Longitudinal Data Research Articles

Unplanned Births and Their Effects on Maternal Health: Findings from the Constances Cohort

Unplanned births remain relatively common, even in regions with high contraceptive prevalence and accessible abortion services, such as contemporary France. Previous studies have shown that unplanned births can have numerous negative consequences for the well-being of mothers and children, including poorer maternal health behaviors, delayed and insufficient prenatal care, and depression during or immediately after pregnancy. However, these studies do not provide conclusive evidence on whether the observed outcomes are a consequence of unplanned births or of the conditions in which they likely occur. Furthermore, scant attention has been given to other dimensions of maternal well-being, such as physical health.This study uses longitudinal data from the French Constances Cohort and applies fixed-effects event study models to examine how women’s self-rated general health and risk of depressive symptoms are affected in the years following an unplanned birth. Results show that women who had an unplanned birth reported a sudden, significant drop in their general health in the year following the birth, particularly among the youngest, while health outcomes following planned births showed a gradual, slight decrease over the time-period considered. The risk of depressive symptoms increased similarly after birth for both unplanned and planned births. This study contributes to the literature by using a longitudinally constructed measure of unplanned births based on pre-birth fertility intentions, rather than commonly used retrospective measures prone to ex post rationalization. It also distinguishes between unwanted and mistimed births while further examining their consequences on medium-term maternal health.

Subjective well-being and attitudes toward outgroup members

There is a large body of literature on the effect of negative attitudes toward outgroup members on target individuals and groups. However, less attention was devoted to their effects on those who hold these attitudes. The current study hypothesized that there would be a bidirectional relationship between subjective well-being and attitudes toward outgroup members. In Study 1, data from the Integrated Values Surveys, the Ecology-Culture Dataset, and the World Happiness Report 2022 were used. In Study 2, a random intercept cross-lagged panel model was implemented using longitudinal data over a six-year period from the GESIS Panel (n = 6783 people living in Germany). Results of Study 1 demonstrated a significant negative relationship between subjective well-being and negative attitudes toward outgroup members both at the country level (118 countries) and at the individual level (663,965 participants in 120 countries around the world). In Study 2, a random intercept cross-lagged panel model revealed significant within-person reciprocal relations between life satisfaction and negative attitudes toward refugees. This study provides evidence for a bidirectional relationship between subjective well-being and negative attitudes toward outgroup members.

Longitudinal links from attachment with mothers and fathers to adolescent substance use: Internalizing and externalizing pathways.

The present study examined whether internalizing and externalizing symptoms may mediate the association between adolescent-mother and adolescent-father attachment and substance use. The sample included 167 adolescents (47% girls) who were assessed at five time points with approximately 1 year between each assessment, beginning in middle adolescence (Mage=14.07) and ending in the transition to young adulthood (Mage=18.39). The adolescents reported their perceived attachment with both their mother and father during middle adolescence (Times 1 and 2), their internalizing and externalizing symptoms during late adolescence (Times 3 and 4), and their alcohol use during the transition to young adulthood (Time 5). The results showed that less secure adolescent-father attachment, but not adolescent-mother attachment, was predictive of heightened externalizing and internalizing symptoms. In turn, heightened externalizing symptoms were predictive of heightened alcohol use. Despite the nonsignificant direct association between adolescent-father attachment and alcohol use, less secure adolescent-father attachment was indirectly predictive of greater alcohol use, mediated through heightened externalizing symptoms. The findings highlight the importance of close and trusting father-adolescent relationships in the development of psychopathology and substance use behaviors. The developmental cascade from a less secure adolescent-father attachment to greater externalizing symptoms and heightened substance use, as well as implications for prevention and intervention of young adult substance use, are discussed. RESEARCH HIGHLIGHTS: The differential pathways from adolescent-mother and adolescent-father attachment to substance use during the transition to young adulthood are not well known. Longitudinal data were used to test whether internalizing and externalizing symptoms may mediate the association between adolescent-mother and adolescent-father attachment and substance use. Less secure adolescent-father attachment predicted heightened internalizing and externalizing symptoms, and less secure adolescent-father attachment predicted greater alcohol use, mediated through heightened externalizing symptoms. The findings suggest that addressing insecure attachment with fathers during adolescence may reduce unhealthy substance use during the transition to young adulthood.

Is Home Care Use Associated With Secondary Health Conditions in Persons With Spinal Cord Injury? Findings From a Longitudinal Cohort Study.

The aims of the study are to describe the use of professional home care in persons with spinal cord injury over a 10-yr period from 2012 to 2022 and to examine how the use of home care is associated with the future occurrence of secondary health conditions and hospitalizations. This is a longitudinal questionnaire survey as part of the Swiss Spinal Cord Injury Cohort Study, conducted in 2012, 2017, and 2022. There were 1549, 1294, and 1158 participants in the three surveys in 2012, 2017, and 2022. The percentage of home care users decreased from 23% in 2012 to 22% in 2017 and 19% in 2022. Fewer home care recipients reported pain (odds ratio 0.52, 95% confidence interval 0.29-0.94), with no reductions in other secondary health conditions, 5 yrs later. Study participants who received home care were more likely to be hospitalized (odds ratio 1.89, 95% confidence interval 1.05-3.42), 5 yrs later. There is a declining trend in the use of home care services among persons with SCI. While home care is associated with lesser development of pain, it does not generally reduce secondary health conditions. Individuals who receive home care are significantly more likely to report hospitalization 5 yrs later.

Labor Market Participation of Persons With Spinal Cord Injury Living in Switzerland Between 2012 and 2022: A Comparison to the General Population.

The aim of the study is to examine labor market participation of individuals with spinal cord injury living in Switzerland over the past decade and in relation to the general population. This is a descriptive study, using longitudinal panel data from the 2012, 2017, and 2022 community surveys of the Swiss Spinal Cord Injury Cohort Study and general population data from the Swiss Labor Force Survey. Employment rates among spinal cord injury responders were 56% in 2012, 61% in 2017, and 64.6% in 2022. Employment gaps between the spinal cord injury and the Swiss general population samples decreased from 22.5% (2012) to 15.4% (2022). This decrease was particularly evident for females, middle-aged individuals, and people with tetraplegia but contrasted with increased employment gaps in Southwestern Switzerland. By identifying increased employment rates among spinal cord injury responders along with decreased employment gaps in relation to the general population, our study indicates improved opportunities for persons with impairments in the Swiss labor market over the past decade. While there is still room to improve the labor market participation prospects among the Swiss spinal cord injury population, particular efforts at the regional level are required to ensure equal labor market participation opportunities for all individuals with spinal cord injury living in Switzerland.

Impact of Psychosocial Resources on Mental Health During the Transition From the Initial Rehabilitation to Community: The Case of Spinal Cord Injury.

This study aims to examine the change in mental health from the clinical to community setting in persons with spinal cord injury and to identify the role of psychosocial resources for this transition. Longitudinal survey self-report data from N = 240 persons with spinal cord injury from the Swiss Spinal Cord Injury Cohort study. Latent profile analysis identified three profiles for each the clinical and the community setting: low, medium, and high mental health load. Latent transition analysis revealed that persons with spinal cord injury were most likely to stay in the same profile from discharge of initial rehabilitation to community, followed by a decrease from a high mental health load to a medium mental health load and an increase from a low mental health load to a medium mental health load. Individuals staying in the low mental health load profile showed significantly higher levels of psychosocial resources compared to individuals increasing to the medium mental health load profile, whereas individuals who stayed in the high mental health load profile showed lower levels of psychosocial resources compared to individuals decreasing to the medium mental health load profile. This study highlights a positive role of psychosocial resources on mental health transitions, underlining the need for strengthening psychosocial resources beyond initial rehabilitation.

Use and Cost Patterns of Antihypertensive Medications in the United States From 1996 to 2021.

Antihypertensive medication use patterns have likely been influenced by changing costs and accessibility over the past 3 decades. This study examines the relationships between patent exclusivity loss, medication costs, and national health policies on antihypertensive medication use. Using 1996 to 2021 Medical Expenditure Panel Survey data of US adults with hypertension taking at least 1 antihypertensive medication, we conducted a cross-sectional analysis. We explored the associations between patent exclusivity loss, per-pill costs, and Medicare Part D enactment on medication use over time, focusing on the most commonly used medications (lisinopril, amlodipine, losartan, hydrochlorothiazide, and metoprolol). The unweighted sample comprised 50 095 US adults (mean age, 62 years; 53% female). The survey-weighted number of adults taking antihypertensive medications increased from 22 million (95% CIs, 20-23 million) to 55 million (95% CI, 51-60 million) between 1996 and 2021. Loss of patent exclusivity led to increased medication fills, notably for lisinopril, amlodipine, and losartan, which all exhibited within-class dominance. However, per-pill cost decreases coinciding with Medicare Part D did not increase the number of individuals treated or the use of specific antihypertensive medications or classes. The increase in antihypertensive medication use over the past decades highlights the significant impact of loss of patent exclusivity on the uptake in the use of specific medications. These findings underscore the complexity of factors influencing medication use, beyond cost reductions alone, and suggest that policies need to consider multiple facets to effectively improve antihypertensive medication accessibility and utilization.

A multiple behaviour temporal network analysis for health behaviours during COVID-19.

The aim of this study was to examine the temporal dynamics of multiple health behaviours (physical activity, alcohol consumption, healthy eating, cigarette consumption, recreational drug use, vaping), and pandemic-related health behaviours (e.g., hand washing, physical distancing) using network psychometrics. The International COVID-19 Awareness and Responses Evaluation (iCARE) study is an international multi-wave observational cohort study of public awareness, attitudes, and responses to public health policies implemented to reduce the spread of COVID-19 on people around the world. A sub-sample of longitudinal data from Canadians (n = 254) was analysed across four waves (February-July 2020). We used temporal network models to fit temporal networks, contemporaneous networks, and between-subject networks from items within the iCARE survey. Positive temporal associations were observed between physical activity and healthy eating, and a bidirectional relationship was evident between outdoor mask use and vaping. A contemporaneous network revealed positive associations between consumption behaviours (vaping, cigarette use, alcohol use, and recreational drug use), and negative associations between physical activity and drug use, and healthy eating and cigarette use. Health behaviours are interconnected and can be modelled as networks or behavioural systems. The application of temporal network analysis to the study of multiple health behaviours is well suited to address key research questions in the field such as 'how do multiple health behaviours co-vary with one another over time'. Future research using time series data and measuring affective and cognitive mediators of behaviour, in addition to health behaviours, has the potential to contribute valuable hypothesis-generating insights.

The global burden of enteric fever, 2017–2021: a systematic analysis from the global burden of disease study 2021

Enteric fever is a major public health challenge in developing countries. We conducted a systematic analysis from the Global Burden of Diseases 2021 Study to provide updated estimates of enteric fever's burden. We presented estimates for incident cases and deaths, age-standardized incidence and mortality rates, years of life lost (YLLs), and case-fatality rates spanning the study period of 2017-2021, stratified by region, country, socio-demographic index (SDI), and age group. Random-effects Poisson regression for longitudinal data was used to estimate the association between SDI and case-fatality rates, adjusting for antimicrobial resistance patterns. In 2021, there were 9.3 million global cases of enteric fever (95% uncertainty interval: 7.3-11.9) and 107.5 thousand deaths (56.1-180.8). The age-standardized incidence rate decreased from 152/100,000 person-years (118-195) in 2017 to 128/100,000 person-years (100-163) in 2021, and the mortality rate decreased from 1.87/100,000 person-years (0.95-3.18) to 1.50/100,000 person-years (0.78-2.54). There were wide geographical differences, with South Asia contributing the most cases and deaths. Age-standardized incidence exceeded the threshold for "high burden" of enteric fever (100/100,000 person-years) in 23 countries in2021.Children under five accounted for 40% of deaths and 47% of YLLs, with incidence and mortality peaking during the second year. Case-fatality was highest in low SDI countries and showed a global trend toward reduction, except among children aged 1-4 years. After adjusting for the prevalence of multidrug resistance, fluoroquinolone non-susceptibility, and third-generation cephalosporin resistance, a higher SDI was associated with a lower case-fatality rate, with a 1.1% (0.7-1.7) reduction for each percentage point increase in SDI. Despite notable improvements, several countries still showed a high burden of enteric fever, which remains a significant global health concern, especially among children under five. Although enhancing water and sanitation systems is crucial, the most significant reductions in the global disease burden are likely to be achieved through broader vaccine coverage. This includes the use of typhoid conjugate vaccines, which are effective in infants and young children and offer extended protection, along with improved data collection and surveillance to guide vaccine distribution efforts across high-incidence areas. This work was partially supported by "Ricerca Corrente" funding from Italian Ministry of Health to IRCCSHumanitas Research Hospital.

Birth expectations, birth experiences and childbirth-related post-traumatic stress symptoms in mothers and birth companions: Dyadic investigation using response surface analysis.

During the perinatal period, women and their birth companions form expectations about childbirth. We aimed to examine whether a mismatch between birth expectations and experiences predict childbirth-related post-traumatic stress symptoms (CB-PTSS) for mothers and birth companions. We also explored the influence of the mismatch between mothers' and birth companions' expectations/experiences on CB-PTSS. Dyadic longitudinal data from the Self-Hypnosis IntraPartumTrial. Participants (n = 469 mothers; n = 358 birth companions) completed questionnaires at 27 and 36 weeks of gestation and 2 and 6 weeks post-partum. We used the measures of birth expectations (36 weeks gestation), birth experiences (2 weeks post-partum) and CB-PTSS (6 weeks post-partum). Correlations revealed that birth expectations were associated with experiences for both mothers and birth companions but were not consistently associated with CB-PTSS. Birth experiences related to CB-PTSS for both mothers and birth companions. The response surface analysis results showed no support for the effect of a mismatch between expectations and experiences on CB-PTSS in mothers or birth companions. Similarly, a mismatch between mothers' and birth companions' expectations or experiences was unrelated to CB-PTSS. Following previous literature, birth expectations were associated with experiences, and experiences were associated with CB-PTSS. By testing the effect of the match between birth experiences and expectations using an advanced statistical method, we found that experiences play a more substantial role than the match between experiences and expectations in CB-PTSS. The impact of birth experiences on CB-PTSS highlights the importance of respectful and supportive maternity care.

Leveraging social capital for digital transformation: The role of ambidexterity and familial orientation

In pursuing competitive advantage, businesses are diffusing new digital technologies across their value chains and, by doing so, catalyzing significant organizational changes in operational processes and strategic approaches. This widespread evolution of businesses has prompted significant questions regarding the drivers of digital change, the necessary resources, and the capabilities that need to be developed or reconfigured. Our paper focuses on the role of intangible resources, such as a firm's social capital, and its specific implications for the broader research on this phenomenon. Consistent with the resource-based view, we examine the role of different types of social capital as critical resources in the digital transformation process while also considering the mediating role of the balancing act of the capabilities around business alignment and adaptation encompassed in the concept of ambidexterity. Furthermore, we introduce a contextual dimension to our study by examining family involvement in a firm as a significant factor influencing how businesses shape their digital goals and allocate resources toward digital transformation. Our paper addresses this under-explored research field by drawing on a large, cross-sectional dataset from an online panel study with 582 management-level participants from UK firms. After testing our hypotheses, our study's results indicate that bridging and bonding social capital resources significantly promote digital transformation by indirectly developing an ambidextrous orientation. Moreover, familial orientation strengthens both relationships. This research is currently one of the few studies providing empirical evidence on the relationship between digital transformation, its antecedents, and, specifically, the role of intangible resources.

Financial Technology, wealth Threshold, and Wealth Inequality in China

This study developed a theoretical model to examine the impact of financial technology (fintech) on wealth inequality from a wealth threshold perspective. The model's numerical simulation results suggest that fintech exerted an inverted U-shaped impact on wealth inequality. Empirical test results based on China Household Finance Survey panel data and fintech index data confirmed this finding. The impact varied across regions and among groups with differing education levels and financial literacy.

Internet use and household financial market participation: evidence from China

ABSTRACT This paper examines the effect of internet use on household financial market participation and its underlying mechanism. Using data from the 2018 China Family Panel Studies (CFPS), we find that households with internet use are 4.7% more likely to participate in financial market investments than those without. The finding remains robust after accounting for endogeneity concerns using the instrumental variables approach and a fresh double machine learning approach. We highlight the phenomenon of ‘digital divide’ that the positive effects of internet use on making financial investments are more substantial among families living in economically developed regions, and families with higher income and education levels than their counterparts. We identify three novel mechanisms through which internet use affects household investment decision-making, namely the internet information mechanism, the internet social mechanism, and the internet enabling mechanism. These findings suggest that expanding digital infrastructure building and increasing the number of internet users are important ways to enhance households’ participation in financial markets.

Posttraumatic Growth: The Role of Health and Financial Difficulties During a Pandemic.

Trauma can have a range of effects on individuals over time, including the potential for positive changes in favorable outcomes commonly referred to as posttraumatic growth. The posttraumatic growth literature has been criticized for various methodological limitations and has largely neglected the exploration of factors that may strengthen or weaken posttraumatic growth trajectories. The present study contributes to this literature by investigating trajectories of five dimensions of posttraumatic growth during the COVID-19 pandemic and the moderating effects of health and financial difficulties on these trajectories. Longitudinal data were collected monthly between July 2020 and December 2022 from a sample of employed adults in Germany, involving N = 1678 participants over 30 time points (Nobservations = 29,552). Results suggest that certain dimensions of posttraumatic growth showed significant increases (i.e., relationships) or decreases (i.e., spirituality and possibilities) over time and that trajectories of three dimensions (i.e., relationships, appreciation, and strength) demonstrated distinctive patterns of interaction with health and financial difficulties. Findings are discussed in the context of the pandemic as well as dynamic theories of personality and identity, with implications for advancing posttraumatic growth theory and research.

Socio-Economic Determinants of the Physical Activity Intensity in Russia

The study examines the socioeconomic, demographic and health determinants of physical activity intensity in Russia. Using RLMS-HSE (the Russia Longitudinal Monitoring Survey – Higher School of Economics) data for 2021 and indicators of six levels of physical activity intensity, the authors built ordered probit models and calculated average marginal effects for a set of covariates. The independent variables were selected based on the SLOTH microeconomic model describing individual preferences towards physical activity.The ordered probit regression provided empirical support for Meltzer and Jena's theoretical framework, stating that as income increases, with the pronounced substitution effect, the time spent on physical activity decreases, and the intensity of physical activity rises. In our study, the thesis about an increase in intensity with rising income was confirmed for a subsample of men but not women. For women, the intensity was determined by factors such as multiple morbidity and residing outside of Moscow or Saint Petersburg, which were not significant in the regressions for men.The study resulted in several important conclusions. The intensity analysis showed that physician recommendations regarding the type of physical activity should consider patients' workload. For individuals with a busy schedule, recommended types of exercise that require a long time (for example, walking) are not suitable. At the same time, intense workouts (such as tennis or gym classes) can be built into the schedule. There is a need to develop sports infrastructure at workplaces, allowing individuals to allocate time for classes without bearing additional time and monetary costs associated with travel to the location of workouts. As income increases, exercising near the workplace allows an individual to increase the intensity of physical activity without incurring additional time costs

Maternal depressive symptoms in and beyond the perinatal period: Associations with infant and preschooler sleep.

(1) To describe sleep in infancy and early childhood among children born to mothers with and without clinically significant depressive symptoms, and (2) to explore the relationships between maternal depressive symptoms and sleep patterns and problems during infancy and early childhood. Secondary analysis of longitudinal data from the Moe Kura: Mother and Child, Sleep and Wellbeing in Aotearoa/New Zealand study. Data was collected in pregnancy (T1), 12 weeks postpartum (T2), and 3 years post-birth (T3). Participants were 262 Māori and 594 non-Māori mother-child dyads. Chi-square and Independent T-tests measured bivariate associations between maternal mood (T1, T2, T3) and child sleep characteristics (T2, T3). Binary logistic regression models examined longitudinal and concurrent associations between maternal depressive symptoms and infant and preschooler sleep. Adjusted models accounted for key socio-demographic variables, as well as infant sleep variables in preschooler models. Bivariate associations were found between prior and concurrent depressive symptomology and many of the infant and preschooler sleep outcomes. In adjusted models, prenatal depressive symptoms remained independently associated with shorter-than-recommended sleep durations in preschoolers. In these models, concurrent depression was also associated with night waking, night LSRSP, and perceived sleep problems at 12 weeks postpartum, and CSHQ-determined and perceived sleep problems at 3 years post birth. Longitudinal and cross-sectional associations were found between maternal depressive symptoms and child sleep. Sleep appears to be one pathway by which maternal depression confers risk for suboptimal child health outcomes. Findings support the need for earlier and better maternal mental health services.

Impact of retirement on physical activity in older adults: a longitudinal analysis from SHARE

Abstract Background With ageing populations experiencing longer post-retirement years, understanding the influence of major life transitions on health behaviours such as physical activity (PA) is crucial. This study assesses how retirement affects PA levels by analysing longitudinal data from the Survey of Health, Ageing and Retirement in Europe (SHARE). Methods This analysis utilised longitudinal data from 2004 to 2020, categorising PA according to WHO guidelines adapted to individual-level data from successive waves of SHARE. The study examined the relative risks (RR) of being physically active, using generalised estimating equation models to compare periods before and after retirement. Results The cohort consisted of 8,998 individuals aged 50 and older who were employed at baseline and retired during the study period. Initially, 83.6% of participants were categorised as physically active. A noticeable decline in PA was observed post-retirement, with RR decreasing from 1.08 (95%CI 1.07-1.10) more than 10 years before retirement to 0.90 (95%CI 0.88-0.93) more than 10 years after retirement, compared to the year of retirement. Previously inactive individuals often increased their PA levels, peaking 3 to 4 years post-retirement (RR increased from 0.79, 95%CI 0.70-0.88, to 1.49, 95%CI 1.36-1.63), suggesting a temporary boost in moderate-intensity PA during the so-called “honeymoon phase”. Conclusions Retirement can serve as a window of opportunity for enhancing leisure-time PA, particularly among those previously inactive. Sustained increases in PA are less common, highlighting the need for targeted interventions to support continued engagement in PA among retirees. The variations in PA trends suggest that interventions should be tailored to individual characteristics such as age at retirement, previous job type, and educational background to maximise public health impact. This study underscores the challenges of ensuring long-term health benefits through PA. Key messages • Retirement significantly alters physical activity patterns, highlighting an opportunity for targeted health interventions. • Longitudinal data reveals critical periods post-retirement that could benefit from public health strategies to maintain or increase physical activity among older adults.

Food insecurity is associated with poor mental health and health risk behaviors in Czechia

Abstract Background Food insecurity (FI) is one of social determinants of health and a profound public health concern because it leads to malnutrition and hence poor health. COVID-19 impacted people disproportionately and exacerbated preexisting social and health inequalities. The aim of this study was to understand how socioeconomic and health conditions affected one’s FI in Czechia during the pandemic using a representative panel of Czech adults from a longitudinal survey, Life During the Pandemic collected in October 2022 as part of a Czech longitudinal survey. Methods A study sample of 1,499 respondents was included in our multiple logistic regression models. Depressive symptoms and anxiety symptoms were assessed using validated questionnaires PHQ-2 and DAG-2. Results Approximately one third of the study sample (32.4%) reported being food insecure. In the fully-adjusted model, adults with depressive symptoms (OR 2.37, CI 1.49, 5.02), problems with sleeping (OR 2.20, CI 1.64, 2.96), smoking habit (OR 2.27, CI 1.68, 3.07), lower fruit and vegetable consumption (OR 1.50, CI 1.04, 2.17) had greater odds of experiencing FI. The lowest income group was associated with 6.2 times greater odds of reporting FI as compared to the high-income group (CI 3.96, 9.67). Lower education attainment also had higher odds of experiencing FI (OR 2.15, CI 1.40, 3.31). Anxiety was not a significant predictor for FI. Conclusions Our study highlights that food insecurity is associated with depressive symptoms, poor quality of sleep and lower consumption of fruit and vegetables among adults. Food insecurity was also evident among lower income and lower education attainment. Funding source This study was supported by The Charles University Grant Agency (GA UK). Key messages • Our study suggests that adults with poor mental health and diet quality experience food insecurity. • Given the established link between a balanced diet and mental well-being, interventions should encompass not only food security but also the quality of dietary intake.

Does perceived neighbourhood social cohesion moderate the effects of demographic changes on health?

Abstract Small-area characteristics changes over time may be associated with health. Various theories have been proposed to explain small-area effects on health, including those concerning social capital. The effect of small-area demographic changes on health are very little researched and hardly anything is known about their relationship with social cohesion. We use longitudinal data from the German Socio-Economic Panel linked to demographic measures at county level. We explore the association between 5-year changes in the proportion of over 65 and of 18-24 years-old and physical and mental health and the moderating effect of perceived neighbourhood social cohesion (SC) using random intercept models. Control variables include education, income, and relevant baseline measures. Included are all participants with a 6-year continuous exposure to their county of residence. The main outcomes are the physical and mental health components (PCS, MCS) of the SF-12. We included 21,239 persons up to 6 times in 399 counties. 5-year changes in the over 65, of 18-24 and measure of SC are associated to MCS and PCS. Significant interactions between the changes in over 65 and social cohesion are seen for MCS only in rural areas and some interactions are significant for PCS. For MCS there are no interactions between social cohesion and the changes in 18-24. Important interaction effects for PCS: for those exposed to a small decrease in 18-24, an increase in one unit of SC is associated with an increase of 1.71 (95% CI[1.38, 2.03]) PCS units. For those exposed to a large decrease this effect is lower by 0.88 PCS units. Exposition to some demographic changes have the effect to reduce the positive effects of perceived social cohesion on health. Thus, demographic changes appear to be health stressors. The evidence shown by this work need to be further developed by considering smaller scales but the evidence opens up a new range of possible intervention targets to reduce health inequalities. Key messages • County level demographic changes are associated with physical and mental health and certain exposition to changes reduce the positive effects of perceived neighbourhood social cohesion. • Social cohesion can be an intervention target to reduce small-area health inequalities, population changes can annulate positive effects. This must be accounted for in public health interventions.

Optimizing elderly care: a data-driven AI model for predicting polypharmacy risk using SHARE data

Abstract Background Polypharmacy (PP), which rises with age, is a growing public health challenge, affecting health outcomes and escalating healthcare expenditure. Leveraging the comprehensive European coverage and longitudinality of the Survey of Health, Ageing, and Retirement in Europe (SHARE) study, we aim to assess PP prevalence patterns and identify the most effective machine learning (ML) model for predicting long-term PP risk, as a foundation for a scalable predictive tool. Methods We used data from participants aged above 50 who were present in wave 6 and at least one of the subsequent three waves of the SHARE study, aiming to predict PP risk at 2, 4, and 6-year intervals. PP was defined as the concurrent use of five or more medications. We analyzed PP prevalence trends and selected the predictor variables from wave 6 using LASSO regression analysis. We evaluated eight ML models, namely, ANN, SVM, DT, RF, GB, XGBoost, LightGBM, and CatBoost, using a rigorous cross-validation strategy to ensure robustness and reliability. Results Our analysis reveals an upward trend in PP prevalence across the surveyed countries, with aggregate figures rising from 34.03% (95% CI 33.1-34.9) in wave 7 to 36.75% (95% CI 35.6-37.9) in wave 8, reaching 39.91 (95% CI 38.9-40.9) in wave 9. Additionally, using the Categorical Boosting ML model to predict PP resulted in overall accuracies of 75.08%, 73.7%, and 71.65% and recall rates of 72.83%, 70.48%, and 67.96% for the 2, 4, and 6-year intervals, respectively. Conclusions This study reveals a rising trend of PP across European countries and demonstrates the potential of using longitudinal data and ML to enhance PP prediction. The tool developed represents a step forward in risk stratification, which would be particularly beneficial in practical settings where family physicians or pharmacists could employ the tool to monitor elderly patients, predict and thus prevent PP, and reduce the negative health and economic impact associated with it. Key messages • The integration of machine learning with longitudinal data presents a significant advance in polypharmacy risk prediction. • The tool we developed highlights the potential translational impact of our findings.