Health after mid-life: the role of adult children's education for older parents' frailty in Europe
Health after mid-life: the role of adult children's education for older parents' frailty in Europe
Abstract:
With increasing longevity in Western societies, the transmission of socio-economic resources between generations becomes bidirectional, from parents to children and vice versa. However, while the impact of family background on children's life opportunities has been the topic of much research, less is known about the reverse influence of children's education on parental health. This study investigates the association between adult children's educational attainment and parental health, considering heterogeneity by gender, age, and societal context. We use eight waves of panel data from the Survey of Health, Ageing and Retirement in Europe spanning two thousand four to twenty twenty-two, tracking eighty-six thousand five hundred parents across twenty-seven countries. We estimate random intercept and growth curve models on a Frailty Index with inverse probability weighting for multivalued treatments to address selection and endogeneity bias. The results show that parents - particularly mothers - with highly educated children have lower frailty. The beneficial role of child education is strongest for older mothers and more pronounced in Southern and Eastern Europe than in Northern Europe, where both cultural traits and universal welfare systems may mitigate educational disparities. We interpret these results in terms of generational equity, suggesting that investments in younger generations' education also benefit older generations' health, particularly in countries with weaker welfare systems and strong family ties.
Introduction
Introduction
An unforeseen consequence of population aging is the bidirectional transmission of social (dis)advantages between generations, from parents to children and from adult children to their older parents. As family sizes shrink, family resources become concentrated among fewer offspring, enabling children to receive a larger share of inheritances, inter-vivo transfers, and other forms of parental investment that promote their economic success, health, and life chances. Concurrently, rising life expectancies extend the period during which adult children can serve as crucial sources of social, practical, and informational support for their aging parents. Studies have acknowledged the influence of the "social foreground", which parallels that of family background, highlighting the increasing role of adult children's social and economic resources in shaping health inequalities in later life. However, while the influences of family background on children's life opportunities have been the topic of much research, less straightforward are the intergenerational consequences of adult children's social and economic resources on their parents' health in later life.
Research on the intergenerational effects of children's education on parental health faces three challenges. First, previous studies often rely on single-country analyses with limited health assessments, typically single-item measures, yielding mixed findings on self-reported general health, mental well-being, cognitive functioning, physical health, and mortality. These contradictions may reflect differences in study design across countries, sample composition, and measures of health employed. As a result, knowledge remains partial and fragmented, as specific health measures are studied in isolation from the broader physical and mental systems that shape vulnerability and dependency in later life.
Second, little attention has been given to potential heterogeneity across the life course, genders, and country contexts. Over the life course, the flow of intergenerational support reverses, with parental roles shifting from "net providers" to "net receivers". As parents age and have fewer resources to cope with external stressors, their children's socio-economic resources - defined here through education - become increasingly important for supporting health care needs and promoting healthy behaviors. These family life courses are presumably highly gendered, with older women having longer life expectancies, stronger intergenerational ties, and greater family responsibilities compared to men. Mothers may be more affected than fathers, both positively and negatively, by their children's education, given their greater investments and emotional attachment to family ties. Additionally, the influence of children's education on parental health could vary across countries due to differences in welfare state characteristics, cultural norms, and healthcare systems. These variations underscore the need for comparative data and analyses across a wide range of individual characteristics and societal contexts.
Third, while theoretical perspectives suggest an "upward" influence of children's education on older parents' health, this relationship may be influenced by endogeneity and unobserved confounders, as parents with higher socioeconomic status are often more capable of providing resources and support for their children's education and tend to experience better health conditions. To reduce the potential sources of bias, recent studies have used propensity score matching and instrumental variable approaches exploiting policy changes that increase schooling age. However, the empirical evidence remains mixed on whether children's education causally affects parental physical and mental health.
For instance, Torres show that an additional year of schooling had minimal or no impact on parents' depressive symptoms. Similarly, Potente found that the causal evidence remains inconclusive, as their quasi-natural experiment exploiting the nineteen seventy-two educational reform in England and Wales - which raised the minimum school-leaving age from fifteen to sixteen - showed no substantive causal effects on children's education on parental self-reported health and long-standing illness. Conversely, using inverse probability of treatment weighting to account for selection bias, Zhang and Silverstein show that children's economic success is associated with the better self-reported health of their parents. These mixed findings suggest that the impacts of children's education on parental health may depend on how accurately pre-existing confounding and selection factors are captured in different empirical approaches.
This study aims to contribute to the literature on the upward transmission of advantages across generations by addressing the three abovementioned challenges. Unlike earlier studies focusing on single health items, we used a forty-deficit Frailty Index as the health outcome. Frailty is a holistic concept of health, defined as a clinically identifiable condition of diminished physiological reserve and increased vulnerability to a wide range of adverse health outcomes, including general health deterioration, hospitalization, and mortality. Since frailty involves a progression from robustness to general health decline in a dynamic (i.e., preventable or delayable) and relatively reversible process, adult children may serve as a latent "safety net" that can bolster the physical, mental, and social reserves of aging parents, promoting feelings of reassurance even before serious health needs arise.
We also address key shortcomings in the literature by examining heterogeneity across parental life course, gender, and country context. Inequalities in health and the need for support are expected to accumulate with age due to the compounding consequences of stressors and resource deficits over time. However, age effects are rarely analyzed systematically in previous studies, which only provide limited insights into how the influence of children's education can evolve as parents age. Similarly, gender differences are often treated as secondary, despite theoretical expectations regarding distinct gendered dynamics in intergenerational exchanges. Lastly, differently from most previous studies examining a single country such as the United States, United Kingdom, Mexico, or Asian countries, we investigate the association between child educational attainment and older parents' health across multiple European countries, where healthcare access and cultural norms around family ties vary remarkably. Leveraging the longitudinal and cross-national comparability of the Survey of Health, Ageing and Retirement in Europe, covering twenty-seven European countries over eight waves, we offer a robust framework for understanding how the broader societal context moderates the role of children's education on parental health.
Finally, our study addresses key methodological challenges by using a "doubly robust" inverse probability of treatment weighting approach to reduce biases from selection and endogeneity. We apply this approach to the context of multiple treatments, specifically different levels of children's education. Approximating experimental conditions, this method reduces the risk of selection and omitted variable bias, thus responding to the longstanding critiques of previous studies for their limited causal validity to evaluate intergenerational spillover effects of education.