Sage Journals: Discover world-class research

Abstract

Subjective cognitive decline (SCD) has been linked to Alzheimer’s Disease in the literature. However, little is known about whether SCD is associated with social/emotional support (SES). To investigate this association, this study utilized the 2015 and 2016 Behavioral Risk Factor Surveillance System data. A study population of 17206 participants aged 45 years and older who responded to both the Emotional Support and Life Satisfaction survey module and the Cognition Decline survey module were included. Of this study population, 11.22% had SCD, and 21.83% reported insufficient SES. A much higher percentage of those with insufficient SES experienced SCD compared to those with sufficient SES (21.15% vs 8.45%, P < .0001). Insufficient SES was significantly associated with SCD (odds ratio = 1.68, 95% confidence interval: 1.37-2.06), after controlling for other factors. Furthermore, this study found certain demographic groups such as female, white, or married groups were more likely to receive sufficient SES.

Keywords

social support emotional support BRFSS subjective cognitive decline Alzheimer disease dementia

Introduction

It is estimated that 5% to 7% of people aged 60 and older have dementia globally, with Alzheimer’s Disease (AD) being the most common form of dementia.¹ Identifying precursors of dementia is a critical step in prevention. Mild cognitive impairment (MCI), particularly the amnestic form of MCI where the memory loss is predominant, is recognized as a precursor for AD, with a conversion rate of 10% to 15% per year.^2
–4

Recent literature has shown increasing evidence that even before MCI, subjective cognitive decline (SCD), often referred to as subjective cognitive impairment, or subjective memory complaints, without objective measurements can “announce dementia” as a “pre-MCI stage of subsequently manifest Alzheimer disease.”^5

-12 A large cohort study following elderly participants for 5 years found that 15% of those with baseline subjective memory deterioration developed dementia compared to 6% of those without subjective memory deterioration.¹³ Another study demonstrated that among 21 individuals who had subjective cognitive worsening complaints, a total of 8 (38%) developed dementia (6 developed AD) within a 3-year follow-up period.¹⁴

Research on SCD focused mainly on biomarkers for people with SCD, or predictive factors for progression to AD, or cognitive outcomes, with little research on risk factors for SCD.^3,11,15,16 Since growing evidence shows people with SCD are at an increased risk of developing AD, identifying risk factors for SCD may help prevent AD by preventing SCD in the first place. Reducing SCD can also lead to other health benefits, because SCD itself is shown to significantly reduce quality of life, shorten life expectancy, and lead to other neurodegenerative diseases and psychiatric diseases.^10,17

Despite the usefulness of identifying risk factors for SCD, limited research exists in the literature. Using large telephone interviews, the study by Chen et al identified several modifiable risk factors for SCD including depression, lower education level, less exercise, and hypertension.¹⁸ Focusing on older African Americans, another study identified the following risk factors for subjective memory change: depression, anxiety, and total number of cerebrovascular risk factors.¹⁹ Peterson et al found household income was associated with SCD.²⁰ These studies have identified some risk factors, but more research is needed to identify additional risk factors for SCD to better prevent the development of SCD and subsequent dementia. Despite the risk factors that have previously been identified, the effects of psychosocial elements (eg, social and emotional support [SES]) on SCD have hardly been investigated. Indeed, psychosocial elements are less studied factors according to a systematic review on risk factors for dementia and (objective) cognitive decline.²¹ This study focuses on one particular psychosocial element, namely SES, which to the best of our knowledge has not been directly explored as a risk factor for SCD in the literature.

There is a need to examine whether insufficient SES is a risk factor for SCD. Social and emotional support is an important factor that influences individuals’ overall well-being, both mentally and physically.^22,23 Multiple studies have suggested that receiving sufficient SES may promote cognitive performance, resilience, and reduce memory loss among those studied.^24

-28 The mechanism of how SES benefits cognitive function has been explored in the literature. People with sufficient SES tend to lead more active social lives with more positive social relationships, which could possibly reduce the likelihood of developing dementia.²⁹ Engaging actively in social activities provides the brain with complex stimuli to prevent cognitive decline.³⁰ In addition, positive social relationships may be a protective factor against depression, which is strongly linked to cognitive decline.³¹

Despite the evidence that SES is beneficial for cognitive performance, it is still unclear whether insufficient SES is a risk factor for SCD. In this article, we will use national US survey data from the Behavioral Risk Factor Surveillance System (BRFSS) to study the association between insufficient SES and SCD. Having sufficient SES is largely a subjective feeling; therefore, this information is best collected by interviews and surveys. The BRFSS data used for this study contain information on participant’s SES status. We hypothesize that insufficient SES is associated with SCD. Additionally, to better inform interventions aimed at improving SES, this study will further investigate which factors are associated with the receipt of sufficient SES.

Method

Data Source

This study used the BRFSS data, which is a large national telephone survey of non-institutionalized US adults conducted by the Centers for Disease Control and Prevention (CDC).³² The BRFSS survey questions consist of (1) required modules, which are required survey components for all 50 states and the District of Columbia; (2) optional modules, which are optional for states to choose from; and (3) state-added modules, which are modules added and used by individual states to collect data of interest to the state.³³ The BRFSS survey conducts more than 400 000 adult interviews annually, and the CDC has various measures in place to ensure high-quality data.³²

The main data fields needed to examine the association between SCD and emotional and SES are contained in two optional modules from BRFSS: the Emotional Support and Life Satisfaction module (ESLS module) and the Cognitive Decline module (CD module). Behavioral Risk Factor Surveillance System data prior to 2011 were not included since they were not comparable to data after 2011 which include cellphone interviews and use of new weighting strategies.³³ For the BRFSS data from 2011 to the most recent data year 2018, only the following states have conducted surveys using both ESLS module and the CD module: Minnesota in 2015, Rhode Island in 2015, and Tennessee in 2016. Therefore, the BRFSS data used for this study were from those three states in either 2015 or 2016.

Variables

A participant’s SCD status was defined by the answer to the question “During the past 12 months, have you experienced confusion or memory loss that is happening more often or is getting worse?” An answer of “yes” was considered as experiencing SCD. An answer of “no” was considered as confirming the absence of SCD. The primary predictor of our interest was SES status, which was assessed by the answer to the question “How often do you get the social and emotional support you need?” With five answer choices to select from: “always,” “usually,” “sometimes,” “rarely,” and “never,” a binary variable was created where “always” and “usually” were considered as having sufficient SES and “sometimes,” “rarely,” and “never” were considered as having insufficient SES. In our sensitivity analysis to see whether the study results were robust to the categorization of SES levels, we used an alternative grouping method, which grouped “sometimes” into sufficient SES.

To control for other variables that could contribute to SCD, we also considered age, gender, race, marital status, employment status, education level, exercise, depressive disorder, and chronic health conditions (coronary heart disease and diabetes) in the model. We chose to study adults aged 45 years and older since cognitive decline can begin as early as middle age.^34,35 Depending on the original answer to the multiple-choice question, marital status was regrouped into four categories: married, divorced, widowed, and single. The divorced group included those who answered “divorced” as well as those who answered “separated” due to the small sample size of “separated.” The category “single” combined individuals who were never married and those who were a member of an unmarried couple. Employment status was also regrouped into the following categories: employed (including “employed for wages” and “self-employed”), out of work (including “out of work for less than one year” and “out of work for more than one year”), homemaker, retired, and unable to work. Education level was divided into three categories: “did not graduate from high school,” “only graduated high school but not higher,” and “graduated from college or technical school” based on the original survey answer. Lastly, respondents’ diabetes status was regrouped into two categories: “having diabetes” and “not having diabetes” (including “no diagnosis of diabetes” and “having prediabetes”). The exercise variable was created based on dichotomous answer to the question “During the past month, other than your regular job, did you participate in any physical activities or exercises such as running, calisthenics, golf, gardening, or walking for exercise?” Observations with missing data were excluded.

Statistical Analysis

SAS version 9.4 was used to perform the statistical analysis.³⁶ The data analysis incorporated the weighting variables to be statistically representative of the overall population according to the CDC’s codebook instruction.³⁷ χ² tests were used to test the associations between SES and various categorical predictors.

Logistic regression models were fit to test the association between SCD and various covariates. Both univariate logistic regression models including one covariate of interest at a time and a multiple logistic regression model including all covariates simultaneously in one model were performed to evaluate the impact of different predictors on SCD. The multiple logistic regression model controlled for other categorical factors including age, gender, race, marital status, education level, employment status, exercise, depressive disorder, and other chronic health issues. In addition, logistic regression was employed to investigate what factors were associated with the receipt of sufficient SES.

Results

A total of 17206 participants from the 2015 and 2016 BRFSS survey were included in this study; 13727 participants from the 2015 survey and 3479 from the 2016 survey, respectively. Table 1 shows the demographic characteristics as well as the percentages of SCD within each subgroup of respondents. The second column in Table 1 summarizes the distribution of various participant characteristics for the entire study population. The third and fourth columns show the percentages of respondents with and without SCD respectively, among those who have the same categorical value listed in column 2.

Table 1.

Demographic Characteristics and Percentages With Subjective Cognitive Decline (SCD).

Variables^a, P Values	All Participants^b (n = 17 206)	% of Respondents With SCD^c (n = 1773)	% of Respondents Without SCD^c (n = 15 433)
Gender; P = .5447	Male (46.99%)	10.96%	89.04%
Gender; P = .5447	Female (53.01%)	11.46%	88.54%
Age; P = .2321	45-54 (30.66%)	10.27%	89.73%
	55-64 (31.84%)	12.08%	87.92%
	65 and older (37.50%)	11.28%	88.72%
Race; P = .0109	White (87.27%)	10.73%	89.27%
	Black (7.47%)	13.32%	86.68%
	Other races (5.26%)	16.52%	83.48%
Marital status; P < .0001	Married (62.28%)	9.14%	90.86%
	Divorced (16.54%)	18.07%	81.93%
	Widowed (12.35%)	11.30%	88.70%
	Single (8.83%)	13.02%	86.98%
Education level; P < .0001	Did not graduate from high school (12.93%)	20.92%	79.08%
	Only graduated from high school but not higher (60.99%)	11.05%	88.95%
	Graduated from college or technical school (26.08%)	6.82%	93.18%
Employment status; P < .0001	Employed (48.41%)	5.56%	94.44%
	Out of work (3.44%)	17.85%	82.15%
	Homemaker (4.97%)	11.08%	88.92%
	Retired (33.39%)	11.39%	88.61%
	Unable to work (9.79%)	36.40%	63.60%
Social and emotional support level; P < .0001	Insufficient (21.83%)	21.15%	78.85%
Social and emotional support level; P < .0001	Sufficient (78.17%)	8.45%	91.55%
Exercise; P < .0001	Exercised in the past month (70.10%)	8.92%	91.08%
Exercise; P < .0001	No exercise in the past month (29.90%)	16.62%	83.38%
Depressive disorder; P < .0001	Yes (20.09%)	29.02%	70.98%
Depressive disorder; P < .0001	No (79.91%)	6.75%	93.25%
Coronary heart disease; P < .0001	Yes (7.68%)	22.42%	77.58%
Coronary heart disease; P < .0001	No (92.32%)	10.29%	89.71%
Diabetes; P < .0001	Yes (16.42%)	18.98%	81.02%
Diabetes; P < .0001	No (83.58%)	9.70%	90.30%

^a The P values are for testing whether the categorical variable under study is associated with SCD status.

^b The percentages under the column “all participants” represent the distribution among all categorical values of a given categorical variable. The percentages under the same categorical variable in this column adds up to 100% vertically. For instance, for the gender variable, 46.99% + 53.01% = 100%.

^c The percentages in the third and fourth columns are the distribution of SCD status for a given subgroup of respondents with the same categorical value of a categorical variable. Those percentages add up to 100% horizontally. For instance, within the male subgroup, 10.96% (with SCD) + 89.04% (without SCD) = 100%.

As shown in Table 1, the study population contained a greater number of females than males (53.01% vs 46.99%, respectively). Each of the three age groups (45-54, 55-64, and 65 and older) accounted for approximately one third of the study population. Since the data originated from the three states with relatively higher percentages of white populations, the non-white population in this study (12.73%) was lower than the national average, even after properly weighting the data.³⁸

Since the study population was older than the general population, a higher percentage of widowed participants (12.35%) and higher percentage without high school diplomas (12.93%) were observed. Among the study population, 33.39% were retired, 48.41% were currently employed, and 9.79%were unable to work.

In total, 21.83% of the study population (n = 3204) reported insufficient SES and 78.17% had sufficient SES. Specifically, for the frequency of receiving SES, 6.18% reported “never,” 3.54% reported “occasionally” and 12.11% reported “sometimes”, while 24.71% responded “usually,” and 53.46% responded “always.”

Columns 3 and 4 in Table 1 show the distribution of SCD status. In total, 11.22% (n = 1773) of the study population experienced SCD. As shown in Table 1, there was no statistical difference in SCD by gender (P = .54), with 10.96% of males and 11.46% of females experiencing SCD. By race, non-whites were more likely to experience SCD (P = .01), with 13.32% of blacks and 16.52% of other race experiencing SCD compared to 10.73% of whites. Subjective cognitive decline status differed by marital status (P < .0001), with those divorced most likely to experience SCD (18.07%). χ² tests showed that SCD status was also distributed differently for other demographic variables including education level and employment status (both P < .0001). The distribution of SCD status was statistically different for various participant characteristics including having exercise (P < .0001), depressive disorder (P < .0001), coronary heart disease (P < .0001), and diabetes (P < .0001).

With respect to the primary risk factor of interest-SES, 21.15% of those with insufficient SES experienced SCD, compared to just 8.45% of those with sufficient SES (P < .0001).

Table 2 shows the logistic regression results to assess the effects of various risk factors for SCD. To examine the effects of different predictors, this study utilized both univariate (columns 3 and 4 in Table 2) and multiple logistic regression models (columns 5 and 6 in Table 2).

Table 2.

Logistic Regression Models for SCD Explained by SES.

Variables	Categories	Univariate Logistic Regression		Multiple Logistic Regression
		SCD		SCD
		ORs (95% CI)	P Values	ORs (95% CI)	P Values
Social and emotional support	Sufficient (Ref)
Social and emotional support	Insufficient	2.90 (2.45-3.44)^a	<.0001	1.68 (1.37-2.06)^a	<.0001
Gender	Female (Ref)
Gender	Male	1.05 (0.89-1.24)	.54	0.84 (0.70-1.00)	.055
Age	45-54 (Ref)
	55-64	1.20 (0.96-1.50)	.11	0.99 (0.78-1.25)	.90
	65 and older	1.11 (0.91-1.36)	.31	1.00 (0.76-1.32)	.99
Race	Other races (Ref)
	White	0.61 (0.42-0.87)^b	.0064	0.73 (0.52-1.02)	.06
	Black	0.78 (0.48-1.25)	.29	0.84 (0.53-1.35)	.47
Marital status	Married (Ref)
	Divorced	2.19 (1.79-2.69)^a	<.0001	1.24 (0.98-1.57)	.07
	Widowed	1.27 (1.03-1.56)^c	.03	0.90 (0.70-1.15)	.41
	Single	1.49 (1.09-2.03)^c	.01	0.97 (0.70-1.33)	.84
Education level	Graduated from college or technical school (Ref)
	Did not graduate high school	3.62 (2.77, 4.72)^a	<.0001	1.43 (1.04-1.97)^c	.03
	Graduated high school but not higher	1.70 (1.44-2.00)^a	<.0001	1.20 (1.00-1.44)	.05
Employment status	Employed (Ref)
	Out of work	3.69 (2.59-5.25)^a	<.0001	2.29 (1.55-3.38)^a	<.0001
	A homemaker	2.12 (1.40-3.20)^b	.0004	2.01 (1.29, 3.14)^b	.002
	Retired	2.18 (1.79-2.65)^a	<.0001	1.95 (1.51-2.52)^a	<.0001
	Unable to work	9.71 (7.59-12.44)^a	<.0001	3.83 (2.88, 5.08)^a	<.0001
Exercise	Exercised in the past month (Ref)
Exercise	No exercise in the past month	2.04 (1.72-2.41)^a	<.0001	1.23 (1.02-1.49)^c	.03
Depressive disorder	No (Ref)
Depressive disorder	Yes	5.65 (4.78-6.68)^a	<.0001	3.80 (3.16-4.57)^a	<.0001
Coronary heart disease	No (Ref)
Coronary heart disease	Yes	2.52 (2.00-3.17)^a	<.0001	1.66 (1.25, 2.19)^b	.0004
Diabetes	No (Ref)
Diabetes	Yes	2.18 (1.80-2.64)^a	<.0001	1.45 (1.18, 1.78)^b	.0005

Abbreviations: CI, confidence interval; ORs, odds ratios; SCD, subjective cognitive decline.

^a P < .0001.

^b P < .01.

^c P < .05.

In the univariate analysis, the following binary variables were statistically significantly associated with SCD: insufficient SES (odds ratio [OR]: = 2.90, 95% confidence interval [CI]: 2.45-3.44), no exercise (OR = 2.04, 95% CI: 1.72-2.41), depressive disorder (OR = 5.65, 95% CI: 4.78-6.68), coronary heart disease (OR = 2.52, 95% CI: 2.00-3.17), and diabetes (OR = 2.18, 95% CI: 1.80-2.64). In the univariate analysis, the following categorical variables with multiple categories were also statistically significant in explaining SCD. For example, being white was less likely to have SCD compared to being of other race (OR = 0.61, 95% CI: 0.42-0.87), whereas being divorced was more likely to experience SCD compared to being married (OR = 2.19, 95% CI: 1.79-2.69). Not graduating from high school was associated with a greater likelihood of experiencing SCD compared to graduating from college or technical school (OR = 3.62, 95% CI: 2.77-4.72). Compared to the employed respondents, those who were out of work (OR = 3.69, 95% CI: 2.59-5.25), unable to work (OR = 9.71, 95% CI: 7.59-12.44), or retired (OR = 2.18, 95% CI: 1.79-2.65) were more likely to experience SCD.

As shown in the last 2 columns in Table 2, in the multiple logistic regression model which included all the covariates of interest, insufficient SES still significantly increased the risk of having SCD (OR_adj = 1.68, 95% CI: 1.37-2.06). With other variables being controlled for in the multiple logistic regression, the following risk factors for experiencing SCD were identified: unable to work (compared to being employed; OR_adj = 3.83, 95% CI: 2.88-5.08), no exercise (OR_adj = 1.23, 95% CI: 1.02-1.49), depressive disorder (OR_adj = 3.80, 95% CI: 3.16-4.57), coronary heart disease (OR_adj = 1.66, 95% CI: 1.25-2.19), and diabetes (OR_adj = 1.45, 95% CI: 1.18-1.78).

To investigate how robust the study results were, we performed some sensitivity analyses. The SES level variable was redefined by grouping “sometimes” into the “sufficient SES” group in sensitivity analysis. The results still showed a strong significant protective effect of sufficient SES against SCD. We also did some subgroup analyses to analyze the SES effect within subgroups stratified by gender or age. Within each subgroup, insufficient SES was consistently significantly associated with an increased risk of SCD.

Since insufficient SES was found to be an important risk factor for SCD based on our analysis, we did a further analysis to investigate what factors are associated with having sufficient SES, in order to inform efforts to improve SES level. As shown in Table 3, results from fitting a multiple logistic regression model indicated that the following factors were associated with a higher likelihood of receiving SES: being female (OR_adj = 1.30, 95% CI: 1.13-1.50), white (compared to other race, OR_adj = 1.46, 95% CI: 1.07-1.98), married (compared to being single, OR_adj = 1.83, 95% CI: 1.45-2.32), having graduated from high school (compared to not graduating from high school, OR_adj = 1.49, 95% CI: 1.18-1.86) or having graduated from college or technical school (compared to not having graduated from high school, OR_adj = 2.25, 95% CI: 1.74-2.90), and employed (compared to unable to work, OR_adj = 1.55, 95% CI: 1.22-1.98), exercise (OR_adj = 1.37, 95% CI: 1.18-1.58), no depressive disorder (OR_adj = 2.28, 95% CI: 1.96-2.66), and no diabetes (OR_adj = 1.22, 95% CI: 1.01-1.46).

Table 3.

Multiple Logistic Regression of the Associations Between Demographic Characteristics and Having Sufficient SES.

	Sufficient SES
	Adjusted ORs (95% CI)	P Values
Gender
Male (Ref)
Female	1.30 (1.13-1.50)^a	.0003
Age
45-54 (Ref)
55-64	1.09 (0.91-1.31)	.34
65 and older	1.08 (0.87-1.33)	.49
Race
Other races (Ref)
White	1.46 (1.07-1.98)^b	.02
Black and African American	0.75 (0.50-1.13)	.16
Marital status
Singled (Ref)
Married	1.83 (1.45-2.32)^c	<.0001
Divorced	0.75 (0.58-0.97)^b	.03
Widowed	1.37 (1.03-1.82)^b	.03
Education level
Did not graduate from high school (Ref)
Only graduated from high school but not higher	1.49 (1.18-1.86)^a	.0007
Graduated from college or technical school	2.25 (1.74-2.90)^c	< .0001
Employment status
Unable to work (Ref)
Employed	1.55 (1.22-1.98)^a	.0004
Out of work	0.89 (0.62-1.29)	.55
A homemaker	1.53 (1.05-2.24)^b	.03
Retired	1.26 (0.98-1.63)	.08
Exercise
No exercise in the past month (Ref)
Exercised in the past month	1.37 (1.18-1.58)^c	< .0001
Depression disorder
Yes (Ref)
No	2.28 (1.96-2.66)^c	< .0001
Coronary heart disease
Yes (Ref)
No	1.00 (0.79-1.26)	.99
Diabetes
Yes (Ref)
No	1.22 (1.01-1.46)^b	.04

Abbreviations: CI, confidence interval; ORs, odds ratios; SES, social and emotional support.

^a P < .01.

^b P < .05.

^c P < .0001.

Discussion

This study shows that insufficient SES is a significant risk factor for SCD. The odds ratio in multiple logistic regression model was 1.68, higher than most other well-recognized risk factors such as no exercise, having history of coronary heart disease, or history of diabetes. Not only is SES statistically significant, but the magnitude of its effect is clinically large, for as many as 21.15% of people with insufficient SES reported SCD, whereas only 8.45% of those with sufficient SES reported SCD. The reasons why insufficient SES is linked to SCD are similar to those that explain why insufficient SES is linked to dementia. The reasons may include that lack of SES is associated with less positive relationships and less engagement in social activities which result in less brain stimulation and a greater degree of depression, whereas more SES is associated with more social activities and better functioning of the brain.^29
-31

To the best of our knowledge, this study is the first to directly investigate the association between SES and SCD using recent large scale US survey data. In the general literature on MCI and dementia, SES is widely recognized as a risk factor for MCI and dementia, though possibly expressed under a variety of terminology representing psychosocial factors, including social support, social engagement, social relations, social network, active social life, informational support, and emotional support.^{21,30,39

-45} The fact that SCD, MCI, and dementia share the same risk factor, that is, SES, highlights the importance of increasing SES level, as it may prevent not only SCD but also MCI and dementia.

In addition, our analysis identified several factors that are associated with a higher likelihood of receiving SES. Some of these factors, such as being employed or having a higher education level, are modifiable which can lead to a higher SES level. Awareness of the detrimental effects of insufficient SES and its risk factors can motivate patients and families to consciously self-evaluate their own SES levels and actively seek more SES. Acknowledging the importance of SES will assist physicians with improving patient care and provide guidance for health interventions and policies that aim to prevent SCD, MCI, and dementia by increasing the level of SES. Various interventions to improve SES have been explored in the literature, including education of spouse, involvement of a support person, and group problem-solving.⁴⁶

In addition to SES, other variables were found to be significant risk factors for SCD in our study, including divorce, out of work, lower education level, no exercise, and history of a depressive disorder, coronary heart disease, and diabetes. Being divorced, out of work, or less educated may indicate smaller social network, less mind-stimulating activities, or a higher likelihood of depression, which may contribute to the development of SCD. No exercise, history of depression, coronary heart disease, or diabetes may lead to physical health problems, which may reciprocally affect one’s mental health and lead to SCD. Interestingly, while older adults are more prone to develop dementia, old age was not found to be a risk factor for SCD in our study population. A likely reason was that the survey questionnaire was only administered to those who were mentally capable of answering surveys, thus elderly people with dementia generally would not have been included as survey respondents.⁴⁷ Caution needs to be taken when interpreting the age effect in our model.

Most of the risk factors identified in this study are consistent with findings in the few related studies on risk factors for SCD. For example, depression and less exercise were also identified as risk factors in the study by Chen et al.¹⁸ Less education was identified to be a risk factor by both Chen et al and Peterson et al.^18,20 Depression, a statistically significant risk factor in our study, was also identified by Sperling et al.¹⁹ However, none of those studies specifically considered SES or any other psychosocial factor as a risk factor, which turned out to be one of the strongest risk factors for SCD in our study.

This study has several limitations. First, the SCD status was based on a respondent’s answer to the survey question, which may be subject to recall bias. However, we believe that this possible bias should be minimal since the respondents were required to be mentally and physically capable to complete the interview.⁴⁷ In addition, the recall bias is probably small because the percentages of respondents who reported having experienced SCD were consistent across all the years in the BRFSS data including the data years that were not included in this study. Second, our data were from three US states and may not be generalizable to the entire US population. Another limitation is that the BRFSS data are cross-sectional with no follow-up over time. Without proper follow-up, it is difficult to know whether a change in SES status over time can result in a change in SCD status, which is of great interest to study.

Despite these limitations, our study using large nationally representative survey data shows a strong association between insufficient SES and SCD, suggesting that obtaining and maintaining sufficient SES may be a critical approach to preventing SCD and possibly subsequent dementia. When designing preventative programs for SCD and dementia, improving SES should be considered by addressing the factors that are associated with insufficient SES.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Xingran Weng

Li Wang

References

Prince

Bryce

Albanese

Wimo

Ribeiro

Ferri

. The global prevalence of dementia: a systematic review and metaanalysis. Alzheimers Dement. 2013;9(1):63–75. e62.

Janoutova

Sery

Hosak

Janout

. Is mild cognitive impairment a precursor of Alzheimer’s disease? Short review. Cent Eur J Public Health. 2015;23(4):365–367.

Summers

Saunders

. Neuropsychological measures predict decline to Alzheimer’s dementia from mild cognitive impairment. Neuropsychology. 2012;26(4):498–508.

Bozoki

Giordani

Heidebrink

Berent

Foster

. Mild cognitive impairments predict dementia in nondemented elderly patients with memory loss. Arch Neurol. 2001;58(3):411–416.

Schmand

Jonker

Hooijer

Lindeboom

. Subjective memory complaints may announce dementia. Neurology. 1996;46(1):121–125.

Reisberg

Gauthier

. Current evidence for subjective cognitive impairment (SCI) as the pre-mild cognitive impairment (MCI) stage of subsequently manifest Alzheimer’s disease. Int Psychogeriatr. 2008;20(1):1–16.

Reisberg

Shulman

Torossian

Leng

Zhu

. Outcome over seven years of healthy adults with and without subjective cognitive impairment. Alzheimers Dement. 2010;6(1):11–24.

Jessen

Amariglio

van Boxtel

, et al. A conceptual framework for research on subjective cognitive decline in preclinical Alzheimer’s disease. Alzheimers Dement. 2014;10(6):844–852.

Reisberg

Prichep

Mosconi

, et al. The pre-mild cognitive impairment, subjective cognitive impairment stage of Alzheimer’s disease. Alzheimers Dement. 2008;4(1 suppl):S98–S108.

10.

Cheng

Chen

Chiu

. From mild cognitive impairment to subjective cognitive decline: conceptual and methodological evolution. Neuropsychiatr Dis Treat. 2017;13:491–498.

11.

Sierra-Rio

Balasa

Olives

, et al. Cerebrospinal fluid biomarkers predict clinical evolution in patients with subjective cognitive decline and mild cognitive impairment. Neurodegener Dis. 2016;16(1-2):69–76.

12.

Selnes

Aarsland

Bjornerud

, et al. Diffusion tensor imaging surpasses cerebrospinal fluid as predictor of cognitive decline and medial temporal lobe atrophy in subjective cognitive impairment and mild cognitive impairment. J Alzheimers Dis. 2013;33(3):723–736.

13.

Wang

van Belle

Crane

, et al. Subjective memory deterioration and future dementia in people aged 65 and older. J Am Geriatr Soc. 2004;52(12):2045–2051.

14.

Jessen

Wiese

Bachmann

, et al. Prediction of dementia by subjective memory impairment: effects of severity and temporal association with cognitive impairment. Arch Gen Psychiatry. 2010;67(4):414–422.

15.

Rabin

Smart

Amariglio

. Subjective cognitive decline in preclinical Alzheimer’s disease. Annu Rev Clin Psychol. 2017;13:369–396.

16.

Galtier

Nieto

Lorenzo

Barroso

. Subjective cognitive decline and progression to dementia in Parkinson’s disease: a long-term follow-up study. J Neurol. 2019;266(3):745–754.

17.

Strand

Knapskog

Persson

, et al. Survival and years of life lost in various aetiologies of dementia, mild cognitive impairment (MCI) and subjective cognitive decline (SCD) in Norway. PLoS One. 2018;13(9):e0204436.

18.

Chen

Siddarth

Ercoli

Merrill

Torres-Gil

Small

. Modifiable risk factors for Alzheimer disease and subjective memory impairment across age groups. PLoS One. 2014;9(6):e98630.

19.

Sperling

Tsang

Williams

Park

Helenius

Manning

. Subjective memory change, mood, and cerebrovascular risk factors in older African Americans. J Geriatr Psychiatry Neurol. 2017;30(6):324–330.

20.

Peterson

Carvajal

McGuire

Fain

Bell

. State inequality, socioeconomic position and subjective cognitive decline in the United States. SSM Popul Health. 2019;7:100357.

21.

Baumgart

Snyder

Carrillo

Fazio

Kim

Johns

. Summary of the evidence on modifiable risk factors for cognitive decline and dementia: a population-based perspective. Alzheimers Dement. 2015;11(6):718–726.

22.

Kim

Sherman

Taylor

. Culture and social support. Am Psychol. 2008;63(6):518–526.

23.

Strine

Chapman

Balluz

Mokdad

. Health-related quality of life and health behaviors by social and emotional support. Their relevance to psychiatry and medicine. Soc Psychiatry Psychiatr Epidemiol. 2008;43(2):151–159.

24.

Glymour

Weuve

Fay

Glass

Berkman

. Social ties and cognitive recovery after stroke: does social integration promote cognitive resilience? Neuroepidemiology. 2008;31(1):10–20.

25.

Ertel

Glymour

Berkman

. Effects of social integration on preserving memory function in a nationally representative US elderly population. Am J Public Health. 2008;98(7):1215–1220.

26.

Fratiglioni

Paillard-Borg

Winblad

. An active and socially integrated lifestyle in late life might protect against dementia. Lancet Neurology. 2004;3(6):343–353.

27.

James

Wilson

Barnes

Bennett

. Late-life social activity and cognitive decline in old age. J Int Neuropsychol Soc. 2011;17(6):998–1005.

28.

Seeman

Lusignolo

Albert

Berkman

. Social relationships, social support, and patterns of cognitive aging in healthy, high-functioning older adults: MacArthur studies of successful aging. Health Psychol. 2001;20(4):243–255.

29.

Wang

Karp

Winblad

Fratiglioni

. Late-life engagement in social and leisure activities is associated with a decreased risk of dementia: a longitudinal study from the Kungsholmen project. Am J Epidemiol. 2002;155(12):1081–1087.

30.

Ellwardt

Van Tilburg

Aartsen

. The mix matters: complex personal networks relate to higher cognitive functioning in old age. Soc Sci Med. 2015;125:107–115.

31.

Schwarzbach

Luppa

Forstmeier

Konig

Riedel-Heller

. Social relations and depression in late life-a systematic review. Int J Geriatr Psychiatry. 2014;29(1):1–21.

32.

Centers for Disease Control and Prevention. Behavioral risk factor surveillance system. 2019. https://www.cdc.gov/brfss/index.html. Accessed December 5, 2019.

33.

Centers for Disease Control and Prevention. The BRFSS data user guide. 2013; https://www.cdc.gov/brfss/data_documentation/pdf/UserguideJune2013.pdf. Accessed February 3, 2019.

34.

Hultsch

Hammer

Small

. Age differences in cognitive performance in later life: relationships to self-reported health and activity life style. J Gerontol. 1993;48(1):P1–11.

35.

Cerhan

Folsom

Mortimer

, et al. Correlates of cognitive function in middle-aged adults. Atherosclerosis risk in communities (ARIC) study investigators. Gerontology. 1998;44(2):95–105.

36.

SAS [computer program]. Version 9.4. Cary, NC: SAS Institute Inc; 2014.

37.

Centers for Disease Control and Prevention. Behavioral risk factor surveillance system: 2015 Codebook Report. 2016; https://www.cdc.gov/brfss/annual_data/2015/pdf/codebook15_llcp.pdf. Accessed December 15, 2018.

38.

Census Bureau

U.S.

Population estimates. 2018; https://www.census.gov/quickfacts/fact/table/US/PST045218. Accessed August 12, 2019.

39.

Lehert

Villaseca

Hogervorst

Maki

Henderson

. Individually modifiable risk factors to ameliorate cognitive aging: a systematic review and meta-analysis. Climacteric. 2015;18(5):678–689.

40.

Dickinson

Bamford

Exley

Emmett

Hughes

Robinson

. Planning for tomorrow whilst living for today: the views of people with dementia and their families on advance care planning. Int Psychogeriatr. 2013;25(12):2011–2021.

41.

Dos Santos

Rocha

Fernandez

de Padua

Reppold

. Association of lower spiritual well-being, social support, self-esteem, subjective well-being, optimism and hope scores with mild cognitive impairment and mild dementia. Front Psychol. 2018;9:371.

42.

Dickinson

Potter

Hybels

McQuoid

Steffens

. Change in stress and social support as predictors of cognitive decline in older adults with and without depression. Int J Geriatr Psychiatry. 2011;26(12):1267–1274.

43.

Holtzman

Rebok

Saczynski

Kouzis

Wilcox Doyle

Eaton

. Social network characteristics and cognition in middle-aged and older adults. J Gerontol B Psychol Sci Soc Sci. 2004;59(6):P278–P284.

44.

Pillemer

Holtzer

. The differential relationships of dimensions of perceived social support with cognitive function among older adults. Aging Ment Health. 2016;20(7):727–735.

45.

Kuiper

Zuidersma

Oude Voshaar

, et al. Social relationships and risk of dementia: a systematic review and meta-analysis of longitudinal cohort studies. Ageing Res Rev. 2015;22:39–57.

46.

Hogan

Linden

Najarian

. Social support interventions. Clin Psychol Rev. 2002;22(3):381–440.

47.

Centers for Disease Control and Prevention. Behavioral risk factor surveillance system 2015 summary data quality report. 2015; https://www.cdc.gov/brfss/annual_data/2015/pdf/2015-SDQR.pdf. Accessed March 20, 2019.

Association Between Subjective Cognitive Decline and Social and Emotional Support in US Adults

Abstract

Keywords

Introduction

Method

Data Source

Variables

Statistical Analysis

Results

Discussion

Footnotes

Declaration of Conflicting Interests

Funding

ORCID iDs

References