Introduction

The global burden of type 2 diabetes mellitus (T2DM) continues to increase, contributing substantially to morbidity, mortality, and healthcare utilization.^1,2 Although T2DM currently has no curative treatment, effective glycemic control remains essential for reducing diabetes-related complications and improving quality of life.^3,4

Over the past several decades, novel pharmacological therapies for T2DM have been developed. Sodium-glucose cotransporter 2 (SGLT2) inhibitors represent an important therapeutic class that promotes glucosuria, thereby improving glycemic control and contributing to weight reduction through urinary calorie loss.^5-8 SGLT2 inhibitors may also reduce systolic blood pressure (SBP), as demonstrated with empagliflozin.⁹ In addition, evidence from recent clinical trials has established their cardiovascular and renal protective effects in patients with T2DM. ^10,11

In the EMPAREG OUTCOME trial, empagliflozin significantly reduced all-cause mortality, cardio-vascular death, and hospitalization for heart failure. These benefits have been attributed to its diuretic effect and preload reduction on the heart.^7,10 Similarly, the CANVAS program demonstrated greater cardiovascular benefits with canagliflozin, including among patients with a history of heart failure.¹² In the DECLARE-TIMI 58 trial, dapagliflozin reduced the risk of hospitalization for heart failure in patients with both preserved and reduced ejection fraction, whereas the reduction in cardiovascular death was significant only among those with reduced ejection fraction.¹³

SGLT2 inhibitors confer renal benefit by reducing oxidative stress and controlling inflammation.⁷ Although an initial decline in estimated glomerular filtration rate (eGFR) may occur, long-term use preserves kidney function and reduces albuminuria, especially in diabetic nephropathy.¹⁴ A meta-analysis confirmed reductions in end-stage renal disease and acute kidney injury.¹⁵ Overall, SGLT2 inhibitors have a favorable safety profile, although genital and urinary tract infections are recognized adverse events associated with their use.^16,17

Real-world observational studies from various countries have supported the findings reported in clinical trials, although some have reported higher rates of adverse events in routine clinical practice.¹⁸ Regional evidence from the Gulf has echoed these results. In Qatar, dapagliflozin significantly improved glycemic control regardless of background therapy, and studies among Emirati populations have likewise demonstrated efficacy in weight loss and HbA1c reduction.^19,20

In Bahrain, the prevalence of T2DM is 16%, with increased morbidity and mortality.²¹ Although numerous studies have evaluated prevalence, risk factors, comorbidities, and outcomes, limited evidence is available regarding the real-world clinical effectiveness of SGLT2 inhibitors in this population.^16,20 Therefore, this study was conducted among adults with T2DM attending Central Diabe-tic Clinics in Bahrain to assess the efficacy and safety of SGLT2 inhibitor therapy in this population.

Materials and Methods

This retrospective observational study was con-ducted over three years. Data collection took place from August 2023 to April 2024, while data analysis was performed periodically throughout 2024–2025. The observation period extended over 32 weeks with measurements at baseline, 16 weeks, and 32 weeks.

Given the retrospective design of the study and the use of existing data, the requirement for informed consent was waived in accordance with institutional ethical guidelines.

Eligibility Criteria

Bahraini adults aged >30 years with uncontrolled T2DM, defined as HbA1c >7% (53 mmol/mol), and an eGFR >45 mL/min/1.73 m², who attended one of five specified Central Diabetic Clinics (CDCs) in Bahrain (Arad, Bilad Al Qadeem, Budaiya, A’ali, or Mohamed Jassim Kanoo) were eligible. Patients initiating SGLT2 inhibitor therapy, irrespective of SGLT2 inhibitor subclass, were included in the exposed group and matched 1:1 with non-exposed controls based on age and sex. Matching was performed irrespective of the affiliated health center using a Microsoft Excel-based matching process. Patients with type 1 diabetes mellitus, those with an eGFR ≤45 mL/min/1.73 m², and women who were pregnant, lactating, or planning pregnancy were excluded.

Data Collection and Management

Following enrollment, medical records of partici-pants in the exposed and non-exposed groups were reviewed via the ISEHA electronic system (Bahrain’s national electronic health record platform). Data were extracted at baseline, 16 weeks, and 32 weeks and included age, sex, body weight, SBP, glycated hemoglobin (HbA1c), and eGFR. Adverse events and treatment discontinuation were identified from CDC clinic notes and laboratory documentation.

Outcome Measures

Primary outcomes were changes in HbA1c, body weight, SBP, and eGFR from baseline to 16 and 32 weeks. Secondary outcomes were incidence of adverse events, including genitourinary infections, and SGLT2 inhibitor treatment discontinuation

Data Management

Data were initially entered into Microsoft Excel, with continuous variables recorded in their original numeric form and categorical variables, including sex, binary outcomes, and adverse events, appropriately coded. Data were securely stored in encrypted folders on both local hard drives and email-accessible files, with access restricted to study investigators to maintain data confidentiality.

Sample Size Determination

The sample size of 282 participants (141 in the exposed group and 141 in the non-exposed group) was calculated using Epi Info™ software assuming a 95% confidence level, 90% power, and an expected outcome incidence of 60% in the exposed group versus 40% in the non-exposed group. Participants were matched 1:1 by age and sex.

Data Analysis

Data were transferred from Microsoft Excel to Python (version 3.11.6) for statistical analysis, which was performed with support from a statistician. Categorical variables were summarized using frequencies and percentages, while continuous variables were summarized using means and standard deviations. Chi-square tests were used to assess the distribution of categorical baseline characteristics between the exposed and non-exposed groups. As several continuous variables did not satisfy the assumptions of normality and homogeneity of variance, non-parametric methods were used to evaluate changes in HbA1c, body weight, systolic blood pressure (SBP), and estimated glomerular filtration rate (eGFR) over time and between groups. The Friedman test was used to assess within-group changes across time points, followed by Wilcoxon signed-rank tests with Bonferroni correction for pairwise comparisons, while the Mann-Whitney U test was used for between-group comparisons. Chi-square tests were used to assess associations between SGLT2 inhibitor exposure and adverse events, while treatment discontinuation rates were summarized descriptively. Analysis of covariance (ANCOVA) was performed to assess between-group differences while accounting for baseline values. Subgroup analyses were conducted to examine the potential influence of insulin and glucagon-like peptide-1 receptor agonist (GLP-1 RA) use. Multi-variable linear regression models were constructed to evaluate factors associated with changes in body weight, HbA1c, SBP, and eGFR, with age, sex, type of SGLT2 inhibitor, and treatment discontinuation status included as predictor variables. All statistical tests were two-sided, and a P value ≤0.05 was considered statistically significant.

Results

A total of 282 participants were included, with 141 each in the exposed and non-exposed groups, matched by age and sex and followed for 32 weeks. Overall, 60.3% were female, and the mean age was approximately 61 years in both groups. Baseline characteristics were generally comparable, although insulin use was more frequent in the non-exposed group (77.2% vs 53.2%), while GLP-1 receptor agonist use was more common in the exposed group (11.3% vs 3.5%). Among exposed participants, 85.8% received empagliflozin and 14.2% received dapagliflozin (Table 1).

At 32 weeks, the exposed group demonstrated significantly greater reductions in body weight and HbA1c than the non-exposed group (both p < 0.001). Weight decreased significantly in the exposed group from baseline to 16 and 32 weeks (p < 0.001), with significantly greater reductions than the non-exposed group at both time points. Clinically meaningful weight loss (≥5% of baseline weight) was achieved by 28.6% versus 10.6% of participants in the exposed and non-exposed groups, respectively (p < 0.001; OR 3.36, 95% CI 1.76–6.43). After adjustment for baseline weight, SGLT2 inhibitor exposure remained independently associated with lower 32-week weight (p < 0.001) (Figures 1 and 2).

HbA1c decreased significantly in the exposed group across follow-up (p < 0.001), with a significantly greater reduction than in the non-exposed group at 32 weeks (p < 0.001). Higher baseline HbA1c was independently associated with greater HbA1c reduction, while SGLT2 inhibitor exposure remained independently associated with greater reduction after adjustment (both p < 0.001) (Figure 3).

eGFR increased significantly in the exposed group by 32 weeks (p < 0.001), with a greater improvement than in the non-exposed group (p = 0.008). However, after adjustment for baseline eGFR, SGLT2 inhibitor exposure was no longer independently associated with eGFR improvement (p = 0.143). The improvement was most evident among insulin users and participants not receiving GLP-1 receptor agonists (Figure 4).

No significant within-group change in SBP was observed over time; however, the exposed group had a modestly greater reduction in SBP at 32 weeks (p = 0.048), particularly among insulin users (p = 0.005).

Overall, 13.4% of participants discontinued SGLT2 inhibitor therapy. The documented reasons included cost, non-adherence, urinary tract infections, intolerance, and vaginal infections, although the reason for discontinuation was unknown in 72.4% of cases. Withdrawal was not significantly associated with age, sex, SGLT2 inhibitor type, or insulin use (Figure 5).

Discussion

Management of T2DM increasingly focuses not only on glycemic control but also on reducing cardiovascular and renal complications. SGLT2 inhibitors have demonstrated favorable metabolic, cardiovascular, and renal effects across randomized clinical trials involving diverse populations. This longitudinal study provides real-world evidence on SGLT2 inhibitor use among Bahraini adults with T2DM attending five Central Diabetic Clinics and demonstrated significant improvements in weight and glycemic control over 32 weeks, with more modest changes in renal function and SBP.

The exposed group achieved a median weight reduc-tion of 4.0 kg over 32 weeks, with a significantly greater proportion achieving ≥5% weight loss than the non-exposed group (28.6% vs 10.6%). These findings are consistent with previous trials, including EMPA-REG OUTCOME and DAPACKD, which reported mean weight reductions of approximately 2-3 kg and 2.5 kg, respectively.^11,22 The significant weight reduction observed as early as 16 weeks may reflect the effects of increased urinary glucose excretion and osmotic diuresis associated with SGLT2 inhibition.²² The smaller additional reduction between 16 and 32 weeks may indicate attenuation of the initial weight-lowering effect, consistent with previously reported metabolic adaptations to SGLT2 inhibitor therapy.²³

SGLT2 inhibitor exposure was also associated with significant reductions in HbA1c at 16 and 32 weeks, with greater reductions observed among participants with higher baseline HbA1c.²⁴ The absence of a significant difference according to insulin use suggests that the observed glycemic improvement was not restricted to patients receiving insulin therapy. These findings are consistent with the established role of SGLT2 inhibitors in glycemic management in T2DM.²⁴

Renal findings require more cautious interpretation. The proportion of exposed participants classified as G1 (eGFR ≥90 mL/min/1.73 m²) increased from 41.1% at baseline to 51.8% at 32 weeks, and eGFR improvement was greater in the exposed group in the unadjusted analysis. However, after adjustment for baseline eGFR, SGLT2 inhibitor exposure was no longer independently associated with eGFR change (p = 0.143). Thus, the observed short-term improvement should not be interpreted as evidence of renal protection and may partly reflect baseline differences and regression to the mean. Longer-term studies are required to determine the effects of SGLT2 inhibitors on renal outcomes in the Bahraini population.

The exposed group also demonstrated a modest but statistically significant greater reduction in SBP at 32 weeks. This finding is consistent with the natriuretic and osmotic diuretic effects associated with SGLT2 inhibition.¹³ However, the magnitude of the difference was small, and residual confounding from concomitant antihypertensive therapy, dietary factors, or lifestyle changes cannot be excluded.

SGLT2 inhibitor discontinuation occurred in 13.4% of exposed participants. The reason for discontinuation was unknown in approximately 72% of cases because of incomplete documentation. Among documented reasons, cost, urinary tract infections, and vaginal infections were the most frequently reported. The cost-related discontinuation may partly reflect the limited availability of SGLT2 inhibitors in primary healthcare centers during the study period. These findings are broadly consistent with real-world evidence describing treatment discontinuation related to tolerability, genitourinary infections, and treatment access.²⁵ Hypoglycemia could not be adequately assessed because of insufficient documentation in the electronic health records.

Strengths and limitations

This study provides real-world evidence regarding SGLT2 inhibitor use among Bahraini adults with T2DM and evaluates multiple metabolic and clin-ical outcomes over 32 weeks. However, several limitations should be considered. The retrospective observational design limits causal inference and introduces the possibility of residual confounding. The study population was restricted to patients attending five CDCs, which may limit generaliza-bility to other populations. In addition, the relatively short follow-up period precluded assessment of long-term cardiovascular and renal outcomes. Incomplete documentation also limited evaluation of the reasons for treatment discontinuation and hypoglycemic events. Further prospective studies with longer follow-up are warranted to evaluate the long-term effectiveness, safety, cardiovascular outcomes, and renal outcomes associated with SGLT2 inhibitor therapy in Bahraini patients with T2DM.

Conclusion

SGLT2 inhibitor use was associated with clinically significant reductions in body weight and HbA1c among Bahraini adults with T2DM, with the great-est weight reduction observed during the first 16 weeks. The findings support their potential role in patients with poor glycemic control and overweight/ obesity. SGLT2 inhibitors were generally well tolerated, although cost and genitourinary infections were important barriers to treatment continuation. Further studies with longer follow-up are needed to evaluate cardiovascular and renal outcomes, hypoglycemia, and strategies to improve treatment persistence and minimize adverse events.

Ethics and Dissemination

The application for ethical approval from the research committee in the primary care sector in the Ministry of Health (MOH) in Bahrain was obtained before initiating the research process.

Confidentiality

All the collected and recorded data were kept confidential following institutional policies and the Health Insurance Portability and Accountability Act (HIPAA). The investigator did not use the collected data and records for any purposes other than conducting this study.

Acknowledgment

The author would also like to thank the supervisor, Mrs. Joanne Lowe, for the review and advice during the research protocol performance.

The author acknowledges the assistance of both Mr. Angus McFadyen and Mr. Hasan Albasri, who helped in providing information about the sample size calculation and data analysis for the study protocol.

Funding and budget

No funding required. The investigators took re­sponsibility for providing the required budget and resources.

Conflict of Interest
Nil