FormalPara Key Summary Points

Why carry out this study?

The use of nonnutritive sweeteners (NNS) in diabetes remains a debatable issue.

There is very little data on the benefits or adverse effects of using small quantities of NNS in daily beverages like coffee and tea from South Asia where the overall carbohydrate consumption, particularly white rice, is very high.

Thus, the present study sought to explore, among Asian Indian with T2D individuals, whether a small reduction in calories by replacing added sucrose in coffee/tea with a NNS (sucralose) has cardiometabolic benefits.

What was learned from this study?

Reduction of 60 kcal from added sucrose over 12 weeks showed no change in glycemic, lipid, or inflammatory measures but resulted in a small reduction in body weight, BMI, and waist circumference.

In individuals with T2D whose carbohydrate intake is high, replacing sugar in coffee/tea with sucralose may not improve glycemia but may have a modest body weight benefit.

Introduction

India is witnessing a huge rise in non-communicable diseases, including type 2 diabetes (T2D) [1]. Indians are at high risk of developing diabetes in their lifetime [2]. The recent ICMR-INDIAB study estimated that 101 million people had diabetes and 136 million had prediabetes in India [3]. A positive association has been shown between the consumption of dietary sugars and cardiometabolic diseases [4, 5]. Intake of sugar is linked to worsening glycemia and inflammatory markers [6]. Studies from India showed an increased risk of T2D with excess carbohydrate intake [7, 8]. Given the negative health consequences, the World Health Organization (WHO) recommended limiting free sugars to less than 10% of total calorie intake and further reduction to less than 5%, for additional health benefits [9].

The American Diabetes Association (ADA) supported the use of nonnutritive sweeteners (NNS) in moderation as a substitute in sugar-sweetened products, provided it reduces the overall calorie and carbohydrate intake [10]. Generally, people consume 7–20% of added sugars in their daily beverages of coffee and tea [11, 12]. A study of the free-living diabetes population in urban India revealed that more than half of this population used added sugar in their coffee and tea, making the beverages a potential daily source of sugar intake in their diets [13]. Sucralose, a NNS derived from sucrose, is globally approved for use in foods and beverages. Sucralose is 600 times sweeter than sucrose and the addition of a very small amount to foods and beverages can provide a similar level of sweetness. Sucralose neither provides calories nor affects blood glucose levels and hence it may be suitable for individuals with diabetes who wish to reduce their calorie or carbohydrate intake [14]. Recently, the WHO cautioned against the use of NNS to control body weight but clearly states that this was mainly meant for people without diabetes [15]. However, the WHO warning has raised concerns among the healthcare stakeholders and the public regarding the use of NNS even among those with T2D. There is very little data on the benefits or adverse effects of using small quantities of NNS in daily beverages like coffee and tea from South Asia where the overall carbohydrate consumption particularly in the form of white rice is very high [8]. This randomized controlled trial (RCT) was therefore conducted to evaluate the effect of replacing added sugar—sucrose in coffee/tea with NNS sucralose for a duration of 12 weeks to look at its effect on cardiometabolic risk factors in Asian Indian adults with T2D.

Research Design and Methods

Study Design

This short-term (12 weeks), parallel-arm RCT was conducted at the Madras Diabetes Research Foundation (MDRF), Chennai, India. Institutional ethics committee approval was obtained before the start of the study (EC reference no. MDRF/NCT/02-03/2021). The trial is registered with the Clinical Trials Registry of India (CTRI/2021/04/032686).

Participants

Men and women with T2D in the age group of 30–50 years, with body mass index (BMI) ≥ 18.5 kg/m2 and HbA1c in the range of 7–12%, who were consuming added sugars (sucrose) in their beverages (coffee or tea) were included in the study. All participants were on a stable dose of oral hypoglycemic agent(s). Those individuals who were willing to follow the study protocol and gave informed consent to use the NNS sucralose for 12 weeks were included in the study.

We excluded from the trial all potential participants with any of the following: (1) acute infections (fever, cough, cold, or diarrhea) for the last 1 month, (2) disorders of liver, thyroid, or other endocrine diseases, (3) diabetic kidney disease, (4) respiratory disorders, cancers, or recent heart attack or stroke, (5) any eating disorders, (6) pregnancy and lactation, or (7) long travel plans. The purpose of the trial, the risks, and possible trial benefits were explained to the participants, and written informed consent was obtained before participation in the trial.

Randomization

The study participants were randomized to either the intervention group or the control group. The intervention group replaced added sugar/sucrose in coffee/tea with tabletop sweetener containing sucralose (in pellet/powder/liquid form) for 12 weeks, which resulted in an average reduction in energy intake of 60 kcal (up to 3 teaspoons of added sugar/sucrose) per day (3 teaspoons of sugar × 5 g = 15 g × 4 kcal = 60 kcal). The control group was advised to continue their routine diet, which included the use of added sugar in coffee or tea. Both groups were advised to continue their usual physical and other lifestyle activities. The randomization sequence was computer-generated. The randomization list was generated to ensure treatment balance by using SAS® software (SAS Institute Inc., Cary, NC, version 9.4). Randomization was stratified according to gender (male and female) with the objective to ensure balance between groups. Equal numbers of participants were randomized to each treatment (intervention or control).

Study Procedures

Eligible participants entered a 1-week run-in period to test their compliance with the regimen, and their preference for tabletop sweetener containing sucralose in either pellet/powder/liquid form. The participants in both groups were habitually consuming added sugar in the form of sucrose in their coffee/tea at baseline, despite being on treatment for T2D. Participants in the intervention group were advised to replace sugar with sucralose-based tabletop sweetener in the pellet (1 pellet = 0.085 g), powder (1 measured spoon = 0.5 g), or liquid form (1 drop = 0.05 ml) per one teaspoon of added sugar/sucrose in coffee or tea without changes in other lifestyle factors (particularly physical activity), and routine medications. Participants were also instructed not to use NNS while preparing Indian sweets or desserts and were further advised against consuming any carbonated beverages containing NNS. All the participants were further instructed to refrain from feasting and fasting during the study period.

Demographic Assessment

An interviewer-administered questionnaire was used to ascertain the demographics, medical history, and lifestyle factors such as dietary habits, physical activity, alcohol consumption, and smoking status.

Anthropometric Measurements

Body weight (BW, in kilograms) was measured by a calibrated electronic weighing scale (OMRON; Omron HBF 212, Tokyo, Japan), and waist circumference (WC, in centimeters) was measured using a non-stretchable measuring tape. Briefly, the participants were requested to stand straight with feet together and WC was measured at the smallest horizontal girth between the costal margins and the iliac crest at the end of expiration. WC was measured twice, and the average was considered for statistical analysis. BMI was calculated as weight (in kilograms)/height (meters squared). Lipid accumulation product (LAP), visceral adiposity index (VAI), and triglyceride (TG)/glucose index were calculated using the following formulae [16, 17]:

  • Lipid accumulation product

    • Men: (WC [cm] − 65) × (TG [mmol/L])

    • Women: (WC [cm] − 58) × (TG [mmol/L])

  • Visceral adiposity index

    • Men: {WC [cm]/(39.68 + [1.88BMI [kg/m2]])}× (TG [mmol/L]/1.03) × (1.31/HDL [mmol/L])

    • Women: {WC [cm]/(36.58 + [BMI [kg/m2] × 1.89])} × (TG [mmol/L]/0.81) × (1.52/HDL [mmol/L]))

  • Triglyceride/glucose index: ln (Fasting triglyceride [mg/dl] × fasting glucose [mg/dl])/2

Anthropometric measurements were repeated at the end of every month for both groups.

Blood Biomarkers

Biochemical parameters were analyzed using a standardized protocol. Participants were advised to fast for 10–12 h overnight prior to the procedure. Blood samples were drawn by venipuncture by a qualified phlebotomist. Plasma and serum samples were separated for glycemic and lipid measurements. Investigations including glycated hemoglobin (HbA1c), fasting plasma glucose serum lipid profile, and inflammatory markers were estimated at baseline and the end of 12 weeks. Details of the biochemical assessments are provided in Appendix 1 (electronic supplementary material).

Clinical Assessment

Blood pressure measurements were recorded at every visit. Participants were requested to sit in an upright position relaxed for a few minutes, and blood pressure levels were then measured in the left arm twice at intervals of 5 min using an electronic OMRON machine (Omron Corporation HEM 7120, Tokyo, Japan) and the mean of the two readings was recorded.

Dietary and Compliance Assessments

As a marker of compliance, 24-h dietary recall was collected from all the participants at periodic intervals. The average of two recalls collected at screening and baseline was used as baseline and the average of six recalls collected during the 12-week study period was used as the end of study assessment. The nutrient data for the consumed diet was obtained from in-house software (EpiNu). Plasma sucralose was assessed in a subsample of 50 participants (25 from each group). The details of the sucralose assessment are described in Appendix 2 (electronic supplementary material). Additionally, data on adverse events were collected. The adverse event questionnaire was designed to capture information such as duration of the event, degree of severity and its effect, and treatment to address the adverse event and report on casualty if any.

Contact by telephone and, when needed, house visits were made to assess study compliance. Participants in the intervention group also returned empty containers/bottles of sucralose to check compliance at follow-up visits.

Study Outcomes

The primary outcome of the study was the change in HbA1c from baseline to week 12. The secondary outcomes included a change in BW, BMI, WC, glycemic profile (fasting plasma glucose), serum lipid profile (total, HDL, and LDL cholesterol and serum triglycerides), and inflammatory markers (high-sensitivity C-reactive protein [hs-CRP; in mg/L] and tumor necrosis alpha [TNFα; in pg/ml]) from baseline to week 12.

Sample Size

A sample size of 105 participants in each of the intervention and control groups, inclusive of 20% dropout, was achieved with 80% power to reject the null hypothesis of equal means with mean difference in HbA1c of 0.4% between the intervention and control group with standard deviation of 0.9% for intervention and 1.0% for control group and with significance level of alpha (p < 0.05) [18,19,20,21].

Statistical Analysis

Statistical analysis was performed using SAS software, version 9.0 (SAS Institute Inc., Cary, NC). Paired t test was performed to determine the changes in primary and secondary outcome measures within the group over time. Monthly anthropometric measures such as BW, BMI, WC, and blood pressure measurements were considered for intention-to-treat analysis using one-way analysis of variance (ANOVA). However, the biochemical parameters were collected only at baseline and end of the 12 weeks and so change in from baseline could be assessed only from those who completed the study. Thus, this is per protocol analysis. Paired t test was performed to determine the changes in primary and secondary outcome measures within the group over 12 weeks. The difference between the two groups was assessed using the generalized linear model. Data was presented as mean ± standard deviation for continuous and n (percentage) for categorical variables. Statistical significance was set at p < 0.05.

Ethical Approval

The study was approved by the Institutional Ethics Committee at Madras Diabetes Research Foundation (EC reference no. MDRF/NCT/02-03/2021) and was conducted in accordance with guidelines in the Declaration of Helsinki. The study was registered in the Clinical Trial Registry of India and Institutional ethics committee approval was obtained before the start of the study (CTRI/2021/04/032686). The purpose of the trial, the risks, and possible trial benefits were explained to the participants, and written informed consent was obtained before participation in the trial.

Results

Out of a total of 300 adults screened, 210 were eligible for the study. A total of 31 participants (14.8%) dropped out of the study for various reasons including unexpected travel, relocation due to a change of job, and other personal reasons. Thus, 179 individuals were included in the final analysis (Fig. S1). The baseline characteristics included both demographic and dietary profiles (with added sucrose, in grams and as a percentage of energy intake [%E]) presented as mean ± SD and showed no significant difference between groups (Table 1). There was no difference with respect to the primary outcome (HbA1c) at the end of 12 weeks, either within or between groups. The difference in HbA1c was 0.1 ± 1.0% in the intervention group and 0.1 ± 1.1% in the control group (non-significant). With respect to the secondary outcomes, the mean BW of participants significantly decreased in the intervention group (within-group change − 0.3 ± 1.6 kg, p = 0.04) compared to the control group (between-group difference − 0.5 kg [95% CI − 1.0, − 0.1], p = 0.02) (Table 2). Additionally, a significant reduction was observed in BMI in the intervention group (within-group change − 0.1 ± 0.6 kg/m2, p = 0.04) at 12 weeks when compared to the control group (between-group difference − 0.2 kg/m2 [− 0.4, 0.0], p = 0.03) (Table 2). Waist circumference significantly decreased in the intervention group (within-group change − 0.9 ± 1.9 cm, p < 0.0001), whereas no change was noted in the control group. The between-group change in waist circumference was significant (− 0.8 cm [− 1.4, − 0.3], p = 0.002). There was no significant change in the fasting plasma glucose or serum lipid parameters in either group (Table 2).

Table 1 Baseline anthropometric, clinical, biochemical, and nutrient profile of the study participants (n = 210)
Full size table
Table 2 Change in anthropometric, clinical, and biochemical profile of the study participants randomized into intervention and control groups (n = 179)
Full size table

Intention to treat analysis was considered for anthropometric and blood pressure measurements that were repeated every 4th week (see Table S1 in the electronic supplementary material for details). The results were similar to the per protocol analysis (Table 2) with regards to a significant reduction in BW, BMI, and WC.

The LAP and VAI significantly decreased in the intervention group (between-group change − 13.2 [− 23.0, − 3.4], p = 0.01 and − 0.5 [− 1.1, − 0.0], p = 0.04, respectively). A significant reduction was also observed in the triglyceride/glucose index in the intervention group compared to the control group (between-group difference − 0.2 [− 0.3, 0.0], p = 0.04) (Table 2).

There was also a significant reduction in the intake of total calories (− 83.4 [− 161.4, − 5.4] kcal, p = 0.04) and available carbohydrates (in grams, p < 0.0001; and %E, p = 0.04) in the intervention compared to the control. Additionally, to the added sucrose intake, a 12-g reduction in white rice consumption also contributed the decrease in from the diet (data not shown). The intake of added sucrose was significantly lower in the intervention group compared to the control group (absolute intake − 14.6 [− 17.0, − 11.2] g and percentage calories − 3.3 [− 4.0, − 2.6] %E, both p < 0.0001) at the end of 12 weeks (Table 3).

Table 3 Change in dietary intake of the study participants randomized into intervention and control group [24-h recall] (n = 179)
Full size table

With respect to inflammatory markers, there were no significant changes in hs-CRP and TNFα in the intervention group (Table 4).

Table 4 Change in inflammatory markers of the study participants in intervention and control groups (n = 169)
Full size table

No sucralose was detected in the plasma of the subsamples in both groups [limit of detection (LOD) was 0.5 ppm, which is 0.5 mg/kg].

The participants did not report any apparent adverse outcomes, such as nausea, vomiting, diarrhea, or headache, during the weekly assessments following the ingestion of NNS sucralose.

Discussion

In the present short-term study, in Asian Indian adults with T2D, replacing added sucrose in coffee and tea with NNS sucralose showed no change in the primary outcome—HbA1c. However, small decreases were observed in body weight, BMI, and waist circumference. The small reduction in calories derived from refined carbohydrates (about 60 kcal) helped to reduce daily total calorie and carbohydrate intake which supports the recent American Diabetes Association (ADA) recommendation [10]. The results are consistent with previous RCTs which had demonstrated that sucralose consumption had no impact on glycemic parameters including plasma glucose levels, HbA1c, plasma glucagon-like peptide 1 (GLP-1), or gastric inhibitory polypeptide (GIP) levels [22]. A meta-analysis of 29 RCTs, including a total of 741 participants, examined the glycemic impact of four NNSs, namely saccharin, aspartame, sucralose, and stevia, of which sucralose and aspartame emerged as two of the most extensively used NNSs. That study also reported that the consumption of NNSs did not result in a change in blood glucose [23].

Our study shows a small improvement in body weight, waist circumference, and BMI over 12 weeks of study. It has been shown that even a subtle imbalance of energy intake and energy expenditure of 1–2% (25–50 kcal) per day could lead to considerable weight gain of 1.3–2.7 kg per year [24]. This underscores that small but consistent efforts in reducing sugar intake can have a cumulative and positive impact on body weight over time, albeit small. This supports the model proposed by Hall et al. [25] that, for an overweight adult, an average reduction of approximately 24 kcal per day will result in an approximate 0.5 kg weight loss in about 1 year. Most RCTs on NNS and weight loss have been reported among overweight/obese individuals [26,27,28] and our study is one of the first to show similar weight loss in adults with T2D. The present study highlights the importance of even such small reductions in daily calories derived from sugar. The changes in anthropometric measures in our study were admittedly small, and therefore is hypothesis-generating. Sustaining the intervention over a longer time might provide more meaningful healthy clinical outcomes like sustained weight loss or, at the very least, prevent weight gain over time by eliminating empty calories from sugars.

In a recent single-arm pilot study among overweight and overweight prediabetic adults in India, the replacement of added sugar with stevia-based tabletop sweetener showed a significant reduction in waist circumference over a 90-day period which is comparable to the present study [29]. Furthermore, a systematic review and meta-analysis of 14 prospective cohort studies showed favorable outcomes when low and no-calorie sweetened beverages substituted sugar-sweetened beverages in people with varying cardiometabolic risk profiles including T2D [30].

LAP and VAI also significantly decreased in the present study. LAP is a product of waist circumference and serum triglycerides and reflects lipid accumulation [31] while VAI is a measure of visceral fat and is the product of waist circumference, BMI, serum triglyceride, and HDL cholesterol levels [32]. The “South Asian” or “Asian Indian” phenotype is associated with abdominal obesity, insulin resistance, and related cardiometabolic disease risk including onset of T2D at a younger age [33]. LAP and VAI are useful biomarkers for elucidating cardiovascular risks among adults with T2D [31, 32]. Additionally, LAP and VAI are potential tools for predicting microvascular complications in people with T2D [34]. In the present study, the replacement of added sucrose in daily beverages showed a significant reduction in waist circumference, and a non- significant reduction in serum triglycerides level, which likely led to a significant reduction in LAP despite no significant improvement in serum lipids.

A randomized crossover trial reported that the consumption of high doses of sucralose (in beverages) is associated with the presence of sucralose in plasma [35]. However, in our study, plasma sucralose was not detected, likely because the levels were below the detectable limits of the assay. However, there is evidence to suggest that sucralose is poorly absorbed, undergoes little metabolism, and is excreted primarily unchanged in the feces of animals and humans [36]. This probably explains why sucralose was not detected in the plasma.

Assessment of the 24-h dietary recall of the participants revealed significant reduction in total energy intake and carbohydrate intake after replacing added sucrose with sucralose for 12 weeks. The integration of sucralose in coffee/tea therefore presents a viable approach to at least partly reduce carbohydrate consumption in South Asians where carbohydrate consumption is very high. This is significant in view of the increased risk of T2D shown with high carbohydrate diets in cross-sectional [7] and longitudinal studies [8, 37] in people of South Asian ancestry.

Human observational studies have reported weight gain and risk of diabetes with consumption of NNS [38]. It is crucial to acknowledge the potential presence of residual confounding in these observational studies, as there might have been inadequate adjustments for various factors pertaining to the characteristics and behaviors of subjects during data analysis. Another limitation of observational study is the possibility of reverse causation which means people who were already at risk of metabolic conditions turn to NNS in their efforts to reduce this risk by cutting down the sugar intake. In addition, the considerable heterogeneity among the study cohorts, potential publication bias, varying outcome measures, the utilization of various types of low-calorie sweeteners, and differing lengths of follow-up periods contribute to significant variability. These factors pose challenges in aggregating and consistently interpreting the results of human studies using NNS [39].

The present study only looked at the use of sucralose in daily beverages like coffee/tea and hence only small amounts of sucralose were consumed. It is difficult to compare our study with earlier studies [40] which mostly evaluated the use of NNS in sugar-sweetened beverages (SSBs) where much large amounts of NNS are consumed. There are several strengths to this study. To our knowledge, this is the first study in Asian Indian adults with T2D to study the replacement of added sucrose with sucralose. The effect of sucralose on inflammatory markers in individuals with T2D has also not been reported earlier in South Asians. The study also has a few limitations. This includes a possibility for recall bias in 24-h dietary recalls, as participants may provide incomplete or inaccurate information about their food intake. The study was conducted in a real-world, free-living setting. Thus, there is a possibility that the results might be influenced by the nature of the treatment and participant behavior although these were minimized by frequent contact with and follow-up of the participants. The sensitivity of the instrument which measured the plasma sucralose was around 0.5 ppm (0.5 mg/kg); therefore, any sucralose presence below this level would not have been picked up by the assay. Finally, it is a short-duration study.

Conclusion

The present short-term RCT conducted among Asian Indians investigated the impact of replacing added sugars/sucrose (contributing about 60 kcal) with sucralose in the daily beverages such as coffee or tea consumed by individuals with T2D. The study shows that there was no effect on HbA1c, but a small beneficial effect on body weight, BMI, and waist circumference. In addition, the consumption of sucralose did not show any change in fasting plasma glucose, serum lipid profile, or inflammatory markers. In people of South Asian and Indian ethnicity in whom carbohydrate consumption is very high due to refined cereals intake which is difficult to culturally change, perhaps one of the small steps could be replacing sugar in coffee/tea with a NNS like sucralose. The results of the present study should be considered as hypothesis-generating and warrant a larger and longer-duration RCT to confirm the safety and efficacy of NNS in general and sucralose in particular, to reduce sugar and refined carbohydrate intake.