The joint structure of major depression, anxiety disorders, and trait negative affect

Abstract

Background:

Dimensional models of psychopathology demonstrate that two correlated factors of fear and distress account for the covariation among depressive and anxiety disorders. Nevertheless, these models tend to exclude variables relevant to psychopathology, such as temperament traits. This study examined the joint structure of DSM-IV-based major depression and anxiety disorders along with trait negative affect in a representative sample of adult individuals residing in the cities of São Paulo and Rio de Janeiro, Brazil.

Methods:

The sample consisted of 3,728 individuals who were administered sections D (phobic, anxiety and panic disorders) and E (depressive disorders) of the Composite International Diagnostic Interview (CIDI) 2.1 and a validated version of the Positive and Negative Affect Schedule. Data were analyzed using correlational and structural equation modeling.

Results:

Lifetime prevalence ranged from 2.4% for panic disorder to 23.2% for major depression. Most target variables were moderately correlated. A two-factor model specifying correlated fear and distress factors was retained and confirmed for models including only diagnostic variables and diagnostic variables along with trait negative affect.

Conclusions:

This study provides support for characterization of internalizing psychopathology and trait negative affect in terms of correlated dimensions of distress and fear. These results have potential implications for psychiatric taxonomy and for understanding the relationship between temperament and psychopathology.

Diagnosis and classification; emotion; epidemiology; mood disorders; unipolar; anxiety disorder; generalized

Introduction

There has been intensive effort to develop a quantitative, empirically based model for psychopathology.¹-⁴ These efforts have been directed at addressing inherent limitations associated with pervasive clinical heterogeneity and diagnostic comorbidity,⁵-⁸ along with the fact that covariance patterns among differing psychopathologies and temperament/personality traits are not addressed by dominant classification systems of mental disorders.⁹,¹⁰

In response to these limitations, different researchers have modeled the underlying structure of common psychiatric conditions to characterize diagnostic overlap using structural equation modeling techniques.¹¹-¹⁵ These studies have consistently shown that a higher-order factor of internalizing proneness accounts for the covariation among unipolar mood and anxiety disorders, whereas a factor of externalizing proneness underlies the covariation among antisocial behavior and substance-use disorders.¹⁶,¹⁷ The internalizing factor is composed - at a next-order structural level - by two subfactors of distress or anxious-misery (comprising unipolar mood disorders, generalized anxiety disorder, and posttraumatic stress disorder) and fear (comprising phobic, panic, and obsessive-compulsive disorders). However, this two-subfactor higher-order internalizing structure is somewhat controversial, with recent evidence examining: 1) the role of latent internalizing and externalizing variables in the development of lifetime comorbidity¹⁸; 2) the structure of common mental disorders in incarcerated offenders¹⁵; and 3) the structure of common and uncommon mental disorders,¹⁹ suggesting a less-differentiated, single-factor solution to the domain of mood and anxiety disorders.

Additionally, the conceptual boundaries of dispositional distress/anxious-misery and fear remain controversial, sometimes being characterized as distinct and sometimes as psychologically indistinguishable phenomena.²⁰ Tellegen & Waller²¹ describe anxiety and depression as specific manifestations of a higher-order dimension of negative emotionality, and fear as pertaining to a higher-order dimension of constraint vs. disinhibition. These personality variables have also been characterized as independent at empirical and conceptual levels. The affective and emotional composite temperament model (AFECT)²²,²³ conceives of distress and fear as independent emotional factors, with distress composed of separable facets that account for the degree of sensitivity to adversity and anxiety, and fear comprising a lower-order trait concept akin to inhibition.²²

By contrast, the Five Factor Model of Personality delineates negative emotions (fear, anxiety, distress, anger) as lower-order facets of a common trait dimension of neuroticism.²⁴,²⁵ Along similar lines, Clark & Watson's temperament model²⁶ posits that fear, anxiety, and distress are manifestations of a common negative affect/activation factor. Furthermore, both models posit that neuroticism/negative affect is common to all unipolar mood and anxiety disorders.²⁴,²⁵ Clark & Watson also showed that symptomatic elements specific to anxiety and depression allowed them to be differentiated: anxiety is characterized by physiological hyperarousal, whereas depression is characterized by anhedonia or low positive affect (PA).²⁶ Watson²⁷ advanced this conception by formulating a hierarchical model in which each unipolar mood and anxiety condition was classified into four groups based on the level of specificity vs. variability attributable to the general negative affect factor: 1) high negative affect conditions with limited specificity; 2) high negative affect conditions with greater specificity; 3) low negative affect conditions with greater specificity; and 4) low negative affect conditions with limited specificity. Within this framework, depressive and anxiety conditions can be differentiated according to the size of this general negative affect factor, and depression is further characterized by low levels of PA (a construct similar to anhedonia).

Few studies have examined the latent structure of internalizing psychopathology in conjunction with temperament/personality traits. One study by Hettema et al.²⁸ examined how genetic and environmental factors shared by trait neuroticism and internalizing disorders accounted for covariation patterns in a large sample of twins. Results indicated substantial overlap between the single-genetic factors that account for individual differences in trait neuroticism and increased liability across the internalizing disorders. Another relevant study that employed the Personality Assessment Inventory (PAI) - a widely used measure of adult personality and psychopathology - delineated internalizing and externalizing factors as distinct factors that saturated directly the covariation patterns among indicators of anxiety, depression, and aggression.²⁹

Taken together, the published evidence indicates that subdivision of the internalizing factor into distress/anxious-misery and fear subfactors requires further examination, particularly if model estimation includes temperament/personality traits. Thus, the current study had three main objectives: 1) to evaluate - via correlational analysis - the hypothesis that trait negative affect (NA) is associated, to differing degrees, with all internalizing psychopathology, whereas PA is a specific feature of depression²⁶,²⁷; 2) to examine the structure of eight DSM-IV-defined unipolar mood and anxiety disorders; 3) to examine the joint structure of DSM-defined internalizing disorders along with a validated measure of trait negative affect.³⁰

Methods

Participants and data collection procedures

The data for the current study were derived from a single-session, population-based cross-sectional survey carried out in the cities of São Paulo and Rio de Janeiro, Brazil. The study was conducted to assess the impact of urban violence on the prevalence of alcohol dependence, unipolar mood and anxiety disorders, and other mental health-related problems. A detailed description of the protocol of this study is provided elsewhere by Andreoli et al.³¹

All participants were assessed in their households by trained non-clinicians using structured questionnaire and interview measures used widely in international research in psychology and psychiatry, including the 20-item Positive and Negative Affect Schedule (PANAS)³⁰ and the Composite International Diagnostic Interview (CIDI) 2.1.³²,³³ All instruments were previously adapted and validated for use in Brazil or, in the case of some inventories including the PANAS, carefully translated to Brazilian Portuguese via standard procedures of translation/back-translation. Participation in the study was voluntary, written informed consent was obtained prior to data collection, and the study protocol was approved by the ethics committee of the Universidade Federal de São Paulo, Brazil.

The resulting sample consisted of 3,728 individuals (1,614 males and 2,114 females), with a mean age of 39.38 years (SD = 15.52, range = 15 to 75 years) and mean educational attainment of 8.79 years of formal schooling (SD = 4.29, range = 0 to 30 years). The racial composition was 43.7% white, 15.6% black, and 36.9% mixed-race. Most participants reported being single (41.5%) or married (41%).

Instruments

The PANAS³⁴ consists of two 10-item mood scales designed to provide independent measures of PA and NA. The PANAS was originally designed as a self-report questionnaire³⁴; however, its administration in a face-to-face interview setting was preferred in the current study to allow the standardization of data gathering procedures regardless of the literacy level of the participant. This interview version of PANAS has been previously validated.³⁰ Respondents were asked to rate, on a 5-point Likert scale (“very slightly or not at all” to “very much”), the extent to which they experienced each particular emotion within a general time-frame (i.e., “in general, in your life as a whole”), yielding trait-oriented scores.

The CIDI 2.1³²,³³ is a structured questionnaire that assesses psychiatric diagnoses via computerized algorithms according to the criteria of the ICD-10 and the DSM-IV. The Brazilian version of the CIDI 2.1 exhibits good levels of internal consistency and acceptable sensitivity and specificity in relation to clinical assessments performed by psychiatrists for most disorders. In the current study, CIDI-2.1 diagnoses were obtained on the basis of the DSM-IV criteria for major depressive disorder (MDD), general anxiety disorder (GAD), posttraumatic stress disorder (PTSD), obsessive-compulsive disorder (OCD), social phobia (SP), specific phobia (SpP), agoraphobia (AP), and panic disorder (PD). All diagnostic variables were coded in binary format as present or absent.

Statistical analysis

Diagnostic structure was first examined using weighted least squares exploratory structural equation modeling (EFA) with geomin oblique rotation. Models with one to three factors were evaluated using root mean square residual (RMR) (≤ 0.05), root mean square error of approximation (RMSEA) (≤ 0.06) values as goodness of fit indexes.³⁵,³⁶ Next, based on published evidence¹¹-¹⁵ and EFA results, the fit of two alternative structural models (single-factor and a two-factor model) were evaluated using weighted least square confirmatory factor analysis (CFA). The fit of these models was compared based on multiple goodness fit indexes: RMSEA; the comparative fit index (CFI); and the Tucker-Lewis index (TLI). Values greater than 0.90 are considered to indicate adequate fit to the data, and those greater than 0.95, close fit for CFI and TLI.³⁵,³⁶

Subsequently, the same procedure described above was implemented using diagnostic and trait NA variables. First, EFA with geomin oblique rotation was used to estimate models with one to three factors, with factors retained based on RMR and RMSEA value considerations. Second, four models were compared: a single-factor model, a two-factor model with NA loading on the distress factor, a two-factor model with NA loading on the fear factor, and a two-factor model with NA loading on both distress and fear factors.

Results

Descriptive statistics, prevalence, and interrelations among variables

The mean PA score for the sample was 30.57 (SD = 8.44) and the mean NA score was 20.92 (SD = 8.85). The least prevalent disorder was PD (2.4%) and the most prevalent was MDD (23.2%). Table 1 displays the prevalence of hierarchy-free MDD and anxiety disorders.

Table 1
Lifetime prevalence of hierarchy-free disorders

Disorder	Prevalence (%)
Specific phobia	18.7
Social phobia	7.3
Agoraphobia	12.9
Panic disorder	2.4
Major depressive disorder	23.2
Generalized anxiety disorder	8.4
Obsessive-compulsive disorder	4.0
Posttraumatic stress disorder	9.7

Tetrachoric correlations among mental disorders were mostly moderate: AP and PD exhibited the highest association (0.66) and GAD and AP the weakest one (0.29). NA displayed mild positive associations with all diagnostic variables (ranging from 0.27 to 0.30), while PA showed a modest negative association with MDD (0.18) and negligible associations with other diagnostic variables. NA and PA themselves were moderately correlated (0.29). Because correlations of PA with diagnostic variables were low and inconsistent, this variable was excluded from structural analysis. Table 2 displays the correlation matrix of the diagnostic and mood variables.

Table 2
Matrix of tetrachoric correlation among evaluated variables

Variables	SpP	SP	AP	PD	MDD	GAD	OCD	PTSD	NA	PA
SpP	1
SP	0.48	1
AP	0.43	0.53	1
PD	0.43	0.48	0.66	1
MDD	0.37	0.26	0.35	0.46	1
GAD	0.39	0.38	0.29	0.39	0.54	1
OCD	0.42	0.45	0.48	0.48	0.45	0.43	1
PTSD	0.50	0.41	0.40	0.50	0.50	0.49	0.44	1
NA	0.27	0.29	0.28	0.28	0.27	0.27	0.28	0.30	1
PA	-0.03	-0.08	-0.08	-0.10	-0.18	-0.07	-0.12	-0.08	0.29	1

• AP = agoraphobia; GAD = general anxiety disorder; MDD = major depression; NA = negative affect; OCD = obsessive-compulsive disorder; PA = positive affect; PD = panic disorder; PTSD = posttraumatic stress disorder; SP = social phobia; SpP = specific phobia.

Factor structure of diagnostic variables

The statistical criteria adopted for the EFA phase identified two correlated factors (r = 0.66) underlying the covariation among internalizing disorders. The first (fear) factor subsumed phobic and panic disorders and the second (distress) factor subsumed SpP, MDD, GAD, OCD and PTSD. Contrary to prediction, SpP displayed appreciable loadings on both factors (≥ 0.32), with a stronger loading for the distress factor. Table 3 shows the retained two-factor solution and reference model-fit statistics.

Table 3
Approximate simple factor structure for diagnostic variables

Diagnoses	Distress factor	Fear factor	Residual variance
MDD	0.75	-0.06	0.50
GAD	0.78	-0.10	0.49
PTSD	0.63	0.15	0.47
OCD	0.47	0.22	0.58
SpP	0.39	0.32	0.57
SP	0.17	0.55	0.54
AP	-0.12	0.89	0.33
Panic	0.15	0.69	0.37
RMR	0.04
RMSEA	0.02

• AP = agoraphobia; GAD = general anxiety disorder; MDD = major depression; OCD = obsessive-compulsive disorder; PTSD = posttraumatic stress disorder; RMR = root mean square residual; RMSEA = root mean square error of approximation; SP = social phobia; SpP = specific phobia.
• Bold type denotes significant correlations (> 0.32).

As shown in Table 4, both models displayed excellent fit to the data. The two-factor model showed slight statistical superiority in relation to the one-factor model and was favored on the basis of previous evidence.¹² Figure 1 depicts the retained model.

Figure 1
Optimal two-factor model for diagnostic variables. GAD = general anxiety disorder; OCD = obsessive-compulsive disorder; PTSD = posttraumatic stress disorder

Table 4
Fit statistics for single- and two-factor confirmatory models using diagnostic variables

Model	Chi-square	DF	SRMR	RMSEA	CFI	TLI
One factor	97.079	18	0.06	0.03	0.96	0.96
Two factors	45.313	17	0.04	0.02	0.99	0.99

• CFI = comparative fit index; DF = degrees of freedom; RMSEA = root mean square error of approximation; SRMR = standardized root mean square residual; TLI = Tucker-Lewis index.

Factor structure of diagnostic and NA variables

RMR (0.037) and RMSEA (0.02) indicated the presence of two correlated factors (r = 0.72) underlying the covariation of diagnostic and NA variables. The first factor subsumed AP, SP, and PD, while the second factor subsumed SpP, MDD, GAD, OCD, PTSD, and NA. Unexpectedly, SpP did not show a significant loading on the fear factor (< 0.32). Table 5 shows the factor structure of joint diagnostic and NA variables based on EFA.

Table 5
Approximate factor structure for diagnostic and trait variables

Variables	Distress factor	Fear factor	Residual variance
NA	0.35	0.11	0.81
MDD	0.74	-0.08	0.53
GAD	0.73	-0.10	0.55
PTSD	0.75	0.04	0.40
OCD	0.54	0.16	0.56
SpP	0.44	0.26	0.58
SP	0.25	0.49	0.53
AP	-0.11	0.92	0.28
Panic	0.16	0.65	0.41
RMR	0.03
RMSEA	0.02

• AP = agoraphobia; GAD = general anxiety disorder; MDD = major depression; NA = negative affect; OCD = obsessive-compulsive disorder; PTSD = posttraumatic stress disorder; RMR = root mean square residual; RMSEA = root mean square error of approximation; SP = social phobia; SpP = specific phobia.
• Bold type denotes significant correlations (> 0.32).

Table 6 shows fit statistics for the single- and alternative two-factor models that were specified. As evident, all models show excellent fit to the data, with the two-factor models displaying slight statistical superiority in relation to the single-factor model. Correlations between the factors in all two-factor models were uniformly high (> 0.85), and associations of NA with the distress factor were more substantial than with the fear factor. Based on EFA results and theoretical accounts, the two-factor model with NA loading on the distress factor was preferred in relation to other models. Figure 2 depicts the retained model.

Figure 2
Optimal two-factor model for joint diagnostic and trait negative affect variables. GAD = general anxiety disorder; OCD = obsessive-compulsive disorder; PTSD = posttraumatic stress disorder

Table 6
Fit statistics for four competing confirmatory models using diagnostic and trait variables

Model	Description	Chi-square	DF	RMSEA	CFI	TLI
One factor	Single-factor	120.612	25	0.032	0.979	0.982
Two factors	NA-distress	70.362	24	0.020	0.990	0.990
Two factors	NA-fear	112.514	24	0.031	0.980	0.983
Two factors	NA-distress/fear	74.109	23	0.024	0.989	0.990

• CFI = comparative fit index; DF = degrees of freedom; RMSEA = root mean square error of approximation; TLI = Tucker-Lewis index.

Discussion

Evaluations of the structure of mental disorders in adult samples have consistently shown that a dimension of internalizing proneness underlies the systematic comorbidity between unipolar mood and anxiety disorders.¹¹-¹⁵ This dimension has often been partitioned into subfactors of distress (also called anxious-misery) and fear.¹² However, some recent evidence has favored a single-factor model.¹⁵,¹⁸ It has also been suggested that a general internalizing factor may account for the covariation among traits related to NA and internalizing disorders.²⁸,²⁹ Nonetheless, very few attempts have been made to address this particular issue empirically. The present study sought to address these gaps by examining the underlying structure of internalizing disorders alone and in conjunction with an NA trait measure.

Correlations among variables of interest indicated moderate overlap in general, with the exception of PA, which showed slight to nonsignificant associations with diagnostic variables. In general, these results are in accordance with previously reported findings and provide additional support for the classification of unipolar mood and anxiety disorders into a single diagnostic spectrum of internalizing/emotional disorders.³⁷ Additionally, this observed pattern of correlations appears relevant to the viability of tenets of Watson's quadripartite model²⁷ regarding the role of NA and PA (or activation) in differentiating depressive and anxiety disorders. Watson²⁷ proposed that depressive mood and anxiety syndromes could be classified and differentiated, based on the level of specificity vs. the degree of variance attributable to the general distress factor (NA), into four groups: 1) high distress symptoms/conditions with limited specificity; 2) high distress symptoms/conditions with greater specificity; 3) low distress symptoms/conditions with greater specificity; and 4) low distress symptoms/conditions with limited specificity. Based on these assumptions, it was expected that NA would show consistent but varying degrees of association with assessed diagnostic conditions, and that PA would show significant associations with MDD alone. As shown in Table 2, these predictions were only partially supported by our data: PA showed negative associations with MDD and null or almost null associations with the other diagnoses, whereas NA showed indistinguishable coefficient values across variables. Based on our findings, PA may be efficient in differentiating MDD from anxiety disorders, but the degree of NA cannot be used to account for differential diagnoses.

Similar to previously reported findings,¹²,¹³ our optimal structural solution indicates that internalizing structure is better conceptualized using a two-factor model of distress and fear tendencies. Still, the way diagnostic conditions were situated within the two-factor model was slightly different here than in other studies,¹²,¹³ OCD was positioned with the distress disorders and SpP loaded on both distress and fear factors (Figure 1). When NA was included in the structural model, SpP was positioned with the distress disorders, not the fear disorders (Figure 2).

These dissimilarities may be attributed to methodological differences. Unlike previous studies,¹²,¹³ we used an exploratory/confirmatory approach - first allowing the data to show the best way to fit the model, and then testing the fit of alternative models via CFA using more stringent criteria. Despite this, it remains conceptually unclear whether OCD is best understood as a fear- or distress-laden disorder. For example, trait fear may be described as a disposition to show flight/fight/freeze responses in the face of an actual threat, whereas trait anxiety (distress/misery-anxiety factor) may be understood as a tendency to experience hypervigilance and discomfort in the perception of potential threats²⁰; OCD symptomatology is clearly characterized by heightened distress and vigilance motivated by the anticipation of threats and negative outcomes,⁹,¹⁰ which seems closer descriptively to trait distress/anxious-misery than fear.

The placement of OCD and SpP as indicators of a distress factor in the current study may have a clinical explanation, i.e., it may be related to the effect of restricting people to places where the feeling that one is able to control the situation predominates. SP, AP, and PD share acute anxiety related to the possibility of being in a place or situation where escape would be difficult if something went wrong (like a panic attack),⁹,¹⁰ which may lead to avoidance of external environments and increased time in familiar surroundings such as one's home. MDD, GAD, and TEPT do not necessarily show a similar pattern: in these disorders, symptomatology tends to be more pervasive and, thus, not have an effect that causes increasing avoidance of outdoor environments. Similarly, OCD and SpP may not prevent people from going outdoors: OCD has the potential to benefit society and work environments,³⁸ which shows its potential viability in outdoor activities; on the other hand, the stimulus associated with a specific phobia may not even be available in the external environment, as is typically the case for those who live in major cities and have specific animal phobias (e.g., of snakes). Thus, the structure reported herein indicates that the organization of internalizing disorders in terms of distress and fear may also be related to the effect these factors have on restricting a person into familiar and controllable environments.

The observed loading of NA on the distress factor was somewhat less robust than the loading of the other disorders on the same factor. This finding, which suggests that NA variability only partially accounts for a common liability factor of distress, is in accordance with Hettema et al.,²⁸ who reported similar results using neuroticism as an indicator of the internalizing factor. Further investigations are required to address the complex relations between temperament, personality, and psychopathology.³⁹

This study has notable strengths and limitations. Among its virtues, we highlight the breadth and representativeness of the sample, which allows the results to be generalized with some confidence to the population of Brazilian adults. Moreover, the instruments used in the evaluations show good evidence of validity and reliability. A limitation that is worth mentioning was the non-availability of the diagnosis of dysthymic disorder in the study sample, which prevented the structural analysis from containing the entire spectrum of unipolar internalizing psychopathology.

In conclusion, our findings showed that two correlated factors of distress and fear may account for the covariation among NA, major depression, and anxiety disorders, and that OCD and SpP are better conceptualized as distress disorders in the current sample. Furthermore, we hypothesize that distress and fear factors impact the life of people in different ways: the fear factor may be associated with avoidance of external environments, while the distress factor may not.

Acknowledgements

Funding for this study was provided by grants from Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (process no. MCT/CNPq 01/2005) and from Fundação de Amparo è Pesquisa do Estado de São Paulo (FAPESP) (process no. 420122/2005-2). HWC, DRL, RAB, MFM, and JJM are CNPq research fellows.

References

• ¹
  Krueger RF, Markon KE, Patrick CJ, Benning SD, Kramer MD. Linking antisocial behavior, substance use, and personality: an intergrative quantitative model of the adult externalizing spectrum. J Abnorm Psychol. 2007;116:645-66.
• ²
  Krueger RF, Deyoung CG, Markon KE. Towards scientific useful quantitative models of psychopathology: the importance of a comparative approach. Behav Brain Sci. 2010;33:163-4.
• ³
  Wright AG, Krueger RF, Hobbs MJ, Markon KE, Eaton NR, Slade T. The structure of psychopathology: Toward an expanded quantitative empirical model. J Abnorm Psychol. 2013;122:281-94.
• ⁴
  Watson D. Rethinking the mood and anxiety disorders: a quantitative hierarchical model for DSM-V. J Abnorm Psychol. 2005;114:522-36.
• ⁵
  Clark LA, Watson D, Reynalds S. Diagnostic and classification of psychopathology: challenges to the current system and future directions. Annu Rev Psychol. 1995;46:121-53.
• ⁶
  Mineka S, Watson D, Clark LA. Comorbidity of anxiety and unipolar mood disorders. Annu Rev Psychol. 1998;49:377-412.
• ⁷
  Widiger TA, Sankis LM. Adult psychopathology: issues and controversies. Annu Rev Psychol. 2000;51:377-404.
• ⁸
  Widiger TA, Clark LA. Towards DSM-V and the classification of psychopathology. Psychol Bull. 2000;126: 946 -63.
• ⁹
  Associação Americana de Psiquiatria. Manual diagnostico estatístico dos transtornos mentais - DSM-IV-TR. 4th ed. Porto Alegre: Artmed; 2002.
• ¹⁰
  Organização Mundial de Saúde (OMS). CID 10. Classificação estatística internacional de doenças e problemas relacionados a saúde. Porto Alegre: Artmed; 1993.
• ¹¹
  Krueger RF, Caspi A, Moffit TE, Silva PA. The structure and stability of common mental disorders (DSM-III-R): a longitudinal-epidemiological study. J Abnorm Psychol. 1998;107:216-27.
• ¹²
  Krueger RF. The structure of common mental disorders. Arch Gen Psychiatry. 1999;56:921-6.
• ¹³
  Vollebergh W, Iedema J, Bijl RV, de Graaf R, Smit F, Ormel J. The structure and stability of common mental disorders: the NEMESIS study. Arch Gen Psychiatry. 2001;58:597-603.
• ¹⁴
  Slade T, Watson D. The structure of common DSM-IV and ICD-10 mental disorders in the Australian general population. Psychol Med. 2006;36:1593-600.
• ¹⁵
  de Carvalho HW, Andreoli SB, Vaidyanathan U, Patrick CJ, Quintana IM, Jorge MR. The structure of common mental disorders in incarcerated offenders. Compr Psychiatry. 2013;54:111-6.
• ¹⁶
  Krueger RF, Markon KE. Reinterpreting comorbidity: A model-based approach to understanding and classifying psychopathology. Annu Rev Clin Psychol. 2006;2:111-33.
• ¹⁷
  Krueger RF, Markon KE. Understanding psychopathology: melding behavior genetics, personality, and quantitative psychology to develop an empirical based model. Curr Dir Psychol Sci. 2006;15:113-7.
• ¹⁸
  Kessler RC, Ormel J, Petukhove M, McLaughlin KA, Green JG, Russo LJ, et al. Development of lifetime comorbidity in the World Health Organization world mental health surveys. Arch Gen Psychiatry. 2011;68:90-100.
• ¹⁹
  Forbush KT, Watson D. The structure of common and uncommon mental disorders. Psychol Med. 2013;43:97-108.
• ²⁰
  Sylvers P, lilienfield SO, LaPrairie JL. Differences between trait fear and trait anxiety: implications for psychopathology. Clin Psychol Rev. 2011;31:122-37.
• ²¹
  Tellegen A, Waller NG. Exploring personality through test construction: development of the multidimensional personality questionnaire. Minneapolis: University of Minnesota; 2007.
• ²²
  Lara DR, Bisol LW, Brumstein MG, Reppold CT, de Carvalho HW, Ottoni GL. The affective and emotional composite temperament model and scale: a system-based integrative approach. J Affect Disord. 2012;140:14-37.
• ²³
  de Carvalho HW, Bisol LW, Ottoni GL, Lara DR. The affective and emotional composite temperament model and scale: psychometric analysis including anxiety and instability subscales. In: Scietific Program and Abstract Book, 3rd International Congress on Neurobiology, Psychofarmacology and Treatment Guidance; 2013: p. 176.
• ²⁴
  Funder DC. Personality. Annu Rev Psychol. 2001;52:197-221.
• ²⁵
  Miller JD, Lyman DR, Widiger TA, Leukefeld C. Personality disorders as extreme variants of common personality dimensions: can the Five Factor Model adequately represent psychopathy? J Pers. 2001;69:253-76.
• ²⁶
  Clark LA, Watson D. Tripartite model of anxiety and depression: psychometric evidence and taxonomic implications. J Abnorm Psychol. 1991;100:316-36.
• ²⁷
  Watson D. Differentiating the mood and anxiety disorders: a quadripartite model. Annu Rev Clin Psychol. 2009;5:221-47.
• ²⁸
  Hettema JM, Neale MC, Myers JM, Prescott CA, Kendler KS. A population-based twin study of the relationship between neuroticism and internalizing disorders. Am J Psychiatry. 2006;163:857-64.
• ²⁹
  Hopwood CJ, Moser JS. Personality Assessment Inventory internalizing and externalizing structure in college students: Invariance across sex and ethnicity. Person Individ Diff. 2011;50:116-9.
• ³⁰
  de Carvalho HW, Andreoli SB, Lara DR, Patrick CJ, Quintana MI, Brassan RA, et al. Structural validity and reliability of the Positive and Negative Affect Schedule (PANAS): evidence from a large Brazilian community sample. Rev Bras Psiquiatr. 2013;35:169-72.
• ³¹
  Andreoli SB, Ribeiro WS, Quintana MI, Guindalini C, Breen G, Blay SL, et al. Violence and post-traumatic stress disorder in Sao Paulo and Rio de Janeiro, Brazil: the protocol for an epidemiological and genetic survey. BMC Psychiatry. 2009;9:34.
• ³²
  Quintana MI, Andreoli SB, Jorge MR, Gastal FL, Miranda CT. The reliability of the Brazilian version of the Composite International Diagnostic Interview (CIDI 2.1). Braz J Med Biol Res. 2004;37:1739-45.
• ³³
  Quintana MI, Gastal FL, Jorge MR, Miranda CT, Andreoli SB. Validity and limitations of the Brazilian version of the Composite International Diagnostic Interview (CIDI 2.1). Rev Bras Psiquiatr. 2007;29:18-22.
• ³⁴
  Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative affect: the PANAS scales. J Pers Soc Psychol. 1988;54:1063-70.
• ³⁵
  Marôco J. Análise de equações estruturais: fundamentos teóricos, softwares e aplicações. Lisboa: Reportnumber; 2010.
• ³⁶
  Byrne BM. Structural equation modeling with Mplus: basic concepts, applications, and programming. New York: Routledge Taylor & Francis Group; 2012.
• ³⁷
  Goldberg DP, Krueger RF, Andrews G, Hobbs MJ. Emotional disorders: cluster 4 of the proposed meta-structure for DSM-V and ICD-10. Psychol Med. 2009;39:2043-59.
• ³⁸
  Polimeni J, Reiss JP, Sareen J. Could obsessive-compulsive disorder have originated as a group-selected adaptive trait in traditional societies? Med Hypotheses. 2005;65:655-64.
• ³⁹
  Andersen AM, Bienvenu OJ. Personality and psychopathology. Int Rev Psychiatry. 2011;23:234-47.

Publication Dates

• Publication in this collection
  15 Apr 2014
• Date of issue
  Oct-Dec 2014

History

• Received
  27 Nov 2013
• Accepted
  2 Jan 2014