Have obesity-related hospital admissions doubled in six years?

When a statistic seems incredible, it probably isn't true

The media usually wait until the New Year - or at least the Christmas perineum - before going all-in with the annual glut of obesity sensationalism. This year, however, The Times has started early. For some reason, the newspaper has become obsessed with this issue. Maybe it is what their readers want or maybe it is what editor Tony Gallagher, formerly of the Daily Mail, thinks they want.

Two weeks ago, it put a story about Henry Dimbleby making a speech on the front page, along with a ridiculous claim about obesity ‘costing’ the UK £98 billion a year. Since then, it has covered some drivel about ultra-processed food ‘hijacking the brain’, a non-story about KFC opposing stupid planning laws and some silliness about the food industry using the ‘Big Tobacco playbook’.

Obesity is the main story on The Times front page once again today, as if it there was nothing else going on in the world. This time, the claim is that the number of obesity-related hospital admissions has doubled in the space of six years.

When statistics like this rise or fall dramatically in a short space of time, it is usually worth looking for a change in methodology. Sure enough, what we have here is a sharp rise in the number of people going to hospital with a secondary diagnosis of obesity. Back in 2009/10, only 142,219 admissions had obesity mentioned as a secondary diagnosis. By 2022/23, this had risen to 1,235,961. In the same period, the rate of obesity among adults has risen only slightly. A ten-fold rise in admissions isn’t very plausible, especially since admissions for which there was a primary diagnosis of obesity haven’t risen at all.

A secondary diagnosis of obesity doesn’t mean that your illness was caused by obesity, nor does it mean that you wouldn’t have got ill if you hadn’t been obese, as the statisticians who collate this data explain…

A secondary diagnosis of obesity does not necessarily indicate obesity as a contributing factor for the admission, but may instead indicate that obesity is a factor relevant to a patient’s episode of care.

For example, if an obese woman goes to hospital to have a baby, that is not an obesity-related admission in the sense that it could have been avoided if she wasn’t obese, but it will be recorded as such if a medic thinks her obesity is relevant to her care. This is not a random example. Maternal care is, by some distance, the most common type of admission to have a secondary diagnosis of obesity recorded.

Recording a secondary diagnosis is essentially optional. A lot depends on whether the medic can be bothered to take a note of it and whether the medic thinks it is relevant. In recent years, the issue of obesity has become more salient and medics have been encouraged to take a note of it (and to record more secondary diagnoses in general). As a result, NHS England statisticians advise people to interpret year-to-year changes with caution…

There is continuing evidence that recording of secondary diagnosis codes is improving over time, which may have contributed (though not fully) to the increases seen in ‘admissions where obesity was a factor’ over the last ten years.

It would not be surprising if the number of hospital admissions that are genuinely related to obesity had risen in the last decade, but the figures cited by The Times do not prove that they have risen and it is extremely unlikely that they have doubled in six years. It is perfectly appropriate to record admissions for which obesity was a contributing factor, but a secondary diagnosis does not necessarily mean that obesity was a contributing factor.

If there had been a dramatic rise in obesity-related admissions, you would expect to see a commensurate rise in obesity-specific admissions, but there is no sign of this in the data. The peak in admissions for which obesity was the primary diagnosis took place in 2011/12 when there were 11,736 (which is too long ago to be included in the graph below). In 2022/23, there were only 8,716. There were 20 million hospital admissions overall in 2022/23, so obesity was the direct cause of only 0.04% of them.

The Times briefly acknowledges this towards the end of its front page story, but with a little get-out clause.

Weight-loss surgery has yet to return to pre-pandemic levels, with 5,099 admissions last year, compared with more than 6,000 five years ago, which some have attributed to pressures on the NHS.

This has meant admissions where the main reason is obesity have yet to return to levels of more than 10,000 seen before Covid-19.

Maybe it is, but the number of admissions wasn’t rising before the pandemic and ‘pressures on the NHS’ clearly haven’t reduced the number of admissions using the broader measure.

In summary, the statistic cited by The Times is more likely to mislead than educate and the suggestion that there has been a dramatic rise in hospital admissions due to obesity in the space of a few years is very likely to be wrong.