Remission of type 2 diabetes can be achieved in short and long-lasting type 2 diabetes: results of a randomized controlled trial in individuals with overweight and obesity

  1. Bettina Schuppelius
  2. Elena Lalama
  3. Jiudan Zhang
  4. Kilian Rüther
  5. Marta Csanalosi
  6. Stefan Kabisch
  7. Anette Christ
  8. Eicke Latz
  9. Olga Pivovarova-Ramich
  10. Knut Mai
  11. Andreas F. H. Pfeiffer

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Background and Aim

A growing body of evidence shows that upon extensive weight loss returning to non-diabetic glucose control is possible in people with recent onset type 2 diabetes (T2D). However, the impact of diabetes duration and different intervention strategies on remission is not clear. Thus, we investigated the remission of T2D in response to three months of two very low-calorie diets (VLCDs) with different macronutrient profiles in individuals with short and long diabetes duration.

Methods

Participants with a BMI >27 kg/m 2 and T2D duration of ≤4 years or ≥8 years were studied before and after following a VLCD strategy (600-800 kcal/day) for three months including discontinuation of antidiabetic medication. Individuals were stratified by diabetes duration and randomly assigned to one of two VLCDs with slightly different macronutrient composition. Phenotyping included mixed meal tolerance test (MMTT), metabolic characterization and assessment of body composition.

Results

Fifty-two (30 women, 22 men) participants were enrolled between September 2020 and November 2022 into the trial and 47 participants completed the intervention. Remission of T2D, defined as plasma fasting glucose levels <126 mg/dl, was achieved in 34 participants (72%). Despite similar weight loss of subjects with a diabetes duration ≤4 years and ≥8 years (-15.2 ± 5.8 kg vs. -13.9 ± 4.8 kg; p = 0.473), subjects with diabetes duration ≥8 years had a 32% lower remission rate (82% vs. 50%; p = 0.027). Remission rates also differed between the two formula diets and were found to be higher with the high-fiber, high-protein, and low-carb, low-fat formula diet (91% vs. 56%; p = 0.008). In addition, individuals that achieved a remission had significantly lower fasting plasma glucose and higher C-peptide levels at baseline.

Conclusions

Our findings show that significant body weight loss through VLCDs can induce T2D remission in nearly three-quarters of participants. Fasting plasma glucose and C-peptide levels, diabetes duration, and used macronutrient profile emerged as important factors for the achievement of diabetes remission, although a considerable remission is still possible after a long duration of T2D.

Article activity feed

  1. PREreview

    This Zenodo record is a permanently preserved version of a PREreview. You can view the complete PREreview at https://prereview.org/reviews/16575046.

    Summary:

    This study aims to determine whether the previously reported remission rates of type 2 diabetes mellitus (T2DM) induced by a very low-calorie diet (VLCD) are influenced by the time elapsed since T2DM diagnosis and the macronutrient composition of the VLCD. This is a pertinent and clinically significant research question, underlining the relevance and timeliness of the investigation. The authors conducted a clinical trial combining an observational case-control design (comparing patients with less than 4 and more than 8 years since the T2DM diagnosis) and a randomized controlled trial (to assess the effects of two formula diets with slightly different macronutrient profiles). 

    The main findings were that (i) diabetes remission was achieved in half the patients with long-term T2DM, which significantly adds to the state of the art, and that (ii) different rates of remission were observed in response to the two VLCD, suggesting that macronutrient composition might play a role in the response to this treatment. Noteworthily, the authors found these results despite having recruited less participants than their sample size target. The latter limited the ability of the study to further explore the data in search of predictors, although they identified some (baseline C-peptide and glucose concentration, body weight and the change in postprandial insulin and fasting triglycerides). Overall, we find the study to be both compelling and clinically relevant. In light of its findings, we offer a series of constructive comments and suggestions aimed at strengthening the manuscript.

    Major Comments:

    One of the main limitations of this study lies in the fact that the diabetes duration groups were not matched for diabetes severity, as acknowledged by the authors. While this may reflect a typical real-world scenario, it nonetheless limits the ability of the study to isolate the effect of T2DM diabetes duration on remission outcomes. It is plausible that glycemic status (reflected by fasting glucose/HbA1c levels) rather than duration per se, primarily influences remission rates. In our view, the authors should further address this limitation when discussing the results. Moreover, the two groups appear to differ substantially in baseline body weight. Although the p-value is only marginally significant, we believe this issue should be further highlighted along the manuscript. 

    Another important point concerns the long-term sustainability of the intervention. Given the likelihood of weight regain following the discontinuation of a VLCD, the potential impact on long-term remission and metabolic outcomes should be acknowledged. 

    The study also contributes to an important gap in the literature regarding the role of VLCD macronutrient composition in T2DM remission. This study's results offer valuable insights into that. However, the two dietary interventions employed were nutritionally quite similar, limiting the applicability of the findings. The manuscript would benefit from further explaining the rationale behind these diets' selection, as well as a comprehensive description of their nutritional composition within the manuscript, even though the formulas might be available online. Additionally, due to the close overlap in macronutrient profiles and the inclusion of only two intervention arms, the analyses presented in Figure 4 may be misleading.

    The conclusions are solidly grounded in the presented data. However, we would recommend revisiting the current title, as it might be somewhat misleading. The main conclusion, stated in the title, does not arise directly from the RCT design. Removing the RCT reference from the title would likely be enough to avoid potential misinterpretation. 

    Overall, the manuscript is outstandingly written. However, it could be slightly improved by: (i) clarifying whether the study employed an open-label design; (ii) providing additional detail regarding the removal of outliers (lines 232-234); and (iii) specifying whether the sample size calculation also accounted for comparisons between dietary interventions, as the current description appears limited to the T2DM duration groups. 

    Minor Comments:

    We would also suggest:

    1. Including percentages alongside the mean and standard deviation on lines 304-305.

    2. Presenting AUC values shown in Table 2 in clinically meaningful units. Specifically, we suggest expressing them per minute, dividing the presented results by 180. 

    3. Incorporating the T2DM duration groups into the graphs shown in Figure 2.

    4. Evaluating the effects of the two dietary interventions stratified by T2DM duration. 

    5. Briefly describing the procedures conducted at each study visit to provide a clearer overview. 

    6. Using "/" rather than "-" to separate interquartile range in Table S2, as the latter may be confused with negative values.

    7. Differentiating drop-outs by T2DM duration groups in Figure 1. 

    Furthermore, although this information is thoroughly described in the manuscript, we recommend incorporating the following key elements into the abstract: (i) macronutrient composition of the dietary interventions (or at least the commercial product names; as the current description lacks sufficient context for interpretation); (ii) number of participants allocated to and completing the trial in each study group; (iii) number of participants within each T2DM duration group; (iv) some key baseline characteristics of the study population; and (v) removal of the MMTT from the methods. 

    Finally, the authors claim that no macronutrient composition has shown superiority for weight loss in the context of VLCDs (lines 80-81). While this is true in the long-term, a previous study showed superiority of a low-CHO diet in the short-term (the length applied in this study): Davis et al. 2009 Diabetes Care.

    Competing interests

    The authors declare that they have no competing interests.

  2. Version published to 10.1101/2025.06.10.25329355 on medRxiv