The study, by Sadanand Karun, Lotus McDougal and Abhishek Singh, examined health expectancy across India and 22 states and union territories using data representing conditions around 2018. At the national level, life expectancy at age 60 was estimated at 17.4 years for men and 19.2 years for women, giving women a female–male survival gap of 1.8 years. But life expectancy alone cannot show whether those extra years are lived in good health. To address that problem, the researchers divided expected remaining life into disability-free life expectancy, or DFLE, and life expectancy with disability, or DLE.

The distinction is technically important. Life expectancy is calculated from age-specific mortality rates and describes how long a population can expect to live under observed death risks. DFLE adds a measure of functional health, estimating the number of years expected to be lived without disability. DLE estimates the remaining years lived with disability. Together, the two measures add up to total life expectancy, but they reveal how survival is distributed between healthier and less healthy states. “A longer life” therefore becomes a more meaningful public-health measure only when researchers ask how many of those years people can expect to live independently.

To generate the estimates, the researchers combined two major Indian data sources. Mortality information came from the Sample Registration System, which provides age-specific death rates through a dual-record process: independent records are cross-checked, and discrepancies are investigated through household revisits. The study averaged mortality rates from 2016 through 2020 to represent 2018. Disability information came from the first wave of the Longitudinal Ageing Study in India, or LASI, a large survey of adults aged 45 and older and their spouses. The analysis included 31,902 people for the national estimates and 24,947 people in the selected states.

Disability was defined using questions based on the Katz scale of independence in activities of daily living. Participants were considered disabled if they reported difficulty lasting at least three months with any one of six basic activities: dressing, walking across a room, bathing, eating, getting in or out of bed, or using the toilet. These tasks capture fundamental self-care and mobility functions rather than isolated symptoms or short-term illness. The investigators constructed abridged life tables for five-year age bands from 60–64 through 85 and older, then used Sullivan’s method to combine survival with age-specific disability prevalence. In simplified terms, the method multiplies the person-years lived in each age interval by the proportion of people observed to be disability-free or disabled.

The results showed that disability prevalence was higher among women than men nationally and in every state included in the analysis. At the national level, the sex gap in DFLE was almost zero—just 0.03 years—meaning that women and men could expect roughly similar absolute amounts of disability-free life from age 60. The contrast was in DLE: women were expected to live 1.77 more years with disability than men. In other words, women’s longer survival largely appeared in the disabled portion of life rather than as a clear advantage in healthy years.

The pattern varied sharply across India. Women’s life-expectancy advantage was greatest in Jammu and Kashmir, where the gap reached 5.06 years, followed by Delhi at 4.78 years, Rajasthan at 4.01 years, Kerala at 3.95 years and Himachal Pradesh at 3.77 years. In Jharkhand, however, the gap was negative by 1.1 years, while Bihar also showed a male advantage in life expectancy. Disability differences were particularly pronounced in West Bengal, where the proportion of older women reporting disability was 13.5 percentage points higher than among men. The smallest disability gap occurred in Rajasthan. These variations demonstrate why national averages can conceal the very different ageing experiences produced by state-level differences in poverty, healthcare, disease burden, living conditions and social structure.

The researchers used two approaches to test whether the paradox was present. First, they compared the share of total life expectancy spent with disability. Nationally, women spent a larger proportion of their remaining life after age 60 in a disabled state than men. The same was true in every assessed state except Odisha. In West Bengal, for example, men spent 34.4 percent of their remaining life in disability compared with 47.8 percent for women. The second approach focused only on women’s additional years of life generated by their mortality advantage. Nationally, 98.4 percent of those additional years were spent with disability. The proportion exceeded 80 percent in Delhi, Kerala, Tamil Nadu, Himachal Pradesh and Uttar Pradesh, and reached 92.9 percent in Uttar Pradesh. In Rajasthan, by contrast, only 9.2 percent of the additional years were spent with disability, while the figure was 16.3 percent in Odisha, 40.2 percent in Haryana and 48.4 percent in Karnataka.

The study’s decomposition analysis helps explain how the gap is produced. Using a stepwise replacement method, the researchers separately estimated the contributions of mortality and disability. The mortality component captures the effect of different survival probabilities, while the disability component captures differences in the prevalence of functional limitations. In most states, both effects increased women’s years lived with disability: women survived longer, but they also had higher disability rates. The disability effect dominated in states including Assam, Maharashtra, West Bengal, Punjab, Jharkhand, Kerala, Himachal Pradesh, Uttar Pradesh, Telangana, Gujarat, Tamil Nadu, Uttarakhand and Chhattisgarh. The mortality effect was larger in Jammu and Kashmir, Delhi, Rajasthan, Odisha, Haryana, Karnataka and Madhya Pradesh. When the researchers examined five-year age groups, mortality contributed more strongly to the DFLE gap at younger older ages, but its contribution generally declined as age increased, while disability effects became more important.

Several biological and social mechanisms could contribute to this pattern, although the study was not designed to establish causes. Women’s biological survival advantage has been linked in previous research to sex differences in immune function and the presence of two X chromosomes. Men, meanwhile, experience higher mortality from conditions and behaviours associated with cardiovascular disease, lung cancer, chronic kidney disease, injuries, alcohol use, smoking and other risks. Women may survive these and other health threats but later develop disabling chronic conditions. After menopause, hormonal changes may increase vulnerability to cardiovascular disease, hypertension, bone disease, osteoarthritis and multiple chronic conditions. These illnesses can be disabling without being immediately fatal, creating precisely the combination of longer survival and poorer functional health captured by the paradox.

Social exposures may be equally significant. Women in India are more likely to spend long periods performing household work, including cooking with polluting fuels. Indoor air pollution has been associated with poorer self-rated health, cognitive function and chronic disease, and its effects may be amplified among older women who have experienced decades of exposure. Women may also face disadvantages in education, income, nutrition, access to medical care and control over household resources, all of which can shape whether chronic disease becomes disabling. At the same time, men may underreport illness because of social expectations that discourage acknowledging vulnerability. If women are more willing to disclose functional problems and men are more likely to conceal them, some of the observed gap could reflect reporting behaviour as well as underlying health differences.

The analysis has important limitations. Mortality data were available at the sub-national level for only 22 of India’s states and union territories, and the researchers could not calculate standard errors or confidence intervals because the raw Sample Registration System data and sample sizes were not publicly available in the necessary form. Disability measures were self-reported and may therefore reflect sex differences in reporting. The study also relied on the first LASI wave, preventing analysis of trends over time. In some states, the oldest age groups contained very few participants, making estimates for ages 80–84 and 85 and above less stable. Comparisons with official life tables suggested that the estimates were broadly robust nationally and in most states, but caution was warranted for Jammu and Kashmir, Himachal Pradesh and Delhi, where differences between the study’s calculations and official estimates were larger.

The findings challenge the use of life expectancy as a standalone measure of progress. A population can gain years of life without gaining an equivalent number of years of independence, mobility or self-care. For India’s rapidly ageing population, the researchers argue that health policy should track DFLE and DLE alongside mortality, and should address disability among women directly. The priorities may need to differ by state: Maharashtra, Assam, West Bengal, Punjab and Telangana showed especially severe disability disadvantages, with women’s years lived with disability exceeding their total mortality advantage. In Delhi, Kerala, Tamil Nadu, Himachal Pradesh and Uttar Pradesh, more than 80 percent of women’s survival advantage was spent with disability. Improving chronic-disease prevention, rehabilitation, accessible healthcare, indoor air quality and social support could help ensure that the years women gain through longer survival are more often lived in good health rather than added to the burden of disability.