Construct validity of the COVID-19 Peritraumatic Distress Index (CPDI) in a Peruvian sample during COVID-19 lockdown and its relationship with anxiety and depression using structural equation modeling

Abstract

Background: The main aims of this study were to adapt the COVID-19 Peritraumatic Distress Index (CPDI) to the Peruvian population and to establish a model explaining depression using CPDI values and anxiety symptoms during the COVID-19 lockdown. Finally, we sought predictive values of the obtained CPDI factors for depression and anxiety as a secondary aim. Materials and Methods: An exploratory factor analysis (n = 300) was performed, followed by confirmatory factor analysis in a second phase (n = 1135). To explain depression scores during the COVID-19 pandemic, we performed structural equation modeling (SEM). Finally, we performed a hierarchical regression model (HRM) to evaluate the amount of explained variance of the CPDI factors above depression, anxiety, and sociodemographic variables. Results: A 2-factor solution (ruminationand stress) for the CPDI (p < 0.001; CFI = 0.99) was found. Concerning the SEM, our model was able to explain 81% of the depression scores (p < 0.001; CFI = 0.98). Finally, in the HRM, rumination could explain 17% additional variance in depression (p < 0.001) and 28% in anxiety (p < 0.001). However, stress showed collinearity with depression and anxiety, not continuing for further HRM analysis. Conclusions: Our results showed a 2-factor solution for the CPDI. Moreover, our SEM model showed that female sex, younger age, and incomplete education (with high COVID-related stress and anxiety) lead to more depression symptoms during the COVID-19 lockdown. Finally, our HRM showed that people who frequently ruminate during the COVID-19 lockdown are more afraid and negatively affected.

Cross-Cultural Validation in Times of COVID-19: An Example Using the COVID-19 Peritraumatic Distress Index (CPDI) among Spanish and Peruvian Populations

“…Participants and procedure present study is a secondary analysis of two databases comprising online survey studies conducted in Peru and Spain during the COVID-19 pandemic. Certain ndings from this survey study have already been reported elsewhere 11,14 . The current study incorporates certain methodological aspects of both studies, primarily pertaining to the recruitment of participants, the description of the online survey, and the CPDI.…”

“…However, since a correlation between the two factors was assumed, a principal component analysis with a promax rotation was performed in this study. Items 7, 8, and 11 were removed from the analysis 14 . When the datasets were combined and analyzed, the same number of factors and a similar distribution of items per factor were obtained (see Table 3).…”

“…Table 3 CPDI items and eliminated items in the Spanish study 11 , Peruvian study 14 , and in the secondary analysis of the combined data. Items that are consistently in the factor are underlined, while the eliminated items in the different studies are written in italics.…”

“…Both surveys were conducted anonymously, and participants were only permitted to complete the survey once. The surveys included questions regarding socio-economic status (i.e., age, gender, education), psychometric data (CPDI scale) 11,14 , and data of past medical history, extracting, in this case, the most frequent disorders in both populations (respiratory diseases, hypertension, hypercholesterinemia, diabetes mellitus, cardiovascular diseases, and anxious-depressive disorders).…”

“…These symptoms include negative cognitive changes, avoidance, compulsive behavior, physical symptoms related to stress, social withdrawal, loss of social functioning, anxiety, and depressive symptoms. CPDI has been validated during the COVID-19 lockdown in different languages worldwide [9][10][11][12] , including European 11 and Latin American Spanish 13,14 . Although both Spanish validation studies showed a bifactorial solution, they showed variations in the items included in each factor and the interpretation of the factors.…”