Introduction

Mental health disorders will affect approximately 20 % of youth with about 14 % in the clinical range (Costello, Mustillo, Erkanli, Keeler, & Angold, 2003; Waddell, Offord, Shepherd, Hua, & McEwan, 2002). The high prevalence and the complexity of the etiology of mental health disorders are concerning; furthermore, mental health disorders can have adverse effects in many areas of an individual’s life including social, cognitive and occupational functioning (Meng & D’Arcy, 2013). Emotional and behavioral disorders can be quite severe and may be best treated with intensive treatment, residential- or home-based. These intensive mental health treatments are often centered on a strengths-based approach that includes psychotherapy and skill building and supportive interventions, and have been associated with improved clinical and functional status of some youth experiencing these disorders (Bond, Drake, Becker, & Mueser, 1999; Lyons, Uziel-Miller, Reyes, & Sokol, 2000; Preyde et al. 2011a). However, the youths’ ratings of their subjective or personal well-being (PWB) are not known nor is it known if there is an association between these youths’ clinical characteristics and their self-reported PWB.

Intensive treatment offered in residential centers or the home is designed to treat youth with moderate to severe emotional and behavioral disorders (EBD), and these youth often have more than one disorder (Barth et al., 2007; Connor, Doerfler, Toscano, Volungis, & Steingard, 2004). Common disorders include conduct disorder, attention deficit hyperactivity disorder (ADHD), oppositional defiance disorder (ODD), post-traumatic stress disorder (PTSD), obsessive–compulsive disorder (OCD), anxiety and mood disorder (Brady & Caraway, 2002; Connor et al., 2004; Preyde et al., 2011a). These youth’s needs are often not adequately managed in other settings which may be indicative of the severity of their emotional and behavioral problems, and the complexity of their domains of living (Cameron, Frensch, Preyde, & Quosai, 2011; Frensch, Cameron, & Preyde, 2009; Preyde, Frensch, Cameron, Hazineh, & Riosa, 2011b; Zelechoski et al., 2013) such as challenges in the home environment, school and community conduct.

Some investigators have found that youth accessing intensive mental health treatment programs generally experience some improvement in clinical outcomes from admission to discharge though many are still within the clinical range upon discharge (Boyer, Hallion, Hammell, & Button, 2009; Briggs et al., 2012; Knorth, Harder, Zandberg, & Kendrick, 2008; Preyde et al., 2011b). Additionally, many youth accessing these intensive interventions (Greenbaum et al., 1996; Preyde et al., 2011a, 2011b, 2013; VanderStoep et al., 2000) have experienced significant adversities in early life and throughout their young lives, including experiences of poverty, family conflict, harsh caregiving, disruptions in living domains and challenges at school both academically and socially, and continued to experience difficulties after treatment. In a review of life satisfaction in youth, Proctor, Linley, & Maltby (2009) indicated that youth experiencing these negative family and environmental factors reported lower PWB than youth without such adversity. Though youth have discussed their continued experiences of mental health disorder (Preyde et al., 2013), no reports of their personal well-being (PWB) have been located.

PWB can be defined as an individual’s subjective evaluation of his or her life with specific regard to affective and cognitive states or a sense of contentment, and satisfaction with life or happiness (Cummins, 2000; Diener, 2000). It is a multi-faceted construct that includes emotional responses and global assessments of life and domain-based satisfaction. PWB is considered a fairly stable mood state of high subjective ratings or dispositional happiness of well-being which is maintained through homeostatic mechanisms termed homeostatically protected mood (HPM; Cummins, 2009; Cummins, Mellor, Stokes, & Lau, 2010). From this perspective, PWB is described as being regulated by psychological mechanisms that evolved to protect mood. Positive moods may be maintained through cognitive restructuring such that negative experiences are re-conceptualized as positive thoughts. Hardships are mitigated by processes of adaptation and habituation, cognitive buffers and positive affectivity. However, considerable adverse conditions can result in a PWB rating that is lower than the normal homeostatic range. Some research has been conducted to test this theory and support has been shown in a convenience sample of adolescents recruited from a school (Tomyn & Cummins, 2011). In addition to HPM, another consideration concerns PWB and mental health disorder.

Intuitively some may consider PWB a component of or the same as mental health; however, the relationship between PWB and mental health disorder has been both supported and refuted in research. It has been suggested that low personal well-being might merge into mental health disorder (Valois, Zullig, Huebner, & Drane, 2004). Research with patients with schizophrenia suggests that there may be a weak association between PWB and symptoms of psychopathology (e.g., Eack & Newhill, 2007; Thorup, Petersen, Jeppesen, & Nordentoft, 2010) and with the self-reported needs in life areas such as accommodation and physical health (Werner, 2012). Similarly, in a sample of men with various mental health diagnoses (Gargiulo & Stokes, 2009), PWB only marginally predicted those with clinical depression, suggesting that the relationship between clinical depression and PWB is complex.

Much of the research with children and adolescents has been conducted with school or normative populations. There is a dearth of research with clinical samples of adolescents; however, in one study, PWB was shown to negatively correlate with distressing side effects of anti-psychotic medication in adolescents (Schimmelmann et al., 2005) as assessed during their first hospitalization for this illness.

Similarly, investigators have often overlooked strengths and positive indicators of development among youth with emotional and behavioral disorders. High PWB scores have been found to be associated with many desirable and enduring outcomes in normative samples. For example, it has been shown that variables associated with very happy people include sociability and increased time spent with others (Diener & Seligman, 2002), prosocial behavior (Lucas, 2001; Thoits & Hewitt, 2001), decreased likelihood of engaging in unhealthy behaviors (Pettit, Kline, Gencoz, Gencoz, & Joiner, 2001), increased likelihood of engaging in healthy behaviors (Lox, Burns, Treasure, & Wasley, 1999), and high immune functioning (Gil et al., 2004). In the current movement toward building resiliency and positive youth development (Damon, 2004) for vulnerable youth such as those accessing intensive mental health treatment, a high PWB may have a role in positive youth development and fostering optimal development (Diener et al., 1999). However, youth who access intensive mental health treatment have typically experienced a multitude of stressors and challenges in their lives which creates a curious question about their perceptions of their personal well-being.

Although children as young as 5 years of age have been found to be capable of accurately and reliably self-reporting their health related quality of life (Varni, Limbers, & Burwinkle, 2007), there is scant research on the PWB of adolescents experiencing mental illness. It may be beneficial to examine the level of PWB in youth with emotional/behavioral disorders and to explore the association between PWB and clinical indicators of mental illness.

The purpose of this study was to explore the level of PWB in youth who accessed intensive mental health treatment (i.e., residential treatment (RT) and home based treatment (HBT). A second purpose was to explore how PWB as captured at 12–18 month post-discharge is predicted by demographic characteristics (i.e. age and gender) and clinical characteristics (i.e., symptom severity and psychosocial functioning) as captured at intake and admission respectively. It was hypothesized that lower clinical scores would be associated with higher PWB.

Methods

This study is a part of a larger longitudinal, observational study on the psychosocial outcomes of children and youth who have accessed RT or intensive HBT in five agencies in Ontario, Canada. Three of these mental health agencies serve children aged 5–12 years upon admission and their families, and two of the mental health agencies serve adolescents aged 12–18 years upon admission and their families. In the original study, the research included administering standardized scales to caregivers and gleaning clinical information from youths’ files. Staff from each of the five agencies contacted families or caregivers of youth who had been discharged from RT or HBT within the past 12–18 months to ask if they would provide contact information to researchers to learn about the study. The contact information of families who were interested was given to researchers and researchers obtained informed consent from participants. If youth were not in the care of the family, a caseworker or foster parent was the respondent for the study. In the original study, approximately 75 % of the families who consented to agency staff to be contacted by research assistants participated in the study (N = 210), 10 % declined, and for the remainder the contact information became obsolete by the time research assistants attempted to make contact. The present study included a subsample of youth (N = 63) who accessed RT (n = 33) and HBT (n = 30) and were contacted 12–18 months post-discharge. Of the original 210 families, 63 youth were located by staff who made the initial contact with youth 12 years old or older to obtain consent to give contact information to researchers. All of the youth contacted by staff who agreed to speak with a research assistant participated in the interviews. Researchers then obtained informed consent and conducted a semi-structured interview with the youth. Youth’s clinical data were gleaned from agency files. Ethics approval was granted from Wilfred Laurier University and the five mental health agencies.

Youth who accessed RT resided in the treatment center 5 days a week and if possible joined their families on weekends, or remained at the treatment centre. Youth in both RT and HBT received similar strengths-based treatments which were individually tailored and consist of individual intervention and parental involvement where possible.

Measures

Clinical data were gleaned from agency files. The Brief Child and Family Phone Interview (3rd version; BCFPI) was used to capture youth’s symptom severity at intake and the Child and Adolescent Functional Assessment Scale (CAFAS; Hodges, 2000) was used to assess psychosocial functioning at admission. Both scales were mandated for use with all patients in Ontario and can be used for a variety of secondary functions such as research, service planning, and comparative analyses.

The BCFPI (Cunningham, Boyle, Hong, Pettingill, & Bohaychuk, 2009) is an interview tool to capture symptom severity relating to three externalizing disorders (i.e. Attention deficit hyperactivity disorder, oppositional defiant disorder, and conduct disorder) and three internalizing disorders (i.e. separating anxiety disorder, general mood, and self-harm) (Cunningham et al., 2009) at intake. The BCFPI has been found to demonstrate acceptable internal consistency reliability with Cronbach’s Alpha values being above 0.75 on all subscales with the exception of the conduct subscale which is 0.68 (Cook et al., 2013; Cunningham et al., 2009). Trained research assistants readministered the BCFPI at about 3 years post-discharge. T-scores are computed, and a score of 70 or higher is indicative of a clinical significance (i.e., a score that is higher than 98 % of the population).

The CAFAS (Hodges, 2000) is the gold standard tool to assess the degree of functional impairment in children and youth with emotional and behavioral disorder. Based on the responses from the youth and possibly others (e.g., caregivers), the clinical interviewer rated youth’s functioning on eight scales; School/Work, Home, Community, Behavior Towards Others, Moods/Emotions, Self-Harmful Behavior, Substance Use, and Thinking. The CAFAS has been found to demonstrate reliability and predictive validity (Hodges & Wong, 1996) and was the best option to measure functioning in youth with mental health disorders (Bates, 2001). The CAFAS can be used for both research and in clinical settings to assess clinical progress or outcome. Subscales scores can range from 0 (minimal or no impairment) to 30 (severe disruption or incapacitation). For the total scale a score of 40–60 is considered moderate impairment, 70–80 is considered marked impairment, and above 90 is considered severe impairment (Hodges, Wong, & Latessa, 1998)

Youths’ PWB was captured with The Personal Well-Being Index-School Children (PWI-SC; Cummins & Lau, 2005), which was used as an interview guide at 12–18 months post-discharge. The domains included school, work, learning, family, people you live with, where you are living, friends, doing things with people outside of your home, money/things you own, future, healthy, life as a whole. A sample question is “How happy do you feel about doing things outside your home?” Participants rated their happiness on each domain on a scale from zero (very dissatisfied) to 10 (very satisfied), and total scores were presented as a mean overall happiness score. Participants were invited to provide qualitative comments on each domain. A score between 60 and 90 is considered in the normal range. The PWI-SC has been found to demonstrate high inter-item reliability with Cronbach’s Alpha reported as 0.82 (Tomyn & Cummins, 2011).

Data Analysis and Sample Size

Demographic information is presented with descriptive statistics. An independent samples t test was conducted to explore the difference between overall PWB among youth who accessed RT and HBT and the assumption of homogeneity was met (p = 0.598). Since there were no differences and since many youth accessing HBT have symptoms at the same level of severity as youth accessing RT (Preyde et al., 2011a), though they have statistically better psychosocial functioning at admission than youth in RT, the two groups were combined to conduct further analyses. To explore the relationship between demographic and clinical characteristics and PWB, a hierarchical multiple regression was conducted. According to Cohen (1992), a sample size of at least 38 is required for a multiple regression analysis with four independent variables with a large effect size, a statistical of power of 0.80 and an alpha level at 0.05. The independent variables were participants’ age, sex, symptom severity (captured by the BCFPI at intake) and psychosocial functioning (captured by the CAFAS at admission). The dependent variable was the mean score of participant’s overall PWB as captured by the PWI-SC at 12–18 month post-discharge. Subsequently, a second hierarchical multiple regression was conducted to control for setting. The independent variables were participants’ type of setting (i.e., RT or HBT), symptom severity (captured by the BCFPI at intake) and psychosocial functioning (captured by the CAFAS at admission). The dependent variable was the mean score of participants’ overall PWB as captured by the PWI-SC at 12–18 month post-discharge. Given the non-statistically significant findings, a subsequent analysis (i.e. Pearson correlation) was performed to explore the relationship between PWB at 12–18 month post-discharge and symptom severity at 3 years post-discharge.

Results

The sample consisted of 63 youth of whom 33 (52.4 %) accessed HBT and 30 (47.5 %) accessed RT (Table 1). Their mean age was 14.61 (SD 1.82) and the majority (43; 68 %) were male. Many youth were attending school (53; 85.5 %) with most attending full time (51; 94.4 %). The mean sample score on personal well-being index was 74.90 (SD 13.90) which suggests that youth overall rated themselves as moderately high in personal wellbeing. Moreover, almost 89 % rated their overall happiness above a score of 60 (out of 100), that is, in the normal range. The mean total CAFAS score was 98.30 (SD 38.72), the mean total BCFPI score was 78.83 (SD 11.56), both of which were in the clinical range.

Table 1 Youth characteristics
Full size table

There was no statistically significant difference in the BCFPI total score between youth who accessed RT (mean 78.44, SD 10.23) and those who accessed HBT (mean 80.91, SD 12.12); t(50) = 0.775, p = 0.442. There was a statistical difference in the CAFAS total score between youth who accessed RT (mean 109.26, SD 31.98) and those who accessed HBT (mean 81.25, SD 33.77; t(57) = −3.252, p = 0.002). There was no statistically significant difference in total PWI scores between youth who accessed RT (mean 73.60, SD 13.20) and youth who accessed HBT (mean 76.20, SD 14.50); t(61) = 0.729, p = 0.469. Despite the difference in functioning, the entire sample was included in the subsequent analyses.

Youth reported high overall levels of PWB (see Table 2). The highest rated domain of personal well-being was “happiness with friends”, with a mean score of 88 (SD 13.8), followed by “happiness with doing things with people outside your home” with a mean score of 81 (SD 21.6). The lowest rated domain of personal well-being was “happiness with people with whom you live”, with a mean score of 55.7 (SD 21.7).

Table 2 Participant’s perceived happiness
Full size table

Some youth had missing clinical data, thus, only participants with no missing data were included in the multiple regression (n = 51). The required assumptions were checked. The assumption of collinearity was met (sex, tolerance = 0.996; age, tolerance = 0.993; BCFPI score, tolerance = 0.969; CAFAS score, tolerance = 0.975). The Durbin–Watson test indicated that the assumption of independent errors was met with a value of 2.058. A histogram and a P–P plot of standardized residuals was produced with SPSS and indicated that the data met the assumption of normally distributed errors. Additionally, a scatterplot of standardized residuals was produced on SPSS and indicated that the data met the assumptions of homogeneity of variance and linearity.

In order to control for age and sex, these variables were entered into the model in the first stage and symptom severity (BCFPI score) and psychosocial functioning (CAFAS score) were entered into the model in the second stage. The results of this analysis revealed that 6.4 % of the variance was explained by age and sex (R2 = 0.064, F(2,48) = 1.642, p = 0.204). With the addition of symptom severity (BCFPI score) and psychosocial functioning (CAFAS score) in the second model, there was an increase of 6.1 %, thus, 12.5 % of the variance was explained (R2 = 0.125, F(2,46) = 1.639, p = 0.181). Moreover, the analysis revealed that sex (β = 0.420, t(50) = 1.617, p = 0.113), age (β = −0.145, t(50) = −1.050, p = 0.299), symptom severity (β = 0.096, t(50) = 0.684, p = 0.497) and psychosocial functioning (β = −0.213, t(50) = −1.525, p = 0.134) did not significantly predict personal well-being scores (Table 3). Since there was no relationship between demographic and clinical scores captured prior to treatment and youths’ perception of well-being at 12–18 month follow-up, subsequent analyses were conducted.

Table 3 Do demographic and clinical characteristics predict perceived well-being scores?
Full size table

A second hierarchical multiple regression was conducted and the type of setting (i.e., RT or HBT) was controlled. The assumption of collinearity was met (treatment, tolerance = 0.851; BCFPI score, tolerance = 0.970; CAFAS score, tolerance = 0.846). The Durbin–Watson test indicated that the assumption of independent errors was met with a value of 2.235. A histogram and a P–P plot of standardized residuals was produced with SPSS and indicated that the data met the assumption of normally distributed errors. Additionally, a scatterplot of standardized residuals was produced on SPSS and indicated that the data met the assumptions of homogeneity of variance and linearity.

Type of setting (RT and HBT) was entered into the model in the first stage and symptom severity (BCFPI score) and psychosocial functioning (CAFAS score) were entered into the model in the second stage. The results of this analysis revealed that 0 % of the variance was explained by treatment setting (R2 = 0.000, F(1,49) = 0.23, p = 0.879). With the addition of symptom severity (BCFPI score) and psychosocial functioning (CAFAS score) in the second model, there was an increase of 6.1 %, thus, 6.1 % of the variance was explained (R2 = 0.061, F(3,47) = 1.009, p = 0.397) by type of treatment and clinical characteristics. Moreover, the analysis revealed that treatment setting (β = 0.079, t(50) = 0.513, p = 0.611), symptom severity (β = −0.230, t(50) = −1.495, p = 0.142) and psychosocial functioning (β = −0.101, t(50) = 0.707, p = 0.483) did not significantly predict personal well-being scores (Table 4).

Table 4 Do type of treatment and clinical characteristics predict perceived well-being scores?
Full size table

Next, a Pearson correlation coefficient was used to explore the relationship between PWB and BCFPI scores captured at 3 years post-discharge. There was no statistically significant relationship between PWB scores (mean 74.90, SD 13.90) and youth’s symptom severity score at 3 year post discharge (mean 70.04, SD 11.55), r(63) = −0.051, p = 0.721.

Discussion

The primary finding of this study was the high proportion of youth (88.9 %) who indicated that they were experiencing happiness in their lives as many reported an overall personal well-being score of 60 or above (M 74.90, SD 13.90). Just for comparison, this mean score is similar to that of normative adolescent samples (Proctor, Linley, & Maltby, 2009). For example, Tomyn and Cummins (2011) reported a mean of 73.61 (SD 14.18). Despite the symptoms of mental health disorders and the often adverse experiences of youth who have accessed intensive mental health treatment, it is possible that their happiness levels may not have been negatively impacted or rather there is some mechanism that preserves a personal sense of well-being for most youth despite these challenges. The non-significant findings might be explained with Cummins theory (HMP): personal well-being appears to have an equilibrium that is set at a rather high positive state and maintained through psychological mechanisms despite the presence of emotional/behavioral disorder and adversity.

This finding supports previous research which suggests that positive emotions can, in fact, co-exist with mental health challenges (Diener & Diener, 1995). In addition, these findings are consistent with the notion that positive mental health and mental health disorders should not be conceived as two ends of one spectrum, as they may be considered separate entities (Diener, 2009; Hatch, Harvey, & Maughan, 2010; Keyes, 2002).

High PWB may be conceived as an indicator of resilience in this sample of youth with multiple challenges. Various benefits associated with high PWB of healthy populations include better health and longevity (Diener & Chan, 2011), relationships with others, increased success in work, higher incomes (Lyubomirsky, King, & Diener, 2005), and stronger immune systems (Howell, Kern, & Lyubomirsky, 2007) compared to less happy counterparts. Additionally, high PWB has been found to ameliorate negative outcomes, including psychological disorders among youth (Park, 2004). Thus, youth with negative life experiences, which have been typically featured among youth who have accessed residential and intensive mental health treatment, can in fact experience happiness despite stressful life circumstances and the experience of a mental health disorder. However, preliminary research (Cameron, Frensch, Preyde, & Quosai, 2011; Frensch et al., 2009; Preyde et al., 2011b) suggests that these youth do not seem to experience some of the benefits that accompany high PWB such as better health and relationships, and occupational success.

Another notable finding from the present study is the similarity of happiness levels between youth who accessed RT compared to youth who accessed HBT. Previously, investigators have indicated that very different populations access RT in comparison to those who access HBT in terms of psychosocial functioning and living arrangements (Preyde et al., 2011a). However, the results of this study indicate that happiness levels between both groups were not dissimilar with the exception of “happiness with where you are living”. This discrepancy between groups may be attributed to the fact that the youth who were accessing HBT may have somewhat greater stability in their home environment in comparison to youth who accessed RT. Additionally, youths’ total CAFAS scores were clinically and statistically different, yet their overall mean happiness levels remained similar. It should also be noted that individual happiness levels have been shown to be stable over time and changes in happiness levels are minimally affected and if so, only temporarily (Lyubomirsky et al., 2005). Genes were reportedly found to account for 80 % of individual happiness levels (Nes et al., 2006) and numerous researchers have indicated that personality largely contributes to happiness levels (DeNeve & Cooper, 1998; Pavot, Diener, & Fujita, 1990; Steel, Schmidt, & Shultz, 2008). Nonetheless, despite the differences between psychosocial functioning and living arrangements in baseline characteristics between both samples, happiness levels may be largely contingent upon individual personalities. Notwithstanding, some investigators have suggested that living conditions may have influences on happiness levels indicating that personality may contribute to happiness but there may be multiple influences on one’s level of happiness (Veenhoven, 1994). Therefore, happiness levels may be conceived, similar to psychopathology, as being largely influenced by the complex interaction among numerous factors, including genetic, personal and environmental contributions.

Clinical Characteristics and Personal Well-Being

In this exploration, it was found that demographic, treatment type, and clinical characteristics did not predict the participants’ personal well-being scores, suggesting that no clear relationship exists between PWB and symptom severity and psychosocial functioning. These findings again support the idea that PWB may be maintained by protective cognitive mechanisms (Cummins) and that positive well-being and mental health disorders should be considered as distinct entities (Huppert & Whittington, 2003; Keyes, 2005). These findings somewhat contrast those of Suldo & Hueber (2004), who did in fact find a relationship where low life satisfaction predicted the occurrence of externalizing problems and high life satisfaction acted as a buffer against the perception of stressful life events among 816 middle and high school students. Likewise, Hatch, Harvey & Maughan (2010) found that oppositional or aggressive behavior in childhood, police contact and truancy in adolescence, and negative social environments (i.e. poor material conditions, poor relationship with parents, contact with services, parental divorce or separation at age seven, and poor relationships with siblings) increased the likelihood of poor PWB and mental health disorders later in life. Although no clear relationship was found between clinical characteristics predicting PWB in the present study, non-significance can be viewed as encouraging since one might conclude that individuals who have accessed intensive mental health treatment for their mental health disorders can experience happiness in many domains of their life. Thus, this finding supports the notions that protective mechanisms may maintain PWB, and that PWB and mental health disorders demonstrated a degree of independence from each other with this sample.

Strengths and Limitations

A major strength of this study was the examination of both clinical characteristics and perceived personal well-being among an adolescent population who accessed intensive mental health treatment. As mentioned, positive factors have often been overlooked in clinical samples of youth with emotional and behavioral disorders. Since mental health disorders can have adverse consequences, this exploration contributed to the understanding of the characteristics of youth who access intensive mental health treatment with regard to PWB.

Limitations of this study should also be acknowledged. First, this population is one of the most vulnerable; they are characterized as having experienced multiple adversities from a young age including a lack of stability which makes locating them for the purposes of research challenging. This difficulty in reaching ‘hard-to-reach” people was reflected in the small sample size; however, the small sample size should not detract too much from the exploratory nature of this article. Moreover, the method of recruitment posed challenges in obtaining follow-up data as agency staff initially contacted participants to gain permission to provide researchers with youths’ contact information. However, when the researchers received the contact information some of this information was no longer valid. Likewise, only youth who were interested in participating were included in this study; hence, the results may not generalize to all youth accessing intensive RT and HBT. Findings of this study may not be generalizable to youth accessing other mental health systems such as in-patient psychiatry, and there was a relatively small sample size and all of the youth were living in Ontario, Canada. Another limitation was that the main construct for this report, PWB, was captured at only at one time point.

Implications for Research

In the future, investigators may consider exploring PWB over an extended period of time in addition to mental health disorders to track fluctuations that may coincide or track trends. Investigators have indicated that happiness or high PWB has been shown to correlate with positive outcomes in many life domains, mainly work, relationships and health (Lyubomirsky et al., 2005). Youth accessing these intensive mental health treatment struggle considerably in these three life domains. Therefore, in the future, investigators might also explore the relationship between PWB and other variables, such as stability in housing and school or employment outlook and quality of social networks. Investigators might also explore the timing of adversity in relation to PWB since many of these youths were born into a life of adversity; that is, they had no or limited other experiences. It should also be noted that youth accessing these intensive mental health treatments (RT or HBT) are dealing with mental health needs that emerged over time and are relatively long-standing. Would PWB be the same for youth who experienced years of optimal developmental conditions followed by adversity and mental health disorder as for youth who had no other experiences?

Implications for Practice

When connecting these findings to practical terms, it is recognized that mental health disorders are often chronic and lifelong conditions. These findings allow us to recognize that these vulnerable youth can identify and report happiness in their lives, despite the significant challenges they experience including their mental health disorder. Strategic interventions may foster increased happiness levels for individual human beings (Norrish & Vella-Brodrick, 2008; Sin & Lyubomirsky, 2009), including those experiencing mental health disorders (Boiler et al., 2013; Fava et al., 2005; Lyubomirsky & Layous, 2013). Many youth may enter RT or HBT with unrealistically elevated or depressed sense of self and well-being and cognitive behavioral interventions are used to challenge distortions. Positive feelings, including happiness, can have enduring positive consequences whereby individuals tend to implicitly think and behave in a positive manner. By doing so, it is believed that strong physical and psychological resources will be created that can be accessed throughout the lives of individuals (Fredrickson, 2001). Thus, in the future investigators and clinicians may want to explore longitudinally PWB throughout admission, intervention and post-discharge phases of care which may have implications for treatment. High PWB may acts as a resource for youth that can positively influence their daily life.

Similarly, with the resilience framework individuals are viewed as active agents who can influence their life circumstances through effective use of resources, despite having experienced adversity (Luthar & Cicchetti, 2000). A resiliency focus in treatment that fosters a sense of hope and optimism in the face of difficult circumstances and a positive life narrative may be good areas upon which to build with individuals in addition to other psychosocial treatments, skill development and strategies to enhance social functioning. Positive interpretation and adjustment to life events may be equally as important as the occurrence of positive life events (MacLeod & Moore, 2000). From a resilience framework, mental health disorders may act as the vulnerability factor, which may adversely influence health outcomes. Whereas, personal well-being interventions may act as a protective factor since they may have the potential to modify the effects of risk in a positive direction. Likewise, interventions that are designed to increase personal well-being may be executed in an economically feasible manner and adopted by various professionals within the broad field of mental health service delivery (Cloninger, 2006; Shaw & Taplin, 2007) with a hope for lasting benefits for youth, specifically those who experiencing mental health disorders.

Conclusion

Youth accessing RT or HBT may experience numerous stressors and have experienced adversity in their lives. However, in this study, promising results were revealed; that is, these youth indicated high levels of perceived personal well-being. This finding supports Cummins’ theory of HPM and suggests that these youth possess a cognitive mechanism that protects their sense of well-being. Moreover, the promotion of mental health by means of strengthening and enhancing personal wellbeing of children and youth is perhaps equally important as treating mental health disorders (Peter, Roberts, & Dengate, 2011). Overall, the findings from this study support the notion that PWB and psychopathology have a degree of independence from each other.