Arsenic at the U.S. legal limit doubled type 1 diabetes risk in children

Children whose mothers drank water at or above 10 micrograms of arsenic per liter during pregnancy were about twice as likely to develop type 1 diabetes, with a hazard ratio of 2.04 and an interval of 1.15 to 3.60.

That 10 figure is not unusual. It is the enforceable American limit, which the EPA writes as 0.010 mg per liter and has enforced since January 2006. The top exposure band in this study is the concentration a fully compliant US utility is permitted to deliver on an ordinary Tuesday.

Here is how large the study was, what it can and cannot establish, and what a household can actually check.

563,871 births, matched home by home

Denmark maintains a register of everyone: who was born, where they lived, and what a doctor recorded afterwards. Researchers matched that against the water supply serving each household, one home at a time, for every singleton born in the country between 2002 and 2012 whose household drew from a public supply.

The scale matters because no American study has come close, and probably never will. This country keeps no register that could support one.

Diagnoses came from national patient and prescription registries rather than questionnaires, so nobody had to remember anything correctly years later. That removes the largest weakness in most environmental health research.

Exposure was matched to the individual home rather than to a regional average standing in for a whole county.

The adjustment list is what makes the result meaningful: calendar year, sex at birth, maternal age, smoking, BMI, education, income, and parental type 1 diabetes. That last one carries the most weight, because type 1 runs in families, and a study omitting it would be measuring inheritance and calling it water.

The number that doubled risk is the number your utility may legally deliver

This is the part worth sitting with. Nobody broke a rule for this finding to happen.

A water company is required to stay under 10 micrograms per liter. A supply delivering 9 is compliant, unremarkable, and produces no violation notice. The Danish cohort found risk roughly doubling at and above that same concentration, with everyone involved in compliance throughout.

That is what makes a Danish study worth American attention. It is not describing a failure. It is describing what the rules permit.

What one study can and cannot settle

This is the first study of its kind, and its authors say so. They describe arsenic as a potential modifiable risk factor for type 1 diabetes rather than an established cause.

That is the right weight to give it. A single cohort, even at this size, establishes that a question is worth asking rather than settling what the answer is. Settling it requires the same linkage run somewhere else and finding the same pattern.

It also remains a Danish cohort: different geology, different treatment, and everything downstream of a tap unlike an American one. Environmental Health News, covering the study, noted that arsenic levels in Danish drinking water are similar to those in the US and other countries, which is what makes the comparison reasonable rather than decorative. What transfers is the exposure range and the shape of the finding, not a number anyone can apply to their own street. The pattern repeats where the compound was used: two childhood cancers came back higher near PFOS contamination.

One additional result is unexplained. Children born above median birth weight showed higher risk across every exposure band, which the authors flag rather than interpret.

Wells are the gap in this

Arsenic is geological. It leaches out of rock and varies from one property to the next in ways no county map resolves, so a neighbour’s clean result says very little about any given well.

Private wells are not necessarily worse. Nobody has looked. The Danish cohort covered households on public supply, which means well owners carry the same exposure question and none of the answers.

Arsenic also sits outside the standard four-test panel, which covers bacteria, nitrate, pH and dissolved solids. None of those touch a metal, and no laboratory adds arsenic unless it is named specifically.

What to check, and when

On a public supply, arsenic is reported by name in the Consumer Confidence Report a utility owes its customers by July 1. That figure is a system-wide result rather than a reading from one kitchen tap.

On a well, it requires a paid test with arsenic named specifically in the order. On most laboratory price lists the add-on is measured in tens of dollars rather than hundreds, and it is worth asking the cost before sending anything.

Where a result comes back high, treatment is specific. Reverse osmosis at a single tap handles arsenic, as do some anion exchange and activated alumina systems. A carbon filter does not, and neither does a standard pitcher, which is where many households believe they have solved a problem they never touched.

Test before a pregnancy rather than during one. The exposure window in this study was gestation, and that is the one window a result cannot be applied to retroactively.