Gum disease tracked 18 to 25% more new diabetes across 28 studies and 300,000 people

A diabetes review appointment checks feet, eyes, kidneys and blood pressure. It does not look in the mouth.

Botelho and colleagues, writing in The Lancet Public Health with WHO funding, screened 45,477 records and pooled 28 longitudinal studies from 16 countries covering over 300,000 participants to ask whether it should. Their answer is that individuals with periodontitis at baseline showed a 18-25% higher occurrence of newly incident type 2 diabetes during follow-up.

The relationship also runs the other way, though less reliably, which is the more interesting half of the finding.

Why longitudinal is the word that matters

Plenty of studies have found that people with diabetes have worse teeth. Almost none of them could say which came first.

That is the gap this review targets. Every included study was longitudinal: it measured oral health at a starting point, then followed people forward and counted who developed diabetes. A snapshot study, which photographs both conditions at once, cannot distinguish a cause from a consequence from a shared origin.

Sequence is not cause, and the authors are careful about the difference. But sequence is the precondition for cause, and this is the first synthesis with enough longitudinal evidence to establish it in both directions.

Three oral measures, one direction

The strongest results all point the same way, from mouth to metabolism.

Periodontitis, the advanced form of gum disease where bone recedes from the tooth, carried the headline 18 to 25% figure. Tooth loss was also associated with newly incident type 2 diabetes over time, at between 12 and 20% higher. And individuals who were edentulous at baseline, meaning they had no natural teeth left, showed a higher occurrence of newly incident type 2 diabetes, at 30% higher.

Three measures of the same underlying thing, ordered by severity, producing effects ordered the same way. That internal gradient is what makes the pattern harder to dismiss than any single estimate.

The asymmetry the authors flag

Bidirectional findings usually get reported as though both arrows were equally solid. Here they are not, and the paper says so.

The authors describe the evidence as asymmetric: people with periodontitis and edentulism are more likely to be diagnosed with type 2 diabetes than people without periodontitis and edentulism in the future, while people with diabetes are more likely to be diagnosed with periodontitis than people without diabetes, but with less consistent evidence.

So the mouth-to-diabetes direction is the better-supported one. That is the opposite of how this relationship is usually described in clinical practice, where gum disease is framed as a complication of diabetes rather than a possible early signal of it.

What could be producing this without any causal link

Two things, and both are serious.

Shared causes come first. Smoking wrecks gums and raises diabetes risk. So does obesity, so does poverty, and so does the absence of regular healthcare. A person who has not seen a dentist in a decade probably has not seen a doctor either, and the diagnosis of diabetes requires someone to look. Adjustment in these studies handles what the original researchers measured, which is never all of it.

Detection is the second. Someone with visibly bad teeth may attract more clinical attention once they do present, which raises the chance a latent diabetes is found and recorded. That would produce this association with no biological link at all.

The proposed mechanism is inflammatory: chronic gum infection maintains a body-wide inflammatory state, and inflammation impairs insulin sensitivity. It is plausible, and this design cannot test it.

What is missing from the picture

Two of the four oral conditions the WHO tracks could not be assessed at all.

Evidence regarding dental caries was scarce and heterogeneous, precluding meta-analysis, and no longitudinal evidence was identified for oral cancer. Cavities are the most common chronic disease on earth and nobody has followed them forward against diabetes in a way that can be pooled. That is a gap in what has been studied, not a finding that nothing is there.

What to do with a dental appointment and a diabetes review

The practical version is unglamorous and cheap: treat the two as connected, because the health system does not.

For someone with diabetes, gum disease is worth raising rather than assuming the dentist knows. For someone with periodontitis and no diabetes diagnosis, this evidence is a reason to have blood sugar checked, particularly alongside the other risk factors. Type 2 diabetes is a disease in which your blood glucose, or blood sugar, levels are too high, and it is frequently present for years before anyone measures it.

The authors’ own conclusion is a systems argument rather than a personal one: their findings support integrating oral health within interdisciplinary diabetes prevention and management strategies. Which is a formal way of saying somebody should look in the mouth.