Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Oct 1.
Published in final edited form as: Neuropsychology. 2019 Jun 17;33(7):975–985. doi: 10.1037/neu0000565

Financial and Health Literacy Discrepancies with Cognition in Older adults

Gali H Weissberger a, S Duke Han a,e,f,g, Lei Yu b,d, Lisa L Barnes b,c,d, David A Bennett b,d, Patricia A Boyle b,c
PMCID: PMC6763386  NIHMSID: NIHMS1038224  PMID: 31204814

Abstract

Objective:

Greater financial and health literacy are associated with better cognition; however, research suggests that some individuals exhibit differences, or discrepancies, in these abilities in old age. We investigated discrepancies between literacy and cognition and factors associated with such discrepancies in older adults without dementia.

Method:

Participants (N=714; age: M=81.4; education: M=15.4; 75.4% female; 5.2% non-white) from the Rush Memory and Aging Project completed cognitive assessments and a financial and health literacy measure that yielded a total literacy score. Participants were characterized into three groups: 1) total literacy scores that are more than one standard deviation (1SD) above cognition (L>C); 2) total literacy scores falling more than 1SD below cognition (L<C); and 3) total literacy within 1SD of cognition (L=C). Logistic regressions were employed to investigate associations between demographic and psychosocial variables and discrepancy group status.

Results:

Of the 714 participants, 24% showed significant discrepancies. In fully adjusted models, in reference to the L=C group, male sex was associated with greater odds of being in the L>C group (OR=2.32, 95% CI=1.33–4.03, p=0.003) and lower odds of being in the L<C group (OR=0.31, 95% CI=0.14–0.66, p=0.002), higher income was associated with lower odds of being in either discrepancy group (L<C OR=0.87, 95% CI=0.79–0.96, p=0.004; L>C OR=0.86, 95% CI=0.76–0.96, p=0.007), and higher trust was associated with lower odds of being in the L>C group (OR=0.92, 95% CI=0.85–0.99, p=0.030).

Conclusions:

Findings support literacy and cognition as partially dissociable constructs and highlight important factors associated with discrepancies between literacy and cognition.

Keywords: literacy, cognition, older adults, discrepancy

Introduction

The population of adults over age 65 in the United States is growing, accounting for nearly 15% of the total population in the year 2015 and estimated to grow to 98.2 million by the year 2060 (US Census, 2017). Understanding factors that contribute to the wellbeing and health of older adults is therefore imperative. One factor that has been linked to health and wellbeing across the lifespan is literacy (Kirsch, 1993; Kutner, Greenburg, Jin, & Paulsen, 2006), a complex and multi-dimensional construct that reflects the ability to use and manipulate written and printed materials across a variety of contexts (UNESCO Education Sector, 2004). Financial and health literacy are two domains of literacy that reflect the ability to access, understand, and utilize information that promote positive financial and health outcomes (Braunstein & Welch, 2002; Institute of Medicine, 2004). Such abilities are particularly important in old age, when individuals are faced with increasingly complex medical issues and financial decisions. Specifically, financial and health literacy are associated with better cognition (J. S. Bennett, Boyle, James, & Bennett, 2012; Federman, Sano, Wolf, Siu, & Halm, 2009; Wilson, Yu, James, Bennett, & Boyle, 2017), lower risk of developing mild cognitive impairment (Wilson et al., 2017), more frequent participation in health promoting behaviors (J. S. Bennett et al., 2012), better physical and mental health status (J. S. Bennett et al., 2012; Taylor, Jenkins, & Sacker, 2009), and better decision making (James, Boyle, Bennett, & Bennett, 2012) in older adults. Conversely, lower health literacy is associated with poorer health outcomes and increased rates of mortality (Baker et al., 2007; Bostock & Steptoe, 2012), and lower financial literacy is associated with increased susceptibility to scams (James, Boyle, & Bennett, 2014). Particularly disconcerting is the fact that many studies report lower health literacy (Baker, Gazmararian, Sudano, & Patterson, 2000; Kirsch, 1993; Kutner et al., 2006) and financial literacy (Agarwal, Driscoll, Gabaix, & Laibson, 2009; Kirsch, 1993; Lusardi & Mitchell, 2014) in older relative to younger age groups. Thus, understanding factors that contribute to low financial and health literacy in older adults is of significant public health importance.

One factor important for literacy is cognition, which declines with age and may partially account for findings of lower health and financial literacy in older adults relative to young adults. Decline in cognitive functioning impacts some abilities necessary for financial and health literacy such as updating and retaining new knowledge, interpreting complex written materials and situations, and performing computations (Boyle, Yu, et al., 2013b). In fact, research investigating the relationship between literacy and cognition in old age has revealed relatively strong associations between the two (J. S. Bennett et al., 2012; Boyle, Yu, et al., 2013b; Federman et al., 2009; Wilson et al., 2017), supporting a role for cognition in maintaining financial and health literacy.

Although it is well established that cognition is important for some aspects of literacy, other factors contribute to financial and health literacy independent of cognition, and these may lead to a discrepancy between the two functions. For example, long-standing factors, such as one’s early exposure to educational resources and specific life experiences, contribute to literacy (Banks, 2010; Boyle, Yu, et al., 2013; Lusardi, Mitchell, & Curto, 2009). Specific to health literacy, the WHO Commission on Social Determinants of Health (Marmot et al., 2008) identified social factors such as income, early life experiences, employment, education, and sex equality, that impact health outcomes later in life (i.e., social determinants of health; Marmot et al., 2008). One factor that may mediate the relationship between social factors and late-life health outcomes is health literacy, as early life factors may impact one’s ability to acquire health literacy (Braveman, Egerter, & Williams, 2011; Marmot et al., 2008). Additionally, affective states such as depression (Smith, 2013) and personality characteristics (Noon & Fogarty, 2007) may alter one’s motivation to maintain and update literacy throughout the lifespan (Mandell & Klein, 2007; Noon & Fogarty, 2007). In prior work (Boyle, Yu, et al., 2013), we proposed and supported a model of literacy in which diverse individual resources contribute to literacy. This model suggests that, in addition to cognition, other factors such as age and early life experiences (i.e., education and word knowledge) have independent effects on literacy (Boyle, Yu, et al., 2013b). Specifically, path analytic findings showed that age has both direct effects on literacy and indirect effects on literacy via declines in episodic memory and executive functioning, and word knowledge and education have independent effects on literacy. Such findings suggest that literacy and cognition are somewhat dissociable constructs and other individual factors including experiential factors contribute to literacy. In the present study, we sought to expand upon this model to include other demographic and psychosocial factors that might contribute to literacy independently of cognition and that might account for the discrepancy between literacy scores and cognitive scores.

Recent work suggests that assessing discrepancies between two functions may be more sensitive to detecting at-risk subgroups than observing differences in performance on different measures of function (Delis et al., 2007; Delis et al., 1992; Fine et al., 2008; S Duke Han, Boyle, James, et al., 2016; Jacobson, Delis, Bondi, & Salmon, 2002; Massman et al., 1993; Strite, Massman, Cooke, & Doody, 1997). For example, a study by Jacobson et al. (2002) found that while elderly controls and preclinical AD patients did not differ in their performances on two neuropsychological measures, discrepancies in performance between the two measures were larger in the preclinical AD group. Findings suggest that use of Cognitive and Literacy Discrepancies 8 difference scores may be more sensitive than analyses of single test means (Jacobson et al., 2002). Thus, the aims of the present study were to (1) determine the number of individuals exhibiting discrepancies between cognition and financial and health literacy in a large non-demented cohort of older adults, and (2) explore demographic and psychosocial factors associated with such discrepancies. To this end, we adopted an established z-transformation discrepancy approach (Delis et al., 2007; Fine et al., 2008; Han, Boyle, James, et al., 2016; Jacobson et al., 2002) in order to identify individuals with total literacy scores that were either better or worse than composite global cognitive scores. We then examined associations of literacy-cognition discrepancy group status with demographic factors (age, education, sex, race, income) and psychosocial factors (depressive symptoms, social isolation, impulsiveness, and trust) using logistic regression.

Based on prior research identifying strong relationships between literacy and cognition (J. S. Bennett et al., 2012; Boyle, Yu, et al., 2013b; Federman et al., 2009; Wilson et al., 2017), we hypothesized that the majority of individuals in our sample would exhibit comparable performances on measures of literacy and cognition (i.e., no discrepancy). However, we predicted that a significant subset would exhibit discrepancies between the two, particularly given evidence that certain contextual factors, such as early life experiences, contribute to literacy (Banks, 2010; Boyle, Yu, et al., 2013b; Lusardi et al., 2009). To this end, we further hypothesized that certain demographic and psychosocial factors would predict group membership as such factors may modify one’s ability or motivation to develop, maintain, and update literacy independently of cognition. Specifically, we hypothesized that older age and lower education would be associated with discrepancy group given recent findings of an effect independent of cognition of these two factors on literacy scores (Boyle, Yu, et al., 2013b; also see Lusardi & Mitchell, 2011). We additionally hypothesized that individuals from lower income brackets and being female would be associated with discrepancy given disparities in access to, and quality of, literacy resources among older persons (Kutner et al., 2006; Lusardi & Mitchell, 2011). Finally, we hypothesized that increased depressive symptoms, social isolation, impulsiveness, and less trust would relate to discrepancy status given that such factors are likely to affect literacy independently of cognition potentially by influencing one’s motivation to maintain literacy (Mandell & Klein, 2007).

Methods

Participants included 714 non-demented older adults enrolled in the Rush Memory and Aging Project, a cohort study of aging and dementia that began in 1997 (D. A. Bennett et al., 2012a). Participants are recruited from local residential facilities, including retirement homes and senior housing facilities, and community organizations in the Chicago metropolitan area. Cognitive impairment is determined by a clinical neuropsychologist with expertise in aging and AD, who reviews data from a detailed battery of cognitive tests, participant background information, and results of a clinical evaluation are done by a physician with expertise in aging and dementia (Bennett et al., 2006). Dementia is diagnosed in accordance with NINCDS/ADRDA criteria (McKhann et al., 1984).

In 2010, a study of decision making was introduced into the Rush Memory and Aging Project and this included a literacy measure. At the time of these analyses, 1147 subjects were alive and eligible to complete the decision making sub-study. Of these, 1060 completed the literacy measure of the decision making sub-study, 27 refused, and 60 had incomplete literacy data. Of the 1060 participants that completed the literacy measure of the decision making sub-study, 55 were deemed to have dementia and 291 did not have complete data on the psychosocial measures of interest, leaving a total of 714 subjects without dementia eligible for these analyses.

Assessment of Cognition

Participants underwent a comprehensive neuropsychological assessment. Nineteen of 21 total measures were used to make a composite score of cognition (D. A. Bennett et al., 2012a). These measures included word list memory (total number of words immediately recalled after each of the three learning trials), word list recall (total words recalled after a delay), and word list recognition from the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) battery, immediate and delayed recall conditions of Logical Memory Story A and the East Boston Story, verbal fluency (animals, fruits/vegetables), the Boston Naming Test, the National Adult Reading Test, forward and backward conditions of the Wechsler Memory Scale-revised Digit Span subtest, Digit Ordering, the Symbol Digit Modalities test, Number Comparison, the Judgment of Line Orientation test, Standard Progressive Matrices, and Stroop color naming and word reading conditions. The Complex Ideational Material test and the Mini Mental Status Examination (MMSE) were used for descriptive and clinical diagnostic purposes only and were not included in the composite score. To create the composite, raw scores from the 19 cognitive measures were converted to z scores using the baseline mean and standard deviation of all subjects enrolled in the Memory and Aging Project. Each participant’s standardized z scores were then averaged to yield a composite global cognition score. The 19 measures used in the global composite score were chosen specifically for their robustness in capturing age-related changes in cognition among older adults. Additionally, combining measures into a composite score minimizes ceiling and floor effects and other sources of measurement error inherent in the measures (Wilson et al., 2002a; 2002b). We required 95% of tests to be valid Cognitive and Literacy Discrepancies 11 in order to derive the global composite score. Psychometric support for the global cognitive score is contained in prior publications and this measure has been used in many prior studies by our group and others (e.g., D. A. Bennett et al., 2012a; Bennett et al., 2005; Bennett, Wilson, Boyle, Buchman, & Schneider, 2012b; Boyle, Wilson, et al., 2013a; Fleischman, Wilson, Bienias, & Bennett, 2005; Han, Boyle, James, et al., 2016; Schneider et al., 2012; Wilson et al., 2002a; 2002b; Wilson, Barnes, & Bennett, 2003).

All study procedures were conducted in accordance with the ethical rules for human experimentation stated in the Declaration of Helsinki and approved by the institutional review board of Rush University Medical Center. Written informed consent was obtained from all participants prior to study participation.

Assessment of Literacy

The financial and health literacy assessment includes 32 questions that query knowledge of health and financial information, concepts, and numeracy, as described in previous work (D. A. Bennett et al., 2012a; J. S. Bennett et al., 2012; James et al., 2012). Briefly, nine questions are devoted to health literacy, and ask about Medicare, following doctors’ prescription instructions, and perceived drug risk among other health topics. Twenty-three questions are devoted to financial literacy, many of which were adapted from the Health and Retirement Survey (Lusardi & Mitchell, 2007; Lusardi & Mitchelli, 2007). Questions include simple monetary calculations (numeracy) and knowledge of financial concepts such as stocks, bonds, “FDIC”, and interest rates. All response options are either true-false or multiple choice (see James et al., 2012 appendix for full assessment). In accordance with previous studies (J. S. Bennett et al., 2012; Boyle, Yu, et al., 2013b; Han, Boyle, Arfanakis, et al., 2016; Han et al., 2014), percent total correct is calculated for healthcare literacy and financial literacy separately, and the two percentages are averaged to determine a total literacy score for each participant.

Demographic variables

Age is calculated based on birthdate. Sex, race, and education were self-reported. For these analyses, race was converted to a binary variable of White and non-White given the low numbers of non-White older adults. Income level was acquired at the time of participants’ baseline assessment via a single question asking participants to select one of ten levels of total family income ($0-$4,999; $5,000 - $9,999; $10,000 - $14,999; $15,000 - $19,999; $20,000 - $24,999; $25,000 - $29,999; $30,000 - $34,999; $35,000 - $49,999; $50,000 - $74,999; $75,000 and over). Income score ranges from 1–10, with 10 indicating the highest income bracket.

Psychosocial variables

Depressive symptoms were assessed using a 10-item version of the Center for Epidemiologic Studies Depression Scale (CES-D; Kohout, Berkman, Evans, & Cornoni-Huntley, 1993; Radloff, 1977). Participants responded ‘yes’ or ‘no’ as to whether they experienced each of ten depressive symptoms in the past week. The score reflects the total number of symptoms endorsed, and ranges from 0–10 with higher scores representing greater depressive symptomatology.

Self-reported ratings of impulsiveness are derived from an 8-item subscale of the neuroticism scale from the NEO Personality Inventory-Revised (Costa & Mac Crae, 1992). Example items include “I rarely overindulge in anything”, “I have trouble resisting my cravings”, “I have difficulty resisting temptation”, “When I am having my favorite foods, I tend to eat too much”, “I seldom give into my impulses.” Items are rated on a five-point Likert scale ranging from ‘strongly disagree’ to ‘strongly agree’. Three items are reverse-scored. The overall score ranges from 0–32 with higher scores indicating more impulsiveness.

Perceived social isolation was assessed using five items from a modified version of the de Jong-Gierveld Loneliness Scale (de Jong-Gierveld, 1987; de Jong-Gierveld & Kamphuls, 1985). Participants are asked to rate items related to feelings of emptiness and loneliness (e.g., “I experience a general sense of emptiness”, “I miss having people around”, “I feel like I don’t have enough friends”) on a 5-point Likert scale ranging from ‘strongly disagree’ to ‘strongly agree’. Ratings are averaged to acquire a total score, which ranges from 1–5, with higher values indicating more loneliness.

Trust ratings were based on 8 self-report questions adapted from the NEO Personality Inventory-Revised (Costa & Mac Crae, 1992) in which higher scores indicate increased levels of trust. Subjects are asked to rate statements on a five-point Likert scale ranging from ‘strongly disagree’ to ‘strongly agree’, with the total score ranging from 0 to 32. Example items include “I tend to be cynical and skeptical of others” and “I believe that most people will take advantage of you if you let them”.

Statistical analyses

Descriptive statistics are provided for all variables included in the analyses across the three discrepancy groups, and groups were compared across all variables of interest using ANOVAs or Kruskal-Wallis tests for non-normally distributed variables, chi-square, and post-hoc Tukey tests. P-values for post-hoc multiple comparisons were adjusted using Tukey’s method for ANOVAs and Dwass-Steel-Critchlow-Fligner’s method for Kruskal-Wallis tests. Using a previously established discrepancy analysis approach (Delis et al., 2007; Fine et al., 2008; Han, Boyle, James, et al., 2016; Jacobson et al., 2002), cognitive composite scores and scores on the literacy measure were z-transformed relative to the analytic sample, and subtracted from each other to yield a discrepancy value between literacy and cognition. Based on discrepancy value, participants were subdivided into one of three groups: (1) literacy scores less than one standard deviation of cognitive abilities (L<C), (2) literacy scores within one standard deviation of cognitive abilities (i.e., roughly equivalent scores; L=C), and (3) literacy scores greater than one standard deviation of cognitive abilities (L>C). To investigate separate associations between discrepancy group status and the nine variables of interest, we ran separate unadjusted logistic regression models with the L=C group as reference for each of the two discrepancy groups. Finally, we repeated the analysis by including all nine demographic and psychosocial variables in the same models. Significance was determined at an alpha-level of 0.05. For the two binary variables of race and sex, females and white individuals were coded as 0, males and non-white individuals were coded as 1 in logistic regression models.

Results

Discrepancy groups and sample characteristics

Of the 714 participants, 173 (24.2%) had a discrepancy between literacy and cognition of more than one standard deviation. Participants who had literacy z-scores that were more than one standard deviation below their cognitive z-scores (L<C) made up 14.3% (n=102) of the total sample; participants who had literacy z-scores that were more than one standard deviation above their cognitive z-scores (L>C) made up 9.9% (n=71) of the total sample. A scatterplot of literacy and cognitive z-scores by discrepancy group is presented in Figure 1.

Figure 1.

Figure 1

Open in a new tab

Global cognition and literacy z-scores by discrepancy group.

Participant demographics and scores on all variables of interest are presented in Table 1. Sample characteristics and scores are provided for the entire sample and for each group separately. Between-group comparisons revealed significant group differences on age (F(2, 711) Cognitive and Literacy Discrepancies 15 = 3.20, p=0.041), education (X2(2) = 6.26, p=.044) income (X2(2) = 18.93, p<0.0001), and trust (X2(2) = 9.99, p=.007). Post-hoc comparisons revealed significantly younger age in the L>C group relative to the L<C group (p<0.05), greater education in the L=C group relative to the L<C group (p=0.04), greater income levels in the L=C group relative to the L<C group (p<.001) and L>C group (p=0.02), and less trust in the L>C group relative to the L<C and L=C groups (both ps<.05). Chi-square test of independence revealed differences between groups in sex breakdown (X2=27.57, p<.0001), but not race (p=0.16). Groups also did not differ with respect to education, number of depressive symptoms, or impulsiveness ratings.

Table 1.

Participant demographics and scores on variables of interest for the whole sample and for each discrepancy group.

Whole Sample (n=714) L<C (n=102) L=C (n=541) L>C (n=71)
M SD M SD M SD M SD
Demographics
Age 81.40 7.41 80.34 8.72 81.36 7.25 83.21 6.27
Education 15.40 3.03 14.88 2.88 15.53 3.01 15.14 3.27
Sex (% female) 75.35% - 92.16% - 74.49% - 57.75% -
Race (% non-white) 5.18% - 9.80% - 4.44% - 4.23% -
Income 7.28 2.36 6.46 2.64 7.50 2.24 6.75 2.48
MMSE 28.35 1.66 28.63 1.43 28.42 1.61 27.39 2.04
Psychosocial
CES-D score 0.86 1.47 0.93 1.82 0.82 1.37 1.03 1.65
Impulsiveness 14.09 3.71 14.24 3.89 14.05 3.67 14.17 3.83
Social Isolation 2.17 0.58 2.14 0.62 2.16 0.57 2.16 0.57
Trust 23.93 3.62 24.03 3.33 24.08 3.58 22.63 4.11
Discrepancy Variables1
Cognitive composite 0.28 0.51 0.54 0.43 0.29 0.49 −0.18 0.49
Literacy score - % correct 69.40% 14.07 56.08% 12.1 70.76% 13.0 78.17% 12.4
Open in a new tab
1

Cognitive composite score and literacy score reflect values prior to the z-transformation done in relation to the analytic sample.

Note: L>C = literacy scores greater than cognitive scores; L=C = literacy scores within 1 standard unit of cognitive scores; L<C = literacy scores less than cognitive scores; M = mean; SD = standard deviation; MMSE = Mini-Mental State Examination; CES-D = Center for Epidemiologic Studies Depression Scale.

Correlates of Discrepancy

Results of logistic regression analyses are presented in Table 2. Associations can be visualized in forest plots (Figure 2). To determine correlates of discrepancy between literacy and cognition, we first examined whether each of the nine variables of interest was significantly associated with either discrepancy group (L>C, L<C) compared to the non-discrepant group (L=C). To this end, we carried out two sets of nine logistic regressions, one per variable per discrepancy group, using L=C as the reference group. Being male (OR=0.25, 95% CI 0.12–0.53, p<0.001) and having higher education (OR=0.93, 95% CI 0.87–1.00, p=0.048) were associated with lower odds of being in the L<C group compared to the non-discrepant, L=C, group. Being male (OR=2.14, 95% CI 1.28–3.56, p=0.004) and older age (OR=1.04, 95% CI 1.00–1.07, p=0.047) were associated with greater odds of being in the L>C group relative to the L=C group. Reporting more trust was associated with lower odds of being in the L>C group compared to the non-discrepant group (OR=0.90, 95% CI 0.84–0.96, p=0.002). Increased perceived social isolation was marginally associated with greater odds of being in the L>C group compared to the L=C group (OR=1.50, 95% CI 0.99–2.26, p=0.053). Higher income was associated with lower odds of being in the L<C discrepant group compared to the non-discrepant group (L<C OR=0.84, 95% CI 0.77–0.91, p<0.0001), and with lower odds of being in the L>C discrepancy group compared to the non-discrepant group (L>C OR=0.88, 95% CI 0.79–0.97, p=0.01).

Table 2.

Association of demographic and psychosocial factors with discrepancies between literacy and cognition.

Unadjusted Models Fully Adjusted Models
L<C L>C L<C L>C
OR 95% wald confidence limits p-value OR 95% wald confidence limits p-value OR 95% wald confidence limits p-value OR 95% wald confidence limits p-value
Demographics
Age 0.98 0.95–1.01 0.201 1.04 1.00–1.07 0.047 0.98 0.95–1.01 0.244 1.03 0.99–1.07 0.134
Education 0.93 0.87–1.00 0.048 0.96 0.88–1.04 0.308 0.98 0.90–1.06 0.586 1.01 0.92–1.10 0.910
Sex 0.25 0.12–0.53 <0.001 2.14 1.28–3.56 0.004 0.31 0.14–0.66 0.002 2.32 1.33–4.04 0.003
Race 1.39 0.94–2.05 0.104 1.24 0.74–2.07 0.412 1.16 0.77–1.74 0.483 1.14 0.66–1.96 0.637
Income 0.84 0.77–0.91 <0.0001 0.88 0.79–0.97 0.010 0.87 0.79–0.96 0.004 0.86 0.76–0.96 0.007
Psychosocial
CES-D score 1.05 0.92–1.21 0.471 1.09 0.94–1.27 0.255 0.99 0.84–1.16 0.890 0.99 0.83–1.18 0.899
Impulsiveness 1.01 0.96–1.07 0.640 1.01 0.94–1.08 0.803 0.99 0.94–1.06 0.832 1.00 0.93–1.07 0.938
Social Isolation 0.93 0.64–1.35 0.705 1.50 0.99–2.26 0.053 0.92 0.60–1.42 0.719 1.17 0.72–1.90 0.528
Trust 1.00 0.94–1.06 0.897 0.90 0.84–0.96 0.002 0.99 0.93–1.06 0.780 0.92 0.85–0.99 0.030
Open in a new tab

Note: L>C = literacy scores greater than cognitive scores; L=C = literacy scores within 1 standard unit of cognitive scores; L<C = literacy scores less than cognitive scores; OR = Odds Ratio; CES-D = Center for Epidemiologic Studies Depression Scale.

Figure 2.

Figure 2

Open in a new tab

Forest plots displaying odds ratios for correlates of L<C discrepancy group (panel A) and L>C discrepancy group (panel B). OR = Odds Ratio; CES-D = Center for Epidemiologic Studies Depression Scale.

We repeated the analysis by including all nine demographic and psychosocial variables in the same models. This revealed that sex, income, and trust remained associated with discrepancy group membership. Specifically, while participants were more likely to be male in the L>C group than the L=C group (OR=2.32, 95% CI 1.33–4.04, p=0.003), males were less likely to be in the L<C group compared to the L=C group (OR=0.31, 95% CI 0.14–0.66, p=0.002). Higher income was associated with lower likelihood of membership in either of the two discrepant groups compared to the L=C group (L<C OR=0.87, 95% CI 0.79–0.96, p=0.004; L>C OR=0.86, 95% CI 0.76–0.96, p=0.007). Finally, higher trust ratings were associated with lower likelihood of L>C group membership compared to L=C membership (OR=0.92, 95% CI 0.85–0.99, p=0.03).

Discussion

This study investigated discrepancies between literacy and cognition, and the factors associated with such discrepancies in a community-based cohort of 714 non-demented older adults. In line with our hypothesis that a subset of individuals would exhibit discrepancies between literacy and cognition, approximately one quarter of participants exhibited significant discrepancies, as defined by literacy scores falling more than one standard unit above or below cognitive scores. Discrepancy status was distributed amongst these individuals, with approximately half exhibiting literacy scores that were worse than their cognitive scores and half exhibiting literacy scores that were better than their cognitive scores. Additionally, certain demographic and psychosocial factors were associated with discrepancy group membership. In fully adjusted logistic regression models, sex, income, and trust were associated with discrepancy group status. With regard to other demographic variables, contrary to our hypothesis, age and education were not associated with discrepancy. With regard to psychosocial variables, trust was associated with discrepancy group status, while perceived social isolation, depressive symptomatology and impulsiveness were not.

As noted, a sizeable subset of non-demented older adults evidenced literacy and cognition discrepancies of greater than one standard unit, and within this subset, approximately half fell into the L>C group and half fell into the L<C group. This suggests that factors beyond cognition contribute to health and financial literacy in older adults, and that literacy and cognition may be partially distinguishable constructs. Literacy itself is a multidimensional construct, reflecting multiple abilities across a variety of contexts (UNESCO Education Sector, 2004). While some components of literacy such as numeracy are associated with cognition (Banks, 2010), other components may be influenced by other demographic or psychosocial factors, thus contributing to a divergence between cognition and literacy in old age. Previous studies by our group also support the notion that literacy and cognition reflect partially different constructs. A path-analytic study by our group found that age had a direct effect on literacy, independent of cognition, suggesting that age affects other behaviors and skills, unrelated to cognition, that influence literacy (Boyle et al., 2013b). Other studies by our group have shown that literacy has positive effects on the health and well-being of non-demented older adults, even after adjusting for cognitive function (J. S. Bennett et al., 2012; James et al., 2012). Taken together, these and the present findings support the notion that literacy and cognition are partially dissociable constructs differentially affected by various psychosocial and demographic factors.

One demographic factor that related to discrepancy between literacy and cognition was income, which unexpectedly predicted literacy-cognition discrepancies in both directions (L>C and L<C). Individuals from lower income brackets are likely to have less access to literacy resources (Lusardi & Mitchell, 2014), thus increasing the likelihood that their literacy scores fall below cognitive scores (L<C). Consistent with this, studies report less use of financial advisers by poorer individuals (Hackethal, Haliassos, & Jappel, 2010). Another possibility, however, is that poor financial and health literacy may limit household income potential, despite a higher level of cognitive functioning. Here, we found that low income predicting membership in the L>C group. One possibility is that there may be a subset of individuals with low income who are motivated to develop advanced literacy for financial and healthcare matters on their own, despite inadequate access to resources. This is consistent with a study that found that motivation significantly explained individual differences in acquiring financial literacy among young adults (Mandell & Klein, 2007). More research is needed to explore these possibilities and understand the bidirectionality of the relationship observed here.

Sex also predicted discrepancy with literacy and cognition. While being female was associated with greater odds of lower literacy relative to cognition, the opposite was observed for males who had greater odds of higher literacy relative to cognition. This latter finding is consistent with our previous finding (Han, Boyle, James, et al., 2016) that being male was associated with greater odds of higher decision-making abilities relative to cognition. Historically in the U.S., males tended to be primary decision makers for family units, thus gaining more experience with decision making matters despite lower cognitive abilities. This notion may also explain the present finding. In fact, research has found men to be more financially literate than women across the globe (Lusardi & Mitchell, 2008, 2011, 2014), possibly due to the necessity of developing strong literacy in order to make relevant household decisions in this age group. Thus, the sample of men in our study may have developed higher literacy despite relatively lower cognitive abilities due to greater experience and exposure to financial and healthcare matters. The opposite effect was also found, with women in our sample exhibiting higher cognition in the face of lower literacy. This finding is consistent with a study that found that even among the most educated women, financial knowledge was low (Mahdavi & Horton, 2014). One possibility for these findings is that women have less experience dealing with financial and healthcare decisions (see Lusardi & Mitchell, 2008), particularly for females of older generations like those in the current sample. These relationships have the potential of diminishing as younger generations age, societal expectations continue to evolve, and women have increasing autonomy in decision making. In fact, research with younger generations may find that this is already the case.

Trust was the only psychosocial factor to predict discrepancy group status. Specifically, lower trust was associated with greater odds of being in the L>C group. One possibility for this relationship is the fact that trust may moderate a person’s efforts for maintaining and updating their own financial and health literacy, such that those with less trust take greater measures to being literate so as to not rely on the knowledge of others. The notion that less trust relates to higher literacy is consistent with economics research into financial market investments suggesting that increased trust relates to increased delegation of financial matters to financial advisers (Cruciani & Rigoni, 2017 for review). Such research suggests that individuals with low levels of trust might be more likely to develop literacy so that they are able to make independent financial and other critical decisions. Conversely, another possibility is that those with higher literacy have greater knowledge of the issues associated with financial and healthcare Cognitive and Literacy Discrepancies 20 institutions, thereby leading to decreased trust of these institutions. Currently, the impact of knowledge on institutional trust remains understudied (for an exception, see PytlikZillig et al., 2017). Of note, this was not a demographically representative sample in that most individuals were white and female. Low trust may differentially impact individuals from different demographic strata, particularly minority groups that face systemic discrimination and reduced access to resources. Examining these relationships in minority groups is an important and necessary avenue for future research to fully understand the impact of trust on literacy and cognition.

Although literacy and cognition appear to be partially dissociable functions, findings from the present study support the notion that cognition is an important correlate of literacy, with over 75% of the study sample exhibiting relatively equal scores between the two functions. This finding is in line with our prior work reporting associations between cognition and literacy (J. S. Bennett et al., 2012; Boyle, Yu, et al., 2013b; Federman et al., 2009; Wilson et al., 2017). A recent study by our group (Boyle, Yu, et al., 2013b) found that declines in both episodic memory and executive function predicted health and financial literacy scores. Cognitive decline may exert its effect on literacy by impacting one’s ability to process and reason through cognitively demanding health and financial materials and situations, encode information learned through these experiences, and apply such information to future decision-making (Boyle, Yu, et al., 2013b). Findings from this study and previous studies linking cognition to literacy have implications for public policy and interventions to improve literacy in old age. For example, redesigning financial and health-related materials so that they are more accessible, comprehensible, and relevant to older adults, especially those with compromised cognition, may be beneficial (Boyle, Yu, et al., 2013b). Such changes may reduce the psychological and economic burden that low literacy and poor decision making have on elderly individuals, their caregivers, and society.

The present findings expand upon the model of literacy proposed by our group (Boyle, Yu et al., 2013) in which diverse individual factors independently impact a person’s financial and health literacy, thus contributing to a divergence between cognition and literacy. This is supported by our finding of a significant subset of individuals whose literacy scores diverge from their cognitive scores. Significant findings with income and sex suggest that a person’s lifetime experiences or lack thereof may affect the degree to which they are able to develop, maintain, and update literacy, possibly via access to resources (low income) and societal expectations (sex). Somewhat consistent with Boyle et al. (2013) who found both direct and indirect effects of age on literacy, in our study age did not predict discrepancy between literacy and cognition, supporting the fact that it is associated with literacy partially via cognition. In contrast, we did not find education to be predictive of discrepancy between literacy and cognition in the fully adjusted model, possibly suggesting that the effect of education on literacy may be mediated by other factors such as income. Our findings are also consistent with Marmot’s (2005) characterization of social determinants of health (e.g., low income) as being ‘causes of the causes’, because they impact downstream behaviors and states (e.g., literacy, stress, health behaviors; Marmot, 2005). Trust may further modify the influence of such contextual factors on literacy by influencing the degree to which individuals are motivated to acquire literacy throughout their lifespan despite these contextual factors, though this could not be tested in the present study. The present data also cannot comment on directionality of findings, which is a notable limitation, and thus future longitudinal studies are necessary to determine whether Cognitive and Literacy Discrepancies 22 this model holds. As previously mentioned, it is possible that literacy-cognition discrepancies impact income and trust rather than the reverse.

Of note, our non-demented sample of older adults included some who met criteria for mild cognitive impairment (MCI). To investigate whether findings differ if we consider only a cognitively non-impaired group of older adults, we ran a follow-up analysis that excluded those in our sample who meet criteria for MCI (N = 90; Boyle et al., 2005; Han et al., 2012); see Supplementary Tables S1 and S2 for demographics and results of these analyses). In doing so, the proportion of those with a discrepancy between literacy and cognition increases from 24.2% in the full sample to 29.7% in the cognitively healthy sample of 624 participants. This increase appears to be mostly driven by an increase in the proportion of individuals falling in the L>C group (from 9.9% in full sample to 15.2% in cognitively healthy sample). In comparison, the proportion of individuals in the L<C group changed from 14.3% in the full sample to 15.2% in the cognitively healthy sample. This suggests that discrepancy between literacy and cognition is even more common among non-impaired individuals (though this was not directly tested). In terms of predictors of discrepancy, some differences between the full sample and the non-impaired sample arose. While low income remains significantly associated with greater odds of being in either discrepancy group (L>C OR = 0.87, p = 0.016; L<C OR = 0.85, p = 0.002), lower trust is no longer significantly associated with increased odds of being in the L>C group after excluding those with MCI (p = 0.275), suggesting that the findings for trust may have been driven by persons with some cognitive impairment. Additionally, male sex remains associated with greater odds of being in the L>C group (OR = 2.20, p = 0.004) and lower odds of being in the L<C group (OR = 0.35, p = 0.013).

The present study has some limitations. First, the sample studied is predominantly white and therefore not generalizable to the larger population of older adults in the United States, or to individuals outside of the United States given that the literacy measure asks questions specifically relevant to health and financial practices within the United States. Evaluating similar questions with more representative groups is important. Additionally, a small amount of measurement error may have partially contributed to discrepancies between literacy and cognition. Another limitation relates to the discrepancy approach utilized in the study. By categorizing participants into three groups, information regarding demographic and psychosocial contributions to different levels of literacy and cognition may be lost. Nevertheless, the present findings offer a unique and arguably more sensitive approach (e.g., see Fine et al., 2008; Jacobson et al., 2002) to understanding how and why the two functions may diverge. Another related limitation is our choice to define a discrepancy as a 1SD or greater difference between literacy and cognition. This choice was made in accordance with past studies that have utilized this approach (e.g., Delis et al., 2007; Delis et al., 1992; Fine et al., 2008; Han, Boyle, James, et al., 2016; Jacobson et al., 2002; Massman et al., 1993; Strite et al., 1997). However, it is important to note that different cut-offs of discrepancy (e.g., 1.5SD difference between literacy and cognition) may yield different results. Future studies may consider exploring the clinical significance of different cutoffs when assessing literacy-cognition discrepancies. Finally, given the cross-sectional nature of the study, causality of relationships is not clear. For example, in contrast to the more obvious interpretation that low income contributes to literacy-cognition discrepancies, the reverse may also be true; literacy-cognition discrepancies may impact income levels, possibly by contributing to suboptimal financial and health decisions. Future longitudinal designs will help to elucidate the temporality of relationships.

Despite these limitations, this study has notable strengths including use of a large, well-characterized community-based sample of older adults, a detailed measure of literacy, comprehensive battery of neuropsychological tests to characterize global cognition, and well-established, validated measures to examine psychosocial factors of interest. The study also utilized a novel approach towards understanding factors that contribute to literacy and cognition discrepancies in old age. Information gleaned from this study informs on the identification of those at high-risk of low literacy (e.g., low income, low trust, female sex). In addition, these findings suggest that interventions that assist individuals in identifying trusted resources for gaining literacy and domain-specific knowledge, ascertain validity of literacy resources, and target underserved populations may be of particular benefit. Such interventions will ultimately prolong independence and improve the health and financial well-being of aging adults.

Supplementary Material

Supplemental Material

Public Significance Statement.

Financial and health literacy are associated with important health outcomes in old age, including cognition and mortality. This study examined discrepancies between literacy and cognition in a large cohort of community dwelling older adults and examined what factors were associated with discrepancies. Findings inform the development of potential interventions targeted towards maintaining or improving literacy and cognition in older adults.

Acknowledgments

This work was supported by the National Institute on Aging at the National Institutes of Health grants R01AG017917 to DAB, R01AG033678 to PAB, and R01AG055430 to SDH.

Footnotes

The authors gratefully thank the Rush Memory and Aging Project staff and participants and declare no competing financial interests.

References

  1. Agarwal S, Driscoll JC, Gabaix X, & Laibson D (2009). The age of reason: Financial decisions over the life cycle and implications for regulation. Brookings Papers on Economic Activity, 2009(2), 51–117. [Google Scholar]
  2. Baker DW, Gazmararian JA, Sudano J, & Patterson M (2000). The association between age and health literacy among elderly persons. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 55(6), S368–S374. [DOI] [PubMed] [Google Scholar]
  3. Baker DW, Wolf MS, Feinglass J, Thompson JA, Gazmararian JA, & Huang J (2007). Health literacy and mortality among elderly persons. Archives of internal medicine, 167(14), 1503–1509. [DOI] [PubMed] [Google Scholar]
  4. Banks J (2010). Cognitive function, financial literacy and financial outcomes at older ages: Introduction. The Economic Journal, 120(548). [Google Scholar]
  5. Bennett DA, Schneider J, Arvanitakis Z, Kelly J, Aggarwal N, Shah R, & Wilson R (2006). Neuropathology of older persons without cognitive impairment from two community-based studies. Neurology, 66(12), 1837–1844. [DOI] [PubMed] [Google Scholar]
  6. Bennett DA, Schneider JA, Buchman AS, Barnes LL, Boyle PA, & Wilson RS (2012a). Overview and findings from the rush Memory and Aging Project. Curr Alzheimer Res, 9(6), 646–663. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bennett DA, Schneider JA, Buchman AS, Mendes de Leon C, Bienias JL, & Wilson RS (2005). The Rush Memory and Aging Project: study design and baseline characteristics of the study cohort. Neuroepidemiology, 25(4), 163–175. doi: 10.1159/000087446 [DOI] [PubMed] [Google Scholar]
  8. Bennett DA, Wilson RS, Boyle PA, Buchman AS, & Schneider JA (2012b). Relation of neuropathology to cognition in persons without cognitive impairment. Annals of neurology, 72(4), 599–609. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bennett JS, Boyle PA, James BD, & Bennett DA (2012). Correlates of health and financial literacy in older adults without dementia. BMC geriatrics, 12(1), 30. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Bostock S, & Steptoe A (2012). Association between low functional health literacy and mortality in older adults: longitudinal cohort study. BMJ, 344, e1602. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Boyle PA, Wilson RS, Aggarwal NT, Arvanitakis Z, Kelly J, Bienias JL, & Bennett DA (2005). Parkinsonian signs in subjects with mild cognitive impairment. Neurology, 65(12), 1901–1906. doi: 10.1212/01.wnl.0000188878.81385.73 [DOI] [PubMed] [Google Scholar]
  12. Boyle PA, Wilson RS, Yu L, Barr AM, Honer WG, Schneider JA, & Bennett DA (2013a). Much of late life cognitive decline is not due to common neurodegenerative pathologies. Annals of neurology, 74(3), 478–489. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Boyle PA, Yu L, Wilson RS, Segawa E, Buchman AS, & Bennett DA (2013b). Cognitive decline impairs financial and health literacy among community-based older persons without dementia. Psychology and aging, 28(3), 614. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Braveman P, Egerter S, & Williams DR (2011). The social determinants of health: coming of age. Annual review of public health, 32, 381–398. [DOI] [PubMed] [Google Scholar]
  15. Costa PT, & Mac Crae RR (1992). Neo Personality Inventory-Revised (NEO PI-R): Psychological Assessment Resources Odessa, FL. [Google Scholar]
  16. Cruciani C, & Rigoni U (2017). Trust and financial literacy. Challenges in ensuring financial competencies
  17. de Jong-Gierveld J (1987). Developing and testing a model of loneliness. Journal of personality and social psychology, 53(1), 119. [DOI] [PubMed] [Google Scholar]
  18. de Jong-Gierveld J, & Kamphuls F (1985). The development of a Rasch-type loneliness scale. Applied psychological measurement, 9(3), 289–299. [Google Scholar]
  19. Delis DC, Lansing A, Houston WS, Wetter S, Han SD, Jacobson M, … Kramer J (2007). Creativity lost: The importance of testing higher-level executive functions in school-age children and adolescents. Journal of Psychoeducational Assessment, 25(1), 29–40. [Google Scholar]
  20. Delis DC, Massman PJ, Butters N, Salmon DP, Shear PK, Demadura T, & Filoteo JV (1992). Spatial cognition in Alzheimer’s disease: Subtypes of global-local impairment. Journal of clinical and experimental neuropsychology, 14(4), 463–477. [DOI] [PubMed] [Google Scholar]
  21. Federman AD, Sano M, Wolf MS, Siu AL, & Halm EA (2009). Health literacy and cognitive performance in older adults. Journal of the American Geriatrics Society, 57(8), 1475–1480. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Fine EM, Delis DC, Wetter SR, Jacobson MW, Jak AJ, McDonald CR, … Bondi MW (2008). Cognitive discrepancies versus APOE genotype as predictors of cognitive decline in normal-functioning elderly individuals: a longitudinal study. The American Journal of Geriatric Psychiatry, 16(5), 366–374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Fleischman DA, Wilson RS, Bienias JL, & Bennett DA (2005). Parkinsonian signs and cognitive function in old age. Journal of the International Neuropsychological Society, 11(5), 591–597. [DOI] [PubMed] [Google Scholar]
  24. Han SD, Arfanakis K, Fleischman DA, Leurgans SE, Tuminello ER, Edmonds EC, & Bennett DA (2012). Functional connectivity variations in mild cognitive impairment: associations with cognitive function. J Int Neuropsychol Soc, 18(1), 39–48. doi: 10.1017/S1355617711001299 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Han SD, Boyle PA, Arfanakis K, Fleischman D, Yu L, James BD, & Bennett DA (2016). Financial literacy is associated with white matter integrity in old age. Neuroimage, 130, 223–229. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Han SD, Boyle PA, James BD, Yu L, Barnes LL, & Bennett DA (2016). Discrepancies between cognition and decision making in older adults. Aging clinical and experimental research, 28(1), 99–108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Han SD, Boyle PA, Yu L, Fleischman DA, Arfanakis K, Leurgans S, & Bennett DA (2014). Financial literacy is associated with medial brain region functional connectivity in old age. Archives of gerontology and geriatrics, 59(2), 429–438. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Jacobson MW, Delis DC, Bondi MW, & Salmon DP (2002). Do neuropsychological tests detect preclinical Alzheimer’s disease: Individual-test versus cognitive-discrepancy score analyses. Neuropsychology, 16(2), 132. [DOI] [PubMed] [Google Scholar]
  29. James BD, Boyle PA, & Bennett DA (2014). Correlates of susceptibility to scams in older adults without dementia. Journal of elder abuse & neglect, 26(2), 107–122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. James BD, Boyle PA, Bennett JS, & Bennett DA (2012). The impact of health and financial literacy on decision making in community-based older adults. Gerontology, 58(6), 531–539. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kirsch IS (1993). Adult literacy in America: A first look at the results of the National Adult Literacy Survey: ERIC. [Google Scholar]
  32. Kohout FJ, Berkman LF, Evans DA, & Cornoni-Huntley J (1993). Two shorter forms of the CES-D depression symptoms index. Journal of aging and health, 5(2), 179–193. [DOI] [PubMed] [Google Scholar]
  33. Kutner M, Greenburg E, Jin Y, & Paulsen C (2006). The Health Literacy of America’s Adults: Results from the 2003 National Assessment of Adult Literacy. NCES 2006–483. National Center for Education Statistics
  34. Lusardi A, & Mitchell OS (2007). Baby boomer retirement security: The roles of planning, financial literacy, and housing wealth. Journal of monetary Economics, 54(1), 205–224. [Google Scholar]
  35. Lusardi A, & Mitchell OS (2008). Planning and financial literacy: How do women fare? American Economic Review, 98(2), 413–417. [Google Scholar]
  36. Lusardi A, & Mitchell OS (2011). Financial literacy around the world: an overview. Journal of pension economics & finance, 10(4), 497–508. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Lusardi A, & Mitchell OS (2014). The economic importance of financial literacy: Theory and evidence. Journal of economic literature, 52(1), 5–44. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lusardi A, Mitchell OS, & Curto V (2009). Financial literacy among the young: Evidence and implications for consumer policy Retrieved from
  39. Lusardi A, & Mitchelli OS (2007). Financial literacy and retirement preparedness: Evidence and implications for financial education. Business economics, 42(1), 35–44. [Google Scholar]
  40. Mahdavi M, & Horton NJ (2014). Financial knowledge among educated women: Room for improvement. Journal of Consumer Affairs, 48(2), 403–417. [Google Scholar]
  41. Mandell L, & Klein LS (2007). Motivation and financial literacy. Financial services review, 16(2), 105. [Google Scholar]
  42. Marmot M (2005). Social determinants of health inequalities. The lancet, 365(9464), 1099–1104. [DOI] [PubMed] [Google Scholar]
  43. Marmot M, Friel S, Bell R, Houweling TA, Taylor S, on behalf of the Commission of Social Determinants of Health. (2008). Closing the gap in a generation: health equity through action on the social determinants of health. The Lancet, 372(9650), 1661–1669. [DOI] [PubMed] [Google Scholar]
  44. Massman PJ, Delis DC, Filoteo JV, Butters N, Salmon DP, & Demadura TL (1993). Mechanisms of spatial impairment in Alzheimer’s disease subgroups: Differential breakdown of directed attention to global-local stimuli. Neuropsychology, 7(2), 172. [Google Scholar]
  45. McKhann G, Drachman D, Folstein M, Katzman R, Price D, & Stadlan EM (1984). Clinical diagnosis of Alzheimer’s disease Report of the NINCDS-ADRDA Work Group* under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology, 34(7), 939–939. [DOI] [PubMed] [Google Scholar]
  46. Noon KL, & Fogarty GJ (2007). Cognitive and personality predictors of financial literacy among adult Australians. Paper presented at the Proceedings of the 42nd Australian Psychological Society Annual Conference: Psychology Making an Impact. [Google Scholar]
  47. PytlikZillig LM, Kimbrough CD, Shockley E, Neal TM, Herian MN, Hamm JA, … Tomkins AJ (2017). A longitudinal and experimental study of the impact of knowledge on the bases of institutional trust. PLoS One, 12(4), e0175387. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Radloff LS (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied psychological measurement, 1(3), 385–401. [Google Scholar]
  49. Schneider J, Arvanitakis Z, Yu L, Boyle P, Leurgans S, & Bennett D (2012). Cognitive impairment, decline and fluctuations in older community-dwelling subjects with Lewy bodies. Brain, 135(10), 3005–3014. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Strite D, Massman PJ, Cooke N, & Doody RS (1997). Neuropsychological asymmetry in Alzheimer’s disease: Verbal versus visuoconstructional deficits across stages of dementia. Journal of the International Neuropsychological Society, 3(5), 420–427. [PubMed] [Google Scholar]
  51. Taylor M, Jenkins S, & Sacker A (2009). Financial capability and wellbeing: Evidence from the BHPS. Financial Services Authority Occasional Paper, 34. [Google Scholar]
  52. UNESCO Education Sector. (2004). The plurality of literacy and its implications for policies and programs: Position paper United National Educational, Scientific and Cultural Organization (UNESCO), Paris. [Google Scholar]
  53. Wilson RS, Barnes L, De Leon CM, Aggarwal N, Schneider J, Bach J, … Evans D (2002b). Depressive symptoms, cognitive decline, and risk of AD in older persons. Neurology, 59(3), 364–370. [DOI] [PubMed] [Google Scholar]
  54. Wilson RS, Beckett LA, Barnes LL, Schneider JA, Bach J, Evans DA, & Bennett DA (2002a). Individual differences in rates of change in cognitive abilities of older persons. Psychology and Aging, 17, 179–193. [PubMed] [Google Scholar]
  55. Wilson RS, Barnes LL, & Bennett DA (2003). Assessment of lifetime participation in cognitively stimulating activities. Journal of clinical and experimental neuropsychology, 25(5), 634–642. [DOI] [PubMed] [Google Scholar]
  56. Wilson RS, Yu L, James BD, Bennett DA, & Boyle PA (2017). Association of financial and health literacy with cognitive health in old age. Aging, Neuropsychology, and Cognition, 24(2), 186–197. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Material
NIHMS1038224-supplement-Supplemental_Material.pdf (48.6KB, pdf)

RESOURCES