Effects of nutrition education using a food-based approach, carbohydrate counting or routine care in type 1 diabetes: 12 months prospective randomized trial

INTRODUCTION: Evidence on the effects of structured nutrition education is weak in adults with type 1 diabetes mellitus (T1D) with moderately impaired glycemic control. Objective was to compare the effects of different types of nutrition education programs on glycemic control, cardiovascular risk factors, quality of life, diet quality and food choices in T1D. RESEARCH DESIGN AND METHODS: A 12 months randomized controlled study conducted at nine diabetes specialist centers with three parallel arms: (i) a food-based approach (FBA) including foods with low glycemic index or (ii) carbohydrate counting (CC) according to today’s standard practice or (iii) individual sessions according to routine care (RC). The primary end point was difference in glycated hemoglobin A1c (HbA1c) between groups at 12 months. RESULTS: 159 patients were randomized (FBA: 51; CC: 52; RC: 55). Mean (SD) age 48.6 (12.0) years, 57.9% females and mean (SD) HbA1c level 63.9 (7.9) mmol/mol, 8% (0.7%). After 3 months, HbA1c improved in both FBA and CC compared with RC. However, there were no significant differences at 12 months in HbA1c; FBA versus RC (−0.4 mmol/mol (1.3), 0.04% (0.1%)), CC versus RC (−0.8 mmol/mol (1.2), 0.1% (0.1%)), FBA versus CC (0.4 mmol/mol (0.3), 0.04% (0.01%)). At 12 months, intake of legumes, nuts and vegetables was improved in FBA versus CC and RC. FBA also reported higher intake of monounsaturated and polyunsaturated fats compared with RC, and dietary fiber, monounsaturated and polyunsaturated fats compared with CC (all p values <0.05). There were no differences in blood pressure levels, lipids, body weight or quality of life. CONCLUSIONS: Nutrition education using an FBA, CC or RC is equivalent in terms of HbA1c and cardiovascular risk factors in persons with T1D with moderately impaired glycemic control. An FBA had benefits regarding food choices compared with CC and RC.