key: cord-0954924-snxjuzta
authors: Yu, Xuexin; Langa, Kenneth M.; Cho, Tsai-Chin; Kobayashi, Lindsay C.
title: Association of Perceived Job Insecurity With Subsequent Memory Function and Decline Among Adults 55 Years or Older in England and the US, 2006 to 2016
date: 2022-04-13
journal: JAMA Netw Open
DOI: 10.1001/jamanetworkopen.2022.7060
sha: 1f1301d0c66e5b6830f039b17087c21e5f1a1c51
doc_id: 954924
cord_uid: snxjuzta

IMPORTANCE: Intensified global economic competition and recent financial crises, including those associated with the COVID-19 pandemic, have contributed to uncertainty about job security. However, little is known about the association of perceived job insecurity with memory function and decline among older adults. OBJECTIVES: To investigate the association between perceived job insecurity and subsequent memory function and rate of memory decline among older adults in the US and England. DESIGN, SETTING, AND PARTICIPANTS: This 10-year prospective population-based cohort study used data from the US Health and Retirement Study (HRS) and the English Longitudinal Study of Ageing (ELSA) collected from 2006 to 2016. Participants included 9538 adults 55 years or older. Data were analyzed from August 1 to 31, 2021. EXPOSURES: Perceived job insecurity (yes vs no) at baseline. MAIN OUTCOMES AND MEASURES: Episodic memory z scores at baseline and rate of decline during the follow-up. RESULTS: Among the 9538 study participants, the mean (SD) age at baseline was 60.97 (6.06) years, and 4981 (52.22%) were women. A total of 2320 participants (24.32%) reported job insecurity at baseline (1088 of 3949 [27.55%] in England and 1232 of 5589 [22.04%] in the US). Perceived job insecurity after 55 years of age was associated with lower baseline memory z scores in the fully adjusted model (β = −0.04 [95% CI, −0.08 to −0.01]) but not with rate of memory decline (β = 0.01 [95% CI, −0.01 to 0.01]). The association appeared to be stronger in the US than in England (job insecurity × US, β = −0.05 [95% CI, −0.11 to 0.02]), but the estimate was imprecise, potentially owing to low statistical power. CONCLUSIONS AND RELEVANCE: The findings of this cohort study suggest that exposure to job insecurity in middle to late life was associated with worse memory function among older adults in the US and England. This association may vary across socioeconomic and social welfare contexts, although future studies with large samples from diverse socioeconomic settings are warranted.

1) Estimating the probability of being observed, in contrast to loss to follow-up (given that the individual survived to the current wave and completed the assessment in the prior wave) Ρr ( ) = (Pr[ = 0 | ( −1) = 0, ( −1) , 0 , ( −1) ]) Eq. (1) Where:

• equals to 0 if the individual is being observed at time j, and equals to 1 if the individual has censored due to loss to follow-up • ( −1) represents memory scores at time j-1

• represents job insecurity status at baseline Cluster and stratification were included to account for complex survey design in HRS and ELSA. Model 1 adjusted for baseline year, baseline age, sex, marital status, race, and foreign-born status Model 2 adjusted for baseline year, baseline age, sex, marital status, race, foreign-born status, occupation, wealth, and education. Model 3 adjusted for baseline year, baseline age, sex, marital status, race, foreign-born status, occupation, wealth, education, alcohol consumption, smoking history, BMI, hypertension, diabetes, stroke, cardiovascular diseases, cancer, and depressive symptoms. IPCWs weights were included to account for potential selective attrition. Model 1 adjusted for baseline year, baseline age, sex, marital status, race, and foreign-born status Model 2 adjusted for baseline year, baseline age, sex, marital status, race, foreign-born status, occupation, wealth, and education. Model 3 adjusted for baseline year, baseline age, sex, marital status, race, foreign-born status, occupation, wealth, education, alcohol consumption, smoking history, BMI, hypertension, diabetes, stroke, cardiovascular diseases, cancer, and depressive symptoms. Model 1 adjusted for baseline year, baseline age, sex, marital status, race, and foreign-born status Model 2 adjusted for baseline year, baseline age, sex, marital status, race, foreign-born status, occupation, wealth, and education. Model 3 adjusted for baseline year, baseline age, sex, marital status, race, foreign-born status, occupation, wealth, education, alcohol consumption, smoking history, BMI, hypertension, diabetes, stroke, cardiovascular diseases, cancer, and depressive symptoms. Model 1 adjusted for baseline year, baseline age, sex, marital status, race, and foreign-born status Model 2 adjusted for baseline year, baseline age, sex, marital status, race, foreign-born status, occupation, wealth, and education. Model 3 adjusted for baseline year, baseline age, sex, marital status, race, foreign-born status, occupation, wealth, education, alcohol consumption, smoking history, BMI, hypertension, diabetes, stroke, cardiovascular diseases, cancer, and depressive symptoms. .001 -0.01 (-0.01 to -0.01) <0.001 Note: Memory scores were z-standardized according to the mean and standard deviation at baseline. Model 1 adjusted for baseline year, baseline age, sex, marital status, race, and foreign-born status Model 2 adjusted for baseline year, baseline age, sex, marital status, race, foreign-born status, occupation, wealth, and education. Model 3 adjusted for baseline year, baseline age, sex, marital status, race, foreign-born status, occupation, wealth, education, alcohol consumption, smoking history, BMI, hypertension, diabetes, stroke, cardiovascular diseases, cancer, and depressive symptoms.

Baseline Characteristics of the Included Study Sample and Excluded Individuals, the US (HRS) and England (ELSA

84) 345 (14.10) 8,218 (28.04) <0.01 c Upper secondary and vocational training 5,007 (52.50) 14,631 (49.92) Tertiary 2,442 (25.60) 4,550 (15.52) Household wealth

07) 4,893 (32.47) <0.01 b Smoking history (Ever vs. never), n (%)

Pearson chi-square test. c Wilcoxon rank-sum tests. d Other racial/ethnic groups in the HRS included Black/African American, American Indian or Alaskan Native, Asian or Pacific Islander, Hispanic/Latino, combination of Black and American Indian, and other; other racial/ethnic groups in the ELSA included Mixed ethnic group