A longitudinal network analysis of interaction factors among Chinese women at high risk for perinatal depression

• Published in: Midwifery Vol. 139; p. 104187
• Main Authors: Huang, Jun, Lin, Yiyang, Fu, Yanqing, Xu, Zelin, Hong, Huilan, Arbing, Rachel, Chen, Wei-Ti, Wang, Anni, Huang, Feifei
• Format: Journal Article
• Language: English
• Published: Scotland Elsevier Ltd 01.12.2024
• Subjects: Central symptoms; Health ecological model; Longitudinal data; Network analysis; Perinatal depression; Health ecological model; Longitudinal data; Central symptoms; Network analysis; Perinatal depression
• ISSN: 0266-6138; 1532-3099; 1532-3099
• DOI: 10.1016/j.midw.2024.104187

Few studies have applied a health ecological model to understand perinatal depression among high-risk women, and existing research remains primarily cross-sectional in nature. This study aimed to explore the interplay among family function, perceived stress, insomnia symptoms, cognitive reactivity subscales (such as hopelessness/suicidality, aggression, control/perfectionism, avoidant coping, and acceptance/coping), mindfulness subscales (including attention, present focus, awareness, and acceptance), physiological indicators (e.g., hgb, 25-hydroxyvitamin D, and HbA1C), and depressive symptoms in Chinese high-risk women during the perinatal period. This was a longitudinal population-based cohort study. This two-wave prospective study was conducted in Fujian Province, China, from December 2021 to January 2023. We used convenience sampling to enroll 368 pregnant patients from obstetrical clinics and inpatient departments of three tertiary hospitals (level 3) in Fuzhou and Quanzhou City, Fujian Province, China. In the statistical analysis, cross-sectional data were analyzed via the contemporaneous network method, and longitudinal data were analyzed via the cross-lagged panel network method. The core symptoms in the depression-related symptom network during the third trimester and three months postpartum were identified as attention (ATT) (strength = 1.02) and acceptance/coping (ACC) (strength = 1.19). All bridge symptoms were shown as depression (EPDS) (bridging strength = 0.07 and 0.09). A comparison between the first and second survey networks showed a reduced edge weight for the association between depressive symptoms and insomnia symptoms (to 0 in the second survey network, diff = -0.18, P < 0.001). Conversely, the association between depressive symptoms and control/perfectionism increased to 0.252 (diff = 0.25, P < 0.001). Through cross-lagged panel network analysis, the EPDS (out strength = 3.68, OEI =3.60) was identified as the most influential symptom and the most predictable symptom (R² = 0.76). Perceived stress (PSS) (in strength = 2.49) and hopelessness/suicidality (HOP) (IEI = 1.96) were identified as the most susceptible symptoms. Cross-sectional network analysis combined with longitudinal network analysis revealed the mechanism of action between symptoms. Attention (ATT) and acceptance/coping (ACC) were identified as the core symptoms in the network of depression-related symptoms during the third trimester and three months postpartum, and the bridge symptoms were both depression (EPDS). In the dynamic network, depression (EPDS) was identified as the most influential and predictable symptom, and perceived stress (PSS) and hopelessness/suicidality (HOP) were identified as the most susceptible symptoms. Targeted interventions focused on attention and coping can reduce stress during pregnancy and enhance postpartum well-being. Strengthening family support and routine screening for symptoms such as stress and depression (EPDS) are crucial for improving maternal mental health globally, particularly in resource-limited settings.