Factor structure and validity of the Inventory of Depression and Anxiety Symptoms-II (IDAS-II) in a chronic back pain treatment-seeking sample.

Abstract

Many patients who seek treatment for chronic back pain are also at a higher risk of having comorbid anxietyand depression-related disorders. Measures of mood and anxiety are routinely used in medical settings to screen for depression-and anxiety-related symptoms. However, factor analyses of other measures of mood and anxiety in medical settings often detect a somatization factor which, in turn, limits their discriminant validity for use across medical settings. The Inventory of Depression and Anxiety Symptoms-II (IDAS-II) is a comprehensive self-report inventory that assesses varying aspects of mood and anxiety. The purpose of this investigation is to examine the three-factor structure and validity of the IDAS-II in a chronic pain treatment-seeking sample. A total of 169 patients completed the IDAS-II and the Minnesota Multiphasic Personality Inventory-2-Restructured Form (MMPI-2-RF) upon admission. Confirmatory factor analyses were computed using the scales of the IDAS-II and zero-order correlations between the IDAS-II factors from the best-fitting model and scale scores of the MMPI-2-RF. Overall, a three-factor structure of the IDAS-II was not supported; instead, a one-factor solution fit best. Using the MMPI-2-RF as external criteria, the one-factor of the IDAS-II correlated highest with the Somatic Complaints scale and the Demoralization scale. Overall, item content on the IDAS-II shares overlap with many symptoms that patients with chronic pain likely would endorse. Discussion about implications of using mood/anxiety measures and models in medical settings that are in line with the Hierarchical Taxonomy of Psychopathology (HiTOP) are discussed. Public Significance StatementMood and anxiety measures and screeners are regularly used in medical settings. This study suggests that measures of mood and anxiety may evidence poor discriminant validity due to somatization that is prevalent in many medical populations.

The Brazilian version of the Inventory of Depression and Anxiety Symptoms–Expanded Version: Internal structure, invariance by sex and race, and associations with pathological personality traits.

“…De la Rosa-Cáceres et al (2020) also observed a second-order factor for the internalizing symptoms. Finally, contrary to all previous studies, Kaye et al (2022) observed a one-factor structure for the inventory. Also, it is important to mention that Cervin et al (2024) found a four-factor solution for a youth sample, which will not be tested in this study, given our focus on adults.…”

“…We performed restricted factor models with half of the sample, replicating the findings from Cervin et al (2024), De la Rosa-Cáceres et al (2020), Irak and Albayrak (2020), Kaye et al (2022), Watson et al (2012), and Wester et al (2022), Petre et al (2023). We also performed models that added a general factor (hierarchical models) for the models found by Irak and Albayrak (2020), Watson et al (2012), and Wester et al (2022), Petre et al (2023).…”

“…Given our large sample size, we split the data set into two random halves using the semTools R package (Jorgensen et al, 2022) to test and cross-validate the factor structure of the Brazilian IDAS-II. With the first half ( N = 1,190), we tested 12 restricted factor models using the 18 IDAS-II scales: (a) one factor (as observed in Kaye et al, 2022), (b) three original factors (Watson et al, 2012—three factors with two cross-loadings), (c) three factors with the Appetite Gain scale relocated (as in Irak & Albayrak, 2020), (d) three factors with three cross-loadings (as in Wester et al, 2022; Petre et al, 2023), (e) three factors without the appetite gain scale and with three cross-loadings (as in Cervin et al, 2024), and (f) a hierarchical model with three first-order factors and one second-order factor (De la Rosa-Cáceres et al, 2020), plus another four models adding a second-order factor to Models b, c, d, and e to verify the plausibility of a hierarchical model with a second-order internalizing factor representing the HiTOP model structure. We also tested two more restricted models derived from an exploratory structural equation modeling (ESEM) and a hierarchical structural equation modeling (Morin & Asparouhov, 2018).…”

“…Also, it is important to mention that Cervin et al (2024) found a four-factor solution for a youth sample, which will not be tested in this study, given our focus on adults. presents all models for adult samples, except Kaye et al’s (2022) one-factor solution.…”

Differences in presurgical MMPI-3 scores across trajectories of recovery from spine surgery.

“…Although these findings are significant, they are limited by recent research indicating that screening instruments, such as the PHQ, are not specific to depression but are actually indicators of general psychological distress rather than major depressive disorder (Katz et al, 2021; McCord & Provost, 2020). Owing to high comorbidity between psychopathology and chronic pain, instruments used in presurgical psychological evaluations (PPEs) must have high levels of sensitivity, specificity, and discriminant validity for the specific constructs that predict recovery (Block et al, 2013), and some instruments have not met these standards (e.g., Kaye et al, 2022). Whereas many instruments assess negative emotionality, the MMPI-3 includes well-validated scales that measure demoralization (i.e., general distress similar to that measured by the PHQ; McCord & Provost, 2020), low positive emotionality (i.e., anhedonia), negative emotionality, as well as relevant narrower constructs such as malaise, neurological complains, family problems, substance abuse, and anger proneness with good discriminant validity (Ben-Porath & Tellegen, 2020a, 2020b).…”