SELECT Trial: Beyond MACE - Secondary Endpoints (2026)

The SELECT trial is famous for one number: a 20% reduction in major adverse cardiovascular events (MACE) in non-diabetic adults with heart disease and obesity on weekly semaglutide 2.4 mg (Lincoff 2023 ^[1]). But that headline result was only the beginning. Over the following two years, the investigators published a stack of prespecified secondary analyses dissecting what else happened across the trial's 17,604 participants — and those papers tell a richer story. Semaglutide improved outcomes in people who already had heart failure regardless of subtype (Deanfield 2024 ^[2]), lowered a composite kidney endpoint by 22% and protected eGFR (Colhoun 2024 ^[3]), produced 10.2% weight loss sustained for four years (Ryan 2024 ^[4]), cut progression to type 2 diabetes from 6.9% to 1.5% (Kahn 2024 ^[6]), and — most strikingly — delivered its cardiovascular benefit largely independent of how much weight people actually lost (Deanfield 2025 ^[5]). This article walks through those secondary endpoints one by one. For the headline trial design and the primary MACE result, start with our SELECT trial overview.

Quick recap: what SELECT was

SELECT (Semaglutide Effects on Cardiovascular Outcomes in People With Overweight or Obesity) randomized 17,604 adults aged 45+ with a BMI of 27 or higher and established cardiovascular disease — prior heart attack, prior stroke, or symptomatic peripheral artery disease — but no diabetes, to weekly subcutaneous semaglutide 2.4 mg or placebo, on top of standard care, for a mean of about 40 months (Lincoff 2023 ^[1]). The primary endpoint, a composite of cardiovascular death, nonfatal myocardial infarction, or nonfatal stroke, fell from 8.0% on placebo to 6.5% on semaglutide — a hazard ratio of 0.80 (95% CI 0.72–0.90; P<0.001). That primary result, and the full trial design, are covered in our overview article. Everything below is the secondary and prespecified subgroup evidence published afterward.

Heart failure: benefit across every subtype

One of the most clinically important secondary papers was the prespecified heart-failure analysis (Deanfield 2024, Lancet ^[2]). At enrolment, 4,286 of the 17,604 participants (24.3%) had a history of investigator-defined heart failure. Among them, 53.0% had heart failure with preserved ejection fraction (HFpEF), 31.4% had heart failure with reduced ejection fraction (HFrEF), and 15.5% had unclassified heart failure. This matters because HFpEF — the "stiff heart" phenotype tightly linked to obesity — historically had almost no effective drug therapy.

In participants with a history of heart failure, semaglutide reduced MACE with a hazard ratio of 0.72 (95% CI 0.60–0.87) and reduced a composite heart-failure endpoint with a hazard ratio of 0.79 (95% CI 0.64–0.98). Crucially, the benefit held across both phenotypes: in HFrEF the MACE hazard ratio was 0.65 (95% CI 0.49–0.87) and in HFpEF it was 0.69 (95% CI 0.51–0.91), with the composite heart-failure endpoint trending the same direction in each subtype (Deanfield 2024 ^[2]). The authors concluded that semaglutide improved outcomes in people with and without clinical heart failure, regardless of heart-failure subtype.

Kidney outcomes: a 22% lower composite endpoint

The prespecified kidney analysis (Colhoun 2024, Nature Medicine ^[3]) is one of the strongest secondary signals in the whole program. The main composite kidney endpoint — death from kidney disease, start of chronic kidney-replacement therapy, persistent eGFR below 15, a persistent ≥50% drop in eGFR, or onset of persistent macroalbuminuria — occurred in 1.8% on semaglutide versus 2.2% on placebo, a hazard ratio of 0.78 (95% CI 0.63–0.96; P=0.02). That is a roughly 22% relative reduction in serious kidney events.

Semaglutide also protected kidney function over time. The treatment benefit on eGFR at 104 weeks was +0.75 mL/min/1.73m² overall (95% CI 0.43–1.06; P<0.001), and the benefit was much larger — +2.19 mL/min/1.73m² (95% CI 1.00–3.38; P<0.001) — in the subgroup who started with reduced kidney function (baseline eGFR below 60). Albuminuria fell too: the urine albumin-to-creatinine ratio dropped about 10.7% overall, and roughly 27% in people who entered with elevated albuminuria (Colhoun 2024 ^[3]). For a population with obesity and vascular disease — both major drivers of chronic kidney disease — this is a meaningful organ-protection signal layered on top of the cardiovascular one.

Body weight: 10.2% loss, sustained for four years

Because SELECT ran for nearly four years, it answered a question shorter obesity trials could not: does the weight loss last? The long-term weight analysis (Ryan 2024, Nature Medicine ^[4]) showed weight loss with semaglutide continued through about week 65 and was then sustained out to 208 weeks (four years). At 208 weeks the mean reductions versus placebo were:

• Body weight: −10.2% with semaglutide vs −1.5% with placebo (P<0.0001).
• Waist circumference: −7.7 cm vs −1.3 cm (P<0.0001).
• Waist-to-height ratio: −6.9% vs −1.0% (P<0.0001).

Categorical responses were substantial. By 104 weeks, among semaglutide-treated participants, 67.8% lost at least 5%, 44.2% lost at least 10%, 22.9% lost at least 15%, and 11.0% lost at least 20% of body weight. More than half (52.4%) improved by at least one BMI category versus 15.7% on placebo, and 12.0% reached a healthy BMI (under 25) versus just 1.2% on placebo. Clinically meaningful weight loss occurred in both sexes and across all races, body sizes, and regions, with women showing a somewhat larger treatment difference (−11.1% vs −7.5% in men) (Ryan 2024 ^[4]). This is the first long-term, four-year demonstration that semaglutide-driven weight loss is durable in a non-diabetic population with established heart disease.

The big one: cardiovascular benefit independent of weight lost

Perhaps the most consequential secondary finding is that SELECT's cardiovascular protection did not track tightly with how much weight a person lost. A prespecified analysis of cardiovascular outcomes by baseline and change in adiposity (Deanfield 2025, Lancet ^[5]) examined whether MACE reduction depended on starting BMI, waist circumference, or the magnitude of weight change on treatment. The benefit was broadly consistent across baseline adiposity and — importantly — was not strongly explained by the amount of weight lost. Combined with the observation that the MACE curves separated early, before much weight had come off, this points to mechanisms beyond fat loss alone: improvements in blood pressure, inflammation, glucose handling, and direct vascular effects (Deanfield 2025 ^[5]).

Glycemia: preventing diabetes before it starts

Although SELECT enrolled people without diabetes, most had prediabetes at baseline — so what happened to their glucose was a natural secondary question. The glycemia analysis (Kahn 2024, Diabetes Care ^[6]) found that by week 156, only 1.5% of semaglutide-treated participants had progressed to biochemical diabetes versus 6.9% on placebo (P<0.0001), with a number-needed-to-treat of about 18.5 to prevent one case. Going the other direction, 69.5% of semaglutide participants were normoglycemic at week 156 versus 35.8% on placebo (P<0.0001) — meaning semaglutide both prevented progression to diabetes and pushed many people with prediabetes back toward normal glucose.

A companion analysis looked at whether baseline HbA1c or change in HbA1c explained the cardiovascular benefit (Lingvay 2024, Diabetes Care ^[7]). Consistent with the adiposity story above, the MACE reduction was broadly present regardless of glycemic starting point — reinforcing that no single intermediate (weight, glucose, or HbA1c) fully accounts for SELECT's cardiovascular result.

Subgroups: sex, BMI, and consistency

Across the secondary analyses, the consistent message is breadth. Weight loss and anthropometric improvement were durable in both men and women and across every BMI category (under 30, 30–35, 35–40, and 40+), with lower rates of serious adverse events on semaglutide in each BMI band (Ryan 2024 ^[4]). Cardiovascular benefit was consistent across baseline adiposity (Deanfield 2025 ^[5]) and baseline HbA1c (Lingvay 2024 ^[7]). Heart-failure benefit held across HFpEF and HFrEF (Deanfield 2024 ^[2]). The pattern — broad, subtype-agnostic benefit that doesn't hinge on any one intermediate marker — is exactly what makes SELECT's secondary data so influential for guidelines and labeling.

What the secondary endpoints add up to

Taken together, SELECT's secondary analyses turn a single cardiovascular headline into a multi-organ story. Semaglutide reduced serious kidney events by 22% and protected eGFR (Colhoun 2024 ^[3]); improved outcomes in people who already had heart failure of either subtype (Deanfield 2024 ^[2]); produced durable, four-year, ~10% weight loss across sexes and body sizes (Ryan 2024 ^[4]); slashed progression to diabetes from 6.9% to 1.5% (Kahn 2024 ^[6]); and delivered cardiovascular benefit that was largely independent of how much weight a person lost (Deanfield 2025 ^[5]). None of this changes the limits of the trial — it was one drug, one dose, a specific high-cardiovascular-risk non-diabetic population, and an industry-sponsored study — but the secondary endpoints substantially broaden what we can say semaglutide does. For the trial design, primary MACE result, and the FDA approval that followed, see our SELECT trial overview.

This article is educational and is not medical advice. Every PMID and numeric claim above was verified against the live PubMed database (esummary/efetch) and the publisher or PMC fulltext on 2026-06-05 before publication. Discuss whether a GLP-1 is appropriate for you with your own clinician.