Which diet is best for type 2 diabetes?

Summary

- An umbrella review was published in November 2021. A meta-analysis pools together all studies on a particular topic. An umbrella review pools together all meta-analyses and systematic reviews, so it really is a definitive review of a topic.

- This review sought to find out which type of diet is best to achieve and maintain weight loss in people with type 2 diabetes (T2D). It found 19 meta-analyses of relevance.

- The review concluded that the greatest weight loss was reported with very low-energy diets, (intake of approximately 400–500 calories per day) and that low-carbohydrate diets were no better for weight loss than higher-carbohydrate/low-fat diets.

- There were several flaws:

1) As often happens in reviews of low-carbohydrate diets, low-carbohydrate diets were not studied. Of the 9 meta-analyses (out of 19) that focused on so-called low-carbohydrate diets, only 2 examined very low-carbohydrate diets (<50/g carbohydrate per day). Others considered up to 45% carbohydrate intake as low carbohydrate.

2) The conclusion that very low-energy diets did best was based on one meta-analysis of very low-energy diets, which demonstrated good results. The nine not-low-carb diets were pooled together for the low-carb finding. Results were lowered in this pooling together, as the average was dragged down by poorer results (in the higher carb meta-analyses).

3) Each meta-analysis compared one diet with something else. The very low-calorie diet interventions (400-500 calories per day) were compared with smaller calorie reduction diets (1,000-1,500 calories per day). Of course, the very low-calorie diet would do better. Comparing, for example, 39% carb with 41% carb is not likely to achieve a difference. The research question – which diet is best? – and the presentation of findings in the umbrella review more than implied that low calorie is better than low carbohydrate, but low calorie wasn’t compared directly with low carbohydrate. Very low calorie was compared with low calorie and not so low carbohydrate was compared with a little bit less low carbohydrate.

- The main results were based on diets of approximately 12 weeks. There were no differences in types of diet undertaken for studies that continued for a year or more.

- There is less evidence for remission of T2D than for weight loss in people with T2D. The DiRECT study was heavily relied upon for findings that very low calorie is best for remission. The review noted the importance of weight loss in achieving remission. The review also observed that remission is easier the shorter the duration of someone’s T2D.

- Very low-calorie diets will do well in the short term (a few weeks). The issue is whether these are healthy, achievable and sustainable in the long term.