Manganese in teeth and neurobehavior: Sex-specific windows of susceptibility

Abstract

Background: Manganese (Mn) is an essential element required for growth and development, but higher body burdens have been associated with neurobehavioral decrements in children.

Objectives: We examined whether prenatal or postnatal Mn measured in deciduous teeth was associated with scores on a test of visuospatial learning and memory.

Methods: Deciduous teeth were collected from 142 participants (ages 10-14years) residing near varied ferro‑manganese industry in Italy. Mn concentrations were measured in prenatal and postnatal tooth regions by laser ablation inductively coupled plasma mass spectrometry (ICP-MS). The Virtual Radial Arm Maze (VRAM), an animal-human analogue task, was used to assess visuospatial learning and memory. We used generalized additive, linear and zero-inflated Poisson mixed regression models to estimate associations between prenatal or postnatal Mn concentrations and repeated measures of all four VRAM outcomes: time, distance, working and reference memory errors. Effect measure modification by sex was examined in stratified models.

Results: U-shaped associations between prenatal Mn and VRAM outcomes were observed among girls only (p_GAMM=0.001 to 0.02 in stratified models). Compared to the mid-tertile of prenatal Mn, girls in the highest tertile took 7.7s [95% CI: -6.1, 21.5] longer to complete the task, traveled 2.3 maze units [0.1, 4.4] farther, and committed more working and reference memory errors (β for count ratio=1.33 [1.01, 1.83]; 1.10 [0.98, 1.24], respectively). This association was not observed among boys. In contrast, for postnatal Mn, no significant associations were found, and patterns were similar for boys and girls.

Conclusions: The prenatal period may be a critical window for the impact of environmental Mn on visuospatial ability and executive function, especially for females.

Keywords: Children; Environmental epidemiology; Manganese; Neurobehavior; Teeth.

Figure 1

Adjusted associations between continuous prenatal tooth Mn concentrations and VRAM measures for girls (A) and boys (B). Results from sex-stratified generalized additive mixed models of VRAM measures of time (T), distance (D), working memory errors (WME), and reference memory errors (RME) as predicted by natural log transformed prenatal tooth manganese concentrations (smoothed; AUC ⁵⁵Mn:⁴³ Ca). Models are adjusted for age, socioeconomic status, Ln transformed blood Pb, videogame use, trial and tooth attrition. Shaded region represents 95% confidence interval. (A) Among girls (n=79): pGAMM <0.01 for T, D and WME; p= 0.69 for RME; (B) Among boys (n=63); T: p=0.37; D: p=0.30; WME: p= 0.58; RME: p=0.66. Tertiles correspond to the following In Mn concentrations: T1: −2.2 to −1.0; T2: −1.0 to −0.7; T3: −0.7 to −0.1 (AUC ⁵⁵Mn:⁴³ Ca).

Figure 2

Adjusted associations between continuous postnatal tooth Mn concentrations and VRAM measures. Results from generalized additive mixed models of VRAM measures of time (T), distance (D), working memory errors (WME), and reference memory errors (RME) as predicted by natural log transformed prenatal tooth manganese concentrations (smoothed; AUC ⁵⁵Mn:⁴³Ca). Models are adjusted for age, socioeconomic status, Ln transformed blood Pb, videogame use, trial and tooth attrition. Shaded region represents 95% confidence interval. (n=137); pGAMMs: T: p=0.09; D: p = 0.33; WME: p=0.88; RME: p=0.06. Tertiles correspond to the following In Mn concentrations: T1: −4.2 to −2.2; T2: −2.2 to −1.9; T3: −1.9 to −0.9 (AUC ⁵⁵Mn:⁴³ Ca).