Occupational value outcomes among people with mental health issues participating in activity-based interventions – The importance of internal and external factors

Abstract Background Perceived occupational value is an important treatment outcome but is still relatively unexplored as a phenomenon. Aim To evaluate whether the Balancing Everyday Life (BEL) intervention for people with mental health issues was effective compared to Standard Occupational Therapy (SOT) regarding improvement on three dimensions of occupational value (concrete, socio-symbolic and self-reward), and how internal factors (self-esteem and self-mastery) and external factors (sociodemographics) were related to occupational value. Methods The study was a cluster RCT (n = 226) and used self-report questionnaires on three occasions (baseline = T1, completed intervention = T2, six-month follow-up = T3). Results No differences between the groups were seen in occupational value change scores. Within-group analyses (T1–T3) indicated a change in the BEL group on concrete value and self-reward value. No change was indicated in the SOT group. Associations showed that self-esteem and self-mastery were correlated with all three aspects of occupational value. Having children was negative for experiencing occupational value, whereas having a friend was positive. None of the correlates could predict changes in occupational value. Conclusion Self-related factors appeared to be integral in occupational value. Significance Since occupational value is essential for a meaningful life, therapists should consider such factors and peer support when supporting people with mental health issues.


Introduction
The way people perceive their everyday activities is important for their well-being and quality of life, as evidenced in research addressing various target groups [1][2][3][4]. The current study focuses on people with mental health issues, and several studies have confirmed such a relationship in this group [5,6]. There is also evidence that interventions based on everyday activities, specifically occupational therapy based on time-use and occupational balance, have a positive effect on well-being [7]. One way of operationalising perceptions of one's everyday activities is to address the value the activities bring to the person, termed occupational value [4,5,8,9]. Occupational value is defined by three constituents, termed dimensions. The concrete dimension encompasses tangible and observable aspects of occupational value, such as the value experienced from working and earning money, or from having cleaned the house. The second dimension, originally termed symbolic but later on reframed into socio-symbolic value [9], is about what the occupation signifies for the person, on a personal, a cultural and sometimes a universal level. For example, enjoying a good meal with friends can have symbolic meaning that taps all three levels, since this activity is interpreted similarly in many different cultures. Occupational value is closely linked with the meaningfulness people find in their everyday life [4,8], which underlines its importance to both occupational therapists and their clients. Yet, occupational value has not been extensively investigated as an occupational therapy outcome, or as a phenomenon, which was an incentive for the current study.
One of the occupational therapy interventions shown to be effective for people with mental health issues is the 16-week Balancing Everyday Life (BEL) program, which was evaluated in an RCT. It is group-based yet individualised and builds upon principles from occupational science and personal recovery. Findings showed that, compared to a group receiving standard occupational therapy (SOT), the BEL participants had improved more after the 16 weeks on occupational engagement, activity level, occupational balance and level of functioning; at a six-month follow-up they had also improved more on quality of life [10]. Overall occupational value was addressed as an outcome in that study, but not the three value dimensions, which are the focus of the current one.
The rationale for the current study was twofold; the primary goal was to investigate additional outcomesthe three occupational value dimensionsfrom the BEL project; the second was to gain more knowledge of occupational value as an outcome measure, such as possible influencing factors. In particular, it would be important to know if the occupational value is related to internal, trait-like variables, such as self-esteem, or with external, sociodemographic factors, or both.
The aim of the study was to evaluate if the BEL intervention was effective compared to SOT regarding improvement on the aforementioned dimensions of occupational value. Part of the aim was to investigate how internal factors (self-mastery and self-esteem) and external factors (sociodemographics in terms of, e.g. age, gender, having children) were related to perceived occupational value among persons participating in the interventions, including change in the occupational value dimensions after completed intervention and at a six-month follow-up.

Methods
The study was part of a comprehensive cluster RCT [10], approved by the Regional Ethical Vetting Board in Lund (Reg. No. 2012/70) and adhering to the Swedish legislation governing research on humans [11] and the Helsinki Declaration of 1975, as revised in 1983 and 2004. The trial was registered with ClinicalTrial.gov. (Reg. No. NCT02619318). The RCT as a whole comprised several outcomes, such as quality of life, level of functioning, occupational engagement [10] and personal recovery [12], as well as a study on predictors [13] and another on implementing BEL in a multi-professional context [14].

Settings and participants
The context for the study was specialist and community-based psychiatry in three healthcare regions in Sweden. Outpatient units within general psychiatry, psychosis care, and community-based mental health centres were invited to the project. Thirty-seven units were identified, but eight were excluded because they participated in other projects or were undergoing reorganisation. The remaining 29 settings consented, and 15 were randomised to the BEL arm and 14 to the SOT arm. Because of illness among staff, one BEL setting could not start the intervention and withdrew.
A gatekeeper (an occupational therapist from the staff) was recruited at each consenting setting to help identify participants meeting the following criteria: need for better balance among everyday activities (e.g. wanting to change their repertoire of activities), aged 18-65 years, substance use disorder not the main diagnosis, and no comorbidity of dementia or developmental disorder. No further exclusion of psychiatric diagnosis was made. All eligible participants were invited and the gatekeeper provided oral and written information about the project. Those who accepted the invitation were then contacted by a research assistant for scheduling appointments for data collection, which took place in the respective settings. Participants received a small payment for participation.
Based on the means and standard deviations from a previous study using the Satisfaction with Daily Occupations instrument [15], we estimated that 65 participants in each group were needed to detect a difference of 0.5 with 80% power at p < 0.05. Expecting 25% attrition, we aimed to include 95 participants from BEL settings and the same number from SOT settings. Since the project was a cluster RCT, and all eligible participants were invited, it was hard to reach the exact number and we arrived at a larger sample than necessary with 226 participants entering the study; 133 from BEL settings and 93 from SOT settings. About 90% of both groups were Swedish-born. Table 1 displays further characteristics. Also, no differences were found between the groups as regards the known sociodemographic and clinical characteristics.
There were 33 dropouts (25%) from baseline to completed intervention in the BEL group and 13 (14%) in the SOT group. This was a statistically significant difference (p ¼ 0.047), but the drop-outs did not differ from completers on known characteristics in terms of sex, age, education level, family situation or diagnosis (p-values ranging between 0.155 and 0.712). Explanations for the dropouts in the BEL group mostly concerned non-completion of the intervention, such as attending only the first group session. Reasons for dropping out in terms of an illness episode or not wanting to complete the data collection were equally common in the two groups. Between the completed intervention and the follow-up there were another 11 dropouts (8%) in the BEL group and 10 (11%) in the SOT group. No statistically significant difference was found (p ¼ 0.527).

The two interventions
Inspired by previous research on activity-based lifestyle interventions [16,17] and in-depth studies on everyday life among people with mental health issues [18][19][20][21], a team of researchers and clinicians developed BEL for use by occupational therapists working in mental healthcare. BEL is group-based (5-8 participants) and comprises 12 themes, described in detail in a manual [22]. Some examples of themes are; my sources for meaning and motivation, activity balance, social life, productive activities, and leisure and relaxation, and ends with two booster sessions without the introduction of new themes. Whereas the number of themes is fixed, feedback from participants and group leaders [23] resulted in the length of the program varying between 16 and 19 weeks depending on the capacity of the participants. In the current study, BEL was given as a 16-week program, one session a week for the 12 themes and 2 biweekly booster sessions. The BEL sessions consist of brief lessons, exercises and discussions. Processes that take place during the program concern self-analysis of wanted or missed activities and exploring goals and strategies for testing self-chosen activities. Targeted activities are then tested in between sessions in the participant's natural environment, and informal peer support is encouraged. The ultimate purpose of BEL is that the participants will be able to reflect on their activity situation and develop strategies for steering their everyday life towards activity balance, in terms of a personalised equilibrium between work and relaxation, social activities and seclusion, etc. The BEL program is led by one or (preferably) two therapists who have undergone a 2-3 day's training program. A study addressing how BEL was implemented in diverse specialised and community-based mental health settings showed high fidelity to the manual and that the intended dose was delivered [24].
The SOT intervention was standard occupational therapy given in specialised and community-based mental health settings. The SOT typically encompassed some form of group intervention, addressing social skills, creative activities, productive activities, or leisure, but some occupational therapists also provided individual therapy with the same contents. The occupational therapists followed the prevailing "best practice", adjusted to the participants' mental health status and support needs.
There were several similarities between the interventions. Both were activity-based and led by occupational therapists who were part of a multi-professional team that could provide a variety of therapies. Most participants in both BEL and SOT also met with other team members and received either or both psychotropic medication and counselling. The differences concerned BEL being a structured and manualized intervention, following the themes, methods, and time frames specified above, whereas SOT represented a variety of means and methods of less intensity.

Data collection
The data collection was based on self-report questionnaires. Research assistants administered the data collection, done on paper. The participants completed the questionnaires themselves but could consult the research assistant if they needed help understanding. Activity value Activity value was assessed by the Occupational Value with predefined items (OVal-pd) instrument [24]. It consists of 18 items with statements regarding the frequency with which the respondent had experienced exemplified variants of occupational value during the past month. Each item starts with "In the past four weeks, I have carried out occupations that … " and is then followed by a specific value experience. Some examples are " … where something important was done" (concrete value), " … which made me feel connected to people (socio-symbolic value) and "because it was a true pleasure to do" (self-reward value). A four-point scale ranging from 1 (not at all) to 4 (very often) is used. The OVal-pd has been found to be psychometrically sound, showing good item fit and homogeneity [24], as well as structural validity and test-retest stability [25].

Internal factors
The internal factors addressed in this study were selfesteem and self-mastery. The Rosenberg self-esteem scale [26] is ten items that reflect different aspects of self-esteem, such as feeling like a person of worth. Five are positively worded and five are negatively. The current study used the yes/no response format, which has been recommended by Oliver and colleagues [27] as preferable with samples that include people with psychosis. The scoring procedure means subtracting the average of the negative items from the average of the positive ones. This results in a selfesteem score that can vary between À 1 and 1. The Swedish version of the Rosenberg Self-esteem scale has been found to possess good internal consistency, good criterion validity, convergent and discriminant validity, as well as sensitivity to change [28]. The Pearlin Mastery Scale [29] has seven items that express whether or not one feels in control of one's life situation. The items are rated according to four rating alternatives, from 1 which indicates the lowest level of mastery to 4 which indicates the highest. This means that the sum score may range between 7 and 28 and a higher score indicates stronger self-mastery. Rasch's analysis of the Swedish version, Mastery-S, has shown adequate reliability and known-groups validity, and the scale represents a logical continuum of the measured construct [30].

External factors
Data on sociodemographics were gathered by a background questionnaire developed for the project. The variables used for the current study concerned age, gender, having children or not, and having a friend or not.
The background questionnaire also addressed clinical information in terms of self-reported diagnosis, perceived health problems and medication. Proceeding from these self-reports, a specialised psychiatrist made ICD-10 diagnoses [31], used for descriptive purposes in the current study.

Data analysis
Non-parametric statistics were used since the questionnaires produced ordinal data [32]. Within-group changes in occupational value were analysed by the Mann-Whitney test or the Kruskal-Wallis test. For comparisons between groups, change scores were calculated by subtracting the baseline occupational value scores from the scores at completion and the follow-up, respectively. Group differences in change were then analysed by the Wilcoxon test. Associations between the inner and outer factors selected as covariates and the occupational value variables were calculated by Spearman rank correlations. The software used was SPSS 28 [33] and the value for statistical significance was set at p < 0.05.

Changes in occupational value after intervention and follow-up
Between-group comparisons based on change scores indicated no statistically significant differences between the BEL and the SOT group regarding change on any of the occupational value dimensions, neither at completed intervention nor at the followup, p-values ranging between 0.178 and 0.747. Table 2 displays within-group analyses showing that the BEL group improved on concrete value and self-reward value from baseline to completed intervention at 16 weeks, whereas the SOT group improved on socio-symbolic value. At the follow-up, the BEL group had increased their scores on all three occupational value dimensions, and the SOT group had increased concrete and self-reward value.
Not shown in Table 2, analyses encompassing variables from all three measurement occasions indicated a statistically significant change in the BEL group on concrete value (v 2 ¼ 6.2; p ¼ 0.046) and self-reward value (v 2 ¼ 10.4; p ¼ 0.005), but no statistically significant change in the SOT group (p ¼ 0.092 and 0.140, respectively). According to these within-group analyses based on all three occasions, none of the groups improved on socio-symbolic value.

Associations between internal and external factors and occupational value
Since no between-group differences were established regarding the occupational value dimensions, analyses of associations between internal and external factors on the one hand and the occupational value scores on the other were performed on the sample as a whole. Selfesteem and self-mastery were both related to all three estimates of occupational value at baseline. All of these baseline associations were highly statistically significant (p < 0.001) and varied between r s ¼0.476 (between selfesteem and socio-symbolic value) and r s ¼0.364 (between self-mastery and self-reward value). As seen in Table 3, improved self-esteem from baseline to completed intervention was related to higher ratings on all three occupational value dimensions. Regarding the change in the correlate of self-mastery, the only statistically significant association at completed intervention was with the change in self-reward value. Focussing on change at the follow-up, there was an association between improved self-esteem and increments in concrete value and self-reward value, but no statistically significant association was seen with a change in sociosymbolic value. Increased self-mastery was positively associated with augmented values for all three dimensions of occupational value. Also shown in Table 3, baseline self-esteem and baseline self-mastery were not statistically significantly related to a change in the occupational value variables at the completed intervention or the six-month follow-up.
External factors, in terms of selected sociodemographic factors, were also analysed for their relationships with occupational value. At both baseline (Z¼-2.43 p ¼ 0.015) and at completed intervention (Z¼-2.0 p ¼ 0.045), having children was associated with lower ratings on self-reward value, but not with any of the other occupational value variables. At the follow-up, those with children scored lower than those without on all three occupational value dimensions (concrete value: Z¼-1.98; p ¼ 0.048; socio-symbolic value: Z¼-2.28; p ¼ 0.023; self-reward value: Z¼-3.1; p ¼ 0.002). Regarding occupational value change, no differences were found between those with children and those without in any of the occupational value dimensions, p-values ranging between 0.543 and 0.918 for changes from baseline to completed intervention and between 0.142 and 0.219 for changes from baseline to the follow-up.
Furthermore, as displayed in Table 4, having a friend was of importance to all of the occupational  value variables, on all three occasions. In all instances, those who had a friend scored higher than those who were without a friend. However, there were no differences between those who had a friend and those who had not regarding change in any of the occupational value dimensions, p-values ranging between 0.195 and 0.795 for change from baseline to completed intervention and between 0.210 and 0.555 for change from baseline to the follow-up. Both sex and age were unrelated to the occupational value dimensions, and this concerned the discrete measurements of occupational value as well as the change variables.

Discussion
Although the BEL group improved on more occupational value dimensions than the SOT group did, there was no difference between the groups regarding the change in any of the dimensions. This is partly in line with an evaluation of an intervention based on time use and patterns of daily occupations for people with stress-related disorders (Redesigning Daily Occupations -ReDO), which used occupational value as an outcome. Unlike the current findings, that study showed that both the intervention group and the comparison group improved during an intervention, but similar to the current study there was no difference in improvement between the groups [34]. The larger RCT, of which the current study is part, showed improvements in the BEL group compared to the group receiving SOT regarding occupational engagement, activity level and occupational balance [10]. The fact that no such group difference was found in the present study might indicate that occupational value, as measured with the OVal-pd questionnaire [24], was less sensitive to change. However, each outcome addresses a specific phenomenon, so variability in results should be expected.
The variables selected to indicate internal factors, self-esteem and self-mastery, were moderately correlated with all occupational value dimensions in a cross-sectional perspective. Furthermore, changes in the internal factors were often related to a corresponding change in the occupational value variables. These findings indicate that a Self-factor is indeed one of the components that make up the phenomenon of occupational value, which has support in previous research as well [5]. None of the selected internal factors was however predictive of change in any of the occupational value variables, which also aligns with previous research [34]. Thus, the overlap between Self-factors and occupational value is not sufficient for the former to be predictive of the latter. The same can be said about one of the external factors addressed, namely having a friend. This variable was consistently related to all of the occupational value dimensions in a cross-sectional perspective but did not predict any aspect of occupational value change. This may indicate that the benefits of having a friend had already had an impact on occupational value perceptions at baseline, and that having a friend did not boost intervention outcomes in this respect. Nevertheless, the findings add to the knowledge about the phenomenon of occupational value, and the social aspect assumed in the ValMO model [8] can be confirmed. Also in line with the model assumptions, the link between having a friend and the occupational value ratings were particularly strong for socio-symbolic value. These findings of associations strengthen not only the assumptions behind the ValMO model but also the validity of the OVal-pd questionnaire.
Having children appeared to be a negative factor for rating an occupational value high, particularly concerning self-reward value. It is likely that a family with children involves more home chores, which may be time-consuming and leave little room for pleasureoriented activities for parents. At the follow-up, however, all aspects of occupational value were rated lower by those who were parents compared to those who were not. This indicates that the everyday chores one could expect among the parents did also not bring much concrete or socio-symbolic value. Importantly, the likelihood of benefitting from the interventions, in terms of increased occupational value, was not affected by having children or not.
Age and sex were not related to perceptions of occupational value in the current study. A study addressing another target group, families with obese children, indicated a possible sex difference. The fathers' but not the mothers' occupational value increased as the children lost weight after a familyoriented intervention [35]. Generally, sociodemographic factors' importance to occupational value is an under-investigated area and should be subjected to further study.
From a mental healthcare perspective, it would be important to assist people with mental health issues in organising everyday life in such a way that they increase their sense of occupational value, shown to be of importance to well-being [5] and a sense of meaningfulness [4,8]. This makes occupational value an important target for both clients and occupational therapists. First, it should be considered when working out the therapy plan. All interventions aimed at stimulating occupational value need to leave much space for activities that align with service users' personal preferences and interests, which is an integral aspect of many occupational therapy programs [23,36,37]. Additionally, based on the current identified cross-sectional associations, access to friendship can boost service users' sense of occupational value. Peer support could be a helpful alternative in this respect, providing peer contact that may not only boost individual outcomes but also increase trust in mental health services [38]. The present findings also indicate that factors of interpersonal character were important in a cross-sectional perspective, and efforts to support service users' self-esteem, self-worth and control over their life situation would be important strategies along with boosting opportunities for accessing valued and satisfying activities. Secondly, the occupational value could be used as an indicator when evaluating the outcomes of occupational therapy. In addition to the OVal-pd instrument used in the current study, which addresses occupational value in people's lives in general, there exists a variant that targets the value generated by a specific occupation, termed OVal-9 [39,40].
This study also generated some new information regarding the OVal-pd as an instrument. First, it identified associations that are in line with the ValMO model, which strengthens the instrument's relevance and construct validity. Second, the fact that there were more consistent indications of positive occupational value change in the BEL group than in the SOT group, including interactions between changes in the interpersonal factors and the occupational value variables, suggests that the OVal-pd has some preliminary sensitivity to change.

Methodological considerations
The internal validity of this study must be seen as good. It was sufficiently powered, was based on solid RCT methodology, and used psychometrically sound instruments. The drop-out rate in the BEL group at the completion of the intervention may seem cumbersome, but the dropouts did not differ from the completers on any of the known characteristics. The study must also be regarded as having external validity in the area of people with mental health issues since the sample was fairly large and diagnostically diversified. The reliability of the study as a whole is hard to assess since this was the first study of its kind, but findings could verify previous research where occupational value was found to be related to selfrelated variables [5,41] and with personality factors [42]. No true predictors were identified, however, only cross-sectional associations and interactions between the selected interpersonal factors and occupational value. This weakens the relevance of the proposed support measures above, although they also have some support in previous research [38,43] The scope of possible predictors was also limited, and future research should address, for example, if and how the support measures proposed above influence occupational value outcomes.

Conclusion
This study pointed to improvements in the BEL group regarding their perceived occupational value, but not greater than the improvements identified in the SOT group. Furthermore, the study yielded new knowledge about occupational value as a phenomenon and as an outcome. Internal, self-related factors appear to be integral in occupational value as a concept, as does an interpersonal aspect in terms of having a friend. This is something occupational therapists and other staff can consider when attempting to support people with mental health issues to move towards an everyday life with more valued and meaningful activities, thereby supporting their well-being and quality of life. Age and sex seemed unrelated to the occupational value dimensions, but are something that deserves further investigation.

Disclosure statement
The author reports no conflict of interests. The author alone is responsible for the content and writing of the paper.

Funding
This work was supported by the Swedish Research Council under Grant K2014-99X-20067-09-4. The funding body had no role in the design, collection, analysis, or interpretation of data, or in writing the manuscript.