Towards an integrative hope-dysfunctional beliefs perspective to personal recovery in schizophrenia: a path analysis

The current study aimed to examine how the above two pathways are meaningfully and empirically connected. First, it examined the relationship between hope, personal recovery, social functioning, and quality of life. Second, it further investigated how defeatist beliefs and asocial beliefs were associated with hope, personal recovery, social functioning, and quality of life. Third, It examined the pathways from defeatist beliefs and asocial beliefs to quality of life, mediated by hope, personal recovery, and social functioning.

First, we made hypotheses regarding correlations. H1: hope, personal recovery, social functioning, and quality of life would be positively intercorrelated. H2: defeatist beliefs and asocial beliefs would negatively correlate with the four constructs in H1. We further hypothesized the four pathways illustrated in Fig. 1. Pathway 1a: asocial beliefs, and subsequently social functioning, would mediate the effect of defeatist beliefs on quality of life. Pathway 1b: asocial beliefs, and subsequently personal recovery, would mediate the effect of defeatist beliefs on quality of life. Pathway 2a: hope, and subsequently personal recovery, would mediate the effect of defeatist beliefs on quality of life. Pathway 2b: hope, and subsequently social functioning, would mediate the effect of defeatist beliefs on quality of life.

A cross-sectional study was conducted by recruiting a convenience sample of 126 informants at six government-sponsored community mental health centers in Hong Kong from July 2020 to June 2021. Inclusion criteria were (a) a diagnosis of DSM-5 schizophrenia spectrum disorders [26], (b) experiencing poor social functioning during the screening period, as indicated by a score of at least 10 in the Social Functioning Questionnaire [27] at the time of recruitment, (c) compliant to medications, (d) aged 18 to 65, (e) able to understand Cantonese or Chinese. Exclusion criteria were (a) neurological disease or damage that would compromise cognitive functioning, (b) physical handicap that would interfere with assessment procedures, or (c) suicidal attempt or ideation in the past three months.

A printed questionnaire containing the following scales was administrated to the informant after written consent was obtained. All data collected were self-reported. Each informant received HKD50 (~ USD6.4) upon questionnaire completion.

Descriptive and correlation analyses were performed using IBM SPSS Statistics version 26. Two informants who had omitted responses to more than 5% of the questionnaire were excluded from the analysis (final n = 124). Multiple imputations were applied to missing data which was missing completely at random (Little’s MCAR test, χ²(2526) = 2417.765, p = 0.94). Then, descriptive analyses were conducted on demographic and measured variables, followed by correlation analyses to identify associations (Pearson's r) between variables.

Structural equation modeling analysis was conducted with R 4.2.2 [39] with package “lavaan” [40]. First, to simplify our model, a latent construct “quality of life” was created to explain covariance between four domains of quality of life, with the model fit tested under a confirmatory factor analysis (CFA; R function “cfa”). Second, a path analysis (R function “sem”) was conducted to examine the overall model fit. Model fit criteria of the CFA and path analysis include chi square (χ²), comparative fit index (CFI), Tucker Lewis Index (TLI), standardized root mean square residual (SRMR), and root mean square error of approximation (RMSEA). Third, bootstrapping analysis with 5,000 resamples were performed to determine the bias-corrected 95% confidence intervals [41] of each mediation paths. To minimize the possibility of Type I error in marginal situations, the significance of the mediation paths was further checked with the test of joint significance, in which the mediation was deemed significant if all individual paths constituting the compound path were significant [42].

Of the 124 informants, 58.9% were female. Their age ranged between 19 and 81 years, with a mean (SD) of 42.70 (13.10). Regarding education, 3.2% of informants graduated from the 6^th grade, 34.7% graduated from the 9^th grade, 37.1% graduated from high school, and 25.0% graduated from tertiary education. Regarding medication, 96.0% had regular medication prescribed by a doctor, 0.8% had no medication, and 3.2% had unstable medication. Meanwhile, 93.5% received periodic case management follow-up, and 6.5% received irregular follow-up. In the previous six months, informants had been hospitalized for a mean (SD) of 2.09 (8.72) days and 0.12 (0.39) times.

Means, standard deviations (SD) and correlations are shown in Table 1. The Mean (SD) of social functioning was 10.84 (3.24), with 65.3% of informants scoring equal to or higher than the cut-off score of 10, indicating poor social functioning [27].

Higher hope, better personal recovery, better functional recovery, and better quality of life in four domains were all associated (all p < 0.05). Besides, having higher defeatist beliefs was associated with poorer social functioning (r = -0.20, p < 0.05), lower hope (r = 0.22, p < 0.05), poorer quality of life (r_physical = 0.29, p < 0.01, r_{psychological} = 0.35, p < 0.001, r_social = 0.33, p < 0.001, r_{environmental} = 0.31, p < 0.001), and higher asocial beliefs (r = -0.43, p < 0.001). However, defeatist beliefs were not correlated with personal recovery. Meanwhile, higher asocial beliefs were associated with poorer social functioning (r = 0.37, p < 0.001), poorer personal recovery (r = -0.25, p < 0.01), poorer quality of life (r_physical = -0.28, p < 0.01, r_{psychological} = -0.22, p < 0.05, r_social = -0.20, p < 0.05, r_{environmental} = -0.24, p < 0.01), although asocial beliefs were not correlated with hope.

The confirmatory factor analysis (CFA) of quality of life showed an excellent fit (χ² = 2.41, p = 0.30, CFI = 1.00, TLI = 0.99, RMSEA = 0.04, 90%CI [0.00, 0.19], SRMR = 0.02).

Figure 2 presents the path diagram, and Table 2 the result of bootstrapping. Model fit was good (χ² = 41.07, p < 0.01, CFI = 0.96, TLI = 0.94, RMSEA = 0.08, 90%CI [0.04, 0.12], SRMR = 0.05), and the whole model explained 75.7% of the variance of the latent factor “quality of life”. Mediation pathways 1a (β = 0.03, p < 0.05, 95% CI [0.01, 0.06]) and pathway 1b (β = 0.03, p < 0.05, 95% CI [0.00, 0.07]) were significant, whereas pathway 2a (β = 0.07, p < 0.10, 95% CI [0.01, 0.16]) and pathway 2b (β = 0.02, p < 0.10, 95% CI [0.00, 0.04]) were marginally significant. Although some confidence intervals included zero when rounded, none of the lower boundaries fell into negative values, and the test of joint significance [42] supported the significance of all four specified paths.

We hypothesized that defeatist beliefs, via asocial beliefs, influence the social functioning and sense of recovery of a person recovering from severe mental illness. The results of path analysis indicated that fewer defeatist beliefs were associated with fewer asocial beliefs, and in turn, predicted higher social functioning (pathway 1a). Interestingly, fewer defeatist beliefs and asocial beliefs were also found to be associated with a greater sense of personal recovery (pathway 1b). These two pathways ultimiately converged into better quality of life for the person. Both findings are consistent with existing literature [10, 14] and seem to suggest that despite suffering from severe mental illness, those with fewer defeatist beliefs may have a more positive outlook on life and be more willing to participate in social and functional activities, leading to better social functioning. Similarly, a person with severe mental illness with a more positive outlook on life will have a better sense of personal recovery (e.g., better positive attitudes towards mental illness and greater hope toward recovery), resulting in better quality of life as a whole. The findings echo a defeatist-asocial belief pathway and reaffirm the importance of helping people with severe mental illness develop strategies to modify their defeatist and asocial beliefs and enhance their overall positive attitudes in life by engaging in activities that provide them with successful experiences and help them build up confidence and motivation to continue to engage in recovery-oriented activities. The Cognitive Therapy-Recover Model initiated by Grant et al. [14] exemplifies this approach to improving the lives of people with severe mental illness.

Supporting our hypotheses, this study also confirms the hope, motivation and strength pathway towards better social functioning and quality of life for people with severe mental illness (Pathway 2a and 2b). Research has repeatedly found hope an indispensable factor underpinning the process of personal recovery [20]. In essence, hope represents “the beliefs that it is possible for someone to regain a meaningful life, despite serious mental illness” ([24]^{(p. S621)}), motivating people to make the change (Pathway 2b). In the hope-personal recovery pathway (Pathway 2a), however, a person does not only embrace a general positive outlook in life, but they also take action to identify and utilize internal and external resources (i.e., strengths) to work towards achieving life goals that enhance their meaning in life. Indeed, the current recovery movement engineered by scholars such as Slade et al. [18] and Rapp et al. [20] has adopted this line of enquiry to improve the lives of people with severe mental illness.

Our results in pathways 1a and 1b revealed that fewer asocial beliefs predicted higher social functioning and higher personal recovery. To the best of our knowledge, this is one of very few studies that has established two relationships stemming from asocial beliefs to social functioning and personal recovery respectively, particularly among an Asian population. Previous research had hypothesized but had not fully, empirically substantiated such possible relationships. One of the thorny issues facing by people with severe mental illness is social withdrawal, which has a strong link to deterioration in social functioning and poor mental health [43]. Our current empirical findings not only affirm the need to develop strategies to enhance the social connectedness of people with severe mental illness but also highlight the importance of fostering positive attitudes and beliefs towards the need for interpersonal relationships among people with severe mental illness. Indeed, recovery from severe mental illness is very much an interpersonal process that requires continuous support from peers, family members, friends, colleagues, religious groups, and community member [44]. These provide people with connectedness and a sense of belonging, opportunities for social learning and practicing skills, instrumental and emotional support, positive feedback and encouragement [45]. Thus, while it is essential to facilitate a person with severe mental illness to be linked to others socially, it is equally important to help them to process the connection in a meaningful way so that they are able to appreciate the benefits through such connection.

As shown by the results, hope mediated the effect of defeatist beliefs on personal recovery and social functioning. The pathway from defeatist beliefs to hope is another new and interesting result in our study. Indeed, this is one of the few studies providing empirical support for the linkage between defeatist beliefs and hope, thus bridging the two dominant lines of enquiry on recovery for people with severe mental illness. This linkage is not difficult to understand given that people with severe mental illness, such as schizophrenia, have neurocognitive deficits in memory and attention. These deficits contribute to possible unsuccessful goal attainment in life, which over time, can give rise to dysfunctional, defeatist attitudes about oneself and one’s performance. These dysfunctional attitudes, in turn, may lead to dissatisfaction and a sense of despair and hopelessness. On the other hand, a reduction in defeatist beliefs can increase the sense of hope among people with severe mental illness and then develop into different hope pathways identifed by our model.

This is one of the first few studies that have attempted to examine the connection between a cognitive deficit and hope-strength connection in the recovery of people with schizhophrenia. The conventional hope pathway in mental health recovery programs would suggest that an increase in hope, say, through building successful experiences, can build a strength-oriented momentum, mitigate negative presumptions and enhance a person’s motivation to make further positive changes [44]. But our findings highlight the significance of defeatist beliefs as underlying factors influencing hope and asocial beliefs, which further lead to poor social functioning and quality of life of people with severe mental illness. It adds another dimension, on top of existing recovery model, to suggest that cognitively changing one’s defeatist beliefs through different means can induce a sense of hope and lead a person to the hope pathway. This provides a diversion to the conventional recovery programs, which claim that irrespective of deficits, the focus on strength building can independently enhance the full recovery of a person. Our study suggests, in line with the cognitive model of negative symptoms postulated by Beck et al. [9], perhaps, the need for a more balanced deficit-and-strength perspective to conceptualize the lives of people with severe mental illness and interventions. While it is important to explore and facilitate a person to use their internal strength and external resources to achieve full recovery, there is also a need to work through various deficits to maximize optimal recovery in a person. Thus, a future recovery-oriented approach should pay attention to both deficits and strengths of the people in both assessment and interventions.

The current study had several limitations. First, the data were collected through a cross-sectional survey. Although path analysis can test our theoretically constructed model against the data, our result could not ascertain the temporal, causal relationship between variables. Further studies with longitudinal design and clinical trials are required to further establish the causality among variables. Second, all the measurements were based on self-reported scales, which could be affected by subjective factors. Some studies have reported that self-reported quality of life and functioning could be affected by insight [46]. Further studies may adopt multi-rater, role-play (e.g., for social functioning) [17], or clinician-reported approaches to mitigate potential bias in self-reported data. Third, the internal consistency of the social functioning questionnaire was poor. We conducted further analyses and found that the low inter-item correlations were not caused by any single item in the scale. In our data, inter-item correlations of the social functioning scale were generally mediocre. Since the eight items in the scale measured perceived functioning in eight aspects of life, it might indicate that, for people with schizophrenia, high functioning in one aspect of life is not associated with high functioning in another aspect. Fourth, the current study did not divide informants into deficit vs. nondeficit groups. Some scholars have suggested deficit and non-deficit schizophrenia are two distinct disorders [47]. Hence, further studies may investigate how heterogeneity within diagnosis or the existence of positive symptoms may affect therapeutic effectiveness. Fifth, there could be a lack of referenced literature from the last five years. Meanwhile, referenced studies in this study were conducted in the last decade, after the publication of cognitive theory of negative symptoms by Grant and Back in 2009 – 2010 [7‐9]. Further studies will be required to enlighten researchers on the association between cognition, negative symptoms, and recovery.