Carbohydrate Counting as a Low-Cost Strategy Associated With Better Glycemic Control in Children With Type 1 Diabetes Mellitus in Resource-Constrained North India: A Cross-Sectional Mediation Analysis.

BACKGROUND: Children with type 1 diabetes mellitus (T1DM) in India face a dual burden of disease-related stress and socioeconomic constraint, with insulin pumps and continuous glucose monitors (CGMs) unaffordable for most families. Carbohydrate (carb) counting is a low-cost, education-based self-management skill linked to improved glycemic control in other populations. We hypothesized that carb counting is associated with better quality of life (QoL) in children with T1DM primarily through improved glycemic control, and tested this pathway using a cross-sectional mediation analysis at a tertiary center in North India. METHODOLOGY: In this hospital-based, cross-sectional study, 50 children with T1DM (age 2-18 years) attending the pediatric endocrinology outpatient clinic of a tertiary care hospital in Dehradun, Uttarakhand, were enrolled. QoL was assessed using the Quality-of-Life Instrument for Indian Diabetes patients (QOLID; 34 items across eight domains). Glycated hemoglobin (HbA1c), self-monitored blood glucose, insulin regimen, and carb-counting practice were recorded. Associations were analyzed using independent-samples t-tests, Pearson correlation, and simple linear regression, with sensitivity analyses (Welch's t-test, Mann-Whitney U), exploratory multivariable linear regression, and Baron-Kenny mediation analysis (Sobel test and bootstrap confidence intervals) used to test whether HbA1c mediates the carb-counting-QoL association. p<0.05 was considered significant. RESULTS: The mean QOLID score was 82.08±9.06%; 19 (38%) children had impaired QoL (score <80%). Financial worries were the lowest-scoring domain (69.50±17.00), impaired in 44 (88%) children, despite 47 (94%) being on basal-bolus insulin therapy and only one (2%) child using an insulin pump. HbA1c showed the strongest negative correlation with total QoL (r=-0.675, p<0.001). Thirty (60%) children practiced carb counting; they had significantly lower HbA1c (7.27±1.02% vs 8.96±2.50%, p=0.0017) and higher total QOLID scores (148.37±12.46 vs 139.40±18.43, p=0.045) than those who did not. Mediation analysis showed that HbA1c fully mediated the carb counting-QoL association: the indirect effect via HbA1c was significant (a×b=9.26, Sobel z=2.87, p=0.0041; bootstrap 95% CI: 2.74-16.82), while the direct effect of carb counting on QoL was negligible and non-significant after accounting for HbA1c (c'=-0.29, p=0.938), with an estimated 103% of the total effect statistically mediated through glycemic control. CONCLUSIONS: Carb counting was associated with better QoL primarily, and statistically fully, through its association with better glycemic control rather than through an independent pathway -- a mechanistically coherent, testable explanation for how a free, education-based skill could plausibly improve QoL in settings where insulin pumps and CGMs remain unaffordable. Financial worry remained the most impaired QoL domain despite near-universal use of low-cost insulin regimens, underscoring the broader economic burden of care. As this mediation pattern is statistical rather than temporally confirmed, prospective or interventional studies are needed before structured carb-counting education can be recommended as an evidence-based strategy in resource-constrained pediatric diabetes care.