Skip to main content
NCBI home page
Physical Therapy logoLink to Physical Therapy
. 2021 Oct 4;101(12):pzab224. doi: 10.1093/ptj/pzab224

Depressive Symptoms Moderate the Relationship Among Physical Capacity, Balance Self-Efficacy, and Participation in People After Stroke

Margaret A French 1, Allison Miller 1, Ryan T Pohlig 2, Darcy S Reisman 1,3,✉
PMCID: PMC8697846  PMID: 34636909

Abstract

Objective

It was previously found that balance self-efficacy mediated the relationship between physical capacity and participation after stroke. The effect of other factors that influence participation, such as depression, on this relationship has not been explored. This study examined the effect of symptoms of depression on the mediated relationship between physical capacity and participation by balance self-efficacy in individuals after stroke.

Methods

In this cross-sectional study, 282 persons with chronic stroke (>6 months) were classified as having either low or high Geriatric Depression Scale scores. This study used a multiple group structural equation model to test moderated mediation by comparing a constrained model (indicating no effect of depression on the mediation) and an unconstrained model (indicating an effect of depression on the mediation). The models were compared using a chi-squared difference test.

Results

The chi-squared difference test suggested that the unconstrained model was a better fit, indicating that depressive symptoms moderated the mediated relationship between physical capacity and participation (χ2(3, N = 282) = 9.0). In the Low Depression group, a significant indirect effect indicated that balance self-efficacy did mediate the relationship between physical capacity and participation. There was no significant indirect effect in the High Depression group.

Conclusion

The results suggest the relationship between physical capacity and participation appears to be mediated by balance self-efficacy in individuals after stroke with low reports of depressive symptoms, but in those with high reports of depressive symptoms, physical capacity and balance self-efficacy are unrelated to participation.

Impact

Targeting balance self-efficacy to improve post-stroke participation may be beneficial only for individuals with low reports of depression. In individuals after stroke with high reports of depression, treatment should include and emphasize the treatment of those depressive symptoms. Additional work further examining these complex relationships is warranted.

Keywords: Community Participation, Depression, Mediation, Moderation, Self-Efficacy, Stroke, Structural Equation Model

Introduction

Approximately 795,000 people in the United States experience a stroke each year, making it one of the leading causes of long-term disability in the United States.1 Community participation after stroke is often reduced.2–5 This reduction in community participation is often accompanied by an increased risk of secondary complications, including diabetes, heart disease, and recurrent stroke and a reduced quality of life.6,7 Thus, improving community participation is critical to improving overall health and quality of life in individuals after stroke. Many factors across a variety of domains, including physical8,9 and psychosocial,10,11 impact community participation. Therefore, to obtain the most accurate impression of a patient’s community participation, rehabilitation professionals should consider all of these factors. However, the relationships between these various factors are complex and not well understood. As a result, it is difficult to design comprehensive rehabilitation interventions that target the relevant factors to improve community participation after stroke. A more complete understanding of the relationships between these factors will allow rehabilitation professionals to integrate the results of assessments of these factors into treatment planning, ultimately allowing for more comprehensive, targeted rehabilitation interventions.

In recent years, the relationships between participation and (1) physical capacity and (2) self-efficacy have drawn significant attention in persons after stroke. In this context, physical capacity is defined as one’s ability to physically complete a task. Balance assessments, such as the Berg Balance Scale (BBS), and walking assessments, such as the Six Minute Walk Test (6MWT), are commonly used measures of physical capacity. Self-efficacy, on the other hand, refers to one’s belief in their ability to perform a task.10,11 In rehabilitation, self-efficacy specifically related to balance has drawn significant attention. One common measure of balance self-efficacy is the Activities-specific Balance Confidence Scale (ABC). Past work is in agreement that greater physical capacity8–10,12–14 and higher balance self-efficacy10,11,15–21 are important predictors of greater community participation.

Impaired physical capacity and poor balance self-efficacy are common after stroke. Given their prevalence, several studies have examined how these 2 factors together affect participation.10,11 Of particular note, French et al found that balance self-efficacy mediated the relationship between physical capacity and participation in individuals after stroke.11 In a mediated relationship, a predictor’s association with the outcome can be attributed to both the direct and indirect effect. The direct effect is the relationship between the predictor and the outcome, whereas the indirect effect is the relationship between the predictor and the outcome through the mediator. The indirect association is quantified by combining an estimate of the predictor’s influence on a third variable (the mediator), with an estimate of that mediator’s influence on the outcome. With this in mind, the significant indirect effect found in French et al suggested that part of the relationship between physical capacity and participation is because physical capacity is related to balance self-efficacy in individuals after stroke. One clinically relevant way that this mediated relationship could manifest is in individuals who have sufficient physical capacity but for some reason exhibit lower community participation than expected. This could be because their balance self-efficacy is low.11 This evidence points to complex relationships among physical capacity, balance self-efficacy, and participation after stroke; however, other factors may further complicate these relationships.

One such factor is symptoms of depression. Depressive symptoms are common after stroke24–26 and are associated with lower physical capacity,27,28 lower balance self-efficacy,29–32 and lower participation.33–36 Although previous studies have examined the bivariate associations between depressive symptoms and physical capacity, balance self-efficacy, and participation, the impact of depressive symptoms on the relationships among these factors has not been examined. Thus, the purpose of this work is to understand the impact of depressive symptoms on the relationships between physical capacity, balance self-efficacy, and participation. We hypothesized that the magnitude of depressive symptoms would affect this mediated relationship by acting as a moderator. A moderator is a variable that affects the strength of the relationship between 2 variables.22,23 In other words, we hypothesized that the relationships between variables within the mediation model may be conditional on an individual’s depressive symptoms score due to its role as a moderator.

Methods

Participants

Data used in this cross-sectional, secondary analysis study were obtained from a preexisting database of community-dwelling adults with stroke at the University of Delaware. This database was part of a larger registry of individuals with stroke who expressed interest in being involved in research. To be included in this database, a trained and licensed physical therapist conducts a standardized 1-time assessment that includes the collection of demographic information and information about stroke characteristics as well as a battery of clinical tests. The purpose of this database is to facilitate screening and recruitment of individuals with stroke for research studies; however, it also generates a large database of clinical metrics that can be used for cross-sectional analyses as done in the current work. To be included in the database, individuals met the following inclusion criteria: (1) history of a stroke more than 6 months prior, (2) at least 21 years of age, (3) able to walk at least 10 m without physical assistance (orthotic and assistive devices allowed), (4) willing and able to attend all sessions, and (5) provide informed consent. The following were exclusion criteria for participation in the preexisting database: (1) unable to stand or walk; (2) resting heart rate outside the range of 40 to 100 bpm; (3) resting blood pressure outside the range of 90/60 to 185/100 mmHg; (4) unable to provide consent as indicated by an inability to answer at least 1 orientation question correctly (item 1b on the National Institutes of Health Stroke Scale) and inability to follow at least one, 2-step command (item 1c of the National Institutes of Health Stroke Scale); (5) cardiac event or cardiac surgery in the past 3 months; (6) chest pain or shortness of breath without exertion; and (7) unexplained dizziness in the last 6 months. Additionally, the participants included in the current work had to have completed all 6 primary measures of interest (discussed in Primary Measures section) and the Geriatric Depression Scale-Short Form (GDS); thus, there were no missing data. All participants signed informed consent approved by the Human Subjects Review Board at the University of Delaware prior to participation. Results from the clinical measures and demographic information were recorded in REDCap (Research Electronic Data Capture), a secure electronic database.37

Theoretical Model

Model Framework

To test our hypothesis regarding the moderating role of depressive symptoms on the relationship between physical capacity, balance self-efficacy, and participation, we developed and subsequently tested a theoretical model (Fig. 1). This theoretical model was based on the work by French et al. The same model was tested in individuals with low reports of depressive symptoms and high reports of depressive symptoms as measured by the GDS (see details in Primary Measures section); however, for simplicity, the model is shown only once in Figure 1. In this theoretical model, “physical capacity” is a latent construct with 4 indicator (observed) variables: 6MWT, Lower Extremity Fugl-Meyer (LEFM), BBS, and Functional Gait Assessment (FGA). Details regarding this latent construct are provided in the Primary Measures section. Balance self-efficacy and participation were measured by the observed variables of ABC and Stroke Impact Scale-Participation subscale (SISP), respectively (Fig. 1). In the theoretical model, which is a mediation model, there are a series of regression pathways, each of which has its own regression coefficient, called a path coefficient, which describes the strength of the relationship between the 2 variables (or construct) in that pathway. There are 4 pathways within our theoretical model: (1) the relationship between physical capacity and ABC, (2) the relationship between ABC and SISP, (3) the relationship between physical capacity and SISP that does not go through the mediator (ie, the direct effect), and (4) the relationship between physical capacity and SISP through ABC (ie, the indirect effect). This indirect effect consists of the relationships between physical capacity and ABC as well as ABC and SISP and is calculated as the product of the coefficients for these 2 pathways.22,23 Conceptually, a significant indirect effect would imply that part of the relationship between physical capacity and SISP occurs because physical capacity is affecting ABC, which in turn affects SISP. Here, we tested if depressive symptoms were a moderator in this mediated relationship by testing this theoretical model in those with high and low symptoms of depression.

Figure 1.

Figure 1

Open in a new tab

Theoretical model. In this work, we hypothesized that the mediated relationship between physical capacity and participation by balance self-efficacy would be moderated by depressive symptoms. In this model, there are 3 pathways of interest (ie, X, Y, and Z) as well as the indirect effect of physical capacity on participation through balance self-efficacy. Here, physical capacity is a latent construct. In this figure, we show the proposed model for a single group; however, we tested this same model in both groups (ie, individuals after stroke with low depressive symptoms and individuals with high depressive symptoms). 6MWT = Six Minute Walk Test; ABC = Activities-specific Balance Scale; BBS = Berg Balance Scale; FGA = Functional Gait Assessment; LEFM = Lower Extremity Fugl Meyer; SISP = Stroke Impact Scale Participation.

Primary Measures Within the Model

The primary measures for this study included the (1) 6MWT, (2) LEFM, (3) BBS, (4) FGA, (5) ABC, and (6) SISP. These measures were selected based on our previous work examining the mediated relationship between physical capacity and balance self-efficacy11 as well as their established validity in the stroke population (see below for details).

The 6MWT, LEFM, BBS and FGA represent various aspects of an individual’s physical capacity. The 6MWT is a submaximal exercise test38 that measures one’s walking endurance38,39 in which individuals walk as far as possible for 6 minutes. The LEFM is a measure of lower extremity impairment specific to stroke and includes an assessment of the participant’s reflexes and synergistic movement patterns.40 The BBS is a 14-item assessment of static balance and fall risk.41 The FGA consists of 10 items used to objectively measure dynamic balance and postural stability during various tasks.42 Previous work has shown that the 6MWT,43 LEFM,44 BBS,41 and FGA42 are valid and reliable measures in individuals with stroke.

The ABC is a 16-item questionnaire that measures balance self-efficacy45–47 or an individual’s belief about their ability to perform tasks that could challenge one’s balance.18,48 Individuals rate their confidence performing various tasks on a 0 (“no confidence”) to 100 (“complete confidence”) scale. The ratings for each item are then averaged to produce a final score that reflects the individual’s overall balance self-efficacy, with lower scores indicating lower balance self-efficacy. This measure has been shown to be valid and reliable in individuals post stroke.47

The SISP subscale is part of the Stroke Impact Scale, which is a questionnaire that assesses health-related quality of life.49 The SISP consists of 8 questions that specifically measure one’s ability to engage in everyday life situations, such as work and recreation,50 thus, reflecting post-stroke participation. Participants rate how often in the past week their stroke has affected their participation in each specific activity on a 1 (“all the time”) to 5 (“none of the time”) scale. Scores are then reported on a 0 to 100 scale, where higher scores reflect better participation. The SISP has been shown to be valid and reliable in individuals with stroke.51–53

Depressive Symptoms

Participants were classified as having either low or high depressive symptoms (LOW and HIGH, respectively) based on scores on the GDS. The GDS is a 15-item self-report depression screening questionnaire in which participants are asked to respond “yes” or “no” to various questions about their feelings over the past week. Previous work has shown that the GDS is a valid and reliable scale in persons with stroke.54,55 We defined those with GDS scores of <6 as having low depressive symptoms (LOW) and those with scores ≥6 as having high depressive symptoms (HIGH). This value has been shown to be the optimal cut-off for detecting clinical depression in individuals after stroke.56

Statistical Analysis

SPSS v26 (IBM, Chicago, IL, USA) and MPlus 8.3 (Muthén & Muthén, Los Angeles, CA, USA)57 were used for statistical analyses. Prior to conducting the analyses detailed below, the data were screened for outliers, and the assumptions of normality and homoscedasticity were checked. Based on the work of French et al that found a significant indirect effect with a sample size of 59 individuals with stroke using similar measures,11 we aimed for a sample size of at least 59 participants to detect a significant indirect effect.

Multiple Group Structural Equation Model

To determine if the presence of depressive symptoms impact the previously found mediated relationship between physical capacity and participation by balance self-efficacy, we used a multiple group structural equation model (SEM). SEM is a series of statistical modeling techniques that allow researchers to test theoretical models that examine the relationships between latent and observed variables; thus, SEM typically includes a combination of factor analyses and path analyses.58 The ability of the theoretical model to reflect the observed data is evaluated using a series of model fit statistics. In multiple group SEM, the basic principles are the same; however, the purpose is to compare the fit of a theoretical model between groups.59 Thus, when testing the theoretical model, the researcher manipulates whether specific pathways are the held constant (ie, constrained) between groups or are allowed to vary (ie, unconstrained) between groups. Then, models in which different paths are constrained and freed are compared using a chi-squared difference test.60 A chi-squared difference test compares different models of varying complexities and allows researchers to determine which model fits best while also being as simple as possible.60

In this work, we compared 2 models, the null model and the moderation model, to determine if the mediated relationship was different between those with high and low symptoms of depression. In the null model, we evaluate a model in which pathways are constrained (ie, the same) between the 2 groups. In the moderation model, we allowed pathways to vary (ie, be unconstrained) between high and low symptoms of depression. In this model, coefficients are estimated separately for each group. These 2 models were then compared using a chi-squared difference test. A significant test would indicate that the moderation model fits the data better, concluding that the pathways differ between those with high and low symptoms of depression. This indicates that depression does moderate the relationship between physical capacity, balance self-efficacy, and participation. In addition to the results of the chi-squared difference test, standardized path coefficients and model fit statistics, including root mean square error of approximation (RMSEA), comparative fit index (CFI), Tucker-Lewis index, and the standardized root mean square residual (SRMR), are reported.

Confirmatory Factor Analysis

Because physical capacity was a latent variable, we also needed to ensure that this construct was stable and not measured differently between groups. To do this, a confirmatory factor analysis was performed prior to conducting our multiple group SEM.61 The 4 indicator variables were selected because they are commonly used in clinical practice and reflect different aspects of physical capacity, specifically balance (BBS, FGA), mobility (6MWT), and physical impairment (LEFM). Within the entire sample, all 4 indicators significantly loaded onto 1 latent variable, which we refer to as “physical capacity.” To ensure that the selected indicators had the same relationship with the latent construct of “physical capacity” in both groups (ie, LOW and HIGH), we used a chi-squared difference test to compare a constrained model, in which the factor loadings of the 4 indicators were the same in both the group with low depressive symptoms and the group with high depressive symptoms, to an unconstrained model that calculated separate factor loadings for each group. A non-significant chi-squared difference test would indicate that the constrained and unconstrained models were not significantly different from each other, evidence that the latent construct of physical capacity was not measured differently between the groups (ie, LOW and HIGH).

Potential Covariates

Because this analysis used cross-sectional data, we also accounted for specific covariates in this model. To determine which covariates to include in our model, we compared demographic information and stroke characteristics between the LOW and HIGH groups. An independent t test was used to test for differences in age and time since stroke between groups. A chi-squared test was used to test for differences in gender, type of stroke, and side of hemiparesis. Variables that were significantly different between groups were included as covariates in the SEM model.

Role of the Funding Source

The funders, National Institutes of Health and Foundation for Physical Therapy Research, played no role in the design, conduct, or reporting of this study.

Results

Participants

Two-hundred and eighty-two participants (age = 62.7 [12.8] years; 164 M; 118 F) were included in this analysis (Tab. 1). Of those 282 participants, 218 had low levels of depressive symptoms (GDS score = 2.2 [1.5]) and 64 had high levels of depressive symptoms (GDS score = 8.5 [2.3]; Tab. 2). In the LOW group, the average age was 64.1 (12.1) years, while in the HIGH group, the average age was 57.9 (14.1) years. This was the only potential covariate that was significantly different between groups (P = .001; Tab. 1); thus, it was the only covariate included in the SEM analysis. Complete demographic information is presented in Table 1. Additionally, there were differences between groups on all clinical characteristics as shown in Table 2. It is important to note that we were specifically interested in potential differences in the relationships between these variables, not differences in group means for specific measures. In addition, where the means of variables differ between groups, the relationships between those variables can either be the same or different between groups, thus warranting the analyses in this work.

Table 1.

Demographic Informationa

All (n = 282) Low Depressive Symptoms
(LOW, n = 218) 		High Depressive Symptoms (HIGH, n = 64) P
Age, y, mean (SD) 62.7 (12.8) 64.1 (12.1) 57.9 (14.1) .001
Sex 164 M; 118 F 124 M; 94 F 40 M; 24 F .47
Time since stroke, mo, mean (SD) 32.6 (47.8) 34.5 (48.1) 26.3 (46.7) .23
Type of stroke Ischemic: 203
hemorrhagic: 41
unknown: 38		 Ischemic: 156
hemorrhagic: 33
unknown: 29		 Ischemic: 47
hemorrhagic: 8
unknown: 9		 .95
Paretic side Left: 130
right: 131
bilateral: 11
unknown: 10		 Left: 96
right: 104
bilateral: 9
unknown: 9		 Left: 34
right: 27
bilateral: 2
unknown: 1		 .34
Open in a new tab

a P values reflect comparisons between the lowdep and highdep groups. Continuous variables were compared using an independent t test, while categorical variables were compared with chi-squared tests.

Table 2.

Clinical Characteristicsa

Characteristic All
(n = 282) 		Low Depressive Symptoms (LOW)
(n = 218) 		High Depressive Symptoms (HIGH)
(n = 64) 		P
Lower extremity Fugl Meyer, mean (SD) 24.4 (7.2) 25.2 (7.1) 21.6 (6.8) <.001
Berg Balance Scale, mean (SD) 44.5 (10.7) 45.3 (10.6) 41.8 (10.7) .01
Six Minute Walk Test, m, mean (SD) 312.1 (168.4) 330.2 (169.8) 250.2 (149.1) .001
Functional gait assessment, mean (SD) 13.6 (6.8) 14.3 (7.0) 11.1 (5.4) .001
Activities-specific balance scale, % mean (SD) 72.6 (21.2) 75.6 (19.9) 62.4 (22.4) <.001
Stroke Impact Scale- participation, mean (SD) 68.7 (23.9) 73.9 (22.0) 50.9 (21.5) <.001
Geriatric Depression Scale, mean (SD) 3.6 (3.2) 2.2 (1.5) 8.5 (2.3) <.001
Open in a new tab

a P values reflect comparisons between the lowdep and highdep groups. Continuous variables were compared using an independent t test, and categorical variables were compared with chi-squared tests.

Confirmatory Factor Analysis

A confirmatory factor analysis was performed to define the latent construct of “physical capacity.” All 4 indicators (ie, BBS, FGA, LEFM, 6MWT) loaded strongly onto the latent construct (data available on request). When comparing the latent construct “physical capacity” between groups, the chi square difference test was not significant (χ2(3, N = 282) = 5.9, P = .12), suggesting that the 4 indicators’ factor loadings did not differ significantly between the groups with low and high depressive symptoms. This finding allowed us to be confident the latent construct “physical capacity” was not measured differently between groups while testing our full model (described below).

Multiple Group SEM

Both the model in which the pathways were constrained and unconstrained fit reasonably well (χ2(34, N = 282) = 106.4, P < .001; RMSEA = 0.12; CFI = 0.94; TFI = 0.93; SRMR = 0.09 and χ2(31, N = 282) = 97.3, P < .001; RMSEA = 0.12; CFI = 0.94; TFI = 0.93; SRMR = 0.08, respectively). A chi-squared difference test comparing these models showed that the model in which the pathways were unconstrained fit significantly better (χ2(3, N = 282) = 9.0, P = .03), suggesting that depressive symptoms moderate the meditated pathway.

The standardized path coefficients, which are interpreted similarly to regression coefficients, and standard errors for the moderated model are presented in Figures 2a and b. The pathway between physical capacity and balance self-efficacy was significant in both groups (LOW: P < .001; HIGH: P < .01). As indicated by the positive and significant path coefficients, greater physical capacity was associated with greater balance self-efficacy in both individuals with low and high depressive symptoms. The pathway between balance self-efficacy and participation was significant in the LOW group (P < .001) but not in the HIGH group (P = .23), indicating that greater balance self-efficacy was associated with greater participation in persons with low depressive symptoms, but not in persons with higher depressive symptoms. Similarly, the direct pathway from physical capacity to SISP was significant in the LOW group (P = .005) but not the HIGH group (P = .70), suggesting that greater physical capacity was associated with greater participation in stroke survivors with low depressive symptoms but not in individuals with higher depressive symptoms. Lastly, the indirect effect was significant in the LOW group (P < .001; Fig. 2a) but not in the HIGH group (P = .24; Fig. 2b). This suggests that part of the effect of physical capacity on participation occurs because physical capacity influences balance self-efficacy, which then influences participation in stroke survivors with lower depressive symptoms. However, in stroke survivors with higher depressive symptoms, this is not the case. Taken together, these findings demonstrate that in those with low depressive symptoms, there is a significant direct and indirect relationship between physical capacity and SISP; however, there is no relationship between physical capacity and SISP, either direct or indirect, in those with high depressive symptoms.

Figure 2.

Figure 2

Open in a new tab

Final model. The mediated pathway of physical capacity on participation through balance self-efficacy is shown for both LOW (a) and HIGH (b) groups. The indirect effect was significant for those in the LOW group but not in for those in the HIGH group. All paths are adjusted for age and standardized path coefficients and their standard errors are presented. For simplicity, the loading factors for the indicators of the latent construct of physical capacity are not shown. 6MWT = Six Minute Walk Test; ABC = Activities-specific Balance Scale; BBS = Berg Balance Scale; FGA = Functional Gait Assessment; LEFM = Lower Extremity Fugl Meyer; SISP = Stroke Impact Scale Participation. *P < .05, **P < .01, and ***P < .001.

Discussion

The purpose of this study was to understand how depression influences the relationship between physical capacity, balance self-efficacy, and participation after stroke. We found that balance self-efficacy mediated the relationship between physical capacity and participation in persons after stroke who had low reports of depressive symptoms but not those with high reports of depressive symptoms, demonstrating the role of depression as a moderator in our mediation model. Additionally, we found that the direct relationship between physical capacity and participation and the relationship between balance self-efficacy and participation was also moderated by the presence of depressive symptoms. This is one of the first studies to account for the moderating effect of depression when examining the relationships between physical capacity, balance self-efficacy, and participation and adds to our understanding of the complex relationships that impact community participation after stroke. The implications of this work are discussed below.

First, we found that balance self-efficacy mediated the relationship between physical capacity and participation in individuals with lower reports of depressive symptoms, thus replicating the findings of French et al in a much larger sample.11 Conversely, we did not find that balance self-efficacy mediated the relationship between physical capacity and participation in individuals with high reports of depressive symptoms, as indicated by the non-significant indirect effect. A large body of literature has found that balance self-efficacy is related to community participation after stroke, suggesting that interventions that improve balance self-efficacy may also lead to improvements in participation in individuals with stroke.10,11,15–21 However, our results suggest that targeting balance self-efficacy to improve participation may be an effective approach only for individuals after stroke with low levels of depressive symptoms and not for individuals after stroke with high reports of depressive symptoms. For example, a clinician working to improve participation in a patient with stroke with low depressive symptoms can target the individual’s balance self-efficacy to improve their participation. However, our results suggest that this approach will be ineffective if the patient reports higher levels of depressive symptoms. Thus, based on our results, understanding a patient’s depressive symptoms has important implications for determining the most effective treatment approach in the context of improving participation.

Two pathways in our model contribute to the indirect effect: the relationship between physical capacity and balance self-efficacy and the relationship between balance self-efficacy and participation (Fig. 1; path X and Y). We found that the relationship between physical capacity and balance self-efficacy was significant in both groups such that individuals with greater physical capacity also had higher balance self-efficacy regardless of the magnitude of depressive symptoms. This is interesting given that depression has been associated with lower physical capacity27,28 and lower balance self-efficacy29–32; however, our results show that the relationship between these 2 constructs is similar regardless of the presence of depressive symptoms. Conversely, the relationship between balance self-efficacy and participation was significant only in the group of individuals after stroke who had low levels of depressive symptoms, which explains the difference in the mediated relationship. Past work has shown that individuals after stroke with depression have lower levels of participation33–36 and balance self-efficacy30; however, here we demonstrate that the relationship between these 2 constructs is different based on the presence of depressive symptoms. Specifically, this work suggests that higher levels of physical capacity are associated with higher levels of participation in individuals who have low reports of depressive symptoms after stroke, whereas this was not the case for individuals with higher levels of depressive symptoms.

Additionally, we found that the direct relationship of physical capacity and participation differed between groups and was significant in individuals with low levels of depression symptoms but not in individuals with higher levels of depression symptoms. It has previously been shown that presence of depression is related to lower levels of physical capacity27,28 and participation33–36; however, the current work also suggests that depression affects the relationship between physical capacity and participation. In other words, these results suggest that individuals after stroke with higher physical capacity also have higher levels of participation when depressive symptoms are low. For individuals who have high reports of depressive symptoms after stroke, higher levels of physical capacity are not significantly associated with higher levels of participation. Taken together, these results suggest that depression, physical capacity, and balance self-efficacy have complex relationships that impact participation. Therefore, screening for depressive symptoms and using a comprehensive multi-disciplinary approach that addresses all of these factors is needed to impact participation.

In addition to differences in the relationship between these variables, we also observed differences in means between groups in our variables of interest (eg, measure of physical capacity, balance self-efficacy, and participation; Tab. 2). These findings support past work that has shown that individuals with depression have lower physical capacity,27,28 balance self-efficacy,30 and participation.33–36 This is a potential confounder in the current work. Although it is possible for groups with different means to have relationships between variables that are either different or not different (ie, slopes that are significantly different or that are not significantly different), the current work cannot determine what role the group differences played. Because we observed no difference in the relationship between physical capacity and balance self-efficacy between groups despite different group means, we hypothesize that the observed differences in the relationship between balance self-efficacy and participation and physical capacity and participation are not merely a result of the differences in group means. This is supported by the fact that despite differences in group means, there was overlap in the distributions of these variables between groups (Tab. 2), suggesting that differences in group means were likely not the sole cause of differences in the relationships between groups. Although this study was not designed to make casual claims, the current work adds significantly to our understanding of the complex relationships between depression, physical capacity, balance self-efficacy, and participation. However, future work designed to draw causal conclusions about these relationships and the impact of mean values would be useful in determining how to use this information to optimize participation after stroke.

Limitations

Although this work provides valuable insight into the complex relationship of physical capacity, balance self-efficacy, depression, and participation, it is not without limitations. One important limitation is that we determined our groups based on scores on the GDS, which may be subject to recall bias because it asks participants to reflect on their feelings over the past week. In addition, although this measure is a commonly used screening tool that has been validated for use in individuals post stroke to guide referrals for further assessment of depression symptoms,55,56 it does not provide a definitive clinical diagnosis of depression. As a result, our conclusions should only be applied to the magnitude of self-reported depressive symptoms and not to those with and without a clinical diagnosis of depression. Although this is a limitation, it may also aid in the application of this knowledge to clinical practice because the GDS is a clinically feasible tool that provides understanding about potential depressive symptoms, whereas obtaining a clinical diagnosis of depression can take time.

Another important factor to note is that there were only 64 participants (22.7%) who were included in the HIGH group based on their GDS score. Past work suggests that approximately one-third of individuals after stroke experience post-stroke depression symptoms.62,63 In individuals between 1 and 5 years post stroke and those more than 5 years post stroke, the reported prevalence is slightly lower at 25% and 23%, respectively.24 As a result, the percentage of individuals in our sample who reported higher depressive symptoms is similar to previously reported values; however, there were even fewer participants (5.7%) who scored greater than 10 on the GDS, which is the score believed to correspond to a clinical diagnosis of depression.55 Thus, it is likely that clinical depression is underrepresented in our sample. Although this may be due to bias in who selects to be involved in research, it would be valuable to replicate these findings in a sample with a higher prevalence of depressive symptoms and more severe reports of depressive symptoms. Despite this potential limitation, it is also important to note that we found differences in the mediated relationship even when a clinical diagnosis of depression was not incorporated into the model. This suggests that depressive symptoms, even in the absence of a clinical diagnosis of depression, should be addressed when aiming to improve participation in individuals after stroke.

In addition, the results of this study may not generalize to individuals outside of the eligibility criteria from which the data were obtained. Specifically, our results may not generalize to persons <6 months post stroke who are unable to walk at least 10 m without assistance. Furthermore, as discussed previously, the relationships among the many factors that influence participation after stroke are likely complex. Thus, there may be additional variables, such as comorbidities and measures that reflect social determinants of health, that serve as moderators and/or mediators but were not included in this study. Additionally, the data for this work were cross-sectional, and as such we cannot make causal claims. To this end, this work represents a first step towards mapping the complex relationships between these many variables that influence participation after stroke; however, additional work that examines more complex models with greater sample sizes that can make causal claims is needed.

In conclusion, we found that balance self-efficacy mediates the relationship between physical capacity and participation in individuals after stroke with lower levels of depressive symptoms but not in those with higher depressive symptoms. This difference is due to symptoms of depression moderating the relationship between balance self-efficacy and participation. As a result, targeting balance self-efficacy as an avenue to improve participation after stroke appears to be primarily appropriate for individuals with low levels of depressive symptoms. In individuals with high levels of depressive symptoms, rehabilitation professionals should prioritize addressing the patient’s depressive symptoms to improve participation. These results point to complex relationships between depression, physical capacity, balance self-efficacy, and participation in individuals after stroke. Continued work to disentangle these complex relationships is needed to determine which interventions are most effective in the correct subset of patients to enable greater improvements in participation after stroke.

Author Contributions

Concept/idea/research design: M.A. French, A. Miller

Writing: M.A. French, A. Miller, R.T. Pohlig

Data collection: M.A. French

Data analysis: M.A. French, A. Miller, R.T. Pohlig

Project management: M.A. French

Fund procurement: D.S. Reisman

Providing participants: D.S. Reisman

Providing facilities/equipment: D.S. Reisman

Consultation: M.A. French, A. Miller

Acknowledgments

The authors would like to thank the individuals with stroke who participated in this work as well as Tamara Wright, who assisted in data collection.

Funding

This work was funded by the National Institutes of Health (ref. no. 5R01HD086362-05 and F31NS111806) and the Foundation for Physical Therapy Research’s Promotion of Doctoral Studies I.

Ethics Approval

This study was approved by the University of Delaware’s Institutional Review Board (1642798-1) and conforms to the standards set by the Declaration of Helsinki.

Disclosures and Presentations

The authors completed the ICMJE Form for Disclosure of Potential Conflicts of Interest and reported no conflicts of interest.

The results of this work were presented as a platform presentation at the American Physical Therapy Association Combined Sections Meeting in February 2021 (held virtually).

References

  • 1. Virani SS, Alonso A, Benjamin EJ, et al. Heart disease and stroke statistics-2020 update: a report from the American Heart Association. Circulation. 2020;141:e139–e596. [DOI] [PubMed] [Google Scholar]
  • 2. Mayo NE, Wood-Dauphinee S, Ahmed S, et al. Disablement following stroke. Disabil Rehabil. 1999;21:258–268. [DOI] [PubMed] [Google Scholar]
  • 3. Lord SE, McPherson K, McNaughton HK, Rochester L, Weatherall M. Community ambulation after stroke: how important and obtainable is it and what measures appear predictive? Arch Phys Med Rehabil. 2004;85:234–239. [DOI] [PubMed] [Google Scholar]
  • 4. Schmid A, Duncan PW, Studenski S, et al. Improvements in speed-based gait classifications are meaningful. Stroke. 2007;38:2096–2100. [DOI] [PubMed] [Google Scholar]
  • 5. Michael KM, Allen JK, Macko RF. Reduced ambulatory activity after stroke: the role of balance, gait, and cardiovascular fitness. Arch Phys Med Rehabil. 2005;86:1552–1556. [DOI] [PubMed] [Google Scholar]
  • 6. Billinger SA, Coughenour E, Mac Kay-Lyonsand MJ, Frederick M. Reduced cardiorespiratory fitness after stroke: biological consequences and exercise-induced adaptations. Stroke Research and Treatment. 2012:2012:959120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Mozaffarian D, Benjamin EJ, Go AS, et al. Heart disease and stroke statistics--2015 update: a report from the American Heart Association. Circulation. 2015;131:e29–e322. [DOI] [PubMed] [Google Scholar]
  • 8. Fulk GD, He Y, Boyne P, Dunning K. Predicting home and community walking activity poststroke. Stroke. 2017;48:406–411. [DOI] [PubMed] [Google Scholar]
  • 9. Fulk GD, Reynolds C, Mondal S, Deutsch JE. Predicting home and community walking activity in people with stroke. Arch Phys Med Rehabil. 2010;91:1582–1586. [DOI] [PubMed] [Google Scholar]
  • 10. Danks KA, Pohlig RT, Roos M, Wright TR, Reisman DS. Relationship between walking capacity, biopsychosocial factors, self-efficacy, and walking activity in persons Poststroke. J Neurol Phys Ther. 2016;40:232–238. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. French MA, Moore MF, Pohlig R, Reisman D. Self-efficacy mediates the relationship between balance/walking performance, activity, and participation after stroke. Top Stroke Rehabil. 2016;23:77–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Awad L, Reisman D, Binder-Macleod S. Distance-induced changes in walking speed after stroke: relationship to community walking activity. J Neurol Phys Ther. 2019;43:220–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Mulder M, Nijland RH, Port IG, Wegen EE, Kwakkel G. Prospectively classifying community walkers after stroke: who are they? Arch Phys Med Rehabil. 2019;100:2113–2118. [DOI] [PubMed] [Google Scholar]
  • 14. Perry J, Garrett M, Gronley JK, Mulroy SJ. Classification of walking handicap in the stroke population. Stroke. 1995;26:982–989. [DOI] [PubMed] [Google Scholar]
  • 15. Durcan S, Flavin E, Horgan F. Factors associated with community ambulation in chronic stroke. Disabil Rehabil. 2016;38:245–249. [DOI] [PubMed] [Google Scholar]
  • 16. Hellstrom K, Lindmark B, Wahlberg B, Fugl-Meyer AR. Self-efficacy in relation to impairments and activities of daily living disability in elderly patients with stroke: a prospective investigation. J Rehabil Med. 2003;35:202–207. [DOI] [PubMed] [Google Scholar]
  • 17. Robinson CA, Shumway-Cook A, Matsuda PN, Ciol MA. Understanding physical factors associated with participation in community ambulation following stroke. Disabil Rehabil. 2011;33:1033–1042. [DOI] [PubMed] [Google Scholar]
  • 18. Salbach NM, Mayo NE, Robichaud-Ekstrand S, Hanley JA, Richards CL, Wood-Dauphinee S. Balance self-efficacy and its relevance to physical function and perceived health status after stroke. Arch Phys Med Rehabil. 2006;87:364–370. [DOI] [PubMed] [Google Scholar]
  • 19. Schmid AA, Van Puymbroeck M, Altenburger PA, et al. Balance and balance self-efficacy are associated with activity and participation after stroke: a cross-sectional study in people with chronic stroke. Arch Phys Med Rehabil. 2012;93:1101–1107. [DOI] [PubMed] [Google Scholar]
  • 20. Torkia C, Best KL, Miller WC, Eng JJ. Balance confidence: a predictor of perceived physical function, perceived mobility, and perceived recovery 1 year after inpatient stroke rehabilitation. Arch Phys Med Rehabil. 2016;97:1064–1071. [DOI] [PubMed] [Google Scholar]
  • 21. Vahlberg B, Cederholm T, Lindmark B, Zetterberg L, Hellstrom K. Factors related to performance-based mobility and self-reported physical activity in individuals 1–3 years after stroke: a cross-sectional cohort study. J Stroke Cerebrovasc Dis. 2013;22:e426–e434. [DOI] [PubMed] [Google Scholar]
  • 22. Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51:1173–1182. [DOI] [PubMed] [Google Scholar]
  • 23. Preacher KJ, Rucker DD, Hayes AF. Addressing moderated mediation hypotheses: theory, methods, and prescriptions. Multivar Behav Res. 2007;42:185–227. [DOI] [PubMed] [Google Scholar]
  • 24. Hackett ML, Pickles K. Part I: frequency of depression after stroke: an updated systematic review and meta-analysis of observational studies. Int J Stroke. 2014;9:1017–1025. [DOI] [PubMed] [Google Scholar]
  • 25. Paolucci S, Gandolfo C, Provinciali L, Torta R, Toso V. The Italian Multicenter Observational Study on post-stroke depression (DESTRO). J Neurol. 2006;253:556–562. [DOI] [PubMed] [Google Scholar]
  • 26. Towfighi A, Ovbiagele B, El Husseini N, et al. Poststroke depression: a scientific statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2017;48:e30–e43. [DOI] [PubMed] [Google Scholar]
  • 27. Blöchl M, Meissner S, Nestler S. Does depression after stroke negatively influence physical disability? A systematic review and meta-analysis of longitudinal studies. J Affect Disord. 2019;247:45–56. [DOI] [PubMed] [Google Scholar]
  • 28. Robinson RG, Starr LB, Kubos KL, Price TR. A two-year longitudinal study of post-stroke mood disorders: findings during the initial evaluation. Stroke. 1983;14:736–741. [DOI] [PubMed] [Google Scholar]
  • 29. Jones F, Riazi A. Self-efficacy and self-management after stroke: a systematic review. Disabil Rehabil. 2011;33:797–810. [DOI] [PubMed] [Google Scholar]
  • 30. Robinson-Smith G, Johnston MV, Allen J. Self-care self-efficacy, quality of life, and depression after stroke. Arch Phys Med Rehabil. 2000;81:460–464. [DOI] [PubMed] [Google Scholar]
  • 31. Salbach NM, Mayo NE, Robichaud-Ekstrand S, Hanley JA, Richards CL, Wood-Dauphinee S. The effect of a task-oriented walking intervention on improving balance self-efficacy poststroke: a randomized, controlled trial. J Am Geriatr Soc. 2005;53:576–582. [DOI] [PubMed] [Google Scholar]
  • 32. Schmid AA, Van Puymbroeck M, Knies K, et al. Fear of falling among people who have sustained a stroke: a 6-month longitudinal pilot study. Am J Occup Ther. 2011;65:125–132. [DOI] [PubMed] [Google Scholar]
  • 33. Barclay R, Ripat J, Mayo N. Factors describing community ambulation after stroke: a mixed-methods study. Clin Rehabil. 2015;29:509–521. [DOI] [PubMed] [Google Scholar]
  • 34. Kossi O, Nindorera F, Adoukonou T, Penta M, Thonnard JL. Determinants of social participation at 1, 3, and 6 months poststroke in Benin. Arch Phys Med Rehabil. 2019;100:2071–2078. [DOI] [PubMed] [Google Scholar]
  • 35. Robinson CA, Shumway-Cook A, Ciol MA, Kartin D. Participation in community walking following stroke: subjective versus objective measures and the impact of personal factors. Phys Ther. 2011;91:1865–1876. [DOI] [PubMed] [Google Scholar]
  • 36. Zhang L, Sui M, Yan T, You L, Li K, Gao Y. A study in persons later after stroke of the relationships between social participation, environmental factors, and depression. Clin Rehabil. 2017;31:394–402. [DOI] [PubMed] [Google Scholar]
  • 37. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. Eng JJ, Dawson AS, Chu KS. Submaximal exercise in persons with stroke: test-retest reliability and concurrent validity with maximal oxygen consumption. Arch Phys Med Rehabil. 2004;85:113–118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39. Fulk GD, Echternach JL. Test-retest reliability and minimal detectable change of gait speed in individuals undergoing rehabilitation after stroke. J Neurol Phys. 2008;32:8–13. [DOI] [PubMed] [Google Scholar]
  • 40. Fugl-Meyer AR, Jaasko L, Leyman I, Olsson S, Steglind S. The post-stroke hemiplegic patient. 1. A method for evaluation of physical performance. Scand J Rehabil Med. 1975;7:13–31. [PubMed] [Google Scholar]
  • 41. Alghadir AH, Al-Eisa ES, Anwer S, Sarkar B. Reliability, validity, and responsiveness of three scales for measuring balance in patients with chronic stroke. BMC Neurol. 2018;18:141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Lin JH, Hsu MJ, Hsu HW, Wu HC, Hsieh CL. Psychometric comparisons of 3 functional ambulation measures for patients with stroke. Stroke. 2010;41:2021–2025. [DOI] [PubMed] [Google Scholar]
  • 43. Flansbjer UB, Holmback AM, Downham D, Patten C, Lexell J. Reliability of gait performance tests in men and women with hemiparesis after stroke. J Rehabil Med. 2005;37:75–82. [DOI] [PubMed] [Google Scholar]
  • 44. Gladstone DJ, Danells CJ, Black SE. The Fugl-Meyer assessment of motor recovery after stroke: a critical review of its measurement properties. Neurorehabil Neural Repair. 2002;16:232–240. [DOI] [PubMed] [Google Scholar]
  • 45. Powell LE, Myers AM. The activities-specific balance confidence (ABC) scale. J Gerontol A Biol Sci Med Sci. 1995;50a:M28–M34. [DOI] [PubMed] [Google Scholar]
  • 46. Salbach NM, Mayo NE, Hanley JA, Richards CL, Wood-Dauphinee S. Psychometric evaluation of the original and Canadian French version of the Activities-Specific Balance Confidence Scale among people with stroke. Arch Phys Med Rehabil. 2006;87:1597–1604. [DOI] [PubMed] [Google Scholar]
  • 47. Botner EM, Miller WC, Eng JJ. Measurement properties of the Activities-Specific Balance Confidence Scale among individuals with stroke. Disabil Rehabil. 2005;27:156–163. [DOI] [PubMed] [Google Scholar]
  • 48. Bandura A. Self-efficacy: The Exercise of Control. New York, NY, USA: W H Freeman/Times Books/Henry Holt & Co; 1997. [Google Scholar]
  • 49. Mulder M, Nijland R. Stroke Impact Scale. J Physiother. 2016;62:117. [DOI] [PubMed] [Google Scholar]
  • 50. World Health Organization . International Classification of Functioning, Disability and Health: Children and Youth Version: ICF-CY. Geneva, Switzerland: World Health Organization; 2007. [Google Scholar]
  • 51. Carod-Artal FJ, Ferreira Coral L, Stieven Trizotto D, Menezes MC. Self- and proxy-report agreement on the Stroke Impact Scale. Stroke. 2009;40:3308–3314. [DOI] [PubMed] [Google Scholar]
  • 52. Guidetti S, Ytterberg C, Ekstam L, Johansson U, Eriksson G. Changes in the impact of stroke between 3 and 12 months post-stroke, assessed with the Stroke Impact Scale. J Rehabil Med. 2014;46:963–968. [DOI] [PubMed] [Google Scholar]
  • 53. Duncan PW, Bode RK, Min Lai S, Perera S. Rasch analysis of a new stroke-specific outcome scale: the Stroke Impact Scale. Arch Phys Med Rehabil. 2003;84:950–963. [DOI] [PubMed] [Google Scholar]
  • 54. Cinamon JS, Finch L, Miller S, Higgins J, Mayo N. Preliminary evidence for the development of a stroke specific geriatric depression scale. Int J Geriatr Psychiatry. 2011;26:188–198. [DOI] [PubMed] [Google Scholar]
  • 55. Sivrioglu EY, Sivrioglu K, Ertan T, et al. Reliability and validity of the Geriatric Depression Scale in detection of poststroke minor depression. J Clin Exp Neuropsychol. 2009;31:999–1006. [DOI] [PubMed] [Google Scholar]
  • 56. Tang WK, Chan SS, Chiu HF, et al. Can the Geriatric Depression Scale detect poststroke depression in Chinese elderly? J Affect Disord. 2004;81:153–156. [DOI] [PubMed] [Google Scholar]
  • 57. Mplus User's Guide [computer program]. Los Angeles, CA, USA: Muthén & Muthén; 1998–2017. [Google Scholar]
  • 58. Hox JJ, Bechger TM. An Introduction to Structural Equation Modeling. Utrecht, the Netherlands: Utrecht University; 1998. [Google Scholar]
  • 59. Kline RB. Principles and Practice of Structural Equation Modelling. 4th ed. New York, NY, USA: The Guilford Press; 2016. [Google Scholar]
  • 60. Widaman KF, Thompson JS. On specifying the null model for incremental fit indices in structural equation modeling. Psychol Methods. 2003;8:16–37. [DOI] [PubMed] [Google Scholar]
  • 61. Brown TA. Confirmatory Factor Analysis for Applied Research. New York, NY, USA: Guilford Publications; 2015. [Google Scholar]
  • 62. Das J, Rajanikant GK. Post stroke depression: the sequelae of cerebral stroke. Neurosci Biobehav Rev. 2018;90:104–114. [DOI] [PubMed] [Google Scholar]
  • 63. Mitchell AJ, Sheth B, Gill J, et al. Prevalence and predictors of post-stroke mood disorders: a meta-analysis and meta-regression of depression, anxiety and adjustment disorder. Gen Hosp Psychiatry. 2017;47:48–60. [DOI] [PubMed] [Google Scholar]

Articles from Physical Therapy are provided here courtesy of Oxford University Press

RESOURCES