Longitudinal Association of Health Satisfaction With Functional Status and Income Satisfaction in Stroke Survivors

Abstract

Importance: Although cross-sectional studies have reported the relationships among performance of activities of daily living (ADLs), income satisfaction, and health satisfaction, longitudinal associations in stroke survivors remain unclear.

Objective: To examine the effects of ADL function and income satisfaction on health satisfaction over time in stroke survivors.

Design: Retrospective longitudinal study designs with latent growth curve models (LGMs) to control for time-varying and time-invariant covariates.

Setting: Community.

Participants: A total of 198 stroke survivors.

Outcomes and Measures: Self-reported ADL performance, income satisfaction, and health satisfaction.

Results: The average age of participants was 70.68 yr (SD = 8.09; 107 men, 91 women). Each LGM showed that an increase in ADL score (standardized β = 0.116∼0.296, all ps < .05) and income satisfaction (standardized β = 0.513∼0.726, all ps < .001) positively predicted health satisfaction over time, even after controlling for time-varying annual income, demographics, and behavioral characteristics. The fit results of all LGMs were within the acceptable range: χ²(47) = 66.378, p = .0327; χ²(47) = 57.742, p = .1355; root-mean-square error of approximation, <0.08; comparative fit index and Tucker–Lewis index, >0.90; standardized root-mean-square residual, <0.05.

Conclusions and Relevance: Because ADLs are significant modifiable health satisfaction factors, occupational therapy practitioners would need to focus more closely on ADL training in clinical settings. Furthermore, referring those clients to appropriate vocational rehabilitation to improve their income satisfaction would be necessary.

Plain-Language Summary: This study recognizes the unique challenges and opportunities that arise when stroke survivors are preparing to return to their communities by emphasizing the significance of ADL training and income satisfaction during this stage of recovery. The study suggests that the therapeutic use of ADL training and income satisfaction could increase health satisfaction for stroke survivors. Therefore, comprehensive ADL training, as an occupational therapy intervention, could be crucial for stroke survivors who are preparing to return to their community from a clinical setting. Furthermore, connecting with vocational rehabilitation could also be important for improving income satisfaction for stroke survivors who are preparing to return to a community.

As an occupational therapy intervention, comprehensive ADL training could be crucial for stroke survivors who are preparing to return to their community from a clinical setting.

In Korea, the number of patients with stroke reached 120,000 in 2021 (Korean Statistical Information Service, 2021). Stroke is a clinically defined syndrome of neurological deficits caused by damage to the blood vessels of the central nervous system (Murphy & Werring, 2020). Stroke is associated with a decline in health-related quality of life (QoL). Abubakar and Isezuio (2012) reported that lower stroke severity was associated with improved health-related QoL. Carod-Artal et al. (2000) identified key factors determining the QoL in stroke survivors, including neurological impairment severity. Cerniauskaite et al. (2012) demonstrated that cerebrovascular disease, particularly severe stroke, has a significant impact on a patient’s QoL, with implications for functional disability factors.

The functional impairment that has the most impact on the health-related QoL of patients with stroke is a decline in performance of their basic activities of daily living (BADLs) and instrumental activities of daily living (IADLs). The performance of activities of daily living (ADLs)—or ADL function, defined as the performance of a combination of BADLs and IADLs—is one of the most basic elements. Accordingly, ADL intervention is part of occupational therapy practice for treating patients with stroke, and several positive effects have been observed, including improvements in BADL performance, IADL performance, and social participation (Steultjens et al., 2003). Moreover, it has been reported that the rehabilitation goal desired by most patients with stroke is ADL and IADL function recovery (Waddell et al., 2016).

Functional disabilities caused by stroke can cause difficulties not only with performing ADLs but also with returning to work. Failure to return to work because of functional impairment from a stroke can create personal income challenges. Furthermore, this loss of income can impede access to rehabilitation services and slow functional recovery (Horner et al., 2003). According to a previous study, among 58 patients with stroke 1 yr after discharge, 83% still had cognitive dysfunction; 20% were dependent in performing ADLs, and 80% had a limited or minimal return to work (Ashley et al., 2019; Edwards et al., 2018; Hofgren et al., 2007; Treger et al., 2007; Wang et al., 2014). Decreased ADL function and reduced income because of stroke are associated with low QoL (Kwon et al., 2018; Li et al., 2023), which can be categorized into physical, material, social, emotional, developmental, and activity QoL (Felce & Perry, 1995). Among these, ADL function and income could be related to both physical and mental well-being. In other words, it goes beyond mere financial considerations and can also affect their psychological well-being and satisfaction. Furthermore, individuals’ or groups’ satisfaction with their income can be shaped when compared with their income expectations. Sometimes, even with a high income, if one fails to meet the higher income expectations, their satisfaction may be low. Conversely, with a lower income, if there are no high income expectations, satisfaction can be high (Clark et al., 2008; D’Ambrosio & Frick, 2007; Ferrer-i-Carbonell & Frijters, 2004). Therefore, both income and income satisfaction can serve as significant variables in the context of both physical and mental well-being.

Several studies reported relationships among ADL function, income satisfaction, and health satisfaction. Health satisfaction is a concept that typically represents a person’s satisfaction with their health-related state. This term can encompass various aspects related to physical health, mental health, and well-being (Diener et al., 1985; Khanna & Tsevat, 2007). Health satisfaction can be influenced by a multitude of factors, including personal experiences, social environments, cultural factors, and more. Kanayama et al. (2016) reported a positive relationship between ADL function and QoL in 40 patients with stroke. Katzan et al. (2018) stated a negative relationship between satisfaction with physical functioning, social role, and income. Kwon et al. (2018) noted that a decrease in the QoL of patients with stroke was related to a reduction in their income status, a decrease in subjective health status, and an increase in activity limitation.

Although previous studies reported significant relationships among ADL function, income satisfaction, and health satisfaction in patients with stroke, these studies were cross-sectional (Ho et al., 2002; Kahneman & Deaton, 2010; Kanayama et al., 2016). A major limitation of cross-section designs is that they cannot indicate causal directions among complicatedly related characteristics. Additionally, this type of research design cannot control for time-varying characteristics over time (e.g., annual income, ADL function, and income satisfaction). Therefore, for examining the effects of ADL function and income satisfaction on health satisfaction, longitudinal study designs would be appropriate. Furthermore, latent growth curve modeling approaches can be used to examine which factors influence the trajectory of health satisfaction over time by controlling for various time-invariant characteristics (e.g., baseline demographics and health behavior characteristics) and time-varying characteristics (e.g., annual income, ADL function, and income satisfaction). However, because health behavior characteristics can change over time (e.g., alcohol consumption, hypertension, diabetes), these characteristics also need to be treated as time-varying factors in latent growth curve models (LGMs). Therefore, in this study, we investigated the longitudinal effects of ADL function and income satisfaction on health satisfaction using an LGM. Our research hypotheses were that higher levels of ADL function will have a positive impact on the trajectory of health satisfaction over time in patients with stroke and that higher levels of income satisfaction will have a positive effect on the trajectory of health satisfaction over time in patients with stroke.

Method

Data Sources

We used data from longitudinal panel databases of the Korean Longitudinal Study on Aging (KLoSA) that were collected by the Korea Employment Information Service. A total of four waves from 2014 to 2020 were used. From 2006 to 2020, the KLoSA collected data from a national sample in every even-numbered year, for a total of eight waves of data being collected (Boo & Chang, 2006). The KLoSA includes information on household membership, health status, employment, income and consumption, assets, subjective expectations and QoL, and mortality rates (Boo & Chang, 2006). We extracted data of individuals with stroke from the 2014–2020 KLoSA databases. The inclusion criteria include patients with stroke and those who responded to all panel surveys from 2014 to 2022.

Dependent Variables

We used health satisfaction as the dependent variable and extracted this information from four time points of the 2014, 2016, 2018, and 2020 KLoSA databases. The KLoSA collected the health satisfaction information using the self-reported question, “How satisfied are you with your health condition?” The closer the score was to 0, the lower the satisfaction with their health condition; the closer the score was to 100, the higher the satisfaction (Boo & Chang, 2006).

Time-Varying Independent Variables for Latent Growth Curve Models 2.a and 2.b

We used two time-varying independent variables, including ADL function (Model 2.a) and income satisfaction (Model 2.b). We extracted each piece of information from the four time points of the 2014, 2016, 2018, and 2020 KLoSA databases. We measured ADL function, defined as a combination of BADL and IADL performance, by assessing the extent to which survey participants were independent in 17 daily tasks: dressing, washing and brushing, bathing or showering, eating, getting in and out of bed, using the toilet, managing bowel control, grooming, performing housework, preparing meals, performing laundry, going out using transportation, going out without using transportation, shopping, managing money, receiving calls, and managing medications. The response categories of the 17 ADL functions were dichotomized (1 = no help needed, and 0 = needed help). We used the total score of the 17 ADL tasks (scores ranged from 0 to 17 points) so that higher scores indicated better ADL function. Income satisfaction was measured using a self-reported 10-point scale; higher scores indicated higher income satisfaction (Boo & Chang, 2006).

Covariates

The baseline demographics and health behavior characteristics were used as time-invariant covariates. Baseline demographics included age (continuous), gender (1 = male, 0 = female), educational attainment (1 = elementary school, 2 = middle school, 3 = high school, 4 = college degree), employment status (1 = employed, 0 = unemployed), residential area (1 = urban, 2 = suburban, and 3 = rural), and annual income (continuous). Because the annual income was inconsistent over the years, we treated this variable as a time-varying covariate. Health behavior characteristics included alcohol consumption (1 = yes, 0 = no), hypertension (1 = having it, 0 = not having it) and diabetes (1 = having it, 0 = not having it). However, alcohol consumption, hypertension, and diabetes, which have the potential to change over time, were additionally used as time-varying independent variables in sensitivity analysis.

Statistical Analysis

We used LGM approaches to elucidate the growth path of dependent variables by estimating growth factors, including the intercept (i) representing the initial value and the slope (s) representing the rate of change over time (Muthén & Muthén, 2007). For LGM estimations, we fixed the intercept loadings at 1.0 to estimate a constant range of individual differences for the initial average value. Van Mierlo et al. (2016) suggested that health satisfaction, as a component of QoL, can vary on the basis of the level of dependence or independence in ADL among patients. Therefore, assuming that the slope loadings have a linear change, the path coefficients of the rate of change factor were fixed at 0.0, 1.0, 2.0, and 3.0 (Muthén & Muthén, 2007). First, we used an unconditional LGM to identify growth patterns in health satisfaction and examine whether the trajectory pattern was well fitted into the unconditional LGM (Model 1).

Next, we added time-varying and time-invariant covariates into the unconditional LGM and investigated the longitudinal associations between ADL function (Model 2.a; see Figure A.1 in the Supplemental Material, available online with this article at https://research.aota.org/ajot) and health satisfaction, as well as between income satisfaction and health satisfaction (Model 2.b; see Figure A.2). Moreover, we examined whether baseline demographics and health behavior characteristics affect latent growth factors (intercept i and slope s). In summary, Model 1 was used to determine how the growth pattern of health satisfaction changes over time. Next, Models 2.a and 2.b were used to examine whether time-varying ADL function and income satisfaction influence the trajectory of health satisfaction over 6 yr and whether time-invariant covariates influence individual differences in latent growth factors (intercept i and slope s) of health satisfaction.

We estimated the parameters of LGMs using maximum likelihood (Muthén & Muthén, 2007). We examined the model fit of LGMs using model fit indices, including χ², comparative fit index (CFI), Tucker–Lewis index (TLI), root-mean-square error of approximation (RMSEA), and standardized root-mean-square residual (SRMR). For χ², a p value of .05 or higher was considered a good fit. Additionally, the model fit was examined using the index (χ²/df), obtained by dividing the test statistic value of χ² by the degree of freedom. A value less than 2.0 was considered a good fit (Cangur & Ercan, 2015). We considered the CFI and TLI model fit criteria with a value of ≥0.95 a good fit, and we considered the RMSEA with a value within 0.08 an acceptable fit (Bentler, 1990; Browne & Cudeck, 1993; Cangur & Ercan, 2015; Hu & Bentler, 1999; Tucker & Lewis, 1973). For SRMR, we considered values within 0.05 as an acceptable fit (Diamantopoulos et al., 2000). We used SAS, Version 9.4, for data management and the statistical program Mplus, Version 8.4, for statistical analyses (Muthén & Muthén, 2017).

Results

The demographic characteristics of the study sample are presented in Table 1. The average age was 70.68 yr (SD = 8.09), and the majority of the sample was male (n = 107; 54.04%). Most participants completed elementary school (n = 102; 51.52%) and were unemployed (n = 152; 76.77%). The urban area was the most common place of residence, with 80 stroke survivors (40.40%), followed by suburban and rural areas with 72 (36.36%) and 46 (23.23%) stroke survivors, respectively. Finally, health satisfaction increased over time from 2014 (M = 47.27, SD = 19.45) to 2020 (M = 51.01, SD = 19.82).

Table 1.

Participant Demographics (N = 198)

Study Variables	n (%)
Age, M (SD)	70.68 (8.09)
Gender
Male	107 (54.04)
Female	91 (45.96)
Educational attainment
Elementary school	102 (51.52)
Middle school	35 (17.68)
High school	43 (21.72)
College degree	18 (9.09)
Employment status
Employed	46 (23.23)
Unemployed	152 (76.77)
Residential area
Urban	80 (40.40)
Suburban	72 (36.36)
Rural	46 (23.23)
Drinking	46 (23.23)
Chronic conditions
Hypertension	137 (69.19)
Diabetes	57 (28.79)
Health satisfaction, M (SD)
2014	47.27 (19.45)
2016	47.37 (19.52)
2018	47.63 (20.75)
2020	51.01 (19.82)

Model 1: Unconditional Latent Growth Curve Model

The estimations of the latent growth factors in the unconditional LGM and its model fit are shown in Table 2. The unconditional LGM demonstrated good model fits, including χ²(5) = 4.778, p = .443, χ²/df = 0.956, RMSEA = 0.000, CFI = 1.000, TLI = 1.000, and SRMR = 0.046. The initial average value (i) of health satisfaction with Model 1 was 46.798 (SE = 1.344, p < .001), and the rate of change (s) was 1.083 (SE = 0.506, p = .032). The rate of change (s) was 1.083, indicating that, on average, health satisfaction increased by 1.083 units per every 2 yr. The covariance of i and s was −37.591 (SE = 13.441, p = .005), indicating a negative relationship between the initial health satisfaction score and its rate of change over time. In other words, the negative correlation indicates that participants with higher baseline health satisfaction scores showed a lower slope value (less change over time) than those with lower baseline health satisfaction scores.

Table 2.

Model 1 Parameter Estimates for Unconditional LGM of Health Satisfaction Change Over Time

Growth Factors (Health Satisfaction)	Unstandardized			Standardized
Growth Factors (Health Satisfaction)	B	SE	p	β	SE	p
Intercept (i) mean	46.798	1.344	<.001**	2.868	0.225	<.001**
Slope (s) mean	1.083	0.506	.032*	0.245	0.122	.045*
Intercept (i) ↔ slope (s) covariance	−37.591	13.441	.005*	−0.522	0.101	<.001**

Note. Model fit: χ²(5) = 4.778, p = .443; χ²/df = 0.956; CFI = 1.000; TLI = 1.000; RMSEA = 0.000; SRMR = 0.046. β = standardized coefficient; B = unstandardized coefficient; CFI = comparative fit index; LGM = latent growth curve model; RMSEA = root-mean-square error of approximation; SRMR = standardized root-mean-square residual; TLI = Tucker–Lewis Index.

*p < .05. **p < .001.

Model 2.a: Latent Growth Curve Model for ADL Function and Health Satisfaction

The parameter estimations of Model 2.a, including the time-varying ADL functions and covariates, are depicted in Table 3. Model 2.a demonstrated marginal model fits, including χ²(47) = 66.378, p = .032; χ²/df = 1.412, RMSEA = 0.046, CFI = 0.939, TLI = 0.904, and SRMR = 0.051. We considered that Model 2.a showed a good model fit (χ²/df < 2.0), even though the p value of the χ² value was .032.

Table 3.

Model 2.a Parameter Estimates for LGM of Health Satisfaction Over Time With Time-Varying (ADL Function) and Time-Invariant Covariates

Model 2.a: Growth Factors (Health Satisfaction)	Unstandardized			Standardized
Model 2.a: Growth Factors (Health Satisfaction)	B	SE	p	β	SE	p
Intercept (i) ↔ slope (s) covariance	−31.911	12.090	.008*	−0.583	0.105	<.001**
Intercept (i) mean	19.495	14.548	.180	1.255	0.942	.183
Slope (s) mean	−1.390	6.108	.820	−0.328	1.474	.820
Study variables (intercept)
Age	−0.069	0.168	.683	−0.036	0.087	.683
Gender	0.160	2.832	.955	0.005	0.091	.955
Educational attainment	4.523	1.335	.001*	0.303	0.087	<.001**
Employment status	7.208	3.127	.021*	0.196	0.084	.019*
Residential area	2.256	1.574	.152	0.113	0.078	.148
Drinking	2.801	3.164	.376	0.076	0.086	.375
Hypertension	−0.374	2.698	.890	−0.011	0.080	.890
Diabetes	−4.540	2.722	.095	−0.132	0.079	.095
Study variables (Slope)
Age	0.067	0.068	.326	0.127	0.129	.327
Gender	0.800	1.134	.480	0.094	0.134	.482
Educational attainment	0.399	0.538	.458	0.098	0.132	.458
Employment status	−1.662	1.249	.183	−0.166	0.125	.184
Residential area	0.195	0.629	.757	0.036	0.116	.757
Drinking	−2.018	1.263	.110	−0.201	0.128	.116
Hypertension	−1.245	1.081	.250	−0.136	0.118	.252
Diabetes	−0.304	1.083	.779	−0.032	0.116	.779
Time-varying covariates
ADL function
2014	0.655	0.314	.037*	0.116	0.056	.037*
2016	1.187	0.270	<.001**	0.223	0.051	<.001**
2018	1.492	0.272	<.001**	0.296	0.054	<.001**
2020	1.188	0.272	<.001**	0.270	0.061	<.001**
Annual income
2014	1.501	0.704	.033*	0.115	0.054	.032*
2016	−0.249	0.577	.666	−0.021	0.048	.665
2018	−1.321	0.666	.047*	−0.099	0.049	.046*
2020	−0.557	0.894	.534	−0.032	0.051	.535

Note. Model fit: χ²(47) = 66.378, p = .032; χ²/df = 1.412; CFI = 0.939; TLI = 0.904; RMSEA = 0.046; SRMR = 0.051. ADL function = performance of a combination of basic and instrumental activities of daily living; β = standardized coefficient; B = unstandardized coefficient; CFI, comparative fit index; LGM = latent growth curve model; RMSEA = root-mean-square error of approximation; SRMR = standardized root-mean-square residual; TLI = Tucker–Lewis Index.

*p < .05. **p < .001.

Model 2.a showed that the initial average value (i) of health satisfaction was 19.495 (SE = 14.548, p = .180), and the rate of change (s) was −1.390 (SE = 6.108), p = .820. Similar to the unconditional LGM, the covariance of i and s was −31.911 (SE = 12.090, p = .008). Positive relationships between ADL function and health satisfaction over time were observed (unstandardized regression coefficients ranged from 0.655 to 1.492; all ps < .05), even after controlling for the time-varying annual income and covariates (Table 3). For the time-invariant covariates in the relationship between ADL function and health satisfaction (Model 2.a in Table 3), we observed a significant difference in the initial average value (i) only in educational attainment (B = 4.523, SE = 1.335, p = .001) and employment status (B = 7.208, SE = 3.127, p = .021; Table 3).

Last, we conducted sensitivity analyses by treating health behavior characteristics (i.e., alcohol consumption, hypertension, diabetes) as time-varying covariates from 2014 to 2020. The sensitivity analysis results were consistent, indicating that the direction and significance of those variables have not changed (Table A.1).

Model 2.b: Latent Growth Curve Model for Income Satisfaction and Health Satisfaction

Similarly, the parameter estimations of Model 2.b, including the time-varying income satisfaction and time-invariant covariates, are depicted in Table 4. Furthermore, Model 2.b showed good model fits, including χ²(47) = 57.742, p = .135; χ²/df = 1.228, RMSEA = 0.034, CFI = 0.982, TLI = 0.972, and SRMR = 0.042. Model 2.b showed that the initial average value (i) of health satisfaction was 13.537 (SE = 11.117, p = .223), and the rate of change (s) was −1.622 (SE = 4.852, p = .738). Similar to the unconditional LGM and Model 2.a, this LGM also demonstrated a negative covariance between i and s, −27.268 (SE = 8.897, p = .002). Additionally, positive relationships between income satisfaction and health satisfaction over time were observed (unstandardized regression coefficients ranged from 0.501 to 0.722; all ps < .001) even after controlling for the time-varying annual income and time-invariant covariates. For the time-invariant covariates in the relationship between income satisfaction and health satisfaction (see Model 2.b in Table 4), we noted a significant difference in the initial average value (i) only in employment status, B = 6.676 (SE = 2.587, p = .010; Table 4).

Table 4.

Model 2.b Parameter Estimates for LGM of Health Satisfaction Over Time With Time-Varying (Income Satisfaction) and Time-Invariant Covariates

Model 2.b: Growth Factors (Health Satisfaction)	Unstandardized			Standardized
Model 2.b: Growth Factors (Health Satisfaction)	B	SE	p	β	SE	p
Intercept (i) ↔ slope (s) covariance	−27.268	8.897	.002*	−0.743	0.080	<.001**
Intercept (i) mean	13.537	11.117	.223	1.205	0.997	.227
Slope (s) mean	−1.622	4.852	.738	−0.431	1.287	.738
Study variables (intercept)
Age	−0.009	0.138	.949	−0.006	0.099	.949
Gender	−0.541	2.331	.817	−0.024	0.103	.817
Educational attainment	1.927	1.128	.088	0.178	0.103	.083
Employment status	6.676	2.587	.010*	0.251	0.095	.008*
Residential area	−0.250	1.322	.850	−0.017	0.092	.850
Drinking	2.656	2.611	.309	0.100	0.098	.309
Hypertension	−0.118	2.211	.957	−0.005	0.091	.957
Diabetes	−2.167	2.253	.336	−0.087	0.090	.334
Study variables (Slope)
Age	0.028	0.058	.636	0.059	0.125	.636
Gender	−0.415	0.985	.673	−0.055	0.130	.673
Educational attainment	0.134	0.484	.782	0.037	0.134	.782
Employment status	−1.397	1.090	.200	−0.157	0.123	.201
Residential area	0.913	0.558	.102	0.189	0.117	.106
Drinking	−1.778	1.101	.106	−0.200	0.125	.111
Hypertension	−0.519	0.935	.579	−0.064	0.115	.579
Diabetes	−0.048	0.950	.959	−0.006	0.114	.959
Time-varying covariates
Income satisfaction
2014	0.501	0.054	<.001**	0.513	0.051	<.001**
2016	0.636	0.048	<.001**	0.622	0.040	<.001**
2018	0.606	0.049	<.001**	0.611	0.041	<.001**
2020	0.722	0.045	<.001**	0.726	0.035	<.001**
Annual income
2014	0.954	0.636	.134	0.073	0.049	.134
2016	−0.449	0.460	.329	−0.039	0.039	.324
2018	−0.203	0.525	.700	−0.015	0.039	.698
2020	−1.083	0.656	.099	−0.062	0.038	.098

Note. Model fit: χ²(47) = 57.742, p = .135; χ²/df = 1.228; CFI = 0.982; TLI = 0.972; RMSEA = 0.034; SRMR = 0.042. β = standardized coefficient; B = unstandardized coefficient; CFI = comparative fit index; LGM = latent growth curve model; RMSEA = root-mean-square error of approximation; SRMR = standardized root-mean-square residual; TLI = Tucker–Lewis Index.

*p < .05. **p < .001.

Discussion

We investigated the association between ADL function and health satisfaction and examined the association between income satisfaction and health satisfaction in patients with stroke using even-numbered-year data from 2014 to 2020. The study findings supported our research hypotheses that higher levels of ADL functions and income satisfaction predict more favorable trajectories of health satisfaction over time in patients with stroke. These findings highlight the significance of promoting ADL function and income satisfaction as integral factors in enhancing long-term health satisfaction outcomes for stroke survivors.

We observed that ADL function and income satisfaction were positively associated with the trajectory pattern of health satisfaction. Although previous studies used cross-sectional designs, they reported similar findings (van Mierlo et al., 2016; Wang et al., 2019). For instance, van Mierlo et al. (2016) reported that ADL function was positively related to QoL in patients with stroke. Moreover, ADL function was a significant factor in the psychosocial change in this patient population (Wang et al., 2019). Further, our longitudinal study using LGMs showed that ADL function was positively associated with health satisfaction over time. Occupational therapy supports patients with stroke in successfully performing ADLs. When patients engage in independent activities and interact with their social environment in their daily lives, their health satisfaction can improve. On the basis of the results of previous studies and the findings of this current study, it can be considered that occupational therapy can contribute to improving patients’ adaptation to daily life and promoting social participation, thereby enhancing their health satisfaction. However, in contrast to the previous cross-sectional studies (van Mierlo et al., 2016; Wang et al., 2019), our LGMs demonstrated that negative correlations were noted between the initial average score of health satisfaction (intercept i) and rate of change (slope s). This negative correlation indicates that people with a higher health satisfaction score showed a slower rate of increase in their health satisfaction score over time (Muthén & Muthén, 2007). According to previous studies, people with stroke may experience rapid improvement during the initial stages of recovery; however, over time, they may encounter limitations in the pace of their recovery (Duncan et al., 1992; Hackett & Anderson, 2000; Kwakkel et al., 2006). This trend suggests a gradual decline in the health satisfaction of stroke patients, emphasizing the importance of long-term health management.

Income satisfaction could also be a significant part of health satisfaction. Our study showed that an increase in income satisfaction positively affected health satisfaction over time. Similarly, Soylu (2023) reported that income satisfaction indirectly predicted psychological, global life satisfaction, domain-specific life satisfaction (satisfaction with social, health, financial, leisure, work, and residence status), and disability status in patients with stroke and in general adults. Additionally, it has been reported that QoL in patients with stroke can be related to income satisfaction (Anđelić Bakula et al., 2011). In other words, our study findings and those of previous studies (van Mierlo et al., 2016; Wang et al., 2019) indicated that ADL function and income satisfaction could be critical components in stroke rehabilitation.

In contrast to the previous studies (Kahneman & Deaton, 2010; Ostir et al., 2008), we adjusted for time-varying annual income when we analyzed the effects of ADL function and income satisfaction on health satisfaction. We noted that ADL function and income satisfaction were positively correlated with health satisfaction over time, despite adjusting for the total annual income. It is interesting to note that a previous study indicated that low annual income does not necessarily indicate low income satisfaction (Ostir et al., 2008). For instance, regardless of economic status, positive emotion can be related to a positive aspect of ADL function, motor, and cognitive status (Ostir et al., 2008). Furthermore, a previous study examined the relationship between income and happiness among 1,000 U.S. residents and reported that high income could not necessarily buy happiness (Kahneman & Deaton, 2010). Therefore, it is deemed essential to facilitate vocational rehabilitation for people who are on the verge of reintegrating into their local communities after a stroke. Vocational rehabilitation, through the process of returning to work, not only enhances the quality of life for patients with stroke but also contributes to heightened income satisfaction merely through engagement in economic activities (Vestling et al., 2003). Furthermore, the repetitive physical activities required in the workplace have the potential to ameliorate the physical functionality of patients with stroke (Arwert et al., 2017). Therefore, it is anticipated that this will not only improve income satisfaction but also augment overall health satisfaction.

In the LGM that examined the association between ADL function and health satisfaction (Model 2.a), only educational attainment and employment status significantly influenced the initial mean score of health satisfaction. Byhoff et al. (2017) conducted a systematic review and observed a positive association between educational attainment and health. They reported that higher educational levels are frequently associated with better health outcomes and higher health satisfaction levels. Additionally, Figueredo et al. (2020) reported that a successful return to work after a health-related absence could positively affect individuals’ health satisfaction and overall well-being. Moreover, there are studies that report a relationship between educational attainment and return to work in patients with stroke (Howard et al., 1985; Treger et al., 2007). For example, patients with stroke who have lower educational attainment may face more difficulties in returning to work than those who have higher educational attainment. In summary, educational attainment and return to work in patients with stroke are critical factors that influence the initial health satisfaction level.

This study had several limitations. The study findings may not be generalized to other health conditions or other adult populations, because we focused on the health satisfaction of patients with stroke using the Korean survey database. Moreover, other limitations include potential biases and inaccuracies arising from self-reporting, including memory recall and social desirability bias, as well as the variable quality and availability of secondary data. These factors may introduce bias in reported levels of income and health satisfaction, and incomplete or inconsistent data could affect the robustness and generalizability of the results. Additionally, both health satisfaction and income satisfaction are subjective measures that can be influenced by unique individual characteristics. Therefore, future studies should consider data from different countries and objectively examine other health conditions and additional factors.

Implications for Occupational Therapy Practice

This study has shown that the health satisfaction level among stroke survivors over time is steadily influenced by ADL function and income satisfaction. The findings of this study have the following implications for occupational therapy practice:

▪ Occupational therapy would be beneficial for evaluating people at the later stages of their recovery to determine whether they would benefit from environmental modifications or assistive devices to improve their ADL function.

▪ Additionally, when establishing a treatment plan, it is important to consider improving the physical function necessary for patients with stroke to return to work.

▪ It is important for clinicians to be aware of the characteristics of functional recovery in stroke survivors, including the fact that people with a higher health satisfaction score may exhibit a slower increase in their health satisfaction score over time.

Conclusion

Our LGM approaches demonstrated the longitudinal trajectory of health satisfaction in stroke survivors and showed a positive association with both ADL function and income satisfaction. It is expected that the health satisfaction of stroke survivors could be effectively improved by intervention plans that incorporate ADL training and address income satisfaction issues. For stroke survivors who are transitioning from a hospital setting back to the community, increasing ADL function and income satisfaction through ADL-focused treatment and linkage with vocational rehabilitation could be critical.

Supplemental Material

Supplementary material for Longitudinal Association of Health Satisfaction With Functional Status and Income Satisfaction in Stroke Survivors

Supplementary material, sj-pdf-1-aot-10.5014_ajot.2024.050410.pdf for Longitudinal Association of Health Satisfaction With Functional Status and Income Satisfaction in Stroke Survivors by Sanghun Nam, Timothy A. Reistetter and Ickpyo Hong in The American Journal of Occupational Therapy

Footnotes

Acknowledgments

This work was supported by the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2021S1A3A2A02096338).

References

Abubakar

S. A.

, & Isezuo

S. A.

(2012). Health related quality of life of stroke survivors: Experience of a stroke unit. International Journal of Biomedical Science, 8, 183–187.

Bakula

M. A.

Kovacević

Sarilar

Palijan

T. Z.

, & Kovač

(2011). Quality of life in people with physical disabilities. Collegium Antropologicum, 35(Suppl. 2), 247–253.

Arwert

H. J.

Schults

Meesters

J. J. L.

Wolterbeek

Boiten

, & Vliet Vlieland

(2017). Return to work 2–5 years after stroke: A cross sectional study in a hospital-based population. Journal of Occupational Rehabilitation, 27, 239–246. https://doi.org/10.1007/s10926-016-9651-4

Ashley

K. D.

Lee

L. T.

, & Heaton

(2019). Return to work among stroke survivors. Workplace Health and Safety, 67, 87–94. https://doi.org/10.1177/2165079918812483

Beal

C. C.

(2010). Gender and stroke symptoms: A review of the current literature. Journal of Neuroscience Nursing, 42, 80–87. https://doi.org/10.1097/JNN.0b013e3181ce5c70

Bentler

P. M.

(1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238–246. https://doi.org/10.1037/0033- 2909.107.2.238

Boo

K. C.

, & Chang

J. Y.

(2006). Korean Longitudinal Study of Ageing: Research design for international comparative studies. Survey Research, 7, 97–122.

Browne

M. W.

, & Cudeck

(1993). Alternative ways of assessing model fit. In Bollen

K. A.

and Long

J. S.

(Eds.), Testing structural equation models (pp. 136–162). Sage.

Byhoff

Hamati

M. C.

Power

Burgard

S. A.

, & Chopra

(2017). Increasing educational attainment and mortality reduction: A systematic review and taxonomy. BMC Public Health, 17, 710. https://doi.org/10.1186/s12889-017-4754-1

10.

Cangur

, & Ercan

(2015). Comparison of model fit indices used in structural equation modeling under multivariate normality. Journal of Modern Applied Statistical Methods, 14, 152–167. https://doi.org/10.22237/jmasm/1430453580

11.

Carod-Artal

Egido

J. A.

González

J. L.

, & Varela de Seijas

(2000). Quality of life among stroke survivors evaluated 1 year after stroke: Experience of a stroke unit. Stroke, 31, 2995–3000. https://doi.org/10.1161/01.str.31.12.2995

12.

Cerniauskaite

Quintas

Koutsogeorgou

Meucci

Sattin

Leonardi

, & Raggi

(2012). Quality-of-life and disability in patients with stroke. American Journal of Physical Medicine and Rehabilitation, 91(13, Supp. 1), S39–S47. https://doi.org/10.1097/PHM.0b013e31823d4df7

13.

Clark

A. E.

Frijters

, & Shields

M. A.

(2008). Relative income, happiness, and utility: An explanation for the Easterlin paradox and other puzzles. Journal of Economic Literature, 46, 95–144. https://doi.org/10.1257/jel.46.1.95

14.

D’Ambrosio

, & Frick

J. R.

(2007). Income satisfaction and relative deprivation: An empirical link. Social Indicators Research, 81, 497–519. https://doi.org/10.1007/s11205-006-0020-0

15.

Duncan

P. W.

Goldstein

L. B.

Matchar

Divine

G. W.

, & Feussner

(1992). Measurement of motor recovery after stroke. Outcome assessment and sample size requirements. Stroke, 23, 1084–1089. https://doi.org/10.1161/01.STR.23.8.1084

16.

Diamantopoulos

Siguaw

J. A.

, & Cadogan

J. W.

(2000). Export performance: The impact of cross-country export market orientation. In Workman

J. P.

Jr. , and Perreault

W. D.

Jr. (Eds.), Marketing theory and applications: Proceedings of the American Marketing Association Winter Conference (Vol. 11, pp. 177–178). American Marketing Association.

17.

Diener

Emmons

R. A.

Larsen

R. J.

, & Griffin

(1985). The Satisfaction With Life scale. Journal of Personality Assessment, 49, 71–75. https://doi.org/10.1207/s15327752jpa4901_13

18.

Edwards

J. D.

Kapoor

Linkewich

, & Swartz

R. H.

(2018). Return to work after young stroke: A systematic review. International Journal of Stroke, 13, 243–256. https://doi.org/10.1177/1747493017743059

19.

Felce

, & Perry

(1995). Quality of life: Its definition and measurement. Research in Developmental Disabilities, 16, 51–74. https://doi.org/10.1016/0891-4222(94)00028-8

20.

Ferrer‐i‐Carbonell

, & Frijters

(2004). How important is methodology for the estimates of the determinants of happiness? Economic Journal, 114, 641–659. https://doi.org/10.1111/j.1468-0297.2004.00235.x

21.

Figueredo

J.-M.

García-Ael

Gragnano

, & Topa

(2020). Well-being at work after return to work (RTW): A systematic review. International Journal of Environmental Research and Public Health, 17(20), 7490. https://doi.org/10.3390/ijerph17207490

22.

Hackett

M. L.

, & Anderson

C. S.

; The Australian Cooperative Research on Subarachnoid Hemorrhage Study Group .(2000). Health outcomes 1 year after subarachnoid hemorrhage: An international population-based study. Neurology, 55, 658–662. https://doi.org/10.1212/WNL.55.5.658

23.

H. K.

Matsubayashi

Wada

Kimura

Kita

, & Saijoh

(2002). Factors associated with ADL dependence: A comparative study of residential care home and community‐dwelling elderly in Japan. Geriatrics and Gerontology International, 2, 80–86. https://doi.org/10.1046/j.1444-1586.2002.00026.x

24.

Hofgren

Björkdahl

Esbjörnsson

, & Sunnerhagen

K. S.

(2007). Recovery after stroke: Cognition, ADL function and return to work. Acta Neurologica Scandinavica, 115, 73–80. https://doi.org/10.1111/j.1600-0404.2006.00768.x

25.

Horner

R. D.

Swanson

J. W.

Bosworth

H. B.

, & Matchar

D. B.

; VA Acute Stroke (VAST) Study Team .(2003). Effects of race and poverty on the process and outcome of inpatient rehabilitation services among stroke patients. Stroke, 34, 1027–1031. https://doi.org/10.1161/01.STR.0000060028.60365.5D

26.

Howard

Till

J. S.

Toole

J. F.

Matthews

, & Truscott

B. L.

(1985). Factors influencing return to work following cerebral infarction. JAMA, 253, 226–232. https://doi.org/10.1001/jama.1985. 03350260078030

27.

, & Bentler

P. M.

(1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1–55. https://doi.org/10.1080/10705519909540118

28.

Kahneman

, & Deaton

(2010). High income improves evaluation of life but not emotional well-being. Proceedings of the National Academy of Sciences of the United States of America, 107, 16489–16493. https://doi.org/10.1073/pnas.1011492107

29.

Kanayama

Tsuchiya

Ono

Yoshimura

, & Kobayashi

(2016). The relationship between satisfaction in activities of daily living and perceived quality of life in recovery rehabilitation inpatients. Japanese Journal of Comprehensive Rehabilitation Science, 7, 1–6. https://doi.org/10.11336/jjcrs.7.1

30.

Katzan

I. L.

Thompson

N. R.

Uchino

, & Lapin

(2018). The most affected health domains after ischemic stroke. Neurology, 90, e1364–e1371. https://doi.org/10.1212/WNL.0000000000005327

31.

Khanna

, & Tsevat

(2007). Health-related quality of life—An introduction. American Journal of Managed Care, 13(Suppl. 9), S218–S223.

32.

Korean Statistical Information Service. (2021). Number of stroke patients. https://kosis.kr/statHtml/statHtml.do?orgId=411&tblId=DT_41104_222

33.

Kwakkel

Kollen

, & Twisk

(2006). Impact of time on improvement of outcome after stroke. Stroke, 37, 2348–2353. https://doi.org/10.1161/01.STR.0000238594.91938.1e

34.

Kwon

Park

J. -H.

Kim

W. -S.

Han

Lee

, & Paik

N. -J.

(2018). Health-related quality of life and related factors in stroke survivors: Data from Korea National Health and Nutrition Examination Survey (KNHANES) 2008 to 2014. PLOS One, 13, e0195713. https://doi.org/10.1371/journal.pone.0195713

35.

Yang

Kuang

Zhou

, & Xu

(2023). Mediating effect of post-stroke depression between activities of daily living and health-related quality of life: Meta-analytic structural equation modeling. Quality of Life Research, 32, 331–338. https://doi.org/10.1007/s11136-022-03225-9

36.

Murphy

S. J.

, & Werring

D. J.

(2020). Stroke: Causes and clinical features. Medicine, 48, 561–566. https://doi.org/10.1016/j.mpmed. 2020.06.002

37.

Muthén

L. K.

, & Muthén

(2007). Growth modeling with latent variables using Mplus. Mplus short courses. Muthén & Muthén.

38.

Ostir

G. V.

Berges

I. -M.

Ottenbacher

M. E.

Clow

, & Ottenbacher

K. J.

(2008). Associations between positive emotion and recovery of functional status following stroke. Psychosomatic Medicine, 70, 404–409. https://doi.org/10.1097/PSY.0b013e31816fd7d0

39.

Soylu

(2023). Predictors of domain-specific and global life satisfaction across the first half of life: Which domains of life satisfaction are the most important for global life satisfaction? Current Psychology, 42, 1–12. https://doi.org/10.1007/s12144-022-04161-0

40.

Steultjens

E. M.

Dekker

Bouter

L. M.

van de Nes

J. C.

Cup

E. H.

, & van den Ende

C. H.

(2003). Occupational therapy for stroke patients: A systematic review. Stroke, 34, 676–687. https://doi.org/10.1161/01.STR.0000057576.77308.30

41.

Treger

Shames

Giaquinto

, & Ring

(2007). Return to work in stroke patients. Disability and Rehabilitation, 29, 1397–1403. https://doi.org/10.1080/09638280701314923

42.

Tucker

L. R.

, & Lewis

(1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38, 1–10. https://doi.org/10.1007/BF02291170

43.

van Mierlo

M. L.

van Heugten

C. M.

Post

M. W.

Hajós

T. R.

Kappelle

L. J.

, & Visser-Meily

J. M.

(2016). Quality of life during the first two years post stroke: The Restore4Stroke Cohort Study. Cerebrovascular Diseases, 41, 19–26. https://doi.org/10.1159/000441197

44.

Vestling

Tufvesson

, & Iwarsson

(2003). Indicators for return to work after stroke and the importance of work for subjective well-being and life satisfaction. Journal of Rehabilitation Medicine, 35, 127–131. https://doi.org/10.1080/16501970310010475

45.

Waddell

K. J.

Birkenmeier

R. L.

Bland

M. D.

, & Lang

C. E.

(2016). An exploratory analysis of the self-reported goals of individuals with chronic upper-extremity paresis following stroke. Disability and Rehabilitation, 38, 853–857. https://doi.org/10.3109/09638288.2015.1062926

46.

Wang

Shang

Yang

Wang

Chang

et al. Jiang, X. (2019). Associations of psychological distress with positive psychological variables and activities of daily living among stroke patients: A cross-sectional study. BMC Psychiatry, 19, 381. https://doi.org/10.1186/s12888-019-2368-0

47.

Wang

Y. C.

Kapellusch

, & Garg

(2014). Important factors influencing the return to work after stroke. Work, 47, 553–559. https://doi.org/10.3233/WOR-131627

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.31 MB