Social determinants of health and cognitive function: A cross‐sectional study among men without dementia

Abstract Background Certain age‐related and medical factors have been associated with cognitive dysfunction; however, less is known regarding social determinants of health. The current study aimed to investigate associations between social determinants of health and cognitive function in a population‐based sample of men without dementia. Methods Data were drawn from the ongoing Geelong Osteoporosis Study (n = 536). Cognitive function was determined using the Cog‐State Brief Battery. Area‐based socioeconomic status (SES) was determined using the Index of Relative Socioeconomic Advantage and Disadvantage, marital status by self‐report, and social support by the Multidimensional Scale of Perceived Social Support, which considers family, friends, and significant others. Results Belonging to a higher SES group, being in a relationship (married/de‐facto) and perceived social support from a significant other and friends were each associated with better overall cognitive function. In regard to the specific cognitive domains, higher SES was associated with better psychomotor function and visual learning, being in a relationship was associated with better working memory, and perceived social support from a significant other was associated with better attention and working memory, with perceived social support from friends associated with better psychomotor function. There were no associations detected between social support from family and any of the cognitive domains. Conclusion Higher SES, being in a relationship, and greater perceived social support from a significant other and friends were associated with better cognitive function. Further studies identifying underlying mechanisms linking social factors with cognition are needed to establish prevention strategies and enhance cognitive health.


BACKGROUND
Over the past century, lifespan has increased dramatically.Consequently, there is a rising prevalence of age-related cognitive decline, such as mild cognitive impairment and dementia (WHO, 2023).Mild cognitive impairment is an intermediate phase between normal ageing and dementia, characterized by a modest decline in cognition greater than expected for an individual's age and education (Albert et al., 2013;Lydon et al., 2022;Petersen et al., 2001).Mild cognitive impairment affects approximately 17% of people over the age of 60 years, with prevalence increasing with increasing age (Petersen et al., 2018).Individuals with mild cognitive impairment are more likely to develop progressive cognitive decline, with an annual conversion rate to dementia of 10%-15% (Petersen et al., 2001(Petersen et al., , 2018;;Petersen, 2011;Petersen, 2016).Dementia is defined as loss of cognitive abilities, being an umbrella term for a number of neurological conditions, with the most common being Alzheimer's disease.Worldwide, ∼55 million people have dementia, with nearly 10 million new cases each year (WHO, 2023).The total number of people with dementia and cognitive decline is projected to reach 78 million by 2030 and 139 million by 2050, with two-thirds of those affected living in low-income and middle-income countries (WHO, 2023).
With these rapidly increasing numbers, there has been a significant focus on modifiable risk and prevention factors related to poor cognitive outcomes (Baumgart et al., 2015).2015).Social determinants of health constitute a growing area of interest, referring to socioeconomic factors such as income, wealth and education, marital status, and social supports, such as friends and family (Braveman & Gottlieb, 2014).
Interactions with social networks, utilizing different skills and participating in a range of tasks with others, have been suggested to foster greater cognitive reserve and increase cognitive function (Nie et al., 2021;Stern, 2009).For example, a study by Bennet et al. (2006) in the United States reported that older adults with Alzheimer's disease who have a larger circle of social connectedness sustained higher working and semantic memory.However, investigations into the role of social factors in relation to overall cognitive function and/or specific cognitive domains including psychomotor function, attention, working memory, and visual learning are limited, with previous studies focusing on older adults with a neurodegenerative disorder.
Thus, we aimed to investigate the relationship between socioeconomic status (SES), relationship status, social support, and their association with both overall and specific domains of cognitive function in a population-based sample of men without dementia.

Participants
Data were collected from an age-stratified, population-based sample of men enrolled in the Geelong Osteoporosis Study (GOS) (Bennett et al., 2006).

Cognitive function
Cognitive status was determined using the Cog-State brief battery (CBB), a computer-based neuropsychology battery for assessing specific cognitive domains for use in epidemiological studies and clinical trials (Fredrickson et al., 2010).Participants were briefed and introduced to the battery and underwent a practice trial prior to each task.The CBB has been previously validated for use in healthy populations and those with mild cognitive impairment and early dementia (Darby et al., 2002;Dingwall et al., 2009;Falleti et al., 2006;Fredrickson et al., 2010;Maruff et al., 2004).Previous studies have also reported on the efficiency, acceptability and test-retest reliability of the CBB (Fredrickson et al., 2010), deeming the CBB a good candidate for research participants with limited familiarity with computers (Fredrickson et al., 2010).The CBB comprises four computerized cognitive tasks, requiring 10-12 min for administration, and consists of (1) a simple reaction time task assessing psychomotor function (detection task [DET]).Psychomotor function refers to processing speed skills and involves the length of time it takes to process information and formulate a response (Anna-Karin et al., 2014).( 2) a choice reaction time task assessing attention (identification task [IDN]).Attention is the ability to focus and concentrate on relevant information while excluding other details (Hennawy et al., 2019).
(3) a one-back task assessing working memory (one-back task [OBK]).Working memory is the ability to hold information in the mind while using that information (Verhaeghen et al., 2003).(4) a continuous recognition task assessing visual learning (onecard learning task [OCL]).Visual learning is the ability to understand and work with visual information (Pal et al., 2016).Reaction time was the primary outcome measure for DET, IDN, and OBK and was calculated according to the speed of performance (log10 million seconds [lmn]), with lower scores indicating greater performance.Reaction time accuracy was the primary outcome measure for OCL and was calculated according to the accuracy of performance and reaction time, with higher scores indicating greater performance.Overall cognitive function (OCF) was generated by combining the four cognitive domains; higher scores indicated better performance, as previously described (Sui et al., 2020).Marital status: Current marital status was determined by self-report and classified into four categories: (1) single, ( 2) being in a relationship (married or de-facto), (3) divorced or separated from marriage, or (4) widowed.

Exposure
Perceived social support: Social support was measured using the Multidimensional Scale of Perceived Social Support (MSPSS) (Zimet et al., 1988).The MSPSS is a self-report scale measuring an individual's perception of their social support.It contains 12 items, four items for each subscale (friends, family, and significant other).There are seven possible responses to each statement scored zero to six, totaling to a maximum score of 72, with higher scores indicating greater perceived social support (Zimet et al., 1988).The MSPSS has been found to be a reliable and valid tool within different age groups and cultural backgrounds, with good to excellent internal consistency and test-retest reliability (Bruwer et al., 2008;Clara et al., 2003;Pedersen et al., 2009;Zimet et al., 1990Zimet et al., , 1988;;Ramaswamy et al., 2009).
Other data: Current cigarette smoking was self-reported.Alcohol intake was obtained from a validated food frequency questionnaire and expressed as grams per day (g/day) (Giles & Ireland, 1996).Physical activity was categorized as "active" if participants reported performing vigorous or light exercise regularly, or categorized as inactive.Weight was measured using electronic scales, height was measured using a Harpenden stadiometer, and body mass index (BMI) was calculated as weight/height 2 (kg/m 2 ).

Socioeconomic status
Unadjusted and age-adjusted data are presented in Compared to the most disadvantaged, increasing advantage was associated with better overall cognitive function, psychomotor function, and visual learning.These patterns persisted after adjustment for age, albeit with some losing significance.BMI, smoking, alcohol intake, and physical activity did not contribute to the models.The effect size for all relationships ranged from 0.004 to 0.016.No associations were evident between SES and attention or working memory before or after adjustment for age (all p > .05).

Marital status
In unadjusted models, married/de-facto was associated with better overall cognitive function and working memory, but not psychomotor function, attention, and visual learning compared to those who are single (Table 2).These associations persisted after adjustment for age, BMI, smoking, alcohol intake, and physical activity.The effect size for all relationships ranged from 0.010 to 0.013.

Perceived social support
Unadjusted and age-adjusted results are presented in Table 2.

Significant other
In unadjusted models, support from a significant other was associated with better attention, but not overall cognitive function, psychomotor function, working memory, or visual learning.Following adjustment for age, support from a significant other was associated with better overall cognitive function, psychomotor function, attention, and working memory, but not visual learning.BMI, smoking, alcohol intake, and physical activity did not contribute to the models.The effect size for all relationships ranged from 0.009 to 0.013.

Family
Before and after adjustment for age, support from family was not associated with any of the cognitive domains (psychomotor function, attention, working memory, and visual learning, all p > .05).

Friends
In unadjusted models, perceived social support from friends was not associated with any of the cognitive domains (psychomotor function, attention, working memory, and visual learning, all p > .05).
In age-adjusted models, social support from friends was associated with better overall cognitive function and psychomotor function, but not attention, working memory, and visual learning compared to low social support from friends.These relationships persisted after further adjustment for BMI, smoking, alcohol intake, and physical activity.The effect size for all relationships ranged from 0.008 to 0.010.

DISCUSSION
In this cross-sectional, population-based study of men, higher SES, being in a relationship, and perceived social support from a significant other and friends were associated with better overall cognitive function.In regard to specific domains, higher SES was associated with better psychomotor function and visual learning.Being married or in a de-facto relationship was associated with better working memory.Perceived social support from a significant other was associated with better attention, and working memory and perceived social support from friends were associated with better psychomotor function.
No associations were evident between being separated/divorced, widowed, or perceived social support from family with any of the cognitive domains.
Similar to our results, albeit in older populations, studies have reported low SES to be associated with poorer cognitive health (Dalstra et al., 2005;Petersen et al., 2021;Wallis et al., 2002;Z. Zhang et al., 2022).A recent cross-sectional study by Z.Moreover, a large cross-sectional study of elderly patients (n = 1420) found that socioeconomic deprivation was associated with lower cognitive function utilizing the Mini-Mental State Examination (Park et al., 2017).Furthermore, Qian et al. (2014) found in a clinical sample of older adults (>65 years) that those identified as having lower SES were less likely to seek help or postponed seeking help for their cognitive health compared to those with higher SES.When investigating specific domains of cognition, we found higher SES was related to psychomotor function and visual learning, suggesting better problem-solving skills and processing capacity.
Supporting a small body of literature in elderly populations (Bae et al., 2015;Chen et al., 2021;Feng et al., 2014;Hui et al., 2020), single individuals were found to be at a greater risk of cognitive decline than those in a relationship.A previous cross-sectional study of elderly Chinese residents aged over 65 years (n = 19,276) found that being divorced, separated, widowed, or single was associated with a greater risk of cognitive decline and dementia compared to residents in a relationship (Chen et al., 2021).
Consistent with our results, a growing body of evidence indicates a positive association between social support and global cognitive function among the elderly (Holtzman et al., 2004;Krueger et al., 2009;Yeh & Liu, 2003).Yeh and Liu (2003) found that higher cognitive function assessed by the Short Portable Mental Status Questionnaire was associated with increased perceived social support in a large sample (n = 4993) of adults aged >65 years.Interestingly, in the current study perceived social support from family members was not associated with cognitive function, with this result being similar to others (Kuiper et al., 2016;Pillemer & Holtzer, 2016).A cross-sectional study by Pillemer and Holtzer (2016) of community-residing older adults reported that perceived emotional and informational support and positive social interactions were all associated with better cognition.
Additionally, previous evidence suggests older people who are more socially supported have higher levels of cognitive function compared to less socially supported individuals (Barnes et al., 2004;Bassuk et al., 1999;Holtzman et al., 2004;Krueger et al., 2009;Yeh & Liu, 2003;Zunzunegui et al., 2003).In a prospective cohort study that followed 1203 individuals without dementia aged 75 and over, it was reported that people with a large social network were at reduced risk of dementia (Fratiglioni et al., 2000).While some studies reported that a larger social network is better for cognitive function (Crooks et al., 2008;Gureje et al., 2011;Scarmeas et al., 2001), others reported the size of social contacts is not strongly associated with cognition in old age (Hughes et al., 2008;).This may be due to the satisfaction of social supports being more important than the number of social connections (Glei et al., 2005;Krueger et al., 2009).Therefore, it is suggested that social contacts and perceived social support from family, friends, and acquaintances are important for fulfilment of different social needs (Yeh & Liu, 2003).
Several mechanisms underlying the link between social determinants of health and cognition have been suggested.First, sustaining social supports requires cognitive strategies which are likely to help build cognitive reserve through cognitive exercise and stimulation, shown to benefit memory and executive function (Cohen, 2016;Luethi et al., 2008;Scarmeas et al., 2006).Additionally, social determinants of health may act as a buffer against stressful situations and promote healthy lifestyle behaviors associated with fluid reasoning, attention, and psychomotor skills (Bae et al., 2015).It has also been suggested that people who accumulate more wealth live in better environments and less stressful conditions, further contributing to better cognitive health in later life (Z.X. Zhang et al., 2009;Z. Zhang et al., 2008).Moreover, people with higher SES often have a higher education and are in professions that require a high level of cognitive functioning (Qian et al., 2014).Thus, individuals with more cognition-demanding jobs may be able to perceive their own cognitive changes, leading them to seek help earlier, lessening cognitive decline (Qian et al., 2014).Furthermore, the marital resource model suggests being married is associated with unique social, psychological, and economic resources that are not typically obtained from other types of relationships, and this in turn promotes cognitive health (Waite & Gallagher, 2000).In addition, the stress model emphasizes that negative aspects of martial disruption, such as divorce and widowhood, create stress and undermine health, and may directly affect overall cognition and domains of working memory, psychomotor skills, and visual spatial abilities (Hughes et al., 2008;Williams & Umberson, 2004;Wilson et al., 2015).
This study has several strengths.First, we examined cognition in a large population-based study of unselected men without dementia spanning the adult age range.Given participants spanned the full adult age range, age was tested as a confounder and/or effect modifier.Second was our ability to adjust for potential confounders that may explain the relationships explored.Third, a comprehensive cognitive test battery was used to assess cognition, which allowed the investigation of specific cognitive domains.However, our study has some limitations.
The study included men only; accordingly, interpretation may not be generalizable to women, and therefore future research may consider collecting comparable data.Furthermore, as the CBB is a computerized battery, it does need to be acknowledged that a lack of familiarity with computers may influence performance.Additionally, as our study utilized participants from the general population, those who participated were more likely to have healthier profiles compared to those who did not.
In conclusion, higher SES, being in a relationship, and greater perceived social support from a significant other and friends were associated with better cognitive function in this population-based sample of men.Such evidence highlights social determinants of health to be associated with cognition in adulthood.Identifying mechanisms and interactions between these exposures on cognitive function is vital for timely treatment and future prevention strategies, as well as targeting those who may benefit more from intervention.
The GOS, originally designed to investigate the epidemiology of osteoporosis, has since expanded to investigate a range of diseases including mental health.Originally, 1540 men aged between 20 and 97 years were selected at random from electoral rolls from the Barwon Statistical Division (BSD) in south-eastern Australia.The inclusion criteria for the study included a listing on the commonwealth electoral roll as a resident of the BSD, and exclusion criteria included residents living in the area <6 months and individuals unable to provide written informed consent.Participants have returned for assessment 5and 15-year post baseline, with the current analysis utilizing data from the 15-year follow-up (2015−2019).Of 603 men who attended the 15-year follow-up at the time of writing, 67 participants did not complete the cognitive assessment, resulting in 536 eligible for inclusion in the current analyses.This study was approved by the Barwon Health Human Research Ethics Committee (00/56), and all participants gave written, informed consent.
Socioeconomic status: Area-based SES was determined by matching residential addresses for each participant to the corresponding Australian Bureau of Statistics 2016 Census Collection District, utilizing the Socio-Economic Index for Areas (SEIFA) value based on census data for each participant (Australian Bureau of Statistics, 2018).SEIFA values demonstrate the characteristics of participants within an area, providing a single measure to rank the level of disadvantage at the arealevel.For our study, SES was determined from the Index of Relative Socioeconomic Advantage and Disadvantage (IRSAD), which accounts for high and low income, and the type of occupation from unskilled employment to professional positions.The IRSAD scores for the participants were categorized into quintiles according to cut-points for the study region, with quintile 1 being the most disadvantaged and quintile 5 the most advantaged.

Table 2
Unadjusted and age adjusted multiple regression models showing associations between socioeconomic status (SES), relationship status and social support, and cognitive function.
, with the lowest SES group (quintile 1, most disadvantaged) held as referent.TA B L E 2*Best model = age, BMI, smoking, alcohol intake and physical activity.
Zhang et al. (2022)reported greater cognitive impairment measured by The Montreal Cognitive Assessment among people aged 65 years and older with lower SES.