India’s COVID-19 Lockdown Widened Nutrition Gaps Between Rich and Poor

India’s first nationwide COVID-19 lockdown disrupted food systems, healthcare access, and livelihoods, raising fears of a national nutrition crisis. A new cross-sectional analysis of the fifth National Family Health Survey (NFHS-5) suggests those fears were only half right: overall levels of underweight, child wasting, and anemia mostly improved or held steady after the lockdown, but the gains were distributed unevenly, concentrating among wealthier households while the poorest improved least. Near-poor men, in particular, saw a significant rise in underweight, and measures of economic inequality in undernutrition widened across the board.

The study, published in BMC Global and Public Health, exploited a rare feature of NFHS-5: its fieldwork ran in two phases, from June 2019 to January 2020 and from January 2020 to April 2021, straddling the lockdown that shut the country down between March 25 and May 31, 2020. The researchers defined a pre-lockdown period covering June 2019 through March 2020 and a post-lockdown period from June 2020 to April 2021, with no data collected during the lockdown itself. Because the survey interviewed different people before and after, the two samples represent independent cross-sections drawn from the same national source population rather than the same individuals followed over time.

The analysis drew on 636,699 households surveyed across all Indian states and union territories. Nutritional and biomarker measurements were collected in specific subsamples, yielding an unweighted pre-lockdown sample of 69,844 men aged 15 to 54, 496,427 women aged 15 to 49, and 153,886 children under five, and a post-lockdown sample of 31,995 men, 227,688 women, and 70,332 children. Underweight in adults was defined as a body mass index below 18.5 kilograms per square meter, wasting in children as a weight-for-height z-score below minus two standard deviations under World Health Organization growth standards, and anemia through hemoglobin thresholds adjusted for altitude and, among adults, smoking status.

Because the pre- and post-lockdown samples differed in geographic and sociodemographic composition, the researchers estimated covariate-adjusted prevalences by marginal standardization, adjusting for age, caste, and religion so that both periods reflect a common population composition. They then fitted survey-weighted multivariable logistic regression models, stratified by wealth quintile, to compare post-lockdown with pre-lockdown odds of each outcome, and used concentration indices and curves to quantify economic disparities. Household wealth was measured with the NFHS-5 wealth index, a composite of asset ownership, housing characteristics, and access to services, divided into five quintiles from Q1, the poorest, to Q5, the wealthiest.

The headline national numbers were broadly reassuring. Adjusted underweight prevalence among men edged up only slightly, from 14.4 percent before the lockdown to 14.6 percent after, while among women it fell from 19.0 percent to 17.5 percent. Wasting among children under five declined from 20.4 percent to 16.3 percent, and anemia prevalence decreased in every group examined: from 25.3 percent to 20.9 percent in men, 58.4 percent to 52.9 percent in women, and 69.7 percent to 63.8 percent in children under five.

Stratification by wealth told a different story. Among adult men, underweight prevalence rose in the poorest quintiles, by 1.0 percentage point in Q1 and 3.0 percentage points in Q2, while falling in the wealthier quintiles. In regression models, the odds of underweight among men increased significantly only in the second quintile, with an adjusted odds ratio of 1.25 (95 percent confidence interval 1.12 to 1.40); the increase in the poorest quintile was not statistically significant (aOR 1.06, 95 percent CI 0.95 to 1.18), and odds declined significantly in the wealthiest quintile (aOR 0.83, 95 percent CI 0.72 to 0.96). Among women, underweight was essentially unchanged in the poorest quintiles (Q1 aOR 1.01, 95 percent CI 0.95 to 1.05) but declined significantly in the third through fifth quintiles and overall (aOR 0.90, 95 percent CI 0.88 to 0.92).

Children showed declines in wasting across all wealth quintiles, but the reductions were smaller among the poorest: minus 3.2 percentage points in both Q1 and Q2, compared with minus 4.3 and minus 5.4 percentage points in Q4 and Q5. In the models, wasting odds fell in every quintile, from an aOR of 0.83 (95 percent CI 0.77 to 0.88) in Q1 and 0.81 (95 percent CI 0.76 to 0.88) in Q2 to 0.66 (95 percent CI 0.57 to 0.76) in Q5. Anemia declined across all wealth quintiles and populations, though the smallest reductions appeared in the richest quintile, and anemia, like underweight and wasting, remained most concentrated in the poorest households throughout.

The equity analysis confirmed the widening gap. All concentration indices were negative, meaning adverse outcomes were disproportionately concentrated among poorer quintiles, and movement farther from zero signals growing inequity. The index for underweight shifted from minus 0.13 to minus 0.16 in men and from minus 0.13 to minus 0.17 in women, while the index for child wasting fell from minus 0.06 to minus 0.10. Trends for anemia were mixed and smaller in magnitude: the index moved from minus 0.09 to minus 0.10 among men, closer to zero among women, from minus 0.04 to minus 0.02, and slightly farther from zero among children, from minus 0.022 to minus 0.024. Concentration curves showed greater deviation from the line of equity after the lockdown, particularly for underweight and wasting.

A sensitivity analysis restricted to the 13 states and union territories contributing data to both periods yielded broadly consistent results. The rise in underweight odds among Q2 men was preserved (restricted aOR 1.26, 95 percent CI 1.09 to 1.45), and anemia declines remained consistent across populations and quintiles. However, wasting improvements among children in Q1 through Q3 were no longer statistically significant in the restricted analysis, with Q1 shifting from a primary aOR of 0.83 to a restricted aOR of 0.93 (95 percent CI 0.86 to 1.02). The authors attribute this attenuation partly to differences in geographic composition, noting that the 13 overlapping states are not representative of the national distribution, and they caution that wasting improvements among children in lower wealth quintiles warrant careful interpretation.

The authors are explicit that the cross-sectional design precludes causal inference and that other changes over the survey period, including state-level policies affecting food access, may have shaped the observed patterns. They also flag a sobering possibility: mortality related to lockdown hardship and COVID-19 infection may have fallen disproportionately on economically vulnerable people, so nutritionally compromised individuals in poorer quintiles could be underrepresented in the post-lockdown sample. That selective survival would bias undernutrition prevalence downward among the poorest groups, meaning the reported widening of disparities may be conservative and the true equity impact greater. Given India’s population size, the authors note that even small shifts translate into enormous absolute numbers, including hundreds of thousands more wasted children and over ten million additional underweight adults.

Why would near-poor men fare worst? The authors hypothesize that households in Q2 may fall above Below Poverty Line eligibility thresholds and therefore missed expanded food grain allocations under the Pradhan Mantri Garib Kalyan Anna Yojana during the lockdown, though they note this hypothesis was not consistent across all outcomes and requires further investigation. Improvements concentrated in wealthier quintiles may reflect a greater capacity to buffer the economic shock and lesser exposure to lockdown disruptions, consistent with evidence that poorer Indian households bore a disproportionate economic burden, but the researchers stress this cannot be tested directly with these data. They also point to a 2020 multi-country survey across 37 nations in which nine out of ten households reported losing more than half their income during the pandemic and eight in ten faced difficulty purchasing food, and to disruptions of school feeding programs and community-based nutrition services as likely contributors.

Anemia’s across-the-board decline, the authors suggest, may partly reflect reduced exposure to intestinal parasites during school closures and hygiene measures, as well as the ongoing Anemia Mukt Bharat initiative, though hemoglobin-based estimates from WHO cutoffs and field capillary measurements may overestimate prevalence in India and cannot distinguish nutritional anemia from other causes. Because nutritional anemia develops gradually, the immediate post-lockdown window may not capture the full impact; longer-term data from Karnataka show anemia prevalence doubling in rural children aged 6 to 12 one year after the lockdown. The authors argue their findings point toward durable safety nets, fortified food access, and uninterrupted nutrition services rather than stopgap relief, citing Kerala’s universal food baskets combining public distribution, community kitchens, and targeted supplementary nutrition as a model of equity-focused response during large-scale shocks.

Subject of Research: Economic disparities in underweight, wasting, and anemia in India before and after the first COVID-19 lockdown

Article Title: Inequitable nutritional impact of India’s first COVID-19 lockdown: a national cross-sectional analysis

Article References: Marri, A. R., Dauphinais, M. R., Martinez, L., Karoly, M., McQuaid, F., & Sinha, P. (2026). Inequitable nutritional impact of India’s first COVID-19 lockdown: a national cross-sectional analysis. BMC Global and Public Health, 4(1), Article 95. https://doi.org/10.1186/s44263-026-00326-2

Image Credits: AI Generated

DOI: 10.1186/s44263-026-00326-2

Keywords: India, COVID-19 lockdown, NFHS-5, nutrition, underweight, child wasting, anemia, wealth inequality, health equity, cross-sectional study, food security, public health