AHA Scientific Sessions 2026 Abstract · Nov 2, 2026
A Coexisting Rise in Obesity, Hyperglycemia, and Thinness with an Apparent Decline in Hypertension: A Paired State-Level Analysis of 70 Million-Plus Indian Adults Across Two Waves of the National Family Health Survey
By Sethi Y, Banach M, Misra A
Background: Low- and middle-income countries are thought to progress through a nutrition transition from undernutrition toward obesity. India's NFHS-6, 2023-24 offers the first opportunity to test whether this is a clean substitution or a simultaneous accumulation of both burdens, and to scrutinize a paradoxical signal—falling hypertension amid rising adiposity and dysglycemia. Methods: We analyzed adult cardiometabolic and nutritional indicators from the published NFHS-6 fact sheets for India and 35 States/Union Territories (Manipur not released), comparing them with NFHS-5 (2019-21). Indicators were thinness (BMI <18.5 kg/m ² ) and overweight/obesity (BMI ≥25, ages 15-49), and elevated random glucose (>140 mg/dl or on treatment) and hypertension (≥140/90 mmHg or on treatment, ages ≥15). Each State/UT served as its own control; changes were tested with paired t, Wilcoxon signed-rank, and sign tests (Holm-adjusted), with Cohen's d_z for effect size. Results: Overweight/obesity rose in 33/35 States/UTs for women (mean +5.8 percentage points [pp]; d_z=1.38; p=1.5х10 ⁻ ⁹ ) and 28/35 for men (+2.9 pp; p=0.004); elevated glucose rose in 30/35 women (+3.6 pp; d_z=1.28) and 29/35 men (+4.0 pp; d_z=1.01; both p<10 ⁻ ⁶ ). Contrary to transition theory, thinness did not fall—men's thinness rose +2.2 pp (p=0.003). Within states, change in thinness and change in obesity were inversely correlated in men (r=–0.46; p=0.005), and both rose together in ~20/35 States/UTs, indicating a distribution widening at both tails rather than a population-wide shift. Glucose burden also diverged geographically (between-state SD 4.4→6.3 in men; p=0.002). Hypertension declined modestly (women –1.2 pp; men –1.6 pp) but did not survive multiplicity correction; the decline was statistically identical in IHCI and non-IHCI states (p≥0.63), and on decomposition was confined to the near-threshold "mildly elevated" band (women –2.5 pp; p<0.001) with no change in severe elevation (p=0.56)—whereas glucose, measured at the same examination, worsened in severity (very-high share 0.49→0.56; p<0.001). Conclusions: Across Indian States/UTs, obesity, hyperglycemia, and adult thinness are rising concurrently, identifying a true double burden of malnutrition that is intensifying and geographically diverging rather than resolving. The apparent improvement in hypertension is not corroborated by its own severity structure or by the program-targeting pattern and is best interpreted as a threshold-concentrated measurement/ascertainment effect plus untargeted treatment, not a genuine vascular gain. These findings argue against framing India's transition as a move from undernutrition to obesity and for regionally differentiated cardiometabolic prevention.