The relationship between resilience and levels of anxiety, depression, and obsessive–compulsive symptoms in adolescents
Abstract
Objective: Mental health problems affect approximately 20% of adolescents. Traditionally, the principal focus has been on vulnerability and risk factors and less on protective factors. The study, therefore, explores the relation between frequent psychiatric symptoms and resilience factors among older adolescents.
Method: The Resilience Scale for Adolescents (READ) was completed by 307 Norwegian high school students (M = 16.4 years) along with the Depression Anxiety Stress Scales, and the Obsessive–Compulsive Inventory-Revised.
Results: Higher resilience scores predicted lower scores on levels of depression, anxiety, stress and obsessive–compulsive symptoms after controlling for age and gender.
Conclusion: This study provides further evidence that it may be fruitful for clinicians and researchers to attend to resilience factors in relation to psychological symptoms among older adolescents. Copyright © 2010 John Wiley & Sons, Ltd.
Key Practitioner Message:
• Resilience is significantly related to psychological symptoms in older adolescents after controlling for age and gender.
• Resilience factors are differently expressed by female and male youths, but overall resilience is equally distributed among the sexes.
• Assessment of resilience factors may provide appropriate targets for interventions among youths.
INTRODUCTION
Clinical research has until recently years focused mainly on risk factors and vulnerability in understanding psychopathology, e.g., increases in the severity and numbers of stressful events increases the probability for developing mental health problems (Egeland & Sroufe, 1993; Hammen, 2003; Heatherington & Elmore, 2003; Zucker, Wong, Puttler, & Fitzgerald, 2003). This focus has not succeeded in explaining all individual variation in mental health in the face of adversity.
A different research approach, developmental psychopathology, strives to engage in a comprehensive evaluation of biological, psychological, social and cultural processes. It attempts to ascertain how the interaction among these multiple levels of analysis may influence individual differences. And thus, the continuity or discontinuity of adaptive or maladaptive behaviour patterns, and the pathways by which normal and pathological development outcomes, may also be achieved (Cicchetti & Dawson, 2002; Cicchetti & Sroufe, 2000). Risk and protective factors have been established at multiple levels of analysis and in multiple domains (Cicchetti, 2006). Originating from this research tradition, the results from resilience research has indicated that the presence of protective factors may moderate the effects of stressors and actually contribute to promoting mental health and prevent the development of psychopathology despite being exposed to significant stressors (Luthar, Cicchetti, & Becker, 2000; Masten & Reed, 2002; Rutter, 2000). Some large prospective longitudinal studies, like the Kauai-study, have shown that a significant minority of high-risk children adapt well and show healthy mental functioning despite experiencing high levels of stress (e.g., Werner & Smith, 1982; 1992; 2001). Even though resilience research originates from the research field of developmental psychopathology, it does not have one established theory, but is rather a collection of empirical findings that have identified numerous protective factors measured with very diverse measures. Among leading researchers, there is a consensus on three overarching categories of protective factors: (1) Personal dispositions, (2) Family cohesion; and (3) Social resources outside the family (e.g., Garmezy, 1993; Werner, 1989, 1993). There are some general gender differences found in larger samples, where females report more social and interpersonal recourses compared with males, and males rate themselves higher on personal dispositions than females (e.g., Werner & Smith, 2001). These findings are found in both adolescent and adult samples (e.g., Hjemdal, Friborg, Stiles, Martinussen, & Rosenvinge, 2006a; Hjemdal, Friborg, Stiles, Rosenvinge, & Martinussen, 2006b).
Mental disorders are likely to be caused by multiple risk factors and processes rather than singular causes (Cicchetti & Sroufe, 2000). Thus, the identification of a causal risk factor will contribute to elucidating only one aspect of a more complex matrix of causes. Within individuals, there are likely to be multiple component processes rather than unitary causes that contribute to psychopathological outcomes (Cicchetti & Blender, 2004). Thus, exploring the relation between resilience and levels of psychopathology promises to add a synergic contribution to the more traditional approach of exploring the relation between psychiatric symptoms and psychopathology. Studies exploring such relations in youth are particularly important as many youths first develop depressive and anxiety symptoms in high school, and early debut is often associated with later adult psychopathology (e.g., Fergusson & Woodward, 2002; Lecrubier, 1998; Patton, Coffey, Posterino, Carlin, & Bowes, 2003; Stein et al., 2001). The development of emotional disorders has been linked to a stress-diathesis hypothesis, and may be particularly interesting as resilience research is also founded on the relation between stress and psychopathology (Compas et al., 1987; Goodyer, 1996; Pine et al., 2002).
Obsessive–compulsive disorder (OCD) has sometimes been seen as somewhat different from other emotional disorders (Tyrer, 1989), perhaps because in up to 40% of cases there is no clear-cut antecedent except for the presence of extremely high standards imposed in childhood (Steketee & Barlow, 2002). As OCD seems somewhat unique among the anxiety disorders because it has been considered as less dependent upon antecedent stress, it is very interesting to investigate the relationship between OCD and resilience.
Depressive and anxiety symptoms are related to internalizing problems. A recent study indicated that a measure of adolescent resilience had higher correlations for internalizing than for externalizing problems (von Soest, Mossinge, Stefansen, & Hjemdal, 2010). This may indicate further reasons to explore the relation between internalization problems and resilience factors.
As other studies with younger adolescents and with adults have demonstrated the interrelationship of resilience with psychiatric symptoms, and therefore, we predicted that resilience in older adolescents would be related to levels of depression, anxiety, stress, and obsessive–compulsive symptoms. In addition, the gender differences found in previous studies were also expected to be present.
METHOD
Participants
High school pupils from one school in a large Norwegian city were invited to fill out the measurements of this study. Three hundred and fifty-three pupils, from the first two years of the three-year high school, received the questionnaires and 307 returned completed questionnaires. Approximately 10% of the students were not present in school on that day and the other 10% did not fill out the questionnaires or filled them out incompletely or in a frivolous manner. There is no information available on the characteristics of those not completing questionnaires. Missing values for a few cases were substituted with subscale averages rounded to whole numbers. If more than 33% of a subscale's items were missing for a particular participant, then the individual was excluded from further analyses. The final sample included 167 males (54%) and 140 females (46%). The average age was 16.4 years (SD = 0.7, range 14–18). This school had pupils that in general can be considered typical for Norwegian high school youths. It did not have programs for the intellectually elite nor have high concentrations of culturally or intellectually challenged youths. The school had separate classes devoted to sports, electrical engineering and music.
Procedure
The pupils were assembled in one of two consecutive obligatory lectures in a large auditorium. Questionnaires were passed out by psychology graduate students prior to a lecture by the second author on mental health problems in youth. Informed consent was obtained at the start of the lecture by the pupils being informed that completion of the questionnaires was voluntary and anonymous. This information was also written on the first page of the questionnaire packet. They were also told that completion of the questionnaires would make the following lecture more personally informative for them. Those feeling a need for more information about their own personal situation were encouraged to contact the school nurse afterwards. Twenty minutes was devoted to the completion of the questionnaires.
Measurements
Resilience Scale for Adolescents (READ)
The READ is a 28-item self-report scale using a five-point Likert scale, with all items positively phrased. The READ is based on the Resilience Scale for Adults (RSA; Friborg, Hjemdal, Rosenvinge & Martinussen, 2003; Hjemdal, Friborg, Martinussen, & Rosenvinge, 2001), which has been shown to possess adequate psychometric qualities (Friborg & Hjemdal, 2004), has established construct validity (Friborg et al., 2005) and predictive validity in a prospective study (Hjemdal et al., 2006b) and in an experimental study (Friborg et al., 2006). Higher scores on the READ reflect higher degrees of resilience. This scale has shown adequate psychometric properties and initial promising validity with younger youths (Hjemdal et al., 2006a). A correlational study using a young adolescent sample (from 13 to 15 years of age) indicated predictive validity for READ. All of its factors significantly and negatively predicted levels of depressive and social anxiety symptoms, when controlling for age, gender and number of negative life events (Hjemdal, Aune, Reinfjell, Stiles, & Friborg, 2007). This may indicate that the protective factors measured with READ are relevant to the prediction of anxiety and depressive symptoms. The READ has essentially the same factor structure as the RSA (Friborg, Barlaug, Martinussen, Rosenvinge, & Hjemdal, 2005; Hjemdal et al., 2006b).
It consists of five subscales: (1) personal competence, (2) social competence, and (3) structured style, as well as subscales for (4) family cohesion; and (5) social resources. These five subscales cover the previously reported three overarching categories for resilience (1) personal dispositions, (2) family cohesion; and (3) social support outside the family.
Previous studies have found significant gender differences, with boys generally reported higher scores on personal competence, while girls reported higher scores on social resources. Similar gender differences have previously been reported in research on resilience (Feingold, 1994; Werner, 1989).
In the current study, READ showed adequate psychometric characteristics with a total Cronbach's alpha of 0.91 and Cronbach's alpha on the five subscales of 0.75, 0.80, 0.61, 0.67, and 0.86, respectively. These findings were similar to those obtained in previous studies with this measure on younger adolescents.
Depression Anxiety Stress Scales (DASS-21)
Both the original 42-item version of the Depression Anxiety Stress Scales (DASS; Lovibond & Lovibond, 1995) and the short form DASS-21 (Antony, Bieling, Cox, Enns, & Swinson, 1998; Henry & Crawford, 2005) have adequate levels of reliability and convergent validity. The short form DASS-21 seems to be an adequate replacement for the full scale version (Antony et al., 1998; Clara, Cox., & Enns, 2001). A shorter version should potentially be more acceptable as it limits the burden on individuals. This may be of particular interest with regards to depressed and anxious participants as well as participants with variable motivation or limited levels of concentration such as youths. According to the DASS website (Lovibond, web page), the downward age limit for the DASS-21 scales is 14 years of age.
The DASS-21 consists of three subscales with seven self-report items in each assessing depression, anxiety and stress (Lovibond & Lovibond, 1995). A four-point severity scale measures the degree to which each symptom description has been experienced over the past week from 0 (not at all) to 3 (very much). In the current study, the Cronbach's alphas for the total score was 0.88, for Stress 0.77, for Anxiety 0.67, and for Depression 0.82. These levels of internal consistency were considered adequate psychometrically.
Obsessive–Compulsive Inventory–Revised (OCI-R)
The 18-items of the OCI-R measures distress by OCD symptoms in the past month (Foa et al., 2002). The OCI-R is based on the 42-item OCI (Foa et al., 1998). A five point Likert scale from 0 (not at all) to 4 (extremely) is used. Six subscales are assessed: (1) washing, (2) checking, (3) obsessions, (4) neutralizing, (5) ordering; and (6) hoarding. Each subscale includes three items. The OCI-R has shown adequate psychometric characteristics with a college-student population in the USA (Hajcak, Huppert, Simons, & Foa, 2004), in a Spanish college sample (Fullana et al., 2005), and was validated in a recent Norwegian study (Solem, Hjemdal, Vogel, & Stiles, in press). The Cronbach's alpha for the OCI-R score was 0.88 in the present study which is considered adequate.
Data Analyses
Coefficients of correlation were applied to investigate inter-relationships among resilience, anxiety, depression, stress and obsessive–compulsive symptoms. Since this is the first time that many of these measures have been employed with this age group, the significance level was set at p < 0.05, two-tailed for exploratory purposes. Finally, separate multiple hierarchical linear regression analyses were performed to assess the degree to which resilience predict variance in symptoms of depression, anxiety, stress and obsessive–compulsive symptoms when controlling for age and gender. To avoid multiple significance effects and in accordance with Bonferroni guidelines, a significance level of p < 0.0125, two-tailed, was employed for these regression analyses.
RESULTS
Levels of Resilience and Psychiatric Symptoms
Table 1 presents the number of items in each scale or factors, scale or factor means (M) and standard deviations (SD) for the measures of resilience, depression, anxiety, stress and obsessive–compulsive symptoms.
| Subscale | Number of items | M | SD |
|---|---|---|---|
| READ total | 28 | 3.87 | 0.52 |
| Personal competence | 8 | 3.81 | 0.58 |
| Social competence | 5 | 3.95 | 0.70 |
| Structured style | 4 | 3.29 | 0.70 |
| Social resources | 5 | 4.36 | 0.64 |
| Family cohesion | 6 | 3.88 | 0.79 |
| DASS-21 total | 21 | 0.38 | 0.38 |
| Depression | 7 | 0.36 | 0.48 |
| Anxiety | 7 | 0.33 | 0.39 |
| Stress | 7 | 0.46 | 0.47 |
| OCI-R total | 18 | 0.57 | 0.52 |
The mean for DASS-21 total scores was approximately the same mean as found in a large adult normal UK population (Henry & Crawford, 2005). Finally, the results obtained from the OCI-R was similar to that of a recent study (Hajcak et al., 2004) with introductory psychology students (OCI-R M = 0.66, SD = 0.52).
Gender Differences for Resilience and Psychiatric Symptoms
Table 2 presents the results for gender differences for resilience and psychiatric symptoms was explored using a t-test. Boys reported significantly higher scores on Personal Competence than girls, while girls reported significantly higher score on Social Resources, Family Cohesion, total DASS-21 scores and on the levels of depressive symptoms than boys.
| Subscale | Boys mean (SD) | Girls mean (SD) | t |
|---|---|---|---|
| READ total | 3.86 (0.51) | 3.90 (0.53) | −0.79 |
| Personal competence | 3.91 (0.53) | 3.69 (0.62) | 3.43*** |
| Social competence | 3.92 (0.75) | 3.99 (0.70) | −0.79 |
| Structured style | 3.47 (0.68) | 3.43 (0.67) | 0.55 |
| Social resources | 4.25 (0.61) | 4.36 (0.64) | −3.40*** |
| Family cohesion | 3.38 (0.77) | 3.99 (0.72) | −2.25* |
| DASS-21 total | 7.27 (7.05) | 9.28 (8.95) | −2.23* |
| Depression | 2.58 (2.87) | 3.90 (3.69) | −2.82** |
| Anxiety | 2.08 (2.32) | 2.59 (3.39) | −1.60 |
| Stress | 2.33 (3.37) | 2.79 (2.36) | −1.24 |
| OCI-R Total | 9.52 (9.68) | 11.64 (9.96) | −1.92 |
- * * p < 0.05,
- ** ** p < 0.01,
- *** *** p < 0.001.
Correlations between Resilience and Psychiatric Symptoms
Table 3 contains correlations among gender, resilience, depression, anxiety, stress and obsessive–compulsive symptoms.
| Measurements | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1. Female gender | – | ||||||||||
| 2. READ total | 0.05 | – | |||||||||
| 3. Pers. comp. | −0.19*** | 0.84*** | – | ||||||||
| 4. Soc. comp. | 0.05 | 0.73*** | 0.59*** | – | |||||||
| 5. Struc. style | 0.06 | 0.70*** | 0.57*** | 0.35*** | – | ||||||
| 6. Soc. res. | 0.19*** | 0.74*** | 0.46*** | 0.46*** | 0.34*** | – | |||||
| 7. Fam. coh. | 0.15** | 0.82*** | 0.53*** | 0.41** | 0.53*** | 0.60*** | – | ||||
| 8. Depression | 0.12* | −0.39*** | −0.40*** | −0.20*** | −0.18*** | −0.35*** | −0.32*** | – | |||
| 9. Anxiety | 0.07 | −0.34*** | −0.33*** | −0.16** | −0.16** | −0.31*** | −0.30*** | 0.49*** | – | ||
| 10. Stress | 0.16** | −0.29*** | −0.32*** | −0.06 | −0.18*** | −0.22*** | −0.28*** | 0.54*** | 0.74*** | – | |
| 11. OCI-R | 0.15** | −0.29*** | −0.25*** | −0.18*** | –0.13* | −0.28*** | −0.25*** | 0.37*** | 0.46*** | 0.49*** | – |
- * Two-tailed * p < 0.05.
- ** ** p < 0.01.
- *** *** p < 0.001.
- Spearmans rho utilized for correlations with gender.
Age was not significantly related to any of the measures and is not reported. Small, but significant positive correlations were found for female gender with obsessive–compulsive symptoms, depression, stress, Social Resources and Family Cohesion. Personal Competence was significantly negatively related to female gender.
Resilience was significantly negatively related to obsessive–compulsive symptoms and psychiatric symptoms as measured with the DASS-21 subscales. The sizes of the correlations were small to moderate. The only exception was a non-significant correlation between Social Competence and Stress. The OCI-R score showed moderate significant positive correlations with DASS-21 scores.
Regression Analyses
Table 4 shows the results for the cross-sectional exploration of resilience as a predictor, in separate multiple hierarchical linear regressions, of depression, anxiety, stress, and obsessive–compulsive symptoms. In the first analysis, age was entered on step one, gender on step two, and the total resilience score was entered on step three. Age and gender were not significantly predictive of levels of psychiatric symptoms among these youths. The only exception was that female gender was significantly related to higher depression scores. However, the results indicated that resilience contributed additional significant variance in the prediction of depression, anxiety, stress and obsessive–compulsive symptoms. Identical separate hierarchical multiple regression analyses were undertaken only substituting the total resilience score with the subscale scores. The results indicated that all subscales were also significantly predictive of most symptoms. However, Social Competence was not predictive of depressive symptoms and Structured Style did not significantly predict obsessive–compulsive symptoms.
| Step | DASS-21 depression | DASS-21 anxiety | DASS-21 stress | OCI-R Total | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| F cha | R2 cha | β | t | F cha | R2 cha | β | t | F cha | R2 cha | β | t | F cha | R2 cha | β | t | ||
| 1 | Age | 2.39 | 0.01 | 0.09 | 1.55 | 0.96 | 0.00 | 0.06 | 0.98 | 0.07 | 0.00 | 0.02 | 0.26 | 0.29 | 0.00 | −0.03 | −0.54 |
| 2 | Gender | 5.92* | 0.02 | 0.13 | 2.43 | 0.91 | 0.00 | 0.05 | 0.95 | 2.39 | 0.01 | 0.09 | 1.55 | 0.22 | 0.00 | −0.03 | 0.47 |
| 3 | READ T | 30.09* | 0.08 | −0.29 | −5.49 | 45.81* | 0.12 | −0.35 | −6.77 | 54.20* | 0.14 | −0.38 | −7.36 | 21.45* | 0.06 | −0.25 | −4.63 |
| 3 | Pers C | 36.20* | 0.10 | −0.32 | −6.02 | 41.72* | 0.11 | −0.34 | −6.46 | 57.21* | 0.15 | −0.39 | −7.56 | 16.17* | 0.05 | −0.22 | −4.02 |
| 3 | SoC | 2.25 | 0.01 | −0.08 | −1.50 | 10.10* | 0.03 | −0.18 | −3.18 | 13.75* | 0.04 | −0.20 | −3.71 | 9.71* | 0.03 | −0.17 | −3.12 |
| 3 | StSy | 12.56* | 0.04 | −0.19 | −3.54 | 14.21* | 0.04 | −0.21 | −3.77 | 14.66* | 0.04 | −0.21 | −3.83 | 1.95 | 0.01 | −0.08 | −1.40 |
| 3 | FaC | 26.68* | 0.08 | −0.27 | −5.17 | 33.02* | 0.09 | −0.31 | −5.75 | 34.19* | 0.10 | −0.31 | −5.85 | 14.73* | 0.05 | −0.21 | −3.84 |
| 3 | SoRe | 15.22* | 0.05 | −0.21 | −3.90 | 36.89* | 0.10 | −0.32 | −6.07 | 42.35* | 0.12 | −0.34 | −6.51 | 19.83* | 0.06 | −0.24 | −4.45 |
- * * p < 0.015. Two-tailed.
- READ T = READ total; Pers C = personal competence; SoC = social competence; StSy = structured style; SoRe = social resources; FaC = family cohesion.
DISCUSSION
The results from the present study illustrate how a measure of resilience may facilitate understanding of the relation between level of psychopathological symptoms and protective factors. Using a resilience measure in the exploration of psychopathology may contribute to a fuller picture of stress and psychopathology in line with a developmental psychopathology approach. Higher resilience scores on personal dispositions, the availability of sources of social support outside the family, as well as perceived levels of family cohesion showed consistent and important significant relationships to levels of depression, anxiety, stress and obsessive–compulsive symptoms when controlling for age and gender differences. The presence of these protective factors is associated with lower levels of stress and psychiatric symptoms.
The gender differences for depressive symptoms found in the present study were expected and were consistent with previous studies (e.g., Wichstrøm, 1999). For the total resilience score, there were no gender differences, but there were, however, gender differences found for three of the subscales of resilience. Female pupils reported significantly lower scores on Personal Competence, but higher scores on Social Resources and Family Cohesion, than males. Similar gender differences in resilience have previously been reported in adolescents (Hjemdal, Friborg, Stiles, Martinussen, & Rosenvinge, 2006a) and in adults (Friborg et al., 2003). These cross-sectional findings are supported by the longitudinal findings of the Kauai study which followed children to adulthood over 32 years (Werner, 1989). Since overall levels of resilience were unrelated to gender differences, the differences obtained may be an indication of gender-related differences of characteristic styles for good adaptation. These findings bear similarities with research findings from psychological research that has explored gender differences in depressive symptoms. Nolen-Hoeksema, Larson and Grayson (1999) found that females carry a triad of vulnerabilities to depressive symptoms compared with males. They are exposed to more chronic stress from their social environment, greater tendency to ruminate when distressed and have lower sense of mastery in their lives. These findings are particularly interesting in view of resilience and gender differences identified here and previously. The READ measures aspects of mastery and ability to abandon rumination with the items included Personal Competence subscale where boys report higher scores than girls. In addition the two interpersonal READ subscales that measure social resources were significantly associated with the girls' mental health, and thus may serve as indirect measures of social stress. This may indicate that if girls have access to healthy social environments, they perceive them as a greater resource than boys, while boys seem to depend on intrapersonal resources to a larger extent. Thus, generally girls that do not have healthy social environments may be more vulnerable to developing mental health symptoms than boys, and boys with low levels of intrapersonal resources may be more vulnerable than girls.
The Cronbach's alpha for the total score and three of the five factors were acceptable, the alphas' for Structured Style and Social Resources were in the lower range of acceptable, and further studies are warranted to explore these two factors in similar samples. The mean levels of resilience obtained with high school adolescents in the present study, were very similar to that obtained in previous studies of Norwegian junior high school adolescents (Hjemdal et al., 2006a). This extends previous findings with younger adolescent samples that found moderate to strong correlations between resilience and measures of depressive symptoms, and small to moderate correlations with measures of stressful life events and social anxiety disorder symptoms, respectively (Hjemdal et al., 2006a, 2007). These findings are now extended and indicate that there is an interesting covariation between levels of protection and more general levels of depressive and anxiety symptoms as well as obsessive–compulsive symptoms.
The correlations between resilience and the psychiatric symptoms in this study were small to moderate. The amount of variance explained in these symptoms was small 0–16%, but similar to those obtained with younger adolescents (Hjemdal, et al., 2007). Total resilience scores accounted for between 9 and 16% of the variance in symptom scores in the regression analyses. In the first prospective study done with college youths (Hjemdal, Friborg, Stiles, Rosenvinge, & Martinussen, 2006b), two resilience factors were significantly predictive of changes in psychiatric symptoms in interaction with life stressors after a 3-month interval. Two aspects of personal competence in adults (Planned Future and Social Competence) explained 7 and 9% of the variance, respectively. It remains to be seen if similar results can be obtained with high school youths. Even though there is a large proportion of the variance in the depressive, anxiety, OCD symptoms and stress that is not explained by READ, one could still argue that the explained variance indicate that measuring protective factors and resilience is important and may add a significant contribution to future studies exploring the relation between mental health and psychiatric symptoms.
The results of the regression analyses revealed that in general, levels of resilience in high school youths are significantly related to levels of depression, anxiety, stress and obsessive–compulsive symptoms. This is also true when controlling for age and gender. However, social competence was not significantly related to symptoms of levels of depressive symptoms. This may be partially explained by the fact that social competence and depressive symptoms were both significantly related to female gender in this sample. Furthermore, the personal disposition associated with having a structured style was not significantly related to the levels of obsessive–compulsive symptoms. In other words, being structured and systematic can characterize individuals with all levels of obsessive–compulsive symptoms. However, the interpretation of these findings is uncertain and both require further study.
This study has one primary limitation related to the administration mode, as questionnaires were filled out in a large auditorium, which may limit the degree of self-disclosure. Furthermore, there was no prospective data collected with this sample. There is no necessary causal relation between the resilience factors and the psychiatric symptoms measured since they are measured at the same time and only are an indication of intercorrelation. However, these factors have been corroborated by previous longitudinal studies like the Kauai study (Werner, 1989), and they showed protective value for the development of psychiatric symptoms in a short-term prospective study with young adults using a related measure of resilience (Hjemdal, Friborg, Stiles, Rosenvinge, et al., 2006b).
In sum, the findings of this study suggest that resilience is significantly related to symptoms of internalizing psychiatric disorders in a high school sample. Taken together with evidence from previous research, this suggests that clinicians and healthcare workers should consider assessing resilience factors in older youths along with mental health problems like depression, anxiety, stress, obsessions and compulsions, and that increasing particular resources of protection may be a way of enhancing mental health or increase stress resistance. Beck (2005, p. 21) indicated that patients who basically function effectively but feel unlovable should take steps to connect to other people. Equally, patients that are helpless and inefficient but likable need to engage in mastery experiences. Based on Beck's statements, an interesting clinical and possibly prevention intervention that may be subject to further research would be to see if resilience interventions should target resilience factors that generally found to be gender-related strengths. Would strengthening intrapersonal resources for boys and interpersonal resources for girls give better results than an inverse intervention? Perhaps the third option with a general strengthening would be indicated.
