Geographic information system mapping and predictors of glycemic control in children and youth with type 1 diabetes: a study from Western India.

Abstract

Objectives: Geographic Information System (GIS) mapping, is a novel way to provide insights into spatial distribution of type 1 diabetes (T1D) and associations between T1D outcomes and potential predictors. We aimed to explore GIS in children with T1D, and identify predictors of poor glycemic control.

Methods: Design: Cross-sectional; Participants: 402 children and youth (187 boys) with T1D. Place of residence (coordinates) of participants were geocoded in GIS. They were divided into two groups living in urban or peri-urban areas using ArcGIS Pro. The characteristics of urban/peri-urban living were linked to sociodemographic and biochemical data and spatial autocorrelation analysis was performed. Association between glycemic control and distance to our unit was studied.

Results: Mean age was 13.2 ± 4.7 years; 196 children were living in urban areas, 206 in peri-urban areas. There was significant difference in HbA1c between groups (Urban 9.9 (9.7, 10.2) %, Peri-urban 10.5 (10.1, 10.8) %) (p=0.004); mean difference 0.5 (0.1, 1.0) with poorer glycemic control and higher prevalence of vitamin D sufficiency in peri-urban and higher prevalence of hypothyroidism in urban areas. There was significant correlation between glycemic control (HbA1c) and distance to our unit r=0.108 (0.023, 0.218) (p=0.031). Individuals with an HbA1c ≥9.5 were residing farther away (58.9 (49.4, 68.5) km) as compared to those with HbA1c <9.5 (44.5 (35.1, 53.9) km) (p<0.05).

Conclusions: Children with T1D when grouped using GIS had differences in glycemic control and comorbidities; peri-urban participants and those residing further away from our unit had poorer glycemic control. Future efforts may be aimed at identifying centers and channelizing resources towards children showing poor glycemic control, thus optimizing disease management.