36th Reves conference
Conference program and abstract book
WEDNESDAY, 14TH MAY
9.00 Opening of the conference (Reves executive board & conference organizers)
9.15 Session 1: Ageing, Health, and Mortality Inequalities: Evidence from Finnish Registers Using Life Table Metrics
Moderator: Marja Jylhä
Pekka Martikainen: Social Inequalities in Health: Life Table Approaches to Research on Aging Populations Using Register-Based Data
Margherita Moretti: Widowhood lifespan and the evolution of widowhood expectancy in Finland over the last three decades
Kaarina Korhonen: Future Trends in Dementia Incidence and Mortality: Projecting the Burden of Dementia on Long-Term Care in Finland by 2040
10.30 Coffee and tea (posters)
10.50 Session 2: Social inequality in life and health expectancy
Moderator: Marc Luy
Emmanuelle Cambois: Trends health expectancies in France: to what extent the social disparities in healthy ageing have been reshuffled over the 2010’s?
Tianyu Shen: Educational Inequalities in Life Expectancy and Disabled Life Expectancy in the U.S.: An Intersectional Cohort Analysis by Race and Sex
Jilei Wu: Gender Differences based on Self-Rated Health Expectancies of Elderly among Provinces of China with a Decade Socioeconomic Development
David Sinclair: Disability Free Life Expectancy: Narrowing the gap between rich and poor
Mateo Farina: Education and Longevity in Older Brazilians: Uncovering Gender Differences in Mortality Patterns
12.20 Group photo
12.30 Lunch (posters)
13.30 Session 3: Mental health and frailty
Moderator: Aïda Solé-Auró
Silvia Loi: Understanding Differential Physical and Mental Health Trajectories of Refugees and Migrants
Stefan Fors: The effect of loneliness on depressive symptoms in the 65+ European population: A longitudinal observational study using SHARE data
Pauliina Halonen: The association between socioeconomic status and dementia in the oldest old – a time series analysis between 2001 and 2022
Yvanna Simon: Association between loneliness profiles and risk of dementia in the Paquid cohort
Anna Tirkkonen: Cognitive Performance in Late Midlife as a Predictor of Frailty over 17 Years: A Longitudinal Birth Cohort Study
15.00 Coffee and tea (posters)
15.30 Session 4: Methods
Moderator: Zachary Zimmer
Erika Banzato: Modeling multimorbidity: exploring diseases interactions and their connection to frailty outcomes
Nicolas Brouard: Challenges in estimating multi-state expectancies from large-scale cross-longitudinal surveys
Rustam Tursun-zade: A demographically motivated index of within-population health inequality
Pietro Belloni: Measuring frailty in the elderly: an indicator based on a combination of classifiers
18.00 Reception at the City Hall (Keskustori 10, 33200 Tampere)
19.30 Conference dinner at Restaurant Tampella (Kelloportinkatu 1, 33100 Tampere)
THURSDAY, 15TH MAY
9.00 Session 5: Stagnation in life expectancy
Moderator: Sari Stenholm
Leah Abrams: Birth Cohort Contributions to Stalling U.S. Mortality Improvements
Hong Mi: Re-estimating Urban-Rural Disparities in Disability Life Expectancy Among Older Population in China Based on Bayesian Methods
José Andrade: Cohort mortality forecasts indicate signs of deceleration in life expectancy gains
Octavio Bramajo: Catching up with stagnation: cause-specific dynamics of change in life expectancy at age 65 in the United States, Canada and Latin America, 2000-2019
Andrew Stokes: The Role of Rising Obesity in the Stagnation of U.S. Life Expectancy
10.30 Coffee and tea (posters)
10.50 Session 6: Support systems and personalized care
Moderator: Pauliina Halonen
Pamela Herd: Burden Reduction in a Social Safety Net Program Reduces Mortality
Giampaolo Lanzieri: Developing European statistics on long-term care
Outi Mäenpää: Rehabilitation needs of persons living with dementia
Anne-Marie Mäkelä: The possibilities of home care clients with dementia influencing their care through goal setting
Puneet Kaur Chehal: The Long-Term Impacts of Medicaid Exposure in Early Childhood on Aging
12.20 Lunch (posters)
13.20 Session 7: Life and health expectancy across populations
Moderator: Mikaela von Bonsdorff
Aïda Solé-Auró: The Unequal Healthy Life Expectancy of men and women.
Rahul Malhotra: Association of sleep and nap duration with total life expectancy, and life expectancy with and without cardiovascular diseases among older adults: A longitudinal population-based study
Marc Luy: Extent of the health disadvantage of Roma people in Europe: Indirect estimations for 17 countries
Alessandro Feraldi: Healthy Working Life Expectancy: Measurements and Socio-Economic Inequalities
Milos Milovanovic: The Years Left Abroad: The impact of emigration on longevity in the Republic of Serbia
14.50 Coffee and tea (posters)
15.20 Session 8: Social inequality in health
Moderator: Linda Enroth
Madelin Gómez-León: Assessing Health Inequalities. Age and Cohort Trajectories Among Older Adults in Europe.
Sarah Åkerman: The role of socioeconomic resources for adhering to the “Blue Zone” lifestyle in contrasting regions in Western Finland and northern Sweden
Karine Pérès: Risk profiles for serious health events according to the first step of the WHO ICOPE approach
Gilbert Brenes-Camacho: Socioeconomic and institutional conditions associated to variance of age at death
Anastasia Lam: Health and wealth trajectories in childless families: how childlessness shapes cumulative inequalities over the life course
16.50 Bio break
17.00 George Myers honorary lecture
Moderator: Emmanuelle Cambois
Jean-Marie Robine: A journey through time over the past 80 years, from 1945 to 2025
18.15 General Assembly (Reves executive board)
FRIDAY, 16TH MAY
9.00 Session 9: Sociodemographic factors and morbidity
Moderator: Kaarina Korhonen
Nosraty Lily: Associations of migration-related and sociodemographic factors with functional mobility and depressive symptoms among midlife and older Russian-speaking migrants
Solveig Argeseanu Cunningham: Quantifying the Lifecourse of Multimorbidity: New Applications of Demographic Methods
Camille Ouvrard-Brouillou: Social vulnerability differently increases the risk of death in men and women: longitudinal analysis over 15 years in the Paquid Study
Emmanuelle Cambois: Gender- and social-specific contribution of chronic diseases to activity limitation in the 60+ population in France
Saara Marttila: Do epigenetic clocks predict future health outcomes in a middle-aged population, and how do they compare to traditional risk factors of cardiometabolic diseases?
10.30 Break
10.50 Session 10: Work, retirement and health
Moderator: Rahul Malhotra
Sari Stenholm: Association of behaviour-related health risk factors with working life expectancy in adults aged ≥50 years
Zachary Zimmer: Trends in productive life expectancy among older adults: An assessment of the expansion or contraction of productive aging
Anu Polvinen: Life course before and after receiving a fixed-term disability pension: an 8-year follow-up study
Thomas Arnhold: Exploring heterogeneities in cognitive functioning at older age: The role of employment histories across European welfare regimes.
Dorly Deeg: Work participation in times of rising state pension age: the decreasing role of health
12.20 Closing remarks (Reves executive board & next conference organizer)
12.30 Lunch
Poster presentations in alphabetical order
Leah Abrams: How trends in working longer shape socioeconomic and racial/ethnic disparities in healthy-working life expectancies in the United States
Florencia Bathory: Onset of depression among people with chronic diseases
Sherin Billy Abraham: Frequency and Predictors of Emergency Department Visits among the Oldest Old in Finland: the Vitality 90+ Study
Félix Blain: Healthy Working Life Expectancy across birth cohorts in the United States
Phillip Cantu: Differential Item Functioning of informant measures of cognitive functioning in the U.S. and Mexico
Laura Kananen: Predicting cardiovascular morbidity and mortality with SCORE2(OP) and Framingham risk estimates in combination with biological aging indicators
Anna Lahti: Associations between social vulnerability and functioning in older age, and the moderating and mediating role of optimism and self-efficacy
Yan Liu: Paradox of Aging? Evidence Based on Happy Life Expectancy and Healthy Life Expectancy in China
Marja Lönnroth: Experiences of Agency Among People with Dementia
Susanna Satuli-Autere: Risk factors predisposing to dementia in individuals with type 1 diabetes
Sini Stenroth: Adversity, perceived stressfulness, and resilience in older men and women – the role of socioeconomic status and functioning
Sara Suikkanen: Developing the customer journey and implementing the new ways of work in the health care and social services for older adults
Konsta Valkonen: Management of vascular risk factors and its association with cognition among older adults
Zixuan Wang: The Social Networks of Chronic Diseases
Jiao Zhang: The Construction and Empirical Research of Healthy City Indicator System from the Perspective of the Theory of Social Determinants of Health: Taking 54 Major Cities as Examples
Huiping Zheng: Decomposing Differences in Cohort Health Expectancy by Cause and Age with Longitudinal Data
Yan Zheng: Living longer in poorer health: Health expectancy in Israel between 2013 and 2022: a cross-national comparison with 15 European countries
Abstracts
Oral presentations in presenting order
Widowhood lifespan and the evolution of widowhood expectancy in Finland over the last three decades
Margherita Moretti1,2, Kaarina Korhonen1,2, Alyson van Raalte2,3, Timothy Riffe4,5,3, Pekka Martikainen1,2
1 Helsinki Institute for Demography and Population Health, University of Helsinki
2 Max Planck – University of Helsinki Center for Social Inequalities in Population Health, University of Helsinki
3 Max Planck Institute for Demographic Research
4 OPIK, University of the Basque Country (UPV/EHU) 5 Ikerbasque (Basque Foundation for Science)
Background: Widowhood is a disruptive event affecting a wide range of life domains, and in ageing societies, increased numbers of individuals are potentially exposed to it. Yet we lack a comprehensive understanding of the demography of widowhood.
Methods: We draw on the entire population of individuals aged 65 and older who resided in Finland from 1987 to 2019 from the Statistics Finland population register. Using information on marital and cohabiting unions, discrete-time event history analysis and incidence-based multistate lifetables, we analyse lifetime risk, mean age at becoming widowed, widowhood expectancy, and variation in years spent widowed, and document gender and educational differences in these metrics over the last three decades in Finland. Additionally, we will show preliminary results of a demographic decomposition of widowhood expectancy, in order to understand the drivers of its change over time.
Results: Our results show that, over time, individuals are less likely to experience widowhood, and when they do, it occurs at older ages. Women have higher widowhood risk, expectancy, and a lower mean age at widowhood than men. Widowhood expectancy for women declined from 8 to 6 years, while for men, it stagnated at around 2 years. Low-educated women faced more widowhood years than highly educated, while the opposite holds for men. The results of the decomposition analysis will highlight the contributions to changes in widowhood expectancy given by improvements in survival, changes in partnership dynamics, and changes in the age and educational differences between partners.
Conclusions: By showing decreased risks, delayed onset, and shorter widowhood expectancy, particularly among women, our results suggest that the current older population may experience reduced exposure to the psychosocial and financial challenges of widowhood, with potentially reduced caregiving burden on families and the state. We will also disentangle how demographic changes affect changes in widowhood.
Future Trends in Dementia Incidence and Mortality: Projecting the Burden of Dementia on Long-Term Care in Finland by 2040
Maria Guzman-Castillo1 , Kaarina Korhonen1,2,*, Michael Murphy3 , Pekka Martikainen1,2
1 Helsinki Institute for Demography and Population Health, Faculty of Social Sciences, University of Helsinki, Finland
2 Max Planck – University of Helsinki Center for Social Inequalities in Population Health, University of Helsinki 3Department of Social Policy, London School of Economics * Presenting author Abstract
Background: We projected the number of individuals with dementia in the community and in residential long-term care (LTC) by education in Finland by 2040.
Methods: A multistate model tracks the population aged 65 years and over through states characterised by the presence of dementia, LTC residence or death, using age-, sex-, education- and year-specific transition probabilities calculated from Finnish registry data under different scenarios.
Results: Under constant transition probabilities, number of individuals with dementia will rise by 74% (from 123,88 to 215,900) between 2018 and 2040. The number of LTC residents is expected to increase by 58% (from 73,900 to 116,800), increasingly suffering dementia. The number of basic-educated with dementia will decrease. Were both dementia incidence and mortality to decline as projected for all-cause mortality, dementia and LTC cases would increase by 67% and 73%, respectively. If the declining trend in LTC use observed prior to 2018 persists, the number of LTC residents would decrease by 10%, but nearly 170,000 individuals with dementia would live in the community in 2040.
Conclusions: Despite optimistic scenarios involving improvements in dementia incidence and mortality rates, this study suggests that demographic pressures on health and social care systems will persist in the coming decades.
Trends health expectancies in France: to what extent the social disparities in healthy ageing have been reshuffled over the 2010’s?
Emmanuelle Cambois, Florian Bonnet (Ined) (Work in progress)
Trends in the chance of survival and healthy ageing are shaped by a country’s general living conditions, the historical trajectories of successive generations, and the individual life courses within these cohorts. Health expectancy differentials across population groups emerge from these intertwined, multi-level dynamics. To capture these complexities, this study examines trends in disability-free life expectancy (DFLE) at the intersection of gender and socio-occupational class (SOC) in France throughout the 2010s. DFLE estimates are derived from the Global Activity Limitation Indicator (GALI), using data from the French waves of EU-SILC (2011-2019) and triennial life tables by SOC. We apply Sullivan’s method to integrate these sources, also accounting for years spent in nursing homes. At this initial stage of analysis, we compare trends and explore both convergences and divergences. Over the 2010s, life expectancy (LE) fluctuated for both sexes. However, women’s DFLE steadily increased, whereas men’s DFLE exhibited irregular patterns. When stratifying by SOC, we found that LE significantly increased only among highly skilled women and manual worker men, with both groups also experiencing a steady rise in DFLE. In contrast, men in middle- and low-skilled occupations experienced stagnating LE and a sharp decline in DFLE. Notably, middle- and low-skilled women had similar LE, but the latter experienced a shorter and stagnating DFLE—a gap that widened over the decade. These shifting trends have reshuffled social disparities, ultimately altering the landscape of gender and SOC differentials in DFLE. These results point out vulnerable groups characterized by shorter and unhealthy ageing or by an absolute expansion of the disability years. We discuss the implications of this evolving picture.
Educational Inequalities in Life Expectancy and Disabled Life Expectancy in the U.S.: An Intersectional Cohort Analysis by Race and Sex
Tianyu Shen1,2, Courtney E. Boen3 and Collin F. Payne2,4
1 Vienna Institute for Demography, Austrian Academy of Sciences
2 School of Demography, Research School of Social Sciences, Australian National University
3 Sociology Department, Brown University
4 Center for Population and Development Studies, T.H. Chan School of Public Health, Harvard University
Substantial disparities in longevity and disability exist between race, sex, and educational groups in the US. However, little research to date has taken an intersectional approach to simultaneously consider inequalities at the intersection of race, sex, and education, particularly from a cohort perspective. It also remains unclear whether and how these educational inequalities have changed within the intersection of race and sex across successive birth cohorts. Utilizing data from the US Health and Retirement Study (HRS) from 1998 to 2020 and multistate modelling, we estimate 22 years of partial cohort life expectancy (LE) and disability-free/disabled life expectancy (DFLE/DLE) by education, race, and sex across three birth cohorts, 1918-1927, 1928-1937, and 1938-1947. Consistent with other studies, there is a clear gender and racial gap in health and mortality, with women living longer but also spending more time unhealthier than men, and the white living longer and healthier than the Black. Disaggregating by education, we reveal some nuances in the gender and racial gap by education, but more importantly, highlight the key contributor to gaps in LE and DFLE between college graduates and those without a high school diploma. Beyond completing high school, a college degree is particularly beneficial for white men, while Black individuals experience more pronounced health benefits from having some college education. We also decompose DFLE/DLE disparities by education across race-sex groups to assess whether they arise from baseline functional impairments accumulated from younger ages or divergent patterns of functional decline with age. Findings reveal these disparities are primarily driven by differences in transitions between disability states and mortality in older ages, emphasizing the impact of early-life education on later-life health and highlighting the potential for targeted interventions in older adulthood to reduce health inequalities.
Gender Differences based on Self-Rated Health Expectancies of Elderly among Provinces of China with a Decade Socioeconomic Development
Wu Jilei & Qiao Xiaochun
Institute of Population Research, Peking University, Beijing 100871
The population health-oriented strategy is the kernel of socioeconomic development, and the key issue of healthy ageing with the rapidly aging population. Based on self-rated health of the elderly obtained by 2010 and 2020 National Population Census and national and provincial life tables of the elderly provided by the National Bureau of Statistics, the elderly health life tables for the nation and provinces are calculated by the Sullivan method, from which health expectancies at exact age 60, self-care expectancies, proportions of health expectancy among life expectancy, and index of quality of life are calculated. The spatial exploratory data analysis and multiscale geographical weighted regression model were utilized for the provincial difference and associated macro-factors on the gender difference of those elder people. The results show that both life expectancy and health expectancies had increased for male and female of the elderly during the last decade. However, there are regional difference by health expectancies between gender. While macro-socio-economic factors explains the association with gender difference of health expectancies among the provinces, the health quality index has better discriminability and shows significant hotspots on the south-east regions. More, with standardization of those socio-economic factors, health resources and education level related to gender healthy expectancies difference, the health quality index has identified the socio-economic factors influence more than the other two factors, which verified the national health-oriented strategy.
Keywords: Healthy Expectancy Life, Gender Difference, Regional Analysis
Disability Free Life Expectancy: Narrowing the gap between rich and poor
David R Sinclair1, Fiona E Matthews2 , Laurie Davies1 , Barbara Hanratty1 , Andrew Kingston1
1Population Health Sciences Institute, Newcastle University, UK
2University of Hull, UK
Background: Health expectancy inequalities in England are stark, with a 16-year gap in Disability-Free Life Expectancy (DFLE) between affluent and deprived areas. Reducing this inequality and extending health expectancy has been a longstanding policy priority, but the optimal strategies remain unclear. This study models potential interventions to extend DFLE and reduce inequalities, identifying high-potential areas and quantifying the scale of changes required. In particular, we consider the interventions required to extend DFLE by five years and reduce inequalities, a previously stated government target.
Methods: We used multistate modelling on longitudinal data from 19,407 individuals aged 50+ in England, integrating disability data from three cohort studies: the English Longitudinal Study of Ageing (ELSA), the Cognitive Function and Ageing Study II (CFAS II), and the Newcastle 85+ study. Scenarios examined how reducing disability risks associated with age and deprivation could extend DFLE.
Results: Interventions targeting deprivation-based inequities yielded modest DFLE gains, extending life expectancy by up to 2.8 years for women and 2.3 years for men in deprived areas. Age-focused interventions in isolation exacerbated the DFLE gap between the most and least deprived populations. In contrast, strategies addressing both age- and deprivation-associated risks showed greater potential. A 30% reduction in age-related disability risks, coupled with equivalent increases in recovery rates and the elimination of deprivation-associated inequalities, increased DFLE by 4.8–8.6 years for men and women aged 50. Women in deprived areas experienced the largest gains.
Conclusions: Extending DFLE by five years requires closing the deprivation gap while reducing age risk. Ageonly approaches may widen inequalities, emphasising the need for integrated, equity focused public health strategies.
Education and Longevity in Older Brazilians: Uncovering Gender Differences in Mortality Patterns
Mateo Farina
University of Texas at Austin
Background: Educational attainment is one of the strongest predictors of mortality risk in later life. However, most research has focused on high-income countries, with more limited understanding of how education shapes mortality risk in low- and middle- income countries (LMIC). Specifically, questions remain of whether differences in educational attainment in primary schooling—the level at which most older adults in LMIC were educated— are associated with lower mortality risk.
Methods: We use the ELSI-Brazil Study, a nationally representative sample of older adults 50 years and older. We examine 11 functional forms to evaluate how educational attainment at different years and education credentials are associated with 6-year mortality (2016-2021) for Brazilian men and women. We also evaluated the sensitivity of the best fitting model to inclusion of childhood selection and adult health and health behaviors.
Results: We found that each year of education was associated with decreased mortality risk for Brazilian women, but not for men. For Brazilian men, we found that mortality risk declined after obtaining a high school diploma. Some of the decrease in mortality risk with higher educational attainment was attenuated by the inclusion of childhood selection, with almost no change after including adult health and health behaviors.
Conclusions: The findings underscore the importance of primary schooling for Brazilian women, but a high school credential for men. More research should examine why gender differences arise with attention to mortality selection into older adulthood and larger socioeconomic conditions shaping the role of education with later life health.
Understanding Differential Physical and Mental Health Trajectories of Refugees and Migrants
Annalisa Busetta1, Silvia Loi2,3, Anna-Kathleen Piereth2,4
- Department of Economics, Business and Statistics (SEAS), University of Palermo, Italy
- Max Planck Institute for Demographic Research
- Max Planck – University of Helsinki Center for Social Inequalities in Population Health, Rostock, Germany and Helsinki, Finland
- Institute of Environmental Medicine, Unit of Epidemiology, Karolinska Institutet, Sweden
Background: Due to the healthy immigrant effect immigrants are healthier than non-immigrant populations. This advantage is a long-lasting effect of positive immigrant selection. However, the positive selection may not hold true for refugees. In this paper, we aim to study the physical and mental health age trajectories of refugees and compare them with those of non-immigrants and first-generation immigrants.
Methods: Using data from the German socio-economic panel, we study physical and mental health age trajectories of immigrants and refugees aged 20-50. Our outcomes are the Physical Component Summary (PCS), and the Mental Component Summary (MCS). We apply random-effect models clustered by individuals to study age trajectories of refugee and immigrants’ health, contrasting them with the German-born population. We evaluate the role of education, and marital status and whether there are any differences in the age-related patterns of mental and physical health that might depend on the duration of stay of immigrants and refugees.
Results: Our findings show that refugees experience faster physical and mental health deterioration with age, compared to immigrants and to the German-born population. We also observe that mental health is better for long-term immigrants and refugees, compared to their shortly arrived counterparts. Education plays a key role in explaining physical health differences, but not mental health ones. Older refugees, recent arrivals, and refugee women display poorer physical and mental health than immigrants and non-immigrants. Long-term refugees show lasting physical health consequences, though their mental health improves over time.
Conclusions: Refugees are more vulnerable than immigrants, experiencing faster physical health deterioration. However, they also appear to be more psychologically resilient, showing improving mental health with longer duration of stay. The long-term effects of the refugee experience seem to contribute to persistent physical health issues, but with increasing duration of stay there is gradual improvement in mental health.
The effect of loneliness on depressive symptoms in the 65+ European population: A longitudinal observational study using SHARE data
Daniëlle Smit, Johan Rehnberg & Stefan Fors
Background: Loneliness has been found to be associated with an increased probability of depressive symptoms in various research studies. Yet, the relationship is complex, and it remains unclear whether loneliness is a cause for depressive symptomatology. This study aims to investigate the possible causal effect of loneliness in the loneliness-depressive symptoms relationship among 65+ older adults in Europe.
Methods: This study analyzed two waves of observational data (2015–2017) from the Study of Health, Aging and Retirement Europe (SHARE) (n = 6808 individuals) and tried to identify a causal effect of direct and indirect loneliness on depressive symptomatology by accounting for confounding of potential unmeasured factors underlying the relationship using an endogenous treatment-effects model.
Results: This study showed that there was a substantial positive association between loneliness in 2015 and depressive symptoms in 2017 in the sample. However, there was no support for the hypothesis that loneliness in 2015 was a cause of depressive symptoms in 2017. In addition, there was no evidence for unmeasured factors confounding the relationship.
Conclusions: Loneliness may not be a cause for depressive symptoms. These findings suggests that here might be other reasons why lonely individuals are at an increased risk of depressive symptoms. Therefore, a shift in focus when aiming to reduce depressive symptoms among lonely older adults may be warranted.
The association between socioeconomic status and dementia in the oldest old – a time series analysis between 2001 and 2022
Pauliina Halonen, Marja Jylhä & Linda Enroth
Background: Higher socioeconomic status (SES) is a protective factor against several age-related chronic diseases, including dementia. The number of people with dementia is increasing, making it a leading cause of disability and increased care needs in old age. Despite the strong association between SES and dementia, there is limited evidence on this association in the oldest old, who are at high risk due to their advanced age. This study integrates whether dementia prevalence varies between SES groups in the oldest old and how the differences have evolved during the 2000s.
Methods: We used the Vitality 90+ data from seven cross-sectional surveys conducted between 2001 and 2022 in Tampere, Finland. Participants were categorized into six SES groups according to their longest-held occupation: upper non-manual, lower non-manual, skilled manual, and unskilled manual workers, housewives, and unknown occupation. Dementia was defined from the survey and national registers. The association between dementia and SES was explored with rate differences over time.
Results: Data included 6,864 participants with median age 92 years. Initially, the most common occupational class was skilled manual workers, but this showed to lower non-manual workers in later surveys, with an increase in upper non-manual workers. Generally, upper and lower non-manual workers had lower dementia prevalence compared to skilled or unskilled manual workers, with the highest dementia prevalence often found in those with unknown occupation. The difference was smallest in 2001 (2.6 pp) and largest in 2018 (10 pp).
Conclusions: Socioeconomic differences in dementia prevalence exist in the oldest ages with slightly increasing trend over the 2000s. The increase in the number of the oldest old may contribute to the increasing SES differences, as this age group becomes less selected in terms of health and sociodemographic factors. We will extend our analysis to concern the changing socioeconomic structure of the study population.
Association between loneliness profiles and risk of dementia in the Paquid cohort
Yvanna Simon, Camille Ouvrard, Marie Sendra, Luc Letenneur, Valérie Bergua, Hélène Amiéva, Jean-François Dartigues and Karine Pérès
Inserm U1219, Team « ACTIVE » (Aging, Chronic disease, Technology, Disability and Environment)– BPH Research Center – University of Bordeaux
Background: At a time when “aging well” has become a societal priority, it is crucial to identify the levers for action that can be used effectively to meet the challenge of aging. For example, loneliness, which particularly affects the most vulnerable, especially the older persons, is worsening year on year at a worrying pace, accentuated by the health crisis and major family and societal changes. While its deleterious effects on health are increasingly well understood, its impact on cognition and dementia remains largely unexplored to date.
Objective. To estimate the risk of incident dementia associated to 4 different loneliness profiles.
Methods: A prospective analysis was carried out on the population-based cohort on aging (Paquid). Loneliness was assessed using the following item of the Center Epidemiologic Studies Depression scale (CESD): “In the past week, how often has the person felt lonely?”. Over the first 8 years of follow-up (4 waves), we defined four profiles of loneliness: never, transient, incident and chronic. We applied multistate models to analyze the relationship between loneliness and the risk of dementia over a 20-year period.
Results: In this study, 1 491 participants were included, 47.2% of whom had never suffered from loneliness, 23.0% from transient loneliness, 13.4% from incident loneliness and 16.4% from chronic loneliness. A total of 540 (36.2%) older adults developed dementia. Older adults suffering from chronic loneliness had a 48% higher risk of developing dementia than those who had never experienced loneliness (HR=1.48; IC=1.10-1.99). Transient and incident loneliness were not associated with a higher risk of dementia.
Conclusions: These findings underline the importance of informing public policy on this issue, given the public health and social cohesion stakes involved. As the feeling of loneliness is potentially reversible, it could provide an interesting lever for action to prevent dementia and promote healthy aging.
Key words: loneliness, dementia, older adults, cohort
Cognitive Performance in Late Midlife as a Predictor of Frailty over 17 Years: A Longitudinal Birth Cohort Study
Anna Tirkkonen, PhD1 , Markus J. Haapanen, MD, PhD2,3,4, Hanna Pajulammi, MD, PhD5 , Jenni Niku, PhD6 , Juulia Jylhävä, PhD4,7, Tuija M. Mikkola, PhD 2,8,9, Eero Kajantie, MD, PhD9,10,11,12, Johan G. Eriksson, MD, PhD2,3,13,14, and Mikaela von Bonsdorff, PhD1,2
1Faculty of Sport and Health Sciences and Gerontology Research Center, University of Jyväskylä, Finland, 2 Folkhälsan Research Center, Helsinki, Finland, 3Department of General Practice and Primary Health Care, University of Helsinki and Helsinki University Hospital, Finland,4Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden,5 Department of Geriatric Medicine, Wellbeing Services County of Central Finland, Jyväskylä, Finland, 6Faculty of Sport and Health Sciences, University of Jyväskylä, Finland,7Faculty of Medicine and Health Technology and Gerontology Research Center, Tampere University, Finland, 8Clinicum Faculty of Medicine, University of Helsinki, Helsinki, Finland, 9Population Health Unit, Finnish Institute for Health and Welfare, Helsinki, Finland, 10Clinical Medicine Research Unit, MRC Oulu, Oulu University Hospital and University of Oulu, Oulu, Finland, 11Department of Clinical and Molecular Medicine, Norwegian University for Science and Technology, Trondheim, Norway, 12New Children’s Hospital, Helsinki University Hospital and University of Helsinki, Helsinki, Finland, 13Singapore Institute for Clinical Sciences, Agency for Science, Technology, and Research, Brenner Centre for Molecular Medicine, Singapore, 14Department of Obstetrics and Gynaecology and Human Potential Translational Research Programme, Yong Loo Lin School of Medicine, National University of Singapore, Singapore.
Background: Cognitive decline and higher levels of physical frailty are associated among older adults. However, it remains unclear whether this potential association is present in late midlife and whether cognitive performance in late midlife is associated with the development of frailty when frailty is assessed with the multidimensional frailty index (FI).
Methods: This study uses data from the Helsinki Birth Cohort Study participants with information on cognitive performance and FI (n=1279, age-range 57-70 years). Reaction time and response accuracy in simple reaction time, choice reaction time, working memory, divided attention, and associated learning were assessed using the CogState assessment in late midlife (mean age 61.5 ± 2.9). A 41-item FI was assessed at three time points across 17 years. The data were analyzed using generalized linear mixed models adjusted for sex, education, and smoking.
Results: The results showed that the FI level was 1.02-1.07-fold per each 1-standard deviation (SD) faster in reaction time (milliseconds) in late midlife in all domains except for simple reaction time in the adjusted models (p<0.001-0.04). Additionally, the level of FI was 0.95-fold per each 1-SD higher in choice reaction time response accuracy (%) among all participants (p=0.004) and 0.94-fold per each 1-SD higher in working memory response accuracy among women (p=0.005) in late midlife. Furthermore, the annual rate of change in FI was 0.97-fold per 1-SD increase in simple reaction time response accuracy among all participants (p=0.003) and 0.97-fold per 1-SD increase in divided attention response accuracy among men (p=0.002).
Conclusion: A consistent association was observed between slower cognitive reaction time and higher FI levels. However, the association between cognitive response accuracy and FI in late midlife was dependent on the cognitive domain being measured. Furthermore, we showed a cognitive domain and trait-specific association between cognitive performance in late midlife and the development of FI.
Modeling multimorbidity: exploring diseases interactions and their connection to frailty outcomes
Erika Banzato, Giovanna Boccuzzo
Department of Statistical Sciences, University of Padova, Padova, Italy
Background: Multimorbidity, defined as the coexistence of two or more chronic diseases, is a key factor in adverse health outcomes, increased healthcare costs, and clinical complexity. Understanding how diseases are interconnected and how they relate to frailty outcomes is essential for improving patient management. Graphical models provide a powerful framework to represent multimorbidity as a network, where nodes correspond to diseases and edges represent their associations.
Methods: We analyzed administrative health data from 213,689 individuals aged 65 or older from the Local Health Unit of the Padova province (Italy). We used chain graphical models to estimate and represent the multimorbidity network along with the relationships between diseases and outcomes. Links are estimated as conditional associations and represent the net pairwise effect between two variables while adjusting for the effect of other variables.
Results: The multimorbidity network highlighted strong associations between specific chronic diseases, revealing underlying patterns of interconnections. Some diseases emerged as central nodes, showing stronger links to frailty outcomes than others. Additionally, frailty outcomes themselves were highly interconnected and closely associated with mortality, suggesting that individuals experiencing these outcomes have a more fragile health status.
Conclusions: Graphical models, particularly chain graphical models, provide a novel approach to studying multimorbidity and its relationship with frailty outcomes. Our findings highlight the complex interplay between diseases and adverse health events, reinforcing the importance of identifying high-risk conditions and their role in shaping health trajectories in older adults.
Challenges in estimating multi-state expectancies from large-scale cross-longitudinal surveys
Nicolas Brouard
French Institute for Populations Studies (INED)
This proposition of communication introduces innovative ways to address the challenges in estimating multi-state expectancies using large-scale longitudinal surveys. We first discuss the challenges in designing longitudinal surveys as well as in examining prevalences of health outcomes considering the rapidly changing health states (i.e., the dynamics of health changing that are captured in the longitudinal surveys). We introduce the method of interpolated Markov chains in order to estimate the incidences of change between states based on the probability of change over a small time interval that is estimated by multinomial regression as a function of age and covariates. More importantly, we discuss how Powell’s algorithm which was useful in optimizing the multinomial regression function to estimate health expectancies for different states based on up to 30 covariates, did not converge when adding more states, covariates or interactions because the likelihood function to be maximized depends on more than 200 variables. We then test the Brent/Praxis algorithm which uses the principal component analysis method to estimate information matrix about probabilities of change, and found that the algorithm is also useful and time efficient.
A demographically motivated index of within-population health inequality
- Rustam Tursun-zade 1; rustam.tursunzade@ehu.eus 2. Timothy Riffe 1, 2, 3; tim.riffe@ehu.eus 3. Alyson van Raalte 4,5. vanraalte@demogr.mpg.de
- OPIK, Department of Sociology and Social Work, University of the Basque Country (UPV/EHU), Leioa, Spain, 48940, Bizkaia,6 Barrio Sarriena s/n. 2. Ikerbasque (Basque Foundation for Science). 3. Laboratory of Population Health, Max Planck Institute for Demographic Research. 4. Lifespan Inequalities Research Group, Max Planck Institute for Demographic Research. 5. Max Planck Helsinki Center for the Social Determinants of Health (MaxHel).
Background: Much has been done to measure within-population lifespan inequality, but less is known about within-population health inequality. Attempts to develop an incidence-based version of such an index are scarce compared to prevalence-based ones.
Methods: Inspired by e-dagger, we propose a new incidence-based index of within-population health inequality. Our index is neatly decomposable into the contributions of different life transitions to total inequality. We calculate transition probabilities from Spanish SILC data using the weighted multinomial logistic regression and derive the fundamental matrix from which we obtain all quantities required to calculate the index.
Results: We offer a demographically motivated, interpretable, and additively decomposable index of incidence-based within-population health inequality, which (i) does not assume the independence of survival and health status and (ii) distinguishes between reversible and irreversible conditions. We relate this index to classic lifespan inequality measures and entropy. The figure below shows the preliminary results of unhealthy life expectancy decomposition. As is the case for e-dagger, contributions to our index from mortality are necessarily positive, while transitions between health states can either increase or decrease inequality depending on the inequality measure.
Figure 1. A survey on living conditions was conducted by the Spanish National Statistical Institute (NIE). Data is shown for females in 2014. The figure shows the marginal (summed over age) contributions in years to Healthy, Unhealthy, and Overall inequality made by the underlying transition.
Conclusions: We offer a synthetic measure of within-population health inequality that can be calculated directly from standard multistate data constructs. The new index measures the power of each transition in state-specific and total life inequality. We consider our index demographically motivated because it can be verbalized as the expected number of years gained or lost of years lived in a state due to each transition. This is the multistate analog of lifetable e-dagger, and we conjecture that it will also relate to a multistate notion of lifetable entropy.
Measuring frailty in the elderly: an indicator based on a combination of classifiers
Pietro Belloni1, Sara Rebottini2,3
1Department of Statistical Sciences, University of Padua, Padua, Italy
2Department of Statistics, Computer Science, Application (DiSIA), University of Florence, Florence, Italy
3Faculty of Education, Free University of Bozen, Bozen, Italy
Background: The ageing population imposes increasing pressure on healthcare systems, requiring more extensive care for older people and early identification of frail individuals to reduce the risk of adverse health events. The aim of this work is to develop an indicator to assess the frailty level of each individual using the administrative health database of ULSS6 Euganea, an Italian healthcare local authority.
Methods: Given the multidimensional nature of frailty, we adopted a multi-outcome approach, considering six outcomes: death, emergency room visits with code red (maximum priority), hip fracture, hospitalization, disability onset, and dementia onset. After selecting a subgroup of frailty determinants for each adverse event using a gradient boosting approach, six classification rules were estimated through outcome-specific logistic regression models. The frailty indicator was created by combining these classification rules, weighted according to their individual predictive capacity with an innovative statistical technique.
Results: The indicator shows good performance across all outcomes and allows for the use of different subgroups of frailty determinants specific to each outcome, including the subject’s gender, which is not used by other state-of-the-art indicators.
Conclusions: This new frailty indicator is capable of capturing the multidimensional nature of frailty in the elderly. It is based on a combination of classifiers of adverse health events, which in turn are based on frailty risk factors. Given the administrative nature of the data used, it can easily be deployed to health authorities who want to keep track of the frailty of the elderly.
Keywords: Health in the elderly, Frailty indicator, Ageing, Determinants of health
Birth Cohort Contributions to Stalling U.S. Mortality Improvements
Leah Abrams,1 Octavio Bramajo, 2 Alyson Van Raalte, 3-5 Mikko Myrskylä, 3-6 Neil Mehta7
- Department of Community Health, Tufts University, Medford, MA, USA
- Sealy Center on Aging, University of Texas Medical Branch, Galveston, TX, USA
- Lifespan Inequalities Independent Research Group, Max Planck Institute for Demographic Research, Rostock, Germany
- Max Planck – University of Helsinki Center for Social Inequalities in Population Health (MaxHel Center), Rostock, Germany
- Max Planck – University of Helsinki Center for Social Inequalities in Population Health (MaxHel Center), Helsinki, Finland
- University of Helsinki, Helsinki, Finland
- Department of Epidemiology, School of Public and Population Health, University of Texas Medical Branch, Galveston, TX, USA
Background: In 2010, U.S. life expectancy plateaued after decades of consistent improvements. While previous research has shown that various age groups and causes of death contributed to the stalled improvements, it remains unclear whether this trend stems from period-specific contextual effects, cohort-specific patterns that impact health outcomes, or a combination of the two.
Methods: Using the Human Mortality Database Cause of Death data, we analyzed cohort mortality trends among adults aged 30-89 years in the United States from 1979 to 2019. We constructed lexis surfaces to examine rates of mortality improvement (ROMIs) across all-cause mortality and three major cause-of-death categories: cardiovascular diseases, cancers, and external causes.
Results: All-cause mortality increased among cohorts born in the 1950s compared to prior cohorts, and then it increased even more among the most recent cohorts, born 1970-1989. Cardiovascular disease mortality displayed adverse trends in those same cohorts but also exhibited a notable period pattern – improving mortality in 2000-2010 and increasing mortality post-2010. Cancer mortality showed consistent improvements for cohorts born after 1920, who smoked less than pre-1920 cohorts, suggesting a progressive mortality decline. However, colon cancer presented a small mortality increase for cohorts born since 1960. While traffic accidents and homicides demonstrated period trends of increasing mortality beginning around 2010, drug poisoning and suicides were predominantly associated with cohort patterns, consistent with prior research on external causes. Additional tests examining 2019-2023 revealed increasing mortality across causes and cohorts in the years affected by the COVID-19 pandemic.
Conclusions: Our findings suggest that the recent stall in U.S. life expectancy improvements is afflicting many birth cohorts and causes of death, indicating a pessimistic outlook for the future of U.S. life expectancy. Nonetheless, the strong cohort patterns identified in our analysis signal that cohort-specific behavioral and social patterns are shaping trends in population health.
Re-estimating Urban-Rural Disparities in Disability Life Expectancy Among Older Population in China Based on Bayesian Methods
Li Yichao, Mi Hong
Shanghai Administration Institute, China School of Public Affairs, Zhejiang University, China Corresponding Email: spsswork@zju.edu.cn
Background: Urban-rural disparities in disability life expectancy (DLE) are critical for assessing long-term care needs and implementing the national strategy to actively address population aging. Current research lacks empirical evidence regarding urban-rural disparities in DLE among Chinese older populations with varying degrees of disability.
Methods: The Bayesian multi-state model with informative priors is constructed to characterize the health transition patterns of urban and rural older populations after adjusting for mortality, based on data from China Health and Retirement Longitudinal Study (CHARLS) 2011–2020 and the Chinese Longitudinal Healthy Longevity Survey (CLHLS) 2008–2014. It further measures DLE and life expectancy with varying degrees of disability.
Results: (1) In 2020, urban older individuals exhibited significantly higher DLE compared to their rural counterparts, driven by disparities in severe DLE. (2) Urban older individuals have a higher proportion of severe DLE, while rural older individuals have a higher proportion of mild DLE. (3) Healthy life expectancy among urban older individuals with healthy status exceeds that of rural counterparts.
Conclusions: Efforts should be made to strengthen the primary healthcare systems in both urban and rural areas to reduce the proportion of life expectancy spent with disabilities. An equitable financing system for long-term care insurance should be established to address urban-rural disparities.
Keywords: Disability Life Expectancy; Health Transition Pattern; Urban-Rural Disparities; Bayesian Multi-State Model; Older Population
Cohort mortality forecasts indicate signs of deceleration in life expectancy gains
José Andrade, Max Planck Institute for Demographic Research
Carlo Giovanni, Camarda Institut national d’études démographiques
Héctor Pifarré i Arolas, La Follette School of Public Affairs, University of Wisconsin – Madison Center for Demography and Ecology Center for Demography of Health and Aging
The dramatic improvements in mortality over the last century have contributed to a sustained increase in life expectancy. However, it remains unclear whether recent increases in life expectancy follow a linear trend or show signs of deceleration. From this perspective, a cohort-based approach—which captures the actual survival experiences of individuals born in a given year—offers valuable insights into this issue. Using data from the Human Mortality Database (HMD), this study estimates cohort life expectancy for individuals born between 1939 and 2000, applying multiple mortality forecasting methods alongside information from the United Nations World Population Prospects 2024 (WPP). Additionally, decomposition techniques help identify key age groups driving changes in life expectancy. Our findings suggest a significant deceleration in cohort life expectancy gains for currently living cohorts, disrupting the steady improvements observed for earlier cohorts (1900–1938). While the mortality improvement rate was 0.52 for the 1900–1938 cohorts, it declined to 0.20 for those born between 1939 and 2000. This slowdown is primarily driven by reduced mortality improvements at very young ages, a trend already evident in our data. Importantly, our results indicate that this deceleration is not due to systematic underestimation by forecasting methods but reflects an emerging reality. In many countries, mortality improvements at 1 younger ages appear to be reaching a lower bound. This raises a crucial question: Can future advancements reverse this slowdown, or is the era of steady life expectancy gains coming to an end?
Catching up with stagnation: cause-specific dynamics of change in life expectancy at age 65 in the United States, Canada and Latin America, 2000-2019
Octavio Bramajo (1) Neil Mehta (2) Mikko Myrskylä (3)
1 Sealy Center on Aging, University of Texas Medical Branch. Galveston, United States
2 Max Planck Institute for Demographic Research, Rostock, Germany
Background: Recent focus on U.S. longevity stagnation at higher ages has focused on comparisons with other high-income countries, with less attention paid to its performance relative to peer nations in the Americas.
Objectives: This study examines changes in life expectancy at age 65 (LE65) in the United States, Canada, and seven Latin American countries—Argentina, Brazil, Chile, Costa Rica, Colombia, Mexico, and Peru—between 2000 and 2019, disaggregated by sex and cause of death.
Research Design and Methods: This observational study used United Nations World Population Prospects data and World Health Organization cause of death data, applying life table methods and linear integral decomposition methods to assess the contribution of causes of death to changes in LE65.
Findings: The United States and Canada experienced substantial increases in LE65 during 2000- 2009 in both sexes, with gains ranging from 1.35 to 2 years. These gains were primarily driven by reductions in cardiovascular disease (CVD) mortality, which contributed 1.2 years to LE65 in both countries, widening the gap with Latin American countries. From 2010-2019, the contribution of CVD to LE65 more than halved, and total gains in LE65 stagnated between 0.7 and 1.1 years. As a result, the gap with most Latin American countries mostly stagnated or even narrowed. The best performer, Chile, surpassed U.S. LE65 during the observation period, while Mexico’s LE65 declined.
Implications: The recent stagnation in LE65 observed across the Americas during 2010-2019 highlights the need for developing national and regional strategies aimed at reducing cardiovascular mortality in the region.
The Role of Rising Obesity in the Stagnation of U.S. Life Expectancy
Andrew C. Stokes, PhD Boston University School of Public Health Email: acstokes@bu.edu
Jennifer Beam Dowd, PhD University of Oxford Email: jennifer.dowd@demography.ox.ac.uk
Virginia W. Chang, MD, PhD New York University School of Global Public Health Email: vc43@nyu.edu Rafeya V. Raquib Boston University School of Public Health Email: rraquib@bu.edu
Neil K. Mehta, PhD The University of Texas Medical Branch Email: nemehta@utmb.edu
Background: U.S. life expectancy improvements have slowed considerably since 2010, with cardiovascular disease (CVD) mortality trends playing a major role. Obesity prevalence has steadily increased since the 1980s, yet its contribution to recent mortality trends remains incompletely understood. This study quantifies the impact of rising obesity on all-cause mortality and life expectancy trends.
Methods: We used data from the National Health and Nutrition Examination Survey (NHANES) linked to mortality records from 1989 to 2019. Cox proportional hazards models were employed to estimate the impact of lifetime maximum body mass index (BMI) on mortality trends, with simulations used to assess excess deaths and life expectancy under counterfactual obesity scenarios.
Results: Rising obesity levels slowed the annual rate of mortality decline by 0.65%, accounting for a 29% reduction in overall mortality improvements. In 2019 alone, obesity contributed to 323,375 excess deaths, and between 1989 and 2019, it accounted for 4.66 million cumulative excess deaths. If obesity levels had remained at 1989 levels, life expectancy at age 40 in 2019 would have been 1.5 years higher.
Conclusions: These findings underscore obesity’s critical role in the stagnation of U.S. life expectancy and highlight the urgent need for policy interventions targeting obesity prevention and management to mitigate further mortality stagnation.
Burden Reduction in a Social Safety Net Program Reduces Mortality
Tracee Saunders1, Jeffrey Hemmeter2, Pamela Herd3, Sebastian Jilke4, Donald Moynihan3, Elana Safran5
1 Department of Political Science, Penn State University
2 Office of Research, Demonstration, and Employment Support in the U.S. Social Security Administration; Washington, D.C.
3 Ford School of Public Policy University of Michigan
4 McCourt School of Public Policy, Georgetown University
5 Office of Evaluation Sciences in the U.S. General Services Administration; Washington, D.C.
Burdensome administrative procedures, ranging from confusion about eligibility guidelines to complicated paperwork. But what happens when eligible individuals don’t take up benefits for which they’re eligible? While we know the short term impacts, such as forgone income and benefits, the long term consequences of these losses remain poorly understood. We examine the mortality impacts of burden reduction – particularly learning costs associated with understanding eligibility and benefits – for the Supplemental Security Income (SSI) program. Using data from a large-scale randomized controlled trial (N=4,016,461) in which informational letters were sent to older adults likely eligible, but not enrolled in SSI, we estimate the mortality effects of older adults’ subsequent enrollment in SSI. The intervention increased SSI awards by an estimated 1.8 percentage points (or a 340 percent increase from a baseline enrollment rate of 0.5 percent). Among those who enrolled in SSI, we estimate a reduction in mortality (Hazard Ratio=0.6101, 95% CI=0.5604–0.6598). These results demonstrate that burden reduction may not only increase access to social welfare programs like SSI in the short term but may have significant downstream impacts by reducing beneficiaries’ mortality risk.
Developing European statistics on long-term care
Giampaolo Lanzieri
Eurostat
Background: To underpin with high-quality statistical information the policy initiatives in the field of long-term care (LTC) in the European Union (EU), the European Commission Expert Group of the European Directors of Social Statistics has set up a Task Force on Long-Term Care statistics (TF LTC). The TF LTC has started in February 2023 and it is planned to work until end 2025. Additional initiatives are being taken in parallel to further improve the statistical information on LTC from existing EU social surveys.
Objectives: The main aim of the TF LTC is to develop consistent EU statistics on long-term care by providing methodological guidance in various domains of long-term care statistics.
Methods: The TF LTC is chaired by Eurostat and is currently composed of experts from 13 EU Member States and 6 international organizations. The TF LTC meets to discuss proposals striving to offer a methodologically sound basis for sustainable LTC statistics, minimising the burden on national data providers and complying with the European Statistics Code of Practice. The TF LTC will consolidate the discussions on the statistics for the various domains of long-term care in a final report with proposed methodological guidelines, that will be submitted for endorsement to the parent Expert Group.
Results: Various LTC topics have been discussed such as: definition of long-term care for statistical purposes, estimation of the population need of long-term care, estimation of LTC workforce, statistics on informal care, guidelines on LTC expenditure (health and social), LTC recipients and beneficiaries. The discussions have been supported by over 20 working papers produced on different LTC subjects by the members.
Conclusion: We provide an overview of the activities, both carried out and forthcoming, to develop a methodological basis for European statistics on long-term care by a dedicated task force composed of experts from various countries and international organizations.
Rehabilitation needs of persons living with dementia.
Outi Mäenpää 1, Johanna Edgren2, Mari Aaltonen2, Jari Pirhonen1 and Jenni Kulmala1
1 Faculty of Social Sciences (Health Sciences) and Gerontology Research Center, Tampere University, Finland
2 Finnish Institute for Health and Welfare, Helsinki, Finland
Background: Rehabilitation is an umbrella term for actions that aim to optimize autonomy, life management, functioning, and well-being of older people. Physical rehabilitation supports physical functioning and daily activities. However, rehabilitation is an underutilized resource in dementia care and health care professionals describe challenges in implementing rehabilitation. Also, scientific knowledge on individual rehabilitation needs is relatively scarce. The aim of this qualitative study was to investigate rehabilitation needs of older home care clients with dementia.
Methods: The study followed a phenomenological approach to provide deep understanding of the rehabilitation needs of home care clients with dementia. Participants were recruited from the organization that provides home care for older people. In total, 12 home care clients with the mean age of 83 years with mild to moderate dementia participated in this study. Data were collected through face-to-face focus group interviews and analyzed using inductive reflexive thematic analysis.
Results: The study participants described the rehabilitation needs through four main themes: 1) Coordination of rehabilitation 2) Social support with distributed agency 3) Individual tailoring of rehabilitation, and 4) Consideration of individual’s characteristics. Rehabilitation was seen as physical activity and physical rehabilitation. Participants felt that rehabilitation could take place both through self-directed activities, such as going for a walk, or participating in rehabilitative activities like group exercises, as well as through planned activities carried out with a rehabilitation professional or home care staff.
Conclusions: This study identified several key elements for successful rehabilitation described by home care clients with dementia. The results highlight the importance of coordination, individually tailored rehabilitation and social support with distributed agency. These results can help health care professionals and family members to implement rehabilitation successfully for older home care clients with dementia.
The possibilities of home care clients with dementia influencing their care through goal setting
Anne-Marie Mäkelä1,2, Jutta Pulkki2, Jokke Häsä 1 and Mari Aaltonen1,2
1Finnish Institute for Health and Welfare, Helsinki, Finland
2Faculty of Social Sciences and Gerontology Research Center (GEREC), Tampere University, Finland
Background: To provide person-centred care, people should have the opportunity to influence their care. Influencing care is part of participating in one’s own life. Furthermore, participation is strongly related to well-being and perceived good life. Despite the increasing number of people with dementia, research on how people with dementia participate in decision-making about their care is scarce. We explore to what extent home care clients with dementia express the goals of care and what type of care goals they set.
Methods: Resident Assessment Instrument (RAI) register data were collected nationally across Finland. The sample included a total of 38,901 home care clients aged 65 or above who had participated a first RAI assessment round in 2023. Of these 44% had a dementia diagnosis (n = 17,124). First, statistical descriptive methods were used to analyse the prevalence of goal setting and the change of this prevalence with declining cognition. Second, the content of goals was analysed using qualitative content analysis.
Results: Of home care clients with dementia, 61% (n =10,342) had set at least one goal for their care. The results show that clients with more impaired cognition set goals less often. The most common goals set by home care clients with dementia were the desire to live at home for as long as possible, access to adequate home care services, and maintaining functional capability and ability to perform daily activities.
Conclusions: These preliminary results suggest that home care should promote multifaceted participation methods for people with dementia to enable them to participate in their own care. Identifying care goals of home care clients with dementia is crucial regarding person-centered care and participation. In addition to identifying goals, home care should strive to include these goals in the implementation of care.
The Long-Term Impacts of Medicaid Exposure in Early Childhood on Aging
Puneet Kaur Chehal
Emory University, US
Medicaid is the U.S. subsidized public health insurance program for low-income uninsured populations. Over time, the program has grown in eligibility to include poor, childless, non-elderly adults whereas historically Medicaid exclusively served pregnant women and children. Researchers have found exciting evidence that Medicaid exposure during sensitive developmental periods influences long term health outcomes but have yet to understand how outcomes improve. In this study, we explore whether improvements due early Medicaid exposure is linked to persistent increases in endowed health over the life course of aging cohorts or whether it largely delays onset of chronic disease, suggesting slower rates of aging. We exploit the phased rollout of Medicaid across states between 1966 and 1970 which resulted in varying levels of program exposure for different birth cohorts, now in adulthood. We use the Panel Study of Income Dynamics (PSID), a semi-annual national longitudinal survey, which allows us to observe year and place of birth to determine early Medicaid exposure and onset of chronic diseases. We use differences in the timing of disease onset between treatment and control cohorts to proxy for differences in aging. In previously reported research findings using the PSID, Medicaid access during early childhood (ages 0–5) leads to an average 3.35 standard deviation improvement in health during adulthood for targeted populations (ages 25–54). Health is measured as a composite disease index reflecting information on high blood pressure, diabetes, heart disease/heart attack, and obesity. An additional 10 years of data are available since these results were published allowing us to explore whether early Medicaid exposure prevented disease altogether or delayed onset and its effect on incidence rates over the life course.
The Unequal Healthy Life Expectancy of men and women. Can education mitigate these inequalities?
Aïda Solé-Auró1
1DemoSoc Research Group, Department of Political and Social Sciences, Universitat Pompeu Fabra, C/ Ramon Trias Fargas, 25-27, 08005, Barcelona, Spain ORCID: 0000-0003-3726-2509
Iñaki Permanyer2,3
2Centre for Demographic Studies (CED), Universitat Autònoma de Barcelona, Carrer de Ca n’Altayó, Edifici E2, 08193, Bellaterra/Barcelona, Spain
3ICREA, Passeig Lluís Companys 23, Barcelona, Spain
That women live longer than men is well know since many decades, as the mortality experiences of males and females differ substantially. Studies on gender differences in morbidity, per contrary, report that women are in worse health than men (Crimmins et al. 2011) and that women spend a higher proportion of their life in poorer health (Luy and Minagawa 2014).
The overall aim of this work is to identify health inequalities. To do so, we assess the contribution of mortality and disability to the gender differences in (un)healthy life expectancy by different education levels based on register diagnosed chronic conditions in Catalonia (Spain). To do so, we use data from the HEALIN cohort (Solé-Auró et al, 2024), a longitudinal population-based dataset encompassing over 1.5 million individuals (accounting for 22% of the total Catalan population) from 2005 up to 2021. This sample is representative in terms of age, sex, and region. All data within the HEALIN cohort are anonymized and comply with Spanish regulations regarding observational studies. Information on the diagnosis history of a large number of chronic diseases, mortality and some demographic characteristics are available. The HEALIN cohort does not provide information on disability and this one comes from the survey of health in Catalonia (ESCA).
We use the death counts and population exposures from the HEALIN cohort. We implement the Sullivan method in different age groups (30-49; 50-64; 65+) and in 2 time periods (2011, 2021) by education level (low, medium and high). In addition, we apply a decomposition method (Adreev et al. 2002) to identify the exact contributions of: a) differences in age-specific mortality rate; b) differences in age-specific disability rates. This analytical strategy aspires to impart vital insights into enigmatic gender inequalities undermining population health.
We observe a persistent gender gap in life expectancy over time, HLE has increased, while UHLE has reduced. Mortality contributes more to the gender gap in HLE than disability but this contribution seems to be reducing over time (men’s mortality seems to be improving). The disability effect mostly explains the reduction of the gender gap in UHLE50. We are working on the results by level of education.
The expected results from this research will be useful for scholars worldwide investigating contemporary health dynamics. It will provide a comprehensive description of the gender health inequalities by education level in Catalonia (Spain), identifying factors contributing to shorter life expectancy as well as life lived in good health. In general, our findings will be useful to guide the development of public policies, as well as promoting territorial and social cohesion.
Keywords: gender differences in (un)health expectancy, education level, morbidity
Association of sleep and nap duration with total life expectancy, and life expectancy with and without cardiovascular diseases among older adults: A longitudinal population-based study
Li Ruoxuan Rosalyn1, Abhijit Visaria2, Stefan Ma3, Rahul Malhotra2,4
1Duke-NUS Medical School, Singapore 2Centre for Ageing Research and Education, Duke-NUS Medical School, Singapore 3 Ministry of Health, Singapore 4Health Services and Systems Research, Duke-NUS Medical School, Singapore
Background: Sleep and nap durations have been linked with many health outcomes, yet their association with health expectancy remains understudied. We investigated the association of nocturnal sleep and daytime nap durations with total life expectancy (TLE), and life expectancy (LE) with and without cardiovascular diseases (CVD), among older adults in Singapore.
Methods: Representative longitudinal data (three waves; 2009-2015) of community-dwelling Singapore residents aged ≥60 years (n=3452) was used. We categorized sleep duration as short (≤6 hours), recommended (7-8 hours), and long (≥9 hours), and nap duration as none, short (≤1 hour) and long (>1 hour). We applied multistate life table methods to estimate TLE, and LE with and without CVD, by sleep duration and nap duration, adjusting for a range of socio-demographic and health variables.
Results: At ages 60, 70, and 80 years, there was no difference in TLE or LE without CVD among those with short, recommended and long sleep. However, long (versus recommended) sleep was associated with a shorter LE with CVD. At ages 60 and 70, the proportion of remaining life without CVD was higher, and with CVD was lower, among those with long (versus recommended) sleep. At ages 60, 70, and 80 years, long nap was associated with a shorter TLE and LE without CVD, versus none or short nap. LE with CVD or proportion of remaining life with and without CVD did not differ by nap duration.
Conclusion: Our findings – similar TLE or LE without CVD between those with suboptimal and recommended nocturnal sleep duration, but lower TLE and LE without CVD among those with excessive daytime napping – reflect the complex relationship between sleep and nap durations and life and health expectancy outcomes. Further research is needed to elucidate underlying mechanisms and guide recommendations for optimal nocturnal sleep and daytime nap durations in older adults.
Extent of the health disadvantage of Roma people in Europe: Indirect estimations for 17 countries
Marc Luy, Tianyu Shen, Paola Di Giulio, Lina Lasar, Tamara Vaz de Moraes Santos
Vienna Institute of Demography of the Austrian Academy of Sciences
According to the most widely cited and generally accepted data from the Council of Europe, life expectancy of Roma people in Europe is between 10 and 15 years lower than that of the general population. However, all existing reports are substantially limited in the comprehensiveness and quality of the available data. Consequently, it is not clear whether the disadvantage in the number of life years is actually that high. The aim of this paper is therefore to provide estimates derived by a different approach and for national subpopulations of Roma people for which no estimates exist so far. We use two Roma surveys from 2019-21 which were conducted by the European Union Agency for Fundamental Rights in 17 European countries and allow the estimation of life expectancy on the basis of proportions of still living parents with the “Orphanhood Method”. The results confirm the large disadvantage of Roma people in life expectancy. The differences in life expectancy at age 30 between national populations and Roma people vary from 4.3 years among men in Portugal to 13.4 years among women in Italy. By including also information on child mortality, the analyses can be extended to provide estimates of life expectancy at birth, using the “flexible two-dimensional mortality model”. The resulting differences in life expectancy at birth between national populations and Roma people vary from 7.0 years among men in Romania to 17.7 years among women in Italy.
Healthy Working Life Expectancy: Measurements and Socio-Economic Inequalities
Alessandro Feraldi and Christian Dudel
Sapienza University of Rome, Italy
Background: Population aging in high-income countries poses significant challenges to the sustainability of social security and pension systems, prompting policies aimed at extending working life. This may have implications for both social and health outcomes. Healthy working life expectancy has emerged as a key metric for understanding the interplay between health and labor force participation in later life. However, the choice of health and work definitions in healthy working life expectancy analyses may substantially influence findings. For example, the magnitude of socio-economic inequalities may vary by health and work definitions, necessitating a deeper investigation into these measures.
Methods: Using U.S. Health and Retirement Study data, and employing a multistate lifetable approach to model individual life courses, we estimate healthy working life expectancy at age 50. We compare healthy working life expectancy based on multiple definitions of health (e.g., physical limitations, self-rated health, and chronic conditions) and employment (e.g., selfreported employment status and labor income) and we explore temporal trends (comparing 2000–2009 to 2010–2020) and disparities by gender and education.
Results: Results will show how different definitions of health and work yield varying estimates of healthy working life expectancy and how disparities in healthy working life expectancy— such as those based on gender and education—differ depending on the measurement approach.
Conclusions: By examining how different definitions of health and work influence findings, we provide guidance for researchers who plan to analyse healthy working life expectancy and can choose between different measurement of health and/or work. Moreover, our findings will help to conduct meaningful comparisons of findings from the literature based on different measurements.
The Years Left Abroad: The impact of emigration on longevity in the Republic of Serbia
Milos Milovanovic, Linköping University, Sweden.
Susanne Kelfve, Jonas Wastesson
Aim: The study aims to investigate the impact of emigration on life expectancy at birth (LE) in Serbian municipalities and to propose a machine learning technique to address missing data issues.
Background: Differences in LE at birth are related to the disparity in economic development. Additionally, we discuss how migration could also, to some extent, reveal inequalities in health (Uprety, 2019; Uprety & Schhuhmann, 2020). The rationale behind this is that due to the selectivity of emigrants (e.g., Lee, 1966; Turra and Elo, 2008), municipalities with a higher share of those who are abroad are going to have lower LE on the aggregate level. Precisely, the municipality level of LE is simply a reflection of the average health status of the remaining population.
Methods: In the case of Serbia, the only source is the decennial population census, and the latest information on population abroad across municipalities is for 2002 and 2011. Therefore, we will use three approaches with different assumptions regarding the change between two known values: linear interpolation; more flexible splines interpolation; and a machine learning technique known as random forest. Given that the study examines clustered data i.e., repeated measures of LE in the same municipality over time further nested into three regions, we use a mixed multilevel (hierarchical) model.
Results: The use of Random Forest enabled a higher level of complexity regarding the relationship between life expectancy and migration. Our model captures a non-linear relationship between life expectancy at birth on one side and both population abroad and return migration on the other.
Conclusions: The research indicates that the relationship between health outcomes and migration is complex, suggesting a nuanced approach to policies aimed at mitigating the adverse effect of extensive emigration.
Assessing Health Inequalities. Age and Cohort Trajectories Among Older Adults in Europe
Madelin Gómez-León & Aïda Solé-Auró
DemoSoc group, Department of Political and Social Sciences, Pompeu Fabra University
Background and Objective: Health inequalities among older adults in Europe present a pressing concern, as disparities in health outcomes are influenced by various factors such as age, gender, socioeconomic status, and geographic location. Understanding the trajectories of these inequalities across different cohorts is crucial with adult mortality steadily postponed to advanced ages, thus concerns have risen about well-being and the quality-of-life gain in later life. In this study we investigate and update health trends among older adults from five distinct European birth cohorts, assessing the health gender gap while evaluating the changes observed across various age groups and generations.
Research Design and Methods: Multivariate logistic regressions are estimated to disentangle cohort and age effects on trends in fair/poor self-perceived health (SPH), Global Activity Limitation Indicator (GALI) and limitations with Activities of Daily Living (ADLs) across birth cohorts. We follow older adults aged 50 to 84 years in 19 European countries over a 18-year period using data from the Survey of Health, Ageing and Retirement in Europe (SHARE).
Results indicate a notable improvement in the three health indicators analyzed across recent birth cohorts. Despite minor cohort differences persisting after accounting for age and sociodemographic variables, improvements were evident. Additionally, although women still report poorer health compared to men, the gender disparity in all health metrics has decreased from older to younger cohorts. The only exception to this trend is observed among males in the Middle and Younger cohorts, who exhibit a greater likelihood of experiencing limitations in Activities of Daily Living (ADL) than their females’ counterparts.
Keywords: health trends, gender health gaps, cohort, Europe, SHARE
The role of socioeconomic resources for adhering to the “Blue Zone” lifestyle in contrasting regions in Western Finland and northern Sweden
Sarah Åkerman1, Dorly Deeg2, Birgitta Olofsson3, Yngve Gustafson4, Johan Niklasson4, Erika Boman5 & Fredrica Nyqvist1
1.Social Policy, Åbo Akademi University, Vasa, Finland 2. Epidemiology and Data Science, Amsterdam University Medical Center, Amsterdam, The Netherlands 3. Nursing, Umeå University, Umeå, Sweden 4. Geriatrics, Umeå University, Umeå, Sweden 5. Nursing, Åland University of Applied Sciences, Mariehamn, Finland
Background: Blue Zones are regions in the world known for extreme longevity. Longevity in the Blue Zones is tentatively explained by the inhabitants’ shared lifestyle, including, amongst other things, a focus on the family, natural movement in everyday life, sense of purpose, and strong community support. The Blue Zone lifestyle is thus adapted both on an individual and community level. Little is known about the role of socioeconomic resources for adhering to alleged healthy lifestyle patterns in old age and whether there is regional variance in these associations. The aim of this study is therefore to investigate the role of socioeconomic position for adherence to the Blue Zone lifestyle in six contrasting regions in Western Finland and Northern Sweden.
Methods: Data is extracted from the multidisciplinary Gerontological Regional Database survey collected in 2021/2022 (N=11 984) in Ostrobothnia, South Ostrobothnia, and Åland, Finland, and Västerbotten, Sweden. Ostrobothnia is divided into two regions based on ethnolinguistic affiliation (Swedishspeaking and Finnish-speaking). Västerbotten is divided into an urban and a remote area. The study conducts regression analyses and investigates interaction effects of education, personal income and region in their effect on adherence to the Blue Zone lifestyle.
Results: Descriptive analyses suggest that the sample on Åland has higher personal income, while the sample in Västerbotten has higher educational level. Preliminary analyses, controlling for sociodemographic resources, show significant regional differences with Swedish-Speaking Ostrobothnia adherring to the Blue Zone lifestyle the most. This region is characterised by high levels of civic engagement, family focus, and religiousness, while regions on Åland and Sweden represent a more secular lifestyle and an appealing physical environment.
Conclusions: This study will add insights on the interplay between individual socioeconomic resources and environmental characteristics in contrasting areas for adopting an alleged healthy lifestyle in old age.
Risk profiles for serious health events according to the first step of the WHO ICOPE approach
Karine Pérès PhD, Ivane Koumetio-Jiatsa MSc, Jeanne Bardinet PhD, Antoine Gbessemehlan PhD, Achille Tchalla PhD MD, Catherine Helmer PhD MD, Luc Letenneur PhD
Inserm U1219 – University of Bordeaux
Background: In response to the aging population, the World Health Organization (WHO) proposes the operationalization of the concept of healthy aging through the ICOPE program, focusing on the early detection of Intrinsic Capacity (IC) declines. The present study aimed at estimating the risks of four adverse outcomes (death, IADL-disability, institutionalization and dementia) associated with IC impairment and to identify the most at-risk profiles (combination of IC impairments).
Methods: Study sample was drawn from two French cohorts of older adults (Three-City Bordeaux and AMI) where the 6 domains of IC (cognition, nutrition, psychology, hearing, vision, and mobility) have been assessed. Three groups of IC impairment have been defined: no IC impairment (IC0), single impairment (IC1) and two impairments or more (IC2+). Cox models were first used to estimate the risks of adverse events associated with these 3 groups. Secondly a conditional inference tree followed by Cox models were used to determine the highest-risk profiles for each event.
Results: Among the 2,964 included participants (median age: 74.7 years, 53.8% of women), 24.9% had one IC impairment (IC1) and 62.3% had two impairments or more (IC2+). Having multiple impairments (IC2+) was associated with higher risk of all outcomes (hazard ratios (HR) between 2 and 3 depending on the considered events), whilst single one was not, except for IADL-disability (HR=1.46, 95%CI=[1.14-1.87]). Some specific profiles were associated with particularly high risks; mobility, cognitive and visual impairments being the most predictive domains.
Conclusions: This study confirms the public health relevance of this approach, but given the high prevalence of individuals detected at the first step of the ICOPE strategy (more than 87%), a refinement of the screening threshold is required, in terms of number of impairments and/or combination of some impairments at greatest risks of events.
Socioeconomic and institutional conditions associated to variance of age at death
Gilbert Brenes-Camacho
Centro Centroamericano de Población y Escuela de Estadística, Universidad de Costa Rica
The variance of age at death is assumed to diminish as health improves in a population. Medical technology, sanitation infrastructure and better healthy habits endue deaths to get concentrated in late life. However, different social, economic and institutional conditions at early and old ages could determine inequalities in the timing of death. The paper explores how socioeconomic status predicts both the mean age at death and its variance using accelerated failure time models with equations for an ancillary parameter. I use three datasets harmonized through the Gateway to Global Aging Data: the Health and Retirement Study (HRS) Rand dataset, the Mexican Health and Aging Study (MHAS) and the Korean Longitudinal Study of Aging (KLOSA). I use one early life socioeconomic status (SES): mother’s education; and three other late life variables: income (classified by quintiles), having public health insurance, and receiving a public pension; the last two variables refer to benefits from social security systems. Years of education –a fifth variable– is only used to analyze mean age at death because collinearity problems in the variance equation. As expected, people with more education live longer. Mother’s education predicts neither the mean nor the variance of age at death in the three countries. Americans and Mexicans –but not Koreans– with higher income live longer too. On the contrary, having public retirement pension income is the only consistent predictor of longer lives and smaller variances at age at death, even after controlling for income. This result highlights that robust social security systems have an impact beyond mere financial security.
Health and wealth trajectories in childless families: how childlessness shapes cumulative inequalities over the life course
Anastasia Lam1,2 and Philipp M. Lersch1,2,3
1 Humboldt-Universität zu Berlin, Berlin, Germany
2 Einstein Center Population Diversity, Berlin, Germany
3 DIW Berlin, Berlin, Germany
Background: Childlessness is increasing worldwide, raising questions of how childless individuals will fare as they get older. Childless individuals generally accumulate more wealth over their life course so might have more resources for healthcare services, and higher wealth is usually associated with better health, but the relationship between childlessness and health is less clear. Therefore, this study aims to examine the bidirectional relationship between health and wealth over time and how it differs in individuals with and without children. Further, as health and wealth are both prone to cumulative (dis-)advantage, we pay particular attention to such processes.
Methods: Using seven waves (2004-2022) of the Survey of Health, Aging, and Retirement in Europe (SHARE), we conduct multivariate growth models to assess joint trajectories of health and wealth over age, by childlessness status. Health is defined using a 58-item frailty index and wealth is couplelevel net non-housing wealth. Both outcomes are percentile rank transformed, with higher values indicating either better health or higher wealth.
Results: We find that health and wealth are positively correlated, indicating that individuals with higher wealth also have better health. Childlessness has a larger effect for individuals who have less wealth, but better health, highlighting the complex relationship between health, wealth, and childlessness. Further, the speed of wealth accumulation decreases with age, while the speed of accumulating health deficits increases, suggesting potential ceiling effects and selective survival, respectively.
Conclusion: These findings have important implications for our understanding of how wealth and health inequalities develop over time within the context of changing family structures. Future analyses will assess whether other health indicators may have different relationships with wealth and childlessness, and how different welfare states might influence the observed patterns due to varying public benefits, social security, and healthcare systems.
Associations of migration-related and sociodemographic factors with functional mobility and depressive symptoms among midlife and older Russian-speaking migrants
Lily Nosraty1,2, Marguerite Beattie 3, Laura Kemppainen1,2, Anne Kouvonen1,4& Sirpa Wrede1
1Faculty of Social Sciences, University of Helsinki, Finland
2Faculty of Social Sciences (Health Sciences) and Gerontology Research Center (GEREC), Tampere Universities, Tampere, Finland
3Faculty of Educational Sciences, University of Helsinki, Finland
4Centre for Public Health, Queen’s University Belfast, Northern Ireland
Background: This study examines how migration-related and sociodemographic profiles are associated with functional mobility and depressive symptoms among midlife and older Russian-speaking migrants in Finland.
Method: We analyzed data from the Care, Health, and Ageing of the Russian-Speaking Minority (CHARM) study, including individuals aged 50+ residing in Finland (Wave 1: 2019, Wave 2: 2022). Two age groups were examined: under 65 years (n = 653) and 65+ years (n = 492). Key variables included sex, education, year and age at migration, migration reason, marital status, local language skills, discrimination, income support, religion, last job, functional mobility (stair walking, walking distance), and depressive symptoms (CES-D scale). Latent Class/Profile Analysis (LCPA) identified subgroups, and the Bolck–Croon–Hagenaars (BCH) method assessed subgroup differences in depressive symptoms and mobility.
Results: Three distinct migration-sociodemographic profiles were identified for Older Adults (65+)
Profile 1 (24%) → Established Manual Workers (Lutheran, Finnish-Educated, No Income Support)
Profile 2 (20%) → High Social Class Status Migrants (Orthodox, Highly Educated, No Income Support)
Profile 3 (56%) → Vulnerable Late Migrants (Orthodox, Lower Education, Income Support, Low Language Proficiency, Ingrian Migration)
Women were more likely to be in Established Manual Workers, while men were more likely to be in Vulnerable Late Migrants.
Younger Migrants (<65)
Profile 1 (47%) → Established Young Migrants (Finnish-Educated, Language Proficiency, No Income Support, Manual Work Background)
Profile 2 (53%) → Vulnerable Young Migrants (Orthodox, No Language Proficiency, Not Married)
Mobility differed significantly between the three profiles of those aged 65 and older (p = 0.007), with the Vulnerable Late Migrants experiencing the greatest limitations. Mobility differences for those under 65 were also statistically significant (p = 0.05), but depression levels did not vary significantly across classes for either age group.
Conclusion: Midlife and older Russian-speaking migrants in Finland exhibit distinct migration-related profiles that affect mobility but not depressive symptom levels.
Quantifying the Lifecourse of Multimorbidity: New Applications of Demographic Methods
Solveig Argeseanu Cunningham Netherlands Interdisciplinary Demographic Institute, The Netherlands Emory University, USA
Fernando Riosmena University of Texas – San Antonio, USA J.
Daniel Zazueta-Borboa Netherlands Interdisciplinary Demographic Institute, The Netherlands
The leading causes of morbidity and mortality worldwide are non-communicable diseases. These are chronic diseases because they develop over time and generally can be conceived of as absorbing states. They tend to accumulate into multimorbidity, a complex and life-shortening combination of conditions.
Research on these diseases has focused typically on one disease at a time, measuring prevalence, characteristics or prognosis. These approaches yield only limited information about risks for and age patterns of disease incidence. Several standard methods in demography can be adapted to bring new perspectives on multimorbidity. We demonstrate the adaptation of demographic methods to develop new measures of multimorbidity: morbidity incidence schedules, morbidity progression ratios, mean age at multimorbidity, and total morbidity rate; these measure the expected number of diseases by age, probabilities to transition from x to x+1 diseases, and cumulative number of diseases an individual would develop given the current disease environment.
We use 2017 data from the Survey of Health, Ageing and Retirement in Europe (SHARE), representative of people aged 50+ in Europe (n~140,000). We define multimorbidity as the presence of 2+ conditions from among hypertension, arthritis, diabetes, cancer, respiratory diseases, heart attack, and stroke.
The average number of diseases increases exponentially from age 50y onwards. By age 60y, Europeans have on average 1.1 chronic diseases. By age 80y, men have on average 1.8 diseases and women 2. The mean age of developing multimorbidity was 64y for both sexes. The probability of moving from one disease to multimorbidity for adults in their 60s is 0.5 and continues to increase with age thereafter. The probability of progressing from 2-3 diseases is 0.45 for adults in their 60s and remains generally flat thereafter; that of progressing from 3-4, 4-5, and 5-6 is around 0.25 across older ages, and slightly lower for progressing from 6-7 diseases.
Social vulnerability differently increases the risk of death in men and women: longitudinal analysis over 15 years in the Paquid Study
Camille Ouvrard, Ivane Koumetio Jiatsa, Antoine Gbessemehlan, Luc Letenneur, Jean-François Dartigues, Hélène Amieva and Karine Pérès
Team « ACTIVE » (Aging, Chronic disease, Technology, Disability and Environment) Inserm U1219 – BPH Research Center – University of Bordeaux
Background: Social vulnerability (SV) is a key determinant of health, influencing mortality risk through socioeconomic and psychosocial mechanisms. While prior research has linked SV to adverse health outcomes, few studies have examined its differential effects by gender. Given the distinct social integration patterns and coping strategies of men and women, understanding these differences is essential for developing targeted public health interventions.
Objective: This study aims to (1) compare SV levels between men and women and (2) assess whether and how SV is differentially associated with mortality over a 15-year follow-up in the PAQUID cohort, a population-based study on aging.
Methods: A total of 3,695 community-dwelling adults aged ≥65 years were analyzed. SV was measured using a 26-item Social Vulnerability Index (SVI) and classified into three levels (low, moderate, high). Cox proportional hazards models estimated the association between SV and all-cause mortality, stratified by gender and adjusted for age and functional limitations. The contribution of six SV subdimensions (e.g., living situation, geographical characteristics, psychological experience, socioeconomic status, leisure activities and social support) was also examined.
Results: Women exhibited higher SV levels than men (40% vs. 21% in the high SV category). High SV was associated with a 40% increased mortality risk in both genders. However, moderate SV significantly increased mortality risk only in men (HR: 1.31, 95% CI: 1.14–1.51; p < 0.001), while women showed resilience until higher SV accumulation. Low leisure activity levels and poor socioeconomic status were key mortality predictors, with men affected even at moderate levels.
Conclusions: Men appear more sensitive to moderate SV, while women require greater SV accumulation to experience increased mortality risk. These findings highlight the need for gendersensitive policies and interventions, particularly targeting social engagement and socioeconomic factors in older men.
Key words: Social vulnerability, gender differences, mortality, older adults, social determinants of health, cohort.
Gender- and social-specific contribution of chronic diseases to activity limitation in the 60+ population in France.
Rahal, Mira – Ecole des hautes études en sciences sociales (EHESS) Geoffard, Pierre-Yves – Paris school of economics (PSE) Cambois, Emmanuelle – Institut national d’études démographiques (INED)
Background: In order to improve our understanding of inequalities in healthy ageing in France, this study examines the contribution of chronic diseases to activity limitation (AL), across gender and socio-occupational categories (SOC). It assesses how the prevalence and the disabling impact of diseases contribute to AL among these groups.
Methods: We used data from the ‘Capabilities, Help, and Resources for Seniors’ surveys (‘Capacités, Aides, et Ressources des Seniors’ – CARE), conducted in France in 2015-16 via face-to-face interviews, on a representative sample of individuals over age 60, living in private households (N= 10,628) and in nursing homes (N= 3262). SOC were based on the self-reported occupation (current or latest for the retired respondents). We examined the gender- and SOC-specific prevalence of self-reported diseases as well as their disabling impact and their contribution to activity limitation (AL), using additive hazard regression models accounting for comorbidities.
Results: Women age 60+ reported higher prevalence of depression, cognitive disorders, and musculoskeletal conditions (MSC), while men reported more cardiovascular diseases (CVD) and diabetes. Skilled and highly-skilled workers reported lower prevalence of diseases compared to farmers, clerical and manual workers. Cognitive disorders and CVD had the highest disabling impacts. CVD and MSC contributed the most to AL overall, except among highly-skilled women for whom depression came second.
Conclusion: Although these associations cannot bring evidence of the causal impact of diseases, our findings suggest a benefit of targeted prevention policies to decrease disability: especially for CVD among men in farmers, low-skilled and self-employed occupations, and for MSC and depression among women across various SOC.
Do epigenetic clocks predict future health outcomes in a middle-aged population, and how do they compare to traditional risk factors of cardiometabolic diseases?
Daria Kostiniuk, Flóra Székely, Joanna Ciantar, Sonja Rajić, Pashupati Mishra, Terho Lehtimäki, Olli Raitakari, Mika Kähönen, Emma Raitoharju, Saara Marttila
Biological age aims to capture the amount of ageing-associated damage that has accumulated over time. By definition, it should be a better predictor of future health than chronological age. Measures based on DNA methylation, so called epigenetic clocks, are considered to be the best measures of biological age available. It has been suggested that these clocks could be used as tools for personal health monitoring as consumer products or in clinical settings.
We wanted to study how epigenetic clocks perform against traditional, easy to measure, risk factors of common chronic diseases, such as sex, BMI and smoking. We utilized the longitudinal Young Finns Study cohort (YFS), including 1141 individuals who, at baseline, were aged 34-49 years and were free of common chronic diseases (cardiovascular diseases, type 2 diabetes, cancer). Our approach was to predict the onset of any of these diseases over a 7-to-9-year follow-up, during which 427 individuals were diagnosed with the above-mentioned conditions.
In our study population, epigenetic clocks PCGrimAge and DunedinPACE predicted onset of disease (OR=1.25, p=0.035 and OR=1.47, p=1.14*10-8 , respectively). However, these associations were no longer statistically significant when adjusted for BMI and smoking. Strikingly, a simpler model consisting only of age, sex, BMI and smoking proved to be more effective at discriminating between individuals with and without disease onset compared to the clock models. Specifically, the area under the ROC curve (AUC) for simpler model was 0.747, significantly higher that the AUCs for PCGrimAge (0.629) and DunedinPACE (0.662) (DeLong’s test p=5.82*10-11 and p=3.81*10-7, respectively).
Our results suggest that, while epigenetic clocks are useful tools in biomedical research, they offer no added value to the health-conscious consumer over the well-established risk factors of common chronic diseases.
Association of behaviour-related health risk factors with working life expectancy in adults aged ≥50 years
Paola Zaninotto* 1, Katriina Heikkilä* 2,3,4, Holendro Singh Chungkham 1,5, Jaana Pentti 3,4,6, Jenni Ervasti 7, Mika Kivimäki 6, 8, Jussi Vahtera3,4, Sari Stenholm 3,4,9 *
Joint first authors 1 Department of Epidemiology and Public Health, University College London, London, United Kingdom 2 Department of Public Health, University of Turku and Turku University Hospital, Turku, Finland 3 Centre for Population Health Research, University of Turku and Turku University Hospital, Turku, Finland 4 Department of Public Health and Welfare, Finnish Institute for Health and Welfare, Helsinki, Finland 5 Psychobiology and Epidemiology Division, Department of Psychology, Stockholm University, Stockholm, Sweden 6 Clinicum, Faculty of Medicine, University of Helsinki, Helsinki, Finland 7 Finnish Institute of Occupational Health, Helsinki, Finland 8 Brain Sciences, University College London, London, United Kingdom 9 Research Services, Turku University Hospital and University of Turku, Turku, Finland
Background: Health is a key determinant of work participation among older adults, but the contribution of behaviour-related health risk factors to working life expectancy (WLE) is unclear. We investigated associations of obesity, alcohol intake, smoking and low level of physical activity with WLE among adults aged 50 years and older.
Methods: Data were from two prospective cohort studies: 9,807 participants of the English Longitudinal Study of Ageing (ELSA) and 65,255 participants of the Finnish Public Sector study (FPS). Obesity, heavy drinking, smoking and low physical activity were self-reported at study baseline. WLE from age 50 up to 70 years was estimated utilising a multi-state modelling, separately for men and women across occupational socioeconomic categories (low, intermediate and high), with adjustment for age.
Results: Our findings suggest that individuals who were obese, consumed higher than recommended amounts of alcohol, smoked and had low levels of physical activity, could expect to work for fewer years than those who did not have these behaviour-related health risk factors. Increasing number of health risk behaviours was associated with shorter WLE across sex and occupational position categories in both studies. The differences in WLEs between those with no health risk behaviours and those with ≥2 health risk behaviours were up to 1.6 years in ELSA and less than 1 year in FPS.
Conclusion: Having multiple behaviour-related health risk factors is linked to shorter WLE at and after age 50 years, a difference that can have important economic implications in societies with ageing populations.
Trends in productive life expectancy among older adults: An assessment of the expansion or contraction of productive aging
Zachary Zimmer
Mount Saint Vincent University
BACKGROUND: Studies of health expectancy customarily focus on physical functioning measures, like disability, and sometimes incorporate pain or chronic illnesses. Few, if any, such studies incorporate broader indicators of social well-being. This leaves unresolved a vital question asked of the health expectancy literature regarding whether improvements in old-age life expectancy result in more years ‘worth living’. To address the gap, we integrate a ‘productive aging’ approach and assess trends in productive life expectancy in the U.S., adjusting for disability and chronic pain.
METHODS: Data is from HRS 1998-2018. Productive aging encompasses work, volunteering, instrumental support and grandparenting assistance. Productive years lived is computed using multistate life tables employing the Interpolative Markov Chains (IMaCh) software. Data includes 70,616 observations of two-year transitions among individuals aged 70+. Expansion versus contraction of productive aging is assessed considering net and relative changes in productive life expectancy among those without and with chronic pain and/or disability.
RESULTS: Estimates of total life expectancy using HRS data indicate increases for men and women from 1998 to 2018 that are consistent with CDC reports. Life expectancy also increased in categories of without and with pain/disability. In 1998, 70-year-old men and women, without pain/disability, expected 41.8% and 35.1% of remaining life in a productive state. Percentages remained fairly stable over a 20-year period. But, relative expectation of productive life for those with pain/disability declined from 27.1% and 22.2% of remaining life for men and women in 1998 to 25.3% and 21.7% by 2018.
CONCLUSIONS: We find stability in relative productive life expectancy for some, but modest declines among those with pain/disability, supporting a contraction of productive aging among those dealing with physical challenges. This supplements the expansion of morbidity argument, suggesting recent gains in life expectancy may occur alongside reductions in share of life that could be considered ‘meaningful’.
Life course before and after receiving a fixed-term disability pension: an 8-year follow-up study
Anu Polvinen
Finnish Centre for Pension
Background: A significant number of working-age individuals leave the labor market prematurely due to health problems or disability. In Finland, approximately 20,000 people start to receive a disability pension each year, with half of them being granted a fixed-term pension with the aim of returning to work. This study investigates various work, retirement, and unemployment patterns among those who received a fixed-term disability pension. It also describes these patterns by individual-level factors as well as income development.
Methods: Individuals aged 30-58 years who received a fixed-term disability pension in 2018 were followed for four years before and after receiving the disability pension, using total register data from Statistics Finland and the Finnish Centre for Pensions. Sequence analysis was used to identify typical life course patterns.
Results: The results showed that return to work after receiving a fixed-term disability pension was relatively rare. The majority (60%) continued to receive a full or partial disability pension four years after receiving a disability pension. One-fifth received a fixed-term disability pension for about 1-3 years, after which many of them appeared to end up being unemployed. The status prior to the disability pension was related to the status after disability. Furthermore, significant differences were observed between the clusters: higher education was associated with a higher likelihood of returning to work, but also of receiving a partial disability pension. Those who ended up receiving a permanent disability pension or were unemployed were more likely to have low education. People with mental disorders rarely returned to work.
Conclusions: This study contributes to the knowledge of the life course of working-age individuals who end up receiving a fixed-term disability pension. In order to prevent disability, it is important to understand which factors are associated with different pathways to work, retirement or unemployment.
Exploring heterogeneities in cognitive functioning at older age: The role of employment histories across European welfare regimes.
Thomas Arnhold1, 2, Daniela Weber1, 2 , and Valeria Bordone1, 3
1POPJUS Program, International Institute for Applied Systems Analysis (IIASA), Wittgenstein Centre for Demography and Global Human Capital (IIASA, OeAW, University of Vienna), Laxenburg, Austria 2Health Economics and Policy Division, Vienna University of Economics and Business, Vienna, Austria 3Department of Sociology, University of Vienna, Vienna, Austria
Background: Maintaining good cognitive functioning is a key determinant of health and wellbeing in older age. Emerging research highlights the protective role of employment throughout the life course in preserving later-life cognition. However, the relationship between (non-)employment and cognition is also influenced by institutional contexts, shaping work and family roles during working age. This study investigates the relationship between employment histories and later-life cognitive functioning, focusing on the moderating role of European welfare regimes. We hypothesize that European welfare regimes (Baltic, Conservative, Central Eastern European, Social Democratic, Southern European) mediate the adverse cumulative impact of full-time employment breaks (Parttime employment, homemaking, unemployment, sickness) on cognitive functioning at older age by shaping social protection policies as well as work and family roles differently.
Methods: Our analysis draws on individual-level retrospective life history data from the Survey of Health, Ageing and Retirement Europe (SHARE), including adults aged 50 to 75 living in 22 European countries (N = 159,631). Using linear random effects growth curve models, we examine interactions between European welfare regimes and employment histories in relation to cognitive functioning (episodic memory and verbal fluency) at older age.
Results: Preliminary results reveal stronger negative associations between working-age part-time employment spells and later-life cognitive functioning for women in Baltic, Central Eastern European, and Southern European welfare contexts. The negative relationship between female homemaking spells and later-life cognition is most pronounced in Central Eastern European welfare regimes. For both genders, Southern European welfare regimes appear to mitigate the adverse relationship between unemployment spells and cognitive functioning.
Conclusions: Welfare regimes may substantially alter cognitive outcomes in later life, particularly for women who experienced part-time employment, homemaking, or unemployment spells. The results contribute to a deeper understanding of cognitive functioning in older age, emphasizing the role of institutional and social contexts in shaping later-life cognitive health.
Work participation in times of rising state pension age: the decreasing role of health
Dorly J.H. Deeg1,2, Maaike van der Noordt3, Astrid de Wind4,2, Mariska van der Horst5, Cécile R.L. Boot4,2
1)Amsterdam University Medical Centers – Vrije Universiteit Amsterdam, Department of Epidemiology and Data Science
2) Amsterdam Public Health Research Institute, Amsterdam, the Netherlands
3) National Institute for Public Health and the Environment (RIVM), Department of Health Knowledge Integration, Center for Health and Society, Bilthoven, The Netherlands
4) Amsterdam University Medical Centers – University of Amsterdam, Department of Public and Occupational Health, Amsterdam, the Netherlands
5) Vrije Universiteit Amsterdam, Faculty of Social Sciences, Department of Sociology, Amsterdam, the Netherlands
Background: This study addresses to which extent workforce participation of older workers is affected by health in a period of rising state pension age (SPA) in the Netherlands.
Methods: Using the Longitudinal Aging Study Amsterdam, we studied three successive periods with baselines 2013, 2016, and 2019, and followups 2016, 2019, and 2022. In each period, we selected the age group 61- 63 years at baseline that had not reached SPA at 3-year follow-up (baseline n’s 109, 172, and 132, respectively). Workforce participation was defined as continuing working and number of hours worked. Health indicators included self-rated health, functional limitations, depressive symptoms, and cognitive ability. Logistic (for transitions) and linear (for working hours) regression models controlled for age, sex, education, and partner status.
Results: Workforce participation was 59%, 58% and 73% at the successive baselines. During the 3-year follow-ups, 58%, 82%, and 72% continued working, respectively. Better self-rated health predicted continued working only in 2019-2022 and prior to the COVID-19 pandemic outbreak. No other health indicators predicted work continuation. In continuing workers, working hours remained stable around 31 hours in 2013-2016 and 2016-2019, but decreased to 26 hours in 2019-2022. Functional limitations and less-than-good self-rated health predicted decrease in working hours only in 2013-2016, but not in other periods. In 2019-2022, a reduction in working hours was significantly greater in workers with fewer depressive symptoms and higher cognitive ability than in workers with more depressive symptoms and lower cognitive ability. In earlier periods, depressive symptoms and cognitive ability were not predictive of change in working hours.
Conclusion: At ages 61 to SPA, health plays a decreasing role over time in exit from the workforce and in hours worked. Motivation to work may be more important. The role of the COVID-19 pandemic in the decrease in workforce participation in 2019-2022 needs further research.
Abstracts
Poster presentations in alphabetical order
How trends in working longer shape socioeconomic and racial/ethnic disparities in healthy-working life expectancies in the United States
Leah R. Abrams PhD, MPH1 Alessandro Feraldi PhD2 & Christian Dudel PhD3-5 1. Department of Community Healthy, Tufts University, Medford, MA, USA (Leah.Abrams@Tufts.edu) 2. Sapienza University of Rome, Italy 3. Max Planck Institute for Demographic Research, Rostock, Germany 4. Federal Institute for Population Research, Wiesbaden, Germany 5. Max Planck–University of Helsinki Center for Social Inequalities in Population Health, Rostock, Germany, and Helsinki, Finland
Background: For decades, Americans enjoyed increasing life expectancy, and for some, extended disability-free life expectancy. In response, Social Security increased the full retirement age. However, mortality progress has since stagnated, Americans with low educational attainment have not benefited from delayed disability, and there remain substantial racial disparities in morbidity and mortality. We examine educational and racial/ethnic differences in expected time spent in healthy work, unhealthy work, healthy retirement, and unhealthy retirement in 1994-2006 and 2008-2020.
Methods: Data on White, Black, and Hispanic men and women of low, medium, and high educational attainment at ages 50 and older came from the Health and Retirement Study. We ran discrete-time multistate models to compute the probability of transitioning between states of health (defined as no functional limitations) and work, which were used to calculate life expectancies after age 50 in each state.
Results: In both time periods, there were marked educational, and to a lesser degree, racial/ethnic disparities in expected years in healthy work and healthy retirement. Compared to 1994-1996, 2008-2020 brought increased years in healthy work for advantaged groups like high education White men and women (+1.34 and +0.95 years respectively). Black and Hispanic people were more likely to increase years in unhealthy work. Only the highest education White and Black women experienced increases in healthy retirement (+0.30 and +0.82 years respectively), while healthy retirement declined in most other groups. Years in unhealthy retirement increased in the later period, especially among Black and Hispanic older Americans across education levels.
Conclusions: The findings reveal how increasing the length of working lives, coupled with trends in health and functioning, is expanding disparities in time spent unhealthy at work and healthy in retirement.
Onset of depression among people with chronic diseases
Florencia Bathory (EDSD – INED), Solveig Cunningham (NIDI), Emmanuelle Cambois (INED)
Background: The prevalence of non-communicable diseases increases with population ageing, causing vulnerability and dependency. Declining health may cause depressive symptoms, for example due to pain, medications, or disability. Chronic conditions and depression are associated for all adult ages and across populations. The associations may be causal in either direction, as depression may be both the result, and potentially the cause of the non-communicable conditions. Depression may also contribute to increasing disability from other diseases. This study focuses on the causal pathways between chronic diseases and depression and we also aim to explore some mediation mechanisms in this relationship. We hypothesise that individuals living with chronic conditions transition more rapidly to depression than those in good health; and that chronic pain, medication and disability mediate this relationship by increasing the risk of depression. Whereas social participation, physical activity and the country of residence may protect from depression.
Methods: We use data from the Survey of Health, Ageing and Retirement in Europe (SHARE), waves 4 to 9, to obtain data on 13 chronic conditions: heart attack, high blood pressure or hypertension, high blood cholesterol, stroke or cerebrovascular disease, high blood sugar or diabetes, chronic lung disease, arthritis, osteoporosis, cancer, stomach or duodenal ulcer, peptic ulcer, Parkinson disease, cataract, hip fracture or femoral fracture. The European-Depression Scale (EURO-D) is used to assess depression. We use multistate models for men and women separately to obtain and compare the transition rates to depression among individuals in good health and individuals living with chronic diseases, controlling for confounders. We examine the effect of mediators in this relationship: chronic pain, activity limitations and medication, following the hypothesis that these may trigger depression; social participation, physical activity and the country of residence as variables that may protect the individuals from the onset of depression status.
Frequency and Predictors of Emergency Department Visits among the Oldest Old in Finland: the Vitality 90+ Study
Sherin Billy Abraham, Jutta Pulkki, Mari Aaltonen, Esa Jämsen, Jani Raitanen, Linda Enroth.
Tampere University, Finland
Background: Emergency department (ED) visits increase with age. However, knowledge of the frequency and drivers of ED visits among the oldest old, and whether patterns differ between home dwellers and round-the-clock care residents, is limited. This knowledge is crucial for functioning of health care systems. Therefore, this study examined the frequency and predictors of ED visits in a 90+ population living at home and in round-the-clock care.
Methods: Data from the Vitality 90+ survey, a population-based study with 1561 respondents in 2014 in Tampere, Finland, was combined with national register data on ED use and mortality until the end of 2017. Predictors of the first ED visit were examined using Cox regression models (4-year cumulative hazard) and predictors of 1–3 and ≥4 ED visits in one year of follow-up with multinomial logistic regression models.
Results: Over the four-year study period, 79% of the participants had at least one ED visit. Home dwellers had higher cumulative hazards of ED visits and were more often frequent ED users (≥4 ED visits) than those living in round-the-clock care. Not receiving home care, multimorbidity, poor subjective health and wellbeing, and limitations in functioning increased the risk of ED visits among home dwellers, while having dementia, limitations in functioning, impaired sensory functions, and less frequent contact with family decreased the risk among round-the-clock care residents. In both groups, the predictors of ED visits were similar in one and four-year follow-ups.
Conclusions: The frequency and predictors of ED visits greatly differ between those living at home and in round-the-clock care. Since most ED visits occurred among those living at home and having poor health but not receiving formal home care, improving the continuity of care and the coverage of home care services could help curb the increase in ED visits among the fast-growing oldest old population.
Healthy Working Life Expectancy across birth cohorts in the United States
Félix Blain
Objectives: This study investigates trends in healthy working life expectancy (HWLE) in the United States amid population aging and recent declines in health among working-age individuals. HWLE, defined as the average number of years expected to be spent healthy and working from age 51, is examined by gender and educational level across three birth cohorts.
Methods: Using longitudinal data from the Health and Retirement Study (HRS), HWLE estimates were calculated for individuals aged 51 and older. Using continuous-time multi-state modelling, trends were analyzed across cohorts born from 1936 to 1955, focusing on differences by gender and educational attainment to assess disparities in HWLE over time.
Results: The findings indicate that HWLE remained stable for most groups but declined among individuals with lower educational attainment. Thus, recent cohorts of older Americans seem unable to extend their working life in good health despite increases in the full retirement age.
Discussion: These findings highlight the need for targeted policies to promote healthier work environments and expanded job opportunities for older adults. Addressing disparities in HWLE, particularly for those with less education, is critical to improving outcomes for future cohorts as they approach retirement age.
Keywords: Active life expectancy, Population aging, Employment, Education, Multistate model.
Differential Item Functioning of informant measures of cognitive functioning in the U.S. and Mexico
Phillip Cantu (UTMB), Joce Jaen (RUSH), Mark Sanderson-Cimino (UCSF), Douglas Tommet (Brown), Keith Widaman (UCR), Richard Norman Jones (Brown), Rebeca Wong (UTHCSA), Alden Gross (Johns Hopkins)
Key findings:
- Informants in Mexico report significantly lower scores on the harmonized Community Screening Instrument for Dementia (CSID) than informants in the U.S. 1.5. We evaluated measurement differences in the CSID between the US and Mexico, finding
- The harmonized informant CSID requires constrained covariances between 2 question sets, “remembering friends’ names” and “remembering family members’ names”, and “difficulty finding the right words” and “using the wrong words”.
- There were significant differences between countries in factor loadings for the questions “difficulty remembering things” and “forgets where things are kept”
Population studies of cognitive functioning rely on survey questionnaires to estimate the prevalence of dementia in nationally representative samples. In addition to direct tests of cognitive functioning, the international studies as part of the Harmonized Cognitive Assessment Protocol, also include informant questionnaires, in which a family member or other person close to the research participant are asked about the participant’s cognitive functioning. While international harmonized measures of directly assessed cognitive have been developed, there are no statistical harmonizations of informant assessments of cognitive functioning. This study investigates differences in how informants respond to questions about cognitive functioning in the US and Mexico.
Using data from the Harmonized Cognitive Assessment Protocol (HCAP) of the Health and Retirement Study (HRS) (n=3,183) in the United States and the Mexican Cognitive Aging Study (MexCOG) (n=1,259) in Mexico, we examined informant reports of cognitive functioning of the Community Screening Instrument for Dementia (CSID). We test for differential item functioning (DIF) using structural equation modeling (SEM) to develop a confirmatory factor analysis (CFA) model of the CSID informant questionnaire in Mexico and the U.S.
We find that a harmonized measure of the CSID has fixed factor loadings between countries but different item thresholds. Our harmonized measure of informant reports of cognitive functioning suggests that informants report less cognitive impairment for older adults in Mexico when controlling for directly assessed cognitive functioning.
Our results suggests that either the harmonized direct assessment of cognition may underestimate cognitive functioning in Mexico or that informants in Mexico are systematically under reporting symptoms of cognitive decline.
Predicting cardiovascular morbidity and mortality with SCORE2(OP) and Framingham risk estimates in combination with indicators of biological aging
Anna Tirkkonen1, Jonathan K.L. Mak2,3, Johan G. Eriksson,4,5,6,7 Pauliina Halonen8,13, Juulia Jylhävä2,9,10, Sara Hägg2 , Linda Enroth8, Jani Raitanen8,11, Iiris Hovatta12, Tuija Jääskeläinen13, Seppo Koskinen13, Markus Haapanen2,4,5, Mikaela B. von Bonsdorff1,4 and Laura Kananen2,8,14
1Faculty of Sport and Health Sciences and Gerontology Research Center, University of Jyväskylä, Finland 2Department of Medical Epidemiology and Biostatistics, Karolinska Institute, Stockholm, Sweden 3Department of Pharmacology and Pharmacy, The University of Hong Kong, Hong Kong, China 4Folkhälsan Research Center, Finland 5Department of General Practice and Primary Health Care, University of Helsinki and Helsinki University Hospital, Finland 6 Institute for Human Development and Potential, Agency for Science, Technology, and Research, Brenner Centre for Molecular Medicine, Singapore 7Department of Obstetrics and Gynaecology and Human Potential Translational Research Programme, Yong Loo Lin School of Medicine, National University of Singapore, Singapore 8Faculty of Social Sciences (Health Sciences) and Gerontology Research Center, Tampere University, Finland 9Faculty of Medicine and Health Technology and Gerontology Research Center, Tampere University, Finland 10Tampere Institute for Advanced Study, Tampere, Finland 11The UKK Institute for Health Promotion Research, Tampere, Finland 12SleepWell Research Program and Department of Psychology, Faculty of Medicine, University of Helsinki, Finland 13Finnish Institute for Health and Welfare, Helsinki, Finland 14Department of Neurobiology, Care Sciences and Society (NVS), Karolinska Institute, Stockholm, Sweden
Background and objective: Previous research assessing whether biological aging (BA) indicators can enhance the risk assessment of CVD outcomes beyond established CVD risk indicators, such as Framingham (FRS) and SCORE2/SCORE2-OP, is scarce. We explored whether BA indicators, namely the Rockwood Frailty-Index (FI), and leukocyte telomere length (TL), improve predictive accuracy of CVD outcomes beyond the traditional CVD risk indicators in general population of middle-aged and older CVD-free individuals.
Methods: Data included 14 118 individuals from three population-based cohorts: TwinGene, Health 2000 (H2000), and the Helsinki Birth Cohort Study (HBCS), grouped by baseline age (<70, 70+). The outcomes were incident CVD and CVD mortality with 10-year follow-up. Risk estimations were assessed using Cox regression and predictive accuracies with Harrell’s Cindex.
Results: Across the three study cohorts and age groups: 1) a higher FI, but not TL, was associated with a higher occurrence of incident CVD (p<0.05), 2) also when considering simultaneously the baseline CVD risk according to FRS or SCORE2/SCORE2-OP(p<0.05), 3) adding FI to the FRS or SCORE2/SCORE2-OP model improved the predictive accuracy of incident CVD. Similar findings were seen for CVD mortality, but less consistently across the cohorts.
Conclusions: We show robust evidence that a higher FI value is associated with an increased risk of incident CVD in middle-aged and older CVD-free individuals, also when simultaneously considering the risk according to the FRS or SCORE2/SCORE2-OP. The FI improved the predictive accuracy of CVD outcomes beyond the traditional CVD risk indicators and demonstrated satisfactory predictive accuracy even when used independently.
Associations between social vulnerability and functioning in older age, and the moderating and mediating role of optimism and self-efficacy
Lahti Anna-Maria, Mikkola Tuija, Tirkkonen Anna, Eriksson Johan, von Bonsdorff Mikaela
Folkhälsan Research Centre, Finland
Background: Although links between social factors, personality and functioning have been established, the interactions of social vulnerability and personal factors impacting later-life functioning remain unclear. We investigated whether social vulnerability is associated with physical and mental functioning in older age, and with the change in functioning over 5 years. We studied whether optimism and self-efficacy act as moderators and/or mediators in the association.
Methods: Physical and mental functioning were measured using the SF-36 Health Survey in 2015 and 2020. Social vulnerability was comprised of several self-reported and register-based social factors. We analyzed the associations between social vulnerability and physical and mental functioning, and the change in them using General Linear Models. Moderation and mediation analysis were run using the PROCESS macro in SPSS. Analyses were adjusted for age and sex.
Results: Social vulnerability was negatively associated with physical functioning in 2015 (β = – 3.8, p<0.001) and 2020 (β = -4.11, p<0.001), and with mental functioning in 2015 (β = -2.16, p<0.001) and 2020 (β = -1.50, p<0.001). Social vulnerability was associated with a steeper decline in physical functioning over 5 years (β = -0.11, p<0.001). Optimism acted as a moderator by buffering the negative association of social vulnerability on physical and mental functioning, and also partly mediated the associations. Self-efficacy acted as a moderator by buffering the negative association between social vulnerability and mental functioning, but not physical functioning. Self-efficacy also partly mediated the association between social vulnerability and physical and mental functioning. A multiple-mediator model showed that self-efficacy no longer mediated the association between social vulnerability and physical functioning when optimism was included.
Conclusions: By impacting social vulnerability we may be able to promote functioning in older age. Social and personal factors need to be acknowledged when targeting and planning health interventions and services for older adults.
Paradox of Aging? Evidence Based on Happy Life Expectancy and Healthy Life Expectancy in China
Yan Liu, PhD Asian Demographic Research Institute, School of Sociology and Political Sciences, Shanghai University, Shanghai, China
Background: The “paradox of aging” indicates that people’s physical and mental health decline while their happiness may remain stable or improve with aging. However, no evidence has examined this paradox considering the mortality effect. Our study aims to address this based on happy life expectancy and healthy life expectancy among older Chinese people.
Methods: Data were drawn from four waves of the Chinese Longitudinal Healthy Longevity Survey (CLHLS) 2008-2018. Life satisfaction was used to reflect happiness in happy life expectancy (HapLE). Healthy life expectancy (HLE) included disability-free life expectancy and cognitive-impairment-free life expectancy. The multi-state interpolated Markov Chain was employed to estimate the life expectancy (LE), HLE, HapLE, and their proportions (HLE% and HapLE%).
Results: (1) We observed the paradox of aging after considering the mortality effect. HapLE and HLE both decreased with increasing age, but HapLE declined slower. Notably, HapLE% significantly rose while HLE% declined with aging. At 65, HLE was higher than HapLE, but this reversed at 80. (2) This paradox of aging was robust regardless of sex, residence, and education. In terms of sex, older women were expected to live proportionally more years with happiness than men, and the gender difference narrowed with aging. Conversely, older men were expected to live proportionally longer with physical and mental health, and this gender disparity initially increased and then decreased. (3) Socioeconomic status explained the paradox of aging. Individuals with higher socioeconomic status enjoyed advantages in both longevity and happiness.
Conclusions: Using HapLE and HLE indicators, we found the paradox of aging. People can have a happy life despite being disabled in ADL or cognitively impaired. Better socioeconomic status was beneficial for realizing higher HapLE through actively adapting and coping with functional decline. Our study offers implications for enhancing longevity and happiness when facing losses in physical and mental health.
Experiences of Agency Among People with Dementia
Ulla Halonen, Marja Lönnroth & Lina Van Aerschot
University of Jyväskylä, Finland
Background: Agency often refers to the ability to influence own life and make choices in relation to available resources, societal structures, and social context. The agency as described by people with dementia (PWD) themselves has been studied rather scarcely, and often it is portrayed as weak, diminishing, and dependent.
Methods: In our study, we examine the agency of PWD through the modalities of agency- perspective (Jyrkämä, 2008). The data is collected with semi-structured thematic interviews with 19 people with dementia aged 68–85 living at home. We investigate how they describe their agency through six modalities (knowing/know-how, being able, wanting, having-to/must, can, feel).
Results: The preliminary analysis shows that the agency of PWD is multifaceted, dynamic and adaptive. Although cognitive decline affects daily life, strong individual agency was identifiable, particularly in the modalities of knowing and wanting. The interviewees actively utilised the skills they had accumulated over their lifetimes and adapted to changing circumstances. In the wanting modality, a strong motivation to maintain independence and social relationships was emphasised. In the being able to-modality, participants emphasised retaining good physical functioning, and compensatory strategies for cognitive challenges. In the can-modality, the importance of surrounding structures (ie. family, financial resources) in enabling agency was highlighted. The results thus emphasise the relationality of agency, constructed in relation to other people, the environment, and available resources. Notably, the “having-to” modality was less frequent in the accounts.
Conclusions: Our study shows that agency does not disappear with dementia, but rather changes. The agency of PWD appears dynamic and adaptive to change. With dementia, agency often becomes shared agency, where loved-ones play a significant role. The findings emphasise the importance integrating social, structural and individual aspects to our understanding of agency.
Risk factors predisposing to dementia in individuals with type 1 diabetes
Susanna Satuli-Autere [1,2], Emma Dahlström [1,3], Fanny Jansson Sigfrids [1,3], Marika I. Eriksson [1,3], Anni Ylinen [1,3], Per-Henrik Groop [1,3‒6], Valma Harjutsalo [1,3], Lena M Thorn [1‒3], on behalf the FinnDiane Study Group
[1] Folkhälsan Research Center, Helsinki [2] Department of General Practice and Primary Health Care, University of Helsinki and Helsinki University Hospital, Finland [3] Research Program for Clinical and Molecular Metabolism, Faculty of Medicine, University of Helsinki [4] Department of Nephrology, University of Helsinki and Helsinki University Hospital, Helsinki, Finland [5] Department of Diabetes, Central Clinical School, Monash University, Melbourne, VIC Australia [6] Baker Heart and Diabetes Institute, Melbourne, VIC, Australia
Background: Life-expectancy in individuals with type 1 diabetes has improved in recent years, increasing their risk of ageing-related diseases. Type 1 diabetes is linked to cognitive decline and dementia, but we lack data on factors contributing to the dementia risk. Thus, the aim was to identify factors predisposing to the development of dementia in individuals with type 1 diabetes.
Methods: The study included 4,256 individuals (mean age 37.9 ± 11.9 years, 52.3% men) with type 1 diabetes from the observational Finnish Diabetic Nephropathy Study. The baseline study visit included a comprehensive evaluation of the participants’ diabetes, metabolic profile, and APOE genotypes. Data on any dementia diagnoses (Alzheimer’s disease, vascular dementia, or other dementias) were retrieved from Finnish Registers until any dementia diagnosis, death, or the end of 2017. Cox regression analysis was used to evaluate the impact of different variables on the development of dementia, accounting for death as a competing risk.
Results: During a median follow-up of 16.8 (IQR14.0‒18.8) years, 80 (1.88%) developed dementia, and 842 (19.8%) died. In the sex-adjusted, Cox regression model, the independently associated factors with any dementia were age [HR 1.16 (95% CI 1.13‒1.19), P<0.001], a previous cardiovascular event [2.24 (1.33‒3.79), P=0.002], the use of antihypertensive medication [2.38 (1.35‒4.20), P=0.003], HDL cholesterol [2.80 (1.59‒4.93), P<0.001], and APO E4/E4 genotype [6.75 (2.96‒15.37), P<0.001]. All, but cardiovascular events, were significant also after accounting for death as a competing risk. We further divided the study participants into quartiles by HDL concentrations, men and women separately. The incidence of dementia was increased in men in the two highest quartiles [HR 3.16 (1.04‒ 9.63), P=0.043] and [4.38 (1.47‒13.02), P=0.008], while not in women.
Conclusions: Higher age, previous cardiovascular event, use of antihypertensive medication, high HDL cholesterol, and APO E4/E4 genotype are independent risk factors for dementia in type 1 diabetes.
Adversity, perceived stressfulness, and resilience in older men and women – the role of socioeconomic status and functioning
Sini M. Stenroth1,2, Sini Siltanen1,2, Markus J. Haapanen2-5, Pirjo Vuoskoski6, Johan G. Eriksson2,3,7,8, Mikaela B. von Bonsdorff1,2
1Gerontology Research Center and Faculty of Sport and Health Sciences, University of Jyväskylä, Jyväskylä, Finland. 2Folkhälsan Research Center, Helsinki, Finland. 3Department of General Practice and Primary Health Care, University of Helsinki and Helsinki University Hospital, Helsinki, Finland. 4Centre for Health Services Research, Faculty of Medicine, The University of Queensland, Princess Alexandra Hospital, Woolloongabba, Queensland, Australia. 5Australian Frailty Network, The University of Queensland, Woolloongabba, Queensland, Australia. 6Faculty of Sport and Health Sciences, University of Jyväskylä, Jyväskylä, Finland. 7Yong Loo Lin School of Medicine, Department of obstetrics and gynecology and Human Potential Translational Research Programme, National University Singapore, Singapore. 8The A*STAR Institute for Human Development and Potential (A*STAR IHDP), Brenner Centre for Molecular Medicine, Singapore.
The prevalence of adversities typically increases with older age. However, few studies have examined how various adversities relate to perceived stressfulness and resilience, the ability to overcome adversity. We investigated adversities older men and women face, assessing how stressful these were and their relation to resilience. Furthermore, we explored whether socioeconomic status (SES) or functioning explained these associations.
The study included 1179 participants from the Helsinki Birth Cohort Study, born between 1934 and 1944. Adversities, perceived stressfulness, and resilience were assessed using the Hardy-Gill Resilience Scale in 2015 or 2017-2018. Adversities were themed using a data-driven content analysis. The perceived stressfulness was measured on a visual analogy scale (0-140mm). Physical and mental functioning were ascertained using the Short Form Health Survey in 2015 or 2017-18. General linear modelling was used to explore perceived stressfulness and resilience across adversity types, separately for men and women.
The five most common adversity types accounted for 78.2 % of all reported adversities: personal illness (n=278, 23.6 %), illness (n=207, 17.6%) or death (n=186, 15.8%) of a close relative, adversity in relationships (n=150, 12.7%), and adversity related to living situation (n=101, 8.6%). Men considered the illness of a close relative (Marginal Mean 105.0, p<.001) and women the death of a close relative (Marginal Mean 117.7, p<.001) to be the most stressful. Among men, the association between personal illness and perceived stressfulness attenuated (p=.272) when adjusted with functioning. Among women, the association between adversity in relationships and resilience became significant (p=.032) after adjusting for SES and functioning. Otherwise, SES or functioning did not explain the associations.
When facing adversity, men experienced less stress and more resilience than women. This may indicate that men handle adversity better than women. SES or functioning had a minimal impact on the associations among types of adversity, perceived stressfulness, and resilience.
Developing the customer journey and implementing the new ways of work in the health care and social services for older adults
Sara Suikkanen1, Jaana Ahl1, Tuula Hämäläinen1, Sari Lehtinen1, Suvi Ollikainen1, Johanna Saarteinen2, Matleena Takaluoma1
1 Faculty of Health care and social services, LAB University of Applied Sciences
2 Faculty of Business, LAB University of Applied Sciences
The need for services for older adults is concentrated in the last years of life. It is crucial to ensure that older clients in social and health care services receive timely, purposeful, functional, inclusive, and meaningful care. Multidisciplinary expertise can influence client flows and ensure sufficient, smooth, and cost-effective services and processes.
Kymenlaakso welfare area and LAB University of Applied Sciences are developing a comprehensive and multidisciplinary approach to assess and support the functional capacity of older adults. The project, running from August 2024 to December 2025, is divided into 1) Workshop phase (8-12/2024) and 2) Side guidance phase (1-12/2025).
In the first phase, five workshops were held to describe customer journey pathways that promote the functional capacity of older adults. Participants included supervisors and staff from emergency rooms, primary care hospital wards, social care discharge units, home care, and home rehabilitation. The first two workshops were held separately for staff and supervisors but covered the same themes and materials. The remaining three workshops involved both groups working together.
During the workshops, main challenges and areas for development in each service were identified. Solutions to these challenges were further developed. Participants considered which customer profiles benefit from each service, what information should be transferred to ensure seamless care, factors that hinder or promote development areas, and how functional capacity would be supported in an ideal situation. Based on these workshops, new customer journey pathways and ways of working were created.
In the second phase, experts from LAB University of Applied Sciences will integrate into daily operations and assist staff and supervisors in the cultural change process towards more activation and support of older adults’ functional capacity. This side guidance is a new way to implement cultural change in social and health care services. Results will be available later this year.
Management of vascular risk factors and its association with cognition among older adults
Konsta Valkonen, Anette Hall, Anna Rosenberg, Riitta Antikainen, Tiina Laatikainen, Hilkka Soininen, Timo Strandberg, Jaakko Tuomilehto, Miia Kivipelto, Alina Solomon, Tiia Ngandu
Finnish Institute for Health and Welfare (THL), University of Eastern Finland (UEF), Finland
Background: Vascular risk factors are common, and they are linked to development of cognitive decline. This study evaluates the management of vascular risks (blood pressure, diabetes, dyslipidemia), and how it associates with cognition over the two-year FINGER multidomain lifestyle intervention.
Methods: Participants (n=1260, aged 60–77, at risk of cognitive decline) were randomly assigned to the multimodal lifestyle intervention group (physical activity, diet, cognitive training, and vascular risk management; n=631) or the control group (n=629) for 2 years. Cognition was assessed using the Neuropsychological Test Battery (NTB). Vascular risk factor management was based on Current Care Guidelines and the participants were grouped based on if they would need drug treatment (based on self-reported disease, medications or measurements), if they were treated and if the treatment goals were met. Linear regression, generalized linear models, and mixed effects models were used.
Results: Suboptimal management of vascular risk factors was common. At baseline, 924 participants were eligible for antihypertensive medication, and 635 among them exhibited poor blood pressure management (treatment target not met). Similarly, 1218 participants qualified for medication for dyslipidemia, with 903 inadequately managed. Additionally, 264 participants met the criteria for diabetes medication, of which 216 were poorly managed/ undiagnosed. Associations were found only between diabetes management and NTB, with participants without diabetes showing higher positive changes over 2 years compared to the well-managed group. The previously reported intervention benefit on cognition did not seem to differ by baseline adequacy of vascular risk management.
Conclusions: Cardiovascular disease (CVD) risk factors were prevalent in this older at-risk population, and their baseline management was suboptimal. The overall lack of significant associations between risk factor management and cognition group may be due to limited variability of both vascular risk factors and cognition, or background factors such as disease duration that are associated with risk management.
The Social Networks of Chronic Disease
Zixuan Wang Utrecht University, The Netherlands
Solveig Argeseanu Cunningham Netherlands Interdisciplinary Demographic Institute, The Netherlands Emory University, USA
Chronic diseases are especially insidious because they tend to accumulate into multimorbidity. There is much to be understood about clustering of chronic diseases, and this information is important for prevention and treatment; for example, which diseases cluster, and why; what are the reasons for clustering, and are the patterns the same across age, sex, and social groups. In this study, we apply social network methods to the clustering and progression of multimorbidity.
Social network analysis has been used to identify relationships among individuals, for example, identifying in a population, who is connected with many or few individuals, who is connected more closely with others, who are the central players. We show what these methods can tell us about how diseases relate to each other.
We use population-representative multi-country data from the 2019 Survey of Health, Ageing and Retirement in Europe (SHARE) to map the social networks of chronic diseases in the European 50y+ population. We examine self-reported diagnosis of heart attack, hypertension, cholesterol, stroke, diabetes, chronic lung disease, arthritis, cancer, respiratory diseases, stomach ulcers, Parkinson’s disease, cataracts, Alzheimer’s disease, rheumatism, osteoporosis, chronic kidney diseases, fractures, obesity, and affective disorders. We measure number of disease communities or clusters; closeness centrality, reflecting distance to all other diseases; betweenness centrality, indicating how much a disease connects different parts of the disease network; and modularity, indicating strength of divisions between disease communities.
We found evidence of three disease communities: the first involving heart conditions, hypertension, cholesterol, and diabetes; the second involving Parkinson’s disease, stroke, and cognitive disorders; and the third involving cancer and conditions relating to digestive, respiratory, and bone health. The diseases with the highest level of closeness and betweenness centrality were obesity, focal in community one, Parkinson’s disease, focal in community two, and cancer, focal in community three.
The Construction and Empirical Research of Healthy City Indicator System from the Perspective of the Theory of Social Determinants of Health: Taking 54 Major Cities as Examples
Zhang Jiao, Lecturer of Beijing City University, Public administration Department; Ph.D of Social Medicine and Healthcare Administration, Renmin University of China; M.B.A., Eastern University
In the accelerating urbanization, the link between urban environments and public health has become more prominent, with urban residents focusing more on health. Constructing healthy cities is vital for China’s long – term population health goals, thus demanding in – depth study.
This study developed a comprehensive healthy city indicator system. Rooted in the theory of social determinants of health, it synthesized concepts from health promotion theory, urban amenities theory, and sustainable development theory. The system comprises six dimensions, further decomposed into 24 specific indicators. A sample of 54 major large cities in China was selected. The entropy weight method was selected to assign weights to the indicators, and confirmatory factor analysis was utilized to validate the rationality of the indicator system.
Findings show large disparities in healthy city development. Mega – cities perform best, followed by large cities, and then other big cities. Regionally, eastern cities are superior, especially in social and economic health – related aspects. The 54 cities are classified into four echelons by comprehensive health scores. There are differences between subsystem scores and the overall index, showing the complexity of healthy city frameworks.
In conclusion, although the overall development trend of healthy cities is positive, challenges such as uneven development across regions, cities, and different health related fields are substantial. To address these issues, enhanced policy interventions and increased capital investment are necessary. Corresponding countermeasures are proposed at macro, meso, and micro levels. This research innovates in its multi-disciplinary approach, diverse methodological application, and practical-oriented results, offering a solid theoretical foundation and practical guidance for healthy city construction and the advancement of the “Healthy China” initiative.
Keywords: Healthy City; Social Determinants of Health; Indicators; Entropy Weight Method
Decomposing Differences in Cohort Health Expectancy by Cause and Age with Longitudinal Data
Tao Sun, Huiping Zheng, Xiaojun Wang
Renmin University of China
Background: Cohort health expectancy, rather than period health expectancy, is a more appropriate measure for describing the health trajectories of specific cohorts. Decomposing the differences in cohort health expectancy by cause (i.e., disease) and age helps us to better understand and alleviate health disparities between sub-cohorts. However, there is a lack of effective decomposition methods for cohort health expectancy.
Methods: We propose a novel longitudinal attribution method that calculates age-cause-specific contributions to disability and death in a cohort during longitudinal follow-up. We then present a new longitudinal decomposition method for the differences in cohort health expectancies based on the attribution results. Finally, we illustrate the proposed methods by decomposing the sex difference in cohort health expectancy using a longitudinal dataset of older Chinese individuals.
Results: The longitudinal attribution results emphasized that cardiometabolic diseases, respiratory diseases, and cancer were the primary contributors to disability and death in the older Chinese population. Decomposing the female-male partial-cohort health expectancy for the 65–69 cohort, we found that the females’ advantage was mainly derived from lower death probabilities due to background, cancer, and respiratory diseases, while the females’ disadvantage was largely due to cardiometabolic diseases, driven by their mortality effects in younger ages and their disability effects in older ages. These findings demonstrate the utility of our methods in identifying key contributors to health disparities in cohort health expectancy.
Conclusions: This study introduced new methods for decomposing the differences in cohort health expectancy, implemented in a user-friendly online application (https://zhenghp.shinyapps.io/PC-HE_decomp/). These methods provide a practical tool for understanding and addressing health disparities.
Living longer in poorer health: Health expectancy in Israel between 2013 and 2022: a cross-national comparison with 15 European countries
Yan Zheng,1 Isaac Sasson1
1 Department of Sociology and Anthropology, Tel Aviv University, Tel Aviv, Israel
Background: The relationship between life expectancy (LE) and health expectancy (HLE) varies considerably across high-income countries. Israel, a high-income country in the Middle-East, is on par with or even surpasses European countries in life expectancy. However, less is known how Israel performs in terms of HLE in cross-national comparison.
Methods: Using data from the Survey of Health, Ageing and Retirement in Europe (SHARE), this study investigated the changes in self-perceived good life expectancy (SPGLE), Activities of Daily Living index (ADL)-free LE, and Instrumental Activities of Daily Living index (IADL)-free LE at 65 years in Israel and 15 European SHARE countries in 2013-2022, compared before and after the COVID-19 pandemic. Changes in HLE were further decomposed to underlying changes in mortality and morbidity.
Results: Within the first period, HLE at age 65 increased in Israel more sharply than LE among both sexes, indicating an absolute compression of morbidity across all health indicators. With few exceptions, other SHARE countries experienced similar changes. In some countries including Israel, compared with the positive contribution from the decreasing mortality, the declining prevalence of unhealthy states posed greater impacts on the rising HLE. However, during the COVID-19 pandemic, many countries experienced declines in HLE to varying degrees, particularly when measured with respect to self-perceived health. In Israel, despite the continuous increase in ADL- free LE, SPGLE declined notably among both sexes, due to the rising prevalence of self-perceived poor health and the rising mortality among males.
Conclusions: Israel stands out compared to other high-income countries in exhibiting the substantial gap between LE and HLE at age 65, particularly among women. This pattern is even more peculiar given its relatively high LE among both sexes, by international standards. Meanwhile, the rising prevalence of observed unhealthy states in some other SHARE countries within both periods needs further attention.