Confirmatory Factor Analysis of the Depression–Anxiety–Stress Scales in Depressed and Anxious Patients

The Depression–Anxiety–Stress Scales (DASS; P. F. Lovibond & S. H. Lovibond, 1995) have shown considerable promise in their ability to differentially assess depression and anxiety symptoms. Most of the work to date has relied on normal or predominantly anxious clinical samples. To extend this research to a predominantly depressed sample, a psychiatric sample from a mood disorders program (N = 439) was used to investigate the factor structure of the DASS. Confirmatory factor analyses indicated that a 3-factor model for the 21-item DASS was supported. A 3-factor model with crossloading items for the 42-item version showed no practical improvement. Further confirmatory analyses evaluated the ability of item subsets from the DASS to represent the constructs specified in the tripartite model (L. A. Clark & D. Watson, 1991). Strong support was obtained for the DASS to represent the construct of anhedonia and physiological hyperarousal, and their relative importance is consistent with predictions of the tripartite model in understanding the relationship between anxiety and depression.     

Depression Anxiety Stress Scales-21: Factor Structure and Test-Retest Invariance,and Temporal Stability and Uniqueness of Latent Factors in Older Adults

This study examined the factor structure and test-retest invariance, and temporal stability and uniqueness of the latent factors (depression, anxiety, and stress) of the Depression Anxiety Stress Scales-21 (DASS-21; Lovibond & Lovibond, 1995) in group a of 269 older adults (age ranging from 60 to 85 years) from the general community. Participants completed the DASS-21 twice, 3 months apart. Confirmatory factor analysis (CFA) of their ratings at Time 1 indicated support for the original 3-factor oblique model (factors for depression, anxiety, and stress). Additional analyses showed support for test-retest invariance for both the measurement (configural, metric and thresholds) and structural (variances and covariances) components of this model. Results also indicated temporal stability and uniqueness of the latent factors. The practical, theoretical, research and clinical implications of the findings are discussed.     

Anxiety, Depression, and Tension/Stress in Children

This study explored whether the structure of negative affect in children is similar to that previously found in adults (S. H. Lovibond & P. F. Lovibond, 1995). A sample of 577 7–14-year-old children completed a 76-item questionnaire designed to measure anxiety, depression, and tension/stress, comprising both exploratory items and the full item content of established measures of childhood anxiety and depression. Exploratory factor analyses indicated that several symptoms that were previously found to be nonspecific in adults also failed to differentiate between anxiety and depression in youth. Further, the core symptoms of depression were shown to be similar to those previously identified in adults. However, the hypothetical constructs of anxiety and tension/stress could not be differentiated. It is possible that these two distinct affective syndromes have not yet emerged in childhood. Alternatively, future refinements of the questionnaire's item content may lead to a more precise assessment of anxiety and tension/stress in children.     

The Personal Disturbance Scale (sAD): normative data and latent structure in a large non-clinical sample

The Personal Disturbance Scale [sAD; Bedford & Foulds (1978) Delusions–Symptoms–States Inventory State of Anxiety and Depression. Windsor: NFER-Nelson] is widely used in diverse settings and yet there are unresolved issues concerning its psychometric properties and normative data for the English speaking version are limited. The sAD was administered to a large sample of the general adult population (N=758). Demographic variables (gender, age, years of education and occupational status) had only very modest influences on sAD scores. Tables are presented for conversion of raw scores on the Anxiety, Depression and Total scales to percentiles. The sAD scales possessed adequate convergent and discriminant validity, as demonstrated by their pattern of correlations with two other measures of depression and anxiety (the DASS and the HADS). Ten competing models of the latent structure of the sAD were derived from theoretical and empirical sources. These models were evaluated using confirmatory factor analysis. The best fitting model (CFI=0.96) had a tripartite structure, and consisted of a general factor of psychological distress/negative affectivity (all items loaded on this factor) plus orthogonal specific factors of anxiety and depression. Correlated errors specified according to previous empirical findings were permitted. The theoretical and practical implications of this latent structure are discussed.     

The Structure of Anxiety and Depression in a Normative Sample of Younger and Older Australian Adolescents

It has been reported that depression and anxiety have overlapping symptoms and are conceptually interrelated. One of the most prominent theoretical developments that explain this association is Clark and Watson’s tripartite model (Journal of Abnormal Psychology, 100:316–336, 1991) that posits these two disorders and negative emotions share a latent component of negative affect (NA). The current study had two aims, (a) to compare a tripartite factor structure against competing models by Confirmatory Factor Analysis (CFA) of the Depression Anxiety and Stress Scales (DASS-21) (Lovibond and Lovibond 1995), and (b) explore the psychometric properties of the DASS-21. The DASS-21 was completed by a representative sample of 4039 young Australians, aged 12–18 years, as part of the South Australian Youth Mental Health Survey (SAYMHS), South Australia, Australia. The best fitting model for the data consisted of anhedonic depression, physiological hyperarousal, and general NA. The psychometric properties of the DASS-21 were excellent with a stable and interpretable factor structure and good internal consistency. The results of the current study suggest that the theoretical tripartite structure of depression and anxiety is robust and applicable among Australian youth. The diagnostic, clinical and theoretical implications of these findings are discussed.     

Using the Depression Anxiety Stress Scale 21 (DASS‐21) across cultures

The DASS‐21 is a well‐established instrument for measuring depression, anxiety, and stress with good reliability and validity reported from Hispanic American, British, and Australian adults. However, the lack of appropriate validation among Asian populations continues to pose concerns over the use of DASS‐21 in Asian samples. Cultural variation may influence the individual's experience and emotional expression. Thus, when researchers and practitioners employ Western‐based assessments with Asian populations by directly translating them without an appropriate validation, the process can be challenging. We conducted a series of rigorous statistical tests and minimized any potential confounds from the demographic information. Following factor analyses, we performed multigroup analysis across six nations to demonstrate consistency of our findings. The advantages of this revised DASS‐18 stress scale are twofold. First, it possesses fewer items, which results in a cleaner factorial structure. Second, it has a smaller interfactor correlation. With these justifications, the revised DASS‐18 stress scale is potentially more suitable for Asian populations. Nonetheless, given limitations, findings should be considered preliminary.     

Replication and Extension of a Hierarchical Model of Social Anxiety and Depression: Fear of Positive Evaluation as a Key Unique Factor in Social Anxiety

Wang, Hsu, Chiu, and Liang (2012, Journal of Anxiety Disorders, 26, 215–224) recently proposed a hierarchical model of social interaction anxiety and depression to account for both the commonalities and distinctions between these conditions. In the present paper, this model was extended to more broadly encompass the symptoms of social anxiety disorder, and replicated in a large unselected, undergraduate sample (n = 585). Structural equation modeling (SEM) and hierarchical regression analyses were employed. Negative affect and positive affect were conceptualized as general factors shared by social anxiety and depression; fear of negative evaluation (FNE) and disqualification of positive social outcomes were operationalized as specific factors, and fear of positive evaluation (FPE) was operationalized as a factor unique to social anxiety. This extended hierarchical model explicates structural relationships among these factors, in which the higher-level, general factors (i.e., high negative affect and low positive affect) represent vulnerability markers of both social anxiety and depression, and the lower-level factors (i.e., FNE, disqualification of positive social outcomes, and FPE) are the dimensions of specific cognitive features. Results from SEM and hierarchical regression analyses converged in support of the extended model. FPE is further supported as a key symptom that differentiates social anxiety from depression.     

Multidimensionality of state and trait anxiety: factor structure of the Endler Multidimensional Anxiety Scales

The Endler Multidimensional Anxiety Scales (EMAS) are derived from an interactional model of personality that proposes that anxiety is a function of the interaction of person and situation variables. The EMAS distinguish between state and trait anxiety and assume that both are multidimensional constructs. The EMAS were administered to 2,009 students in a neutral situation. Three factor analyses were performed to clarify the empirical relation between state and trait anxiety and the variables within the 2 domains. Results support the distinction between state and trait anxiety. Factor analysis of the state items provided support for the separate dimensions of cognitive and autonomic state anxiety. Factor analysis of the trait items provided support for trait anxiety multidimensionality. Four congruent factors were associated with increases in state anxiety in 4 general situations: Social Evaluation, Physical Danger, Ambiguous, and Daily Routines.     

Multi-domain Models of Risk Factors for Depression and Anxiety Symptoms in Preschoolers: Evidence for Common and Specific Factors

Relatively few studies have examined multiple pathways by which risk factors from different domains are related to symptoms of anxiety and depression in young children; even fewer have assessed risks for these symptoms specifically, rather than for internalizing symptoms in general. We examined a theoretically- and empirically-based model of variables associated with these symptom types in a diverse community sample of 796 4-year-olds (391 boys, 405 girls) that included factors from the following domains: contextual (SES, stress and family conflict); parent characteristics (parental depression); parenting (support/engagement, hostility and scaffolding); and child characteristics including negative affect (NA) effortful control (EC) sensory regulation (SR), inhibitory control (IC) and attachment. We also compared the models to determine which variables contribute to a common correlates of symptoms of anxiety or depression, and which correlates differentiate between those symptom types. In the best-fitting model for these symptom types (a) SES, stress and conflict had indirect effects on both symptom types via long-chain paths; (b) caregiver depression had direct effects and indirect ones (mediated through parenting and child effortful control) on both symptom types; (c) parenting had direct and indirect effects (via temperament and SR); and temperament had direct effects on both symptom types. These data provide evidence of common risk factors, as well as indicate some specific pathways/mediators for the different symptom types. EC was related to anxiety, but not depression symptoms, suggesting that strategies to improve child EC may be particularly effective for treatment of anxiety symptoms in young children.     

Does the Beck Anxiety Inventory measure anything beyond panic attack symptoms?

The Beck Anxiety Inventory (BAI) has become a popular measure in anxiety assessment and the BAI does not overlap in content with measures of depression. There is also some factor analytic evidence to support this distinction. However, an inspection of the BAI's content indicates that many of its items resemble, or are identical to, the symptoms of panic attacks listed in the DSM-IV. Further empirical support for this suspicion is provided from the results of a factor analysis of the BAI items and the individual DSM-IV panic symptoms contained in the Panic Attack Questionnaire, using data from a sample of 157 panic disorder patients. A three-factor model (dizziness related, catastrophic cognitions/fear, cardiorespiratory distress) emerged that replicated a three-factor model of panic symptoms identified in earlier work with another panic disorder sample. All but one of the BAI items loaded highly on the three panic symptom clusters and no separate BAI factor was obtained. The BAI appears to be confounded with, or actually measures, panic attacks rather than anxiety in general. Several implications of this finding are discussed.     

Exhaustion is differentiable from depression and anxiety: evidence provided by the SF-36 vitality scale

Stress-related exhaustion is an increasingly focused phenomenon, although scientifically not yet fully conceptualised. This is perhaps partly due to the fact that the distinction between exhaustion and other already established concepts, such as depression, is not clear. This study explores whether exhaustion can be differentiated from depression and anxiety, by means of introducing the SF-36 vitality scale as a strategy for the measurement of exhaustion. The Malm? Shoulder and Neck Study cohort, in this study including 12,607 middle-aged men and women, was utilized. Depression and anxiety were assessed by the general health questionnaire (GHQ). Factor analysis was performed in order to potentially discriminate between the included measures. Factor analysis showed that all four exhaustion items of the (inverted) SF-36 vitality scale loaded on a single factor, separate from the GHQ depression and anxiety factors. These results support the notion that exhaustion can be distinguished from depression and anxiety, providing an empirical foundation for defining exhaustion as a unique concept. Furthermore, since SF-36 has been widely used in clinical and epidemiological studies, the benefits of using the SF-36 vitality scale in the assessment of exhaustion could be substantial.     

Measuring Clarity of and Attention to Emotions

To clarify what is actually measured by the trait version of the State-Trait Anxiety Inventory (STAI; Spielberger, Gorsuch, &; Lushene, 1970), we conducted a confirmatory factor analysis of various models and evaluated convergent and discriminant validity. The best fit was obtained with both a bifactor model, comprising 2 specific factors plus a general factor, and a 1-construct, 2-method model. The total score and the 2 method subscales of the STAI trait version were more strongly correlated with depression than with anxiety. In the bifactor model with 2 specific factors, the depression subscale showed stronger correlations with measures of depression than with measures of anxiety. The correlation of the hypothetical anxiety subscale with measures of depression was equivalent to or higher than its correlation with measures of anxiety. These results suggest that the questionnaire does not strictly evaluate anxiety but, rather, negative affect.     

Structure and Validity of Sluggish Cognitive Tempo Using an Expanded Item Pool in Children with Attention-Deficit/Hyperactivity Disorder

We evaluated the latent structure and validity of an expanded pool of Sluggish Cognitive Tempo (SCT) items. An experimental rating scale with 44 candidate SCT items was administered to parents and teachers of 165 children in grades 2–5 (ages 7–11) recruited for a randomized clinical trial of a psychosocial intervention for Attention-Deficit/Hyperactivity Disorder, Predominantly Inattentive Type. Exploratory factor analyses (EFA) were used to extract items with high loadings (>0.59) on primary factors of SCT and low cross-loadings (0.30 or lower) on other SCT factors and on the Inattention factor of ADHD. Items were required to meet these criteria for both informants. This procedure reduced the pool to 15 items. Generally, items representing slowness and low initiative failed these criteria. SCT factors (termed Daydreaming, Working Memory Problems, and Sleepy/Tired) showed good convergent and discriminant validity in EFA and in a confirmatory model with ADHD factors. Simultaneous regressions of impairment and comorbidity on SCT and ADHD factors found that Daydreams was associated with global impairment, and Sleepy/Tired was associated with organizational problems and depression ratings, across both informants. For teachers, Daydreams also predicted ODD (inversely); Sleepy/Tired also predicted poor academic behavior, low social skills, and problem social behavior; and Working Memory Problems predicted organizational problems and anxiety. When depression, rather than ADHD, was included among the predictors, the only SCT-related associations rendered insignificant were the teacher-reported associations of Daydreams with ODD; Working Memory Problems with anxiety, and Sleepy/Tired with poor social skills. SCT appears to be meaningfully associated with impairment, even when controlling for depression. Common behaviors resembling Working Memory problems may represent a previously undescribed factor of SCT.     

Exploring the factor structure of the Mood and Anxiety Symptom Questionnaire (MASQ)

L. A. Clark and Watson (1991) proposed a tripartite model of anxiety and depression defined in terms of common symptoms relating to general distress, anxiety-specific symptoms of hyperarousal, and depression-specific symptoms of low positive affect and loss of interest. To aid the measurement of and discrimination between anxiety and depression they developed the Mood and Anxiety Symptom Questionnaire (MASQ). Although potentially useful, the MASQ is still in an early stage of development, and there is an indication that a discrepancy exists between items and scales. Therefore, the aim of this study was to explore the items and factors of the MASQ and determine whether a pattern similar to that suggested by L. A. Clark and Watson would emerge. Factor analysis of the MASQ revealed 3 factors relating to general distress, positive affect, and anxious arousal. Inspection of the individual items indicated that some refinements are required.     

Assessment of anxiety and depression in young children: support for two separate constructs

Validated the interrelatedness of depression and anxiety in young children by testing four latent factor models: dual construct, unrelated; dual construct, correlated; single construct; and second-order or higher order analysis to test that depression and anxiety are primary constructs under the higher order factor of general affective distress. Children (N = 86) were ages 6 to 11, with mothers who were HIV-symptomatic or diagnosed with AIDS. Depression and anxiety measures included the Children's Depression Inventory (Kovacs, 1992), selected items from the Dominic-R (Valla, Bergeron, Berube, Gaudet, & St-Georges, 1994), and the Revised Children's Manifest Anxiety Scale (Reynolds & Richmond, 1985). Structural equation modeling was used to test the models. Model 2 (dual construct, correlated) fit the data better than did Models 1 and 3; results for the higher order model were identical to Model 2, suggesting the higher order model is equivalent to the dual-construct model.     

Beyond Negative Affectivity: A Hierarchical Model of Global and Transdiagnostic Vulnerabilities for Emotional Disorders

Background: Negative affectivity (NA) has been linked to anxiety and depression (DEP). Identifying the common factors between anxiety and DEP is important when explaining their overlap and comorbidity. However, general factors such as NA tend to have differential relationships with different disorders, suggesting the need to identify mediators in order to explicate these relationships. Methods: The current study tests a theoretically and empirically derived hierarchical model of emotional disorders including both a general factor (NA) and transdiagnostic risk factors [anxiety sensitivity (AS) and intolerance of uncertainty (IoU)] using structural equation modeling. AS was tested as a mid-level factor between NA and panic disorder/agoraphobia, while IoU was tested as a mid-level factor between NA and social phobia, generalized anxiety disorder, obsessive-compulsive disorder, and DEP. Data from 642 clinical outpatients with a heterogeneous presentation of emotional disorders were available for analysis. Results: The hierarchical model fits the data adequately. Moreover, while a simplified model removing AS and IoU fits the data well, it resulted in a significant loss of information for all latent disorder constructs. Limitations: Data were unavailable to estimate post-traumatic stress disorder or specific phobias. Future work will need to extend to other emotional disorders. Conclusions: This study demonstrates the importance of both general factors that link disorders together and semi-specific transdiagnostic factors partially explaining their heterogeneity. Including these mid-level factors in hierarchical models of psychopathology can help account for additional variance and help to clarify the relationship between disorder constructs and NA.     

Differentiating anxious and depressive self-statements in youth: factor structure of the Negative Affect Self-Statement Questionnaire among youth referred to an anxiety disorders clinic

Conducted a factor analysis on the items from the Negative Affect Self-Statement Questionnaire (NASSQ; Ronan, Kendall, & Rowe, 1994). This analysis yielded 4 factors (Depressive Self-Statements, Anxiety/Somatic Self-Statements, Negative Affect Self-Statements, and Positive Affect Self-Statements) broadly consistent with both the content-specificity hypothesis (Beck & Clark, 1988) and L. A. Clark and Watson's (1991b) tripartite model of anxiety and depression. The association between children's self-talk and measures of trait anxiety and depression was also examined. Self-statements with content theoretically specific to depression were the best predictors of self-reported depressive symptoms, but the results were less clear for trait anxiety. Overall, these results provide evidence for the discriminability of anxious and depressive self-talk in youth and for the utility of the NASSQ as a cognitive assessment instrument.     

Psychometric properties and measurement invariance of Depression,Anxiety and Stress Scales (DASS-21) across cultures

Cross-culturally measurement invariant instruments are useful for the assessment of mental symptoms across cultures. The current study aimed to investigate the psychometric properties and measurement invariance of the Depression, Anxiety and Stress Scales (DASS-21) (DASS) across Pakistan and Germany. German participants were recruited through an online survey (N = 1323), while Pakistani participants were recruited through online survey or paper-and-pencil survey (N = 1841). The DASS-21 showed good reliability, construct and structure validity in both countries. The DASS-21 indicated partial weak (stress subscale) and partial strong measurement invariance (depression and anxiety subscales) between both countries. Latent mean comparison of depression and anxiety symptoms between Pakistani and German students indicated that Pakistani university students experience more symptoms of depression and anxiety. Results suggest that the DASS-21 could be used in Pakistan and Germany, but caution should be taken when making direct comparisons between the two countries.     

The state-trait anxiety inventory, trait version: does it really measure anxiety?

To clarify what is actually measured by the trait version of the State-Trait Anxiety Inventory (STAI; Spielberger, Gorsuch, & Lushene, 1970), we conducted a confirmatory factor analysis of various models and evaluated convergent and discriminant validity. The best fit was obtained with both a bifactor model, comprising 2 specific factors plus a general factor, and a 1-construct, 2-method model. The total score and the 2 method subscales of the STAI trait version were more strongly correlated with depression than with anxiety. In the bifactor model with 2 specific factors, the depression subscale showed stronger correlations with measures of depression than with measures of anxiety. The correlation of the hypothetical anxiety subscale with measures of depression was equivalent to or higher than its correlation with measures of anxiety. These results suggest that the questionnaire does not strictly evaluate anxiety but, rather, negative affect.