Toggle Main Menu Toggle Search

Open Access padlockePrints

Predictors of type 2 diabetes remission in the Diabetes Remission Clinical Trial (DiRECT)

Lookup NU author(s): Alison Barnes, Dr Ahmad Al-Mrabeh, Dr Sviatlana Zhyzhneuskaya, Dr Patrick Welsh, Professor Roy Taylor



This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0).


© 2020 The Authors. Diabetic Medicine published by John Wiley & Sons Ltd on behalf of Diabetes UK. Aim: To identify predictors of type 2 diabetes remission in the intervention arm of DiRECT (Diabetes Remission Clinical Trial). Methods: Participants were aged 20–65 years, with type 2 diabetes duration of <6 years and BMI 27–45 kg/m2, and were not receiving insulin. Weight loss was initiated by total diet replacement (825–853 kcal/day, 3–5 months, shakes/soups), and weight loss maintenance support was provided for 2 years. Remissions (HbA1c <48 mmol/mol [<6.5%], without antidiabetes medications) in the intervention group (n = 149, mean age 53 years, BMI 35 kg/m2) were achieved by 68/149 participants (46%) at 12 months and by 53/149 participants (36%) at 24 months. Potential predictors were examined by logistic regression analyses, with adjustments for weight loss and effects independent of weight loss. Results: Baseline predictors of remission at 12 and 24 months included being prescribed fewer antidiabetes medications, having lower triglyceride and gamma-glutamyl transferase levels, and reporting better quality of life with less anxiety/depression. Lower baseline HbA1c was a predictor at 12 months, and older age and male sex were predictors at 24 months. Being prescribed antidepressants predicted non-remission. Some, but not all effects were explained by weight loss. Weight loss was the strongest predictor of remission at 12 months (adjusted odds ratio per kg weight loss 1.24, 95% CI 1.14, 1.34; P < 0.0001) and 24 months (adjusted odds ratio 1.23, 95% CI 1.13, 1.35; P <0.0001). Weight loss in kilograms and percentage weight loss were equally good predictors. Early weight loss and higher programme attendance predicted more remissions. Baseline BMI, fasting insulin, fasting C-peptide and diabetes duration did not predict remission. Conclusions: Other than weight loss, most predictors were modest, and not sufficient to identify subgroups for which remission was not a worthwhile target.

Publication metadata

Author(s): Thom G, Messow C-M, Leslie WS, Barnes AC, Brosnahan N, McCombie L, Al-Mrabeh A, Zhyzhneuskaya S, Welsh P, Sattar N, Taylor R, Lean MEJ

Publication type: Article

Publication status: Published

Journal: Diabetic Medicine

Year: 2021

Volume: 38

Issue: 8

Print publication date: 01/08/2021

Online publication date: 01/09/2020

Acceptance date: 25/08/2020

Date deposited: 11/01/2021

ISSN (print): 0742-3071

ISSN (electronic): 1464-5491

Publisher: John Wiley & Sons Ltd


DOI: 10.1111/dme.14395

PubMed id: 32870520


Altmetrics provided by Altmetric


Funder referenceFunder name
Diabetes UK. Grant Number: 13/0004691