quintiles. The medians and ranges of manganese intakes are as follows: 1st quintile (lowest), 0.1 (0–0.7); 2nd, 1.6 (0.71–3.8), 3rd, 7.6 (4.0–14.6), 4th, 39.4 (15.6–82.9), and 5th (highest), 172 (83.3–945). (C) IQ is plotted by median of MnH (μg/g) quintiles. The medians and ranges of MnH are as follows: 1st quintile (lowest), 0.2 (0.1–0.3); 2nd, 0.4 (0.31–0.5); 3rd, 0.7 (0.51–0.9); 4th, 1.2 (0.91–1.9); and 5th (highest), 3.2 (1.91–20.7).