This study is a double-blind, randomized, placebo-controlled crossover trial. The crossover design allows each participant to serve as their own control across 3 intervention phases. Each participant will complete a 12-week protocol comprising a 1-week run-in period followed by 3 intervention phases, each separated by a 4-week washout period. A total of 6 study visits, comprising 3 preintervention and 3 postintervention assessments, will capture baseline and postintervention data. During the run-in and washout periods, participants will maintain their habitual diet and physical activity. The study will be conducted at a government primary care facility situated in an urban area. Recruitment will be facilitated through clinic listings and physician referrals. Written informed consent will be obtained after participants receive detailed verbal and written explanations of the study procedures, risks, and rights. The study timeline is summarized in Figure 1.

This study aims to recruit a cohort of clinically stable patients representative of Malaysians with early-phase T2DM. Eligible participants will be males aged 30 to 59 years, with a BMI of 23.0 to 29.9 kg/m², classified as overweight [21]. Participants must have a clinical diagnosis of T2DM for 1 to 10 years, with baseline glycated hemoglobin (HbA_1c) below 10%, and be managed with oral antidiabetic medications only. Individuals with stable, nonsevere chronic conditions, such as hypertension or dyslipidemia, will be included if they have maintained a consistent oral treatment regimen for at least 6 months before enrollment. All participants must have previously received dietary counseling for diabetes and agree to maintain their habitual diabetes-friendly diet and regular physical activity throughout the study. They must abstain from alcohol consumption during the intervention period and be literate in either English or Bahasa Malaysia to ensure full understanding of the study procedures and informed consent materials.

Exclusion criteria focus on lifestyle or dietary patterns that deviate from typical Malaysian practices or that could substantially alter the gut microbiota. Individuals will be excluded if they currently smoke or consume alcohol, have experienced more than 5% body weight change within the past 3 months, or have acute illness or major comorbidities, such as cardiovascular, psychological, neurological, renal, or endocrine disorders other than T2DM. Those with known intolerance or allergy to the test products, adherence to special diets (eg, vegetarian, ketogenic, or intermittent fasting), or use of medications or supplements (eg, probiotics) that may affect intestinal function or microbiota composition will also be excluded. Participation in another clinical trial within the past 3 months will preclude eligibility. Participants may withdraw voluntarily or be withdrawn for safety reasons, protocol deviation, or the need for antibiotic therapy. Diabetes management will continue as per clinical practice. Adverse events will be monitored and managed by the study physician following institutional protocols.

Participants will receive 1 of 3 interventions: sucralose, saccharin, or a placebo, for 7 consecutive days during each intervention phase. Participants will consume the assigned capsule each morning after breakfast with plain water to standardize timing and minimize missed doses. Dosages will be adjusted based on body weight according to national guidelines, with 5 mg/kg for sucralose and 2 mg/kg for saccharin [22]. The placebo, consisting of 500 mg of calcium carbonate, will be administered under identical conditions and is considered physiologically inert at this dose [9].

Randomization and allocation concealment will be conducted by external personnel not involved in recruitment, data collection, or analysis. A computer-generated block randomization sequence will assign participants to 1 of 3 intervention orders (sucralose, saccharin, or placebo) to ensure balanced allocation across participants. All test products will be encapsulated in identical opaque capsules and packaged in indistinguishable small plastic pouches. Labeling will be performed by external personnel according to the intervention sequence, where each package will be marked as “first intervention,” “second intervention,” or “third intervention” to maintain blinding. Both participants and investigators will remain blinded throughout the data collection period. The randomization code will be securely stored and accessible only in the event of a medical emergency.

The primary outcome is the change in insulin sensitivity, assessed using the homeostatic model assessment for insulin resistance (HOMA-IR), derived from fasting glucose and insulin concentrations. The secondary outcome is the change in gut microbiota composition and diversity, analyzed through 16S rRNA gene sequencing. Alpha and beta diversity indexes will be calculated to evaluate within- and between-sample diversity, and differential abundance analysis will be performed to identify microbial taxa associated with each intervention.

The sample size for this crossover trial was estimated based on detecting a medium effect size (Cohen f=0.25), consistent with findings from a recent meta-analysis of soluble fiber supplementation in T2DM, which reported moderate standardized mean differences for HOMA-IR (standardized mean difference −0.58, 95% CI −0.86 to −0.29) [23]. Using G*Power 3.1.9.6, we computed the required sample size for a within-subject repeated-measures ANOVA with 3 conditions (sucralose, saccharin, and placebo), assuming α=.05, and statistical power at 80%, and a moderate correlation (r= 0.5) among repeated measurements [24]. The analysis indicated that a minimum of 28 participants would be required. To account for potential dropout or noncompliance, we will recruit 33 participants.

This study was approved by the Universiti Kebangsaan Malaysia Research Ethics Committee (UKM/PPI/111/8/JEP-2025-322) on July 15, 2025. Procedures follow the Declaration of Helsinki and institutional ethics standards. This study is registered with ClinicalTrials.gov (NCT07124585), first released on August 8, 2025.

Written informed consent will be obtained from all participants before enrolment. Participants will receive a verbal explanation of the study and will be given adequate time to review the participant information sheet. An optional separate consent form will be provided for the storage and future use of deidentified biospecimens in ancillary studies.

Participant privacy and confidentiality will be protected throughout the study. Identifiable information will be collected only as necessary and will be stored separately from research data using unique participant codes. All datasets will be deidentified before analysis, and access to identifiable information will be restricted to authorized study personnel. No identifiable data or images will appear in publications or presentations.

Participants will receive financial compensation for time and transportation. Compensation for harm is not anticipated, as the interventions involve commercially available NNSs administered at doses considered safe. Any adverse events will be documented and managed according to institutional procedures, and medical care will be provided as needed.

Data and biospecimens will be collected at baseline and after each intervention phase by trained research personnel as outlined in the study visit overview (Table 1).

Anthropometric measurements will be obtained using standardized procedures. Body weight will be measured using a calibrated digital scale with participants wearing light clothing, while height will be measured using a stadiometer, ensuring an upright posture and proper head alignment. The average of 2 readings will be recorded for both weight and height, and BMI will be calculated as weight in kilograms divided by height in meters squared (kg/m²). Waist circumference will be measured at the midpoint between the lowest rib and the iliac crest using a nonstretchable measuring tape [25]. Body composition, including body fat and muscle mass, will be assessed using bioelectrical impedance analysis following standard premeasurement guidelines [26]. Blood pressure will be measured twice after a 5-minute rest using an automated sphygmomanometer, and the mean of both readings will be recorded.

For biological specimens, blood samples of approximately 15 mL will be collected from participants at each study visit by trained nurses after a 30-minute rest period. Analyses will include fasting glucose, fasting insulin, glycated hemoglobin (HbA_1c), lipid profile, and kidney function indicators. Insulin resistance will be estimated using the HOMA-IR, derived from fasting glucose and insulin levels [27]. Samples will be transported and processed according to biosafety and quality control procedures to ensure sample integrity for biochemical analysis and future molecular investigations. Stool samples of approximately 1 g will be collected using standardized fecal collection tubes provided to participants. Preintervention samples will be collected a day before each intervention phase, and postintervention samples will be obtained immediately after the final intervention day. Participants will receive detailed written and verbal instructions to ensure correct collection and handling procedures [28].

Sociodemographic and medical history information will be obtained through structured interviews and verified against medical records. Medication use will be reviewed at each study visit. Dietary intake will be assessed using 3-day food records covering 2 weekdays and 1 weekend day. Investigators will review the records during face-to-face sessions using visual aids to ensure completeness and improve portion-size estimation accuracy. Nutrient intake will be analyzed using the Malaysian Food Composition Database [29]. Physical activity will be assessed using the short-form International Physical Activity Questionnaire, available in both English and Malay, which has been validated for the Malaysian population [30,31]. Test product compliance records and adverse event reports will be obtained during each postintervention visit. Any serious adverse event or major complaint will be promptly managed by the study physician and reported to the responsible authorities within 48 hours, with appropriate medical action documented. Leftover blood and stool biospecimens from each visit will be stored at −80 °C in the institutional biobank for future research on disease biomarkers, gene expression, or microbial function, subject to ethics approval and participant consent.

Gut microbiota profiling will be performed using 16S rRNA gene sequencing targeting the V3-V4 region. DNA will be extracted using the Maxwell RSC system (Promega), and library preparation will follow Illumina’s 16S Metagenomic Sequencing protocol [32]. Amplicons will be barcoded, pooled equimolarly, and sequenced on the Illumina MiSeq platform (2×250 bp). Raw reads will be demultiplexed, quality-filtered, and denoised using DADA2 in QIIME 2 to generate amplicon sequence variants and remove chimeras [33,34]. Taxonomic classification will be assigned using a pretrained SILVA 138 database classifier [33]. Microbial diversity will be evaluated using α diversity indexes (Shannon, Faith phylogenetic diversity) and beta diversity metrics (Bray-Curtis, UniFrac distances) [35-38]. Differential abundance analysis will be conducted using appropriate statistical frameworks, such as ANCOM-BC or MaAsLin2, with false discovery rate correction [39]. Downstream analyses will be guided by metadata and study endpoints to identify microbial shifts associated with each intervention.

Clinical data will be presented as mean (SD) unless stated otherwise. Statistical analyses will be performed using SPSS Statistics (version 28.0; IBM Corp) [40]. The normality of continuous variables will be assessed using the Shapiro-Wilk test. Participants’ baseline characteristics will be summarized using descriptive statistics. Differences in glycemic responses across the 3 study arms will be assessed using either 1-way ANOVA or the Kruskal-Wallis test, depending on data distribution. Changes in clinical variables before and after the intervention will be analyzed using repeated-measures ANOVA, with time (before vs after), group (intervention arm), and the time-by-group interaction as factors. Sociodemographic characteristics, dietary intake, and physical activity levels will be considered as covariates where appropriate. A P value of <.05 will be considered statistically significant. Analyses will be conducted according to the intention-to-treat principle, with per-protocol analyses performed for participants who complete the study without major deviations. Missing data will be handled using appropriate imputation methods, and any exploratory analyses will be clearly identified and interpreted with caution.