Residential Context,Social Relationships,and Subjective Well-Being in Assisted Living

Abstract

Adaptive reframing and cumulative inequality theory provide the conceptual framework for investigating associations between four measures of subjective well-being (general, temporal, comparative, and experiential), residential context, and social relationships. Data from 344 cognitively intact assisted living (AL) residents aged 60 years and older interviewed for the Florida Study of Assisted Living were analyzed using logistic regression. Having control over the AL transition, often associated with socioeconomic status, was positively associated with all four dimensions of subjective well-being, consistent with a cumulative inequality framework. Other residential context characteristics (living arrangements prior to AL, private pay, size, licensure status) were less consistently associated with well-being. High-quality staff relationships were associated with temporal well-being, while positive coresident relationships were associated with all four well-being indicators. Compared with preexisting external relationships with family and friends, and consistent with adaptive reframing, social relationships unique to AL were independently and more consistently associated with residents’ perceptions of subjective well-being.

Keywords

well-being long-term care assisted living social support friendship cumulative inequality adaptive reframing

The number of older Americans residing in assisted living (AL) facilities has increased steadily in recent years, and as the industry matures, AL has become more complex and specialized (Street, Burge, & Quadagno, 2009). Leaving a private household and moving to AL is a turning point, a process fraught with change (Young, 1998). Whether the transition from a prior home to AL is regarded with anticipation, resignation, or dread varies from individual to individual. However, all who move must navigate a series of adjustments: finding a suitable AL facility, downsizing, settling in, devising ways to maintain existing social relationships and enacting strategies to establish new ones. The residential transition also triggers unfolding experiential trajectories for AL residents. How residents feel about their daily lives depends on numerous circumstances, including the individualized context surrounding transitions into AL, characteristics of the residential site, individual personality traits and functional capacities, and the daily routines of social relationships associated with life in AL (Ball et al., 2004; Sikorska-Simmons, 2001; Street et al., 2009; Street, Burge, Quadagno, & Barrett, 2007). Our research uses cumulative inequality theory (Ferraro & Shippee, 2009; Ferraro, Shippee, & Schafer, 2009) and the notion of adaptive reframing as the conceptual framework to explore how residential context and social relationships are associated with several dimensions of AL residents’ subjective well-being. Using data from the Florida Study of Assisted Living (FSAL), we analyze the associations between residential context and social relationships and individuals’ perceptions of general well-being, as well as responses to questions invoking explicit temporal, comparative, and experiential dimensions of subjective well-being.

Description of the Problem

Subjective Well-Being in AL

Well-being is a multidimensional concept that reflects rather generalized individual experiences, variously defined and measured (Diener, 2009; Ryff, 1989). Self-reports, whether of well-being or health, incorporate subjective perceptions that only partially capture the objective realities associated with individuals’ daily experiences, yet such measures are important indicators of holistic experiences of health and well-being. For example, self-reports of morbidity predict disability better than physician reports within an elderly population (Ferraro & Su, 2000). Global self-reported health is also a valid predictor of mortality (DeSalvo, Bloser, Reynolds, & Muntner, 2006; Frankenberg & Jones, 2004). Although self-reports of health have obvious objective components (i.e., the presence or absence of an illness, condition, or disease), well-being does not have such obvious objective elements, yet perceptions of well-being are also associated with important outcomes in later life (Diener, 2009). Subjective well-being can be seen as a cognitive-affective process in late adulthood, in which day-to-day affective or emotional states of being (e.g., happiness, affect) are closely associated—but are not identical—with the evaluation or appraisal of a range of well-being conditions (e.g., quality of life, satisfaction with life) Bishop, Martin, & Poon, 2006; Keyes, Shmotkin, & Ryff, 2002).

Measures used to assess well-being are not without problems. Perceived well-being, like its counterpart self-reported health, incorporates myriad unmeasured influences, ranging from personality, mood, and implicit comparisons to unknown reference groups. Responses may be associated with immediate (sometimes transitory) circumstances and recent events that an individual experiences and incorporates (even unconsciously) into his or her assessment of well-being. For these reasons, subjective well-being is an important indication of a person’s general sense of his or her quality of life, but measures of subjective well-being sometimes lack clarity (Diener, 2009; Ryff, 1989). Although unpacking such complexities is beyond the scope of this study, understanding how AL residents’ routine experiences of residential context and social relationships are associated with perceptions of well-being is important enough on its own merits, given the influence of subjective well-being on important later life outcomes (Diener, 2009) and a dearth of research on perceptual dimensions of accumulation that emphasize the linkages between actors’ perceptions of their situations and their structural positions (Ferraro, 2011, pp. 471-472). We use cumulative inequality and adaptive reframing (see below) to explore similarities and differences in the associations among residential contexts and social relationships for four broad indicators of subjective well-being: general, temporal, comparative, and experiential dimensions as perceived by AL residents.

Cumulative Inequality and Adaptive Reframing

For older AL residents, earlier life economic disadvantage constrains choice among AL facilities to less desirable or optimal settings, factors often associated with poorer individual outcomes (Burge & Street, 2010). Cumulative inequality conditions life chances in several important ways. Disadvantage heightens exposure to risk, while advantage raises exposure to opportunities, but the cumulative inequality framework emphasizes that disadvantage and advantage can be experienced simultaneously, albeit in different life domains (Ferraro et al., 2009). Unequal outcomes are sometimes, but not always, tightly linked across life domains. For example, although individuals’ later life negative financial outcomes result from earlier economic disadvantage, the cumulative inequality framework does not presume simultaneous disadvantage in terms of social support.

Using cumulative inequality as a conceptual framework draws attention to how some AL residents may experience suboptimal AL residential contexts stemming from earlier life disadvantages, yet these individuals may have enriching social relationships that represent an accumulation of advantage of social support (Burge & Street, 2010). In contrast, having sufficient financial resources to control AL transition circumstances and to optimize facility choice is no guarantee of advantage in terms of social relationships. In considering the possible influences of residential context and social relationships on well-being, insights from the cumulative inequality theoretical model help distinguish the possible influences of disadvantage and advantage in different aspects of subjective well-being.

When individuals move from community settings into AL, many adaptively reframe subjective assessments of well-being in the face of new circumstances. Adaptive reframing is roughly analogous to response shifting in health-related research: the propensity of individuals to recalibrate their expectations (and hence perceptions of health and well-being) within the context of evolving circumstances (Sprangers & Schwartz, 1999). Applied in health care settings, where patients confront diagnoses and changing or uncertain prognoses, response shifting reflects changes in conceptualizations of their quality of life (Carver & Scheier, 2000; Wilson, 1999). Rather than applying the same prediagnosis standards and values, patients adapt to new circumstances, using coping strategies to reframe and rethink assessments of quality of life (Westerman et al., 2008). Similar age-related adaptive processes occur as individuals adjust to the death of a spouse (Carr, 2004; Dykstra, 1995) or evolving limitations in physical functioning. Adaptive reframing and coping occur for many older individuals when they leave their private households behind and adjust to AL. Even into very old age, individuals actively optimize remaining social relationships and retain adaptive capacities (see Carstensen, Fung, & Charles, 2003; Lang & Carstensen, 1994). Transformations in residential context and social relationships that accompany AL transitions underscore how each may figure in adaptive reframing processes that influence AL residents’ well-being.

Personal characteristics doubtless influence well-being, but environmental factors also play a role. Individuals who once perceived high levels of well-being compared with others in prior community settings may perceive well-being quite differently once in AL, as the transition triggers different routines and likely reference groups. If community-based living heightens an individual’s awareness of functional barriers, moving to AL (where assistive staff members alleviate functional barriers) may improve subjective well-being, all else being equal. Alternatively, if being in AL is experienced as surrounded by residents with intensive care needs that trigger negative perceptions or alienating comparisons, subjective well-being could decline compared with prior community-based living arrangements. Whether AL residents can sustain their pretransition social relationships or have developed postmove ones may also influence perceptions of well-being across a number of dimensions.

Residential Context

Residential context, as we conceptualize it in this research, has two rather distinctive components: a set of individualized experiences surrounding the transition to AL and a set of organizational characteristics associated with particular AL facilities. Individuals move from homes in the community to AL on the basis of different constellations of need, facility proximity, available resources, and existing social networks that influence their subsequent experiences as AL residents. Because the AL industry is largely private pay, elders with greater economic resources have more control and choice over whether and when to move to AL (Hawes, Phillips, Rose, Holan, & Sherman, 2003; Stevenson & Grabowski, 2010). However, even when individual economic circumstances support an array of relatively attractive AL options (Burge & Street, 2010; Street et al., 2009), choice in the AL context is always constrained by need (Carder, Zimmerman, & Schumacher, 2009) and individual circumstance (Burge & Street, 2010).

Having more control and choice may facilitate positive social relationships. Moving to an AL close to a former neighborhood or near family helps AL residents maintain preexisting relationships. Once moved in, with AL staff members who work in care settings that are well-attuned to residents’ particular needs may be better than for individuals who had little choice because of low income (Burge & Street, 2010). Of course, AL staff relationships are unavoidable and coresident relationships are only somewhat voluntary. Specific organizational aspects of AL settings (including AL type and size) shape the composition of coresident populations and limit the array of coresident relationship possibilities (Heumann, 1996; Street et al., 2009), although positive or negative coresident relationships are possible regardless residential context (Burge & Street, 2010). Residents’ ability to develop meaningful social relationships within AL is influenced partly by the “fit” between what residents need and what particular AL facilities have to offer (Ball et al., 2000, 2004). Beyond the residential context that sets the stage for social relationships to endure or develop, the association between relationships and perceived well-being warrants examination, because the quality of social relationships is among the best predictors of health and life expectancy (Dykstra, 2007).

Social Relationships

AL residents experience two distinctive sets of social relationships: those external to and predating the move to AL and new relationships formed in the residential setting. External social connections provide emotional and instrumental support through family relationships and important friendship networks (Ajrouch, Antonucci, & Janevic, 2001; Antonucci & Akiyama, 1995; Crosnoe & Elder, 2002), often independent of financial circumstances. For example, although low income constrains AL facility choice, the same individuals may have such deep and meaningful family relationships and friendships that they are sustained in very positive ways, even in suboptimal residential settings. In contrast, a wealthy AL resident may be able to choose an optimal AL facility but lack satisfying social relationships outside of AL. Such circumstances underscore how inequality does not necessarily proceed in lockstep across life domains. Individuals with active family caregivers who intervene on their behalf may have more positive relationships with staff and coresidents, compared with residents more isolated from or lacking family (Port et al., 2005). AL residents who sustain contact with outside friends may be heartened by the continuing social connection, even as such relationships are reframed to take the realities of AL life into account. A regular phone call may take on the symbolic significance that only face-to-face interaction would have provided in earlier times. Some individuals in later life proactively optimize their remaining social relationships (Carstensen et al., 2003), and satisfaction with enduring friendship and family ties may be associated with AL residents’ positive assessments of some dimensions of well-being.

Later life inevitably spells involuntary social losses as illness and death among friends and relatives cause social connections to dwindle for many older people (Ajrouch, Blandon, & Antonucci, 2005). Moving to AL has its own social cost, as ties to friends in the community can be strained or severed and the availability of paid staff members may limit contact with family members who earlier gave more help. Fewer relationships may lead to loneliness, especially among the very old (Dykstra, 2007), increasing the risk for negative perceptions of well-being for AL residents. Older individuals may interpret social losses in ways that negatively affect perceived well-being.

Internal AL relationships may be more highly associated with subjective well-being than more sporadic external relationships. Although few situations are fertile ground for forging new relationships in later life (Rook, 1991), new social connections are unavoidable in AL. New residents experience an entirely novel cast of AL-specific characters offering fresh relationship potential, expanding social circles to include staff members, coresidents, and even coresidents’ visitors. Relationships formed in AL widen the pre-AL social support network (Pitt, Krieger, & Nussbaum, 2005) and are critical to residents’ adjustment and sense of feeling at home in AL (Port et al., 2005; Sikorska-Simmons, 2001; Street et al., 2007; Van den Hoonaard, 2002).

In the context of evolving relationships, Rook (2009) noted an important distinction between having substitute sources of social support, when individuals replace a lost supportive relationship by activating or intensifying an alternative relationship, and whether the substitute actually compensates for the earlier loss. As she noted, structured opportunities to nurture new friendships and social relationships in old age are rare (Rook, 1991, 2009); yet in AL, new relationships are a certainty. Although cognitively intact AL residents are free to come and go, many residents’ regular activities are almost entirely bounded by the confines of their AL setting, making interactions with staff members and coresidents important components of daily routines. New relationships that fail to adequately compensate for earlier social losses may be insufficient to positively influence well-being, while if compensatory relationships happen to develop, well-being might improve. The potential for developing new social relationships in AL may enhance or detract from well-being, independent of individual characteristics or residential context, another example of how resources in one life domain may counter advantage or disadvantage in another. We explore how residential context and social relationships influence AL residents’ assessments of well-being using FSAL resident survey data.

Methods

The Context of the FSAL

Florida is among the states with the largest and most varied AL industries in the United States (Mollica, Johnson-Lamarche, & O’Keefe, 2005; National Center for Assisted Living, 2009). AL in Florida is a site-specific, residential environment providing meals, supportive services, medication and 24-hour on-site supervision. Approximately 59% of Florida AL facilities are exclusively private pay (they do not accept Medicaid waivers or public assistance), 56.6% have fewer than 20 licensed beds, 19.8% have between 20 and 60 beds, and 23.6% have 61 or more beds (Street, Quadagno, & Burge, 2005). Florida has four licensure categories in state regulations, with licensure level targeted to meeting specialized needs. We use licensure status to distinguish the three Florida AL facility types. Traditional AL facilities with standard licenses provide housing, meals, and personal care services. High-frailty facilities with extended congregate care and/or limited nursing services licenses provide all standard services plus additional physical care, such changing routine dressings, passive range of motion exercises, applying heat and ice caps, cutting toenails, and extra services that help physically and cognitively frail residents age in place. For clients diagnosed with mental illnesses (such as schizophrenia or bipolar disorder), behavioral facilities with limited mental health licenses provide standard AL services plus specialized behavioral care. These AL facility types offer specific sets of services and have distinctive resident population profiles and staffing priorities to meet residents’ needs (Street et al., 2009). For example, FSAL facility surveys record Alzheimer’s prevalence among residents in Florida’s traditional facilities at about 28%, in high-frailty facilities at about 40%, and in behavioral facilities at about 9% (Street et al., 2005). Because behavioral AL settings have a younger resident profile with higher rates of mental illness, there may be fewer opportunities for older residents to develop positive coresident relationships (Street et al., 2009). Florida’s similarities to other states across a range of AL policies, organizational contexts, and diverse resident populations make Florida an appropriate site for exploring how residential context and social relationships influence perceived well-being among AL residents.

The FSAL included face-to-face structured interviews with AL residents throughout the state (n = 681). The FSAL resident sample was broadly comparable with the gender, race/ethnicity, marital status, and age composition of the resident sample in an earlier national AL study (Hawes, Phillips, & Rose, 2000), as well as more recent national research (National Center for Assisted Living, 2009). The first stage of the sampling frame for FSAL used administrative data (n = 1,886) to classify the population of AL facilities by size (small, medium, or large facilities), payer type (exclusively private-pay facilities vs. all others), and spatial distribution (by facility size and payer source) throughout the 11 planning and service areas within Florida. We next selected a purposive sample of facilities to visit (n = 148) that conformed to facility size, payer type, and the spatial distribution of facilities throughout Florida’s planning and service areas to conduct resident interviews. Our analytic sample is restricted to 344 individuals aged 60 years and older, who scored 6 or higher (out of 10) on the Short Portable Mental Status Questionnaire (used to assess cognitive function; Pfeiffer, 1975), and who provided valid data for all variables of interest, from 111 different Florida AL facilities. Approximately 85% of residents we approached agreed to FSAL resident interviews. We present descriptive statistics for the analytic sample in Table 1.

Table 1.

Descriptive Statistics for Analytic Sample (Florida Study of Assisted Living)

Variable	Range for Total Sample	% or M	n or SD	n
Well-being outcomes
Life satisfaction	0 to 1	79.5	267	336
Stable/improved quality of life	0 to 1	75.6	260	344
Others better off	0 to 1	15.3	47	308
Happy most of the time	0 to 1	78.8	249	316
Resident characteristics
Female	0 to 1	72.1	248	344
Married	0 to 1	15.1	52	344
Never married/divorced	0 to 1	22.1	76	344
Widowed	0 to 1	62.8	216	344
Physical function	1 to 3	2.5	0.50	344
Less than HS	0 to 1	27.9	96	344
HS graduate	0 to 1	26.2	90	344
More than HS	0 to 1	45.9	158	344
Residential context
AL transition
Control over move^a	0 to 1	72.4	249	344
Moved from home^b	0 to 1	70.9	244	344
Lived alone before AL^c	0 to 1	56.1	193	344
AL type
Private pay^d	0 to 1	59.6	205	344
Behavioral facility	0 to 1	7.6	26	344
High-frailty facility	0 to 1	38.1	131	344
Traditional facility	0 to 1	54.3	187	344
Small facility (<20 beds)	0 to 1	17.4	60	344
Medium-sized facility (20 to 60 beds)	0 to 1	30.2	104	344
Large facility (≥61 beds)	0 to 1	52.3	180	344
Social relationships
External
Family contact	0 to 4	2.7	1.2	344
Friend contact^e	0 to 1	52.0	179	344
Internal
Coresident relationships	1 to 100	71.0	28.0	344
Staff relationships	1 to 100	89.0	22.3	344

Note: AL = assisted living; HS = high school. Descriptive statistics for all independent variables are calculated using the largest analytic sample (quality of life; n = 344).

Reference category is little or no control over move to AL.

Reference category is moved from some other location (e.g., hospital, another AL facility, nursing home).

Reference category is lived with others (family or friends) before AL transition.

Reference category is use of public funds (e.g., Medicaid) to pay for AL expenses.

Reference category is no contact with friends outside of the AL setting.

Data

Dependent variables

We modeled the effects of residential context and social relationships on four outcomes: general, temporal, comparative, and experiential subjective perceptions of well-being. We used a standard question asked of all residents, “Are you basically satisfied with your life?” (yes = 1, no = 1), to assess general well-being (life satisfaction). We used the measure “How is your quality of life now compared to before you moved [to AL]: better, about the same, or worse?” (better or about the same = 1, worse = 0) to assess a temporal component of subjective well-being (stable or improved quality of life). A third measure asked “Do you feel that most people are better off than you are?” (yes = 1, indicating comparatively lower well-being; no = 0, indicating comparatively greater well-being) for comparative well-being (others are better off). Finally, “Would you say you are happy most of the time?” (yes = 1, no = 0) was the indicator for experiential well-being (happy most of the time).

Independent variables

Past research has demonstrated significant variation in subjective well-being among older adults by gender (Kim & Moen, 2002; Yang, 2008), marital status (Lucas, 2007a; Williams, 2003), educational attainment (Mirowsky & Ross, 2003a, 2003b), and physical health (Lucas, 2007b). Consequently, we control for several resident characteristics, including gender (female = 1, male = 0), three marital status categories (never married or divorced = 1, widowed = 1, married [reference] = 0), and three educational categories (high school graduate = 1, more than high school education = 1, less than high school graduate [reference] = 0). We control for level of resident physical function using an index of standardized z scores that averages residents’ scores on five activities of daily living (needs assistance to eat, dress, transfer, bathe, and toilet; Cronbach’s α = .81), where higher scores represent better physical function. Original codes for these variables ranged from not very hard at all (coded 3) to very hard (coded 1).

Residential context includes three items associated with the individual transition to AL and three with the type of AL facility where residents lived. The first transition variable measures the amount of control residents said they had over the move to AL (some or much control = 1, no control [reference] = 0). The second and third transition variables measure whether residents moved to AL from their own private homes or apartments in the community (coded 1) or from some other setting (another AL facility, a nursing home, or a relative’s home; coded 0 [reference]) and whether they lived alone prior to AL (coded 1) or with others (coded 0 [reference]). For AL type, we distinguish by private pay (exclusively private funds to pay for AL = 1, public funds = 0); facility licensure type, distinguishing behavioral AL facilities (has limited mental health license = 1) and high-frailty AL facilities (has limited nursing services and/or extended congregate care license = 1) from traditional AL facilities (standard license only [reference] = 0); and size, as large (61 or more beds = 1), medium (20 to 60 beds = 1), and small (fewer than 20 beds [reference] = 0) AL facilities.

Social relationships encompass both external and internal social relationships. External relationships are individuals in the community with whom residents had relationships prior to entering AL, including an index of family contact (average of the two items “How often does a family member visit you?” and “How often do you speak on the phone with family?” coded from never/almost never = 0 to daily/almost daily = 4; Cronbach’s α = .78) and friend contact (“Do you have regular contact with friends who do not live here?” coded any contact = 1, none [reference] = 0). Measures of internal relationships include a coresident relationship index that averages responses to six dichotomous items (“Do you regard any of the people who live here as your friends?” “Have you met other residents here with similar interests to yours?” “Do you feel like a member of the family?” “Do you attend most social events?” “Do other residents respect your privacy?” and “Do other residents bother you?” [reversed]; Cronbach’s α = .66) and a staff relationship index that averages responses to five dichotomous items (“Do you feel that you have friends among the staff?” “Do you feel that the staff listens to you?” “Do you feel that the staff shows affection and caring for you?” “Do you feel that the staff shows you respect?” and “Do you feel comfortable discussing health concerns with staff?” Cronbach’s α = .76). For ease of interpretation, both the coresident and staff relationship indexes are multiplied by 100 and range from 0 to 100.

We use the FSAL data in logistic regression models to show associations between residential context and social relationships with a series of four measures assessing different aspects of subjective well-being. For ease of interpretation, we report odds ratios (ORs), and because of the small sample, we assess statistical significance using confidence intervals (CIs) calculated at the 90% confidence level. Because AL facilities were the primary sampling unit and residents’ responses within facilities may be correlated, we calculate CIs using robust standard errors adjusted for clustering of cases within facilities using Stata’s cluster subcommand.

Results

Table 1 provides descriptive characteristics of the sample. Women predominate (72%), and nearly two thirds of the sample is widowed. Approximately 72% of residents said that they had some or complete control over the decision to move to AL, 71% moved to AL from private residences in the community, and 56% lived alone. Almost 60% of FSAL respondents in this sample paid privately for AL expenses, over half (54%) reside in traditional AL facilities with standard-only licenses, and approximately half (52%) live in large facilities (61 or more beds). The typical resident had a mean score of 2.7 for family contact—roughly analogous to seeing or speaking with family members a few times a month—and about half sustained contact with friends in the community. AL residents formed strong internal social relationships, indicated by high mean scores on coresident and staff indexes of 71 and 89 (on 100-point scales), respectively.

Tables 2 through 5 show series of multivariate models investigating the associations of residential context and social relationship variables with four dimensions of subjective well-being. For each dependent variable (one per table) the first model (panel 1) shows associations between residential context (AL transition and AL type) on a particular dimension of well-being, the second model (panel 2) shows associations with social relationships (external and internal), and the final model (panel 3) includes all residential context and social relationships. Each model includes controls for individual resident characteristics (gender, marital status, physical function, and education). For comparisons between models, we report McFadden’s R² to examine explained variance in logistic regression models.

Table 2

Odds Ratios of AL Residents’ General Well-Being (Life Satisfaction) (Florida Study of Assisted Living; n = 336)

Variable	Panel 1	Panel 2	Panel 3
Resident characteristics
Female	1.12	0.72	0.68
	(0.65 to 1.92)	(0.38 to 1.34)	(0.36 to 1.28)
Never married/divorced^a	0.18**	0.44	0.17**
	(0.06 to 0.59)	(0.16 to 1.20)	(0.05 to 0.60)
Widowed^a	0.26**	0.36**	0.15***
	(0.10 to 0.67)	(0.16 to 0.81)	(0.05 to 0.45)
Physical function^b	1.49**	1.29	1.39*
	(1.13 to 1.96)	(0.95 to 1.74)	(1.00 to 1.92)
HS graduate^c	32.06*	1.87	1.58
	(1.00 to 4.24)	(0.93 to 3.75)	(0.71 to 3.53)
More than HS^c	0.94	0.96	0.85
	(0.54 to 1.66)	(0.55 to 1.65)	(0.46 to 1.53)
Residential context
AL transition
Control over move^d	3.64***		2.57**
	(2.20 to 6.03)		(1.39 to 4.76)
Moved from home^e	0.46*		0.35**
	(0.21 to 1.00)		(0.15 to 0.79)
Lived alone before AL^f	1.78		3.00**
	(0.83 to 3.80)		(1.22 to 7.39)
AL type
Private pay^g	0.92		0.988
	(0.52 to 1.61)		(0.54 to 1.82)
Behavioral facility^h	0.96		1.42
	(0.37 to 2.46)		(0.49 to 4.12)
High-frailty facility^h	1.13		1.26
	(0.67 to 1.89)		(0.70 to 2.28)
Medium-sized facility (20 to 60 beds)ⁱ	0.19**		0.18***
	(0.06 to 0.55)		(0.06 to 0.51)
Large facility (≥61 beds)ⁱ	0.28**		0.33*
	(0.10 to 0.80)		(0.11 to 0.94)
Social relationships
External
Family contact		1.18	1.12
		(0.94 to 1.47)	(0.89 to 1.41)
Friend contact^j		1.03	1.02
		(0.64 to 1.66)	(0.63 to 1.64)
Internal
Coresident relationships		1.03***	1.03***
		(1.02 to 1.04)	(1.02 to 1.04)
Staff relationships		1.01	1.01
		(1.00 to 1.02)	(1.00 to 1.02)
McFadden’s R²	.13	.18	.24

Note: AL = assisted living; HS = high school. Confidence intervals reported in parentheses use a 90% confidence level.

Reference category is married.

Physical function is a standardized index of activity of daily living function, which originally ranged from 1 to 3.

Reference category is did not graduate from HS.

Reference category is little or no control over move to AL.

Reference category is moved from some other location (e.g., hospital, another facility, nursing home).

Reference category is lived with others (family or friends) before AL transition.

Reference category is use of public funds (e.g., Medicaid) to pay for AL expenses.

Reference category is traditional (standard license only) facility.

Reference category is small facility(<20 beds).

Reference category is no contact with friends outside of the AL setting.

p < .10. **p < .05.***p < .01.

General Well-Being

Table 2 shows the influences of residential context and social relationships on general well-being, using responses to a question about life satisfaction. Residents with greater control over the AL transition were 3.6 times more likely to be satisfied with life compared with respondents with no control over the transition (OR = 3.6, 90% CI = 2.2 to 6.0), while residents who moved from their own homes were only about half as likely to be satisfied as those who moved from other settings (OR = 0.5, 90% CI = 0.2 to 1.0). Neither AL licensure type nor resident private pay status was associated with life satisfaction. Residents of both medium-sized and large facilities reported lower life satisfaction than residents of smaller facilities (with 19 or fewer residents). This echoes findings from other AL studies showing that residents’ satisfaction is higher in smaller facilities (Ball et al., 2004; Chou, Bouldy, & Lee, 2003; Sikorska, 1999), perhaps because small facilities approximate homelike environments or because closer personal relationships are possible among residents and staff members in smaller places. Positive coresident relationships were significantly associated with greater life satisfaction, and social relationships seem to matter more than residential context for general well-being, demonstrated by the improved model fit (using McFadden’s R²) of panel 2 compared with panel 1. Residential context and social relationships are each independently associated with general well-being (panel 3); even after accounting for social relationships, a positive association between residents’ perceived control and life satisfaction persists, along with a negative association between moving from one’s own home to AL and life satisfaction. The fully elaborated model also shows that residents who lived alone prior to AL are 3 times more likely to be satisfied with their lives compared with those who lived with others prior to AL (OR = 3.0, 90% CI = 1.2 to 7.4).

Temporal Well-Being

Table 3 models show before-and-after temporal self-assessments of well-being, whether quality of life remained stable or improved compared with a decline after the move to AL. Panel 1 shows that residents with greater control over the AL transition were more than 2.5 times as likely to say that quality of life was stable or improved since the move compared with those with no control over the move (OR = 2.5, 90% CI = 1.5 to 4.3). Residents who paid privately were 2.7 times more likely than individuals who had to find facilities that accepted public funds to have stable or improved quality of life after the move (OR = 2.7, 90% CI = 1.7 to 4.3). Compared with traditional facility counterparts, high-frailty facility residents were more than 3 times as likely to experience stable or improved temporal well-being, perhaps a stabilizing influence of receiving appropriate assistive AL services (OR = 3.2, 90% CI = 1.8 to 5.8). Residents with external friend contact (panel 2) were less likely to perceive stable or improved quality of life after the move; such relationships may be reminders of the loss of independence or nostalgia for an earlier social milieu, or the temporal frame may invoke comparison with healthier, community-dwelling elders. Unlike the negative association with external friends, positive coresident and staff relationships seem to be associated with a temporal boost in well-being. The better model fit statistics in panel 1 (vs. panel 2) suggest that residential context may be somewhat more closely associated with perceptions of stable or improved quality of life. However, both residential context and social relationships have independent associations with residents’ temporal well-being in fully elaborated models (panel 3).

Table 3.

Odds Ratios of AL Residents’ Temporal Well-Being (Quality of Life Is Stable or Improved) (Florida Study of Assisted Living; n = 344)

Variable	Panel 1	Panel 2	Panel 3
Resident characteristics
Female	1.09	0.79	0.79
	(0.63 to 1.89)	(0.46 to 1.35)	(0.43 to 1.45)
Never married/divorced^a	0.85	0.71	0.94
	(0.28 to 2.61)	(0.30 to 1.70)	(0.29 to 3.08)
Widowed^a	0.75	0.54	0.70
	(0.28 to 2.03)	(0.25 to 1.14)	(0.25 to 1.94)
Physical function^b	1.72***	1.53***	1.66***
	(1.38 to 2.14)	(1.24 to 1.90)	(1.30 to 2.11)
HS graduate	0.96	1.19	1.11
	(0.49 to 1.88)	(0.59 to 2.41)	(0.52 to 2.39)
More than HS^c	1.68	2.15**	2.22*
	(0.87 to 3.22)	(1.14 to 4.03)	(1.08 to 4.57)
Residential context
AL transition
Control over move^d	2.54***		2.30**
	(1.49 to 4.33)		(1.29 to 4.09)
Moved from home^e	0.69		0.61
	(0.32 to 1.49)		(0.26 to 1.45)
Lived alone before AL^f	0.71		0.80
	(0.38 to 1.32)		(0.41 to 1.56)
AL type
Private pay^g	2.69***		3.03***
	(1.68 to 4.31)		(1.81 to 5.08)
Behavioral facility^h	2.36		2.79
	(0.76 to 7.32)		(0.84 to 9.21)
High-frailty facility^h	3.19***		3.36***
	(1.77 to 5.75)		(1.84 to 6.12)
Medium-sized facility (20 to 60 beds)ⁱ	1.16		1.07
	(0.59 to 2.28)		(0.52 to 2.22)
Large facility (≥61 beds)ⁱ	1.03		1.14
	(0.58 to 1.84)		(0.61 to 2.14)
Social relationships
External
Family contact		0.90	0.91
		(0.72 to 1.14)	(0.71 to 1.17)
Friend contact^j		0.57*	0.50**
		(0.33 to 0.98)	(0.29 to 0.85)
Internal
Coresident relationships		1.02***	1.02**
		(1.01 to 1.03)	(1.01 to 1.03)
Staff relationships		1.01**	1.01**
		(1.00 to 1.02)	(1.00 to 1.02)
McFadden’s R²	.13	.11	.19

Note: AL = assisted living; HS = high school. Confidence intervals reported in parentheses use a 90% confidence level.

Reference category is married.

Physical function is a standardized index of activity of daily living function,which originally ranged from 1 to 3.

Reference category is did not graduate from HS.

Reference category is little or no control over move to AL.

Reference category is moved from some other location (e.g., hospital, another facility, nursing home).

Reference category is lived with others (family or friends) before AL transition.

Reference category is use of public funds (e.g., Medicaid) to pay for AL expenses.

Reference category is traditional (standard license only) facility.

Reference category is small facility(<20 beds).

Reference category is no contact with friends outside of the AL setting.

p < .10. **p < .05.***p < .01.

Comparative Well-Being

Models of comparative well-being (Table 4) are based on an FSAL item asking respondents to respond using a specific referent: “most other people their age.” Residents who lacked control over the AL transition were more likely to say others were better off, as were individuals who depended on public assistance to pay for AL. Large facility residents were nearly 2.5 times more likely than residents of small ones to regard themselves as worse off than most others their age (OR = 2.5, 90% CI = 1.1 to 5.7). Frequent family contact (panel 2) and positive coresident relationships were both positively associated with comparative well-being. When considered separately, the residential context model fits slightly better (McFadden’s R²) than the social relationship model. When controls, residential context, and social relationships are included in the final model (panel 3), the positive association of family contact drops out of statistical significance, while control over the move, private pay, and coresident relationships remain positively associated with residents’ comparative well-being with most others their age.

Table 4.

Odds Ratios of AL Residents’ Comparative Well-Being (Others Are Better Off) (Florida Study of Assisted Living; n = 308)

Variable	Panel 1	Panel 2	Panel 3
Resident characteristics
Female	0.62	0.93	0.88
	(0.30 to 1.24)	(0.43 to 1.97)	(0.40 to 1.91)
Never married/divorced^a	2.88	4.65*	2.27
	(0.52 to 15.91)	(1.07 to 20.17)	(0.35 to 14.64)
Widowed^a	0.91	1.86	0.98
	(0.20 to 4.15)	(0.45 to 7.63)	(0.20 to 4.88)
Physical function^b	0.53***	0.57***	0.56***
	(0.40 to 0.72)	(0.41 to 0.79)	(0.41 to 0.77)
HS graduate	1.15	1.01	1.27
	(0.56 to 2.34)	(0.51 to 1.98)	(0.62 to 2.58)
More than HS^c	0.21***	0.18***	0.18***
	(0.10 to 0.46)	(0.08 to 0.39)	(0.08 to 0.38)
Residential context
AL transition
Control over move^d	0.41**		0.53*
	(0.23 to 0.72)		(0.29 to 0.99)
Moved from home^e	0.72		0.74
	(0.28 to 1.83)		(0.32 to 1.70)
Lived alone before AL^f	1.70		1.42
	(0.62 to 4.62)		(0.55 to 3.64)
AL type
Private pay^g	0.37***		0.30***
	(0.20 to 0.67)		(0.15 to 0.59)
Behavioral facility^h	1.65		1.17
	(0.61 to 4.46)		(0.36 to 3.88)
High-frailty facility^h	1.28		1.03
	(0.68 to 2.22)		(0.56 to 1.91)
Medium-sized facility (20 to 60 beds)ⁱ	1.97		1.90
	(0.81 to 4.79)		(0.76 to 4.75)
Large facility (≥61 beds)ⁱ	2.45*		2.21
	(1.06 to 5.69)		(0.94 to 5.15)
Social relationships
External
Family contact		0.74**	0.79
		(0.60 to 0.92)	(0.61 to 1.03)
Friend contact^j		1.19	1.56
		(0.63 to 2.23)	(0.81 to 3.01)
Internal
Coresident relationships		0.98**	0.98**
		(0.97 to 1.00)	(0.97 to 0.99)
Staff relationships		1.00	1.00
		(0.98 to 1.01)	(0.99 to 1.02)
McFadden’s R²	.26	.24	.30

Note: AL = assisted living; HS = high school. Confidence intervals reported in parentheses use a 90% confidence level.

Reference category is married.

Physical function is a standardized index of activity of daily living function,which originally ranged from 1 to 3.

Reference category is did not graduate from HS.

Reference category is little or no control over move to AL.

Reference category is moved from some other location (e.g., hospital, another facility, nursing home).

Reference category is lived with others (family or friends) before AL transition.

Reference category is use of public funds (e.g., Medicaid) to pay for AL expenses.

Reference category is traditional (standard license only) facility.

Reference category is small facility(<20 beds).

Reference category is no contact with friends outside of the AL setting.

p < .10. **p < .05.***p < .01.

Table 5

Odds Ratios of AL Residents’ Experiential Well-Being (Happy Most of the Time) (Florida Study of Assisted Living; n = 316)

Variable	Panel 1	Panel 2	Panel 3
Resident characteristics
Female	1.08	0.75	0.64
	(0.64 to 1.81)	(0.40 to 1.41)	(0.35 to 1.18)
Never married/divorced^a	1.11	2.64*	1.18
	(0.39 to 3.11)	(1.17 to 5.97)	(0.41 to 3.40)
Widowed^a	1.49	2.11	1.03
	(0.56 to 3.92)	(0.98 to 4.52)	(0.38 to 2.77)
Physical function^b	1.42**	1.24	1.30*
	(1.12 to 1.81)	(0.98 to 1.57)	(1.01 to 1.67)
HS graduate	1.35	0.99	1.02
	(0.76 to 2.37)	(0.55 to 1.78)	(0.52 to 2.02)
More than HS^c	2.19**	2.00*	2.05*
	(1.16 to 4.14)	(1.02 to 3.93)	(1.00 to 4.17)
Residential context
AL transition
Control over move^d	2.21**		1.63
	(1.28 to 3.82)		(0.84 to 3.15)
Moved from home^e	0.24***		0.19***
	(0.10 to 0.59)		(0.08 to 0.45)
Lived alone before AL^f	1.89		3.12**
	(0.86 to 4.16)		(1.40 to 6.96)
AL type
Private pay^g	0.84		0.81
	(0.46 to 1.53)		(0.40 to 1.65)
Behavioral facility^h	0.42		0.60
	(0.16 to 1.13)		(0.18 to 2.01)
High-frailty facility^h	0.55*		0.56
	(0.30 to 0.98)		(0.30 to 1.05)
Medium-sized facility (20 to 60 beds)ⁱ	0.78		0.85
	(0.36 to 1.69)		(0.38 to 1.88)
Large facility (≥61 beds)ⁱ	1.02		1.32
	(0.47 to 2.24)		(0.58 to 3.00)
Social relationships
External
Family contact		1.14	1.09
		(0.92 to 1.41)	(0.88 to 1.35)
Friend contact^j		1.11	1.20
		(0.60 to 2.06)	(0.64 to 2.23)
Internal
Coresident relationships		1.03***	1.03***
		(1.02 to 1.04)	(1.02 to 1.04)
Staff relationships		1.01	1.01
		(1.00 to 1.02)	(1.00 to 1.02)
McFadden’s R²	.09	.16	.21

Note: AL = assisted living; HS = high school. Confidence intervals reported in parentheses use a 90% confidence level.

Reference category is married.

Physical function is a standardized index of activity of daily living function,which originally ranged from 1 to 3.

Reference category is did not graduate from HS.

Reference category is little or no control over move to AL.

Reference category is moved from some other location (e.g., hospital, another facility, nursing home).

Reference category is lived with others (family or friends) before AL transition.

Reference category is use of public funds (e.g., Medicaid) to pay for AL expenses.

Reference category is traditional (standard license only) facility.

Reference category is small facility(<20 beds).

Reference category is no contact with friends outside of the AL setting.

p < .10. **p < .05.***p < .01.

Experiential Well-Being

The measure we used to capture the day-to-day experiential sense of well-being was whether AL residents said that they were happy most of the time. Residents with greater control over the move to AL (panel 1) were more than twice as likely to be happy most of the time compared with residents with less control (OR = 2.2, 90% CI = 1.3 to 3.8). Residents who moved from private homes were far less likely to be happy most of the time compared with residents who moved from other sites (such as another AL or a hospital). High-frailty facility residents were only about half as likely to be happy most of the time as compared to traditional facility residents (OR = 0.6, 90% CI = 0.3 to 1.0). Cognitively intact respondents in high-frailty facilities (with greater proportions of high-needs residents, physical care, and dementia or Alzheimer’s) seem to encounter a residential environment that suppresses being happy most of the time. Residents with strong coresident bonds (panel 2) reported greater experiential well-being compared with respondents with less positive coresident relationships. Social relationships were more closely associated with residents’ experiential well-being than residential context, as model fit comparisons of panel 2 to panel 1 show. The experiential benefits of being happy most of the time, associated with control over the AL transition and coresident friendships, persisted even after controlling for all residential context, social relationship, and control variables (panel 3).

Discussion

Our findings show significant relationships among the AL residential context, social relationships, and four measures of subjective well-being, yet the research has some obvious limitations. Because the FSAL data are cross-sectional, causal analyses are not possible. The sample was purposively (not randomly) generated to ensure that FSAL respondents were drawn from the full range of types, sizes, and geographic distribution of AL facilities around Florida. There may be unmeasured contextual differences across the AL facilities at which administrators cooperated with unrestricted researcher visits to interview residents privately. Although the demographic characteristics of the FSAL respondents are comparable with data from a population-level facility survey in Florida and a national sample of the AL industry, findings reflect the particulars of this sample and should be extended to other populations with that caveat in mind. As the researchers who collected the original data, we are aware that the FSAL respondents we interviewed were likely among the healthier and more gregarious residents of the facilities we visited. The perspectives of less healthy or sociable residents may not be fully represented. Although we interviewed both younger residents (59 years and younger) and older AL residents with moderate to severe cognitive impairments for the FSAL, they are excluded from our analysis because of our focus on older residents, the questionable reliability of recall data, and missing data on one dependent variable (not asked in a shortened version of the FSAL survey used for respondents with moderate to severe cognitive impairment). FSAL data limitations also meant that only single-item measures were available for very complex dimensions of subjective well-being. Insight into the contextual and relational factors that influence subjective well-being in AL residents would benefit greatly both from extensive qualitative studies and by extending the research to include the perspectives of cognitively impaired and younger residents. Even taking these shortcomings into account, the analysis in this study underscores important associations among residential context, social relationships, and AL residents’ perceptions of four dimensions of well-being.

Our key concerns were how and when the particular context of the AL setting may exert its own influences on well-being and whether context and relationships are mutually exclusive or reinforcing. AL residents with greater control and choice had better perceived well-being. Control over the transition was associated with all four dimensions of well-being, and residents who paid privately perceived well-being more positively on two dimensions, compared with residents of facilities that accepted public assistance. Those who can afford to pay simply have more and better choices about when and where they will move, selecting the most appealing facilities (Burge & Street, 2010). Residents of smaller facilities tended to report better general well-being than residents of larger facilities, reflecting perceptions associated with more intimate settings (Ball et al., 2004; Sikorska, 1999). Residents of high-frailty facilities, compared with traditional facilities, were more likely to give positive reports of temporal well-being but also had more negative perceptions of experiential well-being. Elders with the highest level of care needs are clustered in high-frailty facilities, where service packages can stabilize fragile residents, but regular interaction with high-needs coresidents may detract from their experiential well-being. Associations between residential context and multiple dimensions of well-being offers clues into ways residents may adaptively reframe perceptions in particular care settings.

When moves to AL are more immediately need driven or low income constrains choices, having less control undermines subjective well-being. Such disadvantages are consistent with the cumulative inequality framework that predicts negative outcomes associated structured by income and health disadvantages. Also consistent with the cumulative inequality framework, however, is the resilience of elders who draw on a variety of resources in ways that influence their perceptions of well-being. Many appear to adaptively reframe expectations in ways that sustain positive perspectives on several domains of subjective well-being, even in circumstances that might suggest suboptimal residential contexts. This is consistent with earlier research that documents elder resilience in other life domains (Davis, Zautra, Johnson, Murray, & Okvat, 2007; Moen, Sweet, & Hill, 2010).

AL residents who successfully establish positive relationships with staff members and coresidents report better adjustment to AL and feel more at home in new environs (Street et al., 2007). Strong external relationships with family and friends sometimes offer compensatory resources for AL residents who lack other dimensions of advantage, promoting better staff relationships within the AL setting. Opportunities to develop new social relationships in AL or other congregate living situations may offset some of the social support losses typically associated with increasing age (Rook, 2009), although such opportunities also depend on the resident mix within the particular AL setting (Dobbs et al., 2007; Heumann, 1996). Internal relationships seem to be especially salient for elders’ well-being as they adjust to emergent instrumental needs, such as the physical assistance provided by AL staff members and emotional support provided by other coresidents who have also navigated the transition to AL (Burge & Street, 2010; Rook, 2009; Street et al., 2007).

Social relationships demonstrated independent influence in all four domains of AL residents’ subjective well-being, mainly through the mechanism of positive relationships with coresidents. Although external relationships were seldom statistically associated with subjective well-being, that may be an artifact of the limited FSAL measures for external relationships, which document only the frequency of contact, not the quality of those relationships. This seems the most plausible explanation for the lack of statistical association between family contact and perceptions of well-being. Contact with community-dwelling friends was negatively associated with residents’ temporal perceptions of well-being. This may be due to unique comparisons residents invoke when interacting with external friends, perhaps a reminder of something once valued and lost, through moving to AL.

Good social relationships with AL coresidents were independently and positively associated with all four dimensions of well-being. Quality staff relationships were also positively associated with residents’ assessment of temporal well-being. These positive associations between internal social relationships and well-being may, in part, reflect high cognitive capacity and sociality among some FSAL respondents, characteristics that ease the formation of new relationships. Consistent with the cumulative inequality framework, the relevance of internal relationships also underscores the importance of novel, high-quality social relationships, experienced daily, to temper some disadvantages of residential context. Internal social relationships also reflect the capacity for elders to adapt to new residential arrangements and enjoy relationships that enhance subjective well-being, as residents reframe expectations and experiences in the face of the unique circumstances of AL.

Conclusions

Our research emphasizes the fundamentally comparative nature of subjective perceptions of well-being, viewed through the lens of comparisons drawn from the past, made to others, or reflecting a more experiential dimension of well-being. The cumulative inequality framework leaves theoretical space for decoupling advantage and disadvantage across life domains. Our findings of strong positive associations between internal social relationships with subjective well-being reinforce this tenet of cumulative inequality: Disadvantages associated with AL residential context do not necessarily translate into disadvantage in terms of AL residents’ perceptions of subjective well-being. Although our findings suggest the independent importance of strong internal social relationships, residential context shapes many of the opportunities for such relationships to form and flourish in the first place. The most consistent associations among the residential context variables were perceptions of control over the move to AL and the choice that accompanies the ability to pay privately, circumstances that predate transitions to AL, shaped by individual economic and health factors accumulated over lifetimes. In contrast, the most consistent relationship associations with positive well-being were with other AL residents, perhaps providing a window of opportunity for improving well-being if AL settings can enhance opportunities for positive coresident relationships. Our findings suggests that although independent living is very highly valued in American culture, if AL residents have good relationships with coresidents, living independently in the community may be less important for general well-being than having a batch of new friends.

Several issues warrant further exploration. Resident population profiles vary significantly by facility type, and high-frailty facilities have resident populations with greater physical care needs and higher rates of Alzheimer’s and dementia compared with traditional facilities (Street et al., 2009). Although estimates of residents with dementia and Alzheimer’s in AL in Florida vary somewhat across studies and environmental contexts, a sizable proportion of AL residents have diagnoses of cognitive impairment (Zimmerman et al., 2005; Street et al., 2005). To the extent that particular residents find themselves among resident populations that they find alienating or frightening, the quality of coresident relationships could actually undermine perceived well-being. Cognitively intact residents may find limited opportunities to develop meaningful friendships when they reside in facilities with high proportions of residents with dementia, particularly given how residents with dementia are stigmatized (Dobbs et al., 2007). In contrast, residents who find an especially “good fit” between their own care needs, staff capacity to provide essential services, and the broader resident population profile may be primed to form attachments with coresidents and staff members, which are likely to enhance their subjective well-being (Street et al., 2007). Teasing out these associations provides an agenda for additional, more finely grained research, including research that meets the needs for better understanding of both the structural and perceptual dimensions of accumulation of advantage or disadvantage (Ferraro, 2011). Several demographic trends, from shrinking families and increases in lifelong singlehood to population aging, signal that settings such as AL are poised to become ever more important residential care sites. Better understanding of AL environmental factors, such as residential context and social relationships that are associated with subjective well-being, may assist in designing sites and strategies that sustain and enhance the quality of later years for individuals who live in congregate settings.

Footnotes

Acknowledgements

The authors contributed equally to this research. We appreciate helpful comments from Jeralynn Cossman and Jill Quadagno and three anonymous reviewers.

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

This research was partially funded by grants from the National Institute on Aging (NIA 1R03 AG021770-01), the Florida Agency for Health Care Administration (AHCA Contract M0330), and the Florida State University Research Foundation.

References

Ajrouch

K. J.

Antonucci

T. C.

Janevic

M. R.

(2001). Social networks among blacks and whites: The interaction between race and age. Journal of Gerontology: Social Sciences, 56(B), S112-S118.

Ajrouch

K. J.

Blandon

A. Y.

Antonucci

T. C.

(2005). Social networks among men and women: The effects of age and socio-economic status. Journal of Gerontology: Social Sciences, 60(B), S311-S317.

Antonucci

T. C.

Akiyama

(1995). Convoys of social relations: Family and Friendships within a life span context. In Blieszner

Bedford

V. H.

(Eds.), Handbook of aging and the family (pp. 355-372). Westport, CT: Greenwood.

Assisted Living Federation of America, American Association of Homes and Services for the Aged, American Seniors Housing Association, National Center for Assisted Living, National Investment Center for the Seniors Housing and Care Industry. (2009). 2009 Overview of assisted living. Alexandria, Virginia.

Ball

M. B.

Whittington

F. J.

Perkins

M. M.

Patterson

V. L.

Hollingsworth

King

S. V.

(2000). Quality of life in assisted living facilities: Viewpoints of residents. Journal of Applied Gerontology, 19, 304-325.

Ball

M. M.

Perkins

Whittington

F. J.

Connell

B. R.

Hollingsworth

King

S. V.

. . . Combs

B. L.

(2004). Managing decline in assisted living: The key to aging in place. Journal of Gerontology: Social Sciences, 59(B), S202-S212.

Bishop

A. J.

Martin

Poon

L. W.

(2006). Happiness and congruence in older adulthood: A structural model of life satisfaction. Aging and Mental Health, 10, 1-9.

Burge

S. W.

Street

(2010). Advantage and choice: Social relationships and staff assistance in assisted living. Journal of Gerontology Series B: Psychological Sciences and Social Sciences, 65(B), 358-369.

Carder

P. C.

Zimmerman

Schumacher

J. G.

(2009). Understanding the intersection of individual needs and choices with organizational practices: The case of medication management in assisted living. The Gerontologist, 42(4), 463-473.

10.

Carr

(2004). Gender, pre-loss marital dependence, and older adults’ adjustment to widowhood. Journal of Marriage and Family, 66, 220-235.

11.

Carstensen

L. L.

Fung

H. L.

Charles

S. T.

(2003). Socioemotional selectivity theory and emotion regulation in the second half of life. Motivation and Emotion, 27, 103-123.

12.

Carver

C. S.

Scheier

M. F.

(2000). Scaling back goals and recalibration of the affect system are processes in normal adaptive self-regulation: Understanding “response shift” phenomena. Social Science & Medicine, 50(12), 1715-1722.

13.

Chou

S. C.

Bouldy

D. P.

Lee

A. H.

(2003). Factors influencing residents’ satisfaction in residential aged care. The Gerontologist, 43, 459-472.

14.

Crosnoe

Elder

(2002). Successful adaptation in the later years: A life course approach to aging. Social Psychology Quarterly, 65, 309-328.

15.

Davis

M. C.

Zautra

A. J.

Johnson

L. M.

Murray

K. E.

Okvat

H. A.

(2007). Psychosocial stress, emotion regulation, and resilience among older adults. In Aldwin

Park

Spiro

Abeles

(Eds.), Handbook of health psychology and aging (pp. 250-266). New York: Guilford.

16.

DeSalvo

K. B.

Bloser

Reynolds

Muntner

(2006). Mortality prediction with a single general self-rated health question: A metaanalysis. Journal of General Internal Medicine, 21(3), 267-275.

17.

Diener

(2009). Subjective wellbeing. In Diener

(Ed.), The science of wellbeing: The collected works of Ed Diener (pp. 11-48). Amsterdam, the Netherlands: Springer.

18.

Dobbs

Eckert

Rubenstein

Keimig

Clark

Frankowski

Zimmerman

(2007). An ethnographic study of stigma and ageism in residential care or assisted living. The Gerontologist, 48, 517-526.

19.

Dykstra

P. A.

(1995). Loneliness among the never married and the formerly married: The importance of supportive friendships and a desire for independence. Journal of Gerontology: Social Sciences, 50, S321-S329.

20.

Dykstra

P. A.

(2007). Aging and social support. In Ritzer

(Ed.), The Blackwell encyclopedia of sociology (pp. 88-93). Oxford, UK: Blackwell.

21.

Ferraro

K. F.

(2011). Health and aging: Early origins, persistent inequalities. In Settersten

R. A.

Angel

J. L.

(Eds.), Handbook of the sociology of aging (pp. 465-476). New York: Springer.

22.

Ferraro

K. F.

Shippee

T. P.

(2009). Aging and cumulative inequality: How does inequality get under the skin? The Gerontologist, 49, 333-343.

23.

Ferraro

K. F.

Shippee

T. P.

Schafer

(2009). Cumulative inequality theory for research on aging and the life course. In Bengston

V. L.

Gans

Putney

N. M.

Silverstein

(Eds.), Handbook of theories of aging (2nd ed., pp. 423-434). New York: Springer.

24.

Ferraro

K. F.

(2000). Physician-evaluated and self-reported morbidity for predicting disability. American Journal of Public Health, 90(1), 103-108.

25.

Frankenberg

Jones

N. R.

(2004). Self-rated health and mortality: Does the relationship extend to a low income setting? Journal of Health and Social Behavior, 45(4), 441-452.

26.

Hawes

Phillips

C. D.

Rose

(2000). High service or high privacy assisted living facilities, their residents and staff: Results from a national study. Washington, DC: U.S. Department of Human and Health Services.

27.

Hawes

Phillips

C. D.

Rose

Holan

Sherman

(2003). A national survey of assisted living. The Gerontologist, 43, 875-882.

28.

Heumann

(1996). Assisted living in public housing: A case study of mixing frail elderly and younger persons with chronic mental illness and substance abuse histories. Housing Policy Debate, 7, 447-471.

29.

Keyes

C.L.M.

Shmotkin

Ryff

C. D.

(2002). Optimizing well-being: The empirical encounter of two traditions. Journal of Personality and Social Psychology, 82(6), 1007-1022.

30.

Kim

J. E.

Moen

(2002). Retirement transitions, gender, and psychological wellbeing: A life-course, ecological model. Journal of Gerontology: Social Sciences, 57, P212-P222.

31.

Lang

F. R.

Carstensen

(1994). Close emotional relationships in later life: Further support for proactive aging in the social domain. Psychology and Aging, 9(2), 315-324.

32.

Lucas

(2007a). Adaptation and the set-point model of subjective well-being: Does happiness change after major life events? Current Directions in Psychological Science, 16, 75-79.

33.

Lucas

(2007b). Long-term disability is associated with lasting changes in subjective wellbeing: Evidence from two nationally representative longitudinal studies. Journal of Personality and Social Psychology, 92, 717-730.

34.

Mirowsky

Ross

(2003a). Education, social status, and health. New York: Walter de Gruyter.

35.

Mirowsky

Ross

(2003b). Social causes of psychological distress. New York: Walter de Gruyter.

36.

Moen

Sweet

Hill

(2010). Risk, resilience, and life-course fit: Older couples’ strategic adaptations to job displacement. In Fry

P. S.

Keyes

C.L.M.

(Eds.), New frontiers in resilient aging (pp. 283-309). New York: Cambridge University Press.

37.

Mollica

Johnson-Lamarche

O’Keefe

(2005). State residential care and assisted living policy: 2004. Washington, DC: U.S. Department of Health and Human Services.

38.

Pfeiffer

(1975). A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. Journal of the American Geriatrics Society, 23, 433-441.

39.

Pitt

M. J.

Krieger

J. L.

Nussbaum

J. S.

(2005). Finding the right place: Social interactions and life transitions among the elderly. In Ray

E. B.

(Ed.), Health communication in practice: A case study approach (pp. 233-241). New Brunswick, NJ: Routledge.

40.

Port

C. L.

Zimmerman

Williams

C. S.

Dobbs

Preisser

J. S.

Williams

S. W.

(2005). Families filling the gap: Comparing family involvement for assisted living and nursing home residents with dementia. The Gerontologist, 45(Special Issue 1), 87-95.

41.

Rook

K. S.

(1991). Facilitating friendship formation in late life: Puzzles and challenges. American Journal of Community Psychology, 19, 103-110.

42.

Rook

K. S.

(2009). Gaps in social support resources in later life: An adaptational challenge in need of further research. Journal of Social and Personal Relationships, 26(1), 103-113.

43.

Ryff

C. D.

(1989). Happiness is everything or is it? Explorations on the meaning of psychological wellbeing. Journal of Personality and Social Psychology, 57, 1069-1081.

44.

Sikorska

(1999). Organizational determinants of resident satisfaction with assisted living. The Gerontologist, 39, 450-456.

45.

Sikorska-Simmons

(2001). Development of an instrument of resident satisfaction with assisted living. Journal of Applied Gerontology, 20, 57-73.

46.

Sprangers

M. A.

Schwartz

C. E.

(1999). Integrating response shift into health-related quality of life research: A theoretical model. Social Science and Medicine, 48, 1507-1515.

47.

Stevenson

Grabowski

(2010). Sizing up the market for assisted living. Health Affairs, 29, 35-43.

48.

Street

Burge

Quadagno

(2009). The effect of licensure type on the policies, practices, and resident composition of Florida assisted living facilities. The Gerontologist, 49, 211-223.

49.

Street

Burge

Quadagno

Barrett

(2007). The salience of social relationships for resident wellbeing in assisted living. Journal of Gerontology, Series B: Psychological Sciences and Social Sciences, 62, S129-S134.

50.

Street

Quadagno

Burge

(2005). The Florida Assisted Living Study: Final report. Tallahassee: The Florida State University Pepper Institute on Aging and Public Policy.

51.

Van den Hoonaard

D.K.

(2002). Life on the margins of a Florida retirement community: The experience of snowbirds, newcomers, and widowed persons. Research on Aging, 24, 50-66.

52.

Westerman

M. J.

Hak

Sprangers

M.A.J.

Groen

H.J.M.

Van der Wal

The

(2008). Listen to their answers! Response behavior in the measurement of physical and role functioning. Quality of Life Research, 17, 549-558.

53.

Williams

(2003). Has the future of marriage arrived? A contemporary examination of gender, marriage, and psychological wellbeing. Journal of Health and Social Behavior, 44, 470-487.

54.

Wilson

I. B.

(1999). Clinical understanding and clinical implications of response shift. Social Science and Medicine, 48, 1577-1588.

55.

Yang

(2008). Social inequalities in happiness in the United States, 1972-2004: An age period-cohort analysis. American Sociological Review, 73, 204-226.

56.

Young

H. M.

(1998). Moving to congregate housing: The last chosen home. Journal of Aging Studies, 12, 149-165.

57.

Zimmerman

Sloane

P. D.

Williams

C. S.

Reed

P. S.

Preisser

J. S.

Eckert

J. K.

. . . Dobbs

(2005). Dementia care and quality of life in assisted living and nursing homes. The Gerontologist, 45, 133-146.