Depressed mood and Psychiatric anxiety were central symptoms, while Suicidality, Loss of interest, Psychiatric anxiety, and mRS score acted as bridge symptoms in a post-stroke symptom network, with Language mapping to the left superior corona radiata but Suicidality and Loss of interest showing no focal anatomical correlates.
Key Findings
Results
Depressed mood and Psychiatric anxiety were identified as the most central symptoms in the post-stroke symptom network at 3 months.
589 acute ischemic stroke patients were prospectively enrolled in this multicenter cohort study.
Depression, cognition, and dysfunction were assessed via the 17-item Hamilton Depression Rating Scale (HDRS), Mini-Mental State Examination (MMSE), and modified Rankin Scale (mRS), respectively.
Network analysis was used to construct a 3-month post-stroke symptom network integrating depressive symptoms, cognitive impairment, and functional status.
Centrality metrics identified Depressed mood and Psychiatric anxiety as the core nodes with the strongest within-network connections.
Results
Suicidality, Loss of interest, Psychiatric anxiety, and mRS score functioned as bridge symptoms connecting different symptom communities in the post-stroke network.
Bridge symptoms were identified using flow analysis, which quantified the direct association between dysfunction and neuropsychiatric symptoms.
mRS score was positively associated with Suicidality and Loss of interest, indicating that greater functional disability was linked to these specific neuropsychiatric symptoms.
mRS score was negatively associated with Language, suggesting that worse functional outcomes corresponded to poorer language performance.
The bridging role of these symptoms suggests they serve as cross-domain connections between depressive, cognitive, and functional impairment clusters.
Results
Voxel-based lesion-symptom mapping (VLSM) identified a significant lesion cluster in the left superior corona radiata specifically associated with the Language symptom.
VLSM was conducted in an imaging subset of 429 patients from the full cohort of 589.
The left superior corona radiata was the neuroanatomical substrate localized for Language deficits.
Language was identified as a network target through its bridge role and negative association with mRS score.
The left superior corona radiata finding is consistent with known white matter pathways subserving language function.
Results
Suicidality and Loss of interest showed no focal anatomical lesion correlates in VLSM analysis, suggesting reliance on distributed network dysfunction.
Despite being identified as bridge symptoms with clinical significance, neither Suicidality nor Loss of interest mapped to any significant focal lesion cluster.
The absence of focal correlates was interpreted as suggesting these symptoms arise from distributed network dysfunction rather than focal brain damage.
This finding contrasts with Language, which had a clear anatomical substrate in the left superior corona radiata.
The VLSM imaging subset included n = 429 patients, providing adequate power for lesion mapping analyses.
Results
Core and bridge symptoms including Depressed mood, Psychiatric anxiety, Suicidality, Loss of interest, and mRS score were identified as candidate indicators for rehabilitation planning.
The authors propose that targeting central and bridge symptoms may be more effective than treating isolated symptoms in post-stroke rehabilitation.
Suicidality, Loss of interest, and Language were specifically highlighted as promising candidates for monitoring and rehabilitation.
The network approach revealed the interplay among depressive symptoms, cognitive impairment, and dysfunction simultaneously rather than treating them as separate conditions.
The multicenter design with 589 patients supports the generalizability of these network-derived rehabilitation targets.
What This Means
This research suggests that after a stroke, mental health symptoms, thinking difficulties, and physical disability do not exist in isolation but form an interconnected network. By studying 589 stroke patients across multiple hospitals, researchers mapped how symptoms like depressed mood, anxiety, suicidal thoughts, loss of interest, and language difficulties relate to each other and to physical functioning three months after stroke. They found that depressed mood and anxiety were the most central or influential symptoms in this network, while suicidal thoughts, loss of interest, anxiety, and physical disability scores acted as 'bridges' connecting different symptom clusters — meaning these bridging symptoms may be especially important because disrupting them could affect multiple other symptoms at once.
The study also used brain imaging from 429 of these patients to see whether specific brain regions could be linked to the most important network symptoms. Language difficulties were clearly linked to damage in a white matter pathway in the left side of the brain called the superior corona radiata, which is consistent with what is known about language processing. However, suicidal thoughts and loss of interest could not be traced to any single brain region, suggesting these symptoms arise from widespread disruption of brain networks rather than damage to one specific area.
This research suggests that rehabilitation after stroke might benefit from focusing on these identified central and bridge symptoms rather than treating each problem separately. Monitoring symptoms like suicidal thoughts, loss of interest, and language difficulties after stroke may be particularly valuable because of their broad connections to other post-stroke problems. The finding that some psychological symptoms have no single brain location underscores the importance of whole-brain and network-level approaches to understanding and treating post-stroke mental health.
Du Q, Liang W, Fang Y, Li T, Pan C, Li G, et al.. (2026). Symptom Network and Lesion Mapping of Post-Stroke Depression, Cognitive Impairment, and Dysfunction: A Multicenter Cohort Study.. Brain and behavior. https://doi.org/10.1002/brb3.71759