Abstract
Objectives.
To investigate the psychometric properties of the 10-item Social Engagement and Activities Questionnaire (SEAQ) to assess social-group, interpersonal interaction, and solitary activities among low-income, depressed homebound older adults (n=269).
Methods.
We used principal component analysis (PCA) to evaluate the underlying dimensions of the 10-item full SEAQ and a 6-item abbreviated item set. We assessed evidence of validity for the SEAQ by examining relationships between the SEAQ and older adults’ clinical characteristics: perceived social support, disability, and depressive symptoms.
Results.
PCA results showed two components: (1) a general social-group activities engagement component; and (2) a low level of socialization (i.e., strong negative coefficients on the recreational activities and self-enrichment/educational activities and a negative coefficient for interpersonal interaction activities). The general social-group activities engagement component in both the full and abbreviated SEAQ were significantly positively correlated with the full and abbreviated SEAQ and perceived social support, providing evidence for convergent validity, and they were significantly negatively correlated with disability and depressive symptoms, providing evidence for discriminant validity.
Conclusions.
The present study provides evidence of validity for the use of the SEAQ to assess social engagement and activities among low-income, depressed homebound older adults.
Clinical Implications.
The SEAQ may be used in future studies measuring changes in social engagement and activities in these older adults.
Keywords: Homebound older adults, social engagement, social group activities, depression, principal component analysis, validation
Introduction
Medicare considers someone homebound if she/he cannot leave home without assistance and typically cannot leave home (Centers for Medicare and Medicaid Services, 2018). A study of community-dwelling Medicare beneficiaries aged 65+ estimated that between 2011and 2017, 8.3% were persistently homebound, never or rarely (i.e., once a week or less) leaving their home in the preceding month, and 26.2% had a rapid increase in their risk of being homebound over the 7-year period (Xiang et al., 2020). As the numbers of older adults increase, the numbers of disabled, homebound older adults are projected to continue to increase.
Homebound older adults have lower levels of social engagement as disability, especially mobility limitations, makes participation in social activities outside the home more difficult (Rosso et al., 2013; Meek et al., 2018). Given their high medical burden, homebound older adults also suffer from depression at a significantly higher rate than their mobile peers (Musich et al., 2015; Ornstein et al., 2015). In a study of more than 2,200 largely low-income, racial/ethnic minority older adults (aged 50+) who were recipients of home-delivered meals, 38% reported a depression diagnosis (Choi et al., 2019). Depressed mood, anhedonia, and decreased energy in late life depression tend to decrease motivation and increase apathy for social engagement and activities, exacerbating social isolation among homebound older adults (Nelson et al., 2005; Xiang et al., 2020). Disability, depression, and lack of social engagement are likely to be mutual reinforcers for many homebound older adults (Rosso et al., 2013).
Social engagement is defined as participation in either leisure or productive activities that would reinforce social ties and meaningful social roles, which in turn can increase access to resources, enhance resilience, and protect against further mobility declines and disability onset in late life (Berkman et al., 2000; Glass et al., 2006; Okura et al., 2018). Research has shown that higher levels of social engagement are associated with better overall health, greater health-related quality of life, lower rates of loneliness and depression, higher perceived social support, and better cognitive function among older adults (Cherry et al., 2013; Glass et al., 2006; Hajek et al., 2017; James et al., 2011; Krueger et al., 2008; McHugh Power et al., 2019). On the other hand, lower levels of social engagement have been shown to be associated with lower quality of life and perceived social support, higher rates of loneliness and depressive symptoms, and higher rates of disability and mortality in both cross-sectional and longitudinal studies (Bennett, 2002; Buchman et al., 2009; Herbolsheimer et al., 2018; Mendes de Leon et al., 2003; McHugh Power et al., 2019; Okura et al., 2018).
For homebound older adults who tend to be socially isolated, increased social engagement, including visiting family and friends, attending religious services, going out for enjoyment, and attending clubs, can be especially beneficial for their physical and mental health and overall quality of life. However, along with their physical and mental health problems, a lack of transportation often limits their mobility (Szanton et al., 2016). Frequent healthcare visits (e.g., doctor’s appointment, hospitalization, rehabilitation) also take time and energy away from their social engagement (Meek et al., 2018). Among low-income homebound older adults who live in resource-poor, low-income neighborhoods, lack of transportation and financial resources pose even greater barriers to social engagement in beyond their immediate, restrictive living spaces. A study of frail older adults also found that the deficits in their physical (e.g., traffic, unsafe neighborhood, lack of grocery stores) and social (e.g., lack of social networks, recreational/sports facilities, clubs, meeting places) environments were significant factors influencing their social participation (Duppen et al., 2019). Thus, disabled, homebound older adults who are socioeconomically disadvantaged face multiple barriers to social engagement, especially activities outside the home. However, little research has been done on the kind of activities that low-income homebound older adults engage in and their associations with these older adults’ physical, functional, and mental health status.
With growing numbers of homebound older adults, it is important to assess their social engagement. A social engagement measure is also needed as a moderator, mediator, or outcome in clinical interventions that are designed to improve these older adults’ physical and mental health. Given the high rates of depression among homebound older adults, we were especially interested in measuring the level of social engagement and activities in relation to their depression as well as perceived social support and disability. We reviewed the literature attempting to locate social engagement measures for homebound older adults to no avail. Thus, based on our clinical and research experience with low-income, depressed homebound older adults, we developed a 10-item measure—Social Engagement and Activities Questionnaire or SEAQ—that covers a range of social-group activities, interpersonal interaction activities with family and friends, and solitary activities while taking account of their physical and functional health limitations. The 10 items, with follow-up open-ended questions that are intended to identify the type of activities that these older adults engage, were selected through extensive reviews by a team of experts including four PhD gerontologists who have done research on homebound older adults and a dozen case managers of homebound older adults.
The purposes of this paper were to evaluate: (1) the face validity of the SEAQ by categorizing the activities that respondents reported in open-ended questions and examine whether respondents were interpreting the questionnaire as intended; (2) the underlying dimensions of the SEAQ to reduce the ten items to one or more parsimonious variables that can be used in research endeavors; and (3) evidence of validity for the SEAQ by examining relationships between the SEAQ and older adults’ clinical characteristics, including depressive symptoms, perceived social support, and disability. We anticipated that social engagement and activities measured on the SEAQ would be positively associated with perceived social support scores and would be negatively associated with depressive symptom and disability scores. The SEAQ is the first measure specifically developed for low-income, homebound older adults. A validated social engagement/activities scale for growing numbers of low-income, homebound older adults will be an important assessment tool for research on these older adults.
Method
Recruitment and Participants
The subjects were participants in a study of the clinical effectiveness of a short-term, tele-delivered depression treatment. The majority (>95%) of the 277 participants were referred to the study by case managers at a home-delivered meals program in central Texas over a 38-month period (February 2016 – April 2019); the remaining subjects were referred by other aging-service agencies and primary care clinics that serve low-income, homebound adults. Study inclusion criteria included the following: (a) 50+ years of age, (b) homebound status as defined in Medicare, (c) moderate to severe depression defined by scores of ≥ 15 on the 24-item Hamilton Depression Rating Scale (HAMD; Depression Rating Scale Standardization Team, 2003); and (d) willingness to participate. Exclusion criteria were a high suicide risk, probable dementia, bipolar disorder, psychotic disorder, or substance abuse. Participants provided a written informed consent, approved by the authors’ university institutional review board, after the study was explained. All data presented herein were collected at baseline in the subject’s home by trained assessors prior to participation in the clinical trial. Principal component analysis (PCA), the primary analytic technique used in this paper, requires complete cases; thus, we excluded cases missing on one or more SEAQ item using listwise deletion, which is an acceptable missing data treatment when missing data are trivially small (Allison, 2010) as was the case in the present study (2.9% of the participants were excluded under listwise deletion). Among the 277 participants that completed the baseline assessment, 269 cases had complete data on the 10-item SEAQ questionnaire and were included in the analyses presented herein.
Measures
Social engagement/activities.
The SEAQ measures the frequency during the past month (0 = not at all, 1 = just one time, 2 = 2-3 times, 3 = once a week, 4 = more than once a week, and 5 = every day) of the following: (1) religious service attendance; (2) going outside the home (not counting religious service attendance and doctor’s appointment); (3) getting together socially with family, friends or relatives; (4) engagement in recreational activities for fun and relaxation; (5) engagement in gentle or vigorous exercise as a group activity; (6) attendance in meetings of any organized group that are not political in nature; (7) participation in any self-enrichment or educational activities; (8) engagement in activities that are political or social justice in nature; (9) volunteering for religious, charitable, political, health-related, or other organizations including one’s own apartment complex; and (10) informal volunteering for neighbors, friends, or family members (refer to a copy of the SEAQ in the appendix). The SEAQ is scored by adding up the score on each item, yielding a total score ranging from 0 (no engagement in any of the ten activities) to 50 (daily engagement in all ten activities). In order to provide more contextual information, the participants were asked open-ended questions regarding the types of activities that they engaged, if at all, for items #2 (going outside the house), #4 (recreational activities), #6 (attendance in nonpolitical organization meetings), #7 (self-enrichment or educational activities), #8 (political or social justice activities), #9 (formal volunteering), and #10 (informal volunteering). We provided examples of these activities to describe the type of the activities that can be included and facilitate the participants’ recollections. Responses to the open-ended questions were first manually coded and then confirmed with the SPSS word search function to generate specific activity categories and calculate the frequency of each category.
Depressive symptoms.
The 24-item HAMD consists of the GRID-HAMD-21 structured interview guide (Depression Rating Scale Standardization Team, 2003) with three additional items that assess feelings of hopelessness, helplessness, and worthlessness. According to Moberg et al. (2001), these three cognitive symptoms are thought to be more sensitive to depression in older than younger adults.
Perceived social support.
We used the 12-item Multidimensional Scale of Perceived Social Support (MSPSS; Zimet, Dahlem, Zimet, & Farley, 1988) to assess participants’ perceived social support from family (e.g., My family really tries to help me), friends (e.g., I can count on my friends when things go wrong), and significant others (e.g., There is a special person in my life who cares about my feelings). The MSPSS uses a 7-point scale measuring the extent to which respondents agreed with each statement (1 = very strongly disagree; 7 = very strongly agree), with higher scores indicating higher perceived support. Cronbach’s α for the study participants was 0.93. The MSPSS has been validated for older adults (aged 55-82) with a psychiatric disorder and was found to have strong psychometric properties that warrant its usefulness for evaluating perceived adequacy of social support in older adults (Stanley et al., 1998).
Disability.
We assessed disability using the 12-item World Health Organization Disability Assessment Schedule (WHODAS 2.0), which covers degree of difficulty (0 = none; 4 = extreme/cannot do) in six domains of functioning: (a) cognition (learning a new task & concentrating); (b) mobility (standing and walking); (c) self-care (bathing, dressing, emotional state); and (d) getting along (dealing with people that you do not know; maintaining a friendship); (e) life activities (doing housework and other household responsibilities); and (f) participation (joining) (WHO, 2018). Because the in community activities [e.g., festivities, religious or other activities] in the same way as anyone else can were homebound older adults, the last item was revised from “your day-to-day work/school” to “your day-to-day housework in and around the house.” Cronbach’s α for the study participants was .83.
Statistical Analysis
The items comprising the SEAQ were screened prior to data analysis. Although a cutoff criterion of skew < 2 and kurtosis < 7 (Curran et al., 1996) was employed to identify cases that deviated from univariate normality, the skew and kurtosis criterion alone was not used to exclude variables from the analysis. Skewed distributions have the potential to form difficulty factors (e.g., artificial factors; Gibson, 1959) that are a result of their similarity in skew and kurtosis due to an underlying construct. We used PCA to descriptively summarize variables. We did not use statistical inference tests with multivariate normality assumptions (Tabachnick & Fidell, 2012). Nevertheless, skew and kurtosis have the potential to degrade PCA solutions, and we therefore created two sets of variables. The first set used the full SEAQ and the second set contained only the items that met the data screening criteria, henceforth referred to as the abbreviated SEAQ.
The items underlying the SEAQ represent behavioral assessments that can be evaluated relatively objective (i.e., they are tangible, easily rememberable behaviors). Considering this, PCA, which partitions all variance into underlying components, is appropriate to evaluate the number of dimensions as opposed to other data reduction techniques, such as exploratory factor analysis, that typically use items to indirectly measure latent constructs. PCA constructs weighted composites of the variables in the model by partitioning the total variance into components. PCA was implemented using the princomp function from the R stats package (R Core Team, 2019; version 1.8.12) using R version 3.6.1. to determine the number of components underlying the item sets, we used parallel analysis, implemented using the R psych package (Revelle, 2018). Parallel analysis was conducted on both the full and abbreviated item sets.
At the completion of the PCA, the components were correlated with the full and abbreviated SEAQ to assess the extent to which components were related to the original items. We evaluated convergent validity by correlating the components with perceived social support. We evaluated discriminant validity with correlations between the components and depressive symptoms and disability. In addition, we evaluated the key validation relationships between participants that lived alone (n = 134) with participants that lived with another person (n = 135) by comparing covariance matrices between the two groups in a structural equation modeling framework using the R lavaan package (Rosseel, 2012). We included the full SEAQ components with the validation constructs, perceived social support, disability, and depressive symptoms.
Results
Participant Characteristics
Participants averaged 67.4 (SD = 8.8) years old; 69.9% were female; 41.3% were non-Hispanic White, 29.7% non-Hispanic Black, and 29.0% Hispanic; 22.7% were married; 50.2% lived with somebody; 57.3% had at least some college; and 53.9% had a household income of $15,000 or less (Table 1). The mean scores show that they had moderate level of perceived social support and moderate-to-severe levels of disability and depressive symptoms. The overall SEAQ score indicates a low level of engagement and activity. We used the total score for the MSPSS as its three subscale (family, friend, and significant other support) scores were highly correlated.
Table 1.
Sociodemographic and clinical characteristics of participants at baseline (N=269)
| Sociodemographic characteristics | % (n) / M (SD) |
|---|---|
| Age (yrs) | 67.4 (8.8) |
| Race/ethnicity | |
| Non-Hispanic White | 41.3% (111) |
| Non-Hispanic Black | 29.7% (80) |
| Hispanic/other | 29.0% (78) |
| Gender | |
| Male | 30.1% (81) |
| Female | 69.9% (188) |
| Education | |
| < High school | 26.0% (70) |
| High school | 16.7% (45) |
| Some college | 23.4% (63) |
| Associate degree | 10.0% (27) |
| Bachelor's degree or higher | 23.8% (64) |
| Marital status | |
| Married | 22.7% (61) |
| Widowed | 26.0% (70) |
| Divorced/separated | 37.5% (101) |
| Never married | 13.8% (37) |
| Living situation | |
| Living with somebody | 50.2% (135) |
| Living alone | 49.8% (134) |
| Household income | |
| Up to $15,000 | 53.9% (145) |
| $15,001-25,000 | 26.8% (72) |
| $25,001-35,000 | 11.5% (31) |
| $35,001 or higher | 7.8% (21) |
| Self-rated financial status | |
| I really can't make ends meet | 21.9% (59) |
| Just about manage to get by | 61.7% (166) |
| I have enough to get along, and even a little extra | 14.9% (40) |
| Money is not a problem. I can buy pretty much anything I want. | 1.5% (4) |
| Clinical characteristics | |
| Perceived social support. | 53.1 (18.1) |
| Disability | 23.4 (9.1) |
| Depressive symptoms | 23.0 (5.7) |
| Social engagement/activities | 12.2 (6.1) |
Proportion of Participants Reporting SEAQ Activities and Type of Activities
More than three-quarters of participants reported getting out of home to shop (mostly groceries), visit family, and run errands; socialization with family and friends; and engagement in leisure/recreational activities (mostly solitary) in the preceding month (Table 2). A little more than one-third reported religious service attendance, engagement in self-enrichment/educational activities (mostly at home, involving online or phone-based), and informal volunteering (helping neighbors and giving advice to younger generations). Only 7% reported engagement in any group exercise; 17% reported attendance in any nonpolitical organization meetings (mostly Bible study and church-related); 12% reported engagement in political/social justice activities (letter-writing to elected officials, watching political programs on TV); and 7% reported formal volunteering (at church, apartment complex, and other community venues). The types of activities that the participants reported are consistent with what the authors expected when constructing the questions and thus provide evidence of face validity.
Table 2.
Proportion of participants reporting any activity for each SEAQ item and type of activities for selected items
| Item | Activity description (in the past month) | n (%) of 1+ times |
|---|---|---|
| 1 | How often did you attend religious services? | 92 (34.2%) |
| 2 | Not counting religious service and doctor’s appointment, how often did you get out of your apartment building or house and go to other locations? | 215 (79.9%) |
| Shopping | 130 (48.3%) | |
| Visiting family | 55 (20.4%) | |
| Visiting friends for fun/support | 30 (11.2%) | |
| Picking up medications, going to bank, laundromat… | 40 (14.9%) | |
| Walking, walking a dog, going to park, gym… | 18 (6.7%) | |
| Going to movies/shows, eating out, playing games | 45 (16.7%) | |
| Going to senior centers or community cents | 7 (2.6%) | |
| 3 | How often did you get together socially with family, friends, or relatives? | 209 (77.7%) |
| 4 | How often did you engage in any recreational activities for fun or relaxation? | 232 (86.2%) |
| Gardening, going to park, exercising, meditation | 13 (4.8%) | |
| Playing with a pet | 31 (11.5%) | |
| Listening to music | 52 (19.3%) | |
| Playing computer and board games, crossword puzzles | 41 (15.2%) | |
| Arts and crafts, crocheting | 26 (9.7%) | |
| Reading/audio books | 50 (18.6%) | |
| Watching movies, news, sports, soap operas on TV | 144 (53.5%) | |
| Other (going to hair/nail salon, bookstore, bingo and pool games, movie theater; sitting outside to get fresh air) | 18 (6.7%) | |
| 5 | How often did you do gentle or vigorous exercise as a group activity? | 19 (7.1%) |
| 6 | How often did you attend meetings of any organized group that are not political in nature? | 45 (16.7%) |
| Bible study, church & religious meeting | 30 (11.2%) | |
| Apartment complex resident & other meeting | 6 (2.2%) | |
| Hobby club (writing, music, gardening) | 6 (2.2%) | |
| Other (Alcoholics Anonymous, support group for those with post-traumatic stress disorder at a Veterans Affairs clinic) | 3 (1.1%) | |
| 7 | How often did you participate in any self-enrichment program or education? | 95 (35.3%) |
| Online search, listening to TED talk, podcast, reading print materials | 77 (28.6%) | |
| Learning how to use Internet and smart phone | 6 (2.2%) | |
| Going to meetings to learn about programs/resources | 12 (4.5%) | |
| 8 | How often did you engage in activities that are political or social justice in nature? | 32 (11.9) |
| Letter writing, signing petition, voter registration, political rally | 18 (6.7%) | |
| Watching political TV programs & debates, voting | 14 (5.2%) | |
| 9 | How often did you do volunteer work for religious, charitable, political, health-related, or other organizations including apartment complex? | 19 (7.1%) |
| For church or other religious events | 10 (3.7%) | |
| For apartment residents’ council, community center, food pantry | 14 (5.2%) | |
| 10 | How often did you do informal volunteer work for neighbors, friends, or family members? | 102 (37.9%) |
| Family/friend caregiving, babysitting, giving advice for younger generation | 70 (26.0%) | |
| Helping, taking care of, and/or giving advice to neighbors, homeless people | 32 (11.9%) |
Correlation Matrix and PCA Results
The correlation matrix for the 10 SEAQ items (Table 3) shows that religious service attendance was moderately significantly correlated with attendance in nonpolitical meetings and formal volunteering, and two latter activities were moderately correlated with each other, indicating that these three social-group activities are religion-related. Overall, group activities outside the home were correlated with one another, while mostly in-home, solitary activities (i.e., recreational activities and self-enrichment/educational activities) were correlated with each other but not with social-group activities.
Table 3.
Full SEAQ item correlations
| 1. | 2. | 3. | 4. | 5. | 6. | 7. | 8. | 9. | |
|---|---|---|---|---|---|---|---|---|---|
| 1. Religious service attendance | |||||||||
| 2. Get out of home | .19** | ||||||||
| 3. Socialization with family, friends, or relatives | .18** | .27*** | |||||||
| 4. Recreational activities | −.03 | .07 | .06 | ||||||
| 5. Group exercise | .07 | −.05 | .06 | −.00 | |||||
| 6. Attendance in nonpolitical meetings | .45*** | .15* | .15* | .04 | −.05 | ||||
| 7. Self-enrichment program or education | .02 | .08 | .15* | .23*** | −.00 | .06 | |||
| 8. Engagement in political/social justice activities | .11 | .13* | .13* | .02 | −.05 | .26*** | .11 | ||
| 9. Formal volunteering. | .30*** | .11 | .06 | −.02 | .01 | .41*** | −.00 | .18** | |
| 10. Informal volunteering | .16** | .16** | .14* | −.02 | −.04 | .21*** | .20*** | .23*** | .19** |
p < .05
p <.01
p < .001
As shown in Table 4, four items with low frequency activities (any group exercise, attendance in nonpolitical meetings, engagement in political/social justice activities, and formal volunteering) fell below cutoff criteria for both skewness and kurtosis. These four items were dropped in the abbreviated SEAQ. Parallel analysis of both the full SEAQ (ten items) and abbreviated SEAQ (six items) indicated that two components exceeded values from simulated and resampled data. Thus, we used only the first two components from the full and abbreviated PCAs in subsequent validation evaluations. The first component in the full PCA had relatively high coefficients for most items on the scale. The primary exceptions were any group exercise, recreational activities, and self-enrichment/educational activities. The second component was driven by two strong negative coefficients on the recreational activities and self-enrichment/educational activities as well as a negative coefficient for socializing with friends and family (interpersonal interaction activities). Thus, we consider the first component to be a general social-group activities engagement component, and the second component to capture a low level of socialization (i.e., participants with high scores on this component reported little social-group activity).The religious service attendance item loaded approximately equally across both the first and second components. The components in the abbreviated SEAQ PCA largely paralleled the full SEAQ PCA. In the abbreviated SEAQ, all items, except for recreational activities, loaded on the first component, and the second component had a positive coefficient for religious services and strong negative coefficients for recreational activities and self-enrichment or educational activities.
Table 4.
SEAQ items descriptive statistics
| Full SEAQ coefficients of first two components |
Abbreviated SEAQ coefficients of first two components |
||||||||
|---|---|---|---|---|---|---|---|---|---|
| SEAQ item: In the past month, | M | SD | skew | kurtosis | % > 0 | 1 | 2 | 1 | 2 |
| 1. Religious service attendance | 0.90 | 1.37 | 1.12 | −0.28 | 34% | 0.42 | 0.27 | 0.38 | 0.44 |
| 2. Get out of home | 2.26 | 1.57 | 0.00 | −1.12 | 80% | 0.31 | −0.22 | 0.49 | 0.18 |
| 3. Socialization with family, friends, or relatives | 2.01 | 1.55 | 0.32 | −0.95 | 78% | 0.29 | −0.29 | 0.50 | 0.11 |
| 4. Recreational activities | 3.67 | 1.80 | −1.11 | −0.27 | 86% | 0.06 | −0.51 | 0.20 | −0.68 |
| 5. Group exercise | 0.20 | 0.79 | 4.29 | 18.18 | 7% | −0.02 | 0.04 | ||
| 6. Attendance in nonpolitical meetings | 0.40 | 1.01 | 2.51 | 5.11 | 17% | 0.48 | 0.24 | ||
| 7. Self-enrichment program or education | 1.17 | 1.80 | 1.18 | −0.22 | 35% | 0.18 | −0.58 | 0.38 | −0.55 |
| 8. Engagement in political/social justice activities | 0.29 | 0.94 | 3.66 | 13.58 | 12% | 0.33 | −0.07 | ||
| 9. Formal volunteering. | 0.15 | 0.63 | 4.84 | 25.55 | 7% | 0.39 | 0.34 | ||
| 10. Informal volunteering | 1.17 | 1.71 | 1.10 | −0.35 | 38% | 0.35 | −0.13 | 0.43 | 0.09 |
Convergent and Discriminant Validity
Table 5 presents a correlation matrix of the full SEAQ score, abbreviated SEAQ score, perceived social support, disability, depressive symptoms, and the general social-group activities engagement and the low level of socialization components from the full and abbreviated SEAQ scales. The general social-group activities engagement in both the full and abbreviated SEAQ were significantly positively correlated with the full SEAQ and the abbreviated SEAQ and perceived social support, providing evidence for convergent validity, and they were significantly negatively correlated with disability and depressive symptoms, providing evidence for discriminant validity. The low level of socialization component in both the full and abbreviated SEAQ were significantly negatively correlated with the full SEAQ and the abbreviated SEAQ, and neither of them were significantly correlated with perceived social support, disability, or depressive symptoms. A comparison of covariance matrices between participants that lived alone and those that lived with another person using the key validation variables in Table 5 (the full SEAQ components, perceived social support, disability, and depressive symptoms) indicated that these groups did not systematically differ in these relationships (X2[10] = 14.62, p = .147).
Table 5.
Correlations matrix for SEAQ scale totals, SEAQ principal components, and other scale scores
| Full SEAQ |
Abbreviated SEAQ |
Full SEAQ general social-group activities engagement component |
Full SEAQ low level of socialization component |
Abbreviated SEAQ general social-group activities engagement component |
Abbreviated SEAQ low level of socialization component |
Perceived social support |
Disability | |
|---|---|---|---|---|---|---|---|---|
| Abbreviated SEAQ | .95*** | |||||||
| Full SEAQ component 1 | .88*** | .75*** | ||||||
| Full SEAQ component 2 | −.42*** | −.59*** | −.00 | |||||
| Abbreviated SEAQ component 1 | .93*** | .97*** | .80*** | −.48*** | ||||
| Abbreviated SEAQ component 2 | −.12* | −.21*** | .22*** | .74*** | −.00 | |||
| Perceived social support | .21*** | .22*** | .19** | −.05 | .24*** | .07 | ||
| Disability | −.20*** | −.17** | −.21*** | −.01 | −.20** | −.12 | −.11 | |
| Depressive symptoms | −.20*** | −.20*** | −.19** | .08 | −.21*** | −.05 | −.29*** | .35*** |
p < .05
p <.01
p < .001
Discussion
Our analysis provides evidence for the SEAQ’s face validity: responses to open-ended questions about specific activities as follow-up to individual items were consistently the type of activities that the items were intended to capture. Both PCA models also indicated that data reduction was viable for the SEAQ as its items could be parsimoniously combined. Given the similarities between the first and second components (i.e., general social-group activities engagement component and a low level of socialization component, respectively) across the full and abbreviated SEAQ, subsequent discussion will treat the general social-group activities engagement components collectively and will treat the low level of socialization components collectively. The general social-group activities engagement components in both the full and abbreviated SEAQ PCA models exhibited moderate correlations in the predicted direction with each of the validation measures, perceived social support, disability, and depressive symptoms: higher social engagement was associated with higher perceived social support and lower disability and depressive symptoms. The positive correlations thus provide evidence of convergent validity and the negative correlations thus provide evidence of discriminant validity. In contrast, the low level of socialization components exhibited minimal correlations with the validation variables. While initially surprising we believe that the lack of correlation here is because the most prevalent behavior among the participants, engaging in recreational activities, contributed to this component, resulting in a ceiling effect (i.e., nearly all respondents endorsed this activity at a relatively high frequency and the result was that it doesn’t distinguish between individuals).
Because only the general social-group activities engagement components in the PCAs provided evidence of validity, we recommend using this component for research purposes. The very strong correlations between the general social-group activities engagement components and the scale composite score is also an indication that the scale score itself can be used in lieu of principal components. In addition, the correlations between the scale scores and the three validation variables are nearly identical to the correlations between the components and the three validation variables further indicating that they are relatively interchangeable. Furthermore, there was little difference between the full SEAQ and the abbreviated SEAQ in their relationships with validation variables, indicating that either could be used as a composite measure of social engagement.
The analysis presented herein represents the first attempt at data reduction using the SEAQ items among low-income, depressed homebound older adults. While the analysis provides evidence supporting the use of either the full or abbreviated SEAQ, a couple of limitations need to be noted. First, the data screening identified a couple of items (i.e., group exercise and formal volunteering) that exhibited very low behaviors in the present sample which may be a function of respondents’ low mobility and social isolation due to their homebound state and depression. Second, as with the first limitation, the nature of the sample may have contributed to the unidimensionality of the scale. It is possible that multiple components may be necessary in nondepressed samples. Third, the sample was selected from a geographically limited area, i.e., Central Texas, where even winter weather conditions do not pose significant barriers to going outside the home (e.g., going to park). Type of activities among older adults in other regions with more inclement weather conditions are likely to differ substantially. Thus, the SEAQ’s face validity needs to be reexamined for these older adults. Considering these limitations, the SEAQ should be evaluated in a range of older adult samples to assess the type and prevalence of the activities and their generalizability to more mobile and/or nondepressed older adult samples in different geographic areas.
The extant literature supports the need to assess social engagement in older adults. As discussed, higher levels of social engagement are associated with better physical, functional, mental and cognitive health and health-related quality of life among older adults (Cherry et al., 2013; Hajek et al., 2017; James et al., 2011). On the other hand, lower levels of social engagement are associated with higher rates of disability and depression, lower perceived social support and quality of life, and higher rates of mortality (Herbolsheimer et al., 2018; Okura et al., 2018). Research thus indicates that older adults’ engagement in a range of social-group activities, interpersonal interactions with family and friends, and solitary activities should be regularly assessed in conjunction with physical, functional, mental, and cognitive health outcomes. Our findings show that the SEAQ is a potentially useful and valid tool to assess social engagement and activities among homebound older adults, especially those who are low-income and depressed.
The findings also indicate a need for helping socially isolated, depressed older adults engage in more social-group activities at social settings to enhance their social support and reduce their disability and depression. Activities at or related to religious organizations appear to be a source of great social engagement as these activities tend to occur at social-group settings. Attendance in any kind of meetings and formal volunteering also appear to be a great way to promote engagement in social-group activities among these older adults. Although our data showed that these social-group activities were not as frequent as interpersonal-interaction activities with family and/or friends, they may be an important source of meaning in life for the study participants. Interestingly, a recent study of largely non-Hispanic White and highly educated older adults (n=48) with major depressive disorder who participated in a behavioral activation treatment found interpersonal-individual activities with a specific family member or friend, but not social-group or solitary activities, predicted subsequent improvement of depression and increase in behavioral activation (Solomonov et al., 2019). The difference between the study and our study likely stems from the fact that our study participants faced greater barriers to social-group activities because of their homebound state. Their lower socioeconomic status was also likely to have restricted assessable and affordable social participation opportunities (Goll et al., 2015). Thus, social-group activities are more likely to be valued for these vulnerable older adults and were associated with higher perceived social support and lower disability and depressive symptoms.
In order to increase their social engagement, these older adults need to be provided occasions for out-of-home activities and means (e.g., special transportation services) to get there. For mobility-impaired older adults, group exercise programs or other social group activities in their apartment complex should be available and these older adults need to be educated about the benefits of group exercise and other group activities. Neighborhood physical (e.g., walkability) and social environments have also been found to be significant contributors to older adults’ physical activity (Chaudhury et al., 2016). Benefits of volunteering among low-income, racial/ethnic minority older adults in terms of physical and mental health and social connections have been well documented (Morrow-Howell et al., 2014). Disabled, depressed low-income older adults need increased opportunities for formal volunteering at religious, political, and nonpolitical organizations and causes.
In conclusion, the SEAQ was developed as a brief screen of social engagement for low-income, depressed homebound older adults. The SEAQ’s brevity and acceptable validity suggests that it will be useful for measuring social engagement/activities of low-income, depressed homebound older adults in population-based studies or clinical trials for disability, depression, or social support. As the general social-group activities engagement components in the PCAs—mostly social-group activities—provided evidence of validity, we recommend using this component for research purposes.
Clinical Implications.
The SEAQ can be used to better understand social engagement and activities among depressed, low-income homebound older adults, a growing segment of the aging population in both research projects and aging-service agencies.
The SEAQ can be used in intervention planning to identify areas for increasing social engagement in older adults with disability and depression and evaluating intervention effectiveness.
Further research is needed to examine the SEAQ’s potential as a valid assessment tool across a broad range of older adults who may be also at risk of social isolation.
Acknowledgments
Funding
The parent study was funded by the National Institute on Minority Health and Health Disparities (1R01MD009675; PI: N. Choi).
Appendix. 10-item SEAQ
These next questions are to find out about your level of social engagement and activities. Please listen each question carefully when I read it to you and tell me how frequently you are engaged in the activity.
Q1. In the past month, how often did you attend religious services? (IF HESITANT READ LIST)
| Not at all | Just one time | 2-3 times | One a week | More than once a week | Everyday |
Don't know………………….9 Refused………………..8
Q2. In the past month, not counting religious service and doctor’s appointment, how often did you get out of your apartment or house and go to other locations outside your apartment building or house? (Examples include: shopping trips, visiting friends/family/neighbors)
| Not at all | Just one time | 2-3 times | Once a week | More than once a week | Everyday |
Don't know………………….9 Refused………………..8
If done any such activity, what did you do?_____________________________________
Q3. In the past month, how often did you get together socially with family, friends or relatives?
| Not at all | Just one time | 2-3 times | Once a week | More than once a week | Everyday |
Don't know………………….9 Refused………………..8
Q4. In the past month, how often did you engage in any recreational activities for fun and relaxation? (Examples include: watching movies/videos, playing or listening to music, dancing, going to a park, playing board, card or computer games, dominos, or other similar games, doing crossword puzzles, arts and crafts, crocheting, playing with a pet, reading or writing)
| Not at all | Just one time | 2-3 times | Once a week | More than once a week | Everyday |
Don't know………………….9 Refused………………..8
If done any such activity, what did you do?_____________________________________
Q5. In the past month, how often did you do gentle or vigorous exercise as a group activity? (Examples include: walk for exercise, aerobics, yoga, swimming)
| Not at all | Just one time | 2-3 times | Once a week | More than once a week | Everyday |
Don't know………………….9 Refused………………..8
Q6. In the past month, how often did you attend meetings of any organized group that are not political in nature? (Examples include: a bible study, a choir, a committee, board, or resident council, a support group, a sports or exercise group, or a professional society.)
| Not at all | Just one time | 2-3 times | Once a week | More than once a week | Everyday |
Don't know………………….9 Refused………………..8
If attended any such meeting, what kind of meeting(s) did you attend?
___________________________________________________________________________
Q8. In the past month, how often did you engage in activities that are political or social justice in nature? (Examples include: calling/sending emails or mails to elected officials, signing a petition, attending a lobby group meeting, attending a rally, NAACP meetings, Latino/Latina caucus, Juneteenth commemorative parade)
| Not at all | Just one time | 2-3 times | Once a week | More than once a week | Everyday |
Don't know………………….9 Refused………………..8
If done any such activities, what did you do?_____________________________________
Q7. In the past month, how often did you participate in any self-enrichment program or education? (Examples include: attending lecture, informational session, or presentation on and off apartment complex, going to a play or concert, visiting a museum, taking a class, learning something new or researching online.)
| Not at all | Just one time | 2-3 times | One a week | More than once a week | Everyday |
Don't know………………….9 Refused………………..8
If done any such program, what did you do?_____________________________________
Q9. In the past month, how often did you do volunteer work for religious, charitable, political, health-related, or other organizations including your apartment complex?
| Not at all | Just one time | 2-3 times | One a week | More than once a week | Everyday |
Don't know………………….9 Refused………………..8
If done any formal volunteering, what did you do?_____________________________________
How many total hours of formal volunteering did you do? ______________________________
Q10. In the past month, how often did you do informal volunteer work for neighbors, friends, or family members (Examples include: calling, visiting, and/or bringing food or picking up mails for sick/disabled family or friends; dispensing advices to younger generations)?
| Not at all | Just one time | 2-3 times | One a week | More than once a week | Everyday |
Don't know………………….9 Refused………………..8
If done any informal volunteering, what did you do?__________________________________
How many total hours of informal volunteering did you do? ________________________
Footnotes
Disclosure statement
No conflict of interest was reported by the authors.
References
- Allison PD (2010). Missing data. Thousand Oaks, CA: Sage. [Google Scholar]
- Bennett KM (2002). Low level of social engagement as a precursor of mortality among people in late life. Age and Ageing, 31, 165–168. [DOI] [PubMed] [Google Scholar]
- Berkman LF, Glass T, Brissette I, & Seeman TE (2000). From social integration to health: Durkheim in the new millennium. Social Science & Medicine, 51, 843–857. [DOI] [PubMed] [Google Scholar]
- Buchman AS, Boyle PA, Wilson RS, Fleischman DA, Leurgans S,& Bennett DA (2009). Association between late-life social activity and motor decline in older adults. Archives of Internal Medicine, 169, 1139–1146. doi: 10.1001/archinternmed.2009.135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Centers for Medicare and Medicaid Services. (2018). MLN matters. https://www.cms.gov/Outreach-and-Education/Medicare-Learning-Network-MLN/MLNMattersArticles/Downloads/SE1436.pdf
- Chaudhury H, Campo M, Michael Y, & Mahmood A (2016). Neighbourhood environment and physical activity in older adults. Social Science & Medicine,149,104–113. doi: 10.1016/j.socscimed.2015.12.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cherry KE, Walker EJ, Brown JS, Volaufova J, LaMotte LR, Welsh DA, Su LJ, Jazwinski SM, Ellis R, Wood RH, Frisard MI (2013). Social engagement and health in younger, older, and oldest-old adults in the Louisiana Healthy Aging Study. Journal of Applied Gerontology, 32, 51–75. doi: 10.1177/0733464811409034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Choi NG, Sullivan JE, & Marti CN (2019). Low-income homebound older adults receiving home-delivered meals: Physical and mental health conditions, incidence of falls and hospitalizations. Health and Social Care in the Community, 27(4), e406–e416. doi: 10.1111/hsc.12741. [DOI] [PubMed] [Google Scholar]
- Curran PJ, West SG & Finch JF (1996). The robustness of test statistics to nonnormality and specification error in confirmatory factor analysis. Psychological Methods, 1, 16–29. doi: 10.1037//1082-989X.1.1.16. [DOI] [Google Scholar]
- Depression Rating Scale Standardization Team (2003). GRID-HAMD-17, GRID-HAMD-21 structured Interview Guide. International Society for CNS Drug Development, San Diego, CA. [Google Scholar]
- Duppen D, Lambotte D, Dury S, Smetcoren AS, Pan H, De Donder L, & D-SCOPE Consortium. (2019). Social participation in the daily lives of frail older adults: Types of participation and influencing factors. Journal of Gerontology B: Psychological and Social Sciences. pii: gbz045. doi: 10.1093/geronb/gbz045. [Epub ahead of print] [DOI] [PubMed] [Google Scholar]
- Gibson WA (1959). Three multivariate models: Factor analysis, latent structure analysis, and latent profile analysis. Psychometrika, 24, 229–252. [Google Scholar]
- Glass TA, Mendes de Leon CF, Bassuk SS, & Berkman LF (2006). Social engagement and depressive symptoms in late life: Longitudinal findings. Journal of Aging & Health, 18, 604–628. doi: 10.1177/0898264306291017. [DOI] [PubMed] [Google Scholar]
- Goll JC, Charlesworth G, Scior K, & Stott J (2015). Barriers to social participation among lonely older adults: The influence of social fears and identity. PLoS One,10(2):e0116664. doi: 10.1371/journal.pone.0116664. eCollection 2015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Herbolsheimer F, Ungar N, & Peter R (2018). Why is social isolation among older adults associated with depressive symptoms? The mediating role of out-of-home physical activity. International Journal of Behavioral Medicine, 25, 649–657. doi: 10.1007/s12529-018-9752-x. [DOI] [PubMed] [Google Scholar]
- Hajek A, Brettschneider C, Mallon T, Ernst A, Mamone S, Wiese B, Weyerer S, Werle J, Pentzek M, Fuchs A, … AgeCoDe Study Group. (2017). The impact of social engagement on health-related quality of life and depressive symptoms in old age: Evidence from a multicenter prospective cohort study in Germany. Health and Quality of Life Outcomes, 15, 140. doi: 10.1186/s12955-017-0715-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- James BD, Wilson RS, Barnes LL, & Bennett DA (2011). Late-life social activity and cognitive decline in old age. Journal of the International Neuropsychological Society, 17, 998–1005. doi: https://doi-org.ezproxy.lib.utexas.edu/10.1017/S1355617711000531 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Krueger KR, Wilson RS, Kamenetsky JM, Barnes LL, Bienias JL, & Bennett DA (2009). Social engagement and cognitive function in old age. Experimental Aging Research, 35, 1, 45–60. doi: 10.1080/03610730802545028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McHugh Power JE, Steptoe A, Kee F, & Lawler BA (2019). Loneliness and social engagement in older adults: A bivariate dual change score analysis. Psychology and Aging, 34, 152–162. doi: 10.1037/pag0000287. [DOI] [PubMed] [Google Scholar]
- Meek KP, Bergeron CD, Towne SD, Ahn S, Ory MG, & Smith ML (2018). Restricted social engagement among adults living with chronic conditions. International Journal of Environmental Research and Public Health, 15(1), pii: E158. doi: 10.3390/ijerph15010158. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mendes de Leon CF, Glass TA, & Berkman LF (2003) Social engagement and disability in a community population of older adults: The New Haven EPESE. American Journal of Epidemiology, 157, 633–642. doi: 10.1093/aje/kwg028. [DOI] [PubMed] [Google Scholar]
- Moberg PJ, Lazarus LW, Mesholam RI, Bilker W, Chuy IL, Neyman I & Markvart V (2001). Comparison of the standard and structured interview guide for the Hamilton Depression Rating Scale in depressed geriatric inpatients. The American Journal of Geriatric Psychiatry, 9, 35–40. [PubMed] [Google Scholar]
- Morrow-Howell N, Lee YS, McCrary S, & McBride A (2014). Volunteering as a pathway to productive and social engagement among older adults. Health Education & Behavior, 41(1 Suppl), 84S–90S. doi: 10.1177/1090198114540463. [DOI] [PubMed] [Google Scholar]
- Musich S, Wang SS, Hawkins K, & Yeh CS (2015). Homebound older adults: Prevalence, characteristics, healthcare utilization and quality of care. Geriatric Nursing, 36, 445–450. [DOI] [PubMed] [Google Scholar]
- Nelson JC, Clary CM, Leon AC, & Schneider LS (2005). Symptoms of late-life depression: Frequency and change during treatment. American Journal of Geriatric Psychiatry, 13, 520–526. [DOI] [PubMed] [Google Scholar]
- Okura M, Ogita M, Yamamoto M, Nakai T, Numata T, & Arai H (2018). Community activities predict disability and mortality in community-dwelling older adults. Geriatrics and Gerontology International, 18, 1114–1124. doi: 10.1111/ggi.13315. [DOI] [PubMed] [Google Scholar]
- Ornstein KA, Leff B, Covinsky KE, Ritchie CS, Federman AD, Roberts L, Kelley AS, Siu L, & Szanton SL (2015). Epidemiology of the homebound population in the United States. JAMA Internal Medicine, 175, 1180–1186. doi: 10.1001/jamainternmed.2015.1849. [DOI] [PMC free article] [PubMed] [Google Scholar]
- R Core Team (2019). R: A language and environment for statistical computing. R foundation for statistical computing. Retrieved from https://www.R-project.org/. [Google Scholar]
- Revelle W (2018). psych: Procedures for psychological, psychometric, and personality research. Northwestern University. Retrieved from http://CRAN.R-project.org/package=psych. [Google Scholar]
- Rosseel Y (2012). lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48, 1–36. doi: 10.18637/jss.v048.i02 [DOI] [Google Scholar]
- Rosso AL, Taylor JA, Tabb LP, & Michael YL (2013). Mobility, disability, and social engagement in older adults. Journal of Aging and Health, 25, 617–637. doi: 10.1177/0898264313482489. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Solomonov N, Bress JN, Sirey JA, Gunning F Flückiger C, Raue PJ, Areán PA, & Alexopoulos GS (2019). Engagement in socially and interpersonally rewarding activities as a predictor of outcome in “Engage” behavioral activation therapy for late-life depression. American Journal of Geriatric Psychiatry, 27, 571–578. doi: 10.1016/j.jagp.2018.12.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stanley MA, Beck JG & Zebb BJ (1998). Psychometric properties of the MSPSS in older adults. Aging and Mental Health, 2, 186–193. [Google Scholar]
- Szanton SL, Roberts L, Leff B, Walker JL, Seplaki CL, Soones T, Thorpe RJ Jr, & Ornstein KA (2016). Home but still engaged: Participation in social activities among the homebound. Quality of Life Research, 25, 1913–1920. doi: 10.1007/s11136-016-1245-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tabachnick B & Fidell L (2012). Using Multivariate Statistics. Boston, MA: Pearson. [Google Scholar]
- Xiang X, Chen J, & Kim M (2020). Trajectories of homebound status in Medicare beneficiaries aged 65 and older. The Gerontologist, 60, 101–111. doi: 10.1093/geront/gnz023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zimet GD, Dahlem NW, Zimet SG & Farley GK (1988). The multidimensional scale of perceived social support. Journal of Personality Assessment, 52, 30–41. [DOI] [PubMed] [Google Scholar]