Skip to content

Environmental Health Perspectives

Facebook Page EHP Twitter Feed Open Access icon  

Research Volume 125 | 2017

Environ Health Perspect; DOI:0.1289/EHP631

Manganese in Drinking Water and Cognitive Abilities and Behavior at 10 Years of Age: A Prospective Cohort Study

Syed Moshfiqur Rahman,1,2* Maria Kippler,1* Fahmida Tofail,2 Sven Bölte,3,4 Jena Derakhshani Hamadani,2 and Marie Vahter1
Author Affiliations open

1Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden

2International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b), Dhaka, Bangladesh

3Center of Neurodevelopmental Disorders (KIND), Pediatric Neuropsychiatry Unit, Department of Women’s and Children’s Health, Karolinska Institutet, Stockholm, Sweden

4Center for Psychiatry Research, Stockholm County Council, Stockholm, Sweden

PDF iconPDF Version (191 KB)

  • Background:
    Cross-sectional studies have indicated impaired neurodevelopment with elevated drinking water manganese concentrations (W-Mn), but potential susceptible exposure windows are unknown.
    We prospectively evaluated the effects of W-Mn, from fetal life to school age, on children’s cognitive abilities and behavior.
    We assessed cognitive abilities and behavior in 1,265 ten-year-old children in rural Bangladesh using the Wechsler Intelligence Scale for Children (WISC-IV) and the Strengths and Difficulties Questionnaire (SDQ), respectively. Manganese in drinking water used during pregnancy and by the children at 5 y and 10 y was measured using inductively coupled plasma mass spectrometry.
    The median W-Mn was 0.20 mg/L (range 0.001–6.6) during pregnancy and 0.34mg/L (<0.001–8.7) at 10 y. In multivariable-adjusted linear regression analyses, restricted to children with low arsenic (As) exposure, none of the W-Mn exposures was associated with the children’s cognitive abilities. Stratifying by gender (p for interaction in general <0.081) showed that prenatal W-Mn (3 mg/L) was positively associated with cognitive ability measures in girls but not in boys. W-Mn at all time points was associated with an increased risk of conduct problems, particularly in boys (range 24–43% per mg/L). At the same time, the prenatal W-Mn was associated with a decreased risk of emotional problems [odds ratio (OR)=0.39 (95% CI: 0.19, 0.82)] in boys. In girls, W-Mn was mainly associated with low prosocial scores [prenatal W-Mn: OR=1.48 (95% CI: 1.06, 1.88)].
    Elevated prenatal W-Mn exposure was positively associated with cognitive function in girls, whereas boys appeared to be unaffected. Early life W-Mn exposure appeared to adversely affect children’s behavior.
  • Received: 19 January 2016
    Revised: 08 June 2016
    Accepted: 23 June 2016
    Published: 26 May 2017

    Corresponding author: Address correspondence to M. Kippler, Institute of Environmental Medicine, Karolinska Institutet, Box 210, SE-171 77, Stockholm, Sweden. Telephone: +46 8 524 87407, Fax: +468336981, E-mail:

    Supplemental Material is available online (

    The authors declare that they have no actual or potential competing financial interests.

    Note to readers with disabilities: EHP strives to ensure that all journal content is accessible to all readers. However, some figures and Supplemental Material published in EHP articles may not conform to 508 standards due to the complexity of the information being presented. If you need assistance accessing journal content, please contact Our staff will work with you to assess and meet your accessibility needs within 3 working days.

  • PDF icon Supplemental Material PDF (252 KB)

    Note to readers with disabilities: EHP has provided a 508-conformant table of contents summarizing the Supplemental Material for this article (see below) so readers with disabilities may determine whether they wish to access the full, nonconformant Supplemental Material. If you need assistance accessing journal content, please contact Our staff will work with you to assess and meet your accessibility needs within 3 working days.
    PDF icon Supplemental Table of Contents PDF (166 KB)


Manganese (Mn) is an essential element that functions as a cofactor in a number of enzymes and in certain antioxidants, which makes it important during early life development (Mistry and Williams 2011). The primary source of Mn is the diet, which usually provides the required 3.0mg/d for pregnant women and 0.5–2.0 mg/d for children (EFSA 2013). Excess exposure through drinking water is common worldwide (Frisbie et al. 2012; Ljung and Vahter 2007), and there is increasing concern that such exposure may adversely affect the central nervous system, particularly in children (Frisbie et al. 2012; Zoni and Lucchini 2013). Several cross-sectional studies have indicated that elevated Mn concentrations in drinking water (W-Mn) are associated with impaired cognitive abilities, adaptive behaviors, or both in 6- to 13-year-old children (Bouchard et al. 2011; Khan et al. 2011; Oulhote et al. 2014a; Wasserman et al. 2006). Additional studies indicate associations with blood Mn, often with an inverted U-shaped dose–effect relationship (Sanders et al. 2015). However, there appears to be no association between Mn in water and that in blood (Ljung et al. 2009; Rahman et al. 2015).

The timing of exposure may be critical for Mn neurotoxicity because the susceptibility of the brain to toxic insult is known to vary during different phases of development (Grandjean and Landrigan 2014). In addition, exposure may vary over time. Because Mn easily passes through the placenta (Erikson et al. 2007), elevated maternal exposure during pregnancy, combined with increased gastrointestinal absorption (Takser et al. 2004), may lead to excess fetal exposure. Indeed, a few studies have indicated inverse associations between Mn concentrations in umbilical cord blood and child neurodevelopment (Takser et al. 2003; Yu et al. 2014). Similarly, Mn in tooth dentin measured using microspatial analysis, which estimated prenatal and early postnatal exposure, was associated with adverse neurodevelopmental outcomes (Gunier et al. 2015; Mora et al. 2015). In another study that measured Mn levels in pulverized whole teeth, no association of Mn with neurodevelopmental outcomes was observed (Chan et al. 2015). Early childhood may be another critical period of Mn exposure because regulation of intestinal absorption and biliary excretion is not yet fully developed (Erikson et al. 2007). Although breast milk contains very little Mn (Ljung et al. 2009), introduction of food and drinking water may lead to high-level exposure. The present study aimed to prospectively evaluate potential adverse effects of elevated W-Mn, from fetal life to school age, on cognitive abilities and behavior in a large cohort of boys and girls at 10 y of age. Elevated Mn concentrations were found mainly in medium-deep wells, many of which were constructed to decrease exposure to arsenic (As), which was present in many shallow wells (Kippler et al. 2016; Ljung et al. 2009).

Materials and Methods

Study Area and Population

The study involves a large mother–child cohort, covering early pregnancy to 10 y of age. The study was initially nested in a randomized food and micronutrient supplementation trial (MINIMat) conducted in pregnant women living in Matlab, in rural Bangladesh (Persson et al. 2012), to evaluate the health effects of early life exposure to As, which occurs frequently in shallow wells (Vahter et al. 2006). The installation of deeper wells, often with the aim to decrease As exposure, has resulted in elevated W-Mn (Kippler et al. 2016; Ljung et al. 2009), which motivated us to evaluate the potential health consequences. The 1,607 singleton children born within the MINIMat trial between October 2002 and December 2003 were invited for follow-up of child growth and development at 10 y of age, and 95% (n = 1,530) agreed to participate. The main reasons for loss to follow-up were parental refusal and outmigration.

This study was approved by the ethics review committee at International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b) in Bangladesh and by the Regional Ethical Review Board in Stockholm, Sweden. Written consent was obtained from all mothers before enrollment in the initial study and before the test performed at 10 y of age.

Manganese Exposure

Drinking water was sampled during pregnancy and at 5 and 10 y of age. Area-wide screening of water As concentrations (W-As) was performed in 2002–2003 (Rahman et al. 2006), and from this screening we identified the wells used during pregnancy as previously described (Rahman et al. 2013). Out of the 1,530 children with neurodevelopmental assessment and water samples at 10 y, we had maternal water samples for 1,265 children; of these children, 1,162 also had their water sampled at 5 y. At the follow-up investigations, the families were interviewed about all water sources used since the children were born and for how long each source had been used. Thereafter, we collected drinking water samples from available indicated wells, enabling construction of the children’s lifelong W-Mn exposure after measuring Mn in the water samples.

Water Collection and Analysis

Water samples were collected in 20-mL trace element–free polyethylene containers containing 50 µL of nitric acid (69% HNO3; Suprapur, MERCK, Germany) to minimize precipitation of metals. The measurements of Mn in the mothers’ drinking water samples (collected in 2002–2003) were performed during 2008–2009, and the measurements of the children’s drinking water samples (5 y: 2007–2008; 10 y: 2012–2013) were performed during 2010 and 2014, respectively. Element concentrations were measured using inductively coupled plasma mass spectrometry (ICP-MS; Agilent 7700x, Agilent Technologies, Tokyo, Japan) as previously described (Rahman et al. 2013). The analytical performance was validated by analysis of a certified reference material (Table S1). No samples were below the limit of detection (LOD) of Mn (see Table S1). For As, 30 samples were below the LOD of 0.01 µg/L at 5 y, and one sample was below the LOD at 10 y; these concentrations were set to LOD divided by the square root of 2.

Outcome Assessments

The Wechsler Intelligence Scale for Children, Fourth Edition (WISC-IV) (Wechsler 2003) was used to assess the children’s general cognitive abilities (IQ), and more specific cognitive functions were assessed through its subtests. The WISC-IV is the most widely used intelligence battery in clinical practice. In the present study, we used WISC-IV raw scores to exclude bias owing to comparison with foreign culture norms. Ten subtests generate four composite scores: a) Verbal Comprehension (vocabulary, information, and comprehension), b) Perceptual Reasoning (block design, picture concepts, and matrix reasoning), c) Working Memory (digit span and arithmetic), and d) Processing Speed (coding and symbol search). The Full-Scale IQ, which represents a child’s general intellectual ability, was derived from the raw scores of the 10 subtests. All test materials and questionnaires were translated to Bengali. A pilot study with 52 children from the study area was conducted to culturally adapt the tests without changing the underlying constructs. For example, pictures unfamiliar to the rural Bangladeshi children were replaced with similar local pictures, and overly difficult questions were changed: For example, “What is a fossil” was changed to “What is a skeleton,” and “Who was Columbus” was changed to “Who was Rabindranath Tagore.” To assess the children’s cognitive abilities, we recruited four female psychologists with Master’s degrees in psychology; they were trained for 6 wk on the above-mentioned tests. During the training, inter-rater reliabilities were measured between each tester and the trainer until the testers achieved >90% agreement with the trainer. To minimize testers’ bias during data collection, they were also observed by a supervisor (senior female psychologist), who visited the testing from time to time and observed ∼5% of all tests by scoring them independently. Adequate inter-rater reliability (intra-class >0.85) was achieved between each of the testers and the supervisor. Behavior problems were assessed using the parent-reported Strengths and Difficulties Questionnaire (SDQ) for 4–17 y old children (Goodman 2001). The SDQ is a screening test for childhood psychopathology adapted for many different cultures and languages, and it has demonstrated excellent psychometric properties ( The test consists of 25 items composing five scales concerning both negative and positive behaviors of children to be addressed by the caregivers on a 0–2 point scale: conduct problems, hyperactivity/inattention, emotional symptoms, and peer relationship problems, with a maximum score of 40; and prosocial behavior, with a maximum score of 10. Female interviewers with Bachelor’s degrees who had completed two weeks of training visited the families’ homes and conducted the SDQ interviews with the mothers or other female caregivers in the absence of mothers. The testers and interviewers were blinded to Mn concentrations in the children’s drinking water because all water samples were analyzed after the data collection was completed.


Information about the mother’s background characteristics [age, early pregnancy weight, height, date of delivery, education, and hemoglobin concentrations (Hb)] and the children’s anthropometry at birth and exclusive breastfeeding at four months (yes/no) was available from the MINIMat trial (Persson et al. 2012). Maternal education was defined as the number of years at school and was categorized as zero, 1–5, and ≥6 y. During the follow-up of the children, we collected information on the number of siblings, years of education (formal schooling categorized as <3, 3, and >3 y) and type of school [public primary school, Madrasa (Islamic) school, NGO (nonprofit private) school, and English medium (private) school], and the families’ socioeconomic status (SES). SES was estimated via a wealth index constructed from information on family ownership of a number of consumer items, housing structure, and dwelling characteristics (Gwatkin et al. 2000). The index was divided into quintiles wherein the lowest quintile represented the poorest families, and the highest quintile represented the richest families. The children’s weight and height were measured by the testers at health clinics after the developmental assessment was performed. The measures were converted into age- and sex-standardized z-scores [weight for age (WAZ) and height for age (HAZ)] using AnthroPlus software [World Health Organization (WHO), Geneva, Switzerland]. The Hb concentrations (g/L) were measured in peripheral blood (finger prick) using a HemoCue photometer (HemoCue AB, Ängelholm, Sweden).

A modification of the middle childhood version (6–10 y) of the Home Observation For Measurement of Environment (HOME) was used to assess the quality and quantity of stimulation and support for the children at home (Caldwell and Bradley 2003). The HOME consists of 58 questions addressing a) responsibility, b) encouragement of maturity, c) emotional climate, d) learning materials and opportunities, e) enrichment, f) family companionship, g) family integration, and h) physical environment. The questionnaire was tested in a pilot study with 20 children in the study area, and seven questions were dropped because the mothers could not understand them. Raven’s Coloured Progressive Matrices™ test (Raven 2003) was used to measure the mothers’ nonverbal IQ in terms of abstract logical reasoning (Hamadani et al. 2011). The sum of correct answers (from a total of 60 questions) was used as a proxy of the mothers’ IQ. Results from Raven’s matrices have demonstrated good convergence with IQ estimates on the Wechsler scales in normative and clinical samples (Bölte et al. 2009) and were sensibly correlated with education and SES.

Because exposure to As, cadmium, and lead through drinking water and food may affect children’s development (Hamadani et al. 2011; Hu et al. 2006; Kippler et al. 2012), we tested whether urinary concentrations of As metabolites (a marker of exposure to inorganic As), cadmium, and lead at gestational week (GW) 8 and at 5 and 10 y of age influenced the associations of Mn with the children’s cognitive development and behavior.

Statistical Analysis

Bivariate associations between exposures, covariates, and outcomes were assessed using Spearman’s rank correlation coefficient (rs), the Kruskal-Wallis test, the Mann-Whitney U-test, or chi-square tests, depending on the type of data. Scatter plots with Lowess smoothing did not indicate nonlinear associations between W-Mn and the different outcomes. Associations of W-Mn with cognitive abilities (WISC-IV) were explored using linear regression analyses, whereas associations with the problem behavior subtest scores were explored using logistic regression with the following cut-offs: conduct problems (normal <3 and slightly raised ≥3 scores), hyperactivity problems (normal <6 and slightly raised ≥6 scores), peer problems (normal <3 and slightly raised ≥3 scores), emotional problems (normal <4 and slightly raised ≥4 scores), and prosocial behavior (normal >5 and slightly raised ≤5 scores) (AMHOCN 2005).

The models were adjusted for the children’s age at assessment and gender (a priori), for variables known to affect child development or that were significantly associated with the exposure and outcome (p<0.05), and for variables that changed the estimates of W-Mn by ≥10% (maternal IQ and SES, number of siblings, child HAZ at 10 y, education in years, type of school attended, Hb concentrations, total HOME score, the testers of cognitive abilities, and the natural log–transformed urinary As concentrations at each exposure window). In case of strongly associated variables (rs>0.60; i.e. WAZ and HAZ at 10 years of age), we included the variable that had the largest influence on the effect estimates. We excluded one boy with a very high W-Mn concentration (6.5mg/L) at 5 y, which had a disproportionately large influence on the associations. Because W-Mn concentrations were skewed at all sampling occasions, we also repeated the analyses with log2-transfomed W-Mn. Because this analysis gave essentially the same results (significance levels), we chose to use the untransformed concentrations, which also simplified the interpretation. The association between W-Mn and W-As showed a complex pattern (Ljung et al. 2009), with elevated W-Mn concentrations mainly at fairly low W-As concentrations (<20μg/L and vice versa (see Figure S1). We controlled for this pattern by adjusting for urinary As (natural log–transformed), a biomarker of ongoing As exposure (Vahter 2002) in our analyses. In an additional step, we restricted the models to children with water As concentrations <20μg/L (at each exposure time point), where most of the variation in W-Mn occurred (see Figure S1). In sensitivity analyses, we additionally adjusted the models for either food (two groups) or micronutrient supplementation (three groups) or for the combination of both (six groups), which was taken by the mothers during pregnancy.

Because gender differences have previously been indicated in cross-sectional studies exploring associations of Mn exposure and children’s cognitive function (Bouchard et al. 2011; Riojas-Rodríguez et al. 2010), we tested for a multiplicative interaction between W-Mn and child gender (significant at <0.10) and we also repeated the abovementioned analyses stratified by child gender. In girls, scatter plots indicated nonlinear associations of maternal W-Mn with cognitive abilities; therefore, we evaluated these associations using linear spline regression analyses with a knot at the turning point of the curve (3mg/L, according to the scatter plot).

Statistical significance was considered as a 95% confidence interval (CI) that did not include zero or p<0.05 (two-sided). The statistical analyses were conducted using SPSS (version 22.0; IBM Corporation, USA) and STATA (version 11; STATACorp LLC, College Station, TX, USA).


Background Characteristics

The main characteristics of the studied 1,265 children are summarized in Table 1. The age range at testing was narrow: 8.8–10.1 y. The girls were more underweight 46% with WAZ <−2 z-scores) and stunted (30% with HAZ < −2 z-scores) than the boys (41% and 25%, respectively). In general, boys obtained higher scores for verbal comprehension, perceptional reasoning, and working memory, and they showed more conduct problems, hyperactivity, and peer problems, whereas girls had higher scores for processing speed and prosocial behavior, but more emotional problems (Table 1). The mothers of the tested children were 14–45 y of age in early pregnancy. Approximately 30% had a BMI <18.5 kg/m2(range 12.8−35.3 kg/m2), and 32% were anemic as indicated by Hb<110g/L in early gestation (WHO 2011). None of the women had smoked or used alcohol during pregnancy.

Table 1. Background characteristics of the studied children, and their cognitive and behavioral scores at 10 y of age.

Variables All childrena (n=1,265) Boysa (n=656) Girlsa (n=609) p-Valueb
Children at 10 years
 Age at testing (years) 9.5±0.1 9.5±0.1 9.5±0.1 0.61
 Education (%, <3/ 3/ or >3 y) 28/39/33 33/39/29 22/40/38 <0.001
 Number of siblings (%, 1–2/3/>3) 31/42/27 34/41/25 29/44/28 0.12
 WAZ −1.8±1.0 −1.7±1.0 −1.8±1.0 0.081
 HAZ −1.4±0.9 −1.4±0.9 −1.5±1.0 0.23
 HOME 27.0±5.0 26.9±5.1 27.2±4.8 0.18
 Hb (g/L) 119±11.1 119±11.1 118±11.3 >0.15
 Drinking water Mn (μg/L) 339 (4.0, 3,203) 334 (3.1, 2,877) 348 (4.9, 3,435) 0.13
 Drinking water As (μg/L) 2.2 (0.1, 311) 3.6 (0.1, 305) 1.3 (0.1, 319) 0.015
 Full-scale IQ 131±34 133±34 130±33 0.22S
 Verbal Comprehension 36±11 37±11 36±10 0.043
 Perceptual Reasoning 31±12 33±12 30±11 <0.0011
 Working Memory 30±6.1 30±6.2 29±6.0 <0.001
 Processing Speed 34±12 32±11 36±12 <0.001
 Conduct problems 2.9±1.7 3.2±1.7 2.6±1.5 <0.001
 Hyperactivity/inattention 4.5±1.7 4.7±1.7 4.2±1.6 <0.001
 Peer relationship problems 2.1±1.3 2.2±1.3 2.0±1.3 0.0046
 Emotional symptoms 1.6±1.4 1.5±1.4 1.8±1.4 <0.001
 Prosocial behavior 6.6±1.8 6.3±1.8 6.9±1.7 <0.001
Children at 5 yearsc
 Water Mn (μg/L) 228 (8.2, 2,605) 226 (7.9, 2,494) 231 (8.6, 2,748) 0.53
 Water As (μg/L) 3.6 (0.1, 266) 4.5 (0.1, 281) 2.9 (0.1, 264) 0.16
 WAZ 1.8±1.8 −1.8±0.9 −1.9±0.8 0.007
 HAZ −1.6±1.6 −1.6±1.0 −1.6±0.9 0.75
Children at birth
 Gestational age (weeks) 9.1±2.1 39.0±2.2 39.3±2.1 0.004
 Weight (g) 2,688±378 2,688±378 2,643±656 <0.001
 Length (cm) 47.7±2.1 48.0v2.2 47.4±1.9 <0.001
 Age (years) 26.8±5.8 26.9±5.8 26.6±6.2 0.26
 BMI at GW9 (kg/m2) 20.0±2.6 20.0±2.6 20.0±2.6 0.84
 Education in years 4.5±4.0 4.5±4.0 4.5±4.0 0.83
 Maternal IQ 24.7±11.6 24.7±11.8 24.6±11.8 0.87
 Hb (g/L) 116±13.1 116±12.6 115±13.6 0.16
 Drinking water Mn (μg/L) 204 (23, 2,494) 218 (22, 2,251) 200 (25, 2,715) 0.53
 Drinking water As (μg/L) 33 (0.1, 412) 29 (0.1, 411) 35 (0.1, 425) 0.90
Abbreviations: As, arsenic; BMI, body mass index; GW, gestational week; HAZ, height for age; Hb, hemoglobin concentration; HOME, Home Observation for Measurement of Environment; Mn, manganese; SD, standard deviation; SDQ, Strengths and Difficulties Questionnaire; WAZ, weight for age; WISC-IV, Wechsler Intelligence Scale for Children, 4th Edition.

aData shown as the arithmetic mean ±SD or percentage or median (5th and 95th percentile) for As and Mn concentrations.

bDerived by using either Mann-Whitney U-test or χ2 test.

cAvailable for n=1,162.

dSampled at gestational week 14.

The median concentration of prenatal W-Mn was 204 μg/L with a wide range (–1.3–6,550 μg/L). At 5 y, these values were quite similar (median, 228μg/L; range, 0.1–6,550 μg/L), and at 10 y, the median was 339 μg/L (range, 0.1–8,680 μg/L). In total, 810 and 733 of the children had the same water source at 5 and 10 y of age, respectively, as their mother had used during pregnancy (W-Mn prenatally and at 5 y rs=0.76, p<0.001; W-Mn prenatally and 10 y rs=0.58, p<0.001), and 656 of the children had used the same water source from birth to 10 y of age (W-Mn at 5 and 10 y rs=0.74, p<0.001). W-Mn correlated weakly with SES rs=0.16, 0.10, and 0.14, respectively, for water collected prenatally and at 5 and 10 y; p<0.05), but not with child anthropometry measures (WAZ and HAZ at 5 and 10 y; p>0.10).

Manganese and Children’s Cognitive Abilities

In the age- and gender-adjusted analyses, both prenatal and child W-Mn were positively associated with all cognitive ability measures at 10 y of age Table 2). However, adjustment for other covariates, in particular SES, markedly decreased the associations, which were no longer significant. Restricting the analyses to children of women with W-As <20μg/L further reduced the associations (Table 2). Therefore, children with low W-As were selected for the subsequent analyses.

Table 2. Associations of W-Mn (mg/L) during pregnancy and at 5 and 10 y of age with measures of cognitive abilities at 10 y.

Age- and gender-adjusted model Multivariable-adjusted model 1a Multivariable-adjusted model 2b
Exposure windows β (95% CI) p-value β (95% CI) p-value β (95% CI) p-value p-interactionc
W-Mn in pregnancy (n=1,265) (n=1,201) (n=554)
 Full-scale IQ 4.5 (2.5, 6.5) <0.001 1.0 −0.69, 2.7) 0.25 0.42 −1.6, 2.5) 0.69 0.029
 Verbal Comprehension 1.5 (0.87, 2.1) <0.001 0.37 −0.17, 0.92) 0.18 0.070 −0.62, 0.76) 0.84 0.050
 Perceptual Reasoning 1.2 (0.50, 1.9) 0.001 0.26 (−0.40, 0.92) 0.45 0.16 (−0.65, 0.96) 0.70 0.20
 Working Memory 0.61 (0.24, 0.98) 0.001 0.17 (−0.17, 0.51) 0.32 0.072 (−0.33, 0.47) 0.73 0.081
 Processing Speed 1.2 (0.47, 1.9) 0.001 0.20 (−0.47, 0.86) 0.56 0.12 (−0.64, 0.88) 0.76 0.059
W-Mn at 5 y (n=1,162) (n=1,124) (n=705)
 Full scale IQ 2.6 (0.51, 4.6) 0.015 0.23 (−1.5, 2.0) 0.79 −0.37 (−2.3, 1.5) 0.70 0.097
 Verbal Comprehension 0.88 (0.22, 1.5) 0.009 0.096 (−0.46, 0.66) 0.74 −0.10 (−0.73, 0.52) 0.74 0.11
 Perceptual Reasoning 0.87 (0.16, 1.6) 0.017 0.31 (−0.36, 0.99) 0.36 0.10 (−0.64, 0.84) 0.79 0.12
 Working Memory 0.30 (−0.084, 0.68) 0.13 −0.030 (−0.37, 0.31) 0.86 −0.043 (−0.43, 0.34) 0.83 0.013
 Processing Speed 0.53 (−0.20, 1.3) 0.15 −0.15 (−0.83, 0.53) 0.66 −0.32 (−1.0, 0.40) 0.38 0.96
W-Mn at 10 y (n=1,265) (n=1,232) (n=801)
 Full-scale IQ 2.3 (0.70, 3.9) 0.005 0.71 (−0.64, 2.1) 0.30 0.18 (−1.3, 1.6) 0.81 0.44
 Verbal Comprehension 0.57 (0.072, 1.08) 0.025 −0.12 (−0.56, 0.32) 0.59 −0.33 (−0.81, 0.15) 0.18 0.42
 Perceptual Reasoning 0.56 (0.012, 1.1) 0.045 0.12 (−0.41, 0.66) 0.65 −0.050 (−0.63, 0.53) 0.87 0.46
 Working Memory 0.37 (0.081, 0.66) 0.012 0.19 (−0.084, 0.46) 0.17 0.12 (−0.17, 0.40) 0.42 0.43
 Processing Speed 0.78 (0.22 1.3) 0.014 0.52 (−0.016, 1.06) 0.057 0.44 (−0.13, 1.0) 0.13 0.88
Abbreviations: 95% CI, 95% confidence interval; W-Mn, water manganese concentration.

aAdjusted for mother’s IQ, socioeconomic status (SES) (quintiles), child age at intelligence testing (years), gender, education (<3, 3, and >3 y), height for age (HAZ) at 10 y, hemoglobin concentration (Hb) (g/dL), school type [public primary school, Madrasa (Islamic), NGO (nonprofit private), and English medium (private) school], Home Observation for Measurement of Environment (HOME), tester, number of siblings, and urinary arsenic (As) concentrations at each respective time point (natural log–transformed).

bAdjusted as model 1, but restricted to children with water As <20 μg/L at each respective time point.

cIn model 2, we also tested for a multiplicative interaction between W-Mn and gender.

Gender was strongly influential in the models of prenatal W-Mn with different cognitive ability measures, and the interaction between gender and W-Mn was significant for Full-Scale IQ, Verbal Comprehension, Working Memory, and Processing Speed p<0.10; Table 2). When boys were evaluated separately n=288 with W-As <20μg/L), the multivariable-adjusted associations of prenatal or childhood W-Mn concentrations with the different cognitive ability measures were generally inverse, although nonsignificantly (with the exception of Working Memory; Table 3). For girls, the linear spline regression analysis with prenatal W-Mn showed positive associations below the knot at 3mg/L for all cognitive ability measures (Table 3). For Full-Scale IQ and Verbal Comprehension, these associations differed significantly from the inverse associations above the knot (Wald test p<0.05), although only 27 mothers had W-Mn concentrations >3 mg/L. The positive associations of W-Mn concentrations at 5 and 10 y of age were much weaker than those with the prenatal W-Mn and were no longer statistically significant (Table 3). Further adjusting the association of W-Mn at 5 or 10 y with Full-Scale IQ for the mothers’ urinary As concentrations during pregnancy had marginal impact on the estimates (data not shown).

Table 3. Associations of W-Mn (mg/L) during pregnancy and childhood with cognitive abilities at 10 y when restricted to low W-As and stratified by gender.

Boys Girls
Exposure windows β (95% CI) p-value β (95% CI) p-value
W-Mn in pregnancya n=288 n=266/n=27
Full-scale IQ −1.8 (−5.3, 1.7) 0.31 5.2 (1.8, 8.6)/−5.4 (−13, 2.0)b 0.003/0.15
Verbal Comprehension −0.62 (−1.8, 0.53) 0.29 1.5 (0.31, 2.6)/−1.7 (−4.2 , 0.81)b 0.013/0.18
Perceptual Reasoning −0.22 (−1.6, 1.2) 0.76 1.4 (0.033, 2.7)/−1.3 (−4.2, 1.5)b 0.045/0.37
Working Memory −0.32 (−1.0, 0.39) 0.38 0.72 (0.084, 1.4)/−0.51 (−1.9, 0.88)b 0.027/0.47
Processing Speed −0.64 (−1.9, 0.59) 0.31 1.6 (0.29, 3.0)/−1.9 (–4.8, 1.0)b 0.018/0.21
W-Mn at 5 ya,c n=345 n=350
Full scale IQ −3.2 (−6.4, 0.065) 0.055 0.92 (−1.4, 3.3) 0.44
Verbal Comprehension −0.88 (−1.9, 0.17) 0.099 0.26 (−0.53, 1.0) 0.52
Perceptual Reasoning −1.1 (−2.3, 0.18) 0.093 0.65 (−0.27, 1.6) 0.17
Working Memory −0.70 (−1.4, −0.022) 0.043 0.38 (−0.074, 0.84) 0.10
Processing Speed −0.52 (−1.7, 0.67) 0.39 −0.37 (−1.3, 0.57) 0.44
W-Mn at 10 ya n=406 n=395
Full-scale IQ −0.56 (−3.0, 1.8) 0.64 0.70 (−1.2, 2.6) 0.46
Verbal Comprehension −0.59 (−1.4, 0.17) 0.13 −0.16 (−0.78, 0.46) 0.61
Perceptual Reasoning −0.35 (−1.3, 0.60) 0.47 0.17 (−0.55, 0.90) 0.64
Working Memory −0.58 (−0.53, 0.41) 0.81 0.23 (−0.13, 0.59) 0.21
Processing Speed 0.44 (−0.44, 1.3) 0.33 0.45 (−0.31, 1.2) 0.24
Abbreviations: 95% CI, 95% confidence interval; Hb, hemoglobin concentration; W-As, water arsenic concentration; W-Mn, water manganese concentration.

aAdjusted for mother’s IQ, socioeconomic status (SES) (quintiles), child age at intelligence testing (years, one decimal), gender, education <3, 3, or >3 y), height for age (HAZ) at 10 y, hemoglobin concentration (Hb) (g/L), school type [public primary school, Madrasa (Islamic), NGO (nonprofit private), and English medium (private) school], Home Observation for Measurement of Environment (HOME), tester (4 psychologists), number of siblings, and urinary As concentrations at each respective time point (natural log–transformed).

bSpline regression analyses with a knot at 3 mg/L; β and 95% CI for <3mg/L/≥3 mg/L.

cExcluding one outlier with W-Mn=6.5 mg/L.

Manganese and Children’s Behavior

The multivariable-adjusted logistic regression analyses showed that at all time points, W-Mn was significantly associated with an increased risk of conduct problems (Table 4; Model 1). Restricting the analyses to children with low W-As (Model 2) slightly strengthened the risk estimates for conduct problems and low prosocial behavior, which were statistically significant at all exposure time points. Thus, the following analyses concern children using water with <20 μg/L As.

Table 4. Multivariable-adjusted odd ratios (95% confidence intervals) for raised SDQ difficult scores (low prosocial) at 10 y in relation to W-Mn (mg/L) during pregnancy and childhood.

Age- and gender-adjusted model Multivariable-adjusted model 1a Multivariable-adjusted model 2b
Exposure windows OR (95% CI) p-value OR (95% CI) p-value OR (95% CI) p-value p-interactionc
W-Mn in pregnancy n=1,265 n=1,201 n=554
Conduct problems 1.09 (0.96, 1.24) 0.18 1.20 (1.04, 1.39) 0.013 1.29 (1.08, 1.53) 0.005 0.22
Hyperactivity 1.11 (0.97, 1.27) 0.13 1.11 (0.95, 1.29) 0.20 1.03 (0.86, 1.23) 0.74 0.090
Peer problems 0.99 (0.87, 1.13) 0.88 0.99 (0.85, 1.14) 0.85 0.96 (0.81, 1.14) 0.66 0.16
Emotional problems 0.76 (0.60, 0.98) 0.031 0.81 (0.61, 1.07) 0.14 0.70 (0.51, 0.97) 0.034 0.18
Prosocial behaviord 1.10 (0.95, 1.28) 0.21 1.16 (0.97, 1.39) 0.093 1.27 (1.03, 1.57) 0.025 0.35
W-Mn at 5 y n=1,162 n=1,124 n=705
Conduct problems 1.15 (1.01, 1.31) 0.037 1.20 (1.03, 1.39) 0.016 1.26 (1.07, 1.48) 0.005 0.44
Hyperactivity 1.09 (0.95, 1.25) 0.23 1.05 (0.90, 1.23) 0.56 1.06 (0.89, 1.26) 0.49 0.25
Peer problems 0.99 (0.86, 1.13) 0.86 0.97 (0.84, 1.12) 0.66 0.98 (0.83, 1.15) 0.79 0.18
Emotional problems 0.74 (0.57, 0.97) 0.029 0.77 (0.57, 1.03) 0.077 0.72 (0.52, 0.99) 0.045 0.57
Prosocial behaviord 0.98 (0.84, 1.13) 0.77 1.04 (0.88, 1.24) 0.63 1.11 (0.92, 1.34) 0.28 0.60
W-Mn at 10 y n=1,265 n=1,232 n=801
Conduct problems 1.10 (1.00, 1.22) 0.052 1.17 (1.04, 1.31) 0.007 1.18 (1.05, 1.34) 0.007 0.31
Hyperactivity 0.99 (0.89, 1.11) 0.93 0.97 (0.86, 1.11) 0.69 0.96 (0.84, 1.11) 0.61 0.43
Peer problems 0.92 (0.83, 1.03) 0.14 0.91 (0.81, 1.02) 0.12 0.89 (0.78, 1.01) 0.073 0.074
Emotional problems 0.77 (0.63, 0.94) 0.010 0.79 (0.63, 0.99) 0.037 0.83 (0.66, 1.05) 0.12 0.71
Prosocial behaviord 1.09 (0.96, 1.22) 0.17 1.17 (1.01, 1.34) 0.031 1.23 (1.05, 1.43) 0.009 0.25
Abbreviations: 95% CI, 95% confidence interval; OR, odds ratio; SDQ, Strengths and Difficulties Questionnaire; W-Mn, water manganese concentration.

aAdjusted for mother’s IQ, socioeconomic status (SES) (quintiles), child age at intelligence testing (years), gender, education (<3, 3, and >3 years), height for age (HAZ) at 10 years, hemoglobin concentration (Hb) (g/dL), school type [public primary school, Madrasa (Islamic), NGO (nonprofit private), and English medium (private) school], Home Observation for Measurement of Environment (HOME), number of siblings, and urinary arsenic (As) concentration at each respective time point (natural log–transformed).

bAdjusted as model 1, but restricted to children with water As <20 μg/L at each respective time point.

cIn model 2, we also tested for a multiplicative interaction between water Mn and gender.

dOR for low prosocial scores.

We found a significant interaction between gender and prenatal W-Mn for hyperactivity and between gender and W-Mn at 10 y of age for peer problems (Table 4). Stratifying the models by gender (Table 5) indicated slightly stronger associations of prenatal W-Mn with conduct problems in boys (statistically significant) than in girls (not significant). For boys, the obtained odds ratios for conduct problems in relation to W-Mn at 5 and 10 y were slightly lower than those for prenatal exposure. In contrast, prenatal W-Mn was associated with a decreased risk of emotional problems in boys (only 27 cases, out of which 20 also had conduct problems), but for W-Mn at 5 and 10 y of age (37 and 40 cases, respectively), this inverse association was weaker and no longer significant. In girls, W-Mn was mainly associated with an increased risk of low prosocial scores.

Table 5. Multivariable-adjusted odd ratios (95% confidence intervals) of difficult behavior at 10 y in relation to W-Mn (mg/L) during pregnancy and childhood when restricted to low W-As and stratified by child gender.

Boys Girls
Exposure windows OR (95% CI) p-value OR (95% CI) p-value
W-Mn in pregnancya n=288 n=266
Conduct problems 1.43 (1.06, 1.91) 0.017 1.18 (0.95, 1.46) 0.13
Hyperactivity 1.14 (0.88, 1.48) 0.31 0.90 (0.69, 1.18) 0.45
Peer problems 1.11 (0.85, 1.44) 0.45 0.86 (0.68, 1.10) 0.23
Emotional problems 0.39 (0.19, 0.82) 0.013 0.80 (0.54, 1.18) 0.26
Prosocial behaviorb 1.17 (0.87, 1.57) 0.29 1.48 (1.07, 2.06) 0.019
W-Mn at 5 ya n=355 n=350
Conduct problems 1.36 (1.04, 1.79) 0.026 1.21 (0.99, 1.49) 0.066
Hyperactivity 1.16 (0.90, 1.50) 0.25 0.99 (0.77, 1.27) 0.94
Peer problems 1.30 (0.89, 1.44) 0.31 0.90 (0.72, 1.12) 0.33
Emotional problems 0.61 (0.36, 1.03) 0.063 0.72 (0.47, 1.12) 0.15
Prosocial behaviorb 1.10 (0.85, 1.42) 0.48 1.14 (0.85, 1.52) 0.39
W-Mn at 10 ya n=406 n=395
Conduct problems 1.24 (1.01, 1.53) 0.045 1.16 (0.99, 1.36) 0.062
Hyperactivity 1.01 (0.83, 1.23) 0.90 0.94 (0.77, 1.15) 0.54
Peer problems 1.00 (0.82, 1.20) 0.97 0.82 (0.68, 0.99) 0.039
Emotional problems 0.88 (0.62, 1.25) 0.47 0.77 (0.55, 1.08) 0.14
Prosocial behaviorb 1.15 (0.93, 1.41) 0.20 1.34 (1.05, 1.73) 0.021
Abbreviations: 95% CI, 95% confidence interval; OR, odds ratio; W-As, water arsenic concentration; W-Mn, water magnesium concentration.

aAdjusted for mother’s IQ, socioeconomic status (SES) (quintiles), child age at intelligence testing (years), education (<3, 3, and >3 y), height for age (HAZ) at 10 y, hemoglobin (Hb) (g/dL), school type [public primary school, Madrasa (Islamic), NGO (nonprofit private), and English medium (private) school], Home Observation for Measurement of Environment (HOME), number of siblings, and urinary arsenic (As) concentration at each respective time point (natural log transformed).

bOR of low prosocial scores.


This large prospective cohort study indicates that exposure to elevated W-Mn prenatally and at 5 and 10 y of age was associated with a similar increased risk (∼18–29%) of parent-reported conduct problems, particularly in boys. Simultaneously, W-Mn was associated with a lower risk of emotional problems in the boys, most likely because emotional problems occur less frequently in children with conduct problems (frequent temper tantrums or hot tempers, disobedience, fighting with other children, etc.) or are noted less frequently by the parents of such children. In girls, exposure to elevated W-Mn was mainly associated with low scores for prosocial behavior. We found no clear evidence for impairment of cognitive abilities by W-Mn. There was a tendency toward inverse associations in boys, particularly in those with early childhood exposure; however, the effect on Full-Scale IQ corresponded to <0.1 standard deviation (SD) only. On the contrary, prenatal W-Mn was positively associated with all cognitive test scores in girls up to approximately 3 mg/L, above which the positive influence disappeared. The total range of W-Mn at 10 y was 0.001–8.7 mg/L, with 52% exceeding the U.S. lifetime health advisory value of 0.3 mg/L [U.S. Environmental Protection Agency (EPA) 2012]. The U.S. Institute of Medicine has indicated a Tolerable Upper Intake Level of 6 mg Mn/day for children ages 9–13 y (Institute of Medicine 2001).

In support of the present findings, a previous cross-sectional study indicated more classroom behavioral problems at 8–11 y of age (n=211) with increasing W-Mn concentration (Khan et al. 2011), but the time of initiation of the behavioral effects and the potential gender differences were not evaluated. Interestingly, higher perinatal Mn exposure, assessed by concentrations in deciduous tooth dentin, was recently associated with poorer behavioral outcomes in 248 school-age U.S. children (Mora et al. 2015). Unlike the present findings, a few cross-sectional studies have shown inverse associations between W-Mn and cognitive abilities in 6–13 y old children (Bouchard et al. 2011; Khan et al. 2012; Wasserman et al. 2006) (n=362, 840, and 142, respectively). The reason for this discrepancy is not known, and further prospective studies are warranted. Notably, excess Mn exposure is believed to affect dopamine and other neurotransmitters in the brain (Vorhees et al. 2014), and experimental studies indicate an increased susceptibility to Mn very early in life (Beaudin et al. 2013).

The mode(s) of action behind the observed gender differences in the effects of Mn are unknown. It is possible that the underlying mode of action is the same for the observed behavioral effects. The parents’ general scoring of more difficult problems in boys and better prosocial behavior in girls supports the influence of gender-specific norms. Nevertheless, previous studies have indicated gender-related differences in the kinetics of Mn, as shown by higher blood Mn concentrations in girls than in boys before 2–3 y of age (Berglund et al. 2011; Oulhote et al. 2014b) and by markedly higher urinary excretion of Mn in girls than in boys at 6–12 y of age (Berglund et al. 2011). In experimental studies, postnatal exposure to Mn has been shown to alter the levels of monoamines and corticosterone in a sex-dependent manner (Vorhees et al. 2014) and to cause more morphological changes in striatal medium spiny neurons in male mice than in female mice (Madison et al. 2011).

The indicated positive associations of prenatal W-Mn with girls’ cognitive abilities, with no such associations observed in boys, suggest a sex difference during fetal development. Therefore, another hypothesis is that prenatal Mn exposure affects epigenetic and/or hormonal factors. The early fetal epigenome is sensitive to environmental influences, particularly to poor nutrition (Barker et al. 2013), and this sensitivity has been shown to differ by sex (Tarrade et al. 2015). Because it was mainly the prenatal exposure that appeared to be beneficial for the girls’ cognitive development, residual confounding seems unlikely. Gender differences were also apparent with W-Mn at 5 y of age; there was no association of W-Mn with the girls’ cognitive abilities, whereas boys showed mainly inverse associations. Unrelated to W-Mn, girls generally scored lower than boys on the cognitive tests, which is in contrast to the results at 5 y of age (Hamadani et al. 2011). This difference may be related to discrimination against the female gender, resulting in poorer nutrition, less stimulation, and more violence (Adams et al. 2013; VanderEnde et al. 2014).

The strengths of this study include the large sample size and the population-based prospective design with Mn measured in all water sources used by the children and in all those used by their mothers during pregnancy. Based on the lifetime water histories of the children, we also attempted to evaluate the effects of lifetime cumulative exposure. However, because the correlations of W-Mn between years were generally very high (0.7–1.0), the results were essentially the same as those obtained for childhood exposure at 5 and 10 y of age. The cognitive assessment was expressed as raw scores because the WISC-IV has not been standardized for Bangladeshi children, giving rise to low scoring according to the international standard. A limitation of the study is the lack of water samples from the children’s schools, which might have caused nondifferential misclassification of the children’s exposure, particularly at 10 y. At 5 y, only 30% of the children attended school, and then only for a short time each day. However, we observed a strong correlation (rs=0.66; p≤0.001) between the concentration of As in the children’s water and that in their urine, a biomarker of ongoing As exposure, indicating that the included W-Mn concentrations were indeed representative of the children’s exposure. In addition, we only had information on exclusive breastfeeding at four months (yes/no), which showed that only 55% of the infants were still exclusively breastfed at this age. In another study from the same area, infants that were not exclusively breastfed were mainly given cow’s milk (∼25% of the infants at 4 mo), plain water, fruit juice, or semisolid food (all ∼20% at 4 mo) (Saha et al. 2008). Thus, the age at which drinking water was first introduced likely varied among the infants. Another limitation is the lack of an exposure biomarker, in particular blood Mn concentration, which has been used repeatedly in previous studies on child neurodevelopment (Sanders et al. 2015). However, in a subcohort of the mothers, we found no association whatsoever between the concentration of Mn in water and that in blood (Ljung et al. 2009), and because the aim of the present study was to evaluate potential adverse effects of high W-Mn concentrations, we prioritized analyzing all the collected water samples. An additional limitation is that the iron status of the mothers and their children was only assessed via hemoglobin, and more precise markers of iron deficiency such as serum ferritin were only present for a subsample of the mothers.


In conclusion, elevated prenatal and early childhood exposure to W-Mn (mainly from medium-deep wells) appeared to increase the risk of children’s behavior problems at 10 y of age. However, we also found positive associations of prenatal W-Mn and cognitive abilities in girls. Thus, compared with the severe health risks of exposure to elevated concentrations of As, a potent carcinogen and general toxicant that is present in many shallow wells, the effects observed here of elevated Mn concentrations in water from the increasingly common deeper wells seem modest. Potential temporal variations in W-Mn in different types of water sources need to be evaluated.


The authors acknowledge the Swedish Research Council and the Swedish Research Council Formas, the Swedish International Development Cooperative Agency, Public Health Impact of long-term, low-level Mixed Element Exposure in susceptible population strata (PHIME), and Karolinska Institutet for funding the present study. The International Centre for Diarrhoeal Disease Research, Bangladesh (icddr,b) acknowledges the donors providing unrestricted support: the Australian International Development Agency (AusAID), the Government of Bangladesh, the Canadian International Development Agency (CIDA), the Swedish International Development Cooperative Agency (Sida), and the Department for International Development, UK (DFID).

The support of the field staff and the participation of the women and children are highly appreciated.


Adams AM, Rabbani A, Ahmed S, Mahmood SS, Al-Sabir A, Rashid SF, et al. 2013. Explaining equity gains in child survival in Bangladesh: scale, speed, and selectivity in health and development. Lancet 382:2027–2037, PMID: 24268604, 10.1016/S0140-6736(13)62060-7.

AMHOCN 2005 (Australian Mental Health Outcomes and Classification Network). 2005. Strengths and Difficulties Questionnaire Training Manual. Parramatta, NSW, Austrailia: NSW Institute of Psychiatry. [accessed 1 December 2015].

Barker D, Barker M, Fleming T, Lampl M. 2013. Developmental biology: Support mothers to secure future public health. Nature 504:209–211, PMID: 24350368, 10.1038/504209a.

Beaudin SA, Nisam S, Smith DR. 2013. Early life versus lifelong oral manganese exposure differently impairs skilled forelimb performance in adult rats. Neurotoxicol Teratol 38:36–45, PMID: 23623961, 10.1016/

Berglund M, Lindberg AL, Rahman M, Yunus M, Grandér M, Lönnerdal B, et al. 2011. Gender and age differences in mixed metal exposure and urinary excretion. Environ Res 111:1271–1279, PMID: 21962832, 10.1016/j.envres.2011.09.002.

Bölte S, Dziobek I, Poustka F. 2009. Brief report: The level and nature of autistic intelligence revisited. J Autism Dev Disord 39:678–682, PMID: 19052857, 10.1007/s10803-008-0667-2.

Bouchard MF, Sauvé S, Barbeau B, Legrand M, Brodeur ME, Bouffard T, et al. 2011. Intellectual impairment in school-age children exposed to manganese from drinking water. Environ Health Perspect 119:138–143, PMID: 20855239, 10.1289/ehp.1002321.

Caldwell BM, Bradley RH. 2003. Home Inventory Administration Manual. Little Rock, AR:University of Arkansas for Medical Sciences.

Chan TJ, Gutierrez C, Ogunseitan OA. 2015. Metallic burden of deciduous teeth and childhood behavioral deficits. Int J Environ Res Public Health 12(6):6771–6787, 10.3390/ijerph120606771.

EFSA (European Food Safety Authority). 2013. Scientific opinion on dietary reference values for manganese. EFSA J 11:3419–3463.

Erikson KM, Thompson K, Aschner J, Aschner M. 2007. Manganese neurotoxicity: a focus on the neonate. Pharmacol Ther 113:369–377, PMID: 17084903, 10.1016/j.pharmthera.2006.09.002.

Frisbie SH, Mitchell EJ, Dustin H, Maynard DM, Sarkar B. 2012. World Health Organization Discontinues Drinking Water Guideline for Manganese. Environ Health Perspect 120:775–778, PMID: 22334150, 10.1289/ehp.1104693.

Goodman R. 2001. Psychometric properties of the strengths and difficulties questionnaire. J Am Acad Child Adolesc Psychiatry 40:1337–1345, PMID: 11699809, 10.1097/00004583-200111000-00015.

Grandjean P, Landrigan PJ. 2014. Neurobehavioural effects of developmental toxicity. Lancet Neurol 13:330–338, PMID: 24556010, 10.1016/S1474-4422(13)70278-3.

Gunier RB, Arora M, Jerrett M, Bradman A, Harley KG, Mora AM, et al. 2015. Manganese in teeth and neurodevelopment in young Mexican-American children. Environ Res 142:688–695, PMID: 26381693, 10.1016/j.envres.2015.09.003.

Gwatkin DR, Rustein S, Johnson K, Pande RP, Wagstaff A, Amouzou A. 2000. Socio-economic Differences in Health, Nutrition, and Population in Bangladesh. Washington, DC:World Bank. [accessed 1 December 2015].

Hamadani J, Tofail F, Nermell B, Gardner R, Shiraji S, Bottai M, et al. 2011. Critical windows of exposure for arsenic-associated impairment of cognitive function in pre-school girls and boys: a population-based cohort study. Int J Epidemiol 40:1593–1604, PMID: 22158669, 10.1093/ije/dyr176.

Hu H, Téllez-Rojo MM, Bellinger D, Smith D, Ettinger AS, Lamadrid-Figueroa H, et al. 2006. Fetal lead exposure at each stage of pregnancy as a predictor of infant mental development. Environ Health Perspect 114:1730–1735, PMID: 17107860, 10.1289/ehp.9067.

Institute of Medicine (US) Panel on Micronutrients. 2001. Dietary reference intakes for vitamin A, vitamin K, arsenic, boron, chromium, copper, iodine, iron, manganese, molybdenum, nickel, silicon, vanadium, and zinc. Washington, DC:National Academies Press.

Khan K, Factor-Litvak P, Wasserman GA, Liu X, Ahmed E, Parvez F, et al. 2011. Manganese exposure from drinking water and children’s classroom behavior in Bangladesh. Environ Health Perspect 119:1501–1506, PMID: 21493178, 10.1289/ehp.1003397.

Khan K, Wasserman GA, Liu X, Ahmed E, Parvez F, Slavkovich V, et al. 2012. Manganese exposure from drinking water and children’s academic achievement. Neurotoxicology 33:91–97, PMID: 22182530, 10.1016/j.neuro.2011.12.002.

Kippler M, Skröder H, Rahman SM, Tofail F, Vahter M. 2016. Elevated childhood exposure to arsenic despite reduced drinking water concentrations – A longitudinal cohort study in rural Bangladesh. Environ Int 86:119–125, PMID: 26580026, 10.1016/j.envint.2015.10.017.

Kippler M, Tofail F, Gardner R, Rahman A, Hamadani JD, Bottai M, et al. 2012. Maternal cadmium exposure during pregnancy and size at birth: a prospective cohort study. Environ Health Perspect 120:284–289, PMID: 21862444, 10.1289/ehp.1103711.

Ljung KS, Kippler MJ, Goessler W, Grandér GM, Nermell BM, Vahter ME. 2009. Maternal and early life exposure to manganese in rural Bangladesh. Environ Sci Technol 43:2595–2601, PMID: 19452922.

Ljung K, Vahter M. 2007. Time to re-evaluate the guideline value for manganese in drinking water? Environ Health Perspect 115:1533–1538, PMID: 18007980, 10.1289/ehp.10316.

Madison JL, Wegrzynowicz M, Aschner M, Bowman AB. 2011. Gender and manganese exposure interactions on mouse striatal neuron morphology. Neurotoxicology 32:896–906, PMID: 21641932, 10.1016/j.neuro.2011.05.007.

Mistry HD, Williams PJ. 2011. The importance of antioxidant micronutrients in pregnancy. Oxid Med Cell Longev 2011:841749, PMID: 21918714, 10.1155/2011/841749.

Mora AM, Arora M, Harley KG, Kogut K, Parra K, Hernández-Bonilla D, et al. 2015. Prenatal and postnatal manganese teeth levels and neurodevelopment at 7, 9, and 10.5 years in the CHAMACOS cohort. Environ Int 84:39–54, PMID: 26209874, 10.1016/j.envint.2015.07.009.

Oulhote Y, Mergler D, Barbeau B, Bellinger DC, Bouffard T, Brodeur ME, et al. 2014a. Neurobehavioral function in school-age children exposed to manganese in drinking water. Environ Health Perspect 122:1343–1350, PMID: 25260096, 10.1289/ehp.1307918.

Oulhote Y, Mergler D, Bouchard MF. 2014b. Sex- and age-differences in blood manganese levels in the U.S. general population: National Health and Nutrition Examination Survey 2011–2012. Environ Health 13:87, 10.1186/1476-069X-13-87.

Persson LA, Arifeen S, Ekström EC, Rasmussen KM, Frongillo EA, Yunus M, et al. 2012. Effects of prenatal micronutrient and early food supplementation on maternal hemoglobin, birth weight, and infant mortality among children in Bangladesh: the MINIMat randomized trial. JAMA 307:2050–2059,PMID: 22665104, 10.1001/jama.2012.4061.

Rahman M, Vahter M, Wahed MA, Sohel N, Yunus M, Streatfield PK, et al. 2006. Prevalence of arsenic exposure and skin lesions. A population based survey in Matlab, Bangladesh. J Epidemiol Community Health 60:242–248, PMID: 16476755, 10.1136/jech.2005.040212.

Rahman SM, Åkesson A, Kippler M, Grandér M, Hamadani JD, Streatfield PK, et al. 2013. Elevated manganese concentrations in drinking water may be beneficial for fetal survival. PLoS One 8:e74119, PMID: 24066101, 10.1371/journal.pone.0074119.

Rahman SM, Kippler M, Ahmed S, Palm B, El Arifeen S, Vahter M. 2015. Manganese exposure through drinking water during pregnancy and size at birth: A prospective cohort study. Reprod Toxicol 53:68–74, PMID: 25828058, 10.1016/j.reprotox.2015.03.008.

Raven J, Raven JC, Court JH. 2003. Section 1, General Overview. In: Manual for Raven’s Progressive Matrices and Vocabulary Scales. San Antonio:Harcourt Assessment.

Riojas-Rodríguez H, Solís-Vivanco R, Schilmann A, Montes S, Rodríguez S, Ríos C, et al. 2010. Intellectual function in Mexican children living in a mining area and environmentally exposed to manganese. Environ Health Perspect 118:1465–1470, 10.1289/ehp.0901229.

Saha KK, Frongillo EA, Alam DS, Arifeen SE, Persson LA, Rasmussen KM. 2008. Appropriate infant feeding practices result in better growth of infants and young children in rural Bangladesh. Am J Clin Nutr 87:1852–1859, PMID: 18541577.

Sanders AP, Claus Henn B, Wright RO. 2015. Perinatal and childhood exposure to cadmium, manganese, and metal mixtures and effects on cognition and behavior: A review of recent literature. Curr Environ Health Rep 2:284–294, PMID: 26231505, 10.1007/s40572-015-0058-8.

Takser L, Lafond J, Bouchard M, St-Amour G, Mergler D. 2004. Manganese levels during pregnancy and at birth: relation to environmental factors and smoking in a Southwest Quebec population. Environ Res 95:119–125, 10.1016/j.envres.2003.11.002.

Takser L, Mergler D, Hellier G, Sahuquillo J, Huel G. 2003. Manganese, monoamine metabolite levels at birth, and child psychomotor development. Neurotoxicology 24:667–674, PMID: 12900080, 10.1016/S0161-813X(03)00058-5.

Tarrade A, Panchenko P, Junien C, Gabory A. 2015. Placental contribution to nutritional programming of health and diseases: epigenetics and sexual dimorphism. J Exp Biol 218:50–58, PMID: 25568451, 10.1242/jeb.110320.

U.S. EPA (U.S. Environmental Protection Agency). Office of Water. 2012. 2012 Edition of the Drinking Water Standards and Health Advisories. Updated April 2012. [accessed 1 December 2015].

Vahter M. 2002. Mechanisms of arsenic biotransformation. Toxicology 181–182:211–217, PMID: 12505313.

Vahter ME, Li L, Nermell B, Rahman A, El Arifeen S, Rahman M, et al. 2006. Arsenic exposure in pregnancy: a population-based study in Matlab, Bangladesh. J Health Popul Nutr 24:236–245, PMID: 17195565.

VanderEnde K, Amin S, Naved RT. 2014. Community-level correlates of physical violence against unmarried female adolescents in Bangladesh. BMC Public Health 14:1027, PMID: 25277780, 10.1186/1471-2458-14-1027.

Vorhees CV, Graham DL, Amos-Kroohs RM, Braun AA, Grace CE, Schaefer TL, et al. 2014. Effects of developmental manganese, stress, and the combination of both on monoamines, growth, and corticosterone. Toxicol Rep 1:1046–1061, 10.1016/j.toxrep.2014.10.004.

Wasserman GA, Liu X, Parvez F, Ahsan H, Levy D, Factor-Litvak P, et al. 2006. Water manganese exposure and children’s intellectual function in Araihazar, Bangladesh. Environ Health Perspect 114:124–129, PMID: 16393669.

Wechsler D. 2003. Wechsler Intelligence Scale for Children–Fourth Edition (WISC-IV). San Antonio, TX:The Psychological Corporation.

WHO (World Health Organization). 2011. “Haemoglobin concentrations for the diagnosis of anaemia and assessment of severity.” [accessed Day Month Year].

Yu XD, Zhang J, Yan CH, Shen XM. 2014. Prenatal exposure to manganese at environment relevant level and neonatal neurobehavioral development. Environ Res 133:232–238, PMID: 24971720, 10.1016/j.envres.2014.04.012.

Zoni S, Lucchini RG. 2013. Manganese exposure: cognitive, motor and behavioral effects on children: a review of recent findings. Curr Opin Pediatr 25:255–260, PMID: 23486422, 10.1097/MOP.0b013e32835e906b.

WP-Backgrounds Lite by InoPlugs Web Design and Juwelier Schönmann 1010 Wien