Higher purpose in life and education were associated with better cognition among older adults

Background  With aging, some cognitive abilities change because of neurobiological processes. Cognition may also be influenced by psychosocial aspects. Objective  To describe the relationship between a measure of neuroticism, depression symptoms, purpose in life, and cognitive performance in community-dwelling older adults. Methods  This was a cross-sectional analysis based on the data from the second wave of the Frailty in Brazilian Older Adults (FIBRA) study, carried out between 2016 and 2017. The sample consisted of 419 older people (≥ 72 years old) cognitively unimpaired and mostly with low education. The variables of interest were sociodemographic, Neuroticism domain from the NEO-PI-R, Geriatric Depression Scale (GDS), Purpose in Life (PiL) scale, and a cognitive composite score which included the Mini-Mental State Examination (MMSE), and the scores for the sub-items of the Mini-Addenbrooke's Cognitive Examination (M-ACE), namely, Verbal Fluency (VF) – Animal, Clock Drawing Test (CDT), Episodic Memory (name and address). Results  There was a greater number of women (70%), with older age (median = 80 years, IQR = 77-82), and low education (median = 4 years, IQR = 2-5). In the bivariate correlations, years of education (ρ = 0.415; p < 0.001) and PiL (ρ = 0.220; p < 0.001) were positively associated with cognition. Neuroticism (ρ = -0.175; p < 0.001) and depression symptoms (ρ = -0.185; p < 0.001) were negatively associated with cognition. In the logistic regression, after including confounding variables, the associations between cognition and PiL (OR = 2.04; p = 0.007) and education (OR = 1.32; p < 0.001) remained significant. Conclusion  Low PiL and low education levels were associated with worse cognition among older adults. Such results may be of relevance in programs that aim to improve cognition among older adults.


INTRODUCTION
Age-related cognitive change is one of the most studied aspects of aging.With age, there may be a reduction in cognitive performance, with implications for daily life and well-being. 1 Cognitive changes are related to brain aging, that is, the anatomical and physiological changes that occur throughout the life cycle. 2][5] Among psychosocial aspects, studies have suggested that neuroticism, a personality factor related to negative emotional and behavioral reactions to stressful situations, has an association with cognitive performance. 6This aspect of personality, although present in all individuals, when at elevated levels has often been linked to poor cognitive aging in longitudinal studies.For instance, Sutin et al. 7 reported that high scores in neuroticism among respondents were associated with worse performance in domains such as memory, function executive, attention, process speed, visuospatial ability, verbal fluency, and numerical reasoning.Another longitudinal study identified that low neuroticism was associated with less cognitive decline and greater longevity. 3epression is a mood disorder that significantly affects an individual's ability to function and relate to others.Depression symptoms (DS) have commonly been associated with reduced cognitive performance.DS can influence cognitive domains, such as learning and memory, attention and concentration, executive functions, and processing speed. 4Also, in a longitudinal study, lifetime recurrent depression episodes were predictive of worse cognitive status, and a higher risk of cognitive impairment in early old age, 8 although this finding is contested by other studies (e.g., Burhanullah et al. 9 ).
PiL is one of the most important domains of the psychological well-being construct.It refers to the meaning, purpose, and sense of direction that people assign to their own lives. 10mportantly, higher PiL is considered an important predictor of healthy aging.For instance, in previous work, Ryff and   Conclusion Low PiL and low education levels were associated with worse cognition among older adults.Such results may be of relevance in programs that aim to improve cognition among older adults.

PiL and education relate to better cognition
Oliveira et al.
colleagues have documented that higher PiL predicts greater longevity and better health behaviors. 11Recent research has shown that higher PiL can also have a positive impact on cognition.For example, in a longitudinal study, higher PiL was reported as having a protective effect against cognitive decline. 5It was also associated with better coping skills during stressful events and greater longevity. 12In a recent metaanalysis about PiL and cognition, PiL was associated with better performance in episodic memory and verbal fluency tasks. 13he psychosocial aspects described above represent important mental health parameters that can significantly influence cognition among older people. 7,8,12However, these data originated from high-income countries, where socioeconomic conditions and mental healthcare access tend to be significantly better than in middle-and low-income countries.Therefore, we understand that the association between psychosocial aspects and cognition should also be investigated in such adverse contexts.In addition, this association has rarely been explored among the oldest old, and no previous studies about the topic were identified in Brazil.Therefore, the objective of the present study was to investigate the associations between sociodemographic, neuroticism, DS, PiL variables, and cognitive performance in a sample of community-dwelling older adults, 72þ years, residing in the state of São Paulo, Brazil.We hypothesized that lower neuroticism, lower DS, and higher PiL would be associated with better cognitive performance.

Participants and study design
The present study includes cross-sectional analyses of data from the second wave of the Frailty in Brazilian Older Adults (FIBRA) study, carried out between 2016 and 2017.The sample included older Brazilians aged 72þ years living in the city of Campinas and in the Ermelino Matarazzo district in the city of São Paulo, both located in the state of São Paulo.
The inclusion criteria for the follow-up participation were the same as for the baseline assessment: • understanding instructions for tasks, • accepting to participate in the study, and • permanent residence in the census tract.
The exclusion criteria were as follows: • impaired cognitive and communication skills, suggesting cognitive impairment; • permanent or temporary inability to walk; • aphasia or loss of muscle strength due to stroke sequelae; • severe impairment of speech and motricity related to Parkinson's disease in an advanced stage; • severe hearing and visual disability; and • terminal disease.The informants were the older people themselves or a family member residing in the household.For more details regarding the FIBRA study please see Neri et al. 14 Among the 1,284 participants at baseline, there were 192 deaths and 543 losses due to various reasons (participants not located [57.9%], refusals [3.5%], withdrawal of permission during data collection [5.5%], exclusion due to study criteria [1.6%], and presence of risks to interviewers [0.5%]). 14A total of 549 older adults underwent assessment of cognitive status, of which 419 scored above the education-adjusted cut-off score for dementia on the Mini-Mental State Examination (MMSE) and completed the follow-up protocol 14 which included the measures of interest for the present study (►Figure 1).The adopted cutoff scores for the MMSE were: 17 points for illiterates, 22 points for 1-4 years of formal study, 24 points for 5-8 years, and 26 points for those with 9 or more. 15he interviews were conducted by undergraduate and graduate students who were trained to apply the protocols.Recruitment was carried out in older adults' homes according to the list of addresses available in the baseline database.The evaluation was carried out at the residence by a pair of interviewers in one session, with an average duration of 75 minutes.
Participants or caregivers completed and signed an informed consent form.The project was approved by the ethical committee of the State University of Campinas, under the following protocol numbers -CAAE 49987615.3.0000.5404 and 92684517.5.1001.5404.

MMSE
This instrument assesses global cognition, with a maximum score of 30 points. 15

Mini-Addenbrooke's Cognitive Examination (M-ACE)
This is a brief cognitive screening instrument for dementia composed of temporal orientation questions, an Episodic Memory test (name and address), Verbal Fluency (VF)animal category, and the Clock Drawing Test (CDT). 16or the present analyses, the MMSE, and the M-ACE subitem scores for VF, CDT, and the Episodic Memory delayed recall were used to compose a composite score for cognition.The M-ACE total score was not used to avoid the overlap of the temporal orientation sub-items present in the M-ACE and the MMSE.

Neuroticism
It represents the domain of the Revised NEO Personality Inventory (NEO-PI-R).It assesses the characteristics of emotional instability and individual adjustment.This scale consisted of 12 items rated on a Likert scale.The higher the score, the higher the intensity of neuroticism. 6,17The FIBRA study did not collect data on other personality domains besides Neuroticism.

Geriatric Depression Scale (GDS)
This is a 15-item depression screening instrument for older adults. 18A score !6 is considered to be suggestive of major depression.

Purpose in Life (PiL)
This scale is part of the Psychological Well-Being scale developed by Ryff and Keyes. 10,19It examines the sense of meaning in life, beliefs that give meaning to life, and whether the person has goals for living.It consists of ten questions with responses on a Likert scale.The higher the score, the higher the purpose in life.

Statistical analyses
The variables of interest in the study do not have a normal distribution, which is why non-parametric data was used.Therefore, for descriptive analysis of the data, the Mann-Whitey (two categories) and Kruskall-Wallis (three categories or more) tests were used.Mann-Whitney test and the Kruskall-Wallis (with post hoc comparisons, Dwass-Steel-Critchlow-Fligner [DSCF]) to compare continuous variables and the Chi-square test to compare categorical variables.
The dependent variable was the cognitive composite score including the MMSE and the VF, CDT, and Episodic Memory items from the M-ACE.To create the cognitive composite score, the individual test scores were normalized into z-scores.The z-scores for the tests were added and divided by the number of items, to generate a global cognitive score.Considering that the present sample was composed only of participants with unimpaired cognition, according to the MMSE, the composite score was transformed into a dichotomous variable, z 0 (newZ ¼ 0) to represent lower cognitive performance and z > 0 (newZ ¼ 1) to represent higher cognitive performance.
Spearman's correlation tests and logistic regression analyses were used to test the associations among the selected variables.For the logistic regression analysis, the cognitive composite score was the dependent variable, and the sociodemographic and psychosocial variables were included as independent variables.The forced entry model was used to include all the independent variables and the cognitive composite score in the regression equation.
The level of statistical significance adopted for all analyses was p < 0.05.The statistical software Jamovi (Version 2.3.28,2022, Australia) and JASP (Version 0.17.2, 2023, The Netherlands) were used.

RESULTS
The sample was comprised of cognitively unimpaired older adults, mostly with low education.►Table 1 presents the sociodemographic, psychosocial, and cognitive characterization of the total sample and the sample divided by sex.Among the participants, 70% were women, with a high age (median ¼ 80 years) and low educational level (median ¼ 4 years), with 13% of participants without formal education.In addition, 18% of participants lived alone, with a greater number of women in this living condition.Also, 20% of participants had GDS scores suggestive of major depression and women had a higher level of neuroticism than men.
►Table 2 characterizes the sample divided by age groups.There were 185 individuals between 72 and 79 years of age (44%) and 234 older adults 80þ years.The younger group showed a lower score on Neuroticism (p ¼ 0.050).
►Table 3 demonstrates the differences related to educational level (illiterate, 1-4 years of study, 5-8, and 9 þ ).Among those with higher education (9þ years) income was higher and there was a higher percentage of people

PiL and education relate to better cognition
Oliveira et al.
living alone.The higher the level of education of the participants, the higher the cognitive scores.►Table 4 shows the Spearman ρ (rho) bivariate correlations between global cognition (composite score) and the psychosocial and sociodemographic variables.There was a significant correlation between global cognition and education, Neuroticism, PiL, and GDS scores.PiL was negatively correlated with GDS and Neuroticism.Still, Neuroticism was positively correlated with GDS scores; and global cognition and education were also positively correlated.
►Table 5 presents the results of the logistic regression analysis including the variables of interest (sociodemographic, psychosocial, and cognitive variables).This model did not show multilinearity (Tolerance > 0,1; VIF < 10) among the independent variables.The model indicated that education and PiL were significantly associated with global cognitive performance.

DISCUSSION
This study explored the relationships between psychosocial and sociodemographic variables with cognitive performance in adults 72þ years.Among the main results, in bivariate correlations, global cognition was significantly related to education, Neuroticism, GDS, and PiL.In the logistic regression, with psychosocial and sociodemographic variables included as independent variables, global cognition was significantly associated with PiL and number of years of education.
In the sample, a greater presence of women and older people (80þ years) was observed, similarly to what was found in the study by Freitas et al., 20 among other studies, suggesting greater participation of women in gerontological research.The sample consisted mostly of participants with a low educational level.In Brazil, this has been a common finding in community-based studies, as it reflects the   PiL and education relate to better cognition Oliveira et al.  historical reality of lower access to education in older generations. 21,22t was identified that 18% of participants lived alone, with a greater percentage of women in this living arrangement.In the National Health Survey in Brazil, carried out in 2013, it was observed that 15.3% of people aged 60 and over lived alone. 21The change from a cohabitation profile towards a greater number of single-person households in recent decades is related to several factors, such as the demographic and epidemiological transition, changes in family arrangements, shorter duration of marital unions, and fewer children or more childless couples. 23hen the psychosocial variables were compared across sex, it was observed that women had a higher level of Neuroticism.This finding has been commonly reported, although, it is poorly understood, especially among older adults. 24n the present study, participants with low education had worse cognitive performance, and global cognition and educational level were positively correlated, replicating the well-known negative impact of low education on cognition. 25High education is understood as an important marker of cognitive reserve, that is, brain characteristics that allow higher cognitive functioning than expected in the face of pathological processes or aging. 26Therefore, this finding was expected.
Elevated neuroticism has been associated with worse cognitive performance.For example, longitudinal analyses of the Health and Retirement Study (HRS) identified that among individuals with higher neuroticism, there was significantly lower performance in cognition. 7Another study reported that high neuroticism was associated with lower scores on episodic memory. 27In the present study, there was a significant negative correlation between global cognition and neuroticism.However, in the logistic regression, with the inclusion of other psychosocial and demographic variables, the association did not remain significant.
Similarly, cognition has been negatively associated with elevated DS in previous studies. 8Several studies have suggested the negative influence of depression on cognitive performance. 28,29Studies have shown deficits in memory and executive functions among people with major depression. 4,30Studies have also pointed out that higher neuroticism and DS could be associated with faster cognitive decline with age 31,32 and a higher incidence of dementia. 33However, Burhanullah et al. 9 found no association between depression and dementia.In the present analyses, GDS scores were negatively related to global cognition, yet this association was not significant in the multivariable model.
Differently from the other psychosocial variables in the present study, PiL was positively associated with global cognition in the bivariate correlation analyses and the association remained significant in the logistic regression, in the presence of other psychosocial and sociodemographic variables.This finding suggests that the influence of PiL on the cognition of older adults may be greater than the influence of DS and neuroticism.High PiL scores have previously been associated with better cognitive performance, in episodic memory and verbal fluency, and with a lower risk of cognitive decline. 34,35It is possible that older people with higher levels of PiL have a more positive view of aging and present more protective behaviors in relation to their health. 36In turn, such behaviors could be associated with greater cognitive reserve, such as involvement in intellectually stimulating activities.
Logistic regression results revealed that besides higher PiL, higher education was also significantly associated with better global cognition.These findings are in line with the Emory Healthy Aging Study, in which low PiL was identified as a robust predictor of cognitive decline, especially when associated with risk factors such as low education. 35dditionally, it was possible to observe, in the bivariate correlations, that PiL was negatively correlated with GDS and Neuroticism.These findings were similar to those found in previous studies.In Ribeiro et al., 37 high PiL was associated with a lower number of SD and in Kim et al. 38 to a reduced risk for developing major depression.Other authors postulated the idea that higher PiL could lead to greater adoption of healthy behaviors, 39 a more positive view of aging and life, greater resilience, optimism, a sense of self-efficacy, and better functioning of biological systems. 36Likewise, higher PiL would be related to low neuroticism, as it could enable emotional self-regulation in the management of adverse events 39 and also favor positive affects. 37ne of the limitations of this study is its cross-sectional design.Longitudinal surveys, with follow-up data over an extended period, could provide information on the trajectories of these variables and generate a greater understanding of such relationships.In addition, in the present study, cognition was investigated with the use of a composite score based on screening tests, which precluded the investigation of the association with specific cognitive domains.It is important to highlight that the validity and reliability of each measure included in the composite score were not assessed.Unfortunately, the FIBRA study did not include additional domains of the NEO-PI-R, such as conscientiousness, which would have been important to investigate.On the other hand, the present study fills an important gap, as the Brazilian gerontological literature lacks studies involving psychosocial variables and cognitive performance in cohorts of older adults.These data are relevant for cross-cultural comparisons.
In conclusion, the present study investigated the associations between psychosocial variables and cognitive performance among participants from the second wave of FIBRA study.Lower PiL and lower education levels were associated with worse cognition.The association between PiL and global cognitive scores is relevant as it suggests that interventions geared to increase PiL (e.g., Friedman et al. 40 ) might also have a positive effect on cognition.

Figure 1
Figure 1 Recruitment flowchart and composition of the follow-up sample.FIBRA Study, 2016-2017.

Table 5
Logistic regression including the study variables of interest with the cognitive composite score as the dependent variable (n ¼ 419).FIBRA Study, 2016-2017 Geriatric Depression Scale (total score); PiL ¼ Purpose in Life; OR ¼ odds ratio; CI ¼ confidence interval; p-value < 0.05 shows statistical significance.