Publications

Background: Most previous studies of frailty trajectories in older adults focus on the average trajectory and ignore death. Longitudinal quantile analysis of frailty trajectories permits the definition of reference curves, and the application of mortal cohort inference provides more realistic estimates than models that ignore death.

Methods: Using data from individuals aged 65 or older (n=25446) from the Survey of Health, Ageing, and Retirement in Europe (SHARE) from 2004 to 2020, we derived repeated values of the Frailty Index (FI) based on the accumulation of health deficits. We applied weighted Generalized Estimating Equations to estimate the quantiles of the FI trajectory, adjusting for sample attrition due to death, sex, education, and cohort.

Results: The FI quantiles increased with age and progressed faster for those with the highest level of frailty (βa0.9=0.0229, p<0.001; βa0.5=0.0067, p<0.001; H0: βa0.5=βa0.9, p<0.001). Education was consistently associated with a slower progression of the FI in all quantiles (βae0.1=-0.0001, p<0.001; βae0.5=-0.0004, p<0.001; βae0.9=-0.0003, p<0.001) but sex differences varied across the quantiles. Women with the highest level of frailty showed a slower progression of the FI than men when considering death. Finally, no cohort effects were observed for the FI progression.

Conclusions: Quantile FI trajectories varied by age, sex, education, and cohort. These differences could inform the practice of interventions aimed at older adults with the highest level of frailty.

Background: It is currently unclear whether (and when) physical function exhibits a terminal decline phase, that is, a substantial acceleration of decline in the very last years before death.

Methods: 702 deceased adults aged 70 years and older from the Yale PEP Study provided 4 133 measurements of physical function (Short Physical Performance Battery, SPPB) up to 20 years before death. In addition, continuous gait and chair rise subtest scores (in seconds) were assessed. Generalized mixed regression models with random change points were used to estimate the onset and the steepness of terminal decline in physical function.

Results: Decline accelerated in the last years of life in all 3 measures of physical function. The onset of terminal decline occurred 1 year before death for the SPPB, and at 2.5 and 2.6 years before death for chair rise and gait speed test scores, respectively. Terminal declines in physical function were 6-8 times steeper than pre-terminal declines. Relative to those whose condition leading to death was frailty, participants who died from dementia and cancer had an up to 6 months earlier and 3 months later onset of terminal decline in SPPB, respectively.

Conclusions: Terminal decline in physical function among older adults is comparable to the more established terminal decline phenomenon in cognition. Our results provide additional evidence of late-life rapid decline in physical function due to impending death.

Background: Life course epidemiology examines associations between repeated measures of risk and health outcomes across different phases of life. Empirical research, however, is often based on discrete-time models that assume that sporadic measurement occasions fully capture underlying long-term continuous processes of risk.

Methods: We propose (i) the functional relevant life course model (fRLM), which treats repeated, discrete measures of risk as unobserved continuous processes, and (ii) a testing procedure to assign probabilities that the data correspond to conceptual models of life course epidemiology (critical period, sensitive period and accumulation models). The performance of the fRLM is evaluated with simulations, and the approach is illustrated with empirical applications relating body mass index (BMI) to mRNA-seq signatures of chronic kidney disease, inflammation and breast cancer.

Results: Simulations reveal that fRLM identifies the correct life course model with three to five repeated assessments of risk and 400 subjects. The empirical examples reveal that chronic kidney disease reflects a critical period process and inflammation and breast cancer likely reflect sensitive period mechanisms.

Conclusions: The proposed fRLM treats repeated measures of risk as continuous processes and, under realistic data scenarios, the method provides accurate probabilities that the data correspond to commonly studied models of life course epidemiology. fRLM is implemented with publicly-available software.

Multimorbidity-the coexistence of at least two chronic health conditions within the same individual-is an important global health challenge. In high-income countries (HICs), multimorbidity is dominated by non-communicable diseases (NCDs); whereas, the situation may be different in low- and middle-income countries (LMICs), where chronic communicable diseases remain prominent. The aim of this systematic review was to identify determinants (including risk and protective factors) and potential mechanisms underlying multimorbidity from published longitudinal studies across diverse population-based or community-dwelling populations in LMICs. We systematically searched three electronic databases (Medline, Embase, and Global Health) using pre-defined search terms and selection criteria, complemented by hand-searching. All titles, abstracts, and full texts were independently screened by two reviewers from a pool of four researchers. Data extraction and reporting were according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Methodological quality and risk of bias assessment was performed using the Newcastle-Ottawa Scale for cohort studies. Data were summarized using narrative synthesis. The search yielded 1782 records. Of the 52 full-text articles included for review, 8 longitudinal population-based studies were included for final data synthesis. Almost all studies were conducted in Asia, with only one from South America and none from Africa. All studies were published in the last decade, with half published in the year 2021. The definitions used for multimorbidity were heterogeneous, including 3-16 chronic conditions per study. The leading chronic conditions were heart disease, stroke, and diabetes, and there was a lack of consideration of mental health conditions (MHCs), infectious diseases, and undernutrition. Prospectively evaluated determinants included socio-economic status, markers of social inequities, childhood adversity, lifestyle behaviors, obesity, dyslipidemia, and disability. This review revealed a paucity of evidence from LMICs and a geographical bias in the distribution of multimorbidity research. Longitudinal research into epidemiological aspects of multimorbidity is warranted to build up scientific evidence in regions beyond Asia. Such evidence can provide a detailed picture of disease development, with important implications for community, clinical, and interventions in LMICs. The heterogeneity in study designs, exposures, outcomes, and statistical methods observed in the present review calls for greater methodological standardisation while conducting epidemiological studies on multimorbidity. The limited evidence for MHCs, infectious diseases, and undernutrition as components of multimorbidity calls for a more comprehensive definition of multimorbidity globally.

Background: Unmet health needs have the potential to capture health inequality. Nevertheless, the course of healthcare needs fulfilment, and the role of multimorbidity in this process remains unclear. This study assessed the bidirectional transitions between met and unmet health needs and the transition to death and examined the effect of multimorbidity on transitions.

Methods: This study was based on the China Health and Retirement Longitudinal Study, a nationally representative survey in 2011-2015 among 18 075 participants aged 45 and above (average age 61.1; SD 9.9). We applied a multistate survival model to estimate the probabilities and the instantaneous risk of state transitions, and Gompertz hazard models were fitted to estimate the total, marginal and state-specific life expectancies (LEs).

Results: Living with physical multimorbidity (HR=1.85, 95% CI 1.58 to 2.15) or physical-mental multimorbidity (HR=1.45, 95% CI 1.15 to 1.82) was associated with an increased risk of transitioning into unmet healthcare needs compared with no multimorbidity. Conversely, multimorbidity groups had a decreased risk of transitioning out of unmet needs. Multimorbidity was also associated with shortened total life expectancy (TLEs), and the proportion of marginal LE for having unmet needs was more than two times higher than no multimorbidity.

Conclusion: Multimorbidity aggravates the risk of transitioning into having unmet healthcare needs in the middle and later life, leading to a notable reduction in TLEs, with longer times spent with unmet needs. Policy inputs on developing integrated person-centred services and specifically scaling up to target the complex health needs of ageing populations need to be in place.

To date, there is a considerable heterogeneity of methods to score Allostatic Load (AL). Here we propose a comprehensive algorithm (ALCS) that integrates commonly used approaches to generate AL risk categories and assess associations to brain structure deterioration. In a cohort of cognitively normal mid-life adults (n = 620, age 51.3 ± 5.48 years), we developed a comprehensive composite for AL scoring incorporating gender and age differences, high quartile approach, clinical reference values, and current medications, to then generate AL risk categories. Compared to the empirical approach (ALES), ALCS showed better model fit criteria and a strong association with age and sex. ALSC categories were regressed against brain and white matter hyperintensity (WMH) volumes. Higher AL risk categories were associated with increased total, periventricular, frontal, and left parietal WMH volumes, also showing better fit compared to ALES. When cardiovascular biomarkers were removed from the ALSC algorithm, only left-frontal WMHV remained associated with AL, revealing a strong vascular burden influencing the index. Our results agree with previous evidence and suggest that sustained stress exposure enhances brain deterioration in mid-life adults. Showing better fit than ALES, our comprehensive algorithm can provide a more accurate AL estimation to explore how stress exposure enhances age-related health decline.

Objectives: To investigate the availability of Alzheimer's Centers (ACs) in US hospitals.

Methods: Utilizing the American Hospital Association Annual Survey, Area Health Resource File, and US Census (n = 3251), we employed multivariable logistic regression to examine hospital, county, and regional predictors of AC availability.

Results: Large hospitals (>399 beds) had approximately 14 times higher odds of having an AC than small hospitals (<50 beds; OR = 14.0; 95% CI = 6.44 - 30.46). Counties with a higher proportion of Latino residents, relative to non-Latino Whites, had lower odds of having an AC (OR = .05; 95% CI = .01 - .41). Northeastern (OR = 1.92; 95% CI = 1.15 - 3.22) and Midwestern (OR = 2.12; 95% CI = 1.34 - 3.37) hospitals had higher odds of having an AC than Southern hospitals.

Discussion: To address dementia needs and disparities, investment in a national infrastructure is critical.

Background: While prolonged labour market participation becomes increasingly important in ageing societies, evidence on the impacts of entering or exiting work beyond age 65 on cognitive functioning is scarce.

Methods: We use data from two large population-representative data sets from South Korea and the USA to investigate and compare the effects of the labour market (re-)entry and exit by matching employment and other confounder trajectories prior to the exposure. We chose the Korean Longitudinal Study of Aging (N=1872, 2006-2020) for its exceptionally active labour participation in later life and the Health and Retirement Study (N=4070, 2006-2020) for its growing inequality among US older adults in labour participation. We use the matching difference-in-differences (DID) method, which allows us to make causal claims by reducing biases through matching.

Results: We find general positive effects of entering the labour market in South Korea (DID estimate: 0.653, 95% CI 0.167 to 1.133), while in the USA such benefit is not salient (DID estimate: 0.049, 95% CI -0.262 to 0.431). Exiting the late-life labour market leads to cognitive decline in both South Korea (DID estimate: -0.438, 95% CI -0.770 to -0.088) and the USA (DID estimate: -0.432, 95% CI -0.698 to -0.165).

Conclusions: Findings suggest that Korean participants cognitively benefited from late-life labour market participation, while US participants did not. Differences in participant characteristics and reasons for labour market participation may have led to the differential findings. We found the negative effects of exiting the late-life labour force in both countries.

Objectives: This paper models cognitive aging, across mid and late life, and estimates birth cohort and sex differences in both initial levels and aging trajectories over time in a sample with multiple cohorts and a wide span of ages.

Methods: The data used in this study came from the first 9 waves of the English Longitudinal Study of Ageing, spanning 2002-2019. There were n = 76,014 observations (proportion male 45%). Dependent measures were verbal fluency, immediate recall, delayed recall, and orientation. Data were modeled using a Bayesian logistic growth curve model.

Results: Cognitive aging was substantial in 3 of the 4 variables examined. For verbal fluency and immediate recall, males and females could expect to lose about 30% of their initial ability between the ages of 52 and 89. Delayed recall showed a steeper decline, with males losing 40% and females losing 50% of their delayed recall ability between ages 52 and 89 (although females had a higher initial level of delayed recall). Orientation alone was not particularly affected by aging, with less than a 10% change for either males or females. Furthermore, we found cohort effects for initial ability level, with particularly steep increases for cohorts born between approximately 1930 and 1950.

Discussion: These cohort effects generally favored later-born cohorts. Implications and future directions are discussed.

Background and objectives: The existing literature highlights the importance of reading books in middle-to-older adulthood for cognitive functioning; very few studies, however, have examined the importance of childhood cognitive resources for cognitive outcomes later in life.

Research design and methods: Using data from 11 countries included in the Survey of Health, Ageing, and Retirement in Europe (SHARE) data set (N = 32,783), multistate survival models (MSMs) were fit to examine the importance of access to reading material in childhood on transitions through cognitive status categories (no cognitive impairment and impaired cognitive functioning) and death. Additionally, using the transition probabilities estimated by the MSMs, we estimated the remaining years of life without cognitive impairment and total longevity. All models were fit individually in each country, as well as within the pooled SHARE sample.

Results: Adjusting for age, sex, education, and childhood socioeconomic status, the overall pooled estimate indicated that access to more books at age 10 was associated with a decreased risk of developing cognitive impairment (adjusted hazard ratio = 0.79, confidence interval: 0.76-0.82). Access to childhood books was not associated with risk of transitioning from normal cognitive functioning to death, or from cognitive impairment to death. Total longevity was similar between participants reporting high (+1 standard deviation [SD]) and low (-1 SD) number of books in the childhood home; however, individuals with more access to childhood books lived a greater proportion of this time without cognitive impairment.

Discussion and implications: Findings suggest that access to cognitive resources in childhood is protective for cognitive aging processes in older adulthood.