Long-term remission of type 2 diabetes after very-low-calorie restriction and related predictors.

Purpose: Very low-calorie restriction (VLCR) can induce remission of type 2 diabetes mellitus (T2DM), but its long-term remission and related predictors have not been clarified. The aim of present study is to investigate the effect of VLCR in inducing long-term T2DM remission, and the underlying predictors. Methods: A total of 61 participants with T2DM who received 9 days of VLCR from Dec 2012 to Oct 2020 were followed up in Nov 2021, and divided into responders and non-responders groups. Responders were defined as HbA1c < 6.5% over at least 3 months in the absence of pharmacotherapy. Clinical characteristics were compared between responders and non-responders. Potential predictors were examined by logistic regression analyses based on clinical data before and after VLCR. Results: Forty-four participants were successfully followed up, including 19 males and 25 females. Long-term remission was observed in 17 participants (38.64%) after VLCR, with a median 7.83 years. Compared with non-responders, responders had a shorter disease duration, a lower fasting blood glucose (FBG) level, a higher fasting insulin level, and better HOMA-β after VLCR. Besides, acute insulin response (AIR), insulin area under curve in intravenous and oral glucose tolerance test (IVGTT-IAUC and OGTT-IAUC) in responders were higher than those in non-responders after VLCR. Multivariable logistic analysis showed that higher post-VLCR IVGTT-IAUC predicted a longer T2DM remission. Conclusions: After VLCR, more than one third of the participants presented remission over up to 8 years. The improvement of β-cell function, especially the restoration of first-phase insulin-secreting capacity, could prolongate the remission.