key: cord-0026987-womzvkz7
authors: Nakamura, Julia S.; Hong, Joanna H.; Smith, Jacqui; Chopik, William J.; Chen, Ying; VanderWeele, Tyler J.; Kim, Eric S.
title: Associations Between Satisfaction With Aging and Health and Well-being Outcomes Among Older US Adults
date: 2022-02-09
journal: JAMA Netw Open
DOI: 10.1001/jamanetworkopen.2021.47797
sha: 0c14a3743bf5bfaef057b11663fb69467f5c0de7
doc_id: 26987
cord_uid: womzvkz7

IMPORTANCE: Researchers and policy makers are expanding the focus from risk factors of disease to seek potentially modifiable health factors that enhance people’s health and well-being. Understanding if and to what degree aging satisfaction (one’s beliefs about their own aging) is associated with a range of health and well-being outcomes aligns with the interests of older adults, researchers, health systems, and politicians. OBJECTIVES: To evaluate associations between changes in aging satisfaction and 35 subsequent health and well-being outcomes. DESIGN, SETTING, AND PARTICIPANTS: This cohort study used data from the Health and Retirement Study, a national, diverse, and longitudinal sample of 13 752 US adults older than 50 years, to evaluate if changes in aging satisfaction (between combined cohorts from 2008 and 2010 and 4 years later, in 2012 and 2014) were subsequently associated with 35 indicators of physical, behavioral, and psychosocial health and well-being in 2016 and 2018. Statistical analysis was conducted from July 24, 2020, to November 6, 2021. EXPOSURE: Aging satisfaction. MAIN OUTCOMES AND MEASURES: A total of 35 physical (eg, stroke), behavioral (eg, sleep problems), and psychosocial (eg, depression) outcomes were evaluated using multiple linear and generalized linear regression models. Data from all participants, irrespective of how their levels of aging satisfaction changed from the prebaseline to baseline waves, were incorporated into the overall estimate, which was conditional on prior satisfaction. RESULTS: During the 4-year follow-up period, participants (N = 13 752; 8120 women [59%]; mean [SD] age, 65 [10] years; median age, 64 years [IQR, 56-72 years]; 7507 of 11 824 married [64%]) in the highest (vs lowest) quartile of aging satisfaction had improved physical health (eg, 43% reduced risk of mortality [risk ratio, 0.57; 95% CI, 0.46-0.71]), better health behaviors (eg, 23% increased likelihood of frequent physical activity [risk ratio, 1.23; 95% CI, 1.12-1.34]), and improved psychosocial well-being (eg, higher positive affect [β = 0.51; 95% CI, 0.44-0.58] and lower loneliness [β = −0.41; 95% CI, −0.48 to −0.33]), conditional on prebaseline aging satisfaction. CONCLUSIONS AND RELEVANCE: This study suggests that higher aging satisfaction is associated with improved subsequent health and well-being. These findings highlight potential outcomes if scalable aging satisfaction interventions were developed and deployed at scale; they also inform the efforts of policy makers and interventionists who aim to enhance specific health and well-being outcomes. Aging satisfaction may be an important target for future interventions aiming to improve later-life health and well-being.

Frequency of Contact with: Children, Other Family, and Friends. Frequency of contact was measured as the frequency with which participants were in contact with their children, other family, or friends (separately). Participants were asked, "On average, how often do you do each of the following?" 1) "Meet up (include both arranged and chance meetings)," 2) "Speak on the phone," 3) "Write or email," and had the choice of the following 6 responses: 1) ≥3x/week, 2) 1x-2x/week, 3) 1-2x/month, 4) every few months, 5) 1-2x/year, 6) <1x/year or never. 26 The highest value on any of the three modes of contact was taken for each relationship type since contact (regardless of the mode of contact) was the main point of interest. For example, if the respondent did not speak on the phone very often with a given person but met them in person very often, contact was operationalized as being common. Two categories of contact were created: 1) frequent contact: ≥1x/week contact (the reference group) and 2) infrequent contact: <1x/week of contact.

Personality. Personality was assessed with 26 items derived from the Midlife Development Inventory Personality scales (MIDI) and International Personality Item Pool (IPIP): the "Big-5" personality traits (openness to experience, conscientiousness, extraversion, agreeableness, and neuroticism), 27 The goal of MIDI was to create the shortest possible measure that assessed the Big-5 personality traits with high validity and reliability using existing trait inventories. In a pilot study with a probability sample of 1,000 adults aged 30-70, the items with the highest item-tototal correlations and factor loadings were selected for the MIDI. Next, forward regressions were computed to determine the smallest number of items needed to account for more than 90 percent of the total scale variance. For example, items on the conscientiousness scale included "organized," "responsible," "hardworking," and "careless." Response categories ranged from 1 (a lot) to 4 (not at all). Responses were reverse scored so that a higher score indicated higher indication of a given personality trait. All items were averaged to obtain a composite score for each personality trait. 

where the first equality follows by the no-confounding assumption, the second by consistency, and the third by the statistical model.

We considered the idea of creating aggregate measures that combined the incidence of a condition and death due to that condition. However, out of the 14 ways HRS categorizes causes of death, very few categories cleanly mapped onto health conditions we evaluated in this study without a large risk of misclassification error. Thus, we did not pursue this option. Causes of death included deaths due to: 1) Musculoskeletal system and connective tissue; 2) Heart, circulatory and blood conditions; 3) Allergies; hay fever; sinusitis; tonsillitis; 4) Endocrine, metabolic and nutritional conditions; 5) Digestive system (stomach, liver, gallbladder, kidney, bladder); 6) Neurological and sensory conditions; 7) Reproductive system and prostate conditions; 8) Emotional and psychological conditions; 9) Miscellaneous; 10) Other symptoms; 11) Not a health condition; 12) None; 13) Other health condition; 14) Cancers and tumors; skin conditions. a If the reference value is "1," the effect estimate is OR or RR; if the reference value is "0," the effect estimate is β. c An outcome-wide analytic approach was used, and a separate model for each outcome was run. A different type of model was run depending on the nature of the outcome: 1) for each binary outcome with a prevalence of ≥10%, a generalized linear model (with a log link and Poisson distribution) was used to estimate a RR; 2) for each binary outcome with a prevalence of <10%, a logistic regression model was used to estimate an OR; and 3) for each continuous outcome, a linear regression model was used to estimate a β. c All continuous outcomes were standardized (mean=0; standard deviation=1), and β was the standardized effect size. d The analytic sample was restricted to those who had participated in the baseline wave (t1;2012 or 2014). Multiple imputation was performed to impute missing data on the exposure, covariates, and outcomes. All models adjusted for sociodemographic characteristics (age, sex, race/ethnicity, marital status, annual household income, total wealth, level of education). These variables were adjusted for in the pre-baseline wave (t0;2008 or 2010). e The analytic sample was restricted to those who had participated in the baseline wave (t1;2012 or 2014). Multiple imputation was performed to impute missing data on the exposure, covariates, and outcomes. All models adjusted for sociodemographic characteristics (age, sex, race/ethnicity, marital status, annual household income, total wealth, level of education, employment status, health insurance, geographic region), pre-baseline childhood abuse, pre-baseline religious service attendance, pre-baseline values of the outcome variables (diabetes, hypertension, stroke, cancer, heart disease, lung disease, arthritis, overweight/obesity, physical functioning limitations, cognitive impairment, chronic pain, self-rated health, heavy drinking, current smoking status, physical activity, sleep problems, positive affect, life satisfaction, optimism, purpose in life, mastery, health mastery, financial mastery, depressive symptoms, hopelessness, negative affect, perceived constraints, loneliness, living with spouse/partner, contact children Note. Abbreviations: CI, confidence interval; OR, odds ratio; RR, risk ratio. a If the reference value is "1," the effect estimate is OR or RR; if the reference value is "0," the effect estimate is β. b The analytic sample was restricted to those who had participated in the baseline wave (t1;2012 or 2014). Multiple imputation was performed to impute missing data on the exposure, covariates, and outcomes. All models adjusted for sociodemographic characteristics (age, sex, race/ethnicity, marital status, annual household income, total wealth, level of education, employment status, health insurance, geographic region), pre-baseline childhood abuse, pre-baseline religious service attendance, pre-baseline values of the outcome variables (diabetes, hypertension, stroke, cancer, heart disease, lung disease, arthritis, overweight/obesity, physical functioning limitations, cognitive impairment, chronic pain, self-rated health, heavy drinking, current smoking status, physical activity, sleep problems, positive affect, life satisfaction, optimism, purpose in life, mastery, health mastery, financial mastery, depressive symptoms, hopelessness, negative affect, perceived constraints, loneliness, living with spouse/partner, contact children <1x/week, contact other family <1x/week, contact friends <1x/week), personality factors (openness, conscientiousness, extraversion, agreeableness, neuroticism) and the pre-baseline value of the exposure (coded in quartiles). These variables were adjusted for in the wave pre-baseline to the exposure assessment (t0;2008 or 2010). c An outcome-wide analytic approach was used, and a separate model for each outcome was run. A different type of model was run depending on the nature of the outcome: 1) for each binary outcome with a prevalence of ≥10%, a generalized linear model (with a log link and Poisson distribution) was used to estimate a RR; 2) for each binary outcome with a prevalence of <10%, a logistic regression model was used to estimate an OR; and 3) for each continuous outcome, a linear regression model was used to estimate a β. d All continuous outcomes were standardized (mean=0; standard deviation=1), and β was the standardized effect size. *p<0.05 before Bonferroni correction; **p<0.01 before Bonferroni correction; ***p<0.05 after Bonferroni correction (the p-value cutoff for Bonferroni correction is p=0.05/35 outcomes=p<0.001). Note. Abbreviations: CI, confidence interval; OR, odds ratio; RR, risk ratio. a If the reference value is "1," the effect estimate is OR or RR; if the reference value is "0," the effect estimate is β. b The analytic sample was restricted to those who had participated in the baseline wave (t1;2012 or 2014). Multiple imputation was performed to impute missing data on the exposure, covariates, and outcomes. All models adjusted for sociodemographic characteristics (age, sex, race/ethnicity, marital status, annual household income, total wealth, level of education, employment status, health insurance, geographic region), pre-baseline childhood abuse, pre-baseline religious service attendance, pre-baseline values of the outcome variables (diabetes, hypertension, stroke, cancer, heart disease, lung disease, arthritis, overweight/obesity, physical functioning limitations, cognitive impairment, chronic pain, self-rated health, heavy drinking, current smoking status, physical activity, sleep problems, positive affect, life satisfaction, optimism, purpose in life, mastery, health mastery, financial mastery, depressive symptoms, hopelessness, negative affect, perceived constraints, loneliness, living with spouse/partner, contact children <1x/week, contact other family <1x/week, contact friends <1x/week), personality factors (openness, conscientiousness, extraversion, agreeableness, neuroticism) and the pre-baseline value of the exposure (coded in quartiles). These variables were adjusted for in the wave pre-baseline to the exposure assessment (t0;2008 or 2010). c An outcome-wide analytic approach was used, and a separate model for each outcome was run. A different type of model was run depending on the nature of the outcome: 1) for each binary outcome with a prevalence of ≥10%, a generalized linear model (with a log link and Poisson distribution) was used to estimate a RR; 2) for each binary outcome with a prevalence of <10%, a logistic regression model was used to estimate an OR; and 3) for each continuous outcome, a linear regression model was used to estimate a β. d All continuous outcomes were standardized (mean=0; standard deviation=1), and β was the standardized effect size. Note. Abbreviations: CI, confidence interval; OR, odds ratio; RR, risk ratio. a If the reference value is "1," the effect estimate is OR or RR; if the reference value is "0," the effect estimate is β. b The analytic sample was restricted to those who had participated in the baseline wave (t1;2012 or 2014) and had an increase in aging satisfaction between the pre-baseline and baseline wave. Quartile 4 shows the estimates for participants with the largest increases in aging satisfaction from the prebaseline to baseline waves, and other quartiles can be interpreted in the same way. Multiple imputation was performed to impute missing data on the exposure, covariates, and outcomes. All models adjusted for sociodemographic characteristics (age, sex, race/ethnicity, marital status, annual household income, total wealth, level of education, employment status, health insurance, geographic region), pre-baseline childhood abuse, pre-baseline religious service attendance, pre-baseline values of the outcome variables (diabetes, hypertension, stroke, cancer, heart disease, lung disease, arthritis, overweight/obesity, physical functioning limitations, cognitive impairment, chronic pain, self-rated health, heavy drinking, current smoking status, physical activity, sleep problems, positive affect, life satisfaction, optimism, purpose in life, mastery, health mastery, financial mastery, depressive symptoms, hopelessness, negative affect, perceived constraints, loneliness, living with spouse/partner, contact children <1x/week, contact other family <1x/week, contact friends <1x/week), personality factors (openness, conscientiousness, extraversion, agreeableness, neuroticism) and the pre-baseline value of the exposure (coded in quartiles). These variables were adjusted for in the wave pre-baseline to the exposure assessment (t0;2008 or 2010). c An outcome-wide analytic approach was used, and a separate model for each outcome was run. A different type of model was run depending on the nature of the outcome: 1) for each binary outcome with a prevalence of ≥10%, a generalized linear model (with a log link and Poisson distribution) was Note. Abbreviations: CI, confidence interval; OR, odds ratio; RR, risk ratio. a If the reference value is "1," the effect estimate is OR or RR; if the reference value is "0," the effect estimate is β. b The analytic sample was restricted to those who had participated in the baseline wave (t1;2012 or 2014) and had a decrease in aging satisfaction between the pre-baseline and baseline wave. Quartile 4 shows the estimates for participants with the largest decreases in aging satisfaction from the prebaseline to baseline waves, and other quartiles can be interpreted in the same way. Multiple imputation was performed to impute missing data on the exposure, covariates, and outcomes. All models adjusted for sociodemographic characteristics (age, sex, race/ethnicity, marital status, annual household income, total wealth, level of education, employment status, health insurance, geographic region), pre-baseline childhood abuse, pre-baseline religious service attendance, pre-baseline values of the outcome variables (diabetes, hypertension, stroke, cancer, heart disease, lung disease, arthritis, overweight/obesity, physical functioning limitations, cognitive impairment, chronic pain, self-rated health, heavy drinking, current smoking status, physical activity, sleep problems, positive affect, life satisfaction, optimism, purpose in life, mastery, health mastery, financial mastery, depressive symptoms, hopelessness, negative affect, perceived constraints, loneliness, living with spouse/partner, contact children <1x/week, contact other family <1x/week, contact friends <1x/week), personality factors (openness, conscientiousness, extraversion, agreeableness, neuroticism) and the pre-baseline value of the exposure (coded in quartiles). These variables were adjusted for in the wave pre-baseline to the exposure assessment (t0;2008 or 2010). c An outcome-wide analytic approach was used, and a separate model for each outcome was run. A different type of model was run depending on the nature of the outcome: 1) for each binary outcome with a prevalence of ≥10%, a generalized linear model (with a log link and Poisson distribution) was a If the reference value is "1," the effect estimate is OR or RR; if the reference value is "0," the effect estimate is β. b The analytic sample was restricted to those who had participated in the baseline wave (t1;2012 or 2014) and remained within the same aging satisfaction quartile between the pre-baseline and baseline wave. Quartile 4 shows the estimates for participants who remained in the highest quartile of aging satisfaction from the pre-baseline to baseline waves, as compared to participants who remained in the lowest quartile (quartile 1) of aging satisfaction from the pre-baseline to baseline waves. Other quartiles can be interpreted in the same way. Multiple imputation was performed to impute missing data on the exposure, covariates, and outcomes. All models adjusted for sociodemographic characteristics (age, sex, race/ethnicity, marital status, annual household income, total wealth, level of education, employment status, health insurance, geographic region), pre-baseline childhood abuse, pre-baseline religious service attendance, pre-baseline values of the outcome variables (diabetes, hypertension, stroke, cancer, heart disease, lung disease, arthritis, overweight/obesity, physical functioning limitations, cognitive impairment, chronic pain, self-rated health, heavy drinking, current smoking status, physical activity, sleep problems, positive affect, life satisfaction, optimism, purpose in life, mastery, health mastery, financial mastery, depressive symptoms, hopelessness, negative affect, perceived constraints, loneliness, living with spouse/partner, contact children <1x/week, contact other family <1x/week, contact friends <1x/week), personality factors (openness, conscientiousness, extraversion, agreeableness, neuroticism) and the pre-baseline value of the exposure (coded in quartiles). These variables were adjusted for in the wave pre-baseline to the exposure assessment (t0;2008 or 2010). c An outcome-wide analytic approach was used, and a separate model for each outcome was run. A different type of model was run depending on the nature of the outcome: 1) for each binary outcome with a prevalence of ≥10%, a generalized linear model (with a log link and Poisson distribution) was used to estimate a RR; 2) for each binary outcome with a prevalence of <10%, a logistic regression model was used to estimate an OR; and 3) for each continuous outcome, a linear regression model was used to estimate a β. 

Validating mortality ascertainment in the Health and Retirement Study

Documentation of chronic disease measures in the Heath and Retirement Study (HRS/AHEAD)

World Health Organization. Physical status: the use and interpretation of anthropometry: report of a WHO expert committee

Health and Retirement Study imputation of cognitive functioning measures:1992 -2014. Health and Retirement Study

Documentation of cognitive functioning measures in the Health and Retirement Study

Assessment of cognition using surveys and neuropsychological assessment: the health and retirement study and the aging, demographics, and memory study

The aging, demographics, and memory study: study design and methods

A Guttman health scale for the aged

An epidemiology of disability among adults in the United States

Studies of illness in the aged: the index of ADL: a standardized measure of biological and psychosocial function

Assessment of older people: self-maintaining and instrumental activities of daily living

Association among socioeconomic status, health behaviors, and all-cause mortality in the United States

A scale for the estimation of sleep problems in clinical research

The PANAS-X: Manual for the positive and negative affect scheduleexpanded form

The satisfaction with life scale

Distinguishing optimism from neuroticism (and trait anxiety, self-mastery, and self-esteem): a reevaluation of the Life Orientation Test

The structure of psychological well-being revisited

Quartile 3 (n=1545) Quartile 4 (n=1817)

(0.03, 0.15)** 0.18 (0.12, 0.23)*** Negative