The Dimensional Structure of Internalizing Psychopathology: Relation to Diagnostic Categories

Abstract

Recent approaches have aimed to represent the dimensional structure of psychopathology, but relatively little research has rigorously tested subdimensions within internalizing psychopathology. Using adult samples harmonized across three sites ( N = 427), this study tested preregistered models of the dimensional structure of internalizing psychopathology and their relations with current and lifetime depressive and anxiety disorders diagnostic data. Across S-1 bifactor and hierarchical models, we found converging evidence for both general and specific internalizing dimensions. Depression, generalized anxiety disorder (GAD), social anxiety disorder (SAD), and panic attacks were all associated with a general internalizing factor that we posit primarily represents motivational anhedonia. GAD was also associated with a specific anxious apprehension factor, and SAD was associated with specific anxious apprehension and low positive affect factors. We suggest that dimensional approaches capturing shared and specific internalizing symptom facets more accurately describe the structure of internalizing psychopathology and provide useful alternatives to categorical diagnoses to advance clinical science.

Combining dimensional models of internalizing symptoms and repetitive negative thought: Systematic replication, model comparison, and external validation.

“…To investigate the specificity of dimension–disorder relationships, we estimated latent correlations between the factors in the bifactor model and the disorder status variables (see Table 3 and Figure 3). As hypothesized, common internalizing was significantly correlated with MDD, GAD, and ASPD, MDD: β ( SE ) = .45 (.04), p < .001; GAD: β ( SE ) = .48 (.05), p < .001; ASPD: β ( SE ) = .28 (.05), p < .001, confirming previous research employing MDD and GAD diagnoses (Snyder et al, 2022), but the association between the common internalizing factor and ASPD symptoms is novel, suggesting that the common internalizing factor is capturing covariance across indicators that is also relevant to externalizing psychopathology. Indeed, any covariance driven by a general p-factor affecting both internalizing and externalizing disorders should have been partitioned into the common internalizing factor in the current report.…”

“…Replicating Internalizing Symptom Models From Snyder et al (2022) The correlated factors, bifactor, and hierarchical factors parameterizations described by Snyder and colleagues fit the data well (correlated factors: RMSEA = .037, CFI = .964; bifactor: RMSEA = .038, CFI = .963; hierarchical factors: RMSEA = .039, CFI = .960; see Table 2 and Figure 1). Each parameterization provided insights into the latent structure of internalizing psychopathology.…”

“…The preliminary aim of this study was to determine if the internalizing symptoms and repetitive negative thought models described by Snyder et al (2022) and Hur et al (2017) fit within the current sample (for descriptive statistics see Table 1; for correlations between measures see Table S1 in the online supplemental materials). However, prior to model comparisons, we evaluated the degree to which the foundational models were invariant between the sexes.…”

“…To determine how to account for sex in all analyses, we carried out preliminary invariance tests between the sexes in the foundational models central to this report. Tests of configural and scalar invariance were conducted for the bifactor and correlated factors/single factor models from Snyder et al (2022) and Hur et al (2017) as metric invariance tests were not an option with ordinal indicators (Muthén & Muthén, 1998-2017. As the foundational models were found to be noninvariant between the sexes, we did not test for sex differences beyond these invariance tests.…”

“…The current analyses compare this full measurement model to simpler alternatives, with an emphasis on distinguishing between common internalizing and repetitive negative thought. If common internalizing and repetitive negative thought are indeed distinct constructs, then general features of internalizing psychopathology may be better explained by multiple general dimensions, not a single dimension as has often been the case (e.g., Caspi et al, 2014;Snyder et al, 2022) Instead, such a finding would align with theoretical models of internalizing psychopathology that posit distinct somatic and cognitive factors across both depression and anxiety (e.g., Arnau et al, 2001;Ree et al, 2008). If, on the other hand, common internalizing and repetitive negative thought are not distinct constructs, then it suggests commonalities across internalizing psychopathology may be ascribed to a single general factor, not multiple factors.…”