A new study has been appearing in news reports over the past couple of days: a group of researchers found that children drinking water with high levels of manganese had lower IQs than children drinking water with lower levels of manganese. Obviously no parent wants their children to have a lower IQ, but the media reports don’t really provide much information on whether this study indicates any sort of risk to the general population, so I thought I’d take a more in-depth look at the study in the context of what we know about manganese in drinking water supplies.
Manganese (abbreviated Mn) is a naturally occurring metal that is commonly found in groundwater, with highly variable concentrations. It is also found in food and in soil. It is an essential nutrient, and associated with various enzymes in the body, so you need to ingest some manganese. However, at very high exposures it is a neurotoxin (has adverse effects on the brain).
Health Canada evaluated the toxicity of manganese when developing their drinking water guideline, but that was back in 1975. They concluded that manganese was among the least toxic elements. The drinking water guideline they established, 0.15 mg/L, wasn’t actually based on any sort of health effect, but rather is an aesthetic objective based on undesirable taste and the potential for stains on laundry and plumbing fixtures. Historically, manganese toxicity has mainly been associated with inhalation (primarily occupational exposures).
The United States Environmental Protection Agency (US EPA) evaluated the toxicity of manganese in 1995; as part of their review they identified a study in which people exposed to manganese concentrations in water 1.6 to 2.3 mg/L had higher (worse) scores on neurobehavioural tests than people drinking water with lower concentrations of manganese. However, even for the high exposure group, exposure from drinking water was lower than expected exposure from food.
Which leads us to the new study by Bouchard et al., conducted in 8 municipalities in Quebec with a range of manganese concentrations in drinking water ranging from 0.001 mg/L to 2.7 mg/L. They measured the manganese concentrations from kitchen taps (after the filter where one was used) to get a reliable estimate of the exposure concentration. They also estimated manganese exposures from food based on semi-quantitative questionnaires, and measured manganese in hair samples, to get a better picture of overall exposure.
They found that manganese intakes from food were about 1000 times higher than those from water (2335 ug/kg/month from food, 1.6 ug/kg/month from water and 1.9 ug/kg/month from water incorporated into food). However, manganese concentrations in hair appeared to be associated with intake from water, but not with intake from food. They then found that high manganese intake from water was associated with lower IQ; again, manganese intake from food did not appear to affect IQ. The difference in IQ between the highest and lowest exposure groups was over 6 points, and a 10-fold increase in manganese was associated with an IQ decrease of 2.4 points.
The first thing I ask myself when I see results like this is whether there could have been any other factors affecting the results – correlation between two variables (e.g. manganese concentration and IQ) does not necessarily mean that one causes the other. The authors looked at the influence of family income, family structure, maternal education, child birth weight, child’s rank in family, maternal smoking, alcohol consumption during pregnancy maternal depressive symptoms, and other factors, and found they did not affect the results significantly. They also looked at concentrations of lead, arsenic, copper and zinc in tap water, in case another substance was causing the effect, but again did not find a significant association. One thing that did show up was that, on average, manganese concentrations were lower at houses with private wells than those using public wells – perhaps there is some difference between people that have private vs. public wells that could account for IQ differences, but since the study accounted for socioeconomic factors there’s nothing obvious. Given that there are other studies indicating that manganese can have potential neurotoxic effects, the conclusion that the reduced IQ scores are probably a result of manganese is reasonable.
The next question is why the association is only with manganese in drinking water, and not in food. The authors speculate that this could arise from different metabolism of manganese from water than food (chemicals are commonly more readily absorbed by the body from water than food). They also suggest there could be differences in the chemical form of manganese, which could affect its toxicity. Since they found that IQ was better correlated with concentrations in tap water than with exposure from water, another suggestion was that it could actually be inhalation of manganese during showering and other activities that results in the toxic effects – which would be consistent with the neurotoxicity observed in occupational inhalation exposures, though there isn’t really any convincing evidence to support this.
So what are the implications for the general population? In this study, discernible effects on IQ were really only evident at high concentrations (greater than 0.1 mg/L). Most Canadian and US water supplies have lower concentrations than this, but not all – it’s not uncommon for manganese to be naturally high in groundwater wells, and even some surface water bodies (e.g. the Carrot River in Saskatchewan was found to have concentrations up to 1.7 mg/L in an Environment Canada study). There are people being exposed to manganese concentrations in drinking water in the range at which effects were observed in other parts of North America as well.
What can we do about this though? High concentrations of manganese in drinking water are usually natural, so it’s not a case of preventing pollution or cleaning up a contaminated aquifer. About all we can do right now is treat water with high concentrations of manganese to reduce the concentrations. Unfortunately this study by itself probably isn’t enough to tell us what the “safe” concentration in water is. I’d personally be concerned about concentrations above the existing Health Canada aesthetic objective of 0.15 mg/L though. Follow-up studies on other populations are needed to confirm these results and to help narrow in on what concentrations in water start to have adverse effects; it would also be useful if these studies include a more in-depth analysis of the chemical form of manganese – if the toxic effect is due to the form present, then a change to water pH or hardness might significantly reduce (or increase) the effect.
Bouchard, M., Sauvé, S., Barbeau, B., Legrand, M., Brodeur, M., Bouffard, T., Limoges, E., Bellinger, D., & Mergler, D. (2010). Intellectual Impairment in School-Age Children Exposed to Manganese from Drinking Water Environmental Health Perspectives DOI: 10.1289/ehp.1002321