The current study expands the result of the depressive-anxiety symptom network analysis among Chinese adolescents. Moreover, a focus on suicide ideation was added to this network. We investigated which depressive and anxiety symptoms are contributing factors of suicide ideation and compared the difference in comorbidity network structure between adolescents who had suicide ideation and who did not.
Network analysis of depressive-anxiety symptoms demonstrated that psychomotor symptoms were the central and bridge symptoms. “Restlessness” and “Trouble relaxing” was the most prominent central symptoms in this adolescent’s network, while “Restlessness”, “Nervousness”, and “Motor” were the bridge symptoms. Another depressive-anxiety symptom network analysis conducted during the COVID-19 outbreak and the after-peak stages in China showed similar results (Wang et al., 2020). The study indicated that psychomotor symptoms including motor, restlessness, and trouble relaxing maintained the most central symptoms in the network between the outbreak and after the peak. During COVID-19, Chinese middle school students take online courses at home due to temporary regulations of social distance or lockdown. Body-related activities of school-aged adolescents may be significantly reduced during this period since there are no physical education classes, no need to commute to school and no recess to play with classmates. Psychomotor symptoms have a unique significance in the pathophysiology, diagnosis, and therapy of depression (Schrijvers et al., 2008). Meanwhile, core symptoms of anxiety are associated with psychomotor functions, such as restlessness (Spitzer et al., 2006). Since we also recruited the adolescents after the peak stage, this could explain why “Motor” in depressive symptoms or “Restlessness” and “Trouble relaxing” in anxiety symptoms may increase the risk of activation of other depressive and anxiety symptoms in our sample. However, the previous network analysis of depressive and anxiety symptoms in Chinese adolescents did not show the same results (Cai et al., 2022; Liu et al., 2022). In contrast to prior studies, we excluded adolescents who did not have symptoms of depression and anxiety (PHQ-9 < 5 and GAD-7 < 5). The network structures of low-symptom adolescents and high-symptom adolescents were statistically different (Osborn et al., 2020). Our finding is consistent with previous findings of a comorbidity network of depressive and anxiety symptoms in an inpatient sample (Kaiser et al., 2021). In addition, we found a strong link between two psychomotor symptoms in anxiety clusters, which is “Restlessness” and “Trouble relaxing”. This finding is also consistent with the results of depressive and anxiety symptoms network analysis in Chinese patients with epilepsy (Wei et al., 2021). A similar network structure result was found in Chinese adolescents, which demonstrated that the association between 'Restlessness' and 'Trouble relaxing' is relatively strong in both the first and second surveys of depressive-anxiety symptom networks (Liu et al., 2022). Therefore, psychomotor symptoms were more likely to activate other symptoms in our depressive-anxiety symptom network.
"Sad Mood" is another central symptom in this network, which is consistent with the findings of (Cai et al., 2022) in their study of Chinese adolescent depressive and anxiety networks. Moreover, Similar results were found among adolescents in different countries. For example, a network analysis of depressed teenagers in the US showed that “Sad” was one of the most central symptoms and stay stable over the developmental period from 5 to 14 years old (McElroy et al., 2018). This finding indicated that “Sad Mood” is likely to be the key symptom to maintain the depressive-anxiety symptoms network among adolescents. For adolescents in Sub-Saharan Africa, “depressed mood” was one of the most central depression symptoms in the depressive-anxiety symptoms network (Osborn et al., 2020). These studies demonstrate that “Sad Mood” as the most central symptom of adolescent depression and anxiety networks have cross-cultural generalizability. Apart from community samples, similar results were found in large psychiatric samples (Beard et al., 2016; Kaiser et al., 2021). Sad mood represents a negative affect state, which is a hallmark for a diagnosis of major depression based on DSM-5 and ICD-10. Sad mood has been proven to be superior in assessing the risk of an episode or relapse of depression compared to other depressive symptoms. For example, network analysis demonstrated that “Sad mood” was more central than the other 28 depressive symptoms in a large, depressed outpatients’ sample (Fried et al., 2016). The same results have been demonstrated in a longitudinal study with a community sample of adolescents (Georgiades et al., 2006). According to the concept of centrality in network analysis, "Sad mood" may activate other symptoms in the depressive-anxiety network. Therefore, this finding supports that the " Sad mood" is a risk factor for anxiety and depression.
Furthermore, “Sad Mood” was the central node between suicidal ideation and other depressive-anxiety symptoms, which was directly related to suicidal ideation. Moreover, “Sad Mood” was a central symptom in the SI group network, which was less important in the non-SI group network. Adolescents with suicidal ideation who experience a sad mood may be at an increased risk for activation of other anxiety-depressive symptoms. Meanwhile, as a unique bridge symptom in the network of adolescents with suicidal ideation, “Sad Mood” may increase the risk of anxiety. Suicidal ideation in adolescence may develop from a mildly depressed mood in response to acute stressors to a generalized sad mood (Prager, 2009). Based on the mood activation hypotheses and cognitive theory, vulnerable individuals in a sad mood are less likely to show protective bias in their attentional functioning (McCabe et al., 2000). This suggests that those vulnerable adolescents are more likely to neglect positive and neutral content words and bias toward negative-content words when they are in a sad mood (Lau et al., 2004). Meanwhile, sad mood may induce maladaptive cognitions, which are associated with suicidal ideation during depressive episodes (Antypa et al., 2010). Therefore, prevention and treatment of suicidal ideation could focus on reducing the maladaptive cognitions triggered by sad moods. Mindfulness-based therapy has been found to be effective in improving negativity bias (Raes et al., 2009). For example, mindfulness-based stress reduction has shown effectiveness in reducing suicidal ideation, anxiety, and depression (Serpa et al., 2014).
Compared with the non-SI group, “Restlessness” was the most central symptom in the SI group network, which corresponds with the most prominent central symptoms identified in the comorbidity network. Agitated depression is a subtype of major depressive disorder (MDD), with the co-occurrence of depression and psychomotor agitation in an episode (Spitzer et al., 1978). One of the excitatory psychomotor symptoms is restlessness (Akiskal et al., 2005). As a typical symptom of agitated depression, suicidal ideation could be independently predicted by psychomotor activation during MDD (Akiskal et al., 2005). Meanwhile, restlessness is a core symptom of anxiety disorders. Adolescents with anxiety disorders may exhibit emotional processing biases, such as interpreting ambiguous situations as threatening or being attracted by threatening information, which may lead them to experience more negative arousal states including restlessness and trouble relaxing (Osborn et al., 2020). This emotional-processing bias for negative stimuli is also a characteristic of suicidal ideation (Harfmann et al., 2019). Therefore, restlessness may increase the risk of activation of other anxiety and depression symptoms in adolescents with suicide ideation. Physical therapy and pharmacologic treatment, such as rTMS, ziprasidone and haloperidol, were found to be effective in reducing psychomotor symptoms (Citrome et al., 2004; Heath et al., 2018). With the decrease of the central symptom in the depressive-anxiety network, suicidal ideation may also be alleviated.
There are several limitations in this study. First, suicide ideation was measured by one question in PHQ-9 rather than a suicidal ideation scale. Second, it is a cross-sectional study to estimate the undirect network. As these edges do not show a causal relationship between symptoms, it reduces the effectiveness of application in clinical intervention and treatment. Researchers could conduct a longitudinal study with a suicidal ideation scale to investigate the causal relationship between suicide ideation and other symptoms of anxiety and depression in the future.