Abstract
Background and aims: Digital personalized nutrition is increasingly implemented in type 2 diabetes care, but the therapeutic effects appear to be heterogeneous in real-world settings. The aims of this study were to evaluate variations in the glycemic response to a digital personalized dietary intervention and identify baseline determinants of treatment effectiveness. Methods: In this prospective real-world before-after study, 47 adult outpatients with type 2 diabetes received individualized dietary recommendations generated by a web-based platform, with unchanged pharmacotherapy. Glycated hemoglobin (HbA1c), dietary intake, and body mass index were assessed at baseline and after 2-3 months. Subgroup analyses were performed according to baseline HbA1c (<7.0% vs ≥ 7.0%). Results: = 0.003), whereas no meaningful change occurred in patients with well-controlled baseline glycemia. Baseline HbA1c was associated with the magnitude of the glycemic response to the intervention. Conclusions: Digital personalized nutrition in routine clinical practice resulted in heterogeneous glycemic effects rather than a uniform population-level response. Clinically meaningful improvement was primarily observed in patients with elevated baseline HbA1c, supporting a targeted implementation strategy and highlighting the importance of patient stratification in digital dietary interventions.