Effect of glucagon‐like peptide 1 receptor agonists on systolic blood pressure in patients with obesity, with or without diabetes: A systematic review and network meta‐analysis

Summary

The effect of glucagon‐like peptide 1 receptor agonists (GLP‐1RAs) on systolic blood pressure (SBP) in patients with obesity with or without diabetes remains unclear. The aim was to compare the effect of different drug–dose combinations of GLP‐1RAs on SBP. The secondary aim was to assess whether changes in SBP with GLP‐1RAs are associated with weight change. MEDLINE and Cochrane were searched until January 2022 for randomized control trials (RCTs) on patients with obesity, evaluating the impact of semaglutide, liraglutide, efpeglenatide, or exenatide on SBP. Separate analyses were done for trials with and without diabetes. Multivariate meta‐regression assessed if SBP changes with GLP‐1RA varied based on weight change or follow‐up duration. Thirty‐five RCTs were included. Follow‐up duration ranged from 12 to 68 weeks for T2DM and 12–56 weeks for non‐T2DM patients. GLP‐1RAs significantly lowered SBP for all patients (MD = −3.14 [−3.60; −2.68]). Subgroup analysis showed a significantly greater difference in SBP reduction for patients without diabetes (−3.80 [−4.24; −3.37]) when compared with patients with diabetes (−2.13 [−3.27; −1.00]). Among patients with diabetes, liraglutide < 2 mg OD showed the greatest reduction in SBP (−3.78 [−6.27; −1.28]), while efpeglenatide ≤ 6 mg QW showed the greatest reduction in SBP (−6.00 [−9.89; −2.11]) in patients without diabetes. GLP‐1RAs result in mild reductions in SBP in patients with obesity. The change in SBP varies only slightly by the drug–dose combination and appears to be related to the amount of weight loss.

What is already known about this subject

• Glucagon‐like peptide‐1 receptor agonists (GLP‐1RAs) are currently recommended for the treatment of patients with and without diabetes to reduce body weight and hyperglycaemia.

What this study adds

• GLP‐1RAs mildly lower SBP in overweight or obese patients, especially in those without diabetes, with effects linked to weight loss.

• Efpeglenatide's structure allows less frequent dosing, improving compliance and quality of life in patients without diabetes.

• In patients without diabetes, SBP changes were associated with both weight loss and follow‐up time.

1. INTRODUCTION

Glucagon‐like peptide‐1 receptor agonists (GLP‐1RAs) were developed as therapeutic agents for the treatment of type 2 diabetes mellitus (T2DM) due to their antihyperglycemic effects. ¹ , ² However, aside from the benefits of glycaemic control, randomized controlled trials (RCTs) of GLP‐1sRAs in patients with T2DM demonstrated the ability of these agents to reduce body weight. This finding stimulated interest in the use of these agents as potential weight‐loss medications. For example, the STEP 1 trial found once‐weekly semaglutide to incrementally reduce body weight over 68 weeks by 14.9% compared with placebo. ³ Semaglutide and other GLP‐1RAs have subsequently gained regulatory approval as a treatment for weight loss.

Hypertension is a common comorbidity in patients with obesity, with primary hypertension arising in 65%–78% of patients who were obese. ⁴ Co‐existing hypertension and obesity are associated with an increased risk of renal disease, cardiac remodelling, and vasculopathies. ⁵ GLP‐1RAs address obesity by promoting significant weight loss through appetite suppression and improved metabolic regulation, which in turn helps alleviate obesity‐related mechanisms that elevate blood pressure. ⁶ Increased catecholamine concentration, sodium retention, pro‐inflammatory production, decreased nitric oxide, and malfunction in the renin‐angiotensin‐aldosterone system all may contribute to the pathogenesis of obesity‐related hypertension. ⁷ This dual effect positions GLP‐1RAs as a cornerstone therapy in the broader context of obesity management, emphasizing their potential to reduce hypertension and improve cardiovascular outcomes while tackling the root cause of these interlinked conditions. ⁸

Therefore, as GLP‐1RAs are increasingly being prescribed to patients with obesity, it is important to assess their effects on SBP and their correlation with weight loss. Hence, we conducted a systematic review and a network meta‐analysis (NMA) to compare the effect of different drug–dose combinations of GLP‐1RAs on SBP in patients with overweight or obesity, with or without diabetes. Furthermore, we sought to assess, using meta‐regression, whether the change in the length of treatment and weight loss was linked to the change in SBP seen with GLP‐1RAs.

2. MATERIALS AND METHODS

2.1. Data sources and search strategy

This meta‐analysis was conducted in accordance with Preferred Reporting Items for Systematic Review and Meta‐Analyses (PRISMA‐NMA) guidelines. ⁹ Two independent reviewers (AA and AAS) independently searched through the MEDLINE and Cochrane central databases from inception to February 2022. No time or language restrictions were set. The search strategy involved using MeSH terms to determine the different keywords for the GLP‐1RAs coupled with the Boolean operators AND/OR. A detailed search strategy for each database is provided in Table S1. Bibliographies of editorials and relevant reviews from major medical journals and conference proceedings for indexed abstracts were also screened to ensure the inclusion of all relevant trials.

2.2. Study selection

The predefined eligibility criteria for our study were: (a) randomized controlled trials (RCTs); (b) trials that evaluated the efficacy or safety of GLP‐1RAs (semaglutide, liraglutide, exenatide, and efpeglenatide) in patients with obesity with or without diabetes; (c) the comparator arm contained placebo; (d) reported changes in systolic blood pressure (SBP) (mmHg). Exclusion criteria included: (a) trials that evaluated the safety and efficacy of drugs other than GLP‐1RA (e.g., DPP4 inhibitors and SGLT2 inhibitors); (b) trials that enrolled women with obesity, with or without diabetes, who have concomitant polycystic ovary syndrome (PCOS); (c) trials evaluating other GLP‐1RA drugs that did not mention our primary outcome of interest. All articles retrieved from the systematic search were compiled in Endnote Reference Library (Version X8.1; Clarivate Analytics, Philadelphia, Pennsylvania) software, where the duplicates were identified and removed. The remaining articles were then thoroughly reviewed by two reviewers (AA and AAS) to ensure that they met our predefined eligibility criteria. In case of any discrepancy, a third reviewer (MSU) was consulted.

2.3. Data extraction

Two reviewers (AA and AAS) abstracted and verified data from the short‐listed articles on a standardized data collection. The original reference article was reviewed again upon any discrepancies encountered. The mean and standard deviations were extracted and used to calculate weighted mean differences (WMD) with 95% confidence intervals (CI) for the primary outcome. The primary endpoint used for comparison between patients with obesity with and without diabetes was SBP change (mmHg). If standard deviations (SDs) were not reported, we calculated them from standard errors or CI. ¹⁰ Other trial characteristics like the publishing year, sample size, follow‐up duration, drug type and dosage, and background therapy were also extracted (Table S2). In case of any discrepancy, a third reviewer (IS) was consulted. AA and AAS have full access to all the data in the study and take responsibility for its integrity and the data analysis.

2.4. Assessment of risk of bias and certainty in the evidence

The certainty of the evidence was evaluated using the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) approach (GRADEpro GDT) ¹¹ and was classified as high, moderate, low, or very low (Table S3). The five domains of GRADE for rating down certainty in the evidence included risk of bias, imprecision, inconsistency, indirectness, and publication bias. ¹² The risk of bias domain was assessed using the Cochrane Risk of Bias Tool for RCTs ¹³ (Table S4). To make imprecision judgements, we considered an arbitrary 5 mmHg change in blood pressure to be a minimally important difference, realizing that it could be higher. ¹⁴ Thus, a trial with a sample size >800 and CI that does not cross‐5 or + 5 mmHg is considered precise. ¹⁵ We also created an evidence map ¹⁶ that displays the certainty of the evidence and highlights potential areas for future research agendas (Table 1).

TABLE 1.

Evidence map of availability and appraisal of certainty of evidence for systolic blood pressure (SBP) in patients with obesity, with or without diabetes, receiving Glucagon‐like Peptide‐1 Receptor Agonist (GLP‐1RA).

Interventions	Systolic blood pressure (mmHg)
Obese or overweight patients with diabetes
Exenatide 10 μg twice daily
Liraglutide < 2 mg once daily*
Liraglutide > 2 mg once daily*
Semaglutide (3 mg) once daily
Semaglutide (7 mg) once daily
Semaglutide (14 mg) once dally*
Semaglutide < 1 mg once weekly
Semaglutide ≥ 1 mg once weekly*
Obese or overweight patients without diabetes
Efpeglenatide ≤ 6 mg once weekly*
Efpeglenatide ≥ 6 mg twice weekly
Exenatide 10 mg twice daily
Liraglutide < 2 mg once daily*
Liraglutide > 2 mg once daily*
Semaglutide < 1 mg once weekly*
Semaglutide ≥ 1 mg once weekly*

Note: Certainty of Evidence: , Very low; , Low; Moderate; High.

^*Statistically significant change in SBP from placebo.

Two investigators (AA and AAS) performed these assessments with consultation from a third reviewer (MSU) to reconcile differences.

2.5. Statistical analysis

RevMan (version 5.3; Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014) was used for pairwise data analysis statistical analyses. For continuous data, outcomes of interest were presented as WMDs with 95% CIs and were pooled using an inverse variance random‐effects model. A subgroup analysis was conducted, grouping different dosages of drugs. Forest plots were created to visually assess the pooled analysis. Heterogeneity across trials was evaluated using Higgins I ² , where a value <50% was considered acceptable, and a value of ≥75% was considered severe. A p‐value of <.05 was considered significant in all cases.

A frequentist NMA using the random‐effects model was conducted on all treatment comparisons to generate direct or indirect evidence. For the outcome of interest, WMD and 95% CI were pooled, and treatment ranks were generated based on the efficacy of the interventions. The consistency between direct and indirect sources of evidence was examined by the node‐splitting method. All analyses for NMA were performed using the R project for Statistical Computing (v 4.0.1) “netmeta” package (v 1.2–1). A p‐value <.05 was considered significant.

A multivariable meta‐regression was performed to evaluate if the degree of weight loss and length of follow‐up affects SBP in patients with and without diabetes. Comprehensive meta‐analysis software (v 3.3.070) was used to estimate the correlation coefficient (r) and 95% CI. Means and SDs were pooled using a random‐effects model. The proportion of total between‐study variance explained by the model was identified using an analogous index (R ²). A value of R ² < .3 was considered a very weak effect size; .3 < R ² < .5 was considered a weak effect size; .5 < R ² < .7 was considered a moderate effect size; R ² > .7 was considered a strong effect size. ¹⁷

3. RESULTS

3.1. Search results

The initial search revealed 2548 citations. After removing duplicates and screening the titles and abstracts, we included 35 RCTs in the meta‐analysis [21 RCTs containing 10 997 patients with diabetes ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ ; 14 RCTs containing 11 699 patients without diabetes ³ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ , ⁴⁹ , ⁵⁰ , ⁵¹ ]. The PRISMA flow chart Figure 1 summarizes the results of our literature search. The mean age, weight, and SBP were as follows: [with diabetes: 56 years; without diabetes: 43 years]; [with diabetes: 92.3 kg; without diabetes: 104.3 kg]; [with diabetes: 131.1 mmHg; without diabetes 124.1 mmHg)], respectively. Table S2 outlines the baseline characteristics of all included trials. The interventions evaluated in the meta‐analysis include semaglutide [<1 mg, ≥ 1 mg once weekly (QW); 3, 7, 14 mg once daily (QD)], liraglutide [<2 mg; >2 mg QD], exenatide [10 μg twice daily (B.I.D.)], and efpeglenatide [≥6 mg QW; ≤6 mg once biweekly (Q2W)]. Additionally, the minimum follow‐up treatment duration in trials ranged from 12 to 68 weeks for patients with diabetes and from 12 to 56 weeks for patients without diabetes.

FIGURE 1.

PRISMA flow chart outlining literature search.

3.2. Initial findings

When data from all the studies included in the meta‐analysis was pooled, we found that GLP‐1RAs significantly lowered SBP for all patients (MD = −3.14 [−3.60; −2.68]; I ² = 86.2%; p < .01). Severe heterogeneity was observed, due to which a subgroup analysis, based on the patient's diabetes status, was conducted. The analysis showed a significantly greater difference (p < .01) in SBP reduction for patients without diabetes (−3.80 [−4.24; −3.37]; I ² = 83%; p < .01) when compared with patients with diabetes (−2.13 [−3.27; −1.00]; I ² = 86%; p < .01). However, severe heterogeneity could still be observed in both subgroups [Figure 2]. Subgroup analysis of the drugs and their respective doses for patients with obesity with or without diabetes can be found in the Figures S1.1 and S1.2.

FIGURE 2.

Summarized Forest plot displaying Systolic blood pressure (SBP) changes (mmHg) in patients with obesity with and without diabetes receiving Glucagon‐like Peptide‐1 Receptor Agonist (GLP‐1RA) vs placebo^a. (a)—Mean differences (MD) and 95% CIs pooled together to showcase the overall SBP change among all patients (represented by the bottom‐most black diamond), as well as the SBP change in each of the two subgroups (represented by the first and second black diamonds). The presence of the diamond on any particular side of the 0 line in the forest plot indicates that the respective intervention causes a greater reduction in SBP. Non‐intersection of the diamond with the 0 line indicates a significant difference.

3.3. Network meta‐analysis (NMA) results for patients with diabetes

Figure 3A shows the network plots of trials included in this analysis. All GLP‐1RAs significantly reduced SBP except for semaglutide <1 mg QW, semaglutide oral 3 mg QD, and semaglutide oral 7 mg QD, when compared with placebo. Ranking the interventions revealed that the greatest effect was by liraglutide <2 mg QD (MD = −3.78 [−6.27; −1.28]) followed by exenatide 10 μg B.I.D (MD = −3.39 [−5.63; −1.15]), and semaglutide ≥1 mg QW (MD = −3.28 [−5.68; −0.88]) (Table S5(A), Figure S2.1).

FIGURE 3.

Network plot of systolic blood pressure (SBP) change (mmHg) in patients with obesity (A) with diabetes (B) without diabetes receiving glucagon‐like Peptide‐1 Receptor Agonist (GLP‐1RA) vs placebo^a. (a)—Each circular node represents an intervention. The links (edges) represent the relationships between each intervention. The weight (thickness) of the edges corresponds to the number of trials between the interventions.

3.4. Network meta‐analysis (NMA) results for patients without diabetes

Figure 3B shows the network plots of trials included in this analysis. Compared with the placebo, all drugs, except exenatide 10 μg B.I.D and efpeglenatide ≥6 mg Q2W, showed a mean significant decrease in SBP. Ranking the interventions revealed that patients who received efpeglenatide ≤6 mg QW showed the greatest significant reduction in SBP (MD = −6.00 [−9.89; −2.11 mmHg]), followed by semaglutide ≥1 mg QW (MD = −4.54 [−5.69; −3.39 mmHg]), and semaglutide <1 mg QW (MD = −4.51 [−5.49; −3.54 mmHg]) (Table S5(B), Figure S2.2).

3.5. Meta‐regression results for GLP‐1RA with mean weight change and follow‐up

A multivariate meta‐regression random‐effects model quantified the study‐level impact of follow‐up periods and mean weight loss on SBP for patients with and without diabetes treated with semaglutide and liraglutide. Only weight loss (0.0752 [0.0254, 0.1250]; I ² = 88.84%, p < .05) was found to have a significant impact on SBP in patients with diabetes. In patients without diabetes, both mean weight loss (0.1239 [0.0317, 0.2162]; I ² = 98.70%, p < .05) and follow‐up (−0.0245 (−0.0433, −0.0056); I ² = 98.71%, p < .05) were found to have a significant effect on SBP (Table 2).

TABLE 2.

Multivariate meta‐regression for primary outcomes adjusted for mean weight change and follow‐up covariates in patients with obesity, with or without diabetes receiving semaglutide and liraglutide.

Outcome: changes in systolic blood pressure (mmHg)
	Correlation coefficient (95% CI); p	Analogous index (R 2)	Heterogeneity I 2 (%); Cochran's Q model; Tau2
Covariates
Patients with T2DM
Follow‐up (weeks)	0.0034 (−0.0044, 0.0112); p = .39	.00	I 2 = 88.99%, Q = 345.20, Tau2 = 0.1174
Mean weight loss	0.0752 (0.0254, 0.1250); p  < .05	.00	I 2 = 88.84%, Q = 340.56, Tau2 = 0.1135
Patients without T2DM
Follow‐up (weeks)	−0.0245 (−0.0433, −0.0056); p  < .05	.00	I 2 = 98.71%, Q = 1477.00, Tau2 = 0.7302
Mean weight loss	0.1239 (0.0317, 0.2162); p  < .05	.00	I 2 = 98.70%, Q = 1463.60, Tau2 = 0.7935

Note: The bold values highlight statistically significant results (p < 0.05), drawing the reader's attention to key findings and improving visual clarity.

3.6. Certainty of evidence and evidence map

Table S3 displays a summary of the findings and assessment of the certainty of evidence. We found that the majority of certainty of evidence collated for patients with obesity and diabetes for semaglutide was moderate in quality, while the quality of evidence for exenatide was low. In contrast, the majority of certainty of evidence for patients with obesity without diabetes for liraglutide was moderate in quality. However, the quality of evidence for exenatide and efpeglantide was low.

Table 1 is an evidence map summarizing the effect of the included interventions on SBP. Most interventions had a beneficial effect on this outcome. The map also shows the lack of significant effect for a few interventions, mainly in individuals with diabetes. The certainty of evidence across interventions varied from very low to high.

3.7. Risk of bias

Sixteen included trials were considered to have a low risk of bias for all the domains. Nine trials were considered to have a high risk of other bias, while two studies had a high risk of bias in the blinding of participants, personnel, and outcome assessment. Six trials were assigned an unclear risk of bias for allocation concealment, while five trials showed evidence of an unclear risk of bias in selective reporting. Detailed results regarding the assessment of the bias for each selected trial can be found in Table S4.

4. DISCUSSION

This NMA, which included >20 000 patients who were overweight or obesity, revealed several key findings. GLP‐1RAs result in mild reductions in SBP in patients with obesity. The extent of this reduction in SBP varies only slightly with the specific drug–dose combination used and appears to be significantly associated with weight loss achieved in patients with or without diabetes. The meta‐regression results revealed that there were no significant associations between follow‐up time and SBP changes in patients with diabetes, indicating that the changes in SBP were independent. However, in the patients without diabetes, both weight loss and follow‐up time were significantly associated with SBP changes.

GLP‐1 agonists can vary in structure and pharmacokinetics and hence may differ in efficacy. For example, liraglutide is similar to human GLP‐1, which contains a fatty acid component to increase in vivo stability, whereas exenatide shares 53% amino acid sequence identity with human GLP‐1RA. ⁵² The half‐lives of GLP‐1RAs range from 2.4 h (exenatide) to 13 h (liraglutide) ⁵² to 165 h (semaglutide) ⁵³ to 5.6 to 7.5 days (efpeglentatide). ⁴³ Despite these differences, most drug–dose combinations of GLP‐1RAs resulted in a similar degree of SBP reduction. The only noteworthy exception was efpeglenatide ≤6 mg QW, which showed a nominally greater reduction in SBP (MD: −6.00 kg) in patients without T2DM when compared with semaglutide ≥1 mg QW (MD: −4.54 kg) and semaglutide <1 mg QW (MD: −4.51 kg). Efpeglenatide, a novel investigational GLP‐1RA, is chemically conjugated to recombinant human immunoglobulin G4 Fc fragment through a nonpeptidyl linker to CA‐Exendin‐4. Efpeglenatide has merited attention as a promising drug in patients without diabetes, with its current biweekly dosage and potential once monthly dosing, which would be advantageous over the currently available B.I.D., QD, and QW GLP‐1RAs on the market. Further studies are warranted to explore this new drug. ⁵¹

Our meta‐regression results indicated that SBP change was significantly associated with weight loss in patients with and without diabetes. ⁵⁴ GLP‐1RAs promote significant weight loss through appetite suppression and improved metabolic regulation, which helps mitigate obesity‐related mechanisms that elevate blood pressure. However, the independent effects of GLP‐1RAs on systolic blood pressure (SBP) reduction remain to be fully explored. GLP‐1RAs may also have direct cardiovascular protective effects, such as reducing inflammatory markers and improving lipid profiles, which can help mitigate factors contributing to high BP. ⁸ Additionally, GLP‐1RAs have been associated with beneficial renal effects, including reducing albuminuria and improving kidney function. This can be beneficial in patients with hypertension, as renal health is closely linked to BP regulation. ⁸ Although weight loss remains the primary factor driving reductions in BP, the mechanisms outlined above may contribute to BP reduction independent of weight loss. These potential pathways warrant further investigation to determine whether GLP‐1RA could be effectively utilized for blood pressure management in patients without obesity. Future studies should delve deeper into these mechanisms to clarify the broader therapeutic potential of GLP‐1RAs in hypertension treatment.

4.1. Limitations

This meta‐analysis has a few limitations that should be considered while interpreting the results. First, some drugs/dose combinations may have limited power compared with others. Second, although subgroup analyses and meta‐regression were performed, high heterogeneity existed in the patients without diabetes subsets. This can be explained by differences in baseline patient characteristics such as differences in the average ages, gender ratios, sample sizes, treatment duration, and proportion of patients on other antihypertensive medications. Third, many of the trials did not use BP as a primary endpoint. Lastly, our results are limited to the comparison of GLP‐1RA with placebo only; no comparisons were made with active interventions. Additionally, the potential for publication bias should be considered, as the current evidence relies heavily on published studies, which may not fully capture all aspects of GLP‐1RAs' effects on blood pressure.

5. CONCLUSION

GLP‐1RAs result in mild reductions in SBP in patients with overweight or obesity. The change in SBP varies only slightly by the drug–dose combination and appears to be related to the amount of weight loss. Notably, the effect was more pronounced in patients without diabetes, who experienced a greater SBP reduction than those with diabetes. Among patients with diabetes, liraglutide <2 mg once daily showed the greatest reduction in SBP, whereas in individuals without diabetes, efpeglenatide ≤6 mg once weekly showed the greatest reduction in SBP. These insights pave the way for tailored therapeutic approaches and highlight the need for personalized treatment strategies based on diabetes status. A better understanding of these findings is useful as a guide to continuing GLP‐1RA therapy based on patient characteristics, ensuring optimal outcomes for individuals with varying clinical profiles.

AUTHOR CONTRIBUTIONS

Abraish Ali and Asad Ali Siddiqui: Conceptualization, Methodology, Investigation, Writing—Original Draft, Data curation. Muhammad Shariq Usman and Izza Shahid: Supervision, Resources, Writing—Reviewing and Editing. Muhammad Shahzeb Khan: Writing—Reviewing and Editing, Project administration. Prinka Perswani: Writing—Reviewing and Editing. All authors were involved in writing the paper and had final approval of the submitted and published versions.

CONFLICT OF INTEREST STATEMENT

No conflict of interest was declared.

Supporting information

Data S1

Ali A, Siddiqui AA, Usman MS, Shahid I, Khan MS, Perswani P. Effect of glucagon‐like peptide 1 receptor agonists on systolic blood pressure in patients with obesity, with or without diabetes: A systematic review and network meta‐analysis. Clinical Obesity. 2025;15(4):e70012. doi: 10.1111/cob.70012

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are openly available in MEDLINE at https://pubmed.ncbi.nlm.nih.gov/.