Examination of the Factor Structure of Psychopathology in a Mozambican Sample

Abstract

Factor-analytic studies are needed in global samples to advance understanding of psychopathology. Here, we examined the structure of psychopathology and a general psychopathology (p) factor using data from a cross-sectional study of 971 adults (63% women) from Maputo City, Mozambique. We used confirmatory factor analyses of symptoms from 15 psychiatric disorders to test common models of the structure of psychopathology. Models that included Internalizing, Substance Use, and Thought Disorder factors as well as a general p factor fit the data well. Measurement invariance testing revealed that factor loadings on p differed by gender. Higher levels of p, Internalizing, and Thought Disorder factors were associated with greater suicide risk, psychiatric comorbidity, chronic medical illnesses, and poorer functioning. A general p factor and Internalizing, Substance Use Disorder, and Thought Disorder factors were identifiable in this Mozambican sample. Understanding psychopathology dimensions is a step toward building more scalable mental health service approaches globally.

Keywords

factor analysis Mozambique p factor psychopathology

The mental health treatment gap in low and middle-income countries is well documented (Kohn et al., 2004). Progress is being made to understand and improve mental health globally, yet issues persist beyond a lack of resources. Because mental health classification systems were developed in Western contexts, research investigating the nature of psychopathology across cultures and contexts is scarce (Fonagy et al., 2021). Categorical diagnoses developed in Western settings may be inadequate for identifying and treating mental disorders in diverse global populations. Further, these classification systems encourage a single-disorder focus not supported by emerging research (McGrath et al., 2020). Indeed, high rates of comorbidity globally challenge the current emphasis on single-disorder approaches (McGrath et al., 2020). Yet research on shared comorbidities in low and middle-income countries is limited (Kane et al., 2018), which can lead to a narrow focus on categorical disorders, creating barriers to understanding diagnosis and emphasizing siloed treatments that may be ineffective, costly, and difficult to scale.

Transdiagnostic and dimensional approaches to mental health—approaches targeting symptoms across disorders or those that target shared psychopathological mechanisms (Sauer-Zavala et al., 2017)—may enable researchers and practitioners to better understand, identify, and treat comorbidities (Wainberg et al., 2021). These approaches align with calls in global mental health to focus on multiproblem, modular methods that offer one provider tools to address comorbidities (Lovero et al., 2021). In low and middle-income countries, where systems of care are being built, there is a unique opportunity to build efficient comprehensive care services targeting shared symptoms rather than single disorders (Wainberg et al., 2017). Understanding the nature of psychopathology among diverse global populations is one step toward achieving these goals (Kane et al., 2018).

Data-driven methods, such as factor-analytic modeling, are critical tools for identifying shared features across categorical disorders. Factor-analytic studies have found that discrete disorders may be best explained by a smaller number of latent factors (Achenbach & Edelbrock, 1981; Krueger & Markon, 2006) representing distinct families of disorders. These include an internalizing factor (a propensity to mood and anxiety disorders), an externalizing factor (a propensity to antisocial and substance use disorders), and a thought disorder factor (a propensity to schizophrenia, mania, and obsessive compulsive disorder; Caspi et al., 2014). Additionally, these factors are correlated (Wright et al., 2013), which led to identification of a general psychopathology factor (Lahey et al., 2012), often called the p factor (Caspi et al., 2014). The p factor captures shared variation among all forms of mental disorder, accounting for comorbidity and severity across diagnostic categories.

The p factor has been well established in Western, high-income settings (Caspi et al., 2020; Ringwald et al., 2021), but it has rarely been investigated outside of these contexts or in nonpredominantly White samples, with a few exceptions (He & Li, 2021; Martel et al., 2017; Ringwald et al., 2021). The p factor has not been explored in an African sample. Understanding the structure of psychopathology outside of high-income contexts and White samples is important for (a) understanding the nature of psychopathology with diverse populations to develop relevant tools, (b) uncovering shared mechanisms to target in the development of comprehensive measurement and treatment approaches, and (c) informing capacity building in low and middle-income countries that can leverage clinical science to build cost-efficient service systems.

We examined the structure of psychopathology in a sample comprising 971 adults from Maputo, Mozambique. Mozambique is one of the poorest countries in the world, falling in the “low development” category of the Human Development Index. We tested whether factor-analytic models of the structure of psychopathology generalized to a community sample of Mozambican adults. Confirmatory factor analysis (CFA) tested standard models of the structure of psychopathology using clinical interview and self-report measures of 15 psychiatric disorders. We hypothesized that models specifying a general psychopathology p factor and three specific lower-order factors, internalizing, substance use, and thought disorders, would fit the sample data well. We also conducted exploratory analyses to determine whether the factor loadings on p differed between men and women and to examine relations between the psychopathology factors and demographics, physical health, suicide risk, and functioning.

Method

Study location

The study was conducted in Maputo, the capital and most populous city in Mozambique. Data were collected at two primary care clinics and one hospital in Maputo City from May 16 to June 8, 2018. These facilities provide primary care, emergency, and outpatient mental health services. The hospital also provides services for victims of interpersonal violence and inpatient health and psychiatric services.

Participants

Participants included 989 adults who were enrolled in a larger study assessing the validity of mental health screening tools in Mozambican adults. Eligibility criteria included being 18 years of age or older and fluency in Portuguese, the official language of Mozambique (self-reported and interviewer confirmed). Patients and their accompaniers in health facility waiting rooms were invited to participate. All volunteers were taken to a private area for eligibility assessments and informed consent. Participants provided written informed consent as approved by the New York State Psychiatric Institute Institutional Review Board (No. 7479) and the Eduardo Mondlane University Institutional Health Bioethics Council (Institutional Committee on Bioethics for Health (CIBS) Faculty Medicine (FM) of University Eduardo Mondlane and Maputo Central Hospital (MCH)/54/2017). Of the 989 participants, 18 participants with complete nonresponse on symptom measures were excluded from the study (inclusion n = 971; 63% women; 98% Black; age: M = 32.01 years, SD = 11.25).

Measures

Table 1 shows descriptive statistics for all measures. The Mini International Neuropsychiatric Interview (MINI) Plus 4.0.0 (Brazilian version; Amorim, 2000; Sheehan et al., 1998) and a battery of mental health screening tools were administered to participants in a randomized order to assess 15 psychiatric disorders. Lovero et al. (2021) describes the original study procedures and measures. For all measures, we used existing Brazilian or Portuguese translations, and local research team members made minor adjustments for the Mozambican context (e.g., local terms for substances). For the present study, internalizing symptoms of depression, anxiety, somatization, posttraumatic stress disorder, and panic disorder were assessed using the nine-item Patient Health Questionnaire (PHQ-9; Cumbe et al., 2020; Kroenke & Spitzer, 2002), seven-item Generalized Anxiety Disorder (GAD-7) scale (Spitzer et al., 2006), eight-item Somatic Symptom Scale (SSS-8; Gierk et al., 2014), five-item Primary Care Post-Traumatic Stress Disorder Screen (PC-PTSD; Prins et al., 2016), and MINI Panic Disorder symptom count, respectively. We assessed externalizing symptoms of alcohol and cannabis use disorders using the 10-item Alcohol Use Disorders Identification Test (AUDIT; Atkins et al., 2021; Babor et al., 2001) and seven-item Alcohol, Smoking and Substance Involvement Screening Test (ASSIST; World Health Organization [WHO] ASSIST Working Group, 2002). In the original study, in the externalizing domain, only these substance use disorder measures were administered, and antisocial behavior was not assessed in any form; thus, we refer to this domain as substance use disorder symptoms, not externalizing, in this article. Thought disorder symptoms of mania and psychotic disorders were measured using the MINI Mania symptom count and the 20-item Psychosis Screening Questionnaire (PSQ) Thought Interference, Persecution, Strange Experiences, and Hallucinations subscales (Bebbington & Nayani, 1995).

Table 1.

Descriptive Statistics for All Study Variables

Variable	N (%)	Minimum	Maximum	M (SD)	α
Psychiatric symptoms
PHQ-9 Depression	963	0	25	4.36 (4.54)	.82
GAD-7 Anxiety	963	0	21	3.62 (3.77)	.83
SSS-8 Somatization	962	0	29	4.71 (4.91)	.79
PC-PTSD Symptoms	962	0	5	1.01 (1.55)	.84
MINI-4 Panic Symptoms	966	0	16	0.60 (2.57)	.97
ASSIST Cannabis Use	962	0	33	0.52 (3.36)	.87
ASSIST Alcohol Use	962	0	38	3.69 (7.29)	.86
AUDIT Consumption	962	0	12	1.76 (2.48)	.71
AUDIT Dependence	962	0	11	0.52 (1.59)	.70
AUDIT Related Problems	962	0	16	1.19 (2.59)	.66
PSQ Thought Interference	962	0	2	0.27 (0.63)	.85
PSQ Persecution	962	0	3	0.81 (1.20)	.90
PSQ Strange Experiences	962	0	2	0.46 (.76)	.80
PSQ Hallucinations	962	0	2	0.17 (0.53)	.86
MINI Mania Symptoms	966	0	11	1.06 (2.32)	.70
External criteria
Age (years)	971	18	75	32.01 (11.25)
Education (years)	971	0	39	10.77 (4.20)
Chronic medical illness	466 (48)
HIV diagnosis	304 (31)
Current mental health treatment	65 (7)
MINI Diagnoses	971	0	7	1.09 (1.43)
CSSRS Suicidality	949	0	7	0.48 (1.23)	.83
WHO-DAS Cognition	966	0	17	1.04 (2.04)	.69
WHO-DAS Mobility	966	0	13	0.81 (1.94)	.77
WHO-DAS Self-Care	966	0	9	0.32 (1.05)	.65
WHO-DAS Getting Along	965	0	12	0.87 (1.75)	.62
WHO-DAS Life Activities	965	0	21	1.06 (2.67)	.88
WHO-DAS Participation	964	0	20	1.95 (3.90)	.82

Note: PHQ-9 = Patient Health Questionnaire-9; GAD-7 = Generalized Anxiety Disorder-7; SSS-8 = Somatic Symptom Scale-8; PC-PTSD = Primary Care PTSD Screen for DSM-5; MINI-4 = Mini International Neuropsychiatric Interview Plus 4.0; ASSIST = Alcohol, Smoking and Substance Involvement Screening Test; AUDIT = Alcohol Use Disorders Identification Test; PSQ = Psychosis Screening Questionnaire; CSSRS = Columbia Suicide Severity Rating Scale; WHO-DAS = World Health Organization Disability Assessment Schedule 2.0.

Sum scores were calculated for each of the symptom measures. The MINI Plus interview employs skip-out rules (i.e., certain items are not administered if participants did not endorse initial screening items); this may lead to underreporting of symptoms. Therefore, MINI symptom counts were used only to assess disorders that were not otherwise measured by self-report questionnaires, which were not limited by skip-out rules. Because no self-report measures of panic disorder or mania were administered in the original study, we used MINI symptom count scores for these disorders. Specifically, we recoded missing values due to skip-outs as zeros and calculated sum scores of panic disorder and mania symptoms. Based on their distributions, the following sum scores were treated as categorical in subsequent analyses: the PSQ subscales, ASSIST Cannabis Use scale, and MINI Mania and Panic Disorder symptom counts (in the Supplemental Material available online, see the text and Fig. S1 for more details on variable scoring and distributions, respectively).

External criterion variables, those chosen to provide an initial test of construct validity, and demographic variables were also drawn from the sample. These variables reflected potential risk factors for psychopathology, including suicide risk measured with the six-item Columbia Suicide Severity Rating Scale (CSSR-S total sum score; Posner et al., 2011), disability and impairment measured with the 36-item WHO Disability Assessment Schedule (WHO-DAS) 2.0 (five subscales of Cognition, Mobility, Self-Care, Getting Along, Life Activities, and Participation; Federici et al., 2017), presence of chronic health conditions, and self-reported age, years of education, and gender.

Data were collected by Mozambican research assistants using tablets with Redcap software in the clinics. Participants were randomly assigned to the order of MINI versus to screening tools administration.

Statistical analyses

Confirmatory factor analyses

CFA was used to fit correlated-factors, higher-order, bifactor, and one-factor models of the structure of psychopathology (Fig. 1). The correlated-factors model tested the hypothesis that there are latent trait factors that influence a subset of symptoms. Here, we tested three factors representing Internalizing (with loadings from depression, anxiety, somatization, posttraumatic stress disorder, and panic disorder measures), Substance Use Disorders (with loadings from alcohol and cannabis use measures), and Thought Disorder (with loadings from psychosis and mania measures). This model allowed the Internalizing, Substance Use Disorders, and Thought Disorder factors to be correlated. The higher-order model had the same factor structure as the correlated-factors model except that a higher-order p factor was included with loadings on the lower-order Internalizing, Substance Use Disorders, and Thought Disorder factors. The bifactor model tested the hypothesis that the symptom measures reflected both the p factor and the three specific forms of psychopathology that are orthogonal to p. The one-factor model tested whether the specific factors were needed in a simple structural model that assigned each diagnostic symptom only to the p factor. Because there were few participants who endorsed cannabis use (n = 31), CFAs were run with and without cannabis use to determine whether results were robust to their inclusion or exclusion.

Fig. 1.

Four standard confirmatory factor models of the structure of psychopathology: (a) correlated-factors model, (b) higher-order factor model, (c) bifactor model, and (d) one-factor model. Circles are used to denote latent constructs (i.e., general psychopathology (p), Internalizing [INT], Substance Use Disorder [SUD], Thought Disorders [THT] factors), whereas squares are used to denote observed variables. PANIC = Mini International Neuropsychiatric Interview Plus 4.0 panic symptom count; PC-PTSD = Primary Care PTSD Screen for DSM-5; SOM = Somatic Symptom Scale; ANX = Generalized Anxiety Disorder-7; DEP = Patient Health Questionnaire-9; CAN = Alcohol, Smoking, and Substance Involvement Screening Test (ASSIST) Cannabis Use subscale; ALC = ASSIST Alcohol Use subscale; ALC-C = Alcohol Use Disorders Identification Test (AUDIT) Consumption subscale; ALC-D = AUDIT Dependence subscale; ALC-P = AUDIT Related Problems subscale; MANIA = MINI mania symptom count; PSY-H = Psychosis Screening Questionnaire (PSQ) Hallucinations subscale; PSY-S = PSQ Strange Experiences subscale; PSY-P = PSQ Persecution subscale; PSY-T = PSQ Thought Interference subscale.

CFAs were performed in Mplus Version 8.6 (Muthén & Muthén, 2017) using the weighted least squares means and variance adjusted (WLSMV) algorithm. The WLSMV estimator is appropriate for categorical and nonmultivariate normal data and provides consistent estimates when data are missing completely at random with respect to covariates (Asparouhov & Muthén, 2010). We computed standard errors using TYPE=COMPLEX in Mplus, which used a sandwich estimator to account for nonindependence of observations due to cluster sampling within the three study sites. We assessed each model’s fit to the data using the χ² value, the comparative fit index (CFI), the Tucker-Lewis index (TLI), and the root-mean-square error of approximation (RMSEA). Nonsignificant χ² tests indicate good model fit; nonetheless, this test is generally overpowered in large samples such as ours. CFI and TLI values greater than .90 indicate adequate fit; RMSEA scores less than .08 are considered acceptable (Kline, 2015).

Tests of measurement invariance by gender

Of the CFAs that provided adequate fit to the data, we conducted two-group CFAs by gender to determine whether factor loadings were equivalent for men and women. The χ² difference test, using the DIFFTEST function in Mplus, was used to compare a model with loadings equivalent between men and women to a model that allowed those loadings to be freely estimated. Given that the χ² value may be overly sensitive to minor deviations from a perfect model in large samples, we also considered change in alternate fit indices greater than .010 (i.e., RMSEA and CFI; Putnick & Bornstein, 2016).

Structural equation models

Using structural equation models (SEMs), we examined correlations between each latent psychopathology factor (p, Internalizing, Substance Use Disorder, Thought Disorder) and external criterion variables using WITH statements in Mplus. SEMs were conducted using WLSMV estimation and bootstrapping procedures with 1,000 samples (MacKinnon et al., 2004). Standardized estimates and bootstrapped confidence intervals are reported. Given the number of tests, we used p < .01 as the significance threshold.

Results

Confirmatory factor models

We first examined bivariate polychoric correlations among the symptom measures, which were positive and supported testing the four CFA models as planned (Table S1 in the Supplemental Material). The correlated-factors, higher-order, and bifactor models fit the data well (Table 2). The correlated-factors and higher-order models are equivalent and thus demonstrated equivalent fit indices and factor loadings, which were positive and statistically significant (p < .001; Table 3). The highest standardized loading on Internalizing, Substance Use Disorder, and Thought Disorder factors was depression (.740), alcohol-related problems (.845), and hallucinations (.823), respectively. Within the higher-order model, the highest standardized loading on p was for the Thought Disorder factor (.979), followed by the Internalizing (.728) and Substance Use Disorder (.417) factors. In the bifactor model, loadings were positive and statistically significant (p < .001), except for mania, which negatively and nonsignificantly loaded on the specific Thought Disorder factor (β = −0.165, p = .130). The one-factor model did not fit the data well, suggesting that comorbidity is poorly explained by one factor and that unique aspects of disorders are missed in a one-factor solution. Therefore, we did not conduct any further analyses with the one-factor model.

Table 2.

Goodness-of-Fit Statistics for the Four Standard Confirmatory Factor Models

Fit statistic	Correlated-factors model	Higher-order model	Bifactor model	One-factor model
χ²/df	238.364/87	238.379/87	118.498/75	434.344/90
CFI	.940	.940	.983	.864
TLI	.928	.928	.976	.842
RMSEA[90% CI]	.042[.036, .049]	.042[.036, .049]	.024[.016, .033]	.063[.057, .069]

Note: CFI = comparative fit index; TLI = Tucker-Lewis index; RMSEA = root-mean-square error of approximation; CI = confidence interval.

Table 3.

Standardized Factor Loadings for the Four Standard Confirmatory Factor Models

	Correlated-factors model			Higher-order model				Bifactor model				One-factor model
Measure	Internalizing factor	Substance Use Disorder factor	Thought Disorder factor	Internalizing factor	Substance Use Disorder factor	Thought Disorder factor	p factor	Internalizing factor	Substance Use Disorder factor	Thought Disorder factor	p factor	p factor
PHQ-9 Depression	.740			.740				.652			.564	.583
GAD-7 Anxiety	.726			.726				.565			.551	.623
SSS-8 Somatization	.650			.650				.603			.468	.528
PC-PTSD	.699			.699				.279			.570	.593
MINI Panic	.544			.544				.200			.471	.497
AUDIT Consumption		.323			.323				.804		.107	.110
AUDIT Dependence		.707			.707				.720		.285	.338
AUDIT Problems		.845			.845				.726		.331	.356
ASSIST Alcohol Use		.810			.810				.910		.298	.379
ASSIST Cannabis Use		.553			.554				.306		.463	.635
PSQ Thought Interference			.689			.689				.278	.654	.645
PSQ Persecution			.752			.752				.354	.659	.731
PSQ Strange Experiences			.753			.753				.582	.686	.679
PSQ Hallucinations			.824			.823				.225	.746	.677
MINI Mania			.662			.662				-.165	.665	.586
Internalizing							.728
Substance Use Disorder							.417
Thought Disorder							.979

Note: All factor loadings were significant at p < .001 except for the loading of the Thought Disorder factor on Mini International Neuropsychiatric Interview (MINI) Plus 4.0 Mania (p = .130). In the correlated-factors model, the correlations were as follows: Substance Use Disorder and Internalizing: r = .303, Thought Disorder and Internalizing: r = .712, Thought Disorder and Substance Use Disorder: r = .407. PHQ-9 = Patient Health Questionnaire-9; GAD-7 = Generalized Anxiety Disorder-7; SSS-8 = Somatic Symptom Scale-8; PC-PTSD = Primary Care PTSD Screen for DSM-5; AUDIT = Alcohol Use Disorders Identification Test; ASSIST = Alcohol, Smoking and Substance Involvement Screening Test; PSQ = Psychosis Screening Questionnaire.

Although the bifactor model fit the data well, we opted to focus subsequent analyses on the higher-order and correlated-factors model for two reasons: (a) These models do not force unique orthogonal factors in their structure yet still allow for examination of both general and unique sources of variance, and (b) these models are more consistent with the Hierarchical Taxonomy of Psychopathology model (Kotov et al., 2017). Results from measurement invariance testing and SEM analyses with the bifactor measurement model are included in the Supplemental Material (see Tables S2 and S3).

Measurement invariance by gender

Two-group correlated-factors and higher-order models allowing factor loadings to be freely estimated by gender were compared with models that restricted factor loadings to be equivalent between men (n = 364) and women (n = 607). For both the two-group correlated-factors and higher-order models, the χ² difference tests suggested that restricting factor loadings to be equivalent by gender resulted in a significant decrement in fit compared with models allowing factor loadings to be freely estimated (Table 4). However, the change in RMSEA and CFI between the two-group correlated-factors models estimated freely and constrained by gender was small (ΔRMSEA = .001, ΔCFI = .003), suggesting that the significant χ² difference test may have reflected only a small deviation in model fit. Alternatively, ΔRMSEA = .010 and ΔCFI = .019 suggest that the two-group higher-order model constrained by gender did in fact result in a significant decrement in model fit compared with a model freely estimated by gender. In the two-group higher-order model freely estimated for men and women, men showed a higher Substance Use Disorder factor loading on p than did women (Table 5; see also Fig. S2 in the Supplemental Material). Both women and men showed similar Internalizing and Thought Disorder factor loadings on p. These findings demonstrate that the differences in factor loadings between men and women were driven by loadings on the p factor rather than the Internalizing, Substance Use Disorder, or Thought Disorder factors.

Table 4.

Test of Measurement Invariance by Gender of the Two-Group Correlated-Factors and Higher-Order Models

	Test of overall fit			Fit indices relative to prior model
Model	χ²/df	RMSEA	CFI	Δχ²	Δdf	ΔRMSEA	ΔCFI
Correlated factors
Gender free	297.657***/191	.034	.977
Gender constrained	328.200***/206	.035	.974	54.142***	15	.001	.003
Higher-order
Gender free	297.657***/191	.034	.977
Gender constrained	403.874***/209	.044	.958	128.286***	18	.010	.019

Note: Two-group correlated-factors and higher-order models for gender were conducted to determine whether the factor loadings were equivalent for men and women. The difference in χ² between models with equivalent loadings between men and women (“gender constrained”) and models that allowed those loadings to be freely estimated (“gender free”) were tested using the function DIFFTEST in Mplus. Tests of overall model fit and fit indices relative to the prior model are shown. The gender-constrained models fit significantly worse than the gender-free models. As a result, Table 5 reports the standardized factor loadings from the gender-free two-group higher-order model. RMSEA = root-mean-square error of approximation; CFI = comparative fit index.

***

p < .001.

Table 5.

Standardized Factor Loadings From the Gender-Free Two-Group Higher-Order Model

	Men				Women
Measure	Internalizing	Substance Use Disorder	Thought Disorder	p	Internalizing	Substance Use Disorder	Thought Disorder	p
PHQ-9 Depression	.760				.844
GAD-7 Anxiety	.782				.774
SSS-8 Somatization	.697				.684
PC-PTSD	.781				.694
MINI-4 Panic	.667				.527
AUDIT Consumption		.253				.608
AUDIT Dependence		.565				.804
AUDIT Problems		.653				.877
ASSIST Alcohol Use		.759				.815
ASSIST Cannabis Use		.645				.363
PSQ Thought Interference			.688				.743
PSQ Persecution			.629				.763
PSQ Strange Experiences			.800				.790
PSQ Hallucinations			.772				.718
MINI Mania			.615				.726
Internalizing				.864				.882
Substance Use Disorder				.543				.363
Thought Disorder				.732				.732

Note: The factor loadings from the two-group correlated-factors model were the same as the loadings from the two-group higher-order model, therefore they are not shown here. PHQ-9 = Patient Health Questionnaire-9; GAD-7 = Generalized Anxiety Disorder-7; SSS-8 = Somatic Symptom Scale-8; PC-PTSD = Primary Care PTSD Screen for DSM-5; MINI-4 = Mini International Neuropsychiatric Interview Plus 4.0; AUDIT = Alcohol Use Disorders Identification Test; ASSIST = Alcohol, Smoking and Substance Involvement Screening Test; PSQ = Psychosis Screening Questionnaire.

Relationships with external criterion variables

We used SEMs to test relationships between observed external criterion variables and p, Internalizing, Substance Use Disorder, and Thought Disorder factors separately for men and women. We used the higher-order model for examining relations between external criterion variables and p; for testing relations with Internalizing, Substance Use Disorder, and Thought Disorder factors, we employed the correlated-factors model. We used the factor loadings freely estimated for men and women in the two-group models and fixed those loadings in the SEMs.

In both men and women, we found that higher levels of p, Internalizing, and Thought Disorder factors were associated with a greater number of MINI diagnoses, increased suicidality, and poorer functioning in all domains (Table 6; see also Fig. S3 in the Supplemental Material). The Substance Use Disorder factor was largely unrelated to these external criterion variables, except for a positive association with the number of MINI diagnoses.

Table 6.

Relations Between Transdiagnostic Psychopathology Factors and External Criterion Variables

	Men								Women
	p		Internalizing factor		Substance Use Disorder factor		Thought Disorder factor		p		Internalizing factor		Substance Use Disorder factor		Thought Disorder factor
External criterion variable	β	95% CI	β	95% CI	β	95% CI	β	95% CI	β	95% CI	β	95% CI	β	95% CI	β	95% CI
Age (years)	-0.188	[-0.294, -0.071]	-0.083	[-0.174, 0.012]	-0.208	[-0.345, -0.066]	-0.206	[-0.328, -0.071]	-0.049	[-0.147, 0.044]	-0.015	[-0.103, 0.073]	-0.085	[-0.166, -0.015]	-0.053	[-0.165, 0.048]
Education (years)	0.047	[-0.078, 0.153]	0.073	[-0.043, 0.174]	-0.097	[-0.239, 0.042]	0.070	[-0.051, 0.194]	-0.033	[-0.131, 0.076]	-0.011	[-0.089, 0.073]	0.009	[-0.094, 0.129]	-0.094	[-0.188, 0.008]
Chronic medical illness	0.070	[-0.096, 0.239]	0.128	[-0.012, 0.269]	-0.047	[-0.231, 0.163]	0.025	[-0.161, 0.186]	0.224	[0.102, 0.344]	0.216	[0.110, 0.318]	0.006	[-0.108, 0.121]	0.190	[0.062, 0.308]
HIV diagnosis	-0.146	[-0.348, 0.025]	-0.092	[-0.274, 0.050]	-0.260	[-0.475, -0.063]	-0.021	[-0.228, 0.157]	0.033	[-0.102, 0.161]	-0.018	[-0.136, 0.100]	0.031	[-0.086, 0.142]	0.088	[-0.045, 0.220]
Current mental health treatment	0.195	[-0.018, 0.355]	0.244	[0.033, 0.410]	−0.146	[-0.580, 0.094]	0.186	[-0.066, 0.395]	0.206	[-0.022, 0.373]	0.187	[-0.010, 0.336]	0.076	[-0.267, 0.237]	0.140	[-0.109, 0.325]
No. of MINI diagnoses	0.830	[0.715, 0.934]	0.674	[0.529, 0.804]	0.581	[0.404, 0.749]	0.619	[0.521, 0.702]	0.812	[0.721, 0.899]	0.661	[0.567, 0.751]	0.356	[0.229, 0.488]	0.617	[0.553, 0.676]
CSSRS Suicidality	0.398	[0.241, 0.548]	0.426	[0.262, 0.573]	0.022	[-0.140, 0.219]	0.322	[0.193, 0.441]	0.551	[0.412, 0.679]	0.473	[0.358, 0.576]	0.302	[0.147, 0.474]	0.296	[0.218, 0.368]
WHO-DAS Cognition	0.447	[0.291, 0.588]	0.467	[0.310, 0.597]	0.116	[-0.082, 0.313]	0.329	[0.216, 0.435]	0.590	[0.488, 0.695]	0.607	[0.493, 0.725]	0.108	[0.024, 0.202]	0.365	[0.283, 0.449]
WHO-DAS Mobility	0.411	[0.244, 0.560]	0.408	[0.259, 0.550]	0.069	[-0.148, 0.287]	0.317	[0.187, 0.439]	0.455	[0.342, 0.558]	0.486	[0.376, 0.589]	0.085	[-0.003, 0.194]	0.243	[0.151, 0.328]
WHO-DAS Self-Care	0.332	[0.156, 0.517]	0.320	[0.160, 0.478]	0.088	[-0.112, 0.323]	0.279	[0.164, 0.386]	0.368	[0.242, 0.487]	0.374	[0.247, 0.496]	0.051	[-0.024, 0.139]	0.203	[0.100, 0.298]
WHO-DAS Getting Along	0.382	[0.245, 0.508]	0.429	[0.283, 0.560]	-0.009	[-0.176, 0.170]	0.270	[0.160, 0.368]	0.458	[0.354, 0.550]	0.502	[0.391, 0.602]	-0.037	[-0.109, 0.043]	0.280	[0.186, 0.355]
WHO-DAS Life Activities	0.403	[0.262, 0.536]	0.439	[0.295, 0.574]	-0.069	[-0.185, 0.054]	0.207	[0.097, 0.302]	0.429	[0.329, 0.532]	0.422	[0.322, 0.532]	0.024	[-0.036, 0.087]	0.254	[0.169, 0.332]
WHO-DAS Participation	0.546	[0.406, 0.672]	0.590	[0.448, 0.731]	0.102	[-0.062, 0.275]	0.393	[0.283, 0.494]	0.541	[0.435, 0.636]	0.571	[0.470, 0.674]	0.098	[0.007, 0.197]	0.342	[0.261, 0.425]

Note: Estimates that are significant at p < .01 are given in boldface. We used the higher-order model for examining relations between external criterion variables and p; for testing relations with Internalizing, Substance Use Disorders, and Thought Disorder factors, we employed the correlated-factors model. Higher scores on the World Health Organization Disability Assessment Schedule 2 (WHO-DAS) subscales reflect poorer functioning in each domain. CI = confidence interval; MINI = Mini International Neuropsychiatric Interview; CSSRS = Columbia Suicide Severity Rating Scale.

In terms of gender-specific results, men who were high in p, Substance Use Disorder, and Thought Disorder factors were younger; men who were high in Substance Use Disorder were less likely to have an HIV diagnosis. Women who were high in p, Internalizing, and Thought Disorder factors were more likely to have a chronic medical illness. Women, but not men, who were high in Substance Use Disorder exhibited increased suicidality.

Discussion

In a sample of 971 Mozambican adults, we tested four standard confirmatory factor models of the structure of psychopathology including Internalizing, Substance Use Disorder, and Thought Disorder factors as well as a general psychopathology p factor. This is the first examination of the structure of psychopathology in an African sample. We tested four previously established models of the structure of psychopathology (Caspi et al., 2014; Lahey et al., 2017). Consistent with prior research, the correlated-factors, higher-order factor, and bi-factor models fit the data well. The one-factor model did not fit well, suggesting that disorder comorbidity is poorly explained by one factor because there are unique aspects of disorders missed in a one-factor solution. Significant gender differences manifested in expected directions, with a stronger loading of the Substance Use Disorder factor on p in men than women but no differences in the factor loadings of Internalizing or Thought Disorder on p between the genders. Higher levels of p, Internalizing, and Thought Disorder factors were associated with increased suicidality, poorer functioning, and greater comorbidity. Results indicate that the previously observed structure of psychopathology fits within this Mozambican sample.

These findings align with those of factor-analytic studies demonstrating that the structure of psychopathology is relatively stable across countries (de Jonge et al., 2018). A study examining the structure of psychopathology with factor-analytic modeling in Brazil, Colombia, Romania, Poland, and Peru showed that latent factors accounted for shared disorder variance across countries (de Jonge et al., 2018). This mirrors the broader literature in the United States (Ringwald et al., 2021) and global south, such as in Chile (Ignatyev et al., 2019) and Brazil (Martel et al., 2017).

Men had a stronger loading of the Substance Use Disorder factor on p than women. Sex and gender are regarded as important determinants for mental health partly because of the established differential distribution between sexes (Shannon et al., 2019). Although a full discussion on differences in psychopathology of men and women is beyond the scope of this article (see Hartung & Lefler, 2019, for a full discussion of this issue), we will briefly note a few hypotheses that may be relevant for this context. For women, high exposure to adverse life events, such as gender-based violence, increases chances of living with an internalizing disorder (Kuehner, 2017). For men, there are often vital social implications related to alcohol use that reinforce drinking as a means of social engagement, gaining reputational status, and coping (Uy et al., 2014). This may contribute to disproportionate rates of drinking among men and increased likelihood of alcohol-related problems (Rehm et al., 2009). Gender inequity also may play a role, as a 15-country study demonstrated that as equity improved, internalizing and externalizing disorders were more similar in expression among men and women (Seedat et al., 2009).

The patterns of associations between the external criterion variables and psychopathology factors largely generalized across the factors, except for the Substance Use Disorder factor. One explanation for this discrepancy is that of the externalizing disorders, substance use disorders tend to have had lower loadings on the p factor in prior studies of Western adult samples (e.g., Caspi et al., 2014; Romer et al., 2018; Romer & Pizzagalli, 2022). Indeed, unexpectedly, in both men and women, the Substance Use Disorder factor was not associated with other medical conditions or functioning, and it was unrelated to suicidality in men. In this sample, the Substance Use Disorder factor primarily consisted of alcohol use. Other literature has shown that alcohol use is a leading risk factor of mortality among men and has numerous health consequences (Griswold et al., 2018). One explanation for our findings is that the prevalence of alcohol and substance use was low and may not be representative of Mozambicans with substance use disorder symptoms. Given that help seeking is often low among people with substance use problems, those who do seek help may present with different patterns of functioning compared with those who do not seek care (Halsted et al., 2019). Additionally, men who were high on the Substance Use Disorder factor were less likely to have a positive HIV diagnosis. This finding also may be related to a unique aspect of this help-seeking sample given that other literature has shown an association between alcohol use and incident HIV infection (Woolf-King et al., 2013).

Service-system implications

Hierarchical modeling studies, including this study, may be a step toward better assessing and treating the global burden of mental illness through transdiagnostic approaches. Understanding underlying psychopathology dimensions can inform parsimonious services targeting shared disorder features. A comprehensive, dimensional service approach that is scalable is especially useful for public health and aligns with real-world clinical practice. Mozambique, which faces a critical shortage of health resources, has also been at the forefront of adopting innovations to meet mental health needs, such as formalizing task shifting within their national health platform, making dimensional approaches particularly valuable and possible (Fortunato dos Santos et al., 2016). Despite these potential benefits, a review found that one fifth of transdiagnostic interventions were not truly transdiagnostic, suggesting that their application can pose challenges (Fusar-Poli et al., 2019). Understanding the nuances of when to adopt dimensional versus disorder-specific approaches needs further research.

Another area for development is assessment. Research groups developing dimensional assessments have suggested two ways forward. One is a stepwise approach starting with brief screening of higher-order psychopathology before specialized assessment (Lovero et al., 2021). The second is using categorical elements in a dimensional frame to classify symptoms on the basis of severity through comparison with the population mean (Conway et al., 2019). To the latter, in the larger Mozambican study, our team developed a stepwise tool starting with three questions assessing the high-order presence or absence of psychopathology (Lovero et al., 2021). For participants with psychopathology, nine questions assessed and classified lower-order categories: anxiety, depression, traumatic stress reactions (internalizing), substance use disorders, and severe mental illness (thought); classification guided treatment selection. We plan to explore the association of this tool with p and lower-order factors.

Limitations and future directions

This study is a step toward a more diverse understanding of hierarchical models; however, limits exist that can inform future work. First, the data were cross-sectional, curtailing any ability to examine trajectories or sequencing of comorbidities overtime. Longitudinal studies of the structure of psychopathology in African countries are needed to better understand trajectories of psychopathology. This includes exploring the presence of a p factor in youth and across development; examining the predictive ability of p in youth may further inform preventative interventions. Second, we were not able to assess a true externalizing factor because our measures of externalizing psychopathology were limited to substance use disorders. Although we had sufficient indicators to conduct analyses, we do not know whether our results would have differed if we had included measures of antisocial and attention-deficit/hyperactivity disorder symptoms. It will be important to include a range of externalizing symptoms in future studies. Related, demographically, the study did not distinguish between sex and gender; only binary gender options were included. Inclusion of additional locally relevant gender options would provide rich data in future studies.

Another consideration is that measures used were originally developed in Western contexts and may not reflect all symptoms indicative of psychological distress in the Mozambican population. How to balance standardized assessment methods with more culturally specific measurement methods is a tension that many studies grapple with (Bemme & D’souza, 2014). Although standard confirmatory factor models fit the data well in this sample, we were limited to Western measures adapted to Mozambique versus fully idiographic measures of distress (for examples of such measurement, see Kaiser et al., 2013; Osborn et al., 2021; Watson et al., 2020). Much more work is needed to understand explanatory models of mental health in Mozambique, which limited the ability to include such assessment tools in the study. Future work would benefit from the inclusion of more ethnographic exploration to understand conceptions of distress salient to individuals in Mozambique and explore its relationships to these hierarchical models.

This tension presents an important empirical question for further exploration regarding whether and how cultural concepts of distress may or may not map onto latent psychiatric factors (Kohrt et al., 2014). These questions can be tested using similar factor-modeling approaches to investigate how and whether idiographic measures of distress fit within existing factor models. Blending hierarchical modeling with ethnographic measurement has the potential to yield assessments that function to best assess indicators of psychopathology while reflecting the cultural form in which illness may be expressed or shaped. Such a goal still aligns with and can inform efforts to develop briefer screening measures of latent constructs that are rooted in rigorous modeling approaches as well as culturally grounded conceptions of distress. For instance, endorsement of an idiom of distress, such as “thinking too much”—a common idiom in sub-Saharan Africa (Backe et al., 2021)—may correlate with p or a latent factor just as highly as endorsement of more traditionally tested items.

Summary

This study is the first examination of the structure of psychopathology and a general psychopathology p factor in an African sample, the long-term goal of which is to help develop services to better meet mental health needs. Continued epidemiological and clinical psychological science research is needed in Mozambique, the African continent, and low and middle-income countries to advance our understanding of psychopathology and the practical implications of transdiagnostic psychopathology factors, such as p. An opportunity exists to leverage cutting-edge psychopathology research to best meet mental health needs. Through the use of growing systems, such as those in Mozambique, researchers are poised to adopt innovative approaches to mental health services that can advance the field and meet population needs.

Supplemental Material

sj-docx-1-cpx-10.1177_21677026221122773 – Supplemental material for Examination of the Factor Structure of Psychopathology in a Mozambican Sample

Supplemental material, sj-docx-1-cpx-10.1177_21677026221122773 for Examination of the Factor Structure of Psychopathology in a Mozambican Sample by Ali Giusto, Adrienne Romer, Kathryn Lovero, Palmira Fortunado dos Santos, Claire Greene, Lidia Gouveia, Antonio Suleman, Paulino Feliciano, Maria A. Oquendo, Jennifer Mootz and Milton L. Wainberg in Clinical Psychological Science

Footnotes

Acknowledgements

A. Giusto and A. Romer share first authorship of this article. All of the authors thank the Mozambican Ministry of Health, National Directorate of Public Health, Maputo City Health Department, Nampula Province Health Department, and research assistants for facilitating data collection and all study participants for their contribution to this work.

Transparency

Action Editor: Aidan G. C. Wright

Editor: Jennifer L. Tackett

Author Contributions

Ali Giusto: Conceptualization, methodology, writing – original draft, writing – review & editing.

Adrienne Romer: Conceptualization, formal analysis, writing – original draft, writing – review & editing.

Kathryn Lovero: data curation, methodology, writing – original draft, writing – review & editing.

Palmira Fortunado dos Santos: conceptualization, project administration, writing – original draft, writing – review & editing.

Claire Greene: data curation, methodology, writing – original draft, writing – review & editing.

Lidia Gouveia: investigation, project administration, writing – review & editing.

Antonio Suleman: investigation, project administration, writing – review & editing.

Paulino Feliciano: investigation, project administration, writing – review & editing.

Maria A. Oquendo: investigation, methodology, project administration, writing – review & editing.

Jennifer Mootz: investigation, project administration, writing – original draft, writing – review & editing.

Milton L. Wainberg: conceptualization, funding acquisition, investigation, methodology, project administration, writing – original draft, writing – review & editing.

ORCID iDs

Ali Giusto

Adrienne Romer

Supplemental Material

Additional supporting information can be found at

References

Achenbach

T. M.

Edelbrock

C. S.

(1981). Behavioral problems and competencies reported by parents of normal and disturbed children aged four through sixteen. Monographs of the Society for Research in Child Development, 46, 1–82. https://doi.org/10.2307/1165983

Amorim

(2000). Mini International Neuropsychiatric Interview (MINI): Validation of a short structured diagnostic psychiatric interview. Brazilian Journal of Psychiatry, 22, 106–115.

Asparouhov

Muthén

(2010). Weighted least squares estimation with missing data. http://www.statmodel.com/download/GstrucMissingRevision.pdf

Atkins

D. L.

Cumbe

V. F.

Muanido

Manaca

Fumo

Chiruca

Hicks

Wagenaar

B. H.

(2021). Validity and item response theory properties of the Alcohol Use Disorders Identification Test for primary care alcohol use screening in Mozambique (AUDIT-MZ). Journal of Substance Abuse Treatment, 127, Article 108441. https://doi.org/10.1016/j.jsat.2021.108441

Babor

T. F.

Higgins-Biddle

J. C.

Saunders

J. B.

Monteiro

M. G.

(2001). AUDIT: The Alcohol Use Disorders Identification Test: Guidelines for use in primary care (2nd ed.). World Health Organization.

Backe

E. L.

Bosire

E. N.

Kim

A. W.

Mendenhall

(2021). “Thinking too much”: A systematic review of the idiom of distress in sub-Saharan Africa. Culture, Medicine, and Psychiatry, 45, 655–682. https://doi.org/10.1007/s11013-020-09697-z

Bebbington

Nayani

(1995). The Psychosis Screening Questionnaire. International Journal of Methods in Psychiatric Research, 5(1), 11–19.

Bemme

D’souza

N. A.

(2014). Global mental health and its discontents: An inquiry into the making of global and local scale. Transcultural Psychiatry, 51(6), 850–874.

Caspi

Houts

R. M.

Ambler

Danese

Elliott

M. L.

Hariri

Harrington

Hogan

Poulton

Ramrakha

Rasmussen

L. J. H.

Reuben

Richmond-Rakerd

Sugden

Wertz

Williams

B. S.

Moffitt

T. E.

(2020). Longitudinal assessment of mental health disorders and comorbidities across 4 decades among participants in the Dunedin birth cohort study. JAMA Network Open, 3, Article e203221. https://doi.org/10.1001/jamanetworkopen.2020.3221

10.

Caspi

Houts

R. M.

Belsky

D. W.

Goldman-Mellor

S. J.

Harrington

Israel

Meier

M. H.

Ramrakha

Shalev

Poulton

Moffitt

T. E.

(2014). The p factor: One general psychopathology factor in the structure of psychiatric disorders? Clinical Psychological Science, 2(2), 119–137. https://doi.org/10.1177/2167702613497473

11.

Conway

C. C.

Forbes

M. K.

Forbush

K. T.

Fried

E. I.

Hallquist

M. N.

Kotov

Mullins-Sweatt

S. N.

Shackman

A. J.

Skodol

A. E.

South

S. C.

Sunderland

Waszczuk

M. A.

Zald

D. H.

Afzali

M. H.

Bornovalova

M. A.

Carragher

Docherty

A. R.

Jonas

K. G.

Krueger

R. F.

. . . Eaton

N. R.

(2019). A hierarchical taxonomy of psychopathology can transform mental health research. Perspectives on Psychological Science, 14, 419–436. https://doi.org/10.1177/1745691618810696

12.

Cumbe

V. F. J.

Muanido

Manaca

M. N.

Fumo

Chiruca

Hicks

de Jesus Mari

Wagenaar

B. H.

(2020). Validity and item response theory properties of the Patient Health Questionnaire-9 for primary care depression screening in Mozambique (PHQ-9-MZ). BMC Psychiatry, 20, Article 382. https://doi.org/10.1186/s12888-020-02772-0

13.

de Jonge

Wardenaar

K. J.

Lim

C. C. W.

Aguilar-Gaxiola

Alonso

Andrade

L. H.

Bunting

Chatterji

Ciutan

Gureje

Karam

E. G.

Lee

Medina-Mora

M. E.

Moskalewicz

Navarro-Mateu

Pennell

B.-E.

Piazza

Posada-Villa

Torres

. . . Scott

. (2018). The cross-national structure of mental disorders: Results from the World Mental Health Surveys. Psychological Medicine, 48, 2073–2084. https://doi.org/10.1017/S0033291717003610

14.

Federici

Bracalenti

Meloni

Luciano

J. V.

(2017). World Health Organization Disability Assessment Schedule 2.0: An international systematic review. Disability and Rehabilitation, 39, 2347–2380. https://doi.org/10.1080/09638288.2016.1223177

15.

Fonagy

Campbell

Constantinou

Higgitt

Allison

Luyten

(2021). Culture and psychopathology: An attempt at reconsidering the role of social learning. Development and Psychopathology. Advance online publication. https://doi.org/10.1017/S0954579421000092

16.

Fortunato dos Santos

P. F.

Wainberg

M. L.

Caldas-de-Almeida

J. M.

Saraceno

de Jesus Mari

. (2016). Overview of the mental health system in Mozambique: Addressing the treatment gap with a task-shifting strategy in primary care. International Journal of Mental Health Systems, 10, Article 1. https://doi.org/10.1186/s13033-015-0032-8

17.

Fusar-Poli

Solmi

Brondino

Davies

Chae

Politi

Borgwardt

Lawrie

S. M.

Parnas

McGuire

(2019). Transdiagnostic psychiatry: A systematic review. World Psychiatry, 18, 192–207.

18.

Gierk

Kohlmann

Kroenke

Spangenberg

Zenger

Brähler

Löwe

(2014). The Somatic Symptom Scale–8 (SSS-8): A brief measure of somatic symptom burden. JAMA Internal Medicine, 174, 399–407.

19.

Griswold

M. G.

Fullman

Hawley

Arian

Zimsen

S. R. M.

Tymeson

H. D.

Venkateswaran

Tapp

A. D.

Forouzanfar

M. H.

Salama

J. S.

Abate

K. H.

Abate

Abay

S. M.

Abbafati

Abdulkader

R. S.

Abebe

Aboyans

Abrar

M. M.

Acharya

. . . Gakidou

(2018). Alcohol use and burden for 195 countries and territories, 1990–2016: A systematic analysis for the Global Burden of Disease Study 2016. The Lancet, 392, 1015–1035. https://doi.org/10.1016/S0140-6736(18)31310-2

20.

Halsted

Ásbjörnsdóttir

K. H.

Wagenaar

B. H.

Cumbe

Augusto

Gimbel

Manaca

Manuel

J. L.

Sherr

, & with input from the INCOMAS Study Team (2019). Depressive symptoms, suicidal ideation, and mental health care-seeking in central Mozambique. Social Psychiatry and Psychiatric Epidemiology, 54, 1519–1533. https://doi.org/10.1007/s00127-019-01746-2

21.

Hartung

C. M.

Lefler

E. K.

(2019). Sex and gender in psychopathology: DSM–5 and beyond. Psychological Bulletin, 145, 390–409.

22.

J. J.

(2021). Factorial invariance in hierarchical factor models of mental disorders in African American and European American youths. Journal of Child Psychology and Psychiatry, 62, 289–298. https://doi.org/10.1111/jcpp.13243

23.

Ignatyev

Baggio

Mundt

A. P.

(2019). The underlying structure of comorbid mental health and substance use disorders in prison populations. Psychopathology, 52, 2–9. https://doi.org/10.1159/000495844

24.

Kaiser

B. N.

Kohrt

B. A.

Keys

H. M.

Khoury

N. M.

Brewster

A.-R. T.

(2013). Strategies for assessing mental health in Haiti: Local instrument development and transcultural translation. Transcultural Psychiatry, 50, 532–558. https://doi.org/10.1177/1363461513502697

25.

Kane

J. C.

Vinikoor

M. J.

Haroz

E. E.

Al-Yasiri

Bogdanov

Mayeya

Simenda

Murray

L. K.

(2018). Mental health comorbidity in low-income and middle-income countries: A call for improved measurement and treatment. The Lancet Psychiatry, 5, 864–866.

26.

Kline

R. B.

(2015). Principles and practice of structural equation modeling. Guilford Press.

27.

Kohn

Saxena

Levav

Saraceno

(2004). The treatment gap in mental health care. Bulletin of the World Health Organization, 82, 858–866.

28.

Kohrt

B. A.

Rasmussen

Kaiser

B. N.

Haroz

E. E.

Maharjan

S. M.

Mutamba

B. B.

de Jong

J. T.

Hinton

D. E.

(2014). Cultural concepts of distress and psychiatric disorders: Literature review and research recommendations for global mental health epidemiology. International Journal of Epidemiology, 43, 365–406. https://doi.org/10.1093/ije/dyt227

29.

Kotov

Krueger

R. F.

Watson

Achenbach

T. M.

Althoff

R. R.

Bagby

R. M.

Brown

T. A.

Carpenter

W. T.

Caspi

Clark

L. A.

Eaton

N. R.

Forbes

M. K.

Forbush

K. T.

Goldberg

Hasin

Hyman

S. E.

Ivanova

M. Y.

Lynam

D. R.

Markon

. . . Zimmerman

(2017). The Hierarchical Taxonomy of Psychopathology (HiTOP): A dimensional alternative to traditional nosologies. Journal of Abnormal Psychology, 126(4), 454–477.

30.

Kroenke

Spitzer

R. L.

(2002). The PHQ-9: A new depression diagnostic and severity measure. SLACK.

31.

Krueger

R. F.

Markon

K. E.

(2006). Reinterpreting comorbidity: A model-based approach to understanding and classifying psychopathology. Annual Review of Clinical Psychology, 2, 111–133. https://doi.org/10.1146/annurev.clinpsy.2.022305.095213

32.

Kuehner

(2017). Why is depression more common among women than among men? The Lancet Psychiatry, 4, 146–158.

33.

Lahey

B. B.

Applegate

Hakes

J. K.

Zald

D. H.

Hariri

A. R.

Rathouz

P. J.

(2012). Is there a general factor of prevalent psychopathology during adulthood? Journal of Abnormal Psychology, 121, 971–977. https://doi.org/10.1037/a0028355

34.

Lahey

B. B.

Krueger

R. F.

Rathouz

P. J.

Waldman

I. D.

Zald

D. H.

(2017). A hierarchical causal taxonomy of psychopathology across the life span. Psychological Bulletin, 143, 142–186. https://doi.org/10.1037/bul0000069

35.

Lovero

K. L.

Basaraba

Khan

Suleman

Mabunda

Feliciano

dos Santos

Fumo

Mandlate

Greene

M. C.

Fiks Salem

Mootz

J. J.

Mocumbi

A. O.

Duarte

C. S.

Gouveia

Oquendo

M. A.

Wall

M. M.

Wainberg

M. L.

(2021). Brief screening tool for stepped-care management of mental and substance use disorders. Psychiatric Services, 72(8), 891–897. https://doi.org/10.1176/appi.ps.202000504

36.

MacKinnon

D. P.

Lockwood

C. M.

Williams

(2004). Confidence limits for the indirect effect: Distribution of the product and resampling methods. Multivariate Behavioral Research, 39, 99–128.

37.

Martel

M. M.

Pan

P. M.

Hoffmann

M. S.

Gadelha

do Rosário

M. C.

Mari

J. J.

Manfro

G. G.

Miguel

E. C.

Paus

Bressan

R. A.

(2017). A general psychopathology factor (P factor) in children: Structural model analysis and external validation through familial risk and child global executive function. Journal of Abnormal Psychology, 126(1), 137–148.

38.

McGrath

J. J.

Lim

C. C. W.

Plana-Ripoll

Holtz

Agerbo

Momen

N. C.

Mortensen

P. B.

Pedersen

C. B.

Abdulmalik

Aguilar-Gaxiola

Al-Hamzawi

Alonso

Bromet

E. J.

Bruffaerts

Bunting

de Almeida

J. M. C.

de Girolamo

De Vries

Y. A.

Florescu

. . . de Jonge

(2020). Comorbidity within mental disorders: A comprehensive analysis based on 145 990 survey respondents from 27 countries. Epidemiology and Psychiatric Sciences, 29, Article e153. https://doi.org/10.1017/S2045796020000633

39.

Muthén

L. K.

Muthén

B. O.

(2017). Mplus user’s guide (8th ed.).

40.

Osborn

T. L.

Kleinman

Weisz

J. R.

(2021). Complementing standard western measures of depression with locally co-developed instruments: A cross-cultural study on the experience of depression among the Luo in Kenya. Transcult Psychiatry, 58, 499–515. https://doi.org/10.1177/13634615211000555

41.

Posner

Brown

G. K.

Stanley

Brent

D. A.

Yershova

K. V.

Oquendo

M. A.

Currier

G. W.

Melvin

G. A.

Greenhill

Shen

(2011). The Columbia–Suicide Severity Rating Scale: Initial validity and internal consistency findings from three multisite studies with adolescents and adults. American Journal of Psychiatry, 168, 1266–1277.

42.

Prins

Bovin

M. J.

Smolenski

D. J.

Marx

B. P.

Kimerling

Jenkins-Guarnieri

M. A.

Kaloupek

D. G.

Schnurr

P. P.

Kaiser

A. P.

Leyva

Y. E.

(2016). The Primary Care PTSD screen for DSM-5 (PC-PTSD-5): Development and evaluation within a veteran primary care sample. Journal of General Internal Medicine, 31, 1206–1211.

43.

Putnick

D. L.

Bornstein

M. H.

(2016). Measurement invariance conventions and reporting: The state of the art and future directions for psychological research. Developmental Review, 41, 71–90.

44.

Rehm

Mathers

Popova

Thavorncharoensap

Teerawattananon

Patra

(2009). Global burden of disease and injury and economic cost attributable to alcohol use and alcohol-use disorders. The Lancet, 373, 2223–2233.

45.

Ringwald

W. R.

Forbes

M. K.

Wright

A. G.

(2021). Meta-analysis of structural evidence for the Hierarchical Taxonomy of Psychopathology (HiTOP) model. Psychological Medicine. Advance online publication. https://doi.org/10.1017/S0033291721001902

46.

Romer

A. L.

Knodt

A. R.

Houts

Brigidi

B. D.

Moffitt

T. E.

Caspi

Hariri

A. R.

(2018). Structural alterations within cerebellar circuitry are associated with general liability for common mental disorders. Molecular Psychiatry, 23, 1084–1090.

47.

Romer

A. L.

Pizzagalli

D. A.

(2022). Associations between brain structural alterations, executive dysfunction, and general psychopathology in a healthy and cross-diagnostic adult patient sample. Biological Psychiatry Global Open Science, 2, 17–27.

48.

Sauer-Zavala

Gutner

C. A.

Farchione

T. J.

Boettcher

H. T.

Bullis

J. R.

Barlow

D. H.

(2017). Current definitions of “transdiagnostic” in treatment development: A search for consensus. Behavior Therapy, 48, 128–138. https://doi.org/10.1016/j.beth.2016.09.004

49.

Seedat

Scott

K. M.

Angermeyer

M. C.

Berglund

Bromet

E. J.

Brugha

T. S.

Demyttenaere

De Girolamo

Haro

J. M.

Jin

(2009). Cross-national associations between gender and mental disorders in the World Health Organization World Mental Health Surveys. Archives of General Psychiatry, 66, 785–795.

50.

Shannon

Jansen

Williams

Cáceres

Motta

Odhiambo

Eleveld

Mannell

(2019). Gender equality in science, medicine, and global health: Where are we at and why does it matter? The Lancet, 393, 560–569.

51.

Sheehan

D. V.

Lecrubier

Sheehan

K. H.

Amorim

Janavs

Weiller

Hergueta

Baker

Dunbar

G. C.

(1998). The Mini-International Neuropsychiatric Interview (MINI): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry, 59, 22–33.

52.

Spitzer

R. L.

Kroenke

Williams

J. B.

Löwe

(2006). A brief measure for assessing generalized anxiety disorder: The GAD-7. Archives of Internal Medicine, 166, 1092–1097.

53.

P. J.

Massoth

N. A.

Gottdiener

W. H.

(2014). Rethinking male drinking: Traditional masculine ideologies, gender-role conflict, and drinking motives. Psychology of Men & Masculinity, 15(2), 121–128.

54.

Wainberg

M. L.

Lovero

K. L.

Duarte

C. S.

Fiks Salem

Mello Milena Bezuidenhout

Mootz

Feliciano

Suleman

Fortunato dos Santos

Weissman

M. M.

Cournos

Marques

A. H.

Fumo

Mabunda

Alves-Bradford

J.-M. E.

Mello Marcelo Mari

J. J.

Ngwepe

Cidav

Mocumbi

A. O.

. . . Oquendo

M. A.

(2021). Partnerships in research to implement and disseminate sustainable and scalable evidence-based practices (PRIDE) in Mozambique. Psychiatric Services, 72, 802–811. https://doi.org/10.1176/appi.ps.202000090

55.

Wainberg

M. L.

Scorza

Shultz

J. M.

Helpman

Mootz

J. J.

Johnson

K. A.

Neria

Bradford

J.-M. E.

Oquendo

M. A.

Arbuckle

M. R.

(2017). Challenges and opportunities in global mental health: A research-to-practice perspective. Current Psychiatry Reports, 19, Article 28. https://doi.org/10.1007/s11920-017-0780-z

56.

World Health Organization ASSIST Working Group. (2002). The Alcohol, Smoking and Substance Involvement Screening Test (ASSIST): Development, reliability and feasibility. Addiction, 97, 1183–1194

57.

Watson

L. K.

Kaiser

B. N.

Giusto

A. M.

Ayuku

Puffer

E. S.

(2020). Validating mental health assessment in Kenya using an innovative gold standard. International Journal of Psychology, 55, 425–434.

58.

Woolf-King

S. E.

Steinmaus

C. M.

Reingold

A. L.

Hahn

J. A.

(2013). An update on alcohol use and risk of HIV infection in sub-Saharan Africa: Meta-analysis and future research directions. The International Journal of Drug and Alcohol Research, 1(2), 99–110. https://doi.org/10.7895/ijadr.v2i1.45

59.

Wright

A. G.

Krueger

R. F.

Hobbs

M. J.

Markon

K. E.

Eaton

N. R.

Slade

(2013). The structure of psychopathology: Toward an expanded quantitative empirical model. Journal of Abnormal Psychology, 122, 281–294.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.55 MB