The research, led by Anoop Jain of Boston University School of Public Health together with Rockli Kim and S. V. Subramanian of Harvard University, analyzed data from the Longitudinal Aging Study in India, or LASI, a nationally representative survey covering all 36 states and Union Territories except Sikkim. From more than 72,000 respondents, the team focused on 63,564 women and men above the age of 45 who had complete information on the exposures and outcomes of interest. These individuals lived in 2,439 wards and villages, giving the researchers an unusually rich window into how conditions vary not just between people but between the communities they inhabit.

The methodological core of the study is its multilevel design. Rather than treating each respondent as an isolated data point, the researchers built a three-level statistical model in which individuals were nested within households and communities, and communities within states. The exposure of interest was the modeled proportion of people in each community living in households without any toilet, estimated using a Markov Chain Monte Carlo approach to produce an unconfounded community-level measure. Communities were then sorted into quintiles of toilet coverage, ranging from the lowest quintile, where fewer than 62 percent of residents had a toilet, to the highest, where more than 99 percent did.

Depression was measured with four widely used instruments. The first was the ten-item Center for Epidemiological Studies-Depression scale, known as CES-D, which captures symptoms such as feeling depressed, loneliness, low energy, and trouble concentrating over the past week; a score of four or above classified a respondent as having depressive symptoms. The remaining three measures came from the Composite International Diagnostic Interview Short Form, or CIDI-SF, a validated diagnostic screening tool translated into sixteen Indian regional languages. Respondents were classified as having probable major depression using three thresholds: the standard CIDI-SF cutoff, plus two stricter definitions requiring three or five of seven additional symptoms. The mean prevalence of CES-D symptoms in the sample was 28.1 percent, while probable major depression ranged from 7.1 to 8.9 percent depending on the definition used.

The results were striking. After adjusting for age, sex, marital status, education, caste, wealth, place of residence, and even the type of toilet in the respondent’s own household, people living in communities with the lowest toilet coverage had 38 percent higher odds of depressive symptoms on the CES-D scale than those in communities with the highest coverage, with an odds ratio of 1.38 and a 95 percent confidence interval of 1.08 to 1.76. The associations were even stronger for probable major depression: odds ratios of 1.58 for the standard CIDI-SF measure, 1.56 for the three-symptom definition, and 1.72 for the five-symptom definition, all statistically significant. Notably, the highest-coverage communities showed odds ratios close to one, suggesting no elevated risk relative to the reference group.

Perhaps the most provocative finding is what the study did not find. When the researchers examined the respondents’ own household toilet status, the associations with depression were weak and statistically indistinguishable from no effect. The odds ratios for living in a no-toilet household were 1.02 for CES-D and 0.99 for CIDI-SF. In other words, the community context appeared to matter more than the private one. This is a novel result, and the authors propose a plausible mechanism: in communities where nearly everyone lacks a toilet, residents have nowhere else to go, unable to borrow a neighbor’s facility or reach a public toilet block. In better-off communities, households without toilets may still access shared or public options, buffering them from the worst consequences of deprivation.

The cross-level interaction analysis added further nuance. Although none of the formal interaction terms reached statistical significance, the raw prevalence of depressive symptoms rose consistently across all household toilet types as community coverage fell. Among people with twin pit toilets, CES-D prevalence climbed from 12.6 percent in the highest-coverage communities to 27.0 percent in the lowest. Among those with pit latrines, CIDI-SF prevalence rose from 6.5 to 10.5 percent, and even among households with flush toilets, the three-symptom depression measure increased from 6.1 to 9.5 percent. The pattern suggests that a private toilet may offer less protection when the surrounding environment remains visibly degraded.

Why would community sanitation leave such a deep imprint on mental health? The authors point to several candidate pathways rooted in social science theory. One is spatial stigma: the idea that people living in visibly deprived places are marked by where they live, eroding their sense of self-worth, a phenomenon previously linked to depression in Japan and in Mumbai’s legally excluded settlements. Another draws on the broken windows hypothesis, in which visible signs of disorder, here open fecal contamination, signal that a place is neglected and uncontrolled, a perception consistently associated with psychological distress. A third pathway involves social division: sanitation programs such as Community-Led Total Sanitation have explicitly encouraged the use of shame to end open defecation, potentially fracturing community cohesion. The absence of sanitation is also associated with elevated risk of non-partner sexual violence, itself a driver of adverse mental health outcomes.

The study has honest limitations. The data are cross-sectional, collected in 2017 and 2018, so reverse causation and unmeasured confounding cannot be ruled out; communities with poor sanitation likely face other simultaneous deprivations, from poor schooling to caste-based discrimination, that the models could only partially capture. Toilet coverage has also expanded across India since the survey, meaning the analysis may overestimate current deprivation. LASI additionally does not ask whether people without household toilets use community or public facilities as substitutes. The authors call for future survey waves to probe these mechanisms directly, including questions on sanitation-related shame, visual disorder, and social division, and suggest that qualitative methods already used to study spatial stigma in India could illuminate the lived experience behind the statistics.

The stakes are enormous. India’s population over sixty is projected to double from 149 million in 2022 to 347 million by 2050, and 34.7 percent of those over sixty already report difficulty walking 100 yards, meaning many older adults spend most of their time within their immediate neighborhood. If the community around them is saturated with open defecation and the stigma it carries, their mental health may bear the cost. The authors argue that improving community-wide toilet coverage, not merely distributing household latrines, could become a concrete strategy for depression prevention, one that fits within India’s National Program for the Health Care of the Elderly and its vision of a society for all ages. In a field where mental health interventions often focus on the individual, this study is a reminder that sometimes the most powerful medicine for the mind is the environment outside the front door.

Subject of Research: The association between community-level toilet deprivation and depression among older adults in India

Article Title: Association between community toilet deprivation and depression among older adults in India: a multilevel analysis of the Longitudinal Aging Study in India from 2017-2018

Article References: Jain, A., Kim, R., & Subramanian, S. V. (2026). Association between community toilet deprivation and depression among older adults in India: a multilevel analysis of the Longitudinal Aging Study in India from 2017-2018. GeroScience. https://doi.org/10.1007/s11357-026-02537-9

Image Credits: AI Generated

DOI: 10.1007/s11357-026-02537-9

Keywords: sanitation, depression, older adults, India, LASI, multilevel analysis, built environment, mental health, toilet coverage, spatial stigma, public health, aging

The research team, led by Vasim Ahamad of the Department of Migration and Urban Studies at the International Institute for Population Sciences in Mumbai, together with Rahul Rahul, Ram B. Bhagat, and T. V. Sekher, analyzed data from 64,711 respondents aged 45 and above who took part in the first wave of the Longitudinal Ageing Study in India, conducted in 2017 and 2018. LASI, as the survey is known, was carried out with guidance and ethical approval from the Indian Council of Medical Research and is designed to follow its participants over time, making its first wave an unusually rich snapshot of mid-life and old-age wellbeing across the country. Because the study used this existing secondary dataset, no new information was collected directly from respondents, and the analysis was approved as part of the original LASI protocol.

Two design choices shaped what the study could reveal. First, migration status was defined using each respondent’s place of last residence, a standard demographic measure that classifies people as migrants if their current home differs from where they previously lived. Second, perceived discrimination was measured with an everyday discrimination scale, an instrument widely used in social epidemiology that asks people how often they encounter slights such as being treated with less courtesy or respect than others, receiving poorer service, or feeling that others act as if they are not smart or as if they are afraid of them. The everyday discrimination approach captures the routine, cumulative texture of unfair treatment rather than dramatic single events, which researchers consider especially important because chronic low-level mistreatment is thought to wear down health over years and decades.

The headline numbers are striking. Fifty-seven percent of the older adults in the sample were internal migrants who were growing older at their destination rather than in the place where they had spent their earlier lives. After adjusting for a battery of socioeconomic and demographic factors in multivariable logistic regression models, migrants had significantly higher odds of reporting discrimination than non-migrants, with an adjusted odds ratio of 1.38 and a 95 percent confidence interval running from 1.28 to 1.48. In practical terms, even when older migrants and non-migrants were compared at similar levels of income, education, and other characteristics, the experience of having moved still raised the likelihood of perceiving discrimination by a meaningful margin. That statistical adjustment matters: it indicates the effect is not simply an artifact of migrants being poorer or less educated on average, though those disadvantages certainly compound the picture.

The study’s most unexpected result concerns gender. Male migrants faced markedly higher odds of reporting discrimination, with an adjusted odds ratio of 1.59 and a confidence interval of 1.44 to 1.77, compared with an adjusted odds ratio of 1.21 and a confidence interval of 1.08 to 1.36 among female migrants, each relative to non-migrants of the same comparison. This pattern runs counter to a large body of literature in which women report more discrimination than men, and it points to something specific about the male migrant experience in India. One plausible reading, consistent with the structure of Indian migration itself, is that men who move for work enter competitive labor markets, informal settlements, and public spaces where they interact daily with strangers, landlords, employers, and officials, while many women migrate primarily through marriage and become embedded more quickly in household and neighborhood networks. The study reports the pattern without overinterpreting its causes, but the gap between the two confidence intervals is wide enough to suggest a genuine difference rather than statistical noise.

The type of move mattered as much as the fact of moving. Inter-state migrants, who cross linguistic and cultural boundaries when they relocate, reported greater discrimination than intra-state migrants, who move within a state where language, food, and social norms are more familiar. Among migration streams, rural-to-urban migrants fared worse than those moving along other routes, a result that resonates with decades of qualitative research on how city dwellers view newcomers from the countryside. And the timing of the move shaped outcomes too: recent migrants reported a higher level of discrimination than long-settled migrants, suggesting that the sharpest social friction occurs in the years immediately after arrival, when a person is still visibly an outsider, and that discrimination may fade, though not disappear, as migrants put down roots over decades.

Why should perceived discrimination concern public health researchers and not only sociologists? A substantial international literature links the chronic experience of discrimination to elevated stress, higher blood pressure, worse mental health, and earlier biological ageing, mechanisms that operate through sustained activation of the body’s stress response systems. For older adults, whose physiological reserves are already diminishing, this cumulative burden can be particularly consequential. The authors of the new study frame their findings around the idea of dual vulnerability: older migrants in India face disadvantage on two fronts simultaneously, first because of their age in a society where ageing itself can bring diminished respect and exclusion, and second because of their migrant status, which marks them as outsiders in the communities where they now live. Each form of disadvantage reinforces the other, and the everyday discrimination scale used in the study cannot always separate them cleanly, which the researchers acknowledge implicitly by treating the two as intertwined dimensions of the same lived experience.

The scale of the phenomenon gives the findings national significance. Internal migration in India involves more people than international migration to the entire rest of the world combined, with census and survey estimates consistently placing the number of internal migrants in the hundreds of millions. As the country’s population ages, the cohort of people who migrated in their youth and are now entering their later decades will keep expanding, which means the population exposed to whatever social penalties migration carries is growing rapidly. The LASI sample, with more than sixty thousand respondents aged 45 and above, provides an unusually solid foundation for detecting these patterns, and the use of variance inflation factors to check for multicollinearity among the predictors reflects standard epidemiological care in the modeling. The adjusted odds ratios were estimated with confidence intervals that exclude the null value of one in every reported comparison, indicating that the associations are statistically robust at conventional thresholds.

The authors argue that their results underscore the need for appropriate policies and legal frameworks to address and prevent discrimination against older migrants in India. That recommendation arrives at a moment when the country is still building the institutional infrastructure to serve a mobile, ageing population. Migrants frequently lack local documentation, social networks, and familiarity with grievance mechanisms, and older migrants may additionally face barriers in accessing healthcare, pensions, and social protection schemes that are tied to place of residence. The study did not test specific interventions, so the policy implications remain inferential, but the empirical pattern it documents, a consistent elevation in reported discrimination across migration types, streams, and durations, gives policymakers a clear target: the transition period after a move, and the settings where rural-to-urban and inter-state migrants concentrate.

For a study of this kind, the limitations are worth keeping in view. The data come from a single cross-sectional wave, so the analysis captures associations at one point in time rather than changes over the life course, and perceived discrimination is, by definition, a self-reported measure shaped by expectations, memory, and willingness to disclose sensitive experiences. The place of last residence measure also cannot capture the full complexity of circular or repeat migration, which is common in India. Yet even with these caveats, the study makes a contribution that is hard to dismiss: it takes a question usually asked about young workers, namely whether moving within India exposes people to unfair treatment, and asks it about people in the second half of life, where the stakes for health and dignity are arguably higher. The answer it finds, that migrants who are growing old away from home are more likely to feel treated as lesser, turns a demographic statistic into a human problem, and gives researchers a starting point for the longitudinal work that LASI’s later waves will eventually make possible.

Subject of Research: Perceived discrimination among older internal migrants in India

Article Title: The influence of migration status on perceived discrimination among older adults in India: a nationwide cross-sectional study

Article References: Ahamad, V., Rahul, R., Bhagat, R. B., & Sekher, T. V. (2026). The influence of migration status on perceived discrimination among older adults in India: a nationwide cross-sectional study. BMC Public Health. https://doi.org/10.1186/s12889-026-29800-0

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29800-0

Keywords: internal migration, perceived discrimination, older adults, LASI, India, public health, ageing, rural-to-urban migration, inter-state migration, social epidemiology, migration policy, everyday discrimination scale

The study, led by Saurav Basu of ESI-PGIMSR and ESIC Medical College in Kolkata together with Shubhanjali Roy of the Indian Institute of Public Health-Delhi, drew on Wave 1 of the Longitudinal Ageing Study in India, or LASI, a nationally representative survey coordinated by the International Institute for Population Sciences in Mumbai. The researchers analyzed data from 60,640 adults aged 45 years or older, classifying each participant into one of three categories: no obesity, preclinical obesity, or clinical obesity. The classification operationalized the 2025 Lancet consensus definition of clinical obesity at a population scale for the first time, distinguishing people who carry excess adiposity but have not yet developed obesity-related illness from those whose excess fat is already manifesting as diagnosed disease.

Technically, the framework moved beyond the conventional reliance on body mass index alone. The researchers integrated Asian-specific BMI thresholds, which recognize that harmful levels of body fat occur at lower BMI values in South Asian populations than in European ones, with Asian-specific waist circumference cut-offs that capture central or abdominal adiposity, the pattern of fat deposition most strongly linked to metabolic risk. These anthropometric measures were then combined with self-reported physician-diagnosed conditions, including hypertension, diabetes, chronic heart disease, chronic bone or joint disease, and high cholesterol, to determine whether an individual’s excess adiposity had crossed the threshold into clinical obesity.

The headline numbers are striking. Overall, 37.53 percent of older Indian adults had either preclinical or clinical obesity. Of these, 15.00 percent fell into the preclinical category, carrying excess adiposity without a diagnosed obesity-related condition, with a 95 percent confidence interval of 14.27 to 15.75 percent. The remaining 22.53 percent had clinical obesity, meaning their excess body fat was accompanied by one or more physician-diagnosed comorbidities, with a confidence interval of 21.71 to 23.37 percent. The remaining 62.47 percent of participants, with a confidence interval of 61.55 to 63.39 percent, were classified as not having obesity. In other words, for every two older Indians with obesity-related disease already diagnosed, roughly one more carries excess fat that has not yet translated into illness, a hidden reservoir of future risk.

Perhaps more revealing than the prevalence figures are the determinants that emerged from the study’s multivariable regression models, which adjusted for a wide range of sociodemographic, behavioral, and psychosocial factors. Women had nearly double the odds of moving up the obesity severity spectrum compared with men, with an adjusted odds ratio of 1.97. Urban residents had 88 percent higher odds than their rural counterparts, with an adjusted odds ratio of 1.88, a finding consistent with the well-documented role of urban environments in promoting sedentary lifestyles, altered diets, and cardiometabolic risk. Individuals in the richest quintile of household monthly per capita expenditure had 91 percent higher odds than those in the poorest, an adjusted odds ratio of 1.91, underscoring that in India the obesity epidemic currently tracks socioeconomic advantage rather than deprivation.

One of the most intriguing findings concerns the social dimension of obesity. Individuals classified as socially isolated, measured using validated social network assessment tools, had 44 percent higher odds of belonging to a higher obesity category, with an adjusted odds ratio of 1.44. This aligns with a growing international literature linking loneliness and weak social ties to weight gain, poorer health behaviors, and worse chronic disease outcomes. Whether social isolation drives obesity through reduced physical activity, altered eating patterns, chronic stress, or some combination of these pathways remains an open question, but the association persisted even after adjustment for depression scores measured with the CES-D scale and other confounders. The finding suggests that obesity interventions in ageing populations may need to consider social connection as seriously as diet and exercise.

Age itself told a more complex story. Adults aged 70 and above had lower odds of being in a higher obesity category compared with younger participants, with an adjusted odds ratio of 0.80. This pattern may partly reflect survivor effects, as individuals with the most severe obesity-related disease may not survive into the oldest age groups, as well as natural changes in body composition with advancing age. It also reflects the known limitations of BMI in very old adults, where sarcopenia, the age-related loss of muscle mass, can mask adiposity behind a deceptively normal weight. The phenotype-based approach used in this study, which incorporates waist circumference rather than relying on BMI alone, is designed to be more sensitive to these age-related shifts in body composition.

The morbidity burden associated with the obesity phenotypes was substantial. Compared with participants without obesity, those with preclinical or clinical obesity had higher prevalence of hypertension, diabetes, chronic heart disease, chronic bone or joint disease, high cholesterol, and multimorbidity, the co-occurrence of two or more chronic conditions. The gradient across phenotypes is clinically meaningful: preclinical obesity identifies people with excess adiposity who have not yet accumulated diagnosed disease, while clinical obesity identifies those in whom the disease process has already declared itself. This distinction, the authors argue, adds diagnostic value beyond BMI by enabling refined risk stratification at the population level, allowing health systems to target intensive management toward those with established disease while directing early identification and prevention toward those at the preclinical stage.

The biological rationale for this phenotyping is well established. Excess adipose tissue, particularly visceral fat, is not a passive energy store but an active endocrine and inflammatory organ. Chronic low-grade inflammation originating in adipose tissue is a recognized mechanism linking obesity to insulin resistance, type 2 diabetes, fatty liver disease, and cardiovascular disease. In older adults, the interaction between obesity and declining muscle strength further accelerates mobility loss and functional impairment. By distinguishing between adiposity that has and has not yet produced clinical manifestations, the new framework attempts to resolve long-standing debates over the heterogeneous health significance of elevated BMI, including the contested concept of metabolically healthy obesity.

For India, the stakes are considerable. The country is undergoing a well-documented nutrition transition, with rising consumption of ultra-processed foods, declining physical activity, and expanding urbanization driving an epidemic of noncommunicable diseases. Routine obesity diagnosis in Indian clinical practice and national surveys has traditionally leaned heavily on BMI, which fails to capture the metabolic heterogeneity that this study reveals. The authors argue that public health strategies should strengthen early identification of high-risk individuals, integrate waist circumference measurement and morbidity assessment into existing noncommunicable disease screening programs for older adults, and address the social and structural factors, from urban design to social isolation, that shape the obesity burden.

The study’s scale and representativeness lend its conclusions unusual weight, though the authors note that it is based on a secondary analysis of de-identified, publicly available LASI Wave 1 data, with diagnoses relying on self-reported physician diagnoses rather than direct clinical examination. Even with those limitations, the message is clear: obesity in ageing India is common, socially patterned, and already deeply entangled with chronic disease, and recognizing the difference between carrying excess fat and being made sick by it could transform how the country, and other nations facing similar transitions, approach the epidemic. As populations worldwide grow older and heavier, the Indian findings offer both a warning about the scale of hidden preclinical risk and a practical template for identifying, stratifying, and ultimately reducing the morbidity that excess adiposity imposes on ageing societies.

Subject of Research: Prevalence, determinants, and morbidity burden of clinical and preclinical obesity phenotypes among older adults in India

Article Title: Clinical obesity phenotypes among older adults: prevalence, determinants, and morbidity burden from the Longitudinal Ageing Study in India

Article References: Basu, S., & Roy, S. (2026). Clinical obesity phenotypes among older adults: prevalence, determinants, and morbidity burden from the Longitudinal Ageing Study in India. International Journal of Obesity. https://doi.org/10.1038/s41366-026-02226-9

Image Credits: AI Generated

DOI: 10.1038/s41366-026-02226-9

Keywords: obesity, older adults, India, LASI, clinical obesity, preclinical obesity, BMI, waist circumference, multimorbidity, social isolation, noncommunicable diseases, public health

The study, published in BMC Medicine, drew on Wave 1 of the Longitudinal Ageing Study in India, a landmark survey designed to be representative of the country’s population aged 45 and above. The analytical sample focused on individuals aged 50 and older who had at least one child, allowing the researchers to ask a deceptively simple question with unusual rigor: does the number of children a parent has actually cause changes in loneliness, or are the two merely entangled with other factors such as poverty, education, health, and social circumstances?

Answering that question requires more than a simple comparison of lonely parents with different family sizes, because the number of children a couple has is not random. Fertility reflects choices and constraints: wealth, education, religion, region, health, and personality can all shape both how many children people have and how socially connected they feel decades later. Observational studies that simply correlate fertility with loneliness therefore risk producing estimates that are biased in unknown directions. The research team, led by Yanshang Wang of University College London with colleagues at Peking University, Heidelberg University, the International Institute for Population Sciences in Mumbai, and Harvard University, addressed this with an instrumental variable approach using two-stage least squares estimation.

The instrument they exploited was the sex of the firstborn child. In many societies, and particularly in contexts with strong son preference, the sex of a first child influences subsequent fertility decisions: families whose first child is a daughter may go on to have more children than those whose firstborn is a son, in part because of cultural preferences and norms surrounding sons. Crucially, the sex of a firstborn child is essentially random at conception, determined by biological chance rather than by any characteristic of the parents that might independently affect loneliness. By using this natural experiment, the researchers could isolate the portion of variation in family size driven by chance, and trace its consequences for loneliness, avoiding the confounding that plagues ordinary correlations.

The headline result from the full sample was surprising in itself: each additional child was associated with higher loneliness scores in later life, with an estimated effect of 0.076 points on the loneliness scale (a 95 percent confidence interval of 0.019 to 0.132, and a p-value of 0.009). But this average concealed a striking split along gender lines. Among fathers, each additional child significantly reduced loneliness, with an effect of minus 0.095 points (95 percent confidence interval of minus 0.187 to minus 0.003, p equal to 0.043). Among mothers, the pattern ran in the opposite direction and with greater magnitude: each additional child increased loneliness by 0.218 points (95 percent confidence interval of 0.137 to 0.298, p less than 0.001). The authors stress that these are causal estimates, not mere associations, which makes the gender divergence all the more consequential.

When the researchers stratified their analyses by residential setting, the sex-differentiated patterns proved most pronounced in rural India. This geographical texture matters. In rural areas, traditional family structures, stronger son preference, and more constrained access to social infrastructure beyond the household may amplify the gendered division of labour within families. Mothers in such settings often carry the bulk of caregiving and domestic responsibilities, and additional children may extend those burdens deep into parents’ later years, even as adult children migrate away for work. Fathers, by contrast, may reap the social status and network benefits of a larger family without absorbing the same costs. Urban settings, with different norms and greater availability of non-family social ties, appeared to soften these divergent effects.

Exploratory analyses of downstream mechanisms offered further clues about why children might affect mothers and fathers so differently. The researchers examined indicators including social engagement and proximity to children, and found that geography within the family moderated the emotional consequences of family size. Co-residence with children attenuated the adverse effects for mothers by 16.87 percent, and living nearby attenuated them by 22.21 percent. For fathers, proximity worked in the opposite direction, strengthening the protective effect of additional children by 4.52 percent for those co-residing and by 14.00 percent for those with children living close by. In other words, having children nearby transforms what a large family means emotionally, cushioning mothers against the loneliness that additional children would otherwise bring, while amplifying the companionship fathers derive from them.

These findings arrive at a moment when loneliness is increasingly recognised as a public health concern of the first order. Chronic loneliness in older adults is associated with elevated risks of depression, cognitive decline, cardiovascular disease, and premature mortality, prompting several countries to appoint ministers for loneliness and to treat social connection as a measurable policy target. Much of the evidence on family size and wellbeing, however, comes from high-income countries with different family systems, and many studies have been content to report correlations. By delivering causal estimates from a low- and middle-income country with one of the world’s largest ageing populations, this study fills a substantial gap and complicates the received wisdom that bigger families buffer older people against isolation.

The implications are particularly pointed for India, where the population aged 60 and above is projected to grow rapidly in the coming decades and where the joint family, long idealised as a safeguard against the isolation of old age, is being reshaped by urbanisation, migration, and declining fertility. If additional children raise loneliness among mothers, then policies that implicitly treat fertility as a solution to elder isolation, or that assume adult children will reliably supply social support, may be misguided. The authors argue that their results challenge the assumption that larger families uniformly protect older adults from loneliness and instead call for gender-sensitive and community-based approaches to later-life social support, ones that do not lean on family size alone to secure wellbeing in old age.

Like all instrumental variable studies, the analysis rests on assumptions that cannot be verified directly, including the requirement that a firstborn’s sex affects loneliness only through its influence on subsequent fertility. The researchers also note that the downstream analyses of proximity and social engagement were exploratory rather than definitive tests of mechanism. Even so, the study’s design represents one of the most credible attempts to date to pin down the causal effect of childbearing on late-life loneliness, and its central lesson travels well beyond India. Family size is not a one-size-fits-all prescription against isolation. Whether children connect parents to the world or quietly cut them off from it depends on who carries the work of family, and where everyone ends up living.

The data underpinning the study come from LASI, one of the largest longitudinal ageing surveys ever mounted in a low- or middle-income country, which interviews tens of thousands of older Indians across all states and union territories. Because the survey captures detailed information on household composition, health, income, and social participation, it allowed the researchers to construct loneliness measures alongside the family and residential variables needed to probe mechanisms. The instrument itself draws on a well-documented demographic phenomenon: in settings with son preference, firstborn sex shifts completed fertility, a pattern demographers have exploited in other contexts to study the consequences of family size for child health and education.

The asymmetry between mothers and fathers fits a broader literature on gendered ageing. Women typically outlive men, are more often widowed, and tend to have fewer independent sources of income and social contact outside the household, which can leave their emotional wellbeing more tightly coupled to family dynamics. At the same time, mothers in large families may face intensified expectations to provide grandchild care and household labour, obligations that can crowd out other forms of social connection even as they anchor parents within busy, and sometimes physically distant, family networks.

The moderating role of proximity is consistent with research on migration within India, where adult children frequently move to cities for employment, leaving parents behind in origin communities. Such distance can sever the day-to-day contact that converts a large family into a source of companionship, and the study’s finding that nearby residence changes the emotional arithmetic of family size echoes earlier observational work on transnational and internal migration and elder wellbeing.

For policymakers, the results suggest that interventions targeting loneliness in ageing populations should look beyond household structure. Community centres, self-help groups, and social pension schemes that expand older women’s networks outside the family may matter more than encouraging co-residence alone. The authors’ call for gender-sensitive, community-based support aligns with a growing international consensus that social connection in later life is a modifiable determinant of health, one that family size alone cannot be relied upon to secure.

Subject of Research: Causal effect of number of children on loneliness among older parents in India

Article Title: The effect of number of children on loneliness: a nationally-representative instrumental-variable study in India

Article References: Wang, Y., Yang, F., Wayal, S., Bakht, S., Sujata, S., Kern, M., Singh, A., & Bäernighausen, T. (2026). The effect of number of children on loneliness: a nationally-representative instrumental-variable study in India. BMC Medicine. https://doi.org/10.1186/s12916-026-05232-w

Image Credits: AI Generated

DOI: 10.1186/s12916-026-05232-w

Keywords: loneliness, number of children, instrumental variables, India, ageing, BMC Medicine, LASI, gender differences, rural health, social support, fertility, older parents