Long-term weight loss effects of semaglutide in obesity without diabetes in the SELECT trial

Abstract

In the SELECT cardiovascular outcomes trial, semaglutide showed a 20% reduction in major adverse cardiovascular events in 17,604 adults with preexisting cardiovascular disease, overweight or obesity, without diabetes. Here in this prespecified analysis, we examined effects of semaglutide on weight and anthropometric outcomes, safety and tolerability by baseline body mass index (BMI). In patients treated with semaglutide, weight loss continued over 65 weeks and was sustained for up to 4 years. At 208 weeks, semaglutide was associated with mean reduction in weight (-10.2%), waist circumference (-7.7 cm) and waist-to-height ratio (-6.9%) versus placebo (-1.5%, -1.3 cm and -1.0%, respectively; P < 0.0001 for all comparisons versus placebo). Clinically meaningful weight loss occurred in both sexes and all races, body sizes and regions. Semaglutide was associated with fewer serious adverse events. For each BMI category (<30, 30 to <35, 35 to <40 and ≥40 kg m^-2) there were lower rates (events per 100 years of observation) of serious adverse events with semaglutide (43.23, 43.54, 51.07 and 47.06 for semaglutide and 50.48, 49.66, 52.73 and 60.85 for placebo). Semaglutide was associated with increased rates of trial product discontinuation. Discontinuations increased as BMI class decreased. In SELECT, at 208 weeks, semaglutide produced clinically significant weight loss and improvements in anthropometric measurements versus placebo. Weight loss was sustained over 4 years. ClinicalTrials.gov identifier: NCT03574597 .

Fig. 1. Percentage change in mean body weight from baseline through week 208 for all patients in-trial and first on-treatment.

a,b, Observed data from the in-trial period (a) and first on-treatment (b). The symbols are the observed means, and error bars are ±s.e.m. Numbers shown below each panel represent the number of patients contributing to the means. Analysis of covariance with treatment and baseline values was used to estimate the treatment difference. Exact P values are 1.323762 × 10⁻⁹⁴ and 9.80035 × 10⁻¹⁰⁰ for a and b, respectively. P values are two-sided and are not adjusted for multiplicity. ETD, estimated treatment difference; sema, semaglutide.

Fig. 2. Variation in weight loss response for semaglutide andplacebo.

a, Categorical weight loss from baseline at week 104 for semaglutide and placebo. Data from the in-trial period. Bars depict the proportion (%) of patients receiving semaglutide or placebo who achieved ≥5%, ≥10%, ≥15%, ≥20% and ≥25% weight loss. b,c, Percentage change in body weight for individual patients from baseline to week 104 for semaglutide (b) and placebo (c). Each patient’s percentage change in body weight is plotted as a single bar.

Fig. 3. Proportion (%) of in-trial patients achieving sex- and race-specific WC cutoff points for increased metabolic risk according to baseline BMI <35 or ≥35 kg m⁻².

WC cutoff points; Asian women <80 cm, non-Asian women <88 cm, Asian men <88 cm, non-Asian men <102 cm.

Fig. 4. Change in BMI category (healthy, overweight, class I obesity, class II obesity and class III obesity) at baseline and week 104 for in-trial patients.

In the semaglutide group, 12.0% of patients achieved normal weight status at week 104 (from 0% at baseline), compared with 1.2% (from 0% at baseline) for placebo. BMI classes: healthy (BMI <25 kg m⁻²), overweight (25 to <30 kg m⁻²), class I obesity (30 to <35 kg m⁻²), class II obesity (35 to <40 kg m⁻²) and class III obesity (BMI ≥40 kg m⁻²).

Fig. 5. Effect of semaglutide treatment or placebo on mean percentage change in body weight from baseline to week 104 by subgroups.

Data from the in-trial period. N = 17,604. P values represent test of no interaction effect. P values are two-sided and are not adjusted for multiplicity. The dots show estimated treatment differences, and the error bars show 95% CIs. Details of the statistical models are available in Methods. ETD, estimated treatment difference; HbA1c, glycated hemoglobin; MI, myocardial infarction; PAD, peripheral artery disease; sema, semaglutide.

Fig. 6. AEs leading to trial product discontinuation for baseline BMI class (<30 kg m⁻², 30 to <35 kg m⁻², 35 to <40 kg m⁻², and ≥40 kg m⁻²); cumulative incidence for first event over time (left) and cumulative mean number of events over time (right).

Data are in-trial from the full analysis set. sema, semaglutide.

Extended Data Fig. 1. Effect of semaglutide treatment or placebo on waist circumference from baseline to week 104 by subgroups.

Data from the in-trial period. N = 17,604. P values represent test of no interaction effect. P values are two-sided and not adjusted for multiplicity. The dots show estimated treatment differences and the error bars show 95% confidence intervals. Details of the statistical models are available in Methods. BMI, body mass index; CI, confidence interval; CV, cardiovascular; CVD, cardiovascular disease; eGFR, estimated glomerular filtration rate; ETD, estimated treatment difference; HbA1c, glycated hemoglobin; MI, myocardial infarction; PAD, peripheral artery disease; sema, semaglutide.