Temporal stability of depression symptom networks across cultures: Longitudinal analysis of MIDUS and MIDJA

The network approach to psychopathology conceptualizes mental disorders as systems with causally interacting symptoms. While the temporal stability of symptom networks has been examined in single-culture samples, no research has compared stability patterns across cultural contexts.

Using longitudinal data from the Midlife in the United States (MIDUS; N = 673) and Midlife in Japan (MIDJA; N = 197), we estimated depression networks using four CES-D subscales (Depressed Affect, Positive Affect, Somatic Symptoms, Interpersonal Problems) at two time points (T1 and T2). Temporal stability was assessed using the Network Comparison Test and correlations between edge weights and centrality indices.

Depression networks demonstrated high temporal stability in both samples, with edge weight correlations of r = 0.945 (MIDUS) and r = 0.902 (MIDJA) and centrality correlations of r = 0.957 (MIDUS) and r = 0.864 (MIDJA). Positive Affect showed moderate negative connections to other subscales in the American sample but was functionally independent in the Japanese sample across both time points. The Depressed Affect-Somatic connection was the strongest edge in both cultures but notably stronger in Japan.

The MIDJA sample was relatively small, measurement intervals differed between samples, and findings may not generalize beyond these specific midlife populations.

Depression symptom networks show high temporal stability within cultures while exhibiting stable cross-cultural differences in specific-symptom connections. The functional independence of positive affect in Japanese samples has implications for cross-cultural assessments and culturally adapted interventions.