M. Iskandar1, J. Martindale1,2, J.P.W. Bynum1,2, M.A. Davis2,3,4
1. University of Michigan Medical School, Division of Geriatric and Palliative Medicine. Ann Arbor, Michigan, United States; 2. Institute for Healthcare Policy and Innovation, University of Michigan Ann Arbor, Michigan, United States; 3. University of Michigan School of Nursing, Department of Systems, Populations, and Leadership. Ann Arbor, Michigan, United States; 4. University of Michigan Medical School, Department of Learning Health Sciences. Ann Arbor, Michigan, United States
Corresponding Author: Matthew A. Davis, MPH, PhD, University of Michigan, 400 North Ingalls, Ann Arbor, MI 48109-5482, Email: mattadav@umich.edu, Telephone: (734) 764-2814
J Prev Alz Dis 2024;
Published online May 29, 2024, http://dx.doi.org/10.14283/jpad.2024.97
Abstract
Cognitive resilience has emerged as a mechanism that may help explain individual differences in cognitive function associated with aging and/or pathology. It is unknown whether an association exists between family income level and cognitive resilience. We performed a cross-sectional study to estimate the relationship between family income level and high cognitive resilience using the National Health and Nutrition Examination Survey (NHANES) among older adults (age≥60). Logistic regression was used to estimate the association between income level and high cognitive resilience adjusted for other factors. Accounting for differences in education, occupation, and health status, older adults in the highest income category were twice as likely compared to those with very low income to have high cognitive resilience (OR: 1.90, 95% CI: 1.05,3.43). A doseresponse was apparent between income category and high cognitive resilience. The finding that income, above and beyond that of known factors, affects cognitive function is important for future public health strategies that aim to prevent or delay cognitive impairment.
Key words: Dementia, Alzheimer’s disease, cognitive functioning, family income level, education, occupation.
Introduction
The number of older Americans with dementia is expected to increase from the current estimate of 6.7 million to over 13 million by 2050 (1). Among older Americans, a decline in cognitive functioning can be associated with profound changes in well-being and affect a person’s ability to live and function independently. Further, cognitive impairment is associated with substantial financial costs to families, caregivers, and the healthcare system at large (2). The development of new strategies to prevent dementia and promote healthy cognitive aging is now a national public health priority (3).
Cognitive resilience has emerged as a mechanism that may help explain differences in cognitive function associated with aging and/or pathology – it is believed that higher cognitive resilience may delay the development of cognitive symptoms (4). There is interest in understanding what factors influence cognitive resilience throughout the life course. Previous research has shown that educational attainment, a cognitively demanding occupation, healthy behaviors, and leisure activities (reading, writing, hobbies, board games) are all associated with higher cognitive resilience (5, 6). However, one factor potentially related to cognitive resilience that has not been explored is financial income level. Low-income Americans may not have access to healthy lifestyle activities and cognitively demanding leisure activities (7). Also, particularly for those at or below the federal poverty level, financial stress could offset gains in cognitive resilience from other factors such as education. Therefore, among older Americans we examined the association between family household income and cognitive resilience defined as high cognitive functioning independent of age and sex.
Methods
We performed a cross-sectional study to examine family income level and cognitive resilience using data from the National Health and Nutrition Examination Survey (NHANES) among older adults. The NHANES is a cross-sectional, multistage probability interview survey of the civilian noninstitutionalized US population that provides nationally representative estimates; further, it is the premier source of health information used by the Centers for Disease Control and Prevention (8). In NHANES surveys 2011-12 and 2013-14 an in-person cognitive assessment was administered to older participants. To increase sample size for sub-analyses, we combined (i.e., appended) the 2011-12 and 2013-14 NHANES surveys, and therefore our analyses represent a single year.
Older adult NHANES respondents (age ≥ 60 years) who did not have evidence of dementia nor need the assistance of a proxy were eligible to participate in cognitive functioning assessment in the NHANES. Cognitive assessments included word learning and recall (Consortium to Establish a Registry for Alzheimer’s Disease); animal fluency test; and digit symbol substitution test (Wechsler Adult Intelligence Scale, Nine Number-Symbol Pairs, and Sketch to Symbols to Numbers). Performance for each measure was converted into a standard normal deviate (Z-score) and averaged across the three assessments to account for the few missing one of the assessments. We estimated cognitive resilience (i.e., cognitive function independent of age and sex) by using linear regression to predict global cognitive score regressed on age (continuous) and sex (9). Residuals from the linear regression model were used to identify high cognitive resilience defined as those being one standard deviation above the mean. Family income level was estimated based on gross family income as a ratio to the federal poverty level (FPL) for respective survey years. Family income level was collapsed into income categories including Very Low (< 100% FPL), Low (100% to 199% FPL), Middle (200% to 399% FPL), and High (≥ 400% FPL) (10). For each study participant, we collected sociodemographic covariates (race/ethnicity, marital status, education, and occupation) as well as health status and health behaviors (self-reported overall health, diet quality, smoking history, drinking history, and select health conditions). Because education and the cognitive demand of an individual’s occupation are among the strongest predictors of cognitive resilience (5, 6) in unadjusted analyses we examined the percent of older adults across income category separately by education and cognitive demand of occupation. High education was defined as above high school education and low educational attainment was defined as high school or less education. High occupational attainment was identified based on industry and occupational classes recorded by NHANES with higher cognitive demands (Table S1) (11). We compared the characteristics of older adults with high cognitive resilience to those not high using a chi-squared test for proportions and t-test to compare means (Table S2).
Logistic regression was used to estimate the odds ratio for the association between family income category and high cognitive resilience and adjust for covariates. Income category was included as indicator variables (using Very Low as the reference group) and we also examined including income category as a continuous variable in the model to evaluate the statistical significance of a test of trend. While our operational definition of high cognitive resilience was based on an individual having an overall global score (regressed on age and sex) above one standard deviation that represents the top 16% of the distribution based on the normal distribution, we also used linear regression to estimate the association between income category and cognitive resilience as a continuous scale (Table S3). For all analyses, we used complex survey design methods to generate national estimates. Results were based on complete case analysis and missing data were assumed to be missing completely at random. The University of Michigan Health Sciences and Behavioral Sciences Institutional Review Board determined that the study was exempt from review.
Results
Among the estimated 42.8 million older Americans 60 years or older, we estimate 27.3% (11.7 million) have high cognitive resilience. Older adults with high cognitive resilience versus those without high cognitive resilience differed in several ways. Those with high cognitive resilience were more likely to be of Non-Hispanic White race (94.3% versus 76.3%), have higher educational level (88.8% had at least come college versus only 56.3% among those Not High), and were overall healthier as measured by self-reported health status, diet quality, and substance use, p-value < 0.01 for all (Table S2).
Family income category differed by high cognitive resilience status (e.g., 56.0% of high cognitive resilience individuals were of the top income category versus only 33.8% among those without high cognitive resilience, p-value < 0.001). When examined separately by level of education and cognitive demand of occupation, in general, the association between higher income category and high cognitive resilience persisted (Figure 1). The only exception was among older adults with a High School or Less education level where a dose-response was not observed. In adjusted analyses, accounting for sociodemographic differences, education, and health status, being in the top income category was associated with nearly twice the odds of high cognitive resilience compared to those with very low income (OR: 1.90, 95% CI: 1.05, 3.43) (Figure 2). A dose-response was apparent when income category was included in the model as a continuous measure (OR for trend 1.17, 95% CI: 1.02, 1.36). Education had the largest association with high cognitive resilience (OR 4.00, 95% CI: 2.76, 5.78). Self-reported fair or poor health status was associated with a lower likelihood of high cognitive resilience (OR 0.41, 95% CI: 0.26, 0.64). Relative to Non-Hispanic White older adults, other racial groups had lower odds of high cognitive resilience (ORs ranged from 0.34 to 0.22).
Figure 1. Percent of Older Adults with High Cognitive Resilience According to Family Income Level and By Level of Education (A) and Cognitive Demand of Occupation (B)
Abbreviations: FPL, federal poverty level. All analyses weighted to represent the US population. Values represent p-values based on χ-squared test. * Very Low (<100% FPL), Low (100% to 199% FPL), Middle (200% to 399% FPL), High (≥400% FPL). † Based on industry and occupation group codes (deductive reasoning, inductive reasoning, information ordering, perceptual speed, etc)
Figure 2. Odds Ratios for the Association between Family Income Level and High Cognitive Resilience
Abbreviations: CI, confidence interval; FPL, federal poverty level. All analyses weighted to represent the US population. * Very Low (<100% FPL), Low (100% to 199% FPL), Middle (200% to 399% FPL), High (≥400% FPL). † Based on industry and occupation group codes (deductive reasoning, inductive reasoning, information ordering, perceptual speed, etc). § Adjusted for all other factors in table
Discussion
We found that after accounting for differences in education, cognitively demanding occupations, and health status higher family income level is associated with higher cognitive resilience. Older Americans in the highest category of income were nearly twice as likely to have high cognitive resilience. In both unadjusted and adjusted analyses, we found some evidence of a dose-response across income categories.
Although our study is not able to fully explain the association between higher income category and cognitive resilience, one explanation may simply be that individuals with higher income may be more likely to engage in more cognitively demanding leisure activities. Another potential explanation is that income may affect mental health and wellbeing resulting in lower cognitive resilience over time (12). Regardless of the reason, our findings do suggest a potential disparity in cognitive health based on affluence – despite education and occupation level, low income is associated with low cognitive resilience. Lastly, while the findings of race/ethnicity affecting cognitive resilience may be an artifact of bias in cognitive assessments; the association between race/ethnicity and lower cognitive resilience warrants future investigation (13).
Several limitations of our study must be acknowledged. First, we cannot rule out the potential effects of misclassification of family income level and cognitive resilience. Family income level for older adults may not represent income throughout the lifespan. We used validated cognitive assessments to estimate cognitive resilience after accounting for age and sex; however, there is still a potential for misclassification. While age is the strongest risk factor of cognitive decline (14), our study did not use a measure of pathology to estimate cognitive resilience. Second, our study was a cross-sectional observational study, and therefore reverse causality is possible (i.e., that high cognitive resilience causes higher income); nevertheless, identification of the independent association between income and cognitive resilience is important. Lastly, although our analyses accounted for differences in health status and behaviors, we cannot completely rule out residual confounding variables affecting our results.
Cognitive impairment and/or dementia pose a public health crisis that is expected to increase in the U.S. In the absence of effective treatments for dementia, particularly for Alzheimer’s disease (the primary cause of dementia), public health strategies are needed to promote healthy cognitive aging. It is well-known that higher education improves the ability to delay and mitigate the effects of aging and brain pathology through cognitive resilience. However, our study identified a potential disparity regarding affluence and cognitive resilience. From a public health perspective, our results relate to the expansive issue of income inequality in America and demonstrate that income, above and beyond that of known factors, affects cognitive function among older adults.
Acknowledgments: Dr. Davis affirms that everyone who contributed significantly to the work is listed as a co-author.
Conflict of Interest: All authors were supported by grant P30AG066582 from the National Institute on Aging. Dr. Davis received financial support from Regional Anesthesia & Pain Medicine for consulting statistical review. Mr. Iskandar, Mr. Martindale, and Dr. Bynum have no conflicts to report.
Author Contributions: Study concept and design: All authors. Acquisition of subjects and/or data: Martindale and Davis. Analysis and interpretation of data: All authors. Preparation of manuscript: All authors.
Sponsor’s Role: The funders had no role in the study design, data collection, management, and analysis, nor any participation in the preparation, review, and approval of the manuscript. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Funding Information: All authors were supported by the National Institute on Aging (NIA) at the National Institutes of Health [grant number P30AG066582].
References
1. Rajan KB, Weuve J, Barnes LL, McAninch EA, Wilson RS, Evans DA. Population estimate of people with clinical Alzheimer’s disease and mild cognitive impairment in the United States (2020-2060). Alzheimers Dement. 2021;17(12):1966-1975; doi: 10.1002/alz.12362.
2. Hurd MD, Martorell P, Delavande A, Mullen KJ, Langa KM. Monetary costs of dementia in the United States. N Engl J Med. 2013;368(14):1326-1334; doi: 10.1056/NEJMsa1204629.
3. National Plan to Address Alzheimer’s Disease. ASPE. https://aspe.hhs.gov/collaborations-committees-advisory-groups/napa/napa-documents/napa-national-plan
4. Stern Y. Cognitive reserve in ageing and Alzheimer’s disease. Lancet Neurol. 2012;11(11):1006-1012; doi: 10.1016/S1474-4422(12)70191-6.
5. Kesse-Guyot E, Andreeva VA, Lassale C, Hercberg S, Galan P. Clustering of midlife lifestyle behaviors and subsequent cognitive function: a longitudinal study. Am J Public Health. 2014;104(11):e170-e177; doi: 10.2105/AJPH.2014.302121.
6. Scarmeas N, Levy G, Tang MX, Manly J, Stern Y. Influence of leisure activity on the incidence of Alzheimer’s disease. Neurology. 2001;57(12):2236-2242; doi: 10.1212/wnl.57.12.2236.
7. Khullar D, Chokshi DA. Health, income, & poverty: Where we are & what could help. Health Affairs. 2018;10(10.1377); Doi: 10.1377/hpb20180817.901935.
8. National Center for Health Statistics. National Health and Nutrition Examination Survey (NHANES). Centers for Disease Control and Prevention. January 31, 2024. Accessed March 14, 2024. https://www.cdc.gov/nchs/nhanes/index.htm
9. Arenaza-Urquijo EM, Vemuri P. Resistance vs resilience to Alzheimer disease: Clarifying terminology for preclinical studies. Neurology. 2018;90(15):695-703. doi:10.1212/WNL.0000000000005303
10. KFF State Health Facts. Distribution of the Nonelderly Uninsured by Federal Poverty Level (FPL). Accessed March 14, 2024. https://www.kff.org/statedata/
11. Azeez M, Gambatese J, Liu D. Comparison of job cognitive and physical demands in different industries for implications on safety outcomes. InProc., Annual Conf.—Canadian Society for Civil Engineering. Montreal: Canadian Society for Civil Engineering 2019 Jun 12
12. Thomson RM, Igelström E, Purba AK, et al. How do income changes impact on mental health and wellbeing for working-age adults? A systematic review and meta-analysis. Lancet Public Health. 2022;7(6):e515-e528. doi:10.1016/S2468-2667(22)00058-5
13. Barnes LL. Alzheimer disease in African American individuals: increased incidence or not enough data?. Nat Rev Neurol. 2022;18(1):56-62. doi:10.1038/s41582-021-00589-3
14. Livingston G, Huntley J, Sommerlad A, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020;396(10248):413-446; doi: 10.1016/S0140-6736(17)31363-6.
© Serdi 2024