Obesity, diabetes and the GLP-1 drugs

Obesity prevalence has risen for four decades in nearly every high-income country and is still rising in most, though the latest surveys in a few, the United States among them, show a plateau whose cause is argued over. Childhood obesity has risen fastest in the most deprived areas. Type 2 diabetes prevalence rises with age and with obesity, so it grows even where its incidence has levelled off, as it did in several countries in the 2010s; the number living with diabetes is a record everywhere.

Diagnosis has broadened. The move to HbA1c testing and the category of pre-diabetes mean that more people are found earlier, and prevention programmes for those at risk exist in England, the United States and elsewhere, with modest measured effects on progression.

The drugs are the new term. The trial evidence is not in doubt: in adults with obesity, semaglutide and tirzepatide produce weight loss of a size previously seen only with surgery, and the SELECT trial showed a reduction in major cardiovascular events in people with obesity but without diabetes. What is still being established is what happens at population scale. Registry and claims studies from Denmark and the United States show uptake concentrated among people with diabetes, among the insured and the better off, and, in the US, high discontinuation within the first year. Real-world weight loss is smaller than in trials, partly because of dose and persistence. No national statistical series has yet shown a change in the incidence of diabetes or cardiovascular disease attributable to these drugs; the exposed population is still too small and too recent. Oral formulations, price competition and widening eligibility are changing who takes them, and the countries with linked prescribing and outcome data — Denmark above all — will show the population effect first.

The deaths are the hardest part to read. Diabetes is rarely the recorded underlying cause and usually a contributing one, so its mortality is best seen in multiple-cause data; the deaths attributable to obesity depend on a modelled counterfactual and on how body mass is measured, and the estimates differ widely.

Status

building

Direction

strengthening

Would change our view

A national registry showing a measurable fall in cardiovascular events or new diabetes attributable to uptake; persistence data showing the effect is largely lost on discontinuation; or a national prevalence series turning down and staying down.

Trackers

Gaps

Uptake by socioeconomic group outside the US and Denmark; persistence beyond two years; measured rather than self-reported body mass in most surveys; almost nothing yet from Japan or France.