Abstract
OBJECTIVE
To test the effect of substituting plain water (the ideal standard) for habitual artificial sweetened beverage (ASB) intake by people with type 2 diabetes (T2D) on primary measures of diabetes control.
RESEARCH DESIGN AND METHODS
The Study of Drinks with Artificial Sweeteners in People with T2D (SODAS) was conducted at two academic health centers and was a randomized, two-arm, parallel trial with a 2-week run-in period and a 24-week active intervention period. Adults with T2D (n = 181; HbA1c 6.5–8.5%; aged ≥35 years) who regularly consumed commercial ASBs were randomized to receive and consume 24 oz daily for 24 weeks of either 1) a commercial ASB of choice (control) or 2) an unflavored, sparkling or still, bottled or canned water of choice in place of ASBs. The outcomes measures were collected at baseline, 12, and 24 weeks and included the primary measure (HbA1c) and related secondary measures (fructosamine, fasting glucose and insulin, body weight, and continuous glucose monitor metrics).
RESULTS
A total of 179 participants provided complete data over 24 weeks. From baseline to 24 weeks, the mean difference in change of HbA1c was 0.29% (SE 0.12; P = 0.013) higher in the water arm compared with the ASB arm. There were no significant effects on secondary clinical measures, but data were directionally consistent with the primary results.
CONCLUSIONS
For people with T2D and HbA1c <8.5% who regularly consume ASBs, this trial provided no evidence that substituting water would improve glycemic-related clinical care measures over 24 weeks.
Graphical Abstract
Introduction
The global prevalence of adults with diabetes is ∼537 million. More than 38 million people in the U.S. have diabetes, with 90–95% of adults with diabetes having type 2 diabetes (T2D) (1,2). The complexity of managing T2D requires an integrated, evidence-based strategy, due to myriad complications and comorbid conditions being strongly linked to hyperglycemia and glycemic patterns (3,4). Although managing glycemia in T2D is primarily achieved through pharmacotherapy, the integrated approach emphasizes the importance of health-related behaviors, particularly nutritional therapy and dietary recommendations (5–7).
A specific avenue for dietary recommendations that is of high importance and interest to people with T2D but that is lacking direct scientific evidence is how the use of artificial sweeteners (AS) affects clinical measures of diabetes control and underlying mechanisms that, in turn, may affect clinical phenotypes (5,6,8). The vast majority of AS intake in the U.S. diet occurs via artificially sweetened beverages (ASBs), defined as zero-calorie (or near zero) beverages sweetened with Food and Drug Administration–approved artificial sweeteners. People with diabetes are the highest consumers of ASBs, tending to consume them as a replacement for dietary sources of sugar, especially in place of sugar-sweetened beverages (8–12). This is the intended use of ASBs and has been endorsed by clinical and scientific organizations, whereas water consumption is considered the ideal standard in dietary recommendations (6,8). The clinical and public health concern is that there are few data to support or refute the benefit or harm of habitual ASB consumption by people with T2D, and the related evidence in people without diabetes suggests ASBs may affect T2D risk and glycemic measures in a negative manner (13–15).
To address this gap and directly inform standard-of-care guidelines and a specific medical nutritional therapy question for people with T2D, we conducted a randomized parallel trial to test the effect of substituting plain water (the ideal standard) for habitual ASB intake in people with T2D on primary measures of diabetes control. Based on the evidence and dietary recommendations, we hypothesized that consuming water in place of ASBs would reduce HbA1c and other glycemia-related measures in people with T2D.
Research Design and Methods
The Study of Drinks with Artificial Sweeteners in People With Type 2 Diabetes (SODAS) was a two-site, two-arm, parallel trial with a run-in period of 2 weeks and an active intervention period of 24 weeks that randomized 181 adults with T2D (HbA1c 6.5–8.5%), aged 35 years or older, who were usual consumers of commercial ASBs to receive and consume 24 oz daily of either 1) a commercial ASB of choice; or 2) an unflavored, sparkling or still, bottled or canned water of choice in place of ASBs.
Participants
The trial was conducted at the University of California, Irvine, and the University of Minnesota, Minneapolis, between June 2019 and March 2023. The study included adults with T2D who were able to provide informed consent and who met the criteria described in Supplementary Methods.
The trial was performed according to the principles of the Declaration of Helsinki, in accordance with the Medical Research Involving Human Subjects Act and the standards of Good Clinical Practice. The institutional review board of the University of California, Irvine, approved the protocol and amendments on 24 January 2019. The study was registered at ClinicalTrials.gov (identifier NCT03944616) prior to initiation; the protocol and statistical analysis plan are published there. Informed written consent was obtained from all participants before enrollment. There was no participant or public involvement in the design, conduct, and reporting of the trial.
Experimental Protocol
Eligible and interested individuals were invited to enroll in a 2-week run-in period before randomization to strengthen and increase the level of scientific inference for the trial. During the run-in period, participants provided dietary intake data through two, unannounced 24-h dietary recalls; wore both a blinded continuous glucose monitor (CGM) (Abbott Freestyle Libre) and a blinded objective physical activity and sleep monitor (activPAL) to assess both usual glycemic excursions essential to study aims and usual activity and sleep habits; and provided a gut microbiome sample for storage for future analyses. Participants who successfully completed the run-in period and were interested in participating in the full intervention were randomly assigned in a 1:1 ratio to the control group (continue habitual intake of commercial ASBs, 24 oz/day) or the intervention group (unflavored, sparkling or still, bottled or canned water of choice in place of ASBs, 24 oz/day). Randomization was performed with the use of permuted blocks and was stratified by site and sex. The randomization list was created before the start of the study and maintained by an independent statistician during the conduct of the study. The randomization list was not accessible to any other research team members.
Intervention
The intervention was the substitution of the ideal beverage per standard of care and dietary guidelines (i.e., unflavored, still or sparkling, bottled water) for ASBs. Participants were randomized to receive three daily servings (8-oz each, 24-oz total/day) of an ASB or water (unflavored, sparkling or still, bottled or canned water) of choice during the 24-week intervention period. The participants were instructed to consume the 24 oz of beverages provided each day during the intervention period, though they were allowed to consume higher volumes of ASBs if they were randomized to that arm. All participants were allowed to consume water ad hoc. Participants randomized to ASBs had their choice of commercial ASBs and whether the beverages were caffeinated. Participants randomized to water were instructed to avoid intake of ASBs and were provided with bottled or canned water to enhance adherence. If they had previously consumed caffeinated ASBs for reasons related to caffeine ingestion (e.g., alertness), they were counseled about sources of caffeine that are noncaloric and allowed to be consumed per study protocol (e.g., unsweetened teas and coffee), and/or provided caffeinated water or caffeine pills, whatever their preference, to enable optimal participation fidelity and provide them with similar levels of caffeine intake.
The intervention dose of 24-oz/day study beverages was selected because this amount represents the upper portion of usual intake by habitual consumers of ASBs (8,10–12,16). Furthermore, this level of intake aligns with the best methods for study design of interventions investigating the effects of diet where it is necessary to have a moderately large variation of intake by participants to detect an effect within a designated period (17).
The protocols for each arm were identical, except for the intervention beverage (water versus ASB). There was no other dietary counsel or advice provided and no specific instructions on when the beverages were to be consumed, because it was presumed this consideration was immaterial in relation to the primary outcome over 24 weeks, and the intervention was designed to maximize adherence. All beverages for were provided to study participants via home delivery. Details on adherence are reported in Supplementary Methods.
Outcomes
Blood-based Outcomes
Primary and secondary outcome measures were collected under fasting conditions at baseline, 12 weeks, and 24 weeks in the morning after an overnight (>10 h) fast. Clinical personnel performing measurements were blinded to intervention allocation, as were the laboratory personnel performing the specimen analyses. All analyses were performed in batch at the Advanced Research and Diagnostic Laboratory, a fully accredited central biochemistry laboratory at the University of Minnesota.
CGM and Weight Outcomes
Participants wore a blinded Freestyle Libre Pro sensor (Abbott, Chicago, IL), for three, 14-day periods (run-in, intervention weeks 11 and 12, and intervention weeks 23 and 24). At the beginning of each defined period, a sensor was placed on the underside of the participant’s upper arm by the clinical research coordinator. The sensor records glucose levels every 15 min and the following metrics were calculated and reported: 1) time in range (TIR) (the amount of time glucose readings were within a glucose range of 70–180 mg/dL [3.9–10.0 mmol/L]); this is a primary CGM metric (a difference of ≥3% [absolute percentage points] in TIR is considered clinically meaningful for a treatment group difference in mean TIR) (18). 2) time below range (the percentage of time glucose readings were <70 mg/dL [<3.9 mmol/L]); 3) time above range (the percentage of time glucose readings were >180 mg/dL [10.0 mmol/L]); 4) mean glucose (mean glucose concentration during the wear period); and 5) the SD of mean glucose values (a measure of dynamic glucose variability). Body weight was measured to the nearest 0.1 kg with a daily calibrated digital scale when participants were in a fasting state. Participants wore a gown or light clothing and no shoes.
Medication Effect Score
At each study visit, from the run-in period to the completion of the study, participants reported their prescribed pharmacotherapy regimen for glycemia. The medication effect score (MES) of a particular drug dose reflects the decrease in HbA1c that is expected when that specific drug dose is used as monotherapy and is a valid and appropriate measure of diabetes regimen intensity in clinical settings and research (19) (details in Supplementary Methods). The MES was an analytic covariate and a predefined secondary end point representing diabetes regimen intensity.
Statistical Analysis
The primary outcome measure was the between-group difference in HbA1c change over the 24-week intervention. Secondary outcomes were also analyzed comparing change over time between the two arms. The power estimates were based on detecting a clinically significant 0.3% difference in HbA1c percentage units using a superiority framework for comparisons. We assumed the SD of HbA1c to be 1.0%. We set power to 80% and the two-tailed α-level to 0.05 and planned for a repeated measures regression model (0, 12, and 24 weeks) testing the between-arm intervention effect on HbA1c change over 24 weeks (treatment × time interaction), with the baseline to 24-week HbA1c correlation estimated at r = 0.8. These assumptions provided an estimated sample size requirement of 160 participants completing the study. We designed the study for an anticipated participant attrition rate of 20% and aimed to randomize 100 participants to each intervention (n = 200 total) to achieve ≥80% power or higher for the primary aim. Overall, we randomized 181 participants, and 179 participants provided complete data on all measures.
The main analyses for all primary and secondary outcomes were conducted on an intention-to-treat basis, using all randomized participants. The data were analyzed using linear mixed models, with randomization group, categorical time points, and the interaction between randomization group and time, to estimate the difference between mean outcome levels in the intervention and control arms over the 24-week intervention, adjusted for sex, study site, and MES. The analysis included the 0-, 12-, and 24-week measurements available for all the participants and included random effects for individuals to account for intraclass correlation between measurements from the same person at different follow-up times. We carried out per-protocol analyses, along with prespecified sensitivity analyses that excluded participants receiving insulin therapy, stratified by sex, and further adjusted for caffeine intake (estimated via five additional, unannounced 24-h dietary recalls completed over 24 weeks), duration of T2D, type of sweeteners in ASBs consumed, and COVID-19 infection during the intervention. The statistician conducting the analyses was independent of the study and blinded to the intervention allocation. All statistical analyses were performed with SAS 9.4 (SAS Institute, Cary, NC). Statistical significance was set at P < 0.05, using two-tailed tests.
Data and Resource Availability
After publication of the main project’s aims, and secondary projects using the data are complete, the data will be available through request from the corresponding author. Statistical code is available upon request.
Results
Participant Characteristics
Between June 2019 and March 2023, 1,298 people were screened. Of these, 941 people did not meet all inclusion criteria, and 176 declined to participate. As seen in Fig. 1, the remaining 181 participants with T2D were randomized to one of the two conditions: ASB (n = 91) or water (n = 90). The two groups were similar with respect to baseline characteristics (Table 1); the mean age (SD) was 60.1 (10.2) years; just over half (52%) were women; and the mean HbA1c (SD) was 7.19% (0.65). There were no adverse events or other unintended effects of trial interventions or trial conduct on any participant-reported outcomes. There was complete fidelity of the intervention (i.e., each participant received the home delivery of beverages aligned with the intervention arm to which they were randomized). And of the 91 participants randomized to the ASB group, there was complete self-reported adherence by participants (participants reported consuming 94.5% of provided ASBs on average), and all 91 participants provided complete data. Of the 90 participants randomized to drink water, 83 reported being >90% adherent over the 24-week intervention, and 88 participants provided full data.
Figure 1.
SODAS trial flow diagram.
Table 1.
Baseline participant characteristics in SODAS
| Characteristic | Total (N = 181) | ASB (n = 91) | Water (n = 90) |
|---|---|---|---|
| State | |||
| California | 90 | 46 | 44 |
| Minnesota | 91 | 45 | 46 |
| Sex | |||
| Female | 94 | 47 | 47 |
| Male | 87 | 44 | 43 |
| Age (years) | 60.1 (10.2) | 59.6 (10.4) | 60.7(10.1) |
| Race | |||
| African American or Black | 6 | 4 | 2 |
| Hispanic/Latino/Latina | 20 | 9 | 11 |
| White (non-Hispanic) | 135 | 68 | 67 |
| Other (non-Hispanic)* | 20 | 10 | 10 |
| Education | |||
| College or above | 117 | 59 | 58 |
| High school/some college | 64 | 32 | 32 |
| Duration of diabetes (years) | 11.9 (8.3) | 10.1 (7.4) | 13.7 (8.7) |
| HbA1c % | 7.19 (0.9) | 7.19 (1.1) | 7.20 (0.7) |
| MES | 1.77 (1.28) | 1.60 (1.06) | 1.97 (1.45) |
| Uses insulin (n) | 54 | 25 | 29 |
| Modified therapeutic intensity score (blood pressure)† | 0.39 (0.47) | 0.39 (0.51) | 0.39 (0.43) |
| Lipid (statin) therapy intensity‡ | |||
| None | 49 | 30 | 19 |
| Low | 9 | 1 | 8 |
| Moderate | 81 | 37 | 44 |
| High | 42 | 23 | 19 |
| BMI (kg/m2) | 34.3 (7.0) | 35.3 (7.1) | 33.2 (6.7) |
| Usual ASB daily intake (oz/day) | 22.5 (17.6) | 21.4 (15.6) | 23.6 (19.5) |
| Duration (years) of ASB intake | 26.8 (14.6) | 26.7 (15.3) | 27.0 (14.0) |
| Usual commercial ASB composition | |||
| Aspartame/Ace-K | 138 | 69 | 69 |
| Aspartame/Ace-K/sucralose | 35 | 17 | 18 |
| Stevia | 4 | 3 | 1 |
| Sucralose | 4 | 2 | 2 |
Data are reported as mean (SD) unless otherwise indicated. Ace-K, acesulfame potassium.
*Other (non-Hispanic) includes South and Southeast Asian, East Asian (Korean, Chinese, Japanese), Native Hawaiian/Pacific Islander, American Indian or Alaskan Native, or identified with multiple racial or ethnic groups. Because of small numbers, these groups are presented as a collated group to avoid potentially identifying participants.
†The modified therapeutic intensity score is a validated summary measure that accounts for the number of antihypertensive medications in a participant’s regimen and the dose of each medication the participant is receiving relative to the Food and Drug Administration maximally recommended dose modified for the 2017 American College of Cardiology/American Heart Association blood pressure guideline–defined maximum dose for each agent.
‡2018 American Heart Association/American College of Cardiology Guideline on the Management of Blood Cholesterol classification of therapeutic intensity (20).
Outcomes
Primary Outcome
From baseline to 24 weeks, the mean HbA1c decreased from 7.19% to 7.14% in the ASB arm and increased from 7.20% to 7.44% in the water arm, for a mean difference in change of 0.29% (SE 0.12; P = 0.013) higher in the water arm (Table 2).
Table 2.
Summary statistics and effect estimates over 24 weeks of substituting water for ASBs for primary and secondary clinical glycemic and weight outcomes in people with T2D
| ASB (n = 91) | Water (n = 90) | Mean difference in change from baseline-24 week (water − ASB) | |||||||
|---|---|---|---|---|---|---|---|---|---|
| n | Mean | SD | n | Mean | SD | Estimate* | SE | P value | |
| Primary outcome | |||||||||
| HbA1c (%) | |||||||||
| Baseline | 91 | 7.19 | 1.10 | 90 | 7.20 | 0.69 | |||
| 12 weeks | 86 | 7.13 | 1.02 | 84 | 7.43 | 1.02 | |||
| 24 weeks | 91 | 7.14 | 1.19 | 87 | 7.44 | 1.04 | 0.29 | 0.12 | 0.013 |
| Secondary outcome | |||||||||
| Fructosamine (μmol/L) | |||||||||
| Baseline | 90 | 273.2 | 44.8 | 88 | 285.5 | 42.4 | |||
| 12 weeks | 85 | 272.1 | 43.5 | 82 | 286.6 | 47.2 | |||
| 24 weeks | 90 | 273.0 | 51.3 | 88 | 291.6 | 45.8 | 6.22 | 4.92 | 0.21 |
| Fasting glucose (mg/dL) | |||||||||
| Baseline | 90 | 147.5 | 40.0 | 88 | 147.6 | 36.2 | |||
| 12 weeks | 85 | 146.2 | 41.8 | 83 | 149.3 | 44.2 | |||
| 24 weeks | 90 | 149.2 | 44.0 | 87 | 153.7 | 51.3 | 4.71 | 6.43 | 0.47 |
| Fasting insulin (pmol/L) | |||||||||
| Baseline | 89 | 128.0 | 97.1 | 88 | 138.5 | 134.1 | |||
| 12 weeks | 84 | 135.7 | 141.0 | 81 | 134.2 | 145.0 | |||
| 24 weeks | 89 | 143.5 | 145.4 | 86 | 144.2 | 144.2 | −11.8 | 13.4 | 0.38 |
| Weight (kg) | |||||||||
| Baseline | 91 | 102.1 | 21.1 | 90 | 96.1 | 21.5 | |||
| 12 weeks | 91 | 101.5 | 20.4 | 88 | 95.8 | 21.8 | |||
| 24 weeks | 90 | 101.2 | 20.6 | 88 | 95.9 | 21.9 | 1.11 | 0.55 | 0.045 |
Summary statistics are unadjusted means and SD.
*Model estimate adjusted for sex, study site, and MES.
Secondary Glycemic and Weight Outcomes
The intervention effects on secondary glycemic and weight outcomes are also listed in Table 2. From baseline to 24 weeks, the mean fructosamine level (μmol/L) did not change in the ASB arm and increased in the water arm, with a mean difference in change of 6.22 μmol/L higher in the water arm (SE 4.92; P = 0.21). The mean fasting glucose level increased from 147.5 to 149.2 mg/dL in the ASB arm and increased in the water arm from 147.6 to 153.7 mg/dL, resulting in a mean difference in change of 4.71 mg/dL higher in the water arm (SE 6.43; P = 0.47). The mean fasting insulin levels had a mean difference in change of 11.8 pmol/L higher in the ASB arm compared with the water arm (SE 13.4; P = 0.38). Mean weight (kg) was stable in the water arm, whereas there was a modest decrease in the ASB arm; the mean difference in change estimated in the model was 1.11 kg higher in the water arm (SE 0.55; P = 0.045).
To provide context for the primary and main secondary end points in the trial, we also report the formal main analysis of the MES score (predefined secondary end point for change in diabetes regimen intensity). Over 24 weeks, the MES score for participants randomized to ASBs stayed stable at 1.60, and the MES score for participants randomized to the water arm did not change materially (1.97 at baseline to 1.92 at week 24). In the analytically adjusted model, there was a −0.06 lower MES after 24 weeks in those randomized to the water arm compared with those randomized to the ASB arm (P = 0.32).
CGM Outcomes
The primary CGM metric was percent TIR, and this decreased in both arms over the 24 weeks. However, there was a 3.6% greater reduction (SE 3.0; P = 0.29) in the water arm compared with the ASB arm over time. Time above range and mean glucose levels increased in both arms over time, but the change was greater, though nonsignificant, in the water arm compared with the ASB arm (Table 3). Glycemic variability also increased over the course of the intervention, with higher change in the SD of the mean glucose measurements in the water arm than in the ASB arm (2.6; SE 1.7; P = 0.13).
Table 3.
Summary statistics and effect estimates over 24 weeks of substituting water for ASBs regarding CGM metrics in people with T2D
| Metric | ASB (n = 91) | Water (n = 90) | Mean difference in change from baseline to 24 weeks (water − ASB) | ||||||
|---|---|---|---|---|---|---|---|---|---|
| n | Mean | SD | n | Mean | SD | Estimate* | SE | P value | |
| Time in range (%) 70–180 mg/dL (3.9–10.0 mmol/L) | |||||||||
| Baseline | 88 | 76 | 19.2 | 87 | 72.3 | 20.9 | |||
| 12 weeks | 84 | 75.2 | 22.6 | 86 | 66.9 | 26.2 | |||
| 24 weeks | 87 | 72.8 | 23.2 | 83 | 64.7 | 26.5 | −3.6 | 3 | 0.29 |
| Time below range (%) <70 mg/dL (<3.9 mmol/L) | |||||||||
| Baseline | 88 | 3 | 7.5 | 87 | 2.3 | 4.9 | |||
| 12 weeks | 84 | 2.7 | 5.1 | 86 | 2.8 | 5.3 | |||
| 24 weeks | 87 | 2.7 | 5.1 | 83 | 3.6 | 10.3 | 1.5 | 1.3 | 0.26 |
| Time above range (%) >180 mg/dL (>10.0 mmol/L) | |||||||||
| Baseline | 88 | 21.4 | 19.5 | 87 | 25.2 | 21.5 | |||
| 12 weeks | 84 | 22.8 | 24.3 | 86 | 30.2 | 27.1 | |||
| 24 weeks | 87 | 24.7 | 25.1 | 83 | 31.7 | 28 | 2.8 | 3.4 | 0.42 |
| Mean glucose (mg/dL)† | |||||||||
| Baseline | 88 | 147.3 | 42.1 | 87 | 151.4 | 33.1 | |||
| 12 weeks | 84 | 149.4 | 46.4 | 86 | 161.2 | 50.5 | |||
| 24 weeks | 87 | 153.1 | 50.7 | 83 | 162.3 | 47.3 | 4.6 | 5.4 | 0.39 |
| SD of mean glucose (mg/dL)‡ | |||||||||
| Baseline | 88 | 39.6 | 14.7 | 87 | 42 | 13.1 | |||
| 12 weeks | 84 | 40 | 12.9 | 86 | 44.9 | 15.4 | |||
| 24 weeks | 87 | 40.8 | 14.9 | 83 | 45.8 | 16.3 | 2.6 | 1.7 | 0.13 |
Summary statistics are unadjusted means and SD.
*Model adjusted for sex, study site, and MES.
†Mean glucose refers to a measure of the mean 24-h glucose concentration calculated across all recorded glucose readings during the wear period.
‡The SD of mean glucose values is a measure of dynamic glucose variability; SD is strongly correlated with mean glucose.
Per-Protocol and Sensitivity Analyses
The results from per-protocol and sensitivity analyses were all consistent with the main results (Supplementary Table 1). Specifically, adjustment for duration of type 2 diabetes, caffeine intake, class of usual artificial sweetener intake via ASBs, COVID-19 infection, and exclusion of insulin use did not materially alter the findings for the primary outcome. Sex-stratified analyses were directionally consistent, with a greater magnitude of effect in female participants.
Conclusions
This randomized dietary intervention in people with T2D tested the underlying hypothesis informing the general guidance from standard of care and medical nutritional therapy guidelines related to ASB intake in people with T2D. The findings suggest substituting water for ASBs by people with T2D who regularly consume ASBs does not improve any clinical measures related to glycemia or weight over 24 weeks. In fact, the participants randomized to consume water in place of habitual ASB intake had a clinically and statistically significant increase in HbA1c over 24 weeks compared with participants who maintained habitual intake of ASBs, contrary to the hypothesis. Related measures from CGM metrics provide context for underlying glycemic dynamics during the intervention: percent TIR declined at a clinically significant effect size (>3%) (but a nonstatistically significant level) over the course of the intervention in the water arm relative to the ASB arm, and glycemic variability and time above range were increased as well. There was also a statistically significant difference in mean level of weight change in the water arm compared with the ASB arm, but the level is not considered clinically significant (∼1%) (21). The primary outcome findings were consistent in per-protocol and a priori sensitivity analyses, suggesting that the findings were robust to different statistical and analytic assumptions.
There is little evidence to compare with the SODAS results. In a study of 81 adult women (age <50 years) with obesity and T2D (HbA1c <7.2%), only using metformin for diabetes control, and who were usual consumers of ASBs, all participants were part of a 24-week behavioral weight loss intervention, and half were randomized to consume water in place of ASB with lunch. In the secondary glycemic outcomes, there was clinically nonsignificant improvement in HbA1c in the water arm compared with the ASB arm, and statistically significant improvements in fasting glucose and insulin levels and indices of insulin resistance in the water arm compared with the ASB arm (22). Similar results were reported in a 12-week randomized intervention of young adults with overweight and obesity who did not have T2D and who were usual consumers of ASBs. Participants substituting water for ASB had improved glycemic and metabolic measures relative to maintaining ASB intake (23). Although these two studies both focused on regular consumers of ASB, one study included participants at high risk for T2D (but did not have T2D), and the other study included participants with a specific T2D pharmaceutical regimen integrated into a weight loss study, limiting any direct comparisons to the SODAS.
The SODAS was a rigorously conducted, randomized, equicaloric dietary intervention with design and conduct elements that strengthen the inference of the results, despite the results being contrary to the hypothesis and not supporting current medical nutritional therapy and dietary recommendations for people with T2D. The pragmatic design used commercial ASBs consumed, because they are in the real world, to enhance adherence and generalizability of the intervention. Indeed, participants reported ideal levels of adherence to the intervention, but there is no objective measure of adherence to report. ASB intake is a self-selected behavior and consumers tend to have little variation in the beverages they habitually consume (24). Furthermore, commercial ASBs sweetened with one uniform AS compose a negligible proportion of intake (16). Although the actual amount of AS is proprietary information, we characterized individual AS blends for secondary analyses to inform the interpretation of the main effects. Adjustment for the AS blend or type did not alter the primary outcome results, and the study was not designed or powered to detect an effect in individual sweeteners or blends, limiting inference beyond general ASB intake. The supporting data collected over the course of the study (i.e., gut microbiome and biological samples for metabolomics, as well as detailed dietary, physical activity, sleep, and quality of life data) will allow us to examine potential hypothesized biological and behavioral avenues that may contribute to explaining the reported results in the future.
Overall, the results of the SODAs has implications for clinical care and medical nutritional therapy guidelines for people with T2D who are like the study population. Specifically, there was no benefit to substituting water for ASBs by people with T2D who regularly consume ASBs. If anything, the results suggest maintaining usual ASB intake may be a tool to continue to help manage T2D if glycemic measures are controlled and stable. However, replication of these results should be attempted to determine the robustness of the findings, which are contrary to other related data informing the guidelines and the hypothesis. Nonetheless, these results are generalizable to people with T2D and, given the uneven body of evidence on ASB intake, randomized interventions of similar design are merited in different populations to inform scientific, clinical, and public health recommendations on ASB intake.
This article contains supplementary material online at https://doi.org/10.2337/figshare.30559481.
Article Information
The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Duality of Interest. No potential conflicts of interest relevant to this article were reported
Author Contributions. A.O.O. conceived, designed, and supervised the study; wrote the first draft of the manuscript, and critically revised the manuscript. J.C. and L.J. did the analyses and critically revised the manuscript and study documents. S.Ra., S.Ry., and N.R.M. coordinated and conducted the study and data management and critically revised the manuscript. A.E.B., E.S., and A.R. contributed to clinical aspects of design, clinical patient oversight during trial conduct, and critically revised the manuscript. M.A.P. contributed to study design, supervised a study site, critically revised the manuscript and study documents, and interpreted results. All authors approved the final version. A.O.O. attests that all listed authors meet authorship criteria and that no others meeting the criteria have been omitted. A.O.O. and L.J. are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Funding Statement
This work was supported by the National Institutes of Health National Institute of Diabetes and Digestive and Kidney Diseases (grant R01DK117028).
Footnotes
Clinical trial reg. no. NCT03944616, clinicaltrials.gov
Supporting information
References
- 1. Centers for Disease Control and Prevention . National Diabetes Statistics Report, 2024. Accessed 12 August 2024. Available from https://www.cdc.gov/diabetes/php/data-research/index.html.2024
- 2. Sun H, Saeedi P, Karuranga S, et al. IDF Diabetes Atlas: global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045. Diabetes Res Clin Pract 2022;183:109119. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. American Diabetes Association Professional Practice Committee . 1. Improving care and promoting health in populations: Standards of Care in Diabetes—2024. Diabetes Care 2023;47(Suppl. 1):S11–S19 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. American Diabetes Association Professional Practice Committee . 6. Glycemic goals and hypoglycemia: Standards of Care in Diabetes—2024. Diabetes Care 2023;47(Suppl. 1):S111–S125 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Evert AB, Boucher JL, Cypress M, et al. Nutrition therapy recommendations for the management of adults with diabetes. Diabetes Care 2014;37(Suppl. 1):S120–S143 [DOI] [PubMed] [Google Scholar]
- 6. American Diabetes Association Professional Practice Committee. 5. Facilitating positive health behaviors and well-being to improve health outcomes: Standards of Care in Diabetes—2024. Diabetes Care 2023;47(Suppl. 1):S77–S110 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Evert AB, Dennison M, Gardner CD, et al. Nutrition therapy for adults with diabetes or prediabetes: a consensus report. Diabetes Care 2019;42:731–754 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Johnson RK, Lichtenstein AH, Anderson CAM, et al.; American Heart Association Nutrition Committee of the Council on Lifestyle and Cardiometabolic Health; Council on Cardiovascular and Stroke Nursing; Council on Clinical Cardiology; Council on Quality of Care and Outcomes Research; and Stroke Council . Low-calorie sweetened beverages and cardiometabolic health: a science advisory from the American Heart Association. Circulation 2018;138:e126–e140 [DOI] [PubMed] [Google Scholar]
- 9. Dunford EK, Miles DR, Ng SW, Popkin B. Types and amounts of nonnutritive sweeteners purchased by US households: a comparison of 2002 and 2018 Nielsen Homescan purchases. J Acad Nutr Diet 2020;120:1662–1671.e10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Sylvetsky AC, Jin Y, Clark EJ, Welsh JA, Rother KI, Talegawkar SA. Consumption of low-calorie sweeteners among children and adults in the United States. J Acad Nutr Diet 2017;117:441–448.e2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Mackenzie T, Brooks B, O’Connor G. Beverage intake, diabetes, and glucose control of adults in America. Ann Epidemiol 2006;16:688–691 [DOI] [PubMed] [Google Scholar]
- 12. Lin L, Zhu F, Delp EJ, Eicher-Miller HA. Differences in dietary intake exist among U.S. adults by diabetic status using NHANES 2009–2016. Nutrients 2022;14:3284. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Khan TA, Lee JJ, Ayoub-Charette S, et al. WHO guideline on the use of non-sugar sweeteners: a need for reconsideration. Eur J Clin Nutr 2023;77:1009–1013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Lee JJ, Khan TA, McGlynn N, et al. Relation of change or substitution of low- and no-calorie sweetened beverages with cardiometabolic outcomes: a systematic review and meta-analysis of prospective cohort studies. Diabetes Care 2022;45:1917–1930 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. McGlynn ND, Khan TA, Wang L, et al. Association of low- and no-calorie sweetened beverages as a replacement for sugar-sweetened beverages with body weight and cardiometabolic risk: a systematic review and meta-analysis. JAMA Netw Open 2022;5:e222092. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Gardner C, Wylie-Rosett J, Gidding SS, et al.; American Diabetes Association . Nonnutritive sweeteners: current use and health perspectives: a scientific statement from the American Heart Association and the American Diabetes Association. Diabetes Care 2012;35:1798–1808 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Lichtenstein AH, Petersen K, Barger K, et al. Perspective: design and conduct of human nutrition randomized controlled trials. Adv Nutr 2021;12:4–20 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Battelino T, Alexander CM, Amiel SA, et al. Continuous glucose monitoring and metrics for clinical trials: an international consensus statement. Lancet Diabetes Endocrinol 2023;11:42–57 [DOI] [PubMed] [Google Scholar]
- 19. Alexopoulos A-S, Yancy WS, Edelman D, et al. Clinical associations of an updated medication effect score for measuring diabetes treatment intensity. Chronic Illn 2021;17:451–462 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Grundy SM, Stone NJ, Bailey AL, et al. . 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol 2019;73:3168–3209 [DOI] [PubMed] [Google Scholar]
- 21. Horn DB, Almandoz JP, Look M. What is clinically relevant weight loss for your patients and how can it be achieved? A narrative review. Postgrad Med 2022;134:359–375 [DOI] [PubMed] [Google Scholar]
- 22. Madjd A, Taylor MA, Delavari A, Malekzadeh R, Macdonald IA, Farshchi HR. Beneficial effects of replacing diet beverages with water on type 2 diabetic obese women following a hypo-energetic diet: a randomized, 24-week clinical trial. Diabetes Obes Metab 2017;19:125–132 [DOI] [PubMed] [Google Scholar]
- 23. Viveros-Watty PE, López-Franco O, Zepeda RC, et al. Effects on cardiometabolic risk factors after reduction of artificially sweetened beverage consumption in overweight subjects. A randomised controlled trial. Endocrinol Diabetes Nutr. 2022;69:168–177 [DOI] [PubMed] [Google Scholar]
- 24. Bleich SN, Wolfson JA, Vine S, Wang YC. Diet-beverage consumption and caloric intake among US adults, overall and by body weight. Am J Public Health 2014;104:e72–e78 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.