Pooled GLP-1 trials: kidney outcomes move more than the heart ones

Design

systematic review and meta-analysis of 20 randomised placebo-controlled trials, random-effects model

Population

83,004 adults with type 2 diabetes

Primary outcome

major adverse cardiovascular events; composite renal outcome

Effect

MACE RR 0.87 (95% CI 0.83-0.92); renal composite RR 0.80 (0.73-0.88); heart failure hospitalisation RR 0.93 (0.85-1.01), not significant

Twenty randomised placebo-controlled trials in type 2 diabetes, 83,004 patients in total, pooled with a random-effects model. Major adverse cardiovascular events fell by 13% in relative terms (RR 0.87, 95% CI 0.83-0.92), all-cause death by 11% (RR 0.89, 0.84-0.93), cardiovascular death by 12% (RR 0.88, 0.81-0.94), myocardial infarction by 13% (RR 0.87, 0.79-0.96) and stroke by 12% (RR 0.88, 0.81-0.96). The composite renal outcome fell furthest: RR 0.80 (0.73-0.88).

What did not reach significance is as informative. Heart failure hospitalisation (RR 0.93, 0.85-1.01) and coronary revascularisation (RR 0.87, 0.74-1.01) both sit with their upper confidence limit just past 1.0. That is a different picture from the SGLT2 inhibitors, where heart failure is the standout. If a patient's dominant problem is congestion rather than atherosclerosis, this pooled evidence does not make the GLP-1 receptor agonist the obvious first addition.

In practice this supports what guidelines already say, and sharpens the reason for saying it. The renal composite is the largest single effect here, so a falling eGFR or rising albuminuria is a reason to reach for this class, not a reason to hesitate. Cost remains the binding constraint in most Indian practice; where the choice is one agent, let the dominant organ risk decide.

• Record the indication you are treating — atherosclerotic risk, renal decline or weight — so the next reviewer can judge continuation
• Check eGFR and urine albumin-to-creatinine ratio before starting, and at least yearly after
• Where congestion dominates, an SGLT2 inhibitor has the stronger heart-failure evidence
• Counsel on gastrointestinal effects at initiation and at each dose step, not only at the first prescription
• Ask at every visit whether the patient is still taking it — real-world persistence in this class is poor

Why it matters

The kidney, not the heart, is where this class now shows its largest effect — and that changes which patient you start it in first.

Don't overread it

Pooling twenty trials of different agents and durations gives an average effect, not a guarantee that any one drug delivers all six outcomes.

The statistics, in plain English

A risk ratio of 0.87 means 13 fewer events for every 100 that would have occurred, not 13 fewer per 100 patients — the absolute benefit depends on baseline risk, so it is largest in those with established disease. The heart failure estimate (0.93, 0.85-1.01) crosses 1.0 at the top, meaning no effect remains compatible with the data; it is a direction, not a finding. The renal confidence interval (0.73-0.88) sits entirely below 1.0 and is the tightest of the set.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.