This study presents the development and validation of the SMARTCLOTH-Database, a structured food group system designed to support dietary management for individuals with diabetes. The database organises foods into 20 nutritionally coherent groups, distinguishing between raw and cooked forms, based on Spanish dietary patterns and statistical criteria. Nutritional values were calculated using weighted averages derived from national food consumption data and multiple international food composition databases. To validate the system, a computational tool was developed to generate daily menus automatically and compare nutritional outputs against a gold standard reference (BDCA). The comparison showed small mean differences (less than ±22 kcal, < 2 g carbohydrates, < 0.6 g protein, and <1.5 g lipids), with high concordance demonstrated through intraclass correlation coefficients (ICC ≥ 0.75) for carbohydrates and proteins across most analyses. These results indicate that the SMARTCLOTHDatabase provides accurate and reliable estimates of macronutrient content, making it suitable for integration into digital tools aimed at improving dietary self-management in people with diabetes.
García-Rodríguez et al. (Thu,) studied this question.