Safety and efficacy of glucagon-like peptide-1 receptor agonists in individuals with type 2 diabetes mellitus fasting during Ramadan: a systematic review and meta-analysis

Abstract

BACKGROUND

Data on the use of glucagon-like peptide-1 receptor agonists (GLP-1RAs) in individuals with type 2 diabetes mellitus (T2DM) during Ramadan fasting is limited. No meta-analysis has summarized the safety and effectiveness of GLP-1RAs in these situations.

AIM

To evaluate the safety and efficacy of GLP-1RA in patients with T2DM fasting during Ramadan.

METHODS

Electronic databases were systematically searched for relevant studies that featured GLP-1RA in the intervention arm and other glucose-lowering medications in the control arm. The primary outcome was adverse events (AEs) during Ramadan for both groups; other outcomes included changes in glycemic and anthropometric measures during the peri-Ramadan period.

RESULTS

Four studies [three randomized-controlled trials with low risk of bias (RoB) and one prospective observational study with serious RoB] involving 754 subjects were analyzed. GLP-1RA group achieved greater glycated hemoglobin reduction than the non-GLP-1RA group [mean difference (MD): -0.31%, 95%CI: -0.61 to -0.01, P = 0.04, I² = 77%] with a lower risk of documented symptomatic hypoglycemia (risk ratio = 0.38, 95%CI: 0.16 to 0.88, P = 0.02). Any AEs, serious AEs, or AEs that led to treatment discontinuation were comparable between the two groups. The GLP-1RA group experienced greater weight loss compared to the non-GLP-1RA group (MD: -2.0 kg, 95%CI: -3.37 to -0.63, P = 0.004, I² = 95%). There were comparable changes in blood pressure and lipid profile between the two groups. GLP-1RA users experienced higher risks of gastrointestinal AEs, nausea, and vomiting; however, the risks of heartburn, abdominal pain, and diarrhea were similar in both groups.

CONCLUSION

Limited evidence suggests that GLP-1RAs are safe for T2DM management during Ramadan, offering modest benefits in blood sugar control and weight loss. Large multicenter trials are needed to confirm their safety and efficacy in at-risk populations, improving clinical practice decision-making.

Key Words: Type 2 diabetes mellitus; Glucagon-like peptide-1 receptor agonist; Ramadan; Religious fasting; Safety; Meta-analysis

Core Tip: No systematic review and meta-analysis (SR/MA) have previously assessed the safety and efficacy of glucagon-like peptide-1 receptor agonists (GLP-1RAs) in individuals with type 2 diabetes mellitus who fast during Ramadan. This SR/MA of 4 studies with 754 participants identified that GLP-1RAs are safe with only exaggerated gastrointestinal side effects, mainly nausea and vomiting. GLP-1RA also appears to be effective in improving glycated hemoglobin and aiding in weight loss, compared to other glucose-lowering medications. However, other cardiometabolic benefits, such as improvements in blood pressure and lipid profile, were not evident. Larger multinational, randomized trials with broader global participation are needed to establish more definitive clinical practice recommendations for using GLP-1RAs during Ramadan fasting.

INTRODUCTION

Fasting during Ramadan (the ninth month in the Islamic lunar calendar) is one of the five pillars of the Islamic religious faith. During Ramadan, all adult Muslims are required to abstain from consuming food and drink (including medication) from sunrise to sunset[1]. It is estimated that there are approximately 148 million Muslims who live with diabetes worldwide. The fasting duration varies by geographic location and can last up to 19 hours a day in some parts of the world, especially during the summer. The lifestyle changes in Ramadan, especially abstinence from food and drink for long fasting hours and using glucose-lowering drugs (GLDs) for diabetes treatment, may predispose these individuals to develop adverse events (AEs), including dehydration and hypoglycemia[2]. The Diabetes and Ramadan (DaR) 2020 survey, which involved 5865 participants, found that 83.6% of those surveyed fasted during Ramadan, 94.8% fasted for 15 days or more, and 61.9% completed the entire month[3]. Many of these individuals are at risk of developing complications related to fasting and often fast despite medical advice not to do so[4,5]. Consequently, this sizable Muslim population with diabetes that observes fasting during Ramadan needs special attention before and throughout Ramadan.

Glucagon-like peptide-1 receptor agonists (GLP-1RAs) are highly recommended for patients with type 2 diabetes mellitus (T2DM) by all guidelines, given their glycemic, weight, cardiorenal, hepatic, and metabolic benefits[6]. The use of GLP-1RAs continues to increase[7], and it is presumed that many Muslims with T2DM fasting during Ramadan will utilize these medications. The reduced risk of hypoglycemia and the additional benefit of weight loss or weight maintenance confer an extra advantage when continuing their use during Ramadan[6]. According to guidelines from the International Diabetes Federation and the DaR International Alliance, the use of GLP-1RAs confers no additional risk for fasting, a view that has also been reiterated in the American Diabetes Association's recommendations for fasting[6,8]. The guidelines also recommend against initiating GLP-1RAs just before Ramadan, given their tolerability issues, especially gastrointestinal (GI) ones, which may make fasting troublesome for the patient and potentially negate the fast[8].

However, data on GLP-1RA use by patients with T2DM during Ramadan fasting is limited. Due to the limited number of randomized controlled trials (RCTs) and observational studies with small sample sizes specifically addressing the adverse effects and efficacy of GLP-1RAs in individuals with T2DM who fast during Ramadan, we conducted this systematic review and meta-analysis (SR/MA). Our goal was to investigate whether individuals with T2DM using GLP-1RAs during the fasting month of Ramadan achieve glycemic and other metabolic benefits, with or without an increased risk of adverse effects/events.

MATERIALS AND METHODS

Study setting

This SR/MA was conducted following the procedures described in the Cochrane Handbook for Systematic Reviews of Interventions and is reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist[9,10]. The SR/MA has been registered with PROSPERO (CRD42024637706), and the protocol summary can be accessed online.

Ethical compliance

No separate ethical approval was needed for this MA, as such approvals already exist for the individual included studies.

Search strategy

A systematic search was conducted across various databases and registers, including MEDLINE (via PubMed), Embase, Web of Science, and ClinicalTrials.gov. This search spanned from the inception of each database to December 25, 2024. Using the Boolean operators “AND” and “OR” the following terms were searched: (1) “Glucagon-like peptide-1 receptor agonists”; (2) “GLP-1 receptor agonists”; (3) “GLP-1RA”; (4) “Exenatide”; (5) “Lixisenatide”; (6) “Liraglutide”; (7) “Albiglutide”; (8) “Dulaglutide”; (9) “Semaglutide”; (10) “Tirzepatide”; (11) “Ramadan fasting”; (12) “Religious fasting”; (13) “Type 2 diabetes mellitus”; (14) “Type 2 diabetes”; (15) “Type II diabetes”; (16) “T2DM”; and (17) “T2D”. The search terms were applied to titles and abstracts. The aim was to identify published and unpublished studies in English. Additionally, the search involved reviewing references within the published articles retrieved for this study and relevant journals.

Study selection

The Population, Intervention, Comparison, Outcomes, and Study design was used as a framework to formulate eligibility criteria for studies in this SR/MA. The patient population (P) consisted of adults of either sex with T2DM fasting during Ramadan; the intervention (I) was any GLP-1RA either as monotherapy or combined with other GLDs for managing T2DM; the comparison or control (C) group comprised individuals receiving GLDs other than a GLP-1RA either as monotherapy or in combination; the outcomes (O) included safety and/or effectiveness of study drugs during the study period; and RCTs or prospective observational/intervention studies were considered as the study type (S) for inclusion. The studies must have at least two arms: (1) One with any GLP-1RA alone or combined with other GLDs (GLP-1RA group); and (2) The other(s) with other GLDs either as monotherapy or combination therapy (non-GLP-1RA group). Exclusion criteria were case reports or case series, retrospective studies, studies conducted in type 1 or other types of diabetes, including patients with both type 1 and type 2 diabetes, studies conducted among pregnant or lactating women, studies without a non-GLP-1RA control arm, and studies not reporting the outcomes of interest. Four review authors independently identified eligible articles according to the above criteria. Discrepancies in opinion on the inclusion of studies were resolved by consensus.

Data extraction

Four review authors conducted data extraction independently by using standardized data extraction forms. The results were consolidated upon identifying multiple publications from a single study group, and relevant data from each article were incorporated into the analyses. The following data were extracted for all the eligible studies and included in the review: (1) First author; (2) Year of publication; (3) The country where the study was conducted; (4) Study design; (5) Major inclusion criteria of the study subjects; (6) GLDs used in the GLP-1RA and non-GLP-1RA groups; (7) sample size; (8) Number of male and female participants; (9) Mean age; (10) Diabetes duration; (11) Body weight; (12) Glycated hemoglobin (HbA1c); (13) Systolic and diastolic blood pressure (BP); (14) Lipid profile; and (15) AEs, including hypoglycemia. Any disagreements were resolved by consensus.

Dealing with missing data

The necessary supplementary files of the articles were obtained from the websites of the publishing journals. Further information, if needed, was collected from the corresponding authors of the relevant articles via email. All pertinent information collected in this manner was meticulously incorporated into the MA. Moreover, attrition rates, encompassing dropouts, losses to follow-up, and withdrawals, were meticulously examined.

Outcomes analyzed

The primary outcome of interest was the risk of any AEs, including hypoglycemia, in the GLP-1RA group compared to the non-GLP-1RA group. Additional outcomes included the mean difference (MD) of changes in glycated hemoglobin (HbA1c), body weight, systolic and diastolic BP, and lipid profile from baseline between the two groups.

Statistical analysis

The results of the outcomes were reported using risk ratios (RRs) for dichotomous variables and MDs for continuous variables, together with 95% confidence intervals (CIs). The Review Manager computer program, version 7.2.0, was used to generate forest plots, which portrayed the RR or MD for the outcomes; the left side of the forest plot favored the GLP-1RA group, and the right side favored the non-GLP-1RA group[11]. Random-effects analysis models were chosen to address the anticipated heterogeneity resulting from variations in population characteristics and study lengths. The inverse variance statistical method was applied for all instances. The SR/MA encompassed forest plots that integrated data from at least two trials. A significance level of P < 0.05 was used.

Assessment of the risk of bias in the included studies

Three authors independently conducted the risk of bias (RoB) assessment using the Cochrane RoB tool for randomized trials, Version 2 (RoB2), and the RoB in Non-randomized Studies of Interventions, Version 2 (ROBINS-I V2), for the RCTs and non-randomized intervention trials, respectively[12,13]. In discrepancies, the sixth and seventh authors acted as arbitrators to achieve consensus. The RoB VISualization (robvis) web app created RoB plots[14].

Assessment of heterogeneity

The assessment of heterogeneity was initially conducted by studying forest plots. Subsequently, a χ² test was performed using N-1 degrees of freedom and a significance level of 0.05 to determine the statistical significance. The I² test was also employed in the subsequent analysis[15]. The specifics of understanding I² values have been explained in depth elsewhere[16].

RESULTS

Search results

The PRISMA flow diagram of steps in selecting the studies are depicted in Figure 1. The initial search identified 103 articles, narrowed to six after screening titles, abstracts, and subsequent full-text reviews. Finally, four studies involving 754 subjects meeting all the prespecified criteria were included in this SR/MA[17-20]. Two studies were excluded- one did not have a non-GLP-1RA comparator group, and the other was a substudy of an included study[21,22].

Figure 1 Flowchart on study retrieval and inclusion in the meta-analysis. GLP-1RA: Glucagon-like peptide-1 receptor agonist.

Characteristics of included and excluded studies

The specifics of the included and excluded studies are shown in Table 1[17-20] and Supplementary Table 1[21,22], respectively. Three included studies were randomized, parallel-group, open-label, active-controlled trials[17,18,20]; the other was an open-label, single-center, two-arm parallel-group, prospective observational study[19]. All studies had two arms: (1) One with any GLP-1RA alone or combined with other GLDs (GLP-1RA group); and (2) The other with GLDs other than GLP-1RA either as monotherapy or combination therapy (non-GLP-1RA group). In the GLP-1RA arm, two studies used liraglutide[17,20], one used lixisenatide[18], and the other used semaglutide[19]. The follow-up duration of the studies ranged from 12-33 weeks.

Table 1 Baseline characteristics of individual studies and study participants included in the meta-analysis.

Ref.	Study design, trial reg. number (if any)	Major inclusion criteria	Groups	GLD used	n	Sex	Age (years), mean (SD)	Duration of diabetes mellitus (years), mean (SD)	Baseline body weight (kg), mean (SD)	Baseline HbA1c (%), mean (SD)	Study duration
Azar et al[17], 2016, LIRA-Ramadan, 39 sites in 7 countries	Randomized, parallel-group, open-label, active-controlled trial, NCT01917656	Age 18-80 years, HbA1c 7%-10%, body mass index ≥ 20 kg/m2, on stable diabetes treatment (MFN ≥ 1 gm + SU MTD)	GLP-1RA	Liraglutide 18 mg + MFN	171	M: 85, F: 86	54.9 (9.27)	8.0 (5.26)	81.0 (17.1)	8.3 (0.94)	33 weeks
			Non- GLP-1RA	SU MTD + MFN	170	M: 83, F: 87	54.0 (9.33)	7.2 (4.39)	83.1 (16.0)	8.2 (0.91)
Hassanein et al[18], 2019, LixiRam, 16 sites in 5 countries	Phase 4, randomized, open-label, parallel-group, clinical trial, NCT02941367	Uncontrolled diabetes on SU + basal insulin ± one OAD	GLP-1RA	Lixisenatide 20 μgm + basal insulin + MFN	92	M: 40, F: 52	52.6 (9.5)	7.1 (4.8)	76.0 (14.6)	8.7 (0.7)	12-20 weeks
			Non-GLP-1RA	SU + basal insulin + MFN	92	M: 43, F: 49	54.1 (10.6)	7.4 (5.4)	75.0 (11.6)	8.5 (0.7)
Pathan et al[19], 2024, Pathan 2024, Bangladesh Institute of Research and Rehabilitation in Diabetes, Endocrine and Metabolic Disorders, Bangladesh	Open-label, single-center, two-arm parallel-group, prospective observational study	Age > 18 years, on stable GLDs for the last 3 months before enrolment	GLP-1RA	Semaglutide 0.5 mg ± OADs/insulin	55	M: 19, F: 36	49.05 (9.8)	7.34 (5.8)	N/A	7.43 (1.43)	24 weeks
			Non-GLP-1RA	OADs ± insulin, no semaglutide	75	M: 26, F: 49	53.04 (10.9)	12.37 (6.9)	N/A	8.10 (1.56)
Brady et al[20], 2014, Treat 4 Ramadan, 2 sites in United Kingdom	Randomized, parallel-group, open-label, active-controlled trial	Age ≥ 18 years, on stable dose of MFN or MFN + SU/Pioglitazone, HbA1c 6.5%-12%	GLP-1RA	Liraglutide 12 mg + MFN	52	M: 26, F: 26	52.2 (10.7)	N/A	79.0 (11.2)	7.8 (1.0)	16 weeks
			Non-GLP-1RA	MFN + SU	47	M: 24, F: 23	51.5 (11.1)	N/A	86.1 (16.9)	7.6 (1.1)

F: Female; GLD: Glucose-lowering drug; GLP-1RA: Glucagon-like peptide-1 receptor agonist; HbA1c: Glycated hemoglobin; M: Male; MFN: Metformin; MTD: Maximum tolerated dose; N/A: Not applicable; OAD: Oral antidiabetic drugs; SU: Sulfonylurea.

RoB of the included studies

Supplementary Figure 1 depicts the RoB across the four studies included in the MA. The overall RoB, assessed by the RoB2 tool, was low in all three RCTs in this SR/MA. The non-randomized prospective trial included in this SR/MA had a serious overall RoB stemming from bias due to confounding, as evaluated by the ROBINS-I assessment tool. Publication bias evaluation was not conducted due to insufficient RCTs (fewer than 10) in the forest plots[23].

Safety of GLP-1RA

The comparison of the proportions of study subjects in the GLP-1RA and non-GLP-1RA groups experiencing AEs during Ramadan fasting is summarized in Table 2. Identical proportions of study subjects in the GLP-1RA and non-GLP-1RA groups had any AEs, serious AEs, or AEs leading to treatment discontinuation. Study subjects in the GLP-1RA group had higher risks of GI AEs [RR = 5.27, 95%CI: 1.40 to 19.89, I² = 50% (low heterogeneity), P = 0.01], nausea [RR = 14.05, 95%CI: 2.69 to 73.33, I² = 0% (not important heterogeneity), P = 0.002], and vomiting [RR = 5.96, 95%CI: 1.05 to 33.77, I² = 0% (not important heterogeneity), P = 0.04] than those in the non-GLP-1RA group. However, the risks of heartburn, abdominal pain, and diarrhea were not statistically increased in the GLP-1RA group compared to the non-GLP-1RA group. Compared to the non-GLP-1RA group, the GLP-1RA group had a lower risk of documented symptomatic hypoglycemia [RR = 0.38, 95%CI: 0.16-0.88, I² = 45% (low heterogeneity), P = 0.02]; however, the risk of any hypoglycemia was statistically identical in the two groups. None of the participants in either group experienced severe hypoglycemia during Ramadan fasting.

Table 2 Comparison of the safety outcomes in the glucagon-like peptide-1 receptor agonist vs non-glucagon-like peptide-1 receptor agonist arms.

Outcome variables	Number of included studies	Number of participants with outcome/participants analyzed (%)		Pooled effect size, risk ratio (95%CI)	I2 (%)	P value
		GLP-1RA arm	Non- GLP-1RA arm
Any AE	3	54/311	50/314	1.07 (0.76-1.52)	0	0.69
Serious AE	2	2/263	0/262	4.97 (0.24-102.78)	N/A	0.30
AE leading to discontinuation of treatment	2	1/263	0/262	3.00 (0.12-72.70)	N/A	0.50
Gastrointestinal AEs	3	38/318	8/337	5.27 (1.40-19.89)	50	0.01
Heartburn	2	7/147	1/167	4.10 (0.25-68.20)	50	0.33
Nausea	3	17/318	1/337	14.05 (2.69-73.33)	0	0.002
Vomiting	2	9/263	1/262	5.96 (1.05-33.77)	0	0.04
Abdominal pain	2	3/147	0/167	4.61 (0.52-41.28)	0	0.17
Diarrhea	2	5/263	3/262	1.53 (0.40-5.87)	0	0.53
Any hypoglycemia	3	22/318	47/337	0.60 (0.19-1.90)	72	0.39
Documented symptomatic hypoglycemia	3	14/310	39/312	0.38 (0.16-0.88)	45	0.02

AE: Adverse events; GLP-1RA: Glucagon-like peptide-1 receptor agonists; N/A: Not applicable.

Effect on glycemic control

At the end of the studies, change in HbA1c from the baseline was greater in the GLP-1RA than in the non-GLP-1RA group [MD: -0.31%, 95%CI: -0.61 to -0.01, I² = 77% (high heterogeneity), P = 0.04] (Figure 2A)[17-20].

Figure 2 Forest plot demonstrating the mean difference. A: Glycated hemoglobin; B: Body weight in the glucagon-like peptide-1 receptor agonist (GLP-1RA) group vs the non-GLP-1RA group. GLP-1RA: Glucagon-like peptide-1 receptor agonist; MD: Mean difference.

Effects on body weight

Study subjects in the GLP-1RA group achieved a greater body weight reduction than the non-GLP-1RA group [MD: -2.0 kg, 95%CI: -3.37 to -0.63, I² = 95% (high heterogeneity), P = 0.004] (Figure 2B) from baseline to the end of the trials[17-20].

Effect on BP

Study subjects in the GLP-1RA group achieved identical changes in systolic BP [MD: 0.19 mmHg, 95%CI: -4.30 to 4.68, I² = 0% (not important heterogeneity), P = 0.93] (Figure 3A) and diastolic BP [MD: 0.44 mmHg, 95%CI: -2.39 to 3.28, I² = 0% (not important heterogeneity), P = 0.76] (Figure 3B) than the non-GLP-1RA group[19,20].

Figure 3 Forest plot demonstrating the mean difference. A: Systolic blood pressure (BP); B: Diastolic BP; C: Total cholesterol; D: Triglyceride in the glucagon-like peptide-1 receptor agonist (GLP-1RA) group vs the non-GLP-1RA group. GLP-1RA: Glucagon-like peptide-1 receptor agonist; MD: Mean difference.

Effect on lipids

At the end of the studies, the two groups achieved comparable changes in total cholesterol [MD: -19.86 mg/dL, 95%CI: -63.15 to 24.43, I² = 92% (high heterogeneity), P = 0.37] (Figure 3C) and triglyceride [MD: -18.99 mg/dL, 95%CI: -86.83 to 48.84, I² = 74% (moderate heterogeneity), P = 0.58] (Figure 3D) levels from the baseline values[19,20].

Heterogeneity results: Although the results of some analyses (e.g., effects on glycemic control, body weight, and lipids) presented high heterogeneity, sub-analyses could not be performed due to the limited number of studies included in this research.

DISCUSSION

Main findings of this review

This SR/MA analyzed a relatively small dataset from four short-term studies—three RCTs with low RoB and one prospective observational study with serious RoB. The follow-up period ranged from 12 weeks to 33 weeks, involving 754 participants across multiple countries. The SR/MA suggests that compared to other GLDs, the use of drugs in the GLP-1RA class was associated with a statistically significant HbA1c reduction with a significantly lower risk of symptomatic hypoglycemia, without excess risk of serious AEs, and AEs leading to treatment discontinuation. However, as expected, common GI side effects such as nausea and vomiting were observed in the GLP-1RA group without higher risks of heartburn, abdominal pain, and diarrhea. This SR/MA also observed a greater mean body weight reduction from baseline to the end of the trials in participants using GLP-1RA compared to the controls, while there were no significant changes in BP and lipid profiles between the groups.

Implications for clinical practice

Medications in the GLP-1RA class are generally safe and lead to substantial weight loss across all ethnic groups, whether or not they have T2DM, as demonstrated by many studies[24-26]. During the prolonged daytime fasting period of Ramadan, the use of GLDs can be risky due to the exaggerated tendency for hypoglycemia and GI side effects. However, our SR/MA suggested that GLP-1RA use was safe and provided additional benefits of significant weight loss and improvement in HbA1c. Interestingly, protection from hypoglycemia was found among participants, making this drug an attractive option for patients during Ramadan compared to other GLDs, especially insulin and insulin secretagogues. Delayed gastric emptying associated with GLP-1RA could slow down the nutrient flow in the GI tract (GIT) and result in a steady glucose supply to the body from the food prior to initiation of fasting, conferring hypoglycemia protection.

Ramadan fasting often resembles time-restricted eating (TRE) behavior, as in many regions worldwide, fasting lasts more than 12 hours, which is linked to weight loss and metabolic health advantages[27,28]. Whether the use of GLP-1RA in such a situation provides additional metabolic and weight reduction benefits during the Ramadan period is an important research question emerging from our observation in this SR/MA, because even for such a short period TRE, the weight loss advantage observed was relatively great. Moreover, the anorexiant properties of these molecules provide an additional benefit in suppressing food cravings during fasting. Food craving can be an important issue during fasting, especially in T2DM with hyperinsulinemia, and the appetite suppression from GLP-1RA use could offer better (fasting) adherence and quality of life (QoL) to individuals with diabetes during Ramadan. Moreover, the slower and steady nutrient supply from delayed upper GIT motility, as mentioned above with hypoglycemia prevention, adds to better adherence and QoL. Therefore, the higher weight loss and metabolic benefits we observed could be from the effects of TRE on body fat mobilization coupled with reduced energy-rich food consumption due to the anorexiant effects of GLP-1RA use.

As expected, GI AEs were significantly higher in the GLP-1RA group. One of the primary therapeutic effects of the GLP-1RA class of drugs is the slowing of food transit, particularly in the upper GIT. This contributes to the anorexiant properties of these compounds, which often result in nausea and vomiting, increasing the risk of drug discontinuation for some individuals[25,29]. However, treatment discontinuation due to GI AEs was not higher. Furthermore, an exaggerated risk of heartburn, abdominal pain, and diarrhea was not observed as in previous studies of GLP-1RAs[30-32], likely due to altered eating behavior during Ramadan. Total food intake can often be lower than usual during Ramadan fasting, and anorexia due to GLP-1RA use can further reduce the food intake, possibly explaining this interesting finding.

No significant changes in systolic or diastolic BP were observed in our study, unlike other studies that reported improvements in BP with GLP-1RAs[33,34]. This SR/MA also did not indicate significant lipid-lowering benefits for those using GLP-1RA drugs during Ramadan, as seen in other studies involving these medications for patients with or without T2DM[25,35,36]. The discrepancies we observed may be attributed to the relatively shorter duration of the studies and the smaller total number of participants in this SR/MA compared to the aforementioned studies. Another possible explanation is the unique eating pattern of Muslims during "Suhur" and "Iftar" meals before and after their daily fast. These meal practices can often cause spikes in glucose and insulin levels in the body, which can alter the effect of medications. Furthermore, while Islam is a global religion with some common religious standards, marked regional differences in food practices exist. For example, in Arab countries, access to dried fruits such as apricots and dates is easier than in countries like those in South America and sub-Saharan Africa, where the consumption of tubers, bread, and other carbohydrates is more common. Thus, identifying the precise reason for the impact on lipid changes would be challenging without a comprehensive investigation into the dietary changes occurring during Ramadan.

Overall, the GLP-1RA group of agents appears to offer good efficacy and safety for T2DM patients who fast during Ramadan. In addition to the usual pharmacological properties of this drug class, beneficial effects may also relate to the psychosocial impact of Ramadan, characterized by moments of prayer, reflection, and forgiveness that are associated with happiness, reduced stress levels, and lower stress hormones (e.g., cortisol and adrenaline), which can improve insulin resistance and the drug's effect.

Limitations and strengths of our study

We also acknowledge a few limitations. In this SR/MA, the analyzed data is limited to a widely common chronic illness such as T2DM. The follow-up period was relatively short, lasting only 12 weeks to 33 weeks for a lifelong disease, and we do not have the differential effects of the intervention specific to the fasting period and the remaining observation period. Moreover, the study data were collected from a limited number of countries, while Ramadan is a significant religious observance for Muslims worldwide, who have varied cultural, nutritional, and physical practices. Consequently, the wide applicability of our study results may be limited. Moreover, the RoB was low in the three included RCTs, though serious RoB was present in the prospective observational study.

We were unable to assess the socio-economic impact of the efficacy and safety of GLP-RA in this study. Although they are expensive, these drugs may be cost-effective due to their remarkable effectiveness and safety, especially considering the high risk of healthcare costs related to AEs during fasting. Therefore, public health authorities and policymakers should give serious consideration to this option for patients. Despite these limitations, this is the first comprehensive SR/MA to evaluate the effects of GLP-1RAs in patients with T2DM who fast during Ramadan.

CONCLUSION

This SR/MA, based on limited data from a few small-scale studies across various countries, suggests that GLP-1RA may be safe and modestly effective in maintaining glycemic control compared to other GLDs in patients with T2DM during Ramadan fasting. Though the GI AEs appear higher, that did not result in treatment discontinuation. Even if weight loss might be an advantage, other positive cardiometabolic outcomes, such as improvements in lipid profiles and reductions in BP associated with GLP-1RA drugs, were not evident in this SR/MA. Comprehensive multinational RCTs with broader global representation are necessary to strengthen the evidence base for making more robust clinical practice recommendations about the appropriate use of GLP-1RAs during Ramadan.

ACKNOWLEDGEMENTS

We are thankful to Dr. Marina George Kudiyirickal MSc, MJDF-RCS, PhD for providing us the audio core tip of this article