Trajectories of Life Satisfaction Five Years After Medical Discharge for Traumatically Acquired Disability

Transcription

1 Rehabilitation Psychology 2014 American Psychological Association 2014, Vol. 59, No. 2, /14/$12.00 DOI: /a Trajectories of Life Satisfaction Five Years After Medical Discharge for Traumatically Acquired Disability Caitlin L. Hernandez and Timothy R. Elliott Texas A&M University Andrea T. Underhill and Philip R. Fine University of Alabama at Birmingham Jack W. Berry Samford University Mark H. C. Lai Texas A&M University Objectives: We studied the predictive impact of family satisfaction, marital status, and functional impairment on the trajectories of life satisfaction over the first 5 years following medical treatment for traumatic spinal cord injury, burns, or interarticular fractures (total N 662). It was anticipated that fewer functional impairments, being married, and greater family satisfaction would predict higher life satisfaction trajectories. Method: The Functional Independence Measure, the Family Satisfaction Scale, and the Life Satisfaction Index were administered 12, 24, 48, and 60 months postdischarge. Results: Trajectory modeling revealed that greater functional impairment significantly predicted lower life satisfaction, regardless of injury type. However, this association diminished when marital status and family satisfaction were entered into the models. Greater family satisfaction and being married predicted greater life satisfaction across time. Moreover, there was no evidence for increases in life satisfaction trajectories over time: Trajectories were stable across time for all injury groups. Conclusions: Results suggest that being married and greater family satisfaction promote life satisfaction among those who traumatically acquire disability, and these beneficial effects may be more salient than the degree of functional impairment imposed by the condition. Keywords: life satisfaction, trauma, burns, spinal cord injury, family Impact This study is the first to examine life satisfaction trajectories over the first 5 years following medical discharge for a spinal cord or burn injury. Life satisfaction trajectories were stable over time and greater life satisfaction was consistently associated with greater family satisfaction and being married, regardless of injury type. Individuals who are single, separated, or divorced, or who report a family life that may be rigid and inflexible may be at risk for lower life satisfaction years following injury onset, and may require unique interventions and services to facilitate their well-being. This article was published Online First March 10, Caitlin L. Hernandez and Timothy R. Elliott, Department of Educational Psychology, Texas A&M University; Jack W. Berry, Department of Psychology, Samford University; Andrea T. Underhill and Philip R. Fine, Injury Control Research Center, University of Alabama at Birmingham; Mark H. C. Lai, Department of Educational Psychology, Texas A&M University. This research was supported in part by grant No. R49-CE from the U.S. Department of Health and Human Services Centers for Disease Control and Prevention and the National Center for Injury Prevention and Control to University of Alabama at Birmingham s Injury Control Research Center (ICRC). A complete listing of published studies that also report data from the ongoing ICRC health outcomes project is available from the corresponding author. Correspondence concerning this article should be addressed to Timothy R. Elliott, PhD, Department of Educational Psychology, 4225 TAMU, TX A&M University, College Station, TX telliott@ tamu.edu Introduction Life satisfaction is a critical, vital component of quality of life among persons who live with chronic and debilitating injuries. Yet it is a difficult concept to study, particularly in the wake of traumatically acquired disability, as relevant research is often confined to literatures devoted to a specific diagnostic condition in the pursuit of clinical information specific to the condition (and the treating medical specialty). In some areas, such as spinal cord injury (SCI), an extensive literature on life satisfaction now exists (for a recent review, see Post & van Leeuwen, 2012). In contrast, life satisfaction was not mentioned in a recent overview of adjustment following burn injury rehabilitation (Askay & Patterson, 2010), nor was it considered an important clinical outcome in a recent consensus statement (Gibran et al., 2013). Efforts to prospectively study life satisfaction following traumatically acquired disability are typically plagued by attrition due to unstable life circumstances and living arrangements, disparities in access to care and transportation, and personal and social factors that compromise adherence and subsequent follow-up (e.g., prior history of substance abuse, unemployed at time of injury; Corrigan et al., 2003; de Groot, Haisma, Post, van Asbeck, & van der Wonde, 2009; Holavanahalli et al., 2006). Large-scale studies that rely on epidemiological data typically lack essential information necessary to determine the functional impairments that constitute disability (Boyce & Wood, 2011; Powdthavee, 2009). The available research on life satisfaction among persons with traumatically acquired disabilities often lacks theoretical guidance. 183

2 184 HERNANDEZ ET AL. Consequently, there are few systematic, a priori studies of constructs and hypothesized mechanisms and their effects on life satisfaction over time among persons living with severe disabilities. Contemporary theories of well-being and life satisfaction posit that psychological variables may have a stronger influence on life satisfaction than life circumstances and events among people in general (Diener, Lucas, & Scollon, 2006; Frederick & Loewenstein, 1999; Lucas, 2007a). Life satisfaction, like other aspects of well-being and happiness, may be more influenced by ongoing set levels within an individual and less susceptible to events and circumstances than commonly assumed (Diener et al., 2006). Indeed, considerable evidence indicates that subjective well-being may fluctuate in response to major life events both positive and negative (e.g., winning the lottery, death of a spouse) but it typically reverts to pre-event levels within a matter of months (Diener et al., 2006). There is, however, some suggestion that this chronic set point level of well-being may be affected by certain circumstances that are not yet understood (Dunn, Uswatte, & Elliott, 2009; Fujita & Diener, 2005). For example, Resch and colleagues (2009) recently found interesting differences in life satisfaction trajectories over the first 5 years following medical treatment for a traumatic brain injury: Overall, participants decreased in life satisfaction over the 5 years, with the exception of those individuals with the lowest levels of functional impairment. Similarly, Lucas (2007b) found moderate to large decreases in life satisfaction following disability that did not return to predisability levels. These findings, at variance with the hedonic set point models of well-being, may be due to variations in individual set points, the presence of multiple set points within a single individual, and individual differences in adaptation that become apparent over time (Diener et al., 2006; see also Boyce & Wood, 2011). Conceptually, these influences on life satisfaction may be best described as circumstantial factors and intentional activities (Lyubomirsky, Sheldon, & Schkade, 2005). Of these, Dunn et al. (2009) argue that intentional activities ones that reflect behavioral, cognitive, and volitional efforts have the potential to significantly affect life satisfaction following acquired disability. Following traumatic injury, an individual typically experiences sudden (and often dramatic) declines in their ability to participate in routine, meaningful, and goal-directed activities. The degree of functional impairment represents one aspect of intentional activity, and it is often associated with life satisfaction following traumaticonset disability (Dijkers, 2005; Resch et al., 2009). Circumstantial factors such as age, socioeconomic status, and marital status are often included in outcome studies following traumatic injuries. Married individuals tend to report higher life satisfaction than unmarried persons in the years following traumatic disability (Dijkers, 2005; Lucas, Clark, Georgellis, & Diener, 2003). Being married is often associated with more positive quality of life outcomes among people who live with a variety of health conditions (Carr & Springer, 2010). But the reasons and mechanisms through which marriage exerts a positive influence are unclear: Much of the research in disability conceptualizes marital status as a demographic variable, and, as such, it does not convey the behaviors emitted in the relationship that facilitate individual well-being (Lyubomirsky et al., 2005, p. 121). The ability of the family to provide an adaptive, cohesive life for the individual appears to facilitate life satisfaction over time regardless of injury severity, time since injury, and marital status (Dunn & Brody, 2008; Johnson et al., 2010). Emotional bonds between family members and the ability of family members to adapt to changes in roles, relationships, and rules in times of stress may be paramount following disability onset (Olson, Russell, & Sprenkle, 1983). A resilient, cohesive family atmosphere provides a positive experience for the individual with displays of support, reinforcement of self-worth, and increased opportunities for social interaction, positive emotions, and participation in desired activities (Johnson et al., 2010; Klinge, Chamberlain, Redden, & King, 2009). We conducted the current study to examine the relative contributions of marital status, functional impairment, and self-reported family satisfaction (Olson & Wilson, 1982; Olson et al., 1983) on life satisfaction trajectories over the first 5 years following hospitalization for a traumatic SCI, burns, or lower extremity intraarticular fractures (IAFs; fractures that include the surface of a bone contiguous with joint space). Rather than study life satisfaction at a single point in time, we examined the trajectory of life satisfaction over time, maximizing the use of available data from all participants, thus facilitating our ability to identify prospective and co-occurring individual level characteristics that promote or prevent adaptation (Lucas, 2007a, p. 78). We used multilevel modeling (MLM) techniques to study individual-level growth trends in life satisfaction within the context of covarying functional impairment, family satisfaction, and marital status in the first 5 years after treatment for a traumatic, debilitating injury. MLM techniques avoid common statistical obstructions, such as attrition and unequal time intervals, in predicting individual trajectories in longitudinal studies (Kwok et al., 2008). We expected greater functional impairment to be associated with declines in life satisfaction over time, regardless of the type of injury. We also expected being married and greater family satisfaction to be associated with life satisfaction trajectories. Method Participants Prospective participants were a subset of a larger longitudinal study of adjustment following traumatic injury conducted by the Injury Control Research Center at the University of Alabama at Birmingham. Prospective participants included individuals who sustained one of four debilitating injuries (burns, SCI, IAFs of the lower extremities, or brain injury) and who were discharged from a set of hospitals in north-central Alabama. Individuals with brain injuries were studied in previous work and are excluded from this report (see Johnson et al., 2010; Resch et al., 2009). Trained raters identified prospective participants in a review of acute-care medical records at the hospitals. International Classification of Disease, 9th Revision (ICD-9) codes were used to determine medical diagnoses of patients who could be invited to participate (the codes are available upon request from the second author). To be eligible, an individual was admitted to an acute-care setting for at least three days; resided and injured in Alabama; was discharged alive from the hospital between October 1, 1989, and September 30, 1992; was older than 17 years old at time of injury; and agreed to be contacted at prespecified intervals following discharge. Eligible participants were contacted 12 months postdis-

3 TRAJECTORIES OF LIFE SATISFACTION 185 charge by letter, which contained an explanation of the study and included a preaddressed consent-to-contact postcard. Individuals were contacted by phone to obtain consent if the written consent was not initially returned. Consenting participants included 662 (465 men, 197 women) with SCIs, IAFs, or severe burns. The average age at time of injury was years (SD 17.1; range 18 to 96 years). The sample was primarily Caucasian (n 455, or 68.7%; 30.1% was African American). Following consent, a trained interviewer contacted participants to collect data. Caretakers, spouses, or close relatives were interviewed if an individual was unable to complete the survey. Each participant was interviewed by telephone as closely as possible to 12 months following discharge from the acute-care setting. Measures were subsequently administered at 24, 48, and 60 months postdischarge. Variables Marital status. Marital status was recorded at each measurement occasion. Marital status was coded as married, single, divorced, separated, widowed, or other. Functional impairment. To assess cognitive and motor impairments, we used the telephone version of Functional Independence Measure (FIM; Keith, Granger, Hamilton, & Sherwin, 1987). The FIM has 18 questions on a Likert-type rating scale that assess motor functioning (13 items) and cognitive functioning (five items). Higher scores on the FIM indicate greater functional independence. Due to potential ceiling effects on this scale, FIM items were linearized with Rasch scaling procedures to increase item variability and to ensure item quality, stability and reliability (Bond & Fox, 2001; Linacre, 2003). The internal consistency coefficients in this study ranged from.95 to.97. Injury severity. Injury severity was determined with the Abbreviated Injury Scale (AIS; Committee on Injury Scaling, 1985). This measure contains values ranging from 0 (not injured) to6(maximum injury; Committee on Injury Scaling, 1985). Trained raters reviewed the discharge record for each participant to code injury severity. The ICD- MAP was used to convert ICD9-CM coded discharge diagnoses to AIS scores (MacKenzie, Steinwachs, & Shankar, 1989). Family satisfaction. The Family Satisfaction Scale (FSS; Olson & Wilson, 1982) was used to assess family satisfaction. The 14 items are rated on a Likert-type scale (1 dissatisfied, 2 somewhat dissatisfied, 3 generally satisfied, 4 very satisfied, 5 extremely satisfied). Total scores ranged from 14 to 70. Higher scores indicate greater family adjustment, cohesion, and adaptability (Olson et al., 1983). For this study, the FSS was modified because two of the original items (#4 and #5) assessed satisfaction a dependent child may have with parental actions. These items were rewritten to eliminate this focus (see Underhill, Lobello, & Fine, 2004). The internal consistency coefficients ranged from.94 to.97. Life satisfaction. The Life Satisfaction Index A (LSI; Neugarten, Havighurst, & Tobin, 1961) is a 20-item instrument designed to measure psychological well-being, measuring passion for life, mood, and congruence between desired and achieved goals. The total score ranges from 0 to 20. Higher scores indicate greater life satisfaction. The LSI has evidenced acceptable psychometric properties (Adams, 1969; Wallace & Wheeler, 2002). It is considered one of the best available measures of life satisfaction in health outcomes research (McDowell, 2006; p. 235). The internal consistency coefficients ranged from.86 to.90. Data Analysis Preliminary analyses were conducted to examine initial differences within the sample, including basic demographic information and descriptive statistics for each of the three self-report measures. Hierarchical linear modeling (HLM; Raudenbush & Bryk, 2002) was used to examine the influence of family satisfaction and functional independence on life satisfaction growth trajectories 5 years following hospitalization for traumatic injury. HLM accounts for interdependencies within levels of data, accounting for more accurate standard error measurements when ANOVAS would underestimate (Kwok et al., 2008; Resch et al., 2009). The MIXED routine in SPSS, used for fitting multilevel linear growth modeling, was used to analyze these data (Kwok et al., 2008). Several models were conducted to examine the relative contributions of each predictor variable. The explanations for each model and the corresponding equations are contained in the Appendix. Results Preliminary Analyses Individuals with an SCI (n 144; mean age 37.33, SD 15.02) or burns (n 260; mean age 39.97; SD 16.97) were significantly younger than individuals with an IAF (n 260; mean age 44.33, SD 17.83), ps.05. The majority of the sample identified as White (n 455, or 68.7%; Black: n 199, or 30.1%). The disproportionate percentage of White to Black participants was significant across all injury types: SCI, 2 16, df 1, p.01; IAF, , df 1, p.01; burns, , df 1, p.01. Most participants were married at the first measurement point (n 334; 50.5%) and 167 participants were single (25.2%; Table 1). Participants reported being divorced (n 77; 11.6%) or widowed (n 41; 6.2%) at lower rates. Those participants that reported being separated were least frequently observed in this study (n 21; 3.2%). Of the total sample, 75.5% of the participants had moderate to serious injuries (AIS ratings of 2 or 3; Table 2). FIM scores were significantly negatively correlated with AIS in individuals with burns, indicating that with increased ratings of AIS, ratings of FIM decreased in these individuals, r.191, n 739, p.01. However, AIS and FIM scores were not correlated among individuals with IAFs (r.049, n 793, p.169) or SCIs (r.091, n 350, p.090). AIS ratings were not included in modeling trajectories to avoid redundancy and overlapping information. Self-Report Measures The mean scores for the self-report variables by measurement occasion and injury type, and used in subsequent analyses, are presented in Tables 3 and 4. Across all measures and measurement points, 36% of data were missing. Across measures, 25% of data were missing across all time points for both LSI and FIM. However, 59% of data were missing for family satisfaction (FSS)

4 186 HERNANDEZ ET AL. Table 1 Frequencies of Marital Status by Injury Type Across Measurement Periods Injury type Month follow-up Total SCI Marital status Single Married Divorced Separated Widowed Other Total IAF Marital status Single Married Divorced Separated Widowed Other Total Burn Marital status Single Married Divorced Separated Widowed Other Total Note. SCI spinal cord injury; IAF intra-articular fracture. across all measurement occasions. At initial measurement, no differences were found between injury groups on life satisfaction (LSI) and FSS, indicating similar levels of satisfaction with life and family between all injury groups. A repeated-measures MANOVA revealed no significant difference between injury groups and life satisfaction at each time point,.974, p.162. No significant differences were found between married and nonmarried participants in family satisfaction ratings across measurement points using repeated-measures MANOVA,.066, p.653. Table 2 Injury Severity by Injury Type for the Total Sample SCI IAF Burns Total Missing Total Note. 1 minor injury; 2 moderate; 3 severe, not life threatening; 4 severe, life threatening; 5 critical, survival uncertain; and 6 maximum injury (Civil & Schwab, 1988). SCI spinal cord injury; IAF intra-articular fracture. Table 3 Mean Measurement Scores by Injury Type at the First Assessment Injury type LSI FSS FIM SCI Mean n SD IAF Mean n SD Burn Mean n SD Total Mean n SD Note. LSI Life Satisfaction Index; FSS Family Satisfaction Scale; FIM Functional Independence Measure; SCI spinal cord injury; IAF intra-articular fracture. There were significant differences between injury groups for functional impairment (FIM). At initial measurement, individuals with an SCI had significantly greater functional impairment than individuals with an IAF (mean difference on FIM 21.53, SE 1.76) or burns (mean difference 23.96, SE 1.75), ps.05). Across measurement points, the mean life satisfaction (LSI) for the total sample was (SD 4.88). The average family satisfaction score across time for the total sample was (SD 11.59), and the average linearized functional independence measure score was 2.93 (SD 2.02). Table 4 Mean Measurement Scores at Each Measurement Point Month follow-up LSI FSS FIM 12 Mean n SD Mean n SD Mean n SD Mean n SD Total Mean n SD Note. FIM indicates linearized FIM scores. LSI Life Satisfaction Index; FSS Family Satisfaction Scale; FIM Functional Independence Measure.

5 TRAJECTORIES OF LIFE SATISFACTION 187 Growth Models The first multilevel liner growth model analyzed changes in life satisfaction as predicted by functional impairment (FIM scores). Functional impairment significantly predicted life satisfaction, (est).67, SE.09, p.01 (see Table 5). The interaction between functional impairment and time was not significantly predictive of life satisfaction trajectories. Thus, individuals with less functional impairment had higher life satisfaction, and there were no significant changes over time in the relationship between life satisfaction and functional impairment across all injury types. Across all samples and across time, functional impairment predicted life satisfaction trajectories. This initial model, featuring the three predictors of functional impairment, time, and functional impairment by time, accounted for 15% of the variance at the within-participant level and 22% of variance at the betweenparticipants level in life satisfaction trajectories (Snijders & Bosker, 1994, 2012). Multilevel linear growth models predicting the rates of change in life satisfaction were conducted using functional impairment and family satisfaction as the time-variant covariates (see Table 6). Both functional impairment, as measured by total FIM scores (est.68, SE.10, p.01) and family satisfaction (FSS; est.14, SE.02, p.01) significantly predicted life satisfaction. Less functional impairment and greater family satisfaction predicted higher life satisfaction across time. This model accounted for 31% of the explained variance at the withinparticipant level and 44% of variance at the between-participants level in life satisfaction trajectories (an increase of 16% and 22%, respectively, in variance accounted for above and beyond the first model). Combined Analyses Functional impairment, family satisfaction, and injury type. Further analyses incorporating injury type as a time-invariant covariate revealed similar results. Less functional impairment (est.83, SE.11, p.01; Table 7) and greater family satisfaction (est.14, SE.017, p.01) significantly predicted life satisfaction regardless of injury type and time. However, individuals with burns had lower life satisfaction than the other two injury groups across all measurement occasions (est 1.25, SE.34, p.01). This model accounted for 33% of the explained variance at the within-participant level and 47% of variance at the between-participants level in life satisfaction trajectories. Functional impairment, family satisfaction, and marital status. Marital status was incorporated into the models of functional impairment and family satisfaction to predict life satisfac- Table 5 FIM and Estimates of Fixed Effects Parameter Estimate SE T Sig. Intercept Time FIM FIM Time Note. Dependent variable: Life Satisfaction Index. Sig. significance level; FIM Functional Independence Measure. Table 6 FIM, FSS, and Estimates of Fixed Effects Parameter Estimate SE T Sig. Intercept Time FSS FIM FIM Time FSS Time FSS FIM Time Note. Dependent variable: Life Satisfaction Index. FSS Family Satisfaction Scale; FIM Functional Independence Measure; Sig. significance level. tion. Individuals who were married were used as the basis for comparison. Being single or separated significantly predicted lower life satisfaction, regardless of injury type (single: est 1.46, SE.32; separated: est 3.53, SE.76), ps.05 (Table 8). Greater family satisfaction remained predictive of higher life satisfaction (est.12, SE.03, p.01) regardless of marital status. However, once marital status was included in the model, functional impairment was no longer significantly predictive of life satisfaction, and there was no significant effect of functional impairment on life satisfaction over time. This model accounted for 34% of the explained variance at the within-participant level and 49% of variance at the betweenparticipants level in life satisfaction trajectories. Functional impairment, family satisfaction, injury type, and marital status. A final model incorporating all measures of family satisfaction, functional impairment, injury type, and marital status was conducted. Family satisfaction remained a significant predictor of life satisfaction (est.12, SE.03, p.01; Table 9), but functional impairment was not a significant predictor. Across injury types, burn injuries were significantly associated with lower life satisfaction (est 1.29, SE.33, p.01). Other injury types were not associated with life satisfaction. Being single, divorced, or separated significantly and negatively predicted life satisfaction (single: est 1.51, SE.32, p.01; divorced: est 1.15, SE.40, p.01; separated: est 3.54, SE.76, p.01). Overall, family satisfaction, marital status, and injury status were the strongest predictors of life Table 7 Injury Type, FIM, FSS, and Estimates of Fixed Effects Parameter Estimate SE T Sig. Intercept Time FSS SCI a Burns a FIM FIM Time FSS Time FSS FIM Time Note. Dependent variable: Life Satisfaction Index. FIM Functional Independence Measure; SCI Spinal Cord Injury; FSS Family Satisfaction Scale. a Reference injury type: intra-articular fracture.

6 188 HERNANDEZ ET AL. Table 8 Marital Status, FIM, FSS, and Estimates of Fixed Effects Parameters Estimate SE T Sig. Intercept Time FIM Single a Divorced a Widowed a Separated a FSS FIM Time FSS Time FIM FSS FIM FSS Time Note. Dependent variable: Life Satisfaction Index. FIM Functional Independence Measure; FSS Family Satisfaction Scale; Sig. significance level. a Reference marital status: married. satisfaction in these models. Being married or widowed and having an SCI or IAF were associated with higher life satisfaction. This model accounted for more of the explained variance than any of the previous models: 36% of the explained variance at the withinparticipant level and 52% of variance at the between-participants level in life satisfaction trajectories. Thus, this final model accounted for 30% additional variance at the between-participants level and 21% additional variance at the within-participant level above and beyond the variance accounted for in the first model (with three predictors). Discussion Lower functional impairment, higher family satisfaction, and being married were significantly associated with trajectories of higher life satisfaction over the first 5 years following medical discharge for an SCI, burn injury, or severe lower-extremity fractures. Regardless of the type of injury, greater functional abilities were associated with more optimal life satisfaction. In addition, the degree of functional impairment was relatively stable across time for each disability group: No significant changes were observed in FIM scores over time. These findings are consistent with our position that individuals with greater functional impairment may have less ability to perform certain tasks and decreased opportunities to engage in desired and valued activities. However, our results also found the relation of functional impairment to life satisfaction diminished upon adding other predictors in later models that accounted for greater variance in life satisfaction trajectories. The relationship between impairment and life satisfaction was not significant in the equations in which marital status and family satisfaction were included as predictors. Family satisfaction, indicative of a flexible, resilient, and cohesive family (Olson, 2011), was predictive of greater life satisfaction in every model in which it was included. Cohesive and resilient families may facilitate positive emotions and adaptive behaviors that promote well-being in individuals following injury (Ryff, 1989). These features, consistent with increased family satisfaction and positive emotion, might help individuals following onset of injury, in that they may inoculate against stress and increase the likelihood of the person generalizing positive experiences to intentionally engage in more positive acts (Dunn et al., 2009). Unfortunately, few studies investigate the quality of family relationships as a predictor of adjustment following injury onset. Most research conceptualizes family in terms of social support (Müller, Peter, Cieza, & Geyh, 2012), and more work is needed to understand how family cohesion and resilience impact overall adjustment following traumatic and disabling injuries. Similarly, marital status was also significantly predictive of life satisfaction trajectories in the two models in which it was included. Being single, divorced, or separated was significantly associated with lower life satisfaction across time and across each injury type. Occasionally conceptualized as a proxy of support, marriage is often associated with well-being following injury (for burns, see Patterson, Ptacek, Cromes, Fauerbach, & Engrav, 2000; for SCI, see Putzke, Elliott, & Richards, 2001). It is possible that being married accounts for more of the everyday support from day-today interactions with spouses (van Leeuwen et al., 2012; Vaux, 2000). Family members, in general, assume many responsibilities for care and assistance following medical discharge to the community, and many people with acquired disability indicate that partners and family members are more influential than professional staff in helping them adjust in the first year following the onset of disability (Rogers & Kennedy, 2000). A positive family environment may also influence adjustment postinjury by facilitating participation in desired activities including leisure pursuits, social integration, and general mobility (Carr & Springer, 2010; Freedman, Stafford, Schwarz, Conrad, & Cornman, 2012). In some cases, family members may impede or undermine quality of life with misguided attempts at support, or by promoting a critical, angry, or generally unconstructive environment (e.g., EnglandKennedy & Horton, 2011). Restrictions in desired activities are associated with depression among persons with chronic health conditions and their family caregivers (Mausbach et al., 2011). Nevertheless, use of the marital status variable provides no information about the mechanisms by which it influences quality Table 9 Injury Type, Marital Status, FIM, FSS, and Estimates of Fixed Effects Parameters Estimate SE T Sig. Intercept Time FIM SCI a Burns a Single b Divorced b Widowed b Separated b FSS FIM Time FSS Time FIM FSS FIM FSS Time Note. Dependent variable: Life Satisfaction Index. FIM Functional Independence Measure; SCI spinal cord injury; FSS Family Satisfaction Scale; Sig. significance level. a Reference injury type: intra-articular fracture. b Reference marital status: married.

7 TRAJECTORIES OF LIFE SATISFACTION 189 of life in the years following traumatic injury (an influence that is independent from, and perhaps more persistent than, functional impairment). Rather than attribute positive, palliative effects of marriage on adjustment following injury, the pattern of our results might suggest that being single, divorced, or separated may compromise well-being after traumatic injury. These individuals could be socially isolated or lack intimacy or social contact with others. These are issues to examine in future research. We also found that burn injury was a significant predictor of life satisfaction trajectories. The burn sample had more functional ability than the other two groups. Consequently, the significantly lower level of life satisfaction observed among individuals with burn injuries, compared with the rest of the study participants, was unexpected. The relationship between burn injury and life satisfaction remained significant in the equations that included marital status and family satisfaction (and these equations accounted for the greater amounts of variance accounted for in the life satisfaction trajectories). This pattern suggests that factors other than the potential to engage in intentional activities may have greater impact on life satisfaction among persons who incur severe burns than we assumed, such as stigma, body image, social functioning, and pain (Corry, Pruzinsky, & Rumsey, 2009). Our theoretical approach to this study and our use of MLM to examine life satisfaction trajectories provides information about the relative influence of several clinically important variables on these trajectories. There are other approaches that rely on latent growth mixture models to reveal considerable heterogeneity in the ways people adjust following major live events (Bonanno, Westphal, & Mancini, 2011). From that work we know individual differences are apparent in the first year following traumatic injury and optimal levels of adjustment may occur and stabilize within 3 to 6 months following injury for most individuals (Bonanno, Kennedy, Galatzer-Levy, Lude, & Elfström, 2012; de Roon-Cassini, Rusch, Mancini, & Bonanno, 2010; Klein et al., 2011). Our first measurement occasion occurred 1 year postdischarge, a point at which levels of life satisfaction had likely leveled off for our participants. Thus, the significant and dynamic changes in life satisfaction among our participants occurred prior to their consent to be in our study. From our perspective, both approaches provide valuable information about the demographic, social, and personality factors that promote positive reactions to major life events (Lucas, 2007a, p. 78), like traumatic, debilitating injury. There are several limitations of our study that merit consideration. We relied on self-report instruments for all measured variables. The LSI, despite its utility in past research (McDowell, 2006), was first published in 1961 and may be considered dated by some researchers. The data were collected from participants who were treated and discharged between the years 1989 to 1992 in one southeastern state in the United States. We do not know the degree to which these features affect our results or their generalizability. Other variables associated with life satisfaction following disability were excluded from this study. Participation in desired activities is an important predictor of life satisfaction, and there is emerging evidence that it may mediate the prospective relationship of functional impairment to life satisfaction following traumatic injury (Erosa, Berry, Elliott, Underhill, & Fine, in press; Kalpinski et al., 2013). Our unexpected finding that burn injury was predictive of lower life satisfaction regardless of functional impairment, family satisfaction, and marital status implicates other mechanisms unique to these injuries that warrant scrutiny in future study, such as stigma, body image, and pain. Finally, future research should attend to the mechanisms by which family satisfaction exerts a positive influence on quality of life in the years following traumatic disability. References Adams, D. L. (1969). Analysis of life satisfaction index. Journal of Gerontology, 24, doi: /geronj/ Askay, S., & Patterson, D. (2010). Psychological rehabilitation in burn injuries. In R. G. Frank, B. R. Caplan, & M. Rosenthal (Eds.), Handbook of rehabilitation psychology (2nd ed., pp ). Washington, DC: American Psychological Association Press. Bonanno, G. A., Kennedy, P., Galatzer-Levy, I. R., Lude, P., & Elfström, M. L. (2012). Trajectories of resilience, depression, and anxiety following spinal cord injury. Rehabilitation Psychology, 57, doi: /a Bonanno, G. A., Westphal, M., & Mancini, A. D. (2011). Resilience to loss and potential trauma. Annual Review of Clinical Psychology, 7, doi: /annurev-clinpsy Bond, T. G., & Fox, C. M. (2001). Applying the Rasch model: Fundamental measurement in the human sciences. Mahwah, NJ: Lawrence Erlbaum. Boyce, C. J., & Wood, A. M. (2011). Personality prior to disability determines adaptation: Agreeable individuals recover lost life satisfaction faster and more completely. Psychological Science, 22, doi: / Carr, D., & Springer, K. W. (2010). Advances in families and health research in the 21st century. Journal of Marriage and the Family, 72, doi: /j x Civil, I. D., & Schwab, C. W. (1988). The abbreviated injury scale, 1985 revision: A condensed chart for clinical use. The Journal of Trauma, 28, Committee on Injury Scaling. (1985). The Abbreviated Injury Scale: 1985 Revision. Morton Grove, IL: American Association for Automotive Medicine. Corrigan, J. D., Harrison-Felix, C., Bogner, J., Dijkers, M., Terrill, M. S., & Whiteneck, G. (2003). Systematic bias in traumatic brain injury outcome studies because of loss to follow-up. Archives of Physical Medicine and Rehabilitation, 84, doi: /apmr Corry, N., Pruzinsky, T., & Rumsey, N. (2009). Quality of life and psychosocial adjustment to burn injury: Social functioning, body image, and health policy perspectives. International Journal of Psychiatry, 21, doi: / de Groot, S., Haisma, J. A., Post, M. W. M., van Asbeck, F. W. A., & van der Woude, L. H. V. (2009). Investigation of bias due to loss of participants in a Dutch multicentre prospective spinal cord injury cohort study. Journal of Rehabilitation Medicine, 41, doi: / de Roon-Cassini, T. A., Rusch, M. D., Mancini, A. D., & Bonanno, G. A. (2010). Psychopathology and resilience following traumatic injury: A latent growth mixture model analysis. Rehabilitation Psychology, 55, doi: /a Diener, E., Lucas, R. E., & Scollon, C. N. (2006). Beyond the hedonic treadmill: Revising the adaptation theory of well-being. American Psychologist, 61, doi: / x Dijkers, M. P. (2005). Quality of life of individuals with spinal cord injury: A review of conceptualization, measurement, and research findings. Journal of Rehabilitation Research & Development, 42, doi: /JRRD

8 190 HERNANDEZ ET AL. Dunn, D. S., & Brody, C. (2008). Defining the good life following acquired physical disability. Rehabilitation Psychology, 53, doi: /a Dunn, D. S., Uswatte, G., & Elliott, T. (2009). Happiness, resilience and positive growth following disability: Issues for understanding, research and therapeutic intervention. In S. J. Lopez (Ed.), The Oxford handbook of positive psychology (2nd ed., pp ). New York, NY: Oxford University Press. doi: /oxfordhb/ EnglandKennedy, E. S., & Horton, S. (2011). Disability, participation, and subjective well being among older couples. Social Science & Medicine, 73, doi: /j.socscimed Erosa, N. A., Berry, J. W., Elliott, T. R., Underhill, A., & Fine, P. R. (in press). Predicting quality of life five years after medical discharge for traumatic spinal cord injury. British Journal of Health Psychology. doi: /bjhp Frederick, S., & Loewenstein, G. (1999). Hedonic adaptation. In D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: The foundations of hedonic psychology (pp ). New York, NY: Sage. Freedman, V. A., Stafford, F., Schwarz, N., Conrad, F., & Cornman, J. C. (2012). Everything that I thought they would be, they weren t : Family systems as support and impediment to recovery. Social Science & Medicine, 74, doi: /j.socscimed Fujita, F., & Diener, E. (2005). Life satisfaction set point: Stability and change. Journal of Personality and Social Psychology, 88, doi: / Gibran, N. S., Wiechman, S., Meyer, W., Edelman, L., Fauerbach, J., Gibbons, L.,... Wiggins, B. (2013). American Burn Association consensus statement. Journal of Burn Care & Research, 34, doi: /bcr.0b013e31828cb249 Holavanahalli, R. K., Lezotte, D. C., Hayes, M., Minhajuddin, A., Fauerbach, J., Engrav, L.,... Kowalske, K. (2006). Profile of patients lost to follow-up in the burn injury rehabilitation model systems longitudinal database. Journal of Burn Care & Research, 27, doi: / 01.BCR Johnson, C. L., Resch, J. A., Villarreal, V., Elliott, T. R., Kwok, O.-M., Berry, J. W., & Underhill, A. T. (2010). Family satisfaction predicts life satisfaction trajectories over the first five years after traumatic brain injury. Rehabilitation Psychology, 55, doi: /a Kalpinski, R., Williamson, M. L. C., Elliott, T. R., Berry, J. W., Underhill, A. T., & Fine, P. R. (2013). Modeling the prospective relationships of impairment, injury severity and participation to quality of life following traumatic brain injury. BioMed Research International, 2013, article doi: /2013/ Keith, R. A., Granger, C. V., Hamilton, B. B., & Sherwin, F. S. (1987). The functional independence measure: A new tool for rehabilitation. Advances in Clinical Rehabilitation, 1, Klein, M. B., Lezotte, D. C., Heltshe, S., Fauerbach, J., Holavanahalli, R. K., Rivara, F. P.,... Engrav, L. (2011). Functional and psychosocial outcomes of older adults after burn injury: Results from a multicenter database of severe burn injury. Journal of Burn Care & Research, 32, doi: /bcr.0b013e a Klinge, K., Chamberlain, D. J., Redden, M., & King, L. (2009). Psychological adjustments made by postburn injury individuals: An integrative literature review. Journal of Advanced Nursing, 65, doi: /j x Kwok, O. M., Underhill, A. T., Berry, J. W., Luo, W., Elliott, T. R., & Yoon, M. (2008). Analyzing longitudinal data with multilevel models: An example with individuals living with lower extremity intra-articular fractures. Rehabilitation Psychology, 53, doi: / a Linacre, J. M. (2003). A user s guide to Winsteps/Ministep: Rasch-model computer programs. Chicago, IL: University of Chicago. Lucas, R. E. (2007a). Adaptation and the set-point model of subjective well-being: Does happiness change after major life events? Current Directions in Psychological Science, 16, doi: /j x Lucas, R. E. (2007b). Long-term disability is associated with lasting changes in subjective well-being: Evidence from two nationally representative longitudinal studies. Journal of Personality and Social Psychology, 92, doi: / Lucas, R. E., Clark, A. E., Georgellis, Y., & Diener, E. (2003). Reexamining adaptation and the set point model of happiness: Reactions to changes in marital status. Journal of Personality and Social Psychology, 84, doi: / Lyubomirsky, S., Sheldon, K. M., & Schkade, D. (2005). Pursuing happiness: The architecture of sustainable change. Review of General Psychology, 9, doi: / MacKenzie, E. J., Steinwachs, D. M., & Shankar, B. (1989). Classifying trauma severity based on hospital discharge diagnoses: Validation of an ICD9-CM to AIS 85 conversion table. Medical Care, 27, doi: / Mausbach, B. T., Chattillion, E. A., Moore, R. C., Roepke, S., Depp, C. A., & Roesch, S. (2011). Activity restriction and depression in medical patients and their caregivers: A meta-analysis. Clinical Psychology Review, 31, doi: /j.cpr McDowell, I. (2006). Measuring health: A guide to rating scales and questionnaires (3rd ed.). New York, NY: Oxford University Press. Müller, R., Peter, C., Cieza, A., & Geyh, S. (2012). The role of social support and social skills in people with spinal cord injury: A systematic review of the literature. Spinal Cord, 50, doi: /sc Neugarten, B. L., Havighurst, R. J., & Tobin, S. S. (1961). The measurement of life satisfaction. Journal of Gerontology, 16, doi: /geronj/ Olson, D. (2011). FACES IV and the circumplex model: Validation study. Journal of Marital and Family Therapy, 37, doi: /j x Olson, D. H., Russell, C., & Sprenkle, D. H. (Eds.). (1983). Circumplex model: Systemic assessment and treatment of families (2nd ed.). New York, NY: Haworth. Olson, D. H., & Wilson, M. (1982). Family inventories: Inventories used in a national survey of families across the family life cycle. In D. H. Olson, H. I. McCubbin, H. Barnes, A. Larsen, M. Muxen, & M. Wilson (Eds.), Family social science (pp ). St. Paul, MN: University of Minnesota. Patterson, D. R., Ptacek, J. T., Cromes, F., Fauerbach, J. A., & Engrav, L. (2000). Describing and predicting distress and satisfaction with life for burn survivors. Journal of Burn Care and Rehabilitation, 21, doi: / Post, M. W., & van Leeuwen, C. M. C. (2012). Psychosocial issues in spinal cord injury: A review. Spinal Cord, 50, doi: /sc Powdthavee, N. (2009). What happens to people before and after disability? Focusing effects, lead effects and adaptation in different areas of life. Social Science & Medicine, 69, doi: /j.socscimed Putzke, J. D., Elliott, T., & Richards, J. S. (2001). Marital status and adjustment one year post-spinal cord injury. Journal of Clinical Psychology in Medical Settings, 8, doi: /a: Raudenbush, S. W., & Bryk, A. S. (2002). Hierarchical linear models: Applications and data analysis methods (2nd ed.). Thousand Oaks, CA: Sage. Resch, J. A., Villarreal, V., Johnson, C. L., Elliott, T. R., Kwok, O., Berry, J. W. & Underhill, A. (2009). Trajectories of life satisfaction in the first 5 years following traumatic brain injury. Rehabilitation Psychology, 54, doi: /a

9 TRAJECTORIES OF LIFE SATISFACTION 191 Rogers, B., & Kennedy, P. (2000). A qualitative analysis of reported coping in a community sample of people with spinal cord injuries: The first year post discharge. SCI Psychosocial Process, 13, Ryff, C. D. (1989). Happiness is everything, or is it? Explorations on the meaning of well-being. Journal of Personality and Social Psychology, 57, doi: / Snijders, T. A. B., & Bosker, R. J. (1994). Modeled variance in two-level models. Sociological Methods & Research, 18, Snijders, T. A. B., & Bosker, R. J. (2012). Multilevel analysis: An introduction to basic and advanced multilevel modeling (2nd ed.). Thousand Oaks, CA: Sage. Underhill, A. T., Lobello, S. G., & Fine, P. R. (2004). Reliability and validity of the family satisfaction scale with survivors of traumatic brain injury. Journal of Rehabilitation Research and Development, 41, doi: /jrrd van Leeuwen, C. M. C., Post, M. W. M., van Asbeck, F. W. A., Bongers- Janssen, H., van der Woude, L., de Groot, S., & Lindeman, E. (2012). Life satisfaction in people with spinal cord injury during the first five years after discharge from inpatient rehabilitation. Disability and Rehabilitation: An International, Multidisciplinary Journal, 34, doi: / Vaux, A. (2000). Social support: Theory, research, and intervention. New York: NY: Praeger. Wallace, K. A., & Wheeler, A. J. (2002). Reliability generalization of the Life Satisfaction Index. Educational and Psychological Measurement, 62, doi: / Initial analyses examined the overall relationship of functional impairment on life satisfaction in individuals who have sustained a spinal cord injury, burns, or an intra-articular fracture. This one-level model is represented as LSI ti 0i 1i Time ti 2i FIM ti 3i FIM ti * Time ti e ti In this equation, life satisfaction was the outcome measure (LSI); i represents an individual, with t representing a specific time point; 1i represents slope parameters that represent linear rates of change over time, and 2i and 3i represent the linear relationship between the respective variable and LSI controlling for other variables. Interaction effects, noted as 4i, 5i, 6i, and 7i, represent linear rates of change over time as a function of respective variables. The individual intercept is represented by 0i, and e ti represents within-individual error. More complex models followed to assess the combined effect of family satisfaction and functional impairment on life satisfaction over time: LSI ti 0i 1i Time ti 2i FIM ti 3i FSS ti 4i FIM ti * Time ti 5i FSS ti * Time ti 6i FIM ti * FSS ti 7i FIM ti * FSS ti * Time ti e ti Further building upon initial models incorporated a multilevel model, incorporating injury type as a higher order predictor of life satisfaction, to examine possible differences in growth trajectories by injury group. This equation was structured as follows: Level 1: LSI ti 0i 1i Time ti 2i FIM ti 3i FSS ti 4i FIM ti * Time ti 5i FSS ti * Time ti 6i FIM ti * FSS ti 7 i FIM ti * FSS ti * Time ti e ti Appendix Description of Statistical Models Level 2: 0i Injury Type U 0i 1i 10 2i 20 3i 30 4i 40 5i 50 6i 60 7i 70 The combined model is as follows: LSI ti Injury Type U 0i 10 * Time ti 20 FIM ti 30 * FSS ti 40 FIM ti * Time ti 50 FSS ti * Time ti 60 FIM ti * FSS ti 70 FIM ti * FSS ti * Time ti e ti LSI ti is an outcome measure of life satisfaction for participant i at time t, and 00 is a Level 2 estimate of the mean population value for initial status. 10 is the average rate of change in LSI, and 20 and 30 are the average relation between LSI and FIM, and LSI and FSS, respectively. Between-individual error is represented by U 0i, and within-individual error is represented by e ti. 40, 50, and 70 examine the potential FIM and FSS by time interaction effects, and 60 examines the interaction of FIM and FSS with LSI. A combined equation modeled the complex relationship of functional impairment and family satisfaction with the additional variables of injury type and marital status: (Appendix continues)