Type 2 diabetes mellitus is a chronic metabolic condition in which persistent hyperglycemia develops through several interacting abnormalities, including insulin resistance, reduced pancreatic beta-cell function, and inappropriate hepatic glucose production. Continued exposure to elevated glucose increases the risk of microvascular and macrovascular complications, so achieving adequate glycemic control early in the course of disease is a major therapeutic objective [1,2].

Management of type 2 diabetes increasingly follows an individualized approach that combines lifestyle measures with pharmacological treatment. Selection of drug therapy is influenced not only by glucose-lowering capacity, but also by the likelihood of hypoglycemia, adverse-effect profile, coexisting illnesses, treatment complexity, and patient preferences [3]. Metformin remains a widely used foundational therapy because it has demonstrated efficacy, a low risk of hypoglycemia when used alone, low cost, and extensive clinical experience [4,5]. Its major antihyperglycemic action is suppression of hepatic glucose production, with additional improvement in insulin sensitivity and peripheral glucose handling. Gastrointestinal symptoms such as nausea, abdominal discomfort, and diarrhea are recognized limitations that may affect continuation of treatment in some patients [6].

Vildagliptin is an oral DPP-4 inhibitor that reduces the degradation of endogenous incretin hormones. By extending incretin activity, it promotes glucose-dependent insulin release and reduces inappropriate glucagon secretion [7]. Monotherapy trials have shown clinically relevant reductions in HbA1c with a low intrinsic risk of hypoglycemia in the absence of insulin secretagogues [8]. Comparative trials have reported sustained glycemic improvement with both vildagliptin and metformin, although the relative magnitude of HbA1c change and gastrointestinal tolerability can vary with the study population, treatment dose, baseline glycemia, and duration of exposure [9,10].

Although evidence comparing these agents is substantial internationally, direct evidence from Indian tertiary-care hospitals is less extensive, particularly among patients beginning treatment with monotherapy. Postprandial glucose deserves attention because meal-related glucose excursions contribute to overall glycemic exposure and may remain abnormal despite improvement in fasting glucose. Comparing fasting glucose, two-hour postprandial glucose, HbA1c, and treatment-emergent adverse events can therefore help describe the short-term response profile of the two drugs.

Accordingly, the present study compared the efficacy and safety of vildagliptin with metformin in newly diagnosed adults with type 2 diabetes mellitus treated at a tertiary-care hospital in Tirupati, India. The predefined efficacy measures were changes in fasting blood glucose, two-hour postprandial blood glucose, and HbA1c during a 6-month period, while safety was evaluated through systematic recording of adverse drug reactions.

Study design and setting: A prospective, randomized, controlled, single-center investigation was performed in the General Medicine outpatient department of Sri Venkateswara Ram Narayan Ruia Government General Hospital, Tirupati, which is affiliated with Sri Venkateswara Medical College. The study was conducted from December 2023 through August 2024, and each participant was followed for six months. The documented allocation method was based on participant numbering: odd numbers were assigned to the vildagliptin arm and even numbers to the metformin arm. Ethical approval and consent: The study protocol was approved by the Institutional Ethics Committee of S.V. Medical College, Tirupati. The archived approval communication is dated March 1, 2023 and states that the protocol had been reviewed during the committee meeting of February 24, 2023; the application reference is recorded as Lr. No. 12/2023. The exact formal IEC approval/reference number should be cross-checked with the original ethics records before manuscript submission. All participants provided written informed consent. No external sponsor supported the study; the principal investigator financed the work independently. Participants: Adults 30–60 years of age with newly diagnosed type 2 diabetes mellitus were considered eligible after providing written informed consent. Participants of both sexes were included. Patients were excluded when diabetic retinopathy, nephropathy or neuropathy, substance abuse, hepatic or renal disease, allergy to either study medication, pregnancy or lactation, or an immunocompromised condition such as HIV infection or previous renal transplantation was present. Sample size calculation: For sample-size estimation, the study used a two-sided alpha of 0.05 and 80% statistical power, with an assumed standard deviation of 0.98 and an anticipated clinically meaningful between-group HbA1c difference of 0.8 percentage points. The stated equation was n = 2(Zα + Zβ)^2σ^2/d^2. Substitution of Zα=1.96, Zβ=0.84, σ=0.98, and d=0.8 produced n=23.53, which was rounded to 24 participants per group. A 10% allowance for nonresponse and an additional 10% allowance for attrition led to a planned sample of 30 participants in each group, giving a total target of 60. Randomization, treatment, and follow-up: The 60 eligible participants were divided equally into Group A (vildagliptin) and Group B (metformin) according to the odd-even numbering procedure. Group A received oral vildagliptin 50 mg once daily, one-half hour before breakfast, whereas Group B received oral metformin 500 mg once daily, one-half hour before breakfast. Treatment was continued throughout the study period, with evaluations scheduled at baseline, 3 months, and 6 months. Six participants discontinued during follow-up—two from the vildagliptin group and four from the metformin group—and were replaced using the same eligibility and allocation process so that 30 participants remained in each arm. Therefore, the reported analysis represents the maintained final cohort rather than an intention-to-treat population. Outcome assessment: Fasting blood glucose, two-hour postprandial blood glucose, and HbA1c were measured at baseline and at the 3- and 6-month visits. Approximately 5 mL of fasting venous blood was obtained before breakfast for measurement of fasting glucose and HbA1c, and a further 2 mL sample was collected two hours after breakfast for postprandial glucose estimation. Fasting and postprandial glucose were analyzed by the glucose oxidase-peroxidase method, while HbA1c was measured using high-performance liquid chromatography. Safety surveillance covered allergy, hypoglycemia, gastrointestinal intolerance, dehydration, headache, lactic acidosis, and any other adverse events reported during follow-up. Statistical analysis: The dataset was entered into Microsoft Excel 2019 and analyzed using IBM SPSS Statistics version 25.0. Continuous variables were expressed as mean ± standard deviation, whereas categorical variables were summarized as frequencies and percentages. Paired t-tests were specified for within-group pre/post comparisons and independent-samples t-tests for between-group comparisons, with p≤0.05 taken as the threshold for statistical significance. For the manuscript analysis, the 6-month between-group comparisons for fasting glucose, postprandial glucose, and HbA1c were independently recalculated from the reported means, standard deviations, and n=30 per group using two-sided independent-samples t-tests with pooled variance. Because adverse-event counts were sparse, two-sided Fisher exact tests were considered appropriate for those comparisons. Reference ranges used in figures: The graphical displays used standard reference benchmarks solely to aid interpretation. The fasting glucose figure shows the commonly used normal fasting plasma glucose interval of 70–99 mg/dL. The postprandial graph uses a normal two-hour glucose value of less than 140 mg/dL. For HbA1c, the figures show the normal threshold of less than 5.7% together with the 7.0% treatment target used in the study tables. These benchmarks were provided for context and were not applied as eligibility criteria. The sample-size equation was n = 2(Zα + Zβ)²σ²/d². Using Zα = 1.96, Zβ = 0.84, σ = 0.98, and d = 0.8 yielded n = 23.53, which was rounded to 24 participants per arm. After allowing a combined 20% for nonresponse and attrition, the planned enrollment was set at 30 participants per group.

During the study period, 101 patients were screened. Before enrollment, 41 were excluded: seven had chronic kidney disease, seven had hepatic dysfunction, eight had thyroid disorders, two had type 1 diabetes mellitus, two were pregnant, and 16 did not provide written informed consent. The remaining 60 participants were enrolled and allocated equally to the two treatment groups. During follow-up, six participants discontinued—two from the vildagliptin arm and four from the metformin arm. These withdrawals were replaced using the same eligibility and allocation procedure, leaving 30 participants in each group for the maintained final cohort shown in Figure 1.

The two groups had similar overall age and sex distributions. Vildagliptin was given to 16 males (53.3%) and 14 females (46.7%), whereas the metformin group comprised 14 males (46.7%) and 16 females (53.3%). In both arms, the 51–60-year category was the largest, accounting for 14 participants (46.7%) in each group. Participants aged 41–50 years numbered 10 (33.3%) in the vildagliptin group and 12 (40.0%) in the metformin group; those aged 31–40 years numbered six (20.0%) and four (13.3%), respectively.

Fasting blood glucose decreased over the 6-month observation period in both study arms. In participants receiving vildagliptin, the mean value changed from 175.97 ± 61.99 mg/dL at baseline to 135.80 ± 32.38 mg/dL at month 3 and 118.83 ± 14.86 mg/dL at month 6. The corresponding metformin values were 170.77 ± 51.28, 137.53 ± 32.02, and 124.20 ± 28.75 mg/dL. The reported within-group baseline-to-6-month reductions were statistically significant for both vildagliptin (p=0.0001) and metformin (p=0.0012). The 6-month difference between groups was not statistically significant, with a mean difference of −5.37 mg/dL (95% CI −17.19 to 6.46; p=0.367).

Two-hour postprandial glucose also fell in both groups, with a larger absolute decline observed in the vildagliptin arm. Mean values in the vildagliptin group decreased from 269.00 ± 90.23 mg/dL at baseline to 199.53 ± 40.44 mg/dL at 3 months and 167.53 ± 17.29 mg/dL at 6 months. Metformin-treated participants showed a corresponding change from 294.93 ± 100.08 to 215.87 ± 48.79 and 201.73 ± 45.46 mg/dL. For vildagliptin, the reported comparisons for baseline to 3 months, 3 to 6 months, and baseline to 6 months were statistically significant (p=0.0010, p=0.0012, and p=0.000006, respectively). In the metformin group, baseline-to-3-month and baseline-to-6-month comparisons were significant (p=0.0002 and p=0.0106), whereas the 3- to 6-month comparison was not (p=0.8152). At 6 months, the between-group difference was −34.20 mg/dL (95% CI −51.97 to −16.43; p=0.0003).

HbA1c declined in each treatment arm during follow-up. The mean HbA1c in the vildagliptin group fell from 8.04 ± 0.65% at baseline to 7.40 ± 0.50% at 3 months and 7.03 ± 0.40% at 6 months, corresponding to an absolute decrease of 1.01 percentage points. In the metformin group, the mean changed from 7.88 ± 0.81% to 7.65 ± 0.68% and then to 7.57 ± 0.66%, an absolute decrease of 0.31 percentage points. The HbA1c table used one-sample comparisons against a 7.0% reference target rather than a direct paired analysis of change; therefore, those p-values are not interpreted here as evidence of within-group change. Based on independent-samples t-test recalculation from the reported summary statistics, the 6-month between-group difference was −0.54 percentage points (95% CI −0.82 to −0.26; p=0.0003).

Overall safety findings were similar in that no hypoglycemia, dehydration, lactic acidosis, or other serious adverse event was reported. Allergy occurred in three participants in each group (10.0% in both arms). Headache was reported by four participants (13.3%) receiving vildagliptin and five (16.7%) receiving metformin. Gastrointestinal intolerance occurred in four participants (13.3%) in the vildagliptin group and in none of the metformin-treated participants. No recorded adverse event led to discontinuation of treatment. Because event numbers were small, differences in adverse-event frequencies should be interpreted cautiously.

TABLE 1. BASELINE DISTRIBUTION OF AGE AND SEX

TABLE 2. GLYCEMIC PARAMETERS AT BASELINE, 3 MONTHS, AND 6 MONTHS

TABLE 3. REPORTED WITHIN-GROUP COMPARISONS OF GLYCEMIC MEASURES

TABLE 4. ADVERSE DRUG REACTIONS OBSERVED DURING FOLLOW-UP

FIGURE 1. FLOW OF PARTICIPANT SCREENING, ENROLLMENT, ALLOCATION, AND REPLACEMENT.

FIGURE 2. CHANGE IN MEAN FASTING BLOOD GLUCOSE DURING THE 6-MONTH FOLLOW-UP.

FIGURE 3. CHANGE IN MEAN TWO-HOUR POSTPRANDIAL BLOOD GLUCOSE DURING THE 6-MONTH FOLLOW-UP.

FIGURE 4. CHANGE IN MEAN HbA1c DURING THE 6-MONTH FOLLOW-UP.

This study evaluated vildagliptin and metformin as monotherapies in 60 newly diagnosed adults with type 2 diabetes mellitus over six months. Both treatment groups demonstrated improvements in fasting glucose, postprandial glucose, and HbA1c. The response pattern differed between groups: vildagliptin was associated with a larger observed HbA1c reduction and a lower 6-month postprandial glucose value, whereas fasting glucose values did not differ significantly between treatments. The stronger postprandial reduction observed with vildagliptin is pharmacologically plausible. DPP-4 inhibition preserves endogenous incretin action, thereby enhancing glucose-dependent insulin secretion and reducing glucagon release, effects that are particularly relevant in the post-meal state [7]. In this cohort, postprandial glucose fell by approximately 37.7% from baseline with vildagliptin and by 31.6% with metformin. The calculated 6-month between-group difference was about 34 mg/dL and remained statistically significant when derived from the reported summary statistics. Nevertheless, the present study was not designed to test mechanistic endpoints, so the pharmacological explanation should be regarded as supportive rather than definitive. The HbA1c findings also warrant careful interpretation. The observed reduction was 1.01 percentage points with vildagliptin and 0.31 percentage points with metformin, with a statistically significant difference between groups at six months based on the recalculated analysis. This result does not establish that vildagliptin is generally superior to metformin. Earlier head-to-head work has produced different findings; in a large one-year randomized comparison, both treatments lowered HbA1c, but metformin showed a numerically greater mean reduction and vildagliptin did not satisfy the predefined noninferiority criterion [9]. The current result should therefore be considered specific to the population, treatment doses, and study procedures used here. Differences from larger comparator studies may partly reflect the characteristics of this cohort. The study enrolled a relatively small number of participants at one center and restricted eligibility to newly diagnosed adults aged 30–60 years. Both therapies were administered at fixed doses that were relatively low—vildagliptin 50 mg once daily and metformin 500 mg once daily—and no dose escalation was performed. Baseline postprandial glucose was numerically higher in the metformin arm, while baseline HbA1c was slightly higher in the vildagliptin arm. These baseline differences should be considered when interpreting the magnitude of the observed treatment effects. The observed safety profile was generally reassuring: there were no recorded hypoglycemic events requiring intervention and no serious adverse events. The low intrinsic risk of hypoglycemia with vildagliptin is consistent with its glucose-dependent insulinotropic action [11,12]. Published pooled safety analyses have also described broadly comparable overall adverse-event rates between vildagliptin and comparator treatments [13]. Although gastrointestinal intolerance is a recognized limitation of metformin therapy [6], the present dataset recorded such intolerance only among participants receiving vildagliptin. This study-specific observation should not be generalized beyond the present sample. Several features strengthen the study design. It used prospective follow-up with direct comparison of two monotherapies, assessed three clinically relevant glycemic outcomes, repeated measurements over a six-month period, and documented adverse events systematically. The sample-size calculation was based explicitly on an anticipated HbA1c difference and included allowances for nonresponse and attrition. In addition, the study contributes data from a tertiary-care hospital in South India, a setting for which direct head-to-head evidence remains relatively limited. The principal limitations are the modest sample size and the single-center setting, both of which reduce precision and limit external validity. The results indicate that, within this study, the vildagliptin group had a larger reduction in HbA1c and a lower 6-month postprandial glucose value, while fasting glucose outcomes were comparable. The pattern is consistent with the known pharmacology of DPP-4 inhibition, but it should be regarded as hypothesis-generating because of the small sample, use of fixed low doses, replacement of participants who withdrew, and lack of participant-level data in the archived analysis.

Among newly diagnosed adults included in this prospective randomized study, both vildagliptin and metformin were associated with improved glycemic indices during six months of treatment. The vildagliptin arm showed a greater observed reduction in HbA1c and a lower mean postprandial glucose value at six months, whereas fasting glucose did not differ significantly between groups. No hypoglycemia or serious adverse events were recorded, and both treatments were generally well tolerated in this cohort.

The observed differences should be interpreted within the limits of this study rather than generalized to all patients with type 2 diabetes mellitus. Larger multicenter randomized studies that are prospectively registered, use participant-level intention-to-treat analyses, and apply standardized dose titration would be useful to determine whether the present findings are reproducible across broader clinical populations.

DECLARATIONS FOR SUBMISSION:

Funding: The study received no external funding or sponsorship.

Conflicts of interest: The author reports no conflicts of interest related to this study.

Ethics approval: The study received approval from the Institutional Ethics Committee of S.V. Medical College, Tirupati. The archived approval letter is dated March 1, 2023 and records approval/reference Lr 12/2023.

Consent: Written informed consent was obtained from every participant.