The role of social and intellectual activity participation in older adults' cognitive function

(1)

Archives of Gerontology and Geriatrics 107 (2023) 104891

Available online 5 December 2022

The role of social and intellectual activity participation in older adults’

cognitive function

Irene Fern´andez

^a^,^*

, Adri´an García-Moll´a

^b

, Amparo Oliver

^a

, Noemí Sans´o

^c

, Jos´e M. Tom´as

^a

aDepartment of Methodology for the Behavioral Sciences, Faculty of Psychology, University of Valencia, Spain

bUniversity of Valencia, Valencia, Spain

cDepartment of Nursing and Physiotherapy, University of the Balearic Islands, Balearic Islands Health Research Institute (IDISBA), Spain

A R T I C L E I N F O Keywords:

Successful aging Social participation Cognitive reserve Cognitive function

A B S T R A C T

Theoretical background: A challenge of the ageing of the population is cognitive performance, given its association to optimal ageing. Documented predictors of cognition have included socio-demographics, education or physical factors. However, the association of social and intellectual activity participation to cognition has been less studied.

Aim: This study presents a predictive model of cognitive functioning including these alternative factors as well as more seminal ones to explain cognition in old age.

Materials and methods: The sample was composed by 45475 older adult participants in the 8th Wave of the Survey of Health, Aging and Retirement in Europe, that took place between 2019 and 2020. A correlational design was specified to test the effects of age, gender, years of education, physical inactivity, number of chronic diseases, social activity participation and intellectual activity participation on temporal orientation, numeracy, verbal fluency and memory. A completely a priori Structural Equation Model with latent variables was tested.

Results: The sample had an average of 70 years of age, was well-educated and physically active and engaged in reading. There was a higher proportion of females. The model showed an optimal fit to the data, explaining 8.7%-36.0% of the different cognitive components’ variance. Age, years of education and intellectual activity displayed the largest effects across the cognitive domains.

Conclusions: Findings suggest that social and intellectual activity participation are of relative importance to predict cognition in old age, even when considering other well-documented factors affecting older adults’

cognitive functioning.

1. Introduction

Aging of the population is an increasing phenomenon, as projections estimate the World’s older adult population will nearly double by 2050 (WHO, 2019). A study by Cao et al. (2020) examined life expectancy, healthy life expectancy (i.e. the amount of years lived in good health) and the gap between both during the period 1995-2017. Their results showed increases in all three terms, meaning that the increase in healthy life expectancy is not proportional to the increase in life expectancy, which implies that nowadays people live longer than before, but not in better health. Additionally, it is also well established that increments of life expectancy are associated to increased prevalence of cognitive decline (Lourenco et al., 2018) and cognitive impairment, in general, poses a burden on the quality of life of individuals who suffer from it

(Lawson et al., 2016). Moreover, successful aging is determined by high functioning in the physical, cognitive, psychological and social dimensions of the individual (Cheng, 2014).

Given the relevance of cognitive function in older age, the aim of this study is to present a predictive model of cognitive function, including social and intellectual activity participation as well as more seminal ones, namely age, gender, education and physical health. Limited attention has been paid to the relative impact of cognitive and social activity participation onto cognition, while also considering well- established physical predictors of cognition.

During the late nineties, engagement in social activities started attracting attention for the study of old people’s cognition. Since then, several correlational studies (Choi, 2020; Fu et al., 2018; Krueger et al., 2009; Litwin & Stoeckel, 2016) have found greater engagement in social

* Corresponding author.

E-mail address: [email protected] (I. Fern´andez).

Contents lists available at ScienceDirect

Archives of Gerontology and Geriatrics

journal homepage: www.elsevier.com/locate/archger

https://doi.org/10.1016/j.archger.2022.104891

Received 4 October 2022; Received in revised form 1 December 2022; Accepted 2 December 2022

(2)

activities to be related to improved cognitive scores. In the same vein, longitudinal studies (Glei et al., 2005; Hwanget al., 2018; Miceli et al.

2019; Seeman et al., 2011) have also found that participating in social activities is associated to higher levels of cognitive function. Participa- tion in social activities is understood as people’s engagement in social activities like religious/voluntary groups, social/sports clubs, neigh- borhood meetings and such (Bourassa et al., 2017; Miceli et al., 2019).

Evidence shows that social activity is related to cognitive function along the life cycle, and that this relationship is particularly evident for older adults (Seeman et al., 2011). Additionally, Hultsch et al. (1999) carried out a longitudinal study that found engagement in intellectually chal- lenging activities to be associated to cognitive maintenance over time, and maintenance of participation in these activities to indicate less changes in older adults’ cognition.

Within the theoretical framework of the cognitive enrichment hypothesis, some authors (Bourassa et al., 2017; Hertzog et al., 2008) have argued that engaging in social and intellectual activities contributes to the maintenance of cognitive skills in old age because these activities imply cognitive-demanding contexts that represent opportunities for cognitive enrichment. Evidence, hence, seems to suggest that participation in social and intellectual activities could be beneficial for older adults’ cognitive function. In this line, Li et al. (2020) showed more participation in social and intellectual activities to be associated with the most favourable cognitive function trajectory group. Based on the literature, the first hypothesis of this study is that social and intellectual activity participation will have a positive association to older adults’

cognitive function. Moreover, the study will further investigate whether the effect on cognitive function differs as a function of the type of activity and the cognitive component.

Some other factors studied in relation to cognitive function that will be considered in this study include age, gender, education and physical health. Increased age has been associated to worsened cognitive performance, while some evidence has also been found about lower levels of cognitive function in women than men (Bourassa et al., 2017; Lour- enco et al., 2018; Pan et al., 2020). Regarding education, previous literature points towards a positive relationship between the amount of years of education and older adults’ cognitive function (Cadar et al., 2017). Finally, physical health factors contemplated in the literature on older adults’ cognition are physical activity and the number of chronic diseases endured by an individual. The former has been positively associated to cognitive function in the old age (Buchman et al., 2012;

Tsai & Chang, 2019), while the later has been negatively related to cognitive functioning among older people (Gorelick et al., 2011; Lim et al., 2016; Lourenco et al., 2018). The second hypothesis of the study is that education will have a positive effect on older adults’ cognitive function. The final and third hypothesis is that age, physical inactivity, the number of chronic diseases and being female will show a negative effect on older people’s cognition.

2. Method

2.1. Sample and procedure

This study uses data from the 8th wave of the Survey of Health, Aging and Retirement in Europe (SHARE; Börsch-Supan, 2021; Börsch-Supan et al., 2013) which took place between October 2019 and August 2020, except for the period between March 2020 and May 2020 in which fieldwork was suspended due to the outbreak of the COVID-19 pandemic. The SHARE project is a longitudinal study whose target are representative adult populations aged 50 years or older from several European countries plus Israel. Partners of people aged 50 or older are also part of the target population of SHARE, regardless of age. The project employs a probabilistic and stratified sampling procedure with four sampling stages. Further details are offered in Malter and Börsch-Supan (2017). The Ethical Approval for gathering of the data used in this study was obtained by the SHARE project and it can be

publicly consulted at: http://www.share-project.org/fileadmin/pdf_

documentation/MPG_Ethics_Council_SHARE_overall_approval_29.05.20 20__en_.pdf.

A total of 46733 individuals participated in the 8th Wave of SHARE.

From them, we selected only those that were capable of answering the cognitive function tests. The final sample consisted of 45475 participants, aged between 32 and 103 years old, and their sociodemographic characteristics are presented in Table 1. A total of 27 countries were represented in this wave: Austria (3.3%), Germany (6.2%), Sweden (5.1%), Netherlands (4.2%), Spain (4.4%), Italy (4.5%), France (5.4%), Denmark (4.7%), Greece (6.4%), Switzerland (4.1%), Belgium (4.3%), Israel (1.9%), Czech Republic (5.8%), Poland (4.4%), Luxembourg (2.0%), Hungary (1.7%), Slovenia (5.2%), Estonia (6.4%), Croatia (2.6%), Lithuania (3.1%), Bulgaria (2.0%), Cyprus (1.1%), Finland (2.5%), Latvia (1.7%), Malta (1.7%), Romania (2.8%), and Slovakia (2.2%).

2.2. Instruments

Cognitive function was operationalized by means of temporal orientation, memory, numeracy and verbal fluency. More detailed information about the specific cognitive tests employed is available in Mehrbrodt et al. (2019).

- Temporal orientation was measured with four indicators reflecting the respondent’s orientation to month, year, day of week and date.

Each indicator was dichotomously measured as correct (1) or incorrect (0). Higher scores indicate better temporal orientation.

- Memory, as measured by the 10-word recall test (Harris & Dowson, 1982). Participants were read ten words and they were then asked to recall them at two time points: immediately after having the words read (recent recall) and after having completed the verbal fluency and numeracy tests (delayed recall). Answers are coded on a scale from 0 to 10, representing the number of correctly evoked words on each occasion.

- Verbal fluency is a measure of executive function (Dewey & Prince, 2005). It consisted of a simple test in which participants were asked to report all the animals they could think of in 60 s. The total amount of animals elicited by the participant was recorded.

- Numeracy is also a measure of executive function (Cragg & Gilmore, 2014; Dewey & Prince, 2005). In the study, it was measured with a sequence of five questions in which the participant was asked to subtract 7, starting at 100. The starting question was: “Now let’s try some subtraction of numbers. One hundred minus 7 equals what?”.

After the respondent’s answer, he or she was asked “And 7 from that?” four consecutive times. Answers were coded as correct (1) or incorrect (0).

Activity participation deemed activities the respondent had participated in during the last year. The following activities were considered:

to do voluntary/charity work, to attend educational/training courses, to attend sport/social/other clubs, to take part in a political/community- related organization, to read books/magazines/newspapers, to do

Table 1

Descriptive characteristics of the sample.

Characteristics Mean ± SD or n (%)

Age (years) 70.04 ± 9.32

Gender (women) 26264 (57.8)

Marital status

Married and living with spouse 26108 (57.4)

Widow/widower 6945 (15.3)

Divorced 3320 (7.3)

Other 2907 (6.4)

Missing 6195 (13.6)

Years of education 11.30 ± 4.13

(3)

word or number games, to play cards/chess/similar. To reduce the number of variables to be included in the model, a series of exploratory factor analyses were performed in Mplus 8.7 (Muth´en & Muth´en, 1998-2017). The best fitting exploratory analyses yielded two factors with excellent fit: RMSEA = .027 [90%CI .024 - .030], CFI = .993, and SRMR = .030. The first factor represented social activities, given that it summarised the scores of volunteerism, attending educational/training courses, joining sport/social/other clubs and taking part in a political/community-related organizations. The second factor was formed by activity participation related to intellectual activities, namely reading books/magazines/newspapers, doing word or number games and playing cards/chess/similar. Factor scores for the two dimensions were calculated and employed in subsequent models. These two sum- mary measures, social and intellectual activities, represented activity participation in the model.

Measures of physical health included respondents’ number of chronic diseases and a measure of physical inactivity. Number of chronic diseases was measured asking the respondents whether a doctor had ever told them they had any of the following conditions: heart attack, high blood pressure, high blood cholesterol, stroke, diabetes, chronic lung disease, cancer, stomach/duodenal ulcer, Parkinson’s disease, cataracts, hip/femoral fracture, other fractures, Alzheimer’s disease/

dementia, other affective/emotional disorders, rheumatoid arthritis, osteoarthritis or other rheumatism, and chronic kidney disease. Re- sponses were coded as 1 for “yes” and 0 for “no”. For this study, the sum of all chronic diseases was used. Physical inactivity was measured asking how frequently the respondent performed activities requiring a moderate level of energy such as gardening, cleaning the car, or going for a walk. It was measured in a 4-point Likert scale with 1 “More than once a week”, 2 “Once a week”, 3 “One to three times a month”, and 4 “Hardly ever, or never”.

Finally, measures regarding age, gender (0 = female, 1 = male) and years of education were also employed. SHARE data is collected with a computer-assisted personal interview (CAPI) by a trained interviewer.

Personal interviews at the respondents’ residence are carried out with the CAPI instrument installed on a laptop computer. Further information on SHARE data collection is available in B¨orsch-Supan and Jürges (2005).

2.3. Statistical analyses

First, descriptive statistics of the variables involved in the study were calculated. In order to establish a predictive model of cognitive functioning, as measured by temporal orientation, memory, verbal fluency and numeracy, a Structural Equation Model (SEM) was tested. This model postulated a series of antecedent variables (age, gender, years of education, physical activity, chronic diseases, social activity participation and intellectual activity participation) and tested their effect on cognitive function components. The proposed theoretical model can be consulted in Fig. 1.

The method of estimation was Weighted Least Squares Mean and Variance corrected (WLSMV), given that the data were not multivariate normal and some of the indicators or variables were ordinal (Finney &

DiStefano, 2006). Model fit was assessed with the most recommended indexes (Kline, 2016; Tanaka, 1993): the chi-square statistic (χ²), the Comparative Fit Index (CFI), the Root Mean Error of Approximation (RMSEA), and the Standardized Root Mean Residual (SRMR). Good fit of the model to the data is usually considered when the CFI is higher than .90, and the RMSEA and the SRMR are lower to 0.08, while excellent fit is considered with CFI over .95 and RMSEA/SRMR lower than .05 (Hu &

Bentler, 1999). To assess the predictive capacity of our model, Cohen’s (2013) f²effect size was employed. This effect size measure is based in R-square and is considered to represent a large effect with a value equal to or higher than .35, a medium effect with a value equal to or higher than .15 and a small effect with a value equal to or higher than .02.

Analyses were performed using Mplus 8.6 (Muth´en & Muth´en, 1998-2017).

3. Results

Descriptive statistics of the variables considered in the present study are displayed in Table 2 for the general sample. Additional descriptive statistics of these variables per country is available in the Supplemental Material.

The theoretical model proposed in Fig. 1 was estimated and tested.

Overall fit was very good: χ²(44) = 1939.63, p < .001, CFI = .985, RMSEA = .031 90% CI [.030 - .032], SRMR = .030. Correlations among

Fig. 1. Theoretical a priori model. Correlations among exogenous variables not shown for clarity.

(4)

exogenous variables are reported in Table 3. Standardized coefficients, standard errors and associated p-values are shown in Table 4 and correlations among dependent variables are shown in Table 5.

Regarding the predictors of temporal orientation in the old age, all of them were statistically significant, except gender. Age, physical inactivity and number of chronic diseases were negatively related to temporal orientation, while years of education, social activity and intellectual activity were positively related. Considering the size of the effects, the most relevant factor explaining older adults’ temporal

orientation was age (β = -.340), followed by intellectual activity (β = .167). The power of years of education’s (β = .089) and social activity’s (β = .078) were similar for explaining temporal orientation’s variance.

Physical health factors contemplated in the model presented medium to small effects for both statistically significant predictors (β = -.119 for physical inactivity and β = -.025 for number of chronic diseases).

On its part, all the predictors contemplated in the model had a statistically significant relation to memory, with age, being male, physical inactivity and number of chronic diseases, being negatively related to memory. Regarding the relative impact of the predictors, age (β = -.323), intellectual activity (β = .210) and years of education (β = .190) had the largest significant effects, while small effects were found for number of chronic diseases (β = -.065), physical inactivity (β = -.077) and gender (β = -.086). Social activity (β = .134) had a moderate relative contribution in explaining memory’s variance.

Concerning numeracy, all predictors showed a statistically significant relationship. Age, physical inactivity and chronic diseases showed a negative effect, while the remaining effects were positive. The main predictors of numeracy were years of education (β = .132) and intellectual activity (β = .140). The rest of the predictors had very small Table 2

Descriptive statistics of the variables under study.

Characteristics Mean ± SD or n (%)

Age (years) 70.04 ± 9.32

Gender (women) 26264 (57.8%)

Years of education 11.30 ± 4.13

Frequency of moderate physical activity

More than once a week 29302 (64.4%)

Once a week 6728 (14.8%)

One to three times a month 3279 (7.2%)

Hardly ever, or never 6149 (13.5%)

Missing 17 (< 0.1%)

Number of chronic diseases 1.91 ± 1.61

Social activity participation

Volunteerism 7301 (16.1%)

Educational/training course 4625 (10.2%)

Sport/social/other clubs 12095 (26.6%)

Political/community organizations 2736 (6.0%)

Intellectual activity participation

Read books/magazines/newspapers 32648 (71.8%)

Number games 19605 (43.1%)

Play cards/chess/similar 12621 (27.8%)

Temporal orientation (correct)

To day of month 40754 (89.6%)

To month 44665 (98.2)

To year 44361 (97.6%)

To day of the week 44149 (97.3%)

Numeracy (correct)

Subtraction 1 42608 (93.7%)

Verbal fluency 20.25 ± 7.74

Memory

Recent recall 5.26 ± 1.75

Delayed recall 3.86 ± 2.14

Table 3

Correlation coefficients and standard errors among exogenous variables. All correlations are statistically significant (p < .001). Note: gender coded 0 = fe- male, 1 = male.

(1) (2) (3) (4) (5) (6)

(1) Age 1

(2) Gender Coefficient .030 1 Standard

error .005

(3) Years of

education Coefficient -.209 .068 1 Standard

error .005 .005

(4) Physical

inactivity Coefficient .209 -.038 -.139 1 Standard

error .004 .005 .005

(5) Number of chronic diseases

Coefficient .275 -.046 -.126 .189 1 Standard

error .004 .005 .005 .004

(6) Social

activity Coefficient -.115 .018 .248 -.225 -.118 1 Standard

error .005 .005 .004 .006 .005

(7) Intellectual

activity Coefficient -.053 -.073 .267 -.248 -.042 .303 Standard

error .005 .005 .005 .005 .005 .004

Note: Gender (men coded 1).

Table 4

Standardised coefficients, standard errors and p-values of the tested model.

Note: gender coded 0 = female, 1 = male.

Standardised

coefficient Standard error p- value Temporal orientation

Age -.340 .008 <.001

Gender -.010 .010 .289

Years of education .089 .010 <.001

Physical inactivity -.119 .008 <.001

Number of chronic diseases -.025 .010 .004

Social activity .078 .010 <.001

Intellectual activity .167 .009 <.001

Numeracy

Age -.089 .005 <.001

Gender .066 .005 <.001

Number of chronic diseases -.036 .005 <.001

Verbal fluency

Age -.212 .003 <.001

Gender .015 .004 <.001

Memory

Age -.323 .004 <.001

Gender -.086 .004 <.001

Table 5

Correlation coefficients and standard errors among dependent variables. All correlations are statistically significant (p < .001).

(1) (2) (3)

(1) Temporal orientation 1

(2) Numeracy Coefficient .241 1

Standard error .010

(3) Verbal fluency Coefficient .433 .149 1

Standard error .007 .004

(4) Memory Coefficient .521 .240 .406

Standard error .007 .005 .003

(5)

effects.

Finally, all predictors displayed statistically significant effects on verbal fluency. Age, physical inactivity and number of chronic diseases were negatively related to verbal fluency, while being a male, years of education and participation in social and intellectual activities were positively associated to verbal fluency. The largest estimated effects explaining verbal fluency’s variance were age (β = -.212), years of education (β = .171), intellectual activity (β = .247) social activity (β = .132) and physical inactivity (β = -.108). On the opposite, gender (β = .015) and number of chronic diseases (β = -.021), albeit statistically significant, represented the smallest effects.

Overall, the model adequately fitted the data, and significantly predicted the different components of older adults’ cognitive function, being verbal fluency and memory the components that were better explained (28.5% of the variance of verbal fluency and 36.0% of the variance of memory) by the predictors considered in the model, followed by temporal orientation (25.8% of explained variance) and by numeracy (8.7% of explained variance). Regarding the effect sizes, effects of predictors on memory, verbal fluency and temporal orientation were large (f²=.56 for memory, f²=.40 for verbal fluency and f²=.35 for temporal orientation), while these effects were small in size in the case of numeracy (f²=.09).

4. Discussion

Several factors have been related to older adults’ cognition, both from a cross-sectional and a longitudinal perspective. Be that as it may, to the best of our knowledge, there was no evidence comparing the effects of social and intellectual activity participation simultaneously on different components of cognition. The study by Krueger et al. (2009) did test the effects of social and intellectual activities onto cognition, but cognition was considered as a global component. However, evidence by other studies suggested that there is heterogeneity in the sizes of the effects of a single predictor onto distinct components of cognitive function (Bourassa et al., 2017; Litwin & Stoeckel, 2016).

The present study proposed a predictive model of older adults’

cognitive function, differentiating among memory, temporal orientation, and executive function, to test the effects of social and intellectual activity participation, controlling for other well-established predictors of cognitive function, namely age, gender, education and physical health. The integration of these factors constitutes the major contribution of this study, while differentiating among different cognitive domains rather than using a measure of global cognition further contributes to the study of the heterogeneity of each predictor’s effect size onto distinct components of cognitive function.

Social activity participation had been previously associated to better cognitive performance (Choi, 2020; Fu et al., 2018; Glei et al., 2005;

Hwang et al., 2018; Krueger et al., 2009; Litwin & Stoeckel, 2016;

Miceli et al. 2019; Seeman et al., 2011). Results in this study showed that, when controlling for intellectual activity and other predictors, social activity had a moderate effect in explaining memory and verbal fluency. Further, statistically significant effects of social activities on temporal orientation and numeracy were also found, but the magnitude of these effects was relatively small compared to that of other predictors.

Hence, it seems that social activity is positively associated to older adults’ cognitive function, but the degree of association differs among the different components of cognition.

Compared to previous literature, present findings generally revealed diminished effects of social activity participation on cognitive domains.

Previous literature employed either a measure of global cognition (Choi, 2020; Fu et al, 2018; Hwang et al., 2018; Krueger et al., 2009;

Miceli et al., 2019) or memory (Litwin & Stoeckel, 2016), instead of differentiating among different cognitive domains, with the exception of the study by Seeman et al. (2011) that considered executive function and episodic memory separately. Present results are in line with those reported by Seeman et al. (2011) in regard to the relative lower effect of

social engagement to executive function compared to memory, but these authors employed other measurement instruments. Therefore, the methods employed to operationalize cognition across studies does not allow for a clear comparison of specific effects.

One possible explanation for the differential effects of social activity onto the different domains of cognition could be that social activities do not occur in situations where temporal orientation and numeracy are specially needed. Therefore, according to cognitive enrichment hypothesis (Hertzog et al., 2008), social activities would imply cognitive-demanding contexts but such contexts would be more demanding of other cognitive domains, such as verbal fluency and memory.

Intellectual activity showed meaningful predictive power for temporal orientation, numeracy, verbal fluency and memory, with a moderate to high substantive relative contribution to each of the cognitive components’ explained variance. This result is in line with the study by Hultsch et al. (1999) and offers promising perspectives regarding older adults’ cognition preservation. For example, one way to intervene in cognitive function maintenance could be to implement informal intellectual activities such as the ones contemplated in this study. However, as pointed out by Hultsch et al. (1999), it is also possible that only people with good levels of cognition decide to participate in such activities, given that the cross-sectional nature of the data does not allow to establish causal relationships. Among other previous studies, only that of Krueger et al. (2009) considered cognitive activities separately from social ones. However, these authors included cognitive activity participation as a control variable and did not report the specific effect. They did report, however, a decrease of social activity participation’s effect onto global cognition when including cognitive and physical activity participation as covariates.

Finally, the effects of the rest of the factors contemplated in the model were generally as reported in the literature. Notwithstanding, physical predictors showed somewhat unexpected results, as their effects were much smaller than those previously reported (for example:

Buchman et al., 2012; Gorelick et al., 2011; Lim et al., 2016; Lourenco et al., 2018; Tsai & Chang, 2019). Previous literature did not consider the effects of physical health factors together with those of social and intellectual activity participation. Therefore, it is possible that when examining all these effects simultaneously, the relative impact of physical health is diminished compared to that of activity participation.

Another contribution of this study is that our model provided evidence on how variables differently predict distinct components of cognitive function. For example, physical health factors showed slightly higher effects in temporal orientation and verbal fluency, while they virtually had no explicative capacity over numeracy. Activity participation, both social and intellectual, had a considerable effect on memory and verbal fluency. It is also worth mentioning that in the case of numeracy and verbal fluency, intellectual activity presented the biggest predictive effect. Executive function has been strongly associated to fluid intelligence (Van Aken et al., 2016), which is the ability to adapt and be flexible in cognitive-demanding contexts (Cattell, 1963).

Therefore, intellectual activity participation could be serving as training of such fluid abilities, in line with the cognitive enrichment hypothesis (Hertzog et al., 2008).

Although the model presents relevant contributions and was established using a representative sample of older adults, which allows for generalization of results to the older adult population, this study also presents limitations. The first one refers to the measures employed in the study. Data used in this study was extracted from a large-scale survey tapping different topics (i.e. SHARE) and therein assessment had to be carried out with short and coarse measures. Accordingly, social and intellectual activities were measured in terms of occurrence regardless of its nature. Future studies examining the effect of activities’ frequency would contribute to tighten the gap of knowledge regarding the effects of these predictors on older adults’ cognition. Moreover, the study design was cross-sectional, which entails that causal attributions cannot

(6)

be made. Hence, reverse causality of the effects is possible and the authors encourage future research to provide causality of the relationships found in this study from a longitudinal perspective. The directionality of the effects examined in the study was hypothesized from previous longitudinal evidence of these effects. Finally, data recollection took place between October 2019 and August 2020 but had to be interrupted between March 2020 and May 2020 because of the outbreak of the COVID- 19 pandemic. Individuals who were assessed after the outbreak may differ to those who were assessed before it. Research on the impact of the COVID-19 pandemic show declines in activity levels (De Pue et al., 2021;

Lebrasseur et al., 2021) as well as in cognition (Tondo et al., 2021;

Amieva et al., 2022), as consequences of, among other factors, the restrictions and safety measures against COVID-19. Therefore, although the relationships reported in the present study are bolstered by previous evidence, they could still be confounded by the occurrence of the pandemic and additional studies are needed to ensure these findings.

5. Conclusions

This study found participation in activities to be significantly associated to different components of cognition. Intellectual activity was one of the most important predictors of all four considered components of cognition (i.e. temporal orientation, numeracy, memory and verbal fluency), although effect sizes were bigger for verbal fluency and memory. Social activity, in turn, significantly predicted all cognitive dimensions but its effects were smaller. All in all, results highlight the relevance of social and intellectual activity participation for cognition, even after controlling for the effects of other well-established associated factors of cognition such as physical activity or years of education.

Further study is needed, however, to analyse these associations longi- tudinally, as to provide insight about the directionality of the effects.

Funding statement

Irene Fern´andez is the recipient of grant PRE2019-089021 funded by MCIN/AEI/10.13039/501100011033 and by “ESF Investing in your future”. This research is supported by project PID2021-124418OB-I00 funded by MCIN/AEI/10.13039/501100011033 and by “ERDF A way of making Europe”.

Availability of data and material

The data that support the findings of this study are available from the SHARE Project but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of the SHARE Project.

CRediT authorship contribution statement

Irene Fernández: Conceptualization, Methodology, Software, Formal analysis, Writing – original draft. Adrián García-Mollá:

Conceptualization, Writing – original draft, Data curation. Amparo Oliver: Writing – review & editing, Supervision, Funding acquisition.

Noemí Sans´o: Conceptualization, Writing – review & editing. Jos´e M.

Tom´as: Methodology, Formal analysis, Writing – review & editing, Project administration, Funding acquisition.

Declaration of Competing Interest None.

Acknowledgements

The SHARE data collection has been primarily funded by the Euro- pean Commission through FP5 (QLK6-CT-2001-00360), FP6 (SHARE-I3:

RII-CT-2006-062193, COMPARE: CIT5-CT-2005-028857, SHARELIFE:

CIT4-CT-2006-028812) and FP7 (SHARE-PREP: N^◦211909, SHARE- LEAP: N^◦227822, SHARE M4: N^◦261982). Additional funding from the German Ministry of Education and Research, the Max Planck Society for the Advancement of Science, the U.S. National Institute on Aging (U01_AG09740-13S2, P01_AG005842, P01_AG08291, P30_AG12815, R21_AG025169, Y1-AG-4553-01, IAG_BSR06-11, OGHA_04-064, HHSN271201300071C) and from various national funding sources is gratefully acknowledged (see www.share-project.org).

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.archger.2022.104891.

References

Amieva, H., Retuerto, N., Hernandez-Ruiz, V., Meillon, C., Dartigues, J. F., & P´er`es, K.

(2022). Longitudinal study of cognitive decline before and after the COVID-19 pandemic: Evidence from the PA-COVID survey. Dementia and Geriatric Cognitive Disorders, 51(1), 56–62. https://doi.org/10.1159/000521999

B¨orsch-Supan, A. (2021). Survey of health, ageing and retirement in Europe (SHARE) Wave 8. Release version: 1.0.0. SHARE-ERIC. Data set. https://doi.org/10.6103/

SHARE.w8.100

B¨orsch-Supan, A., & Jürges, H. (2005). The survey of health, aging and retirement in Europe – Methodology. Mannheim Research Institute for the Economics of Aging (MEA.

B¨orsch-Supan, A., Brandt, M., Hunkler, C., Kneip, T., Korbmacher, J., Malter, F., Schaan, B., Stuck, S., & Zuber, S. (2013). Data resource profile: The survey of health, ageing and retirement in Europe (SHARE). International Journal of Epidemiology, 42, 992–1001. https://doi.org/10.1093/ije/dyt088

Bourassa, K. J., Memel, M., Woolverton, C., & Sbarra, D. A. (2017). Social participation predicts cognitive functioning in aging adults over time: Comparisons with physical health, depression, and physical activity. Aging & Mental Health, 21, 133–146.

https://doi.org/10.1080/13607863.2015.1081152

Buchman, A. S., Boyle, P. A., Yu, L., Shah, R. C., Wilson, R. S., & Bennett, D. A. (2012).

Total daily physical activity and the risk of AD and cognitive decline in older adults.

Neurology, 78, 1323–1329. https://doi.org/10.1212/wnl.0b013e3182535d35 Cadar, D., Robitaille, A., Clouston, S., Hofer, S. M., Piccinin, A. M., & Muniz-Terrera, G.

(2017). An International evaluation of cognitive reserve and memory changes in early old age in 10 European countries. Neuroepidemiology, 48, 9–20. https://doi.

org/10.1159/000452276

Cao, X., Hou, Y., Zhang, X., Xu, C., Jia, P., Sun, X., Sun, L., Gao, Y., Yang, H., Cui, Z., Wang, Y., & Wang, Y. (2020). A comparative, correlate analysis and projection of global and regional life expectancy, healthy life expectancy, and their GAP:

1995–2025. Journal of Global Health, 10(2). https://doi.org/10.7189/

jogh.10.020407

Cattell, R. B. (1963). Theory of fluid and crystallized intelligence: A critical experiment.

Journal of Educational Psychology, 54, 1–22. https://doi.org/10.1037/h0046743 Cheng, S. T. (2014). Defining successful aging: The need to distinguish pathways from

outcomes. International Psychogeriatrics, 26(4), 527–531. https://doi.org/10.1017/

S1041610213001713

Choi, M. (2020). Association between social participation and cognitive function among community-dwelling older adults living alone: Analysis of a nationally

representative survey. International Journal of Nursing Practice, 26, e12841. https://

doi.org/10.1111/ijn.12841

Cohen, J. (2013). Statistical power analysis for the behavioral sciences. Academic press.

Cragg, L., & Gilmore, C. (2014). Skills underlying mathematics: The role of executive function in the development of mathematics proficiency. Trends in Neuroscience and Education, 3, 63–68. https://doi.org/10.1016/j.tine.2013.12.001

De Pue, S., Gillebert, C., Dierckx, E., Vanderhasselt, M. A., De Raedt, R., & Van den Bussche, E. (2021). The impact of the COVID-19 pandemic on wellbeing and cognitive functioning of older adults. Scientific Reports, 11(1), 1–11. https://doi.org/

10.1038/s41598-021-84127-7

Dewey, M. E., & Prince, M. J. (2005). Cognitive function. In A. B¨orsch-Supan, A. Brugiavini, H. Jürges, J. Mackenbach, J. Siegrist, & G. Weber (Eds.), Health, ageing and retirement in Europe – First results from the Survey of Health, Ageing and Retirement in Europe (pp. 118–125). Mannheim Research Institute for the Economics of Aging (MEA).

Finney, S. J., & DiStefano, C. (2006). Non-normal and categorical data in structural equation modeling. In G. R. Hancock, & R. O. Mueller (Eds.), Structural equation modeling: A second course (pp. 269–314). Greenwich, CT: Information Age.

Fu, C., Li, Z., & Mao, Z. (2018). Association between social activities and cognitive function among the elderly in China: A cross-sectional study. International Journal of Environmental Research and Public Health, 15, 231. https://doi.org/10.3390/

ijerph15020231

Glei, D., Landau, D., Goldman, N., Chuang, Y., Rodríguez, G., & Weinstein, M. (2005).

Participating in social activities helps preserve cognitive function: An analysis of a longitudinal, population-based study of the elderly. International Journal of Epidemiology, 34(4), 864–871. https://doi.org/10.1093/ije/dyi049

Gorelick, P. B., Scuteri, A., Black, S. E., DeCarli, C., Greenberg, S. M., Iadecola, C., Launer, L. J., Laurent, S., Lopez, O. L., Nyenhuis, D., Petersen, R. C., Schneider, J. A.,

(7)

Tzourio, C., Arnett, D. K., Bennett, D. A., Chui, H. C., Higashida, R. T., Lindquist, R., Nilsson, P. M., … Seshadri, S. (2011). Vascular contributions to cognitive impairment and dementia: A statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke, 42, 2672–2713.

https://doi.org/10.1161/str.0b013e3182299496

Harris, S. J., & Dowson, J. H. (1982). Recall of a 10-word list in the assessment of dementia in the elderly. The British Journal of Psychiatry, 141, 524–527. https://doi.

org/10.1192/bjp.141.5.524

Hertzog, C., Kramer, A., Wilson, R., & Lindenberger, U. (2008). Enrichment effects on adult cognitive development. Psychological Science in the Public Interest, 9, 1–65.

https://doi.org/10.1111/j.1539-6053.2009.01034.x

Hu, L., & Bentler, P. M. (1999). Cut-off criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. https://doi.org/10.1080/10705519909540118

Hultsch, D. F., Hertzog, C., Small, B. J., & Dixon, R. A. (1999). Use it or lose it: Engaged lifestyle as a buffer of cognitive decline in aging? Psychology and Aging, 14, 245–263.

Hwang, J., Park, S., & Kim, S. (2018). Effects of participation in social activities on cognitive function among middle-aged and older adults in Korea. International Journal of Environmental Research and Public Health, 15, 2315. https://doi.org/

10.3390/ijerph15102315

Kline, R. B. (2016). Principles and practice of structural equation modelling. Guilford Press.

Krueger, K., Wilson, R., Kamenetsky, J., Barnes, L., Bienias, J., & Bennett, D. (2009).

Social engagement and cognitive function in old age. Experimental Aging Research, 35, 45–60. https://doi.org/10.1080/03610730802545028

Lawson, R. A., Yarnall, A. J., Duncan, G. W., Breen, D. P., Khoo, T. K., Williams- Gray, C. H., Barker, R. A., Collerton, D., Taylor, J. P., & Burn, D. J. (2016). Cognitive decline and quality of life in incident Parkinson’s disease: The role of attention.

Parkinsonism and Related Disorders, 27, 47–53. https://doi.org/10.1016/j.

parkreldis.2016.04.009

Lebrasseur, A., Fortin-Bédard, N., Lettre, J., Raymond, E., Bussières, E. L., Lapierre, N., Faieta, J., Vincent, C., Duchesne, L., Ouellet, M. C., Gagnon, E., Tourigny, A., Lamontagne, M.È., & Routhier, F. (2021). Impact of the COVID-19 pandemic on older adults: Rapid review. JMIR Aging, (2), 4. https://doi.org/10.2196/26474 Li, H., Li, C., Wang, A., Qi, Y., Feng, W., Hou, C., Tao, L., Liu, X., Li, X., Wang, W.,

Zheng, D., & Guo, X. (2020). Associations between social and intellectual activities with cognitive trajectories in Chinese middle-aged and older adults: A nationally representative cohort study. Alzheimer’s Research & Therapy, 12(1). https://doi.org/

10.1186/s13195-020-00691-6

Lim, S. L., Gao, Q., Nyunt, M. S. Z., Gong, L., Lunaria, J. B., Lim, M. L., Ling, A., Lam, C. S- P., Richards, A. M., Ling, L. H., & Ng, T. P. (2016). Vascular health indices and cognitive domain function: Singapore longitudinal ageing studies. Journal of Alzheimer’s Disease, 50, 27–40. https://doi.org/10.3233/jad-150516

Litwin, H., & Stoeckel, K. J. (2016). Social network, activity participation, and cognition.

Research on Aging, 38, 76–97. https://doi.org/10.1177/0164027515581422 Lourenco, J., Serrano, A., Santos-Silva, A., Gomes, M., Afonso, C., Freitas, P., Paul, C., &

Costa, E. (2018). Cardiovascular risk factors are correlated with low cognitive

function among older adults across Europe based on the SHARE database. Aging and Disease, 9, 90–101. https://doi.org/10.14336/ad.2017.0128

Malter, F., & B¨orsch-Supan, A. (2017). SHARE Wave 6: Panel innovations and collecting Dried Blood Spots. Munich Center for the Economics of Aging (MEA).

Mehrbrodt, T., Gruber, S., & Wagner, M.. Scales and multi-item indicators [Online document]. http://www.share-project.org/fileadmin/pdf_documentation/SHARE_Sc ales_and_Multi-Item_Indicators.pdf.

Miceli, S., Maniscalco, L., & Matranga, D. (2019). Social networks and social activities promote cognitive functioning in both concurrent and prospective time: Evidence from the SHARE survey. European Journal of Ageing, 16, 145–154. https://doi.org/

10.1007/s10433-018-0486-z

Muthén, L. K., & Muthén, B. O. (1998–2017). Mplus User’s Guide (8th Ed.). Muthén &

Muth´en.

Pan, X., Zhang, D., & Shi, L. (2020). Longitudinal evidence on grandparenting and cognitive function in older adults in China. Journal of Intergenerational Relationships.

https://doi.org/10.1080/15350770.2020.1742845

Seeman, T. E., Miller-Martinez, D. M., Merkin, S. S., Lachman, M. E., Tun, P. A., &

Karlamangla, A. S. (2011). Histories of social engagement and adult cognition:

Midlife in the U.S. Study. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 66, 141–152. https://doi.org/10.1093/geronb/gbq091 Tanaka, J. S. (1993). Multifaceted conceptions of fit in structural equation models. In

K. A. Bollen, & J. S. Long (Eds.), Testing structural equation models (pp. 10–39). Sage.

Tondo, G., Sarasso, B., Serra, P., Tesser, F., & Comi, C. (2021). The impact of the COVID- 19 pandemic on the cognition of people with dementia. International Journal of Environmental Research and Public Health, (8), 18. https://doi.org/10.3390/

ijerph18084285

Tsai, H., & Chang, F. (2019). Associations of exercise, nutritional status, and smoking with cognitive decline among older adults in Taiwan: Results of a longitudinal population-based study. Archives of Gerontology and Geriatrics, 82, 133–138. https://

doi.org/10.1016/j.archger.2018.12.008

Van Aken, L., Kessels, R. P. C., Wingbermühle, E., Van Der Veld, W. M., & Egger, J. I. M.

(2016). Fluid intelligence and executive functioning more alike than different? Acta Neuropsychiatrica, 28, 31–37. https://doi.org/10.1017/neu.2015.46

WHO. (2019). World population ageing 2019. Highlights. United Nations: Department of Economic and Social Affairs (ST/ESA/SER.A/430).

Further readings

Christensen, H., Korten, A., Jorm, A. F., Henderson, A. S., Scott, R., & Mackinnon, A. J.

(1996). Activity levels and cognitive functioning in an elderly community sample.

Age and Aging, 25, 72–80.

Singh-Manoux, A., Richards, M., & Marmot, M. (2003). Leisure activities and cognitive function in middle age: Evidence from the Whitehall II study. Journal of Epidemiological Community Health, 57, 907–913.