Younger adults made up a striking share of COVID deaths in Manipur, records show

More than one in ten COVID-19 deaths recorded in the plains districts of Manipur, a northeastern Indian state, occurred among adults aged 18 to 44 years, according to a retrospective analysis of hospital records and Medical Certificates of Cause of Death published in Discover Public Health. Of the 1,000 verified deaths analysed, 126 (12.6%) were younger adults and 874 (87.4%) were adults over 44, a distribution that the study’s authors describe as evidence of pronounced age-related inequities shaped by social and structural factors rather than biological age alone.

The research team, led by H. Nirendrakumar Singh and colleagues, drew on officially reported mortality data from the Directorate of Health and Family Welfare, Government of Manipur. During the study period, from March 2020 to December 2021, 1,982 COVID-19 deaths were officially reported in the state, but the analysis was restricted to 1,000 deaths from five plain districts — Imphal East, Imphal West, Bishnupur, Thoubal and Kakching — chosen for data completeness, accessibility of hospital records and verification of certification information.

The dataset captured socio-demographic details, comorbidities, clinical course, laboratory parameters and indicators of hospital care, and compared younger adults (18–44 years) with older adults (over 44 years) using descriptive statistics and chi-square tests, with a multivariable logistic regression model for the socio-demographic factors.

The demographic pattern of the deaths was uneven across the state. Imphal West accounted for 431 deaths and Imphal East for 280, while Thoubal recorded 107 and Bishnupur and Kakching 91 each. Yet the proportion of deaths occurring among younger adults varied markedly by district: Kakching had the highest share at 23.1%, followed by Thoubal at 14.0%, while Imphal East had the lowest at 9.3%. The district-level association with age group was statistically significant overall (p = 0.017) and among males (p = 0.011), but not among females. The authors caution that the study was not designed to investigate the mechanisms behind these geographic differences, which could reflect variation in healthcare access, health-seeking behaviour, population structure or local transmission dynamics.

One of the most striking findings concerned chronic disease. Overall, 553 deaths (55.3%) occurred in individuals with no documented chronic condition, and younger adults accounted for 87 deaths (15.7%) among those without chronic disease compared with only 39 deaths (8.7%) among those with at least one recorded comorbidity — a difference the authors describe as statistically significant (p < 0.001). Diabetes and hypertension showed significant associations with age group, with deaths among younger adults more frequent in those without these conditions, while chronic lung disease, chronic kidney disease, chronic liver disease, stroke, heart disease and cancer did not reach statistical significance, although all recorded deaths with stroke, heart disease or cancer occurred among older adults in small numbers.

The multivariable analysis produced some unexpected results. Educational attainment was among the strongest independent predictors: adults with secondary education (10–11 years) had roughly four times the adjusted odds of belonging to the younger adult death group compared with those with no formal education (AOR = 4.01; 95% CI 1.97–8.15), and those with higher education (≥ 12 years) had 3.74 times the odds (95% CI 1.95–7.16), both statistically significant. The authors suggest this counterintuitive pattern may reflect more workplace interactions, greater occupational mobility or other unmeasured exposure-related factors rather than education itself, and they urge caution and further investigation. By contrast, the number of children was strongly inversely associated with younger adult deaths — those with two, three, or four or more children had significantly lower adjusted odds (AOR = 0.18, 0.09 and 0.04 respectively) — but the authors interpret this as a demographic life-course pattern rather than a protective effect of parenthood.

Several factors that appeared significant in descriptive comparisons lost their association after adjustment. Occupation was significantly associated with age category in the bivariate analysis (p = 0.006), with younger adult deaths more common among those not in the workforce (15.4%) and those in sales, services, household and domestic work (14.2%), and least common among professional, technical, managerial and clerical occupations (5.2%), yet none of these categories remained independently significant after adjustment, though two estimates were of borderline significance. Marital status and family income followed the same trajectory: unmarried status (56.5% of younger adult deaths) and absence of family income (16.1%) were prominent descriptively but attenuated after controlling for other characteristics. Sex was not independently associated with younger adult mortality (AOR = 0.84; 95% CI 0.52–1.33), nor was place of residence, although urban residents had non-significantly lower adjusted odds than rural residents (AOR = 0.67; 95% CI 0.42–1.05).

Laboratory findings, available only for a subset of cases, pointed to marked inflammatory and coagulation activation across all age groups. C-reactive protein levels, recorded for 55 cases, were elevated among younger adults — mean 60.1 in males and 63 in females — and substantially lower in the 45–64 year group, though these differences were not statistically significant (p = 0.62). Median oxygen saturation at admission was lower among younger adults (74 for men, 77.5 for women) than among those aged 45 and over (80 for both sexes), with a small sex difference across groups. The authors argue that elevated inflammatory and coagulation markers among younger patients suggest severe thrombotic and inflammatory processes were not limited to elderly patients, supporting early laboratory evaluation regardless of age, while stressing that these associations cannot be treated as causal in a retrospective observational study with likely non-random missing laboratory data.

Hospital course differed little between age groups once patients were admitted. ICU admission (p = 0.524), mechanical ventilation (p = 0.850), duration of hospital stay (p = 0.144) and timing of admission relative to test confirmation (p = 0.207) showed no statistically significant age-group differences. ICU care was provided to 13.2% of younger adults and 86.8% of older adults, and ventilatory support to 12.8% and 87.2% respectively, proportions the authors attribute to the overall age distribution of deaths rather than differential treatment. Deaths occurred across all durations of hospital stay, from same-day admissions to stays longer than ten days, without a significant age pattern. The authors conclude that differences in mortality patterns were more closely related to underlying patient characteristics than to variations in in-hospital treatment, and that disparities may arise largely in pre-hospital stages such as testing, referral and admission delays.

The study also documented serious deficiencies in death certification: using WHO criteria for completeness and accuracy, apart from one certificate in each age group, none were complete and the certifications were yet to be accurate, a finding with implications for the reliability of mortality surveillance. The authors list further limitations, including the retrospective design without a control group, unadjusted comparisons that may be confounded, possible under-documentation of comorbidities in hospital records, hospitalization bias that likely overrepresents individuals who accessed healthcare facilities, and the inability by design to account for confounders such as healthcare access, testing availability, referral delays and treatment-seeking behaviour. Within those constraints, the analysis — intended to inform public health surveillance and equity-oriented planning rather than establish causal relationships — identifies younger adults, including those without apparent comorbidities, as a group that may require greater attention in risk stratification and pandemic preparedness in the region.

Subject of Research: Age-related patterns in COVID-19 mortality in Manipur, India, based on hospital records and death certificates

Article Title: Evidence of age-related inequities in COVID mortality in the Indian state of Manipur using hospital records and death certificates

Article References: Singh, H. N., Nandeibam, A. R., Singh, K. J., Singh, J. P., Anand, T., Singh, K. R., & Wahengcha, Y. (2026). Evidence of age-related inequities in COVID mortality in the Indian state of Manipur using hospital records and death certificates. Discover Public Health, 23(1), Article 1720. https://doi.org/10.1186/s12982-026-02681-9

Image Credits: AI Generated

DOI: 10.1186/s12982-026-02681-9

Keywords: COVID-19 mortality, Manipur, India, younger adults, age-related inequity, death certification, comorbidities, social determinants of health, D-dimer, C-reactive protein, retrospective study, hospital records