What this is
- This systematic review evaluates the efficacy and safety of glucagon-like peptide-1 receptor agonists (GLP-1 RAs) for obesity management in adults, both with and without type 2 diabetes (T2DM).
- It includes 22 randomized controlled trials (RCTs) with a total of 41,757 participants, focusing on weight loss outcomes and adverse effects.
- Key findings reveal that semaglutide and tirzepatide are particularly effective, while liraglutide shows modest efficacy.
Essence
- GLP-1 RAs, especially semaglutide and tirzepatide, are effective for weight management in adults with and without T2DM, with notable gastrointestinal side effects.
Key takeaways
- Tirzepatide 15 mg leads to a weight loss of 9.5 kg in adults with T2DM, with 72% of participants losing ≥ 5% of their baseline weight.
- Semaglutide 2.4 mg results in a weight loss of 14.9% in non-diabetic adults, demonstrating significant efficacy in this population.
- GLP-1 RAs are associated with higher rates of gastrointestinal side effects compared to placebo, including nausea (14%-28% vs. 5%-10%) and vomiting (6%-12% vs. 2%-4%).
Caveats
- Heterogeneity among trials limits the ability to pool data and compare outcomes directly, affecting generalizability.
- Variability in lifestyle interventions across studies may influence weight loss outcomes, complicating the interpretation of efficacy.
Simplified
1. Introduction
The treatment of obesity extends beyond weight reduction; it requires addressing the underlying pathophysiologic mechanisms exacerbated by obesity, which contribute to complications, comorbidities, and mortality. These complications may include type 2 diabetes (T2DM), dyslipidemia, arterial hypertension, metabolic dysfunction–associated steatotic liver disease (MASLD), and obstructive sleep apnea (OSA) [1]. Effective obesity management is a critical component in preventing T2DM and reducing the risk of cardiovascular disease or advanced chronic liver disease [2, 3]. Beyond weight loss, emerging data demonstrate metabolic benefits of obesity therapies, including reduction in hepatic steatosis, improvements in endothelial function, and favorable effects on heart failure [4, 5].
Efforts to reverse the obesity epidemic encompass interventions including lifestyle changes, pharmacotherapy, endoscopic interventions, and surgery. Comprehensive lifestyle modifications, including diet and exercise, are the cornerstone of all weight management interventions [6, 7]. However, the available evidence shows that the resulting weight loss of 1–6 kg is challenging to maintain [8 –10]. For such individuals, antiobesity medications may be used as an adjunctive pharmacotherapy.
Glucagon-like peptide-1 receptor agonists (GLP-1 RAs), recognized for their antidiabetic properties, have been investigated as antiobesity drugs. They increase insulin secretion, reduce appetite, delay gastric emptying, and modulate dopamine reward pathways to decrease cravings and food intake [9, 11, 12]. These medications bind to GLP-1 receptors in the central nervous system, pancreas, and intestines to regulate hunger and satiety [9, 12]. The US Food and Drug Administration (FDA) has approved daily liraglutide 3 mg, weekly semaglutide 2.4 mg, and recently, weekly tirzepatide for obesity management [13 –15]. Tirzepatide, a dual glucose-dependent insulinotropic polypeptide (GIP) and GLP-1 RA, enhances insulin secretion and sensitivity while acting on additional nutrient-stimulated hormone pathways [11, 15 –17]. The availability of high-dose GLP-1 RAs in both oral and injectable formulations offers promising individualized treatment options.
While clinical trials of varying sizes have been conducted to investigate the weight loss efficacy of these drugs, individual studies have varied in design and population and have reported inconsistent weight loss effects [10, 11, 17 –32]. Few head-to-head studies have compared the efficacy and safety of GLP-1 RAs. There is also substantial variability regarding weight loss in patients with diabetes compared to patients without diabetes. This systematic review aimed to provide an overview of the weight loss efficacy and adverse event profile of GLP-1 RAs in two populations: adults with overweight/obesity with and without T2DM.
2. Methods
2.1. Overview
We initially aimed to conduct a network meta-analysis to compare the efficacy and safety of GLP-1 RAs for obesity management, integrating direct and indirect evidence from multiple randomized controlled trials (RCTs). However, substantial heterogeneity across trials, including variations in study design, participant characteristics, and outcome measures, made this approach unfeasible. The lack of direct head-to-head RCT comparisons between GLP-1 RAs further hindered a robust network meta-analysis. As a result, we opted for a narrative systematic review to present findings from each study, organizing key trial features and population characteristics that may influence drug effectiveness and interpretation.
2.2. Data Sources and Searches
Embase, MEDLINE, and Cochrane were systematically searched from inception to April 1, 2025, for trials assessing GLP-1 RA efficacy and safety in weight management in adults with overweight/obesity, both with and without T2DM. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed [33]. The protocol for this study was registered on PROSPERO (CRD42023466527). Search terms included in our literature search are shown in eTable 1. We restricted the search to clinical trials published in the English language and involving human subjects. Risk of bias for individual studies was assessed using the Cochrane Risk of Bias 2 tool (Supporting Information, eFigure 2). To minimize the potential for publication bias, we conducted a comprehensive search of multiple databases and reviewed reference lists of included studies and relevant review articles.
2.3. Study Selection
Our inclusion criteria for clinical trials were as follows: (1) phase 3 or 4 RCTs comparing GLP-1 RAs to each other or placebo; (2) minimum treatment duration of 40 weeks to assess efficacy and safety; (3) participants aged 18 or older; (4) BMI ≥ 30 kg/m2 or ≥ 27 kg/m2 with obesity-related comorbidities (e.g., hypertension, dyslipidemia, OSA, and cardiovascular disease); and (5) primary or secondary outcome of change in body weight. We excluded open-label, crossover, phase 1 or 2 trials, and studies focused on specific populations like polycystic ovary syndrome.
2.4. Data Extraction and Synthesis
In this systematic review, absolute weight change and percentage total body weight loss (%TBWL) were the primary efficacy outcomes. Efficacy benchmarks were considered met if the mean weight loss difference between the treatment and placebo groups was at least 5%, or if the proportion of patients losing ≥ 5% of baseline body weight in the active group was at least 35% and approximately double that of the placebo group [34]. Secondary outcomes focused on safety, including gastrointestinal side effects (nausea, vomiting, diarrhea, and constipation), pancreatitis, and serious adverse events (as defined by the Medical Dictionary for Regulatory Activities). Side effect data were extracted and presented as risk differences (RDs) with 95% confidence intervals (CIs). RD was chosen to directly compare the proportion of adverse events between treatment and control groups, allowing for a clearer interpretation of the absolute risk increase or decrease associated with GLP-1 RA therapy. Although baseline characteristics were extracted, subgroup analyses based on these variables were not conducted.
2.5. Quality Assessment
Two reviewers (Jena Velji-Ibrahim and Dhruvil Radadiya) independently extracted data from the RCTs using a standardized form, recording study details, sample size, efficacy, safety outcomes, GLP-1 RA type, dose, and follow-up period. Baseline characteristics (age, sex, BMI, waist circumference, and A1c) were also noted. Conflicts were resolved by consultation with the third author (Kalpit Devani). To minimize bias and ensure accurate reporting, we consulted Qureshi et al. [35]. Two reviewers assessed the risk of bias using the Cochrane Risk of Bias Tool Version 2 [36], with any conflicts resolved by Kalpit Devani.
2.6. Role of Funding Source
The authors did not receive any funding or financial support for any part of the study. Therefore, our study and its design are free from any influence by the pharmaceutical companies that manufacture the drugs compared in this analysis.
2.7. Ethics Statement
This systematic review was confirmed by the Human Investigation Committee (IRB) of Prisma Health to be exempt from IRB approval because it utilized only publicly available, de-identified data from previously published RCTs and did not involve direct interaction with human subjects or access to private, identifiable information.
3. Results
The study selection process is shown in eFigure 1 and eTable 1. A total of 22 RCTs were included, with durations ranging from 40 to 160 weeks and sample sizes from 282 to 17,604 participants. All trials were double-blinded. 18 RCTs evaluated the efficacy of a GLP-1 RA compared with placebo: daily liraglutide 3 mg (5 RCTs) [21, 22, 27, 28, 37], daily liraglutide 1.8 mg (3 RCTs) [21, 22, 29], weekly efpeglenatide 2 mg (1 RCT) [38], weekly efpeglenatide 4 mg (1 RCT) [38], weekly efpeglenatide 6 mg (1 RCT) [38], weekly albiglutide 30 mg (1 RCT) [39], weekly albiglutide 50 mg (1 RCT) [39], daily semaglutide 25 mg (1 RCT) [18], daily semaglutide 50 mg (2 RCTs) [18, 32], weekly semaglutide 2.4 mg (5 RCTs) [10, 24, 26, 40, 41], weekly tirzepatide 5 mg (2 RCTs) [17, 20], and weekly tirzepatide 10 and 15 mg (3 RCTs) [17, 20, 30]. Other comparisons included multidose weekly dulaglutide [42], weekly semaglutide 2.4 mg versus daily liraglutide 3 mg [19], weekly semaglutide 2.4 mg versus weekly semaglutide 1 mg [24], multidose weekly tirzepatide with weekly semaglutide 1 mg [25], and multidose weekly tirzepatide with weekly dulaglutide [31].
All but one of the included studies were Phase 3 RCTs, with a Phase 4 RCT by Pi-Sunyer and colleagues [27]. Most studies used subcutaneous injections, except for two that utilized oral semaglutide [16, 32]. A total of 41,757 participants (with or without T2DM) with obesity (BMI ≥ 30 kg/m2 or ≥ 27 kg/m2 with obesity-related comorbidities) were included. Participant ages ranged from 45 to 58 years, with 38.7% male participants. Baseline characteristics, including body weight, BMI, and HbA1c, are summarized in Table 1.
3.1. Weight Loss in Participants With Diabetes
Figure 1 illustrates the absolute weight change and %TBWL with each GLP-1 RA compared to placebo in participants with diabetes. Significant weight loss was observed with all GLP-1 RAs except efpeglenatide and albiglutide. Tirzepatide 15 mg was associated with the most weight loss, followed by tirzepatide 10 mg. Weight loss with tirzepatide 5 mg and semaglutide 2.4 mg was similar. Efficacy benchmarks were met if the difference in mean weight loss between the treatment and placebo groups was at least 5%, or if at least 35% of patients lost ≥ 5% of baseline body weight, approximately double that of the placebo group [34].
3.1.1. Albiglutide
A 52-week double-blind trial found no significant weight loss difference between albiglutide (30 mg and 50 mg) and placebo [39]. Neither efficacy benchmark was met. Participants had a lower BMI, with 33 kg/m2 in the placebo group and 34 kg/m2 in the albiglutide groups, compared to other studies. Males comprised 57% in the placebo and albiglutide 30 mg group and 51% in the albiglutide 50 mg group. Participants received diet and exercise counseling at each study visit.
3.1.2. Efpeglenatide
Efpeglenatide was studied in a 56-week trial that did not include lifestyle modifications. Here, the baseline BMI of participants was 34 kg/m2. Efpeglenatide was associated with significant weight loss compared to placebo, but the effect slightly decreased from weeks 30–56 [38]. Neither benchmark was met in this study.
3.1.3. Dulaglutide
One study demonstrated that higher doses of dulaglutide (3.0 mg and 4.5 mg) resulted in substantially greater weight loss than the 1.5 mg dose. However, both benchmarks were not met. Lifestyle modifications were not part of the trial protocol. This trial had a near-equal percentage of male and female participants. Participants had an average BMI of 34 kg/m2. They were permitted to continue metformin use.
3.1.4. Liraglutide
The satiety and clinical adiposity-liraglutide evidence (SCALE) Diabetes trial, a 56-week phase 3a study, compared various doses of liraglutide to placebo. The absolute weight changes were 6.4 kg with liraglutide 3 mg, 5.0 kg with liraglutide 1.8 mg, and 2.2 kg with placebo [22]. The mean weight loss difference (< 5%) did not meet the efficacy benchmark, but > 35% of subjects in the liraglutide group lost ≥ 5% of baseline body weight. Lifestyle interventions included dietary counseling and 150 min of weekly physical activity. This study only had 21% and 22% males in the GLP-1 RA group and placebo group, respectively. Insulin was used as rescue medication.
3.1.5. Semaglutide
The semaglutide treatment effect in people with obesity (STEP) 2 trial, a double-blind, double-dummy, phase 3 superiority study, showed that both efficacy benchmarks were met with semaglutide 2.4 mg [24]. A significant proportion (68.6%) of patients lost ≥ 5% of baseline body weight. Lifestyle interventions included dietary counseling and 150 min of weekly physical activity. Insulin use as rescue medication was not allowed.
Only one study compared high doses of oral semaglutide in diabetics, comparing semaglutide 25 mg and 50 mg to the 14 mg dose [18]. Over 35% of participants lost ≥ 5% of weight with high doses. Lifestyle modifications were not included, and participants could continue other oral T2DM medications.
3.1.6. Tirzepatide
The study of tirzepatide in participants with T2DM (SURPASS) 1 trial, the shortest at 40 weeks, compared tirzepatide to placebo and reported weight losses of 7.0 kg, 7.8 kg, and 9.5 kg with tirzepatide 5 mg, 10 mg, and 15 mg, respectively. This was the only study where both weight loss benchmarks were met, although participants had a lower baseline BMI of 32 kg/m2 compared to 35–37 kg/m2 in other studies [30]. There were a similar number of male and female participants. In a 72-week double-blind, randomized, placebo-controlled study, both benchmarks were met with tirzepatide 10 mg and 15 mg [17]. This was the only study that allowed participants to continue using insulin and oral T2DM medications, improving the generalizability to patients who start a GLP-1 RA while already on insulin. Lifestyle interventions were similar to the SCALE and STEP 2 trials, including dietary counseling and 150 min of physical activity per week.
3.1.7. Comparison of Different Medications
Three studies directly compared different GLP-1 RAs. A double-blind, double-dummy study compared oral semaglutide 14 mg daily to liraglutide 1.8 mg daily and placebo. Although the weight loss difference between GLP-1 RAs and placebo was < 5%, 44.7% of participants taking oral semaglutide lost ≥ 5% of baseline body weight [29]. Adherence was a concern due to different administration routes. Participants were permitted to take metformin or an SGLT2 inhibitor. An open-label, 40-week phase 3 trial comparing tirzepatide 5 mg, 10 mg, and 15 mg to oral semaglutide 1 mg showed that > 35% of participants lost ≥ 5% of baseline body weight in all groups. However, only tirzepatide 15 mg met both benchmarks, as considerable weight loss was achieved with semaglutide 1 mg. Blinding was challenging due to device differences and dose-escalation schemes [25]. Participants were allowed to take metformin. The SURPASS J-mono trial, a double-blind, multicenter, randomized phase 3 study in Japanese patients, compared tirzepatide to dulaglutide, showing that both benchmarks were met with tirzepatide, though the homogeneous population may limit generalizability [31]. This study included participants with the lowest BMI of all studies, being in the overweight, not obese range.
3.2. Weight Loss in Participants Without Diabetes
Figure 2 illustrates absolute weight change and %TBWL with each GLP-1 RA compared to placebo in a population without diabetes. All GLP-1 RAs were associated with significant weight loss compared to placebo. Tirzepatide 15 mg was associated with the most weight loss, followed closely by tirzepatide 10 mg. Similar weight loss was achieved with tirzepatide 5 mg, semaglutide 2.4 mg, and oral semaglutide 50 mg.
3.2.1. Liraglutide
Four trials compared liraglutide 3 mg to placebo, one of which included liraglutide 1.8 mg [21]. Two were phase 3a [27, 28] and two were phase 3b trials [21, 37]. In all studies, > 35% of participants lost ≥ 5% of baseline body weight; however, only two studies met both efficacy benchmarks with a 5% mean difference in weight loss between groups [21, 27]. Astrup and colleagues [21] and Le Roux and colleagues [28] provided long-term efficacy data (160 weeks). Both studies included lifestyle modifications, but Wadden and colleagues [37] integrated intensive behavioral therapy, which may limit generalizability. This study showed considerable placebo group weight loss, reducing the difference between groups. All three studies had predominantly female populations, with Wadden and colleagues including 16% and 17% males in the liraglutide and placebo groups, respectively [21, 28, 37]. Additionally, Le Roux and colleagues only included participants with prediabetes [28].
3.2.2. Semaglutide
One multicenter, double-blind study compared semaglutide 2.4 mg to placebo without lifestyle modifications in participants without T2DM. This trial, the largest in terms of participants, lasted 104 weeks and had mostly male participants (72%), who were older (average age 62 years), and had the lowest average participant BMI of 33 kg/m2 [41].
Three trials (STEP 1, 3, and 5) compared semaglutide 2.4 mg to placebo and included lifestyle modifications [10, 26, 40]. STEP 1 had the largest sample size [26], while STEP 3, with intensive behavioral therapy, demonstrated the smallest weight loss difference [10]. All three studies had a mostly female population. While STEP 1 and 3 were 68 weeks, STEP 5 was 104 weeks. These trials were pivotal for semaglutide market approval, as efficacy benchmarks were met with all trials.
One study compared oral semaglutide 50 mg to placebo over 68 weeks, aligning with STEP 1 and 3 durations [32]. This study included lifestyle modifications and had a mostly female population.
3.2.3. Tirzepatide
A single 72-week study comparing tirzepatide to placebo also met both benchmarks [20]. Participants in this study had a similar baseline BMI and percentage of male participants to the other trials in nondiabetics.
3.2.4. Comparison of Different Medications
The randomized, open-label, phase 3b STEP 8 study compared semaglutide 2.4 mg to liraglutide 3 mg directly in individuals without T2DM [19]. It found that semaglutide resulted in significantly greater weight loss compared to liraglutide in adults with overweight or obesity. The weight loss and adverse effect profiles of semaglutide and liraglutide were similar to those reported by Wilding and colleagues [26] and Pi-Sunyer and colleagues [27].
3.3. Side Effects
Figure 3 illustrates the RD of various side effects of GLP-1 RAs compared to placebo. In both the T2DM and non-T2DM populations, there was no greater risk of pancreatitis with GLP-1 RAs compared to placebo. In both diabetic and nondiabetic populations, all GLP-1 RAs except for albiglutide and efpeglenatide 2 mg were more likely than placebo to be associated with gastrointestinal adverse effects such as nausea, vomiting, diarrhea, and constipation. Tirzepatide showed comparable RDs across all doses. The RD of serious adverse events was comparable between GLP-1 RAs and placebo.
3.4. Heterogeneity of Trials
The inability to pool data in our analysis is due to the heterogeneity among the included studies. Differences in study design, population characteristics, GLP-1 RA doses, treatment duration, and the presence of lifestyle interventions contributed to this variation. For instance, some trials focused on individuals with T2DM, while others included nondiabetic participants, leading to differences in baseline characteristics like BMI and comorbidities. The studies also used various GLP-1 RAs at different doses, including daily and weekly formulations, complicating direct comparisons. Additionally, the inclusion of lifestyle interventions such as diet and exercise varied, influencing weight loss outcomes. The use of rescue medications, like insulin or oral T2DM drugs, further affected generalizability, preventing data pooling for a unified analysis.
4. Discussion
In this systematic review encompassing 22 RCTs with 41,757 adults, we aimed to provide an overview of the efficacy and safety of GLP-1 RAs for weight management in adults with and without T2DM. Our findings suggest that while most GLP-1 RAs offer substantial weight loss benefits, both efficacy benchmarks were met if the mean weight loss versus placebo was ≥ 5%, or if ≥ 35% of participants in the treatment group lost ≥ 5% of baseline body weight. Using these criteria, semaglutide 2.4 mg and tirzepatide 5 mg, 10 mg, and 15 mg met efficacy benchmarks in both populations, while semaglutide 50 mg met benchmarks in adults without T2DM. We found a comparable safety profile between GLP-1 RAs, although a higher incidence of gastrointestinal side effects was observed with liraglutide 3 mg.
Liraglutide 3 mg, while showing modest weight loss, has not consistently met efficacy benchmarks compared to newer GLP-1 RAs such as semaglutide and tirzepatide [24, 26, 40]. However, it has demonstrated benefits in reducing major cardiovascular events, reducing the risk of new-onset persistent macroalbuminuria, and slowing the decline in estimated glomerular filtration rate in patients with T2DM [43]. With its patent expiration and the availability of a generic form, liraglutide may become a more cost-accessible option despite its lower efficacy. This shift might increase its use despite its relatively lower efficacy, highlighting the ongoing need to balance cost, accessibility, and clinical outcomes in obesity management.
Beyond weight loss, GLP-1 RAs have beneficial systemic effects. They reduce the risk of major adverse cardiovascular events and preserve renal function, including reductions in kidney failure and slowing the decline in estimated glomerular filtration rate [4]. Emerging evidence also suggests that GLP-1 RAs reduce hepatic steatosis and liver inflammation, which is particularly relevant given the high prevalence of MASLD and MASH in individuals with obesity [5, 44].
Previous studies, and this systematic review, confirm that individuals with T2DM face greater challenges in losing weight compared to those without diabetes. Although semaglutide 2.4 mg and all doses of tirzepatide met weight loss benchmarks in both T2DM and non-T2DM populations, weight reduction was generally less pronounced in adults with T2DM. This may reflect lower baseline body weight, concomitant use of medications such as sulfonylureas and insulin that promote weight gain [2], reduced glycosuria, altered gut microbiome, and genetic predisposition to weight gain [45, 46]. Additionally, individuals with T2DM may be older and may have struggled with obesity for longer than those without T2DM.
Emerging therapies offer potential improvements in efficacy and broader treatment options, particularly for patients with diabetes, where obesity's pathophysiology is more complex. For example, orforglipron, a once-daily, oral, nonpeptide GLP-1 RA, may provide a competitive alternative to oral semaglutide as it does not require fasting [47]. Notably, retatrutide has demonstrated superior weight loss compared to tirzepatide in early phase trials [48]. By activating GLP-1, GIP, and glucagon receptors, it exerts effects across multiple metabolic pathways, regulating appetite, energy expenditure, and glucose homeostasis, which may provide enhanced efficacy, particularly in patients with obesity and T2DM. Retatrutide represents a promising advancement in pharmacologic therapy for obesity, offering potential for greater weight reduction than currently available GLP-1 RAs, although long-term safety and real-world effectiveness remain under investigation. Overall, new agents may improve efficacy while addressing a wider range of therapeutic needs and patient preferences, offering more tailored options for managing obesity and its comorbidities.
Limitations to the intertrial comparisons include the variability in lifestyle interventions, with some studies including intensive behavioral therapy while others did not incorporate formal lifestyle modifications [10, 37]. Real-world studies of adherence to these lifestyle interventions are needed, as these may not be representative of adherence in the general population. Given the variability in follow-up duration, shorter follow-up periods in certain trials may underestimate long-term efficacy and safety. Although formal assessment of publication bias was limited by the small number of studies for each GLP-1 RA and dose, we minimized potential bias by conducting a comprehensive search of multiple databases and reviewing reference lists of included studies and relevant reviews.
Demographics should also be considered when interpreting results [49]. For example, STEP 2, SURMOUNT-2, and SURPASS-1 included 51%, 51%, and 48% female participants, respectively. However, STEP 3, STEP 4, and SURMOUNT-1 included 81%, 79%, and 68% female participants, respectively. This discrepancy is important as females have been reported to respond more favorably to GLP-1 RAs, with some studies suggesting greater weight loss in females compared to males [50, 51]. Additionally, while STEP 1 and STEP 2 were designed to have a greater proportion of Asians than other trials [52], they still had a predominantly Caucasian population. One study included a 100% Asian population [31]. Future research should focus on underrepresented populations and gender-based differences to improve generalizability.
Considering safety, similar gastrointestinal side effects across GLP-1 RAs suggest these adverse events are inherent to the class and linked to their mechanisms of action. Previous RCTs and meta-analyses reported higher gastrointestinal side effects with subcutaneous semaglutide compared to liraglutide [11, 19]. Our systematic review, which assessed safety alongside efficacy, found similar gastrointestinal side effects across GLP-1 RAs, with liraglutide 3 mg showing the highest incidence. Notably, GLP-1 RAs had a similar incidence of pancreatitis and serious adverse events compared to placebo, consistent with previous meta-analyses [11, 16].
In conclusion, this systematic review highlights the efficacy and safety of GLP-1 RAs for weight management, and our findings support the use of GLP-1 RAs, particularly semaglutide 2.4 mg and all doses of tirzepatide, as highly effective strategies for weight management in adults with and without diabetes. Semaglutide 50 mg was also effective in the nondiabetic population. As liraglutide is off-patent, its reduced cost could increase its accessibility. Future research should prioritize head-to-head comparisons, long-term cardiovascular and hepatic outcomes, and personalized approaches considering gender and ethnicity.