Study Compares GLP-1 Drugs on Their Benefits, Harms

Variable weight-loss and cardiovascular benefits were seen with GLP-1 drugs in a systematic review and network meta-analysis, but effects on quality of life were less clear.

Across 262 randomized studies published through November 2025, moderate- to high-certainty evidence showed weight loss with the following GLP-1 drugs at 1 year versus lifestyle modification alone:

• Tirzepatide (Zepbound): mean difference -14.9% (95% CI -16.0 to -13.9)
• Cagrilintide-semaglutide (CagriSema): mean difference -14.8% (95% CI -16.9 to -12.7)
• Oral semaglutide (Rybelsus): mean difference -10.9% (95% CI -12.7 to -9.1)
• Orforglipron (Foundayo): mean difference -9.9% (95% CI -12.4 to -7.5)
• Subcutaneous semaglutide (Wegovy): mean difference -9.8% (95% CI -10.6 to -9.1)

Subcutaneous semaglutide (risk ratio [RR] 0.43, 95% CI 0.21-0.84) and tirzepatide (RR 0.49, 95% CI 0.27-0.88) were also both associated with reduced heart failure risk, while subcutaneous semaglutide stood alone in being tied to a reduced risk of all-cause mortality (RR 0.81, 95% CI 0.72-0.93) and myocardial infarction (RR 0.72, 95% CI 0.61-0.85), reported Sheyu Li, MD, of West China Hospital and Sichuan University in Chengdu, China, and colleagues.

Less certain was the evidence that these drugs could reduce kidney failure or improve quality of life beyond a subjective threshold, they noted in The BMJ.

“Obesity drugs produce variable weight loss at 1 year, with larger benefits generally accompanied by greater harms and discontinuation,” the authors wrote. “Most agents do not improve quality of life meaningfully and few show cardiovascular benefits. Decisions in clinical practice should consider trade-offs between benefits and harms within the context of shared decision making.”

However, on the U.K. Science Media Centre website, experts urged careful consideration of what the data actually show.

Hamid Merchant, PhD, MPharm, of the University of East London, disagreed with the message that most obesity drugs do not improve quality of life or heart health based on this meta-analysis.

“A more accurate interpretation would be that while obesity medicines vary in their benefits and harms, several produce substantial weight loss, some demonstrate important cardiovascular benefits, and the quality-of-life findings depend heavily on how clinically meaningful improvement is defined,” Merchant wrote.

“The findings do not show that obesity medications have no wider health benefits,” agreed Marie Spreckley, PhD, of the University of Cambridge in England. “Rather, they highlight that while the evidence for weight loss is strong, evidence for some longer-term outcomes is still developing and differs considerably between individual medications.”

In any case, the harms evident in the meta-analysis included gastrointestinal events with naltrexone-bupropion, oral semaglutide, orforglipron, and tirzepatide (RR 3.1 to 4.2). An increased risk of fatigue was also detected with naltrexone-bupropion (RR 8.9; absolute increase 331 per 1,000 people over 1 year), orforglipron (RR 3.4; 100 more per 1,000), and CagriSema (RR 3.2; 92 more per 1,000).

Notably, tirzepatide was associated with the largest reductions in fat mass (by 25.7%), as well as lean mass (by 8.3%).

As for next-generation GLP-1 medications (e.g., ecnoglutide, mazdutide, retatrutide), there was very low- to low-certainty evidence for their weight reduction effects to date.

“As the number of new drugs increases, physicians must now choose the right obesity treatment for each patient, individualizing the balance of benefits and adverse effects,” wrote Hamlet Gasoyan, PhD, MPH, and Michael Rothberg, MD, MPH, both of the Cleveland Clinic, in an accompanying editorial.

“Right now, physicians and patients can choose among lifestyle modification, metabolic and bariatric surgery, and rapidly expanding pharmacotherapy options,” they noted. “Physicians could use this up-to-date synthesis of evidence to guide their patients towards the treatments that are most suitable for them based on their preferences for outcomes and adverse events.”

On the Science Media Centre website, Sonya Babu-Narayan, MBBS, clinical director at the British Heart Foundation, said that “the findings reinforce that GLP-1 medicines, prescribed by a doctor, are an important option for some people and that they work best alongside healthy eating, regular physical activity including strength exercises, and other healthy behaviors.”

For their systematic review and meta-analysis, Li and colleagues selected studies that had randomized participants to at least 12 weeks of an obesity therapy versus control (lifestyle modification, placebo, or another drug). Eligible enrollees were adults with overweight or obesity, with or without cardiovascular or metabolic complications.

Study interventions included tirzepatide, CagriSema, ecnoglutide, mazdutide, retatrutide, semaglutide (oral and subcutaneous), survodutide, orforglipron, phentermine-topiramate (Qsymia), liraglutide (Saxenda), beinaglutide, naltrexone-bupropion, loxenatide, orlistat, exenatide, SGLT-2 inhibitors, dulaglutide (Trulicity), and metformin.

The 262 trials enrolled 99,791 eligible patients. Median age was 49 years, 63.3% were women, and median body mass index was 34.7. Follow-up typically lasted a median 26 weeks.

Beyond the GLP-1 medications, the appetite suppressant-anticonvulsant phentermine-topiramate was also tied to weight loss (mean difference -8.1%, 95% CI -9.7 to -6.5) in the study.

The quality-of-life analysis included 43 trials with 45,663 participants. For a clinically meaningful quality-of-life improvement, the investigators selected a threshold of 10 points on the Short Form 36 (SF-36), a self-reported patient survey that measures health-related quality of life, resulting in little to no effects detected among study drugs.

The SF-36 cutoff nevertheless remains “a judgement-based criterion,” noted Merchant. “Alternative thresholds have been proposed in the literature and may have led to different interpretations of the data. Therefore, these findings should not be interpreted as evidence that patients experience no meaningful day-to-day benefits from treatment.”

As for safety, the absolute risk increase for discontinuation due to any adverse event ranged from 19 (orlistat) to 94 (orforglipron) per 1,000 people over 1 year, Li’s group reported.

Gasoyan and Rothberg cautioned that the meta-analysis did not include individual-level data, leading to an inability to adjust for individual factors that modify the effectiveness of specific drugs.