Centrality of symptoms in networks of depression and comorbid psychopathology among adults aged 50 years and older: A systematic review and examination of influencing factors
Abstract
Background: Network analysis has increasingly been adopted over the past decade to study the structure of depressive symptoms in middle-aged to older populations, with ‘central’ symptoms often taken to suggest potential intervention targets. However, questions have been raised regarding the stability and clinical significance of centrality metrics. This systematic review aimed to examine the centrality of depressive symptoms among adults aged 50 years and older across network studies, with a focus on identifying its influencing factors.
Methods: We conducted systematic searches via PubMed, PsycINFO, and Web of Science from database inception to February 15, 2025, with one additional study identified during peer review. The National Institutes of Health Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies was adapted to assess the quality of included studies. Symptoms were ranked according to their centrality. Factors associated with the presence of any DSM-based depressive symptom, non-DSM depressive symptom, and comorbid symptom/associated factor as among the top three central nodes, were examined using Firth penalised multivariable logistic regression.
Results: Fifty-seven studies comprising 176 networks were reviewed, with 17 different depression scales adopted. Six of them were ‘strong’ in terms of their quality. ‘ Depressed mood ’ was the most central DSM-based depressive symptom, although its rate of being among the most central nodes dropped from 80.6% in depressive symptoms-only networks to 59% in comorbidity and extended networks. ‘Lack of happiness’ was among the most central non-DSM depressive symptoms. Meanwhile, comorbid symptoms and risk/protective factors were among the most central nodes in comorbidity and extended networks, with anxiety symptoms of ‘ uncontrollable worry’, ‘ trouble relaxing ’, and ‘ nervousness ’ showing the highest node and bridge centrality. ‘Tired in the morning’ , ‘lack companionship’ , ‘alone’ , and ‘isolated’ were other commonly identified bridge symptoms. Networks constructed during the COVID-19 pandemic was associated with a lower odds of DSM-based depressive symptoms (aOR=0.22, CI = 0.07–0.64) and a higher odds of comorbid symptoms/associated factors (aOR=3.41, CI = 1.31–9.45) being most central, while networks constructed among clinical/subclinical samples were associated with a lower odds of comorbid symptoms/associated factors being most central (aOR=0.22, CI = 0.08–0.56). Those adopting DSM-based depression scales showed a trend of DSM-based depressive symptoms being most central across networks (aOR=3.10, CI = 0.97–11.20).
Conclusion: Substantial heterogeneity was observed in the centrality of symptoms in networks of depressive symptoms, with network types, environmental changes during time of data collection, and the choice of depression scales being influencing factors. Clinical implications and future directions in the context of these findings are discussed.
Related
- Identity Centrality and Trauma- and Stressor-related Psychopathology: A Systematic Review and Meta-analysis
- Self-Identity and Trauma- and Stressor-related Psychopathology: A Systematic Review
- Current situation and influencing factors of knowledge, attitudes, and behaviors related to depression among parents of adolescents with depression
- Study on influencing factors and configurational pathways of hope level in adolescent patients with depression
- Factors influencing baseline electrocardiographic parameters in rats: A systematic review of experimental studies published from 1960 to the present
BibTeX
@article{wong2026centrality,
title = {{Centrality of symptoms in networks of depression and comorbid psychopathology among adults aged 50 years and older: A systematic review and examination of influencing factors}},
author = {Stephanie Ming Yin Wong and Zuna Loong Yee Ng and Yun-Lin Wang and Jasmine Oi Lan Yeung and Tianyin Liu and Dara Kiu Yi Leung and Wai Chi Chan and Terry Yat Sang Lum and Gloria Hoi Yan Wong},
journal = {Translational Psychiatry},
year = {2026},
month = oct,
doi = {10.1038/s41398-026-04460-1},
url = {https://www.nature.com/articles/s41398-026-04460-1},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Centrality of symptoms in networks of depression and comorbid psychopathology among adults aged 50 years and older: A systematic review and examination of influencing factors
AU - Stephanie Ming Yin Wong
AU - Zuna Loong Yee Ng
AU - Yun-Lin Wang
AU - Jasmine Oi Lan Yeung
AU - Tianyin Liu
AU - Dara Kiu Yi Leung
AU - Wai Chi Chan
AU - Terry Yat Sang Lum
AU - Gloria Hoi Yan Wong
PY - 2026
DA - 2026/10/06
JO - Translational Psychiatry
DO - 10.1038/s41398-026-04460-1
UR - https://www.nature.com/articles/s41398-026-04460-1
AB - Background: Network analysis has increasingly been adopted over the past decade to study the structure of depressive symptoms in middle-aged to older populations, with ‘central’ symptoms often taken to suggest potential intervention targets. However, questions have been raised regarding the stability and clinical significance of centrality metrics. This systematic review aimed to examine the centrality of depressive symptoms among adults aged 50 years and older across network studies, with a focus on identifying its influencing factors. Methods: We conducted systematic searches via PubMed, PsycINFO, and Web of Science from database inception to February 15, 2025, with one additional study identified during peer review. The National Institutes of Health Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies was adapted to assess the quality of included studies. Symptoms were ranked according to their centrality. Factors associated with the presence of any DSM-based depressive symptom, non-DSM depressive symptom, and comorbid symptom/associated factor as among the top three central nodes, were examined using Firth penalised multivariable logistic regression. Results: Fifty-seven studies comprising 176 networks were reviewed, with 17 different depression scales adopted. Six of them were ‘strong’ in terms of their quality. ‘ Depressed mood ’ was the most central DSM-based depressive symptom, although its rate of being among the most central nodes dropped from 80.6% in depressive symptoms-only networks to 59% in comorbidity and extended networks. ‘Lack of happiness’ was among the most central non-DSM depressive symptoms. Meanwhile, comorbid symptoms and risk/protective factors were among the most central nodes in comorbidity and extended networks, with anxiety symptoms of ‘ uncontrollable worry’, ‘ trouble relaxing ’, and ‘ nervousness ’ showing the highest node and bridge centrality. ‘Tired in the morning’ , ‘lack companionship’ , ‘alone’ , and ‘isolated’ were other commonly identified bridge symptoms. Networks constructed during the COVID-19 pandemic was associated with a lower odds of DSM-based depressive symptoms (aOR=0.22, CI = 0.07–0.64) and a higher odds of comorbid symptoms/associated factors (aOR=3.41, CI = 1.31–9.45) being most central, while networks constructed among clinical/subclinical samples were associated with a lower odds of comorbid symptoms/associated factors being most central (aOR=0.22, CI = 0.08–0.56). Those adopting DSM-based depression scales showed a trend of DSM-based depressive symptoms being most central across networks (aOR=3.10, CI = 0.97–11.20). Conclusion: Substantial heterogeneity was observed in the centrality of symptoms in networks of depressive symptoms, with network types, environmental changes during time of data collection, and the choice of depression scales being influencing factors. Clinical implications and future directions in the context of these findings are discussed.
SN - 2158-3188
ER -