Does Pollution Cause Diabetes? Sorting the Strong Evidence From the Thin

It's a fair question, and it usually comes from a real place: someone develops diabetes without the risk factors they expected to see — not overweight, reasonably active, no obvious family history — and starts looking around at the air, the water, the plastic. The honest answer is that pollution is a genuine contributor to type 2 diabetes at the population level, that some exposures have far better evidence behind them than others, and that the word "cause" is doing more work in that question than the science can quite support.

Why "cause" is the hard word here

You cannot run the experiment that would settle this. Nobody can randomly assign one group of people to drink contaminated water for twenty years and another to drink clean water, then compare their HbA1c. So the evidence comes from observing populations, and observation can only ever say that an exposure travels with a disease — never, on its own, that it produced it.

What researchers do instead is look for the fingerprints of a real effect: does more exposure mean more disease, does the association survive adjusting for weight and income and diet, and is there a biological mechanism that makes sense? Where all three line up, epidemiologists are willing to attribute a share of a disease burden to an exposure. That's exactly what the Global Burden of Disease programme has done for air pollution — and the resulting number is not small.

A 2022 analysis in The Lancet Planetary Health (opens in a new tab)00122-X/fulltext) attributed about 13.4% of type 2 diabetes deaths worldwide to outdoor PM2.5 and a further 6.5% to household air pollution — roughly a fifth of the global burden between them. In that analysis, outdoor fine particles ranked as the third leading risk factor for type 2 diabetes globally, behind only high blood glucose itself and high body-mass index. So this belongs in the same conversation as weight and diet, not in a footnote.

The exposure most people overlook is indoors

Outdoor smog gets the attention every winter, and it deserves it — we've covered what it does and how to get through a smog season without losing your daily walk. But that 6.5% figure for household air pollution is a separate exposure with a very different profile, and in Pakistan it is arguably the more relevant one.

The World Health Organization (opens in a new tab) estimates that around 2.1 billion people — a quarter of the world — cook on open fires or inefficient stoves burning wood, animal dung, crop waste, coal or kerosene, and attributes about 2.9 million deaths a year to the resulting air. Those deaths are overwhelmingly cardiovascular: 32% ischaemic heart disease and 23% stroke, the same two conditions diabetes already raises your risk of.

The exposure is also unevenly distributed within a household. WHO notes plainly that women and children, being the ones who cook and gather fuel, carry the greatest share of it. If there is a woman in your family who has spent decades at a chulha in an unventilated kitchen, she has an exposure history that no amount of checking the outdoor AQI captures — and it is one of the few environmental exposures on this page that a household can genuinely change.

• A switch to LPG or natural gas for daily cooking removes the exposure at source, which is why it's the intervention public health bodies push hardest.
• Where that isn't possible, ventilation matters more than it sounds: cooking with a window or door open, or moving the stove outside or under a covered outdoor space, meaningfully cuts what gets breathed in.
• Kerosene is included in WHO's list of polluting fuels — a kerosene stove or lamp is not a clean alternative to wood.
• Indoor smoke from other sources adds to the same load: cigarettes, sheesha, coal angeethis for warmth, and mosquito coils burning in a closed room.

Arsenic in groundwater: the Pakistan-specific one

This is the exposure that most deserves to be better known here, because it is both well-evidenced and geographically specific. Arsenic occurs naturally in the rock and sediment under the Indus Plain. Decades of heavy irrigation in a dry climate have driven evaporation that raises soil pH, and higher pH makes arsenic more soluble — so it ends up dissolved in the groundwater that tube wells and hand pumps draw from.

A 2017 mapping study published in Science Advances, reported by AAAS (opens in a new tab), sampled groundwater at nearly 1,200 sites across the country and estimated that as many as 60 million people relying on Indus Plain groundwater are at risk. Concentrations around 500 µg/L were measured near Lahore and Muzaffargarh, and above 200 µg/L in the south including near Hyderabad. The WHO guideline value (opens in a new tab) is 10 µg/L.

The diabetes link is stated directly by WHO, which lists cardiovascular disease and diabetes among the conditions long-term arsenic exposure has been associated with. The clearest human data comes from Bangladesh, where a study of 1,004 adults (opens in a new tab) found roughly double the odds of type 2 diabetes among people drinking water above 50 µg/L compared with those at or below it, with risk rising further among those exposed for more than ten years. That study was cross-sectional, so it can't establish sequence, and its authors say so. But the association is strong and consistent at the higher exposure levels seen across this region.

What's actually actionable: find out whether your area has a known problem. Provincial public health and water agencies, and some NGOs, do arsenic testing, and WHO points to source testing and marking safe versus unsafe wells as the intervention that reduces exposure fastest. If your household drinks from a hand pump or private tube well in Punjab or Sindh, that test is worth chasing down. If your water does test high, the fix is a different source or a treatment system rated for arsenic — a standard candle filter is not one.

Plastics and everyday chemicals: real question, thin evidence

This is where the internet gets loudest and the science gets quietest. Endocrine-disrupting chemicals — BPA in plastics, phthalates, pesticide residues, non-stick coatings — are a plausible concern because they interact with hormone signalling, and laboratory work does show effects on beta cells and insulin pathways.

But an evidence-based review in Diabetologia (opens in a new tab) that went through this literature found more than 40 cross-sectional studies and only seven prospective ones. Its verdict was moderate evidence for one exposure — DDT and its metabolite p,p'-DDE — and for BPA, phthalates and PFASs, evidence it described as scarce for exactly that reason. Cross-sectional studies measure the chemical and the disease at the same moment, which cannot tell you which came first, and in metabolic disease that matters: body fat stores some of these compounds, so people with more of it may show higher levels as a consequence rather than a cause.

None of that means the concern is baseless — it means the evidence isn't yet strong enough to rank this alongside cooking smoke or arsenic. Reasonable, low-cost habits are fine to adopt if they cost you nothing: not microwaving food in plastic containers, not leaving bottled water in a hot car, washing produce. What isn't reasonable is spending money you don't have on detox products or expensive replacements while a hand pump of untested water sits in the courtyard.

Ranking these honestly

One thing worth holding onto: every exposure on this list raises risk in a population, and none of them determines an individual outcome. The things that move your own numbers week to week — what you eat, whether you move, how you sleep, whether you take your medication — remain the things with the shortest distance between action and result. Environmental risk is real, and it's mostly addressed at the level of a household or a city rather than a day.

Your own pattern is the most useful data you have

Population statistics can't tell you what's moving your readings. A few weeks of logged glucose alongside meals, activity and sleep usually can — and it's the fastest way to find the levers that actually respond to you.

Start logging with Diatic (opens Diatic on Google Play in a new tab)