Associations of Social Support With Physical and Mental Health Symptom Burden After COVID-19 Hospitalization Among Older Adults.

BACKGROUND: Despite significant support system disruptions during the coronavirus 2019 (COVID-19) pandemic, little is known about the relationship between social support and symptom burden among older adults following COVID-19 hospitalization. METHODS: From a prospective cohort of 341 community-living persons aged ≥60 years hospitalized with COVID-19 between June 2020 and June 2021 who underwent follow-up at 1, 3, and 6 months after discharge, we identified 311 participants with ≥1 follow-up assessment. Social support prehospitalization was ascertained using a 5-item version of the Medical Outcomes Study Social Support Survey (range, 5-25), with low social support defined as a score ≤15. At hospitalization and each follow-up assessment, 14 physical symptoms were assessed using a modified Edmonton Symptom Assessment System inclusive of COVID-19-relevant symptoms. Mental health symptoms were assessed using Patient Health Questionnaire-4. Longitudinal associations between social support and physical and mental health symptoms, respectively, were evaluated through multivariable regression. RESULTS: Participants' mean age was 71.3 years (standard deviation, 8.5), 52.4% were female, and 34.2% were of Black race or Hispanic ethnicity. 11.8% reported low social support. Over the 6-month follow-up period, low social support was independently associated with higher burden of physical symptoms (adjusted rate ratio [aRR], 1.26; 95% confidence interval [CI], 1.05-1.52), but not mental health symptoms (aRR, 1.14; 95% CI, 0.85-1.53). CONCLUSIONS: Low social support is associated with greater physical, but not mental health, symptom burden among older survivors of COVID-19 hospitalization. Our findings suggest a potential need for social support screening and interventions to improve post-COVID-19 symptom management in this vulnerable group.

A Joint Model for Disability, Self-Rated Health, and Mortality Among Medicare Beneficiaries-Differences by Chronic Disease and Race/Ethnicity.

OBJECTIVES: Quantifying interdependence in multiple patient-centered outcomes is important for understanding health declines among older adults. METHODS: Medicare-linked National Health and Aging Trends Study data (2011-2015) were used to estimate a joint longitudinal logistic regression model of disability in activities of daily living (ADL), fair/poor self-rated health (SRH), and mortality. We calculated personalized concurrent risk (PCR) and typical concurrent risk (TCR) using regression coefficients. RESULTS: For fair/poor SRH, highest odds were associated with COPD. For mortality, highest odds were associated with dementia, hip fracture, and kidney disease. Dementia and hip fracture were associated with highest odds of ADL disability. Hispanic respondents had highest odds of ADL disability. Hispanic and NH Black respondents had higher odds of fair/poor SRH, ADL disability, and mortality. PCRs/TCRs demonstrated wide variability for respondents with similar sociodemographic-multimorbidity profiles. DISCUSSION: These findings highlight the variability of personalized risk in examining interdependent outcomes among older adults.

Associations between dementia staging, neuropsychiatric behavioral symptoms, and divorce or separation in late life: A case control study.

Dementia can be difficult for married couples for many reasons, including the introduction of caregiving burden, loss of intimacy, and financial strain. In this study, we investigated the impact of dementia staging and neuropsychiatric behavioral symptoms on the likelihood of divorce or separation for older adult married couples. For this case-control study, we used data from the National Alzheimer's Coordinating Center (NACC) Uniform dataset (UDS) versions 2 and 3. This dataset was from 2007 to 2021 and contains standardized clinical information submitted by NIA/NIH Alzheimer's Disease Research Centers (ADRCs) across the United States (US). This data was from 37 ADRCs. We selected participants who were married or living as married/domestic partners at their initial visit. Cases were defined by a first divorce/separation occurring during the follow-up period, resulting in 291 participants. We selected 5 controls for each married/living as married case and matched on age. Conditional logistic regression estimated the association between overall Neuro Psychiatric Inventory (NPI) score and severity of individual symptoms of the NPI with case/control status, adjusted for education, the CDR® Dementia Staging Instrument score, living situation, symptom informant, sex, and race. Separate analyses were conducted for each symptom. Multiple comparisons were accounted for with the Hochberg method. Later stage of dementia was negatively associated with divorce/separation with an adjusted odds ratio (AOR) = 0.68 (95%CI = 0.50 to 0.93). A higher overall NPI score was positively associated with divorce/separation AOR = 1.08 (95% CI = 1.03 to 1.12,). More severe ratings of agitation/aggression, depression/dysphoria, disinhibition, and elation/euphoria were associated with greater odds of divorce/separation. Among older adults in the US, a later stage of dementia is associated with a lower likelihood of divorce or separation, while having more severe neuropsychiatric behavioral symptoms of agitation/aggression, depression/dysphoria, disinhibition, and elation/euphoria are associated with a higher likelihood of divorce or separation.

Longitudinal associations between cognitive functioning and depressive symptoms among couples in the Mexican Health and Aging Study.

OBJECTIVE: To examine the bidirectional associations between older adult spouses' cognitive functioning and depressive symptoms over time and replicate previous findings from the United States (US) in Mexico. DESIGN: Longitudinal, dyadic path analysis with the actor-partner interdependence model. SETTING: Data were from the three most recent interview waves (2012, 2015, and 2018) of the Mexican Health and Aging Study (MHAS), a longitudinal national study of adults aged 50+ years in Mexico. PARTICIPANTS: Husbands and wives from 905 community-dwelling married couples (N = 1,810). MEASUREMENTS: The MHAS cognitive battery measured cognitive function. Depressive symptoms were assessed using a modified nine-item Center for Epidemiologic Studies Depression Scale. Baseline covariates included age, education, number of children, limitation with any activity of daily living, limitation with any instrumental activity of daily living, and pain. RESULTS: As hypothesized, there were significant within-individual associations in which one person's own cognitive functioning and own depressive symptoms predicted their own follow-up cognitive functioning and depressive symptoms, respectively. In addition, a person's own cognitive functioning predicted their own depressive symptoms, and a person's own depressive symptoms predicted their own cognitive functioning over time. As hypothesized, there was a significant partner association such that one person's depressive symptoms predicted more depressive symptoms in the partner. CONCLUSION: Findings from this study of older Mexican couples replicates findings from studies of older couples in the US, showing that depressive symptoms in one partner predict depressive symptoms in the other partner over time; however, there was no evidence for cognition-depression partner associations over time.

Rationale, Design, and Characteristics of the VALIANT (COVID-19 in Older Adults: A Longitudinal Assessment) Cohort.

BACKGROUND: Most older adults hospitalized with COVID-19 survive their acute illness. The impact of COVID-19 hospitalization on patient-centered outcomes, including physical function, cognition, and symptoms, is not well understood. To address this knowledge gap, we collected longitudinal data about these issues from a cohort of older survivors of COVID-19 hospitalization. METHODS: We undertook a prospective study of community-living persons age ≥ 60 years who were hospitalized with COVID-19 from June 2020-June 2021. A baseline interview was conducted during or up to 2 weeks after hospitalization. Follow-up interviews occurred at one, three, and six months post-discharge. Participants completed comprehensive assessments of physical and cognitive function, symptoms, and psychosocial factors. An abbreviated assessment could be performed with a proxy. Additional information was collected from the electronic health record. RESULTS: Among 341 participants, the mean age was 71.4 (SD 8.4) years, 51% were women, and 37% were of Black race or Hispanic ethnicity. Median length of hospitalization was 8 (IQR 6-12) days. All but 4% of participants required supplemental oxygen, and 20% required care in an intensive care unit or stepdown unit. At enrollment, nearly half (47%) reported at least one preexisting disability in physical function, 45% demonstrated cognitive impairment, and 67% were pre-frail or frail. Participants reported a mean of 9 of 14 (SD 3) COVID-19-related symptoms. At the six-month follow-up interview, more than a third of participants experienced a decline from their pre-hospitalization function, nearly 20% had cognitive impairment, and burdensome symptoms remained highly prevalent. CONCLUSIONS: We enrolled a diverse cohort of older adults hospitalized with COVID-19 and followed them after discharge. Functional decline was common, and there were high rates of persistent cognitive impairment and symptoms. Future analyses of these data will advance our understanding of patient-centered outcomes among older COVID-19 survivors.

Activities of Daily Living Needs and Support in Adult Child-Parent Dyads.

This study examined whether changes in middle-aged children's perceptions of their parents' activities of daily living needs (ADL needs) were associated with changes in the mutuality of support in their relationship. A group of 366 middle-aged children in Waves 1 (2008) and 2 (2013) of the Family Exchanges Study self-reported providing and receiving tangible, emotional, and informational support to and from their n = 468 parents. Increased perceived parental ADL needs were associated with increased provision of tangible and informational support to parents but not with changes in support received. Increases in perceived parental ADL needs were associated with higher incongruence for all three support types (the child providing more support than they receive).

The Contribution of Chronic Conditions to Hospitalization, Skilled Nursing Facility Admission, and Death: Variation by Race.

Multimorbidity (≥2 chronic conditions) is a common and important marker of aging. To better understand racial differences in multimorbidity burden and associations with important health-related outcomes, we assessed differences in the contribution of chronic conditions to hospitalization, skilled nursing facility admission, and mortality among non-Hispanic Black and non-Hispanic White older adults in the United States. We used data from a nationally representative study, the National Health and Aging Trends Study, linked to Medicare claims from 2011-2015 (n = 4,871 respondents). This analysis improved upon prior research by identifying the absolute contributions of chronic conditions using a longitudinal extension of the average attributable fraction for Black and White Medicare beneficiaries. We found that cardiovascular conditions were the greatest contributors to outcomes among White respondents, while the greatest contributor to outcomes for Black respondents was renal morbidity. This study provides important insights into racial differences in the contributions of chronic conditions to costly health-care utilization and mortality, and it prompts policy-makers to champion delivery reforms that will expand access to preventive and ongoing care for diverse Medicare beneficiaries.

The geriatrics research instrument library: A resource for guiding instrument selection for researchers studying older adults with multiple chronic conditions.

Background: After the passage of the 21st Century Cures Act in the U.S., the Inclusion Across the Lifespan policy eliminates upper-age limits for research participation unless risk justified. Broader inclusion will necessitate the use of reliable instruments in research that characterize the health status and function of older adults with multiple chronic conditions. As there is a plethora of such instruments, the Geriatrics Research Instrument Library (GRIL) was developed as freely available online resource of data collection instruments commonly used in gerontological research. GRIL has been revised and updated by the Advancing Geriatrics Infrastructure and Network Growth (AGING) Initiative, a joint endeavor of the Health Care Systems Research Network (HCSRN) and the Older Americans Independence Centers (OAICs). Methods: Extensive PubMed literature searches and domain expert feedback were utilized to inventory and update GRIL through the addition of instruments and compiling of instrument metadata. GRIL is hosted on the National Institute on Aging OAIC Coordinating Center website with a platform utilizing Microsoft Structured Query Language (SQL) and an Adobe ColdFusion application server. Tracking statistics are collected using Google Analytics. Results: Presently, GRIL includes 175 instruments across 18 domains, including instrument metadata such as instrument description, copyright information, completion time estimates, keywords, available translations, and a link and reference to the original manuscript describing the instrument. The GRIL website includes user-friendly features such as mobile platforming and resource links. Conclusions: GRIL provides a user-friendly public resource that facilitates clinical researchers in efficiently selecting appropriate instruments to measure clinical outcomes relevant to older adults across a full range of domains.

Patterns of Caregiving Among Older Adults With and Without Dementia: A Latent Class Analysis.

OBJECTIVES: We identify common patterns of caregiving or "care types" among older adults with and without dementia. Prior research has focused on primary caregivers and on the independent effects of individual caregiving attributes. We examine multiple caregivers of older adults, including the primary caregiver, and how multiple caregiving attributes co-occur to shape caregiving types. METHODS: We link 2015 care recipient (N = 1,423) and unpaid caregiver data (N = 2,146) from the National Health and Aging Trends Study and the National Study of Caregiving. Latent class analysis of caregiving attributes, representing care intensity and regularity, and various care activities, was used to construct care types. Multinomial logistic regression was used to examine if the recipients' dementia status and caregivers' background characteristics predicted membership in care types. RESULTS: Five distinguishable care types were identified. Caregivers who were female, adult children, or coresidents, those caring for persons with dementia, and those who had paid help had higher odds of being in the more demanding care types. Conversely, older, White caregivers and those with support for their caregiving activities were less likely to be in a demanding care type. DISCUSSION: Care types can help us understand sources of heterogeneity in caregiving and effectively target caregiver support services and interventions.

Mealtime Behavior and Depressive Symptoms in Late-Life Marriage.

This study examined whether one spouse's mealtime behaviors were associated with their own and their partner's depressive symptoms among older, married couples. We examined gender differences in these associations and tested marital satisfaction as a mediator of these associations. 101 couples self-reported mealtime behavior (number of meals, snacks, fast-food meals, and meals eaten alone), depressive symptoms, and marital satisfaction. Results of the Actor Partner Interdependence Model revealed a statistically significant actor effect of number of fast-food meals on depressive symptoms and a significant partner effect of number of fast-food meals and number of meals eaten alone on depressive symptoms. There were gender differences. Husbands' marital satisfaction mediated the effect of meals eaten alone on depressive symptoms. Wife's marital satisfaction mediated the effect of the husband's meals eaten alone, and wife's number of fast-food meals on the wife's depressive symptoms. Findings have implications for dyadic interventions to improve depressive symptoms.

Examining Health Care Mobility of Transgender Veterans Across the Veterans Health Administration.

Purpose: Transgender veterans are overrepresented in the Veterans Health Administration (VHA) compared with in the general population. Utilization of multiple different health care systems, or health care mobility, can affect care coordination and potentially affect outcomes, either positively or negatively. This study examines whether transgender veterans are more or less health care mobile than nontransgender veterans and compares the patterns of geographic mobility in these groups. Methods: Using an established cohort (n = 5,414,109), we identified 2890 transgender veterans from VHA electronic health records from 2000 to 2012. We compared transgender and nontransgender veterans on sociodemographic, clinical, and health care system-level measures and conducted conditional logistic regression models of mobility. Results: Transgender veterans were more likely to be younger, White, homeless, have depressive disorders, post-traumatic stress disorder (PTSD), and hepatitis C. Transgender veterans were more likely to have been health care mobile (9.9%) than nontransgender veterans (5.2%) (unadjusted odds ratio = 2.02, 95% confidence interval = 1.73-2.36). In a multivariable model, transgender status, being separated/divorced, receiving care in less-complex facilities, and diagnoses of depression, PTSD, or hepatitis C were associated with more mobility, whereas older age was associated with less mobility. For the top three health care systems utilized, a larger proportion of transgender veterans visited a second health care system in a different state (56.2%) than nontransgender veterans (37.5%). Conclusions: Transgender veterans were more likely to be health care mobile and more likely to travel out of state for health care services. They were also more likely to have complex chronic health conditions that require multidisciplinary care.

Individual Heterogeneity in the Probability of Hospitalization, Skilled Nursing Facility Admission, and Mortality.

BACKGROUND: Multimorbidity is common in adults aged 65 and older and is associated with health care utilization and mortality, but most methods ignore the interrelationship among concurrent outcome nor provide person-specific probabilities. METHOD: A longitudinal cohort of 5300 older Americans from the 2011-2015 rounds of the National Health and Aging Study was linked to Center for Medicare and Medicaid Services claims. Odds ratios for 15 chronic conditions adjusted for sociodemographic factors were estimated using a joint model of hospitalization, skilled nursing facility (SNF) admission, and mortality. Additionally, we estimated the person-specific probability of an outcome while currently at risk for other outcomes for different chronic disease combinations demonstrating the heterogeneity across persons with identical chronic conditions. RESULTS: During the 4-year follow-up period, 2867 (54.1%) individuals were hospitalized, 1029 (19.4%) were admitted to a SNF, and 1237 (23.3%) died. Chronic kidney disease, dementia, heart failure, and chronic obstructive pulmonary disease had significant increased odds for all 3 outcomes. By incorporating a person-specific random intercept, there was considerable range of person-specific probabilities for individuals with hypertension, diabetes, and depression with dementia, (hospitalization: 0.14-0.61; SNF admission: 0.04-0.28) and without dementia (hospitalization: 0.07-0.44; SNF admission: 0.02-0.15). Such heterogeneity was found among individuals with heart failure, ischemic heart disease, chronic kidney disease, hypertension, hyperlipidemia, and osteoarthritis with and without Medicare. CONCLUSIONS: This approach of joint modeling of interrelated concurrent health care and mortality outcomes not only provides a cohort-level odds and probabilities but addresses the heterogeneity among otherwise similarly characterized persons identifying those with above-average probability of poor outcomes.

Diffusing capacity in normal-for-age spirometry and spirometric impairments, using reference equations from the global lung function initiative.

BACKGROUND: Reference equations from the Global Lung Function Initiative (GLI) are now available for both spirometry and diffusion. However, respiratory phenotypes defined by GLI-based measures of diffusion have not yet been evaluated in GLI-based normal-for-age spirometry or spirometric impairments. METHODS: We evaluated cross-sectional data from 2100 Caucasians, aged 40-85 years. GLI-based spirometric categories included normal-for-age and the impairments of restrictive-pattern and three-level severity of airflow-obstruction (mild, moderate, severe). GLI-based diffusion included diffusing capacity of the lung for carbon monoxide (DLCO) and measured components of alveolar volume (VA) and transfer coefficient (KCO): DLCO = [VA]x[KCO]. Using multivariable regression models, adjusted odds ratios (adjORs) for DLCO, VA, and KCO < lower limit of normal (LLN) were calculated for spirometric impairments, relative to normal-for-age spirometry. RESULTS: Relative to normal-for-age spirometry, the restrictive-pattern increased the adjORs (95% confidence intervals) for DLCO and VA < LLN-4.61 (3.62, 5.85) and 15.53 (11.8, 20.4), respectively, but not for KCO < LLN-1.02 (0.79, 1.33). Also relative to normal-for-age spirometry, airflow-obstruction from mild to severe increased the adjORs for DLCO < LLN-from 1.22 (0.80, 1.86) to 6.63 (4.91, 8.95), for VA < LLN-from 1.37 (0.85, 2.18) to 7.01 (5.20, 9.43), and for KCO < LLN-from 2.04 (1.33, 3.14) to 3.03 (2.29, 3.99). Notably, in normal-for-age spirometry, 34.5%, 19.7%, and 25.3% of participants had DLCO, VA, or KCO < LLN, respectively. CONCLUSION: Abnormal diffusion is most prevalent in spirometric impairments but also occurs in normal-for-age spirometry. These results further inform the respiratory phenotypes of GLI-based spirometric categories and, in turn, the spirometric evaluation of respiratory disease.

Antihypertensive medications and physical function in older persons.

BACKGROUND: To further inform benefits and risks of medications on physical function in aging populations, we have evaluated the associations of antihypertensive (antiHTN) class and number used with skeletal muscle function, mobility, sedentary time, and symptoms in older persons. METHODS: Using baseline data from the Lifestyle Interventions and Independence in Elder (LIFE) study (N = 1567, mean age 78.9 years) and multivariable models, we evaluated cross-sectional associations of antiHTN class and number used with physical measures and symptom questionnaires. AntiHTN class included diuretics, angiotensin converting enzyme inhibitors (ACEi), angiotensin receptor blockers (ARB), calcium channel blockers (CCB), and beta blockers (BB). Physical measures included respiratory muscle weakness (maximal inspiratory pressure), grip weakness (dynamometer), impaired lower extremity proximal muscle strength (chair stands), impaired balance (three-stage test), slow gait (400 m walk), mobility impairment (Short Physical Performance Battery), and high sedentary time (accelerometry). Symptoms included dyspnea and fatigue. Covariates included clinical characteristics and non-antiHTNs. RESULTS: Use of any antiHTN was highly prevalent (n = 1248 [79.6%]). In the antiHTN subgroup, each antiHTN class was well represented (ranging 36.6%-62.7%) and included use of three or more antiHTNs (32.0%). In adjusted models, the only statistically significant associations were use of BB and three or more antiHTNs with high sedentary time: odds ratios (95% confidence intervals) 1.44 (1.12, 1.85) and 1.52 (1.04, 2.23), respectively. CONCLUSION: Use of BB and three or more antiHTNs yielded 44% and 52% increased odds of accelerometry-defined high sedentary time, respectively. Notably, high sedentary time is a risk factor for adverse health outcomes. Thus, future work should evaluate whether high sedentary time mitigates benefits or increases risks, regarding antiHTN use in aging populations.

Re-evaluation of the Uplift Clinical Trial Using Age-Appropriate Spirometric Criteria.

BACKGROUND: The clinical trial of tiotropium in COPD, UPLIFT, enrolled adults with a mean age of 65 years and moderate-to-severe airflow obstruction, based on criteria from the Global Initiative for Chronic Obstructive Lung Disease (GOLD). For the UPLIFT cohort, however, GOLD-based criteria are not age-appropriate. RESEARCH QUESTION: Will the use of more age-appropriate criteria for airflow obstruction from the Global Lung Function Initiative (GLI) modify the spirometric classification of the UPLIFT cohort and, in turn, the mortality effect of tiotropium in COPD? STUDY DESIGN AND METHODS: Baseline spirometric classifications were first cross-tabulated by GLI- and GOLD-based criteria. Next, in GLI- and GOLD-based airflow obstruction, modified intention-to-treat analyses evaluated differences in time to death over 4 years, comparing tiotropium vs placebo. Because treatment response may differ by COPD severity, the mortality effect also was evaluated within stratum defined by GLI- and GOLD-based moderate and severe airflow obstruction. RESULTS: Of 5,898 participants with GOLD-based airflow-obstruction, staged as moderate in 2,739 (46.4%) and severe in 3,156 (53.5%), GLI-based criteria established airflow obstruction in 5,750 (97.5%), staged as moderate in 795 (13.5%) and severe in 4,947 (83.9%). Relative to placebo, tiotropium yielded statistically nonsignificant adjusted hazard ratios (adjHRs) (95% CI) for death of 0.91 (0.80-1.04) and 0.91 (0.79-1.03) in GLI- and GOLD-based airflow obstruction, respectively. However, statistically significant effect modification was observed, but only in GLI-based moderate and severe airflow-obstruction, with tiotropium yielding adjHRs for death of 0.53 (0.34-0.81) and 0.99 (0.86-1.13), respectively. The P value for interaction was .007. INTERPRETATION: Mortality reduction by tiotropium was only statistically significant in GLI-based moderate airflow-obstruction, a group that was underrepresented in UPLIFT because of severity misclassification by the original GOLD-based enrollment criteria.

CRcoder: An Interactive Web Application and SAS Macro to Support Personalized Clinical Decisions.

INTRODUCTION: Electronic health care data offer an opportunity to improve clinical decision making through advanced statistical analyses of longitudinal observations. OBJECTIVE: To describe a Web application and SAS/STAT macro (SAS Institute Inc, Cary, NC) for computing joint models to estimate the typical and personalized risk of 2 concurrent binary outcomes. METHODS: Features of the Web application design include uploading longitudinal files formatted with constant or time-varying covariates, specification of 2 binary outcomes, specification of a propensity model for treatment, and joint and separate models of the outcomes. In addition we designed an SAS macro for conducting the analysis. Fitting of joint and separate statistical models was implemented using a model specified in the Web application, with subsequent processing by the SAS macro. To illustrate the fitting of models, a sample of older adults with comorbid hypertension and chronic obstructive pulmonary disease from the Medical Expenditure Panel Survey was created to examine the association between polypharmacy (use of ≥ 5 medication classes) and limitations in social activities and mobility. RESULTS: Relative to separate models, the joint models typically estimated attenuated associations between explanatory variables and the 2 outcomes with smaller standard errors. These joint models yielded estimates of personalized concurrent risk and typical concurrent risk. DISCUSSION: Clinical decision making based on electronic health data can be improved using joint modeling to generate an individual's probability of concurrent risk. CONCLUSION: This user-friendly software performs the advanced statistical analyses needed to estimate typical and personalized concurrent risks.

FEV1 as a Standalone Spirometric Predictor and the Attributable Fraction for Death in Older Persons.

BACKGROUND: Commonly used thresholds for staging FEV1 have not been evaluated as standalone spirometric predictors of death in older persons. Specifically, the proportion of deaths attributed to a reduced FEV1, when staged by commonly used thresholds in L, percent of predicted (% pred), and Z scores, has not been previously reported. METHODS: In 4,232 white persons ≥ 65 y old, sampled from the Cardiovascular Health Study, FEV1 was stratified as stage 1 (FEV1 ≥ 2.00 L, ≥80% pred, and Z score ≥-1.64), stage 2 (FEV1 1.50-1.99 L, 50-79%pred, and Z score -2.55 to -1.63), and stage 3 (FEV1 < 1.50 L, < 50% pred, and Z score < -2.55). Notably, a Z score threshold of -1.64 defines normal-for-age lung function as the lower limit of normal (ie, 5th percentile of distribution), and accounts for differences in age, sex, height, and ethnicity. Next, adjusted odds ratios and average attributable fractions for 10-y all-cause mortality were calculated, comparing FEV1 stages 2 and 3 against stage 1, expressed in L, % pred, and Z scores. The average attributable fraction estimates the proportion of deaths attributed to a predictor by combining the prevalence of the predictor with the relative risk of death conferred by that predictor. RESULTS: FEV1 stage 2 and 3 in L, % pred, and Z scores yielded similar adjusted odds ratios of death: 1.40-1.51 for stage 2 and 2.35-2.66 for stage 3. Conversely, FEV1 stages 2 and 3 in L, % pred, and Z scores differed in prevalence: 12.8-28.6% for stage 2 and 6.4-17.5% for stage 3, and also differed in the adjusted average attributable fraction for death: 3.2-6.4% for stage 2 and 4.5-9.1% for stage 3. CONCLUSIONS: In older persons, the proportion of deaths attributed to a reduced FEV1 is best stratified by Z score staging thresholds because these yield a similar relative risk of death but a more age- and sex-appropriate prevalence of FEV1 stage.

Personalized and typical concurrent risk of limitations in social activity and mobility in older persons with multiple chronic conditions and polypharmacy.

PURPOSE: We define personalized concurrent risk (PCR) as the subject-specific probability of an index outcome within a defined interval of time, while currently at risk for a separate outcome, where the outcomes are not mutually exclusive and can be jointly modeled with a shared random intercept. We further define typical concurrent risk as the risk obtained by setting the random intercept to null. METHODS: Drawing data from the Medical Expenditure Panel Survey (cohorts 2008-2013), we jointly model limitations in social activity and mobility over two years among older community-dwelling persons with both hypertension and chronic obstructive pulmonary disease. The joint model uses inverse probability of treatment weighting based on each participant's baseline propensity of polypharmacy (≥5 classes of medication). RESULTS: Even among participants with the same covariates, older persons with multiple chronic conditions exhibit wide-ranging heterogeneity of the treatment effect from polypharmacy, a risk factor for negative health outcomes among older persons. The magnitude of the PCRs is dominated by the value of the subject-specific random effect. CONCLUSIONS: Estimates of PCR and typical concurrent risk can be calculated from national or institutional data sets and may facilitate the practice of personalized care for older patients with multiple chronic conditions.

Treatment of Hypertension in Complex Older Adults: How Many Medications Are Needed?

Background: Many older adults with hypertension receive multiple antihypertensives. It is unclear whether treatment with several antihypertensive classes results in greater cardiovascular benefits than fewer antihypertensive classes. Objectives: We investigated (a) the longitudinal associations between treatment with ≥ 3 versus 1-2 classes and death and major adverse cardiovascular events (MACE) and (b) whether these associations varied by the presence of mobility disability. Methods: We included 6,011 treated hypertensive adults ≥65 from the Medical Expenditure Panel Survey (MEPS), a nationally representative community sample. Times to MACE and death were compared between those receiving ≥3 versus 1-2 classes using multivariable proportional hazards regression. We used inverse probability of treatment weighting to account for indication and contraindication bias. Results: There were no significant differences in the risk of mortality (hazard ratio [HR] = 0.96, p = .769) or MACE (HR = 1.10, p = .574) between the exposure groups, and there were no significant exposure × mobility disability interactions. Discussion: We found no benefit of ≥3 versus 1-2 antihypertensive classes in reducing mortality and cardiovascular events in a representative cohort of older adults, raising concern about the added benefit of additional antihypertensives in the real world.

Joint modeling of concurrent binary outcomes in a longitudinal observational study using inverse probability of treatment weighting for treatment effect estimation.

PURPOSE: Correlated healthcare utilization outcomes may be encoded as binary outcomes in epidemiologic studies. We demonstrate how to account for correlation between concurrent binary outcomes and confounding by person characteristics when estimating a treatment effect in observational studies. METHODS: We present a joint shared-parameter model, weighted by inverse probability of treatment weights (IPTW) to account for confounding. The model is evaluated in a simulation study that emulates the Medical Expenditure Panel Survey data and compared with a covariate-adjusted joint model and with separate outcome models (IPTW weighted and covariate adjusted). RESULTS: For the IPTW-weighted joint model, relative bias in the estimated treatment effect on outcome 1 ranged from -0.057 to -0.033 and outcome 2 from -0.077 to -0.043. For the covariate-adjusted joint model, relative bias ranged from -0.010 to -0.083 for outcome 1 and from -0.087 to -0.110 for outcome 2. The covariate-adjusted joint model estimated the effect more closely than the covariate-adjusted separate model. The IPTW-weighted joint model estimated the effect more closely for outcome 1. CONCLUSIONS: The IPTW-weighted joint model handles correlation between binary outcomes, adjusts for confounding, and estimates the treatment effect accurately in observational studies. We illustrate the contribution of person-specific effects in estimating personalized risk.