Identification of Key Depressive Symptoms in Pregnant Women: Simulating Interventions

Symptoms with the highest strength centrality in mental disorders are regarded as potential intervention targets, and their suitability as intervention targets remains uncertain. This study employs the NodeIdentifyR algorithm (NIRA), a novel network intervention method, to examine how alleviating or aggravating specific symptoms affects the network's sum scores. A total of 701 pregnant women from the U.S. National Health and Nutrition Examination Survey (2003-2023) were included. Depressive symptoms were assessed using the Patient Health Questionnaire-9, and pregnancy status was confirmed via urine tests or self-report. The Ising model was used to construct the symptom network, and NIRA was applied to simulate both alleviating and aggravating interventions. The results revealed that the centrality analysis identified "guilt or low self-worth" as the central symptom with the highest strength value. Alleviating interventions targeting "fatigue" had the greatest reduction in symptom network activation, resulting in the largest decrease in the projected symptom sum score by 52.53%, indicating it as a potential target for alleviating depressive symptoms. Aggravating interventions targeting "guilt or low self-worth" had the greatest increase in the severity of depressive symptoms, causing the largest increase in the symptom sum score by 70.97%, suggesting its potential as a prevention target. Aggravating interventions caused greater overall changes in symptom activation than alleviating interventions, with both affecting the same symptoms in different ways. Notably, "fatigue" is primarily associated with physiological and other somatic factors rather than emotional symptoms, highlighting the need for cautious interpretation of fatigue-targeted interventions.