STEP 10 reverses prediabetes in 81% of participants

Results from STEP 10, a phase 3 trial of once-weekly injectable semaglutide 2.4 mg, were published July 29, 2024, showing that 81% of adults with obesity and prediabetes had blood sugar back in the normal range after 52 weeks, compared with 14% of those on placebo ^[1].

The trial enrolled 207 adults at 30 sites in Canada, Denmark, Finland, Spain and the UK between September 16 and December 29, 2021 ^[1]. To qualify, participants had to be 18 or older with a BMI of 30 kg/m² or higher and prediabetes by UK National Institute for Health and Care Excellence criteria — an HbA1c of 6.0–6.4% (42–47 mmol/mol) or a fasting plasma glucose of 5.5–6.9 mmol/L at screening ^[1]. They were randomly assigned 2:1 to semaglutide 2.4 mg (138 people) or placebo (69 people), and everyone received diet and physical activity counseling for 52 weeks, followed by a 28-week period off treatment ^[1].

What the numbers showed

At the start, the average participant was 53 years old, weighed 111.6 kg (about 246 pounds), had a BMI of 40.1 kg/m², a waist circumference of 120.1 cm, an HbA1c of 5.9% and a fasting glucose of 5.9 mmol/L ^[1]. Seventy-one percent were female and 88% were White ^[1].

After 52 weeks, body weight fell 13.9% with semaglutide versus 2.7% with placebo, an estimated treatment difference of 11.2 percentage points (95% CI −13.0 to −9.4; p<0.0001) ^[1]. On the second primary endpoint — reverting to normoglycemia, defined as HbA1c below 6.0% and fasting glucose below 5.5 mmol/L — 103 of 127 people (81%) on semaglutide met the mark versus nine of 64 (14%) on placebo, an odds ratio of 19.8 (95% CI 8.7 to 45.2; p<0.0001) ^[1].

On safety, serious adverse events occurred in 12 semaglutide participants (9%) and six placebo participants (9%) ^[1]. Adverse events leading to stopping treatment occurred in eight people (6%) on semaglutide versus one (1%) on placebo ^[1]. The authors reported no new safety signals and said the tolerability profile matched earlier studies and the GLP-1 receptor agonist class generally ^[1].

The trial was funded by Novo Nordisk, and three of the authors are employees and shareholders of the company; several others reported consulting fees, research grants or speaking honoraria from Novo Nordisk and other drugmakers ^[1].

Why it matters for patients

Prediabetes is common among adults with obesity, and STEP 10 is the first large randomized look at what semaglutide 2.4 mg does specifically in that group ^[1]. The headline finding is that four out of five treated participants no longer met the blood sugar definition of prediabetes after a year, versus roughly one in seven on placebo plus lifestyle counseling ^[1]. The investigators concluded the results "support the potential use of semaglutide 2·4 mg as a treatment option for individuals with obesity and prediabetes to achieve reversion to normoglycaemia" ^[1].

Some important caveats come straight from the trial's design. The endpoint was blood sugar returning to a normal range at week 52 — not a measure of whether type 2 diabetes was prevented over years ^[1]. The study population was mostly female (71%) and mostly White (88%), which limits how well the results describe a more diverse US population ^[1]. And the trial was relatively small, with 207 participants total ^[1].

The biggest open question is durability. STEP 10 included a 28-week off-treatment follow-up period after the 52 weeks of dosing, but the published abstract does not report what happened to weight or blood sugar during that stretch ^[1]. Whether normal glucose levels held after people stopped the drug is not answered in the material available here.

What happens next

The trial is listed as complete on ClinicalTrials.gov under NCT05040971 ^[1]. The results appeared in The Lancet Diabetes & Endocrinology, published online July 29, 2024, with an accompanying commentary by M. Bergman titled "Prediabetes remission in people with obesity" in the journal's September 2024 issue ^[1]. Whether these findings lead to any change in how semaglutide is labeled or prescribed for prediabetes is not addressed in the source.