Study Protocol for the Effects of Artificial Intelligence (AI)-Supported Automated Nutritional Intervention on...

Nutritional intervention is effective in improving glycemic control in patients with type 2 diabetes but requires large inputs of manpower. Recent improvements in photo analysis technology facilitated by artificial intelligence (AI) and remote communication technologies have enabled automated evaluations of nutrient intakes. AI- and mobile-supported nutritional intervention is expected to be an alternative approach to conventional in-person nutritional intervention, but with less human resources, although supporting evidence is not yet complete. The aim of this study is to test the hypothesis that AI-supported nutritional intervention is as efficacious as the in-person, face-to-face method in terms of improving glycemic control in patients with type 2 diabetes.

This is a multicenter, unblinded, parallel, randomized controlled study comparing the efficacy of AI-supported automated nutrition therapy with that of conventional human nutrition therapy in patients with type 2 diabetes. Patients with type 2 diabetes mainly controlled with diet are to be recruited and randomly assigned to AI-supported nutrition therapy (n = 50) and to human nutrition therapy (n = 50). Asken, a mobile application whose nutritional evaluation has been already validated to that by the classical method of weighted dietary records, has been specially modified for this study so that it follows the recommendations of Japan Diabetes Society (total energy restriction with proportion of carbohydrates to fat to protein of 50-60, 20, and 20-30%, respectively).

The primary outcome is the change in glycated hemoglobin levels from baseline to 12 months, and this outcome is to be compared between the two groups. The secondary outcomes are changes in fasting plasma glucose, plasma lipid profile, body weight, body mass index, waist circumference, blood pressures, and urinary albumin excretion. The results of this randomized controlled trial will fill the gap between the demand for support of AI in nutritional interventions and the scientific evidence on its efficacy.

UMIN000032231.