Poster #14 - Loneliness or Fatigue? Gender Difference in Initial Symptom of Adolescent Depression: Cross-Lagged Panel Network Analysis

• In Event: Poster Session 20

Abstract

Background: Adolescence is sensitive developmental period for depression (Leussis & Andersen, 2008), and approximately 34% of adolescents reported depressive symptoms (Shorey et al., 2022). According to the network theory, individual depressive symptoms interplay with each other to form depression (Borsboom & Cramer, 201). The identification of depressive symptom structure, especially the symptoms with great influences on other symptoms (i.e., central output symptoms) could help understand the core mechanisms underlying the development of depression and shed light on targeted interventions (Borsboom, 2017). Gender differences in depression also emerge in adolescence. Girls have shown a higher prevalence of depression and higher severity of specific depressive symptoms than boys (Hyde & Mezulis, 2020; Mullarkey et al., 2019). Although gender differences in depression have been addressed in some epidemiological studies (e.g., Girgus & Yang, 2015), whether and how girls and boys differ in the symptom structure of depression have been largely unexplored. Thus, our study aimed to explore the longitudinal interplay of depressive symptoms in adolescence and gender difference in the longitudinal network of depression by applying a prospective Cross-Lagged Panel Network (CLPN) analysis.
Methods: Participants (N = 2524, Mage = 12.30, 45.8% girls, at T1) were extracted from the China Family Panel Study (Institute of Social Science Survey, 2019; http://www.isss.pku.edu.cn/cfps), a nationally representative survey of Chinese children and adolescents. They self-reported depressive symptoms using Epidemiological Studies Depression Scale (CES-D-8; Radloff, 1977) in 2016 (T1) and 2018 (T2). The CES-D-8 used 8 items to specify the depression: felt depressed, everything was an effort, sleep was restless, felt happy (reverse-scored), felt lonely, felt sad, enjoyed life (reverse-scored), and could not get ‘going’. Data analyses were conducted in R version 4.2.0 (RStudio 2022.02.3). Longitudinal symptom structure was estimated with the CLPN model (Epskamp, 2020), comprised of estimating auto-regressive and cross-lagged coefficients between depressive symptoms, and out- and in-expected influence (EI) centrality of each symptom. Gender differences were examined for symptom centrality, edge strength, and network density (Funkhouser et al., 2021).
Results: In general, controlling for all other symptoms and demographic covariates, “lonely” and “not get going (fatigue)” at T1 were most predictive of other symptoms at T2 and reported high out-EI (Estimate = .44 and .36, respectively); whereas symptoms related to “depressed” and “happy (anhedonia)” at T2 were easily predicted by other symptoms at T1 and reported high in-EI (Estimate = .27 both). Additionally, “lonely” with the highest out-EI (Estimate = .61) was considered the most initial symptom for girls whereas “fatigue” was for boys (Estimate = .29). Girls presented more and stronger bidirectional associations between depressive symptoms and had denser network connectivity than boys, implying girls’ greater vulnerability to the persistence and exacerbation of depression.
Conclusions: Interpersonal loneliness exerts the strongest influence on girls whilst somatic fatigue was most critical for boys, revealing the necessity of prioritizing the two central symptoms in early interventions for different genders. Girls’ interpersonal-centered depression and boys’ somatic-centered depression suggest gender-specific characteristics in the development and persistence of adolescent depression, and possibly enlighten the targeted intervention programs.

Authors

• Sihan Liu, Beijing Normal University

• Haining Ren, Arizona State University
  Presenting Author

• Yijia Li

• Sinan Fu, Faculty of Psychology, Beijing Normal University, China

• Zhuo Rachel Han, Beijing Normal University