Association between Body Mass Index and Glycated Haemoglobin (HbA1c) in Newly Diagnosed Type 2 Diabetes Mellitus: A Cross-Sectional Analysis from a Tertiary Care Centre in Kanpur, India

Abstract

Background: Obesity is regarded as the dominant modifiable driver of type 2 diabetes mellitus (T2DM), and lifetime-risk estimates from Western cohorts rise steeply across the body mass index (BMI) spectrum. Whether the severity of hyperglycaemia at the moment of diagnosis tracks with BMI in an Indian population — a population characterised by thin-fat phenotype, early beta-cell failure and diabetes occurring at lower BMI — is far less certain.

Objectives: To describe the BMI distribution and the glycated haemoglobin (HbA1c) distribution of newly diagnosed T2DM patients attending the out-patient department of a tertiary care teaching hospital in Kanpur, India, and to test whether HbA1c category is associated with BMI class.

Methods: This was a cross-sectional analysis of the baseline demographic and laboratory dataset of 410 newly diagnosed, treatment-naïve adults with T2DM enrolled in a randomised controlled trial (CTRI/2024/10/074542) between October 2024 and September 2025. Height, weight and BMI were recorded using standardised anthropometry, and BMI was classified into five categories. HbA1c was measured on venous blood and grouped into five glycaemic bands (<6.5%, 6.5–7.4%, 7.5–8.9%, 9.0–9.9%, ≥10.0%). Continuous variables are summarised as median and inter-quartile range (IQR), with class-specific medians, IQRs and observed ranges displayed graphically; the association between BMI class and HbA1c category was tested using Pearson's chi-square test of independence, with p < 0.05 considered statistically significant.

Results: The cohort comprised 410 newly diagnosed patients. HbA1c at diagnosis was high across the whole cohort (median 8.1%, IQR 7.0–9.3%), and 15.4% of patients presented with HbA1c ≥10.0%. BMI was distributed across all five classes, with a substantial proportion of patients in the normal-weight (33.7%) and underweight (5.9%) categories in addition to overweight (29.8%) and obese (30.7%) categories. HbA1c category showed no statistically significant association with BMI class (χ² = 14.3028, degrees of freedom = 16, p = 0.576164; Cramér's V = 0.093). Median HbA1c spanned less than one percentage point across the five BMI classes (8.0–8.9%), the inter-quartile ranges overlapped almost completely, and BMI and HbA1c were essentially uncorrelated on their continuous scales (Spearman ρ = −0.002, p = 0.97).

Conclusion: In this North Indian tertiary-care population, newly diagnosed T2DM occurred — and presented with comparable glycaemic severity — across the entire BMI spectrum. BMI alone did not discriminate patients with severe from those with mild hyperglycaemia at diagnosis. These findings support ethnicity-specific screening strategies that do not rely on a BMI threshold as a gate to diabetes testing, and they reinforce the argument that body-fat distribution and beta-cell reserve, rather than BMI, drive the Indian T2DM phenotype.