- Research article
- Open Access
- Published: 05 December 2019
BMC Geriatrics volume 19, Article number: 343 [2019] Cite this article
13k Accesses
25 Citations
8 Altmetric
Metrics details
Abstract
Background
Five-factor model [FFM] personality traits have been associated consistently with risk of Alzheimer’s disease and related dementias [ADRD]. Less is known about how these traits are associated with functioning in specific domains of cognitive function in older adulthood.
Methods
Participants [N = 2865] were drawn from the 2016 Harmonized Cognitive Assessment Protocol sub-study of the Health and Retirement Study [HRS]. Participants completed a battery of cognitive tasks that measured performance in five domains: Memory [eight tasks], speed-attention-executive [five tasks], visuospatial ability [three tasks], fluency [one task], and numeric reasoning [one task]. Participants completed an FFM personality measure as part of the regular HRS assessment in either 2014 or 2016. Linear regression was used to examine the association between the traits and each cognitive task and composite scores for the five domains, controlling for age, sex, race, ethnicity, and education. We also tested whether the associations were moderated by these sociodemographic factors or mental status.
Results
Neuroticism was associated with worse performance on all of the cognitive tasks. Conscientiousness was associated with better performance across all five cognitive domains, although not necessarily with every task. Openness and Agreeableness were associated with better performance in all domains, except for numeric reasoning. Extraversion was associated with better speed-attention-executive and fluency. There was no robust evidence that the association between personality and cognition was moderated by sociodemographic characteristics or global cognitive function.
Conclusions
Personality traits have pervasive associations with functioning across five cognitive domains. Consistent with the literature on personality and risk of ADRD, Neuroticism and Conscientiousness were associated with cognitive performance in the expected direction in all domains. Extraversion was the only trait that showed domain-specific associations. The present research supports models of personality and health in the context of cognition and suggests that personality is associated with intermediate markers of cognitive health.
Peer Review reports
Background
Five Factor Model [FFM] personality traits [1] are associated consistently with significant cognitive impairment. In particular, individuals who score higher in Neuroticism [the tendency to experience negative emotions and stress] or lower in Conscientiousness [the tendency to be organized, disciplined, and responsible] are at greater risk of both mild cognitive impairments [2] and of Alzheimer’s disease [3]. Much of the work on personality and cognition in older adulthood has focused on global cognitive outcomes. And yet there are many cognitive domains that contribute to overall cognitive health [4]. The present research addresses five common domains of function [5]: Episodic memory, speed-attention-executive, visuospatial ability, fluency, and numeric reasoning. Episodic memory is memory for specific events in time and place. It includes memory for personal events that range from the distant past to events that just happened in the current moment. It is often measured with word lists that participants recall immediately and after a short delay. Speed and attention are functions that assess how quickly someone can respond to a stimulus [speed] and how well they can attend to the stimulus [attention]; executive function includes these basic functions as well as cognitive flexibility. Visuospatial ability is the ability to visualize, rotate, and manipulate shapes in more than one dimension. Fluency is the ability to produce and use words correctly. Finally, numeric reasoning is the ability to manipulate numbers and includes basic arithmetic.
There are a number of reasons why personality traits may be associated with performance on cognitive tasks. Individuals who score higher in Neuroticism, for example, tend to be anxious and vulnerable to stress [6] and have self-presentational concerns around other people [7]. Such anxiety and self-conscientiousness are likely to inhibit performance on tasks administered in the presence of a tester. And indeed, individuals higher in Neuroticism tend to remember fewer words [8], respond slower on response time tasks [9, 10], have worse visuospatial performance [9], and produce fewer words on fluency tasks [11]. The associations between Extraversion [the tendency to be outgoing, sociable, and active] and cognition, in contrast, appear to be more domain specific. Extraversion has been associated with faster performance [10] and greater verbal fluency [11], whereas the association with episodic memory is more mixed [8,9,10], and it tends to be unrelated to visuospatial ability and numeric reasoning. These domain-specific associations are consistent with characteristics of this trait that include talkativeness [12] and vigor [13]. Cognitive flexibility and verbal abilities are core characteristics of Openness [14]. And, as expected, this trait tends to be related to better performance on tasks that include a verbal component [15] and on tasks that require cognitive flexibility [16, 17]. Agreeableness [the tendency to be trusting and empathetic] is sometimes associated with greater dementia risk [18] but is not associated consistently with performance on cognitive tasks [19]. Finally, Conscientiousness is associated with achievement striving and organization and a lifestyle that supports maintaining cognitive health across adulthood [20]. As such, it tends to be associated with better performance on a range of tasks [8, 11], but the associations are not always consistent [19]. For example, some find positive associations between Conscientiousness and better performance in tasks that measure speed, attention, and executive function [9] and others find no relation [10, 17, 21]. These differences may be due, in part, to differences in sample size [effects are generally modest and require large sample sizes for adequate power], differences in sample population [e.g., older versus younger adults; clinical versus nonclinical populations], and/or differences in measurement of both personality and tasks that measure speed, attention, and executive function. More broadly, and for these reasons, it is challenging to synthesize the literature on personality and measures of performance in these five cognitive domains. Focusing on large sample studies [N > 1000] that include validated measures of personality and cognition, the strongest evidence that personality is associated with specific cognitive domains is for the domains of memory and verbal fluency. In large samples of older adults, for example, higher Neuroticism tends to be associated with worse memory performance, whereas higher Openness and Conscientiousness tend to be associated with better memory [8, 22]. Further, our recent meta-analysis on personality and verbal fluency [meta-analytic N > 85,000] indicated that higher Neuroticism was associated with lower fluency whereas higher Extraversion, Openness, and Conscientiousness were associated with greater fluency [11]. As such, stronger hypotheses can be made for memory and fluency based on this previous literature than for speed-attention-executive, visuospatial ability, and numeric reasoning.
The literature on personality and cognition suggests that the traits may have differential associations with different aspects of cognitive function. Previous research on personality and cognition has tended to focus on one or two personality traits [typically Neuroticism and Extraversion] and/or individual cognitive functions [e.g., episodic memory]. Although there is growing recognition of the importance of including all five traits and tasks from multiple cognitive domains, such studies remain relatively rare. The present research sought to unify the literature by examining the association between all FFM traits and five common domains of cognitive function in a relatively large sample of older adults. The large sample also allowed us to examine whether the associations varied by sociodemographic characteristics. Based on the literature on personality and cognition, we expected higher Neuroticism to be associated with worse performance on all of the cognitive tasks and higher Conscientiousness and Openness to be associated with better performance. In contrast, we expected higher Extraversion to be associated with better performance on speed and fluency tasks. We did not expect Agreeableness to be associated with the cognitive tasks. In addition to the main effect, we examine whether the association between personality and cognition is moderated by age, sex, race, ethnicity, education, or global cognitive function.
Method
Participants and procedure
Participants were part of the 2016 Harmonized Cognitive Assessment Protocol [HCAP], an ongoing sub-study of the Health and Retirement Study [HRS]. Participants were selected to be a part of the HCAP assessment if they were 65 years or older and completed the 2016 interview of the HRS. Of the eligible participants, a subset was randomly selected and invited to participate in the HCAP assessment [N = 5500]. A total of 3496 participants completed at least some part of it. Participants could have the HCAP assessment administered in either English or Spanish. Participants who also completed the personality measure in either the 2014 or 2016 regular HRS assessment were selected for analysis. A total of 2865 participants had complete data on personality and some measures in HCAP. The sample was on average 76.49 [SD = 7.36] years old, with 66% of the sample between the ages of 65–79, 30% between the ages of 80–89, and 4% over the age of 90. The sample was 60% female, 14% African American, 9% Hispanic, and had an average of 12.93 [SD = 2.98] years of education. By comparison, the Census estimates that the US population aged 65 and older is 56% female, 9% African American, 8% Hispanic, and that 84% of this population had completed high school [23].
Analytic samples ranged from 2456 [numeric reasoning] to 2814 [CERAD immediate recall, verbal fluency] based on missing data across the cognitive tasks. Compared to participants who had the personality assessment available, participants without the personality assessment and thus not included in the analyses [n = 629] were older [d = .24, p < .01], more likely to be Hispanic [χ2 = 38.16, p < .01], more likely to be a race other than white [χ2 = 59.85, p < .01], and had fewer years of education [d = .38, p < .01]; there was no difference in participant sex. Further, there were differences on all of the cognitive tasks, with ds that ranged from .30 [p < .01; Backward Count] to .71 [p < .01; Mini Mental State Examination]; across all cognitive tasks, participants who also had the personality assessment performed better than participants who did not have personality. More information on the HCAP assessment, sampling, and how to obtain the data can be found at //hrs.isr.umich.edu/news/2016-harmonized-cognitive-assessment-protocol-hcap-early-version-10. The Health and Retirement Study make their data available to the public, but access to the HCAP data requires an additional authentication process to verify the identity of the person and institute requesting access to the data. We acquired this administrative permission to access the HCAP data. Information about how to access the HCAP data can be found at //hrs.isr.umich.edu/data-products/cognition-data. The Institutional Review Board at the Florida State University approved this research [protocol #IRB00000446, “Secondary Data Analysis of Public Health Databases”].
Measures
Personality
Participants completed the Midlife Development Inventory [MIDI [24]] as part of the Leave-Behind Questionnaire in either 2014 or 2016. The MIDI assesses FFM personality traits with 26 adjectives. Items on the MIDI measured Neuroticism [e.g., moody; alpha = .71], Extraversion [e.g., talkative; alpha = .75], Openness [e.g., creative; alpha = .80], Agreeableness [e.g., helpful; alpha = .79], and Conscientiousness [e.g., organized; alpha = .71]. Items were rated on a scale from 1 [a lot] to 4 [not at all] and reverse scored in the direction of the trait label. The mean was taken across items for each trait [range 1–4].
Cognition
HCAP is an extensive assessment of cognitive function that covers the domains of episodic memory, speed-attention-executive, visuospatial ability, fluency, and numeric reasoning. Detailed information about test administration and scoring can be found in Weir and colleagues [25].
Episodic memory
Participants completed several measures of episodic memory, measured as immediate recall, delayed recall, and recognition. Participants completed the CERAD Word List Learning and Recall Task. Participants were presented visually with a list of 10 words, two seconds at a time for each word. Participants read each word and after the last word were asked to recall as many words from the list as possible [immediate recall]. After a short delay, participants were again asked to recall as many words from the list as possible [delayed recall]. Finally, participants were shown 10 target words and 10 foils and were asked to indicate which words were on the original list [recognition]. Participants also completed two story memory tasks: Brave Man and the Wechsler Memory Scale Logical Memory I. Both tasks involved being read a passage and being asked to report back the main points of the story immediately and after a short delay. The Logical Memory task also included a recognition test, in which participants were asked 15 yes/no questions about the story.
Speed-attention-executive
This domain was assessed with several tasks. In the Letter Cancellation Test, participants were given one minute to cross out as many “P” and “W” letters as possible from a large grid of letters. The score was the last letter gotten to at the one-minute mark. Participants completed the Backward Count task as a measure of processing speed in which they counted backward from 100 as fast as possible. The count of numbers said in 30 s was the score. The Symbol-Digit Modalities Test had random geometric figures and a separate key that paired numbers with each figure. Participants were asked to substitute a number for each figure on the sheet of paper. The score was the number of correct pairings made in 90 s. The Trail Making Test had two parts. Part A was a sheet of numbers in circles on a page and participants were asked to connect the consecutively numbered circles as fast as possible. Trails B included letters as well as numbers and participants had to switch between numbers and letters as quickly as possible. For both parts, the outcome was time to complete the task [i.e., slower times indicated worse performance].
Visuospatial ability
The CERAD Constructional Praxis task required participants to copy geometric forms that varied in difficulty both immediately and after a short delay. Note that performance on the delay task reflects visual memory as well as visuospatial ability. Raven’s Standard Progressive Matrices were geometric pictures with a small section missing. Participants were asked to choose the correct picture from a set that correctly completed the picture. Participants completed 17 pictures from the total Raven’s test.
Fluency
Fluency was measured with a semantic verbal fluency task. Specifically, participants were asked to name as many animals as possible in 60 s.
Numeric reasoning
The HRS number series was a measure of numeric reasoning. Participants were presented with a series of numbers with one or two numbers missing. Participants were asked to identify the missing numbers. The test was not timed and participants could take as much time as necessary to complete it. The test was adaptive, such that items got more or less difficult depending on participants’ responses.
Global cognitive function
Finally, participants completed the Mini-Mental State Examination [MMSE] as a measure of global cognitive function [26]. The MMSE was used to test cognitive status as a moderator of the relation between personality and the cognitive tasks and not as an outcome. The total MMSE score was dichotomized into no impairment [≥24, coded as 0] and any impairment [≤23, coded as 1].
Covariates
Covariates were self-reported age in years, sex [female = 1, male = 0], race [African American = 1 [dummy variable 1], other/unknown = 1 [dummy variable 2] both compared to white = 0], Hispanic ethnicity [1 = yes, 0 = no], and education in years. Some participants chose to have the HCAP administered in Spanish. We included language of administration [0 = English, 1 = Spanish] as a covariate.
Statistical approach
We used linear regression to examine the association between personality and the cognitive measures. Specifically, each cognitive task was predicted separately by each trait, controlling for age, sex, race, ethnicity, education, and language of HCAP administration. For the domains with multiple tasks [episodic memory, speed-attention-executive, visuospatial ability], in addition to the individual tasks, the score for each task was standardized and then the mean taken across the tasks within the domain as an overall measure of the domain [Trails A and Trails B were multiplied by − 1 to reverse the direction of the scoring to be consistent with the scoring of the other tasks in this domain]. We then tested whether the association between personality and each domain was moderated by age, sex, race, ethnicity, education, or mental status. Finally, we did a threshold analysis to test whether the traits were associated with performance below a threshold. Specifically, we tested whether personality was associated with risk of performing at least one standard deviation below the mean [a relatively arbitrary but standard cutoff for threshold analyses] for each of the five domains. For all analyses, p was set to