Adnan Agha, Sozan Yasser Rakha, Bachar Afandi, Nishan Sudheera Kalupahana, Juma M Alkaabi
The mean AdultCarbQuiz score was 26.2 ± 4.0 out of 43 (60.9%). Domain analysis revealed strong carbohydrate food recognition (domain mean 80.1% correct) but poor quantification skills (20-27.5% correct for portion estimation) and meal-level counting (2.5-15% correct). Diabetes distress was present in 37.5% of the participants, and impaired hypoglycaemia awareness was present in 20.0% of the participants. Knowledge of carbohydrate counting showed a moderate inverse correlation with diabetes distress (r = -0.468, 95% CI -0.68 to -0.18, p = 0.002), which strengthened further after adjusting for HbA1c and Gold score (partial r = -0.579, 95% CI -0.77 to -0.30, p < 0.001). In exploratory analyses, younger age (r = -0.404, p = 0.010) and shorter diabetes duration (r = -0.344, p = 0.030) were associated with higher knowledge scores. No association was observed between the knowledge scores and HbA1c levels (r = 0.179, p = 0.295).
INTRODUCTION: Carbohydrate counting is central to dietary self-management in patients with type 1 diabetes mellitus (T1DM). However, objective assessments of carbohydrate counting knowledge and its relationship with psychosocial outcomes are scarce in Gulf populations. This study assessed carbohydrate counting knowledge, diabetes distress, and hypoglycaemia awareness in patients with T1DM attending a tertiary diabetes centre in the United Arab Emirates.
METHODS: This was a cross-sectional study of 40 patients with T1DM (age ≥ 14 years, diabetes duration ≥ 12 months) attending the Tawam Hospital Diabetes Clinic, Al Ain, UAE. Carbohydrate counting knowledge was assessed using the Arabic version of the 43-item AdultCarbQuiz questionnaire. Diabetes distress was measured using the two-item Diabetes Distress Scale (DDS-2), and hypoglycaemia awareness was measured using the Gold score. Pearson's correlations, one-way analysis of variance (ANOVA) across knowledge quartiles, and partial correlations adjusted for HbA1c and Gold score were performed.
RESULTS: The mean AdultCarbQuiz score was 26.2 ± 4.0 out of 43 (60.9%). Domain analysis revealed strong carbohydrate food recognition (domain mean 80.1% correct) but poor quantification skills (20-27.5% correct for portion estimation) and meal-level counting (2.5-15% correct). Diabetes distress was present in 37.5% of the participants, and impaired hypoglycaemia awareness was present in 20.0% of the participants. Knowledge of carbohydrate counting showed a moderate inverse correlation with diabetes distress (r = -0.468, 95% CI -0.68 to -0.18, p = 0.002), which strengthened further after adjusting for HbA1c and Gold score (partial r = -0.579, 95% CI -0.77 to -0.30, p < 0.001). In exploratory analyses, younger age (r = -0.404, p = 0.010) and shorter diabetes duration (r = -0.344, p = 0.030) were associated with higher knowledge scores. No association was observed between the knowledge scores and HbA1c levels (r = 0.179, p = 0.295).
DISCUSSION: In this cross-sectional sample, carbohydrate counting knowledge was independently and inversely associated with diabetes distress, regardless of glycaemic control. Specific knowledge gaps in carbohydrate quantification and meal-level counting are actionable targets for educational intervention in this population. These findings highlight the value of addressing both nutritional knowledge and emotional well-being in T1DM management; whether improving carbohydrate-counting knowledge reduces distress requires longitudinal and interventional study.