Preventing comorbidity between distress and suicidality: a network analysis

Abstract

Suicidality among individuals between 10 and 35 years of age may be poised to exert massive burdens on society through decreased economic productivity and increased incidence of chronic physical conditions in the individuals' later years, thereby necessitating early prevention of suicide. While research suggests that the pathway to suicidality may begin from episodes of psychological distress, such pathway may involve complex interplays between intermediary psychiatric symptoms and external stimuli that are not easily delineated through conventional means. This study applies the network approach to psychopathology to elucidate this complexity. Comorbidity between psychological distress and suicidality in 1968 community-dwelling individuals is analyzed with regularized partial correlation networks to identify their bridge symptoms and links. Temporal relationships between symptoms are analyzed through temporal symptom network formed from 453 individuals who completed subsequent follow-up surveys. Network analysis shows that feelings of hopelessness and the presence of suicidal ideation are the strongest bridge symptoms in the comorbidity symptom network, and form the only prominent link that bridges psychological distress and suicidality. Effects of sleep troubles, anxiety, and poor social relationships on suicidal ideation appear to be mediated by hopelessness. The same observations hold among individuals with and without diagnoses of psychiatric disorders, as well as young people (10-24 year-olds) and young adults (25-35 year-olds). The edge between hopelessness and suicidal ideation remains the strongest bridge link after controlling for effects of symptoms from the previous time point. Findings here provide an evidence base for both professional training in caregiving professions as well as gatekeeper training in community members to emphasize more on how to effectively recognize hopelessness, and instill hope, in young people and young adults for various types of distress.

Keywords: Psychiatric disorders; Public health; Risk factors; Social sciences.

Fig. 1. Distress-suicidality symptom network constructed from responses of 1968 individuals in the baseline 2018 survey.

Edge thickness represents the magnitude of the partial correlation between symptoms. Blue nodes and orange nodes represent suicidality and distress symptoms respectively. All edge weights are positive.

Fig. 2. Values of centrality indices in the full-sample symptom network.

CHQ1 Headache, CHQ2 Heart problems, CHQ3 Chest problems, CHQ4 Numbness, CHQ5 Sleep troubles, CHQ6 Carrying too much burden, CHQ7 Losing confidence, CHQ8 Nervous and tense, CHQ9 Worry about family & friends, CHQ10 Life is hopeless, CHQ11 Poor relationship with family & friends, CHQ12 Not hopeful about future, SIDAS1 Frequency of suicidal thoughts, SIDAS2 Control over suicidal thoughts, SIDAS3 Closeness to attempting suicide, SIDAS4 Tormented by suicidal thoughts, SIDAS5 Interference in daily activities due to suicidal thoughts.

Fig. 3. Temporal network of distress and suicidality symptoms constructed from responses of 453 individual over surveys in 2018, 2019, and 2020.

Network structures of distress and suicidality symptoms constructed from responses of 453 individual over surveys in 2018, 2019, and 2020: (A) lag-1 temporal network, (B) contemporaneous network, and (C) between-subjects network. Green and red edges denote positive and negative correlations respectively. Edge thickness reflects the magnitude of correlation. Blue nodes and orange nodes represent suicidality and distress symptoms respectively.