Summary: In this post-hoc subgroup analysis of the double-blind, placebo-controlled STEP 6 trial (401 Japanese and Korean adults with overweight or obesity), once-weekly subcutaneous semaglutide 2.4 mg produced clinically relevant weight loss across every baseline subgroup, with mean change from baseline ranging from -9.40% to -16.42%. Significant treatment-by-subgroup interactions were detected for sex, and for type 2 diabetes and dyslipidaemia at the 2.4 mg dose, but these exploratory signals do not establish that any single factor reliably predicts response.
PICO Summary
| Element | Detail |
|---|---|
| Population | 401 Japanese and Korean adults (148 female, 253 male) with overweight or obesity, including a subset of Japanese participants with type 2 diabetes; post-hoc subgroup analysis of the double-blind, randomised STEP 6 trial (Japan and South Korea). |
| Intervention | Once-weekly subcutaneous semaglutide, 2.4 mg or 1.7 mg, for 68 weeks. Subgroups defined by baseline body weight, BMI, age, sex, glycaemic status, dyslipidaemia, and hypertension. |
| Comparison | Placebo (subcutaneous), within the same randomised trial. The comparator was placebo, not an active weight-loss agent. |
| Outcome | Mean change from baseline in body weight with semaglutide 2.4 mg was clinically relevant across all subgroups, ranging from -9.40% to -16.42%; estimated treatment differences favoured both semaglutide doses over placebo. Significant treatment-by-subgroup interactions versus placebo at week 68: sex (semaglutide 1.7 mg, p = 0.0008; 2.4 mg, p = 0.0005); type 2 diabetes (2.4 mg only, p = 0.0381); dyslipidaemia (2.4 mg only, p = 0.0181). Confidence intervals, absolute risk reduction, and number needed to treat are not applicable to this continuous-outcome subgroup analysis and were not the focus of the report. |
Expert Commentary
This is a post-hoc subgroup analysis of an existing randomised trial, and it should be read as hypothesis-generating rather than confirmatory. The headline message is reassuring and consistent with the parent study: semaglutide 2.4 mg lowered body weight by roughly 9 to 16 percent across a broad range of baseline characteristics in an East Asian population, so the drug works across the demographic spectrum rather than only in selected patients. The more nuanced claim, that sex, type 2 diabetes, and dyslipidaemia modify the size of that response, rests on interaction tests that were not pre-specified as the trial’s primary aim. The single most important limitation is precisely this multiplicity: numerous subgroups and two doses were examined, several interactions were significant only at the 2.4 mg dose, and isolated p-values around 0.02 to 0.04 are fragile once the number of comparisons is considered. The analysis was funded by the manufacturer, and several authors are company employees, which warrants the usual caution even though the effect sizes are biologically plausible rather than implausibly large. Can I use this with my patients? Yes, for confirming that an East Asian adult with overweight or obesity, including those with type 2 diabetes, can expect meaningful weight loss on semaglutide; no, not for selecting or excluding patients on the basis of these subgroup signals. Clinicians should treat the subgroup differences as questions for prospective study, not as prescribing rules.
References
Kadowaki T, Lee SY, Ogawa W, Nishida T, Overvad M, Tobe K, et al. Clinical characteristics affecting weight loss in an East Asian population receiving semaglutide: A STEP 6 subgroup analysis. Obes Res Clin Pract. 2025;18(6):457-464. doi:10.1016/j.orcp.2025.01.002
