Suicide remains one of the leading causes of death among adolescents worldwide, and major depressive disorder is among the strongest known risk factors. Clinicians have long recognized that suicidal ideation and suicide attempts are not simply points on a single continuum; many adolescents who think about suicide never attempt it, while some attempts occur with little reported forewarning. Understanding the neural mechanisms that separate these states is therefore considered critical for early risk detection. Yet most neuroimaging research has treated suicidal adolescents as a single group, blurring precisely the distinction that matters most for prevention.

The research team, led by Zhenggen Lin, Hongyu Zheng, and Rujing Zha, with corresponding authors Xinyou Lv and Hui Zhong, recruited 106 adolescents diagnosed with major depressive disorder according to DSM-5 criteria, together with 23 age-matched healthy controls. Within the depressed group, participants were carefully categorized into four subgroups: those with suicidal ideation only, those with a history of suicide attempts but no current ideation, those with both ideation and attempts, and those with neither. This four-way stratification is what gives the study its unusual diagnostic power, allowing the investigators to ask not merely whether suicidal adolescents differ from healthy ones, but whether the act of attempting suicide leaves a different neural fingerprint than the thought alone.

Each participant underwent resting-state fMRI, a technique that measures spontaneous blood-oxygen-level-dependent fluctuations while the subject lies quietly in the scanner without performing any task. Rather than examining isolated brain regions, the researchers applied graph theoretical analysis, a mathematical framework that treats the brain as a network of nodes and edges. Each region of interest, defined using the Automated Anatomical Labeling atlas, becomes a node, and the temporal correlation between regional activity patterns defines the strength of the connections, or edges, between them. From this network representation, the team computed measures such as nodal clustering coefficient, nodal degree centrality, nodal efficiency, and local efficiency, each of which captures a different aspect of how well a given node is wired into the wider network.

The results converged on one structure with striking consistency. Adolescents with current suicidal ideation, whether or not they had also attempted suicide, showed a reduced nodal clustering coefficient in the left parahippocampal gyrus, indicating that this region’s local neighborhood of connections had become less densely interconnected. The same groups also displayed reduced nodal degree centrality in the right parahippocampal gyrus, meaning that this hub maintained fewer strong connections to the rest of the brain overall. In network terms, the parahippocampal gyrus appeared to be losing both its local cohesion and its global influence precisely in those adolescents whose depression was accompanied by active thoughts of suicide.

Seed-based functional connectivity analysis sharpened the picture further. When the researchers placed seeds in the bilateral parahippocampal gyri and examined correlations with the rest of the brain, the suicidal ideation group stood apart from both the attempt-only group and the healthy controls: their parahippocampal connectivity with the lateral temporal cortex was significantly reduced. The lateral temporal cortex, encompassing regions such as the middle and superior temporal gyri, is involved in semantic processing, auditory association, and the interpretation of complex perceptual input. Weakened coupling between a memory-association hub and this perceptual-linguistic belt may, the authors suggest, reflect altered mnemonic–perceptual integration in adolescents ruminating on suicide.

Perhaps the most clinically provocative finding involved the sensorimotor network. Functional connectivity between the parahippocampal gyrus and sensorimotor regions, including the precentral and postcentral gyri and the paracentral lobule, was significantly lower in both the suicidal ideation group and the combined ideation-plus-attempt group compared with depressed adolescents without suicidality and with healthy controls. The sensorimotor network governs the planning and execution of movement and the bodily sense of self. Reduced dialogue between memory-related and motor-related circuitry could plausibly relate to differences in how suicidal thoughts become, or fail to become, translated into action, although the authors are careful to frame this as an interpretive hypothesis rather than a demonstrated mechanism.

The clinical associations add weight to the imaging results. Severity of suicidal ideation, measured with the Beck Scale for Suicide Ideation, and depressive symptoms, rated on the 17-item Hamilton Depression Rating Scale, were incorporated into the analyses, and the network alterations were linked to current suicidal ideation rather than to depression severity alone. This specificity matters: if the parahippocampal changes merely tracked how depressed the adolescents were, they would be of little use for suicide risk assessment. Instead, the pattern appears to tag the suicidal state itself, raising the possibility that parahippocampal connectivity could serve as an imaging biomarker to complement clinical interviews and self-report scales, which remain the mainstay of risk evaluation but are limited by reliance on disclosure.

The authors are candid about the limitations of their work. The study is cross-sectional, capturing a single moment in time, so it cannot determine whether the observed network changes precede suicidal ideation, result from it, or reflect some third factor. The four diagnostic subgroups were small and unequal in size, which limits statistical power and the generalizability of the subgroup comparisons. Resting-state fMRI measures correlations, not causal communication, and motion artifacts, although controlled through framewise displacement quality checks and standard preprocessing pipelines in SPM12, DPABI, GRETNA, and related toolboxes, remain an inherent concern in adolescent imaging. The authors explicitly call for validation in larger, longitudinal, and multimodal studies that combine imaging with genetic, behavioral, and ecological momentary assessment data.

Even with those caveats, the study opens a genuinely new window on one of psychiatry’s most urgent questions. By showing that the parahippocampal gyrus, a region rarely at the center of suicide research, behaves differently in adolescents who are thinking about suicide versus those who have acted, the work challenges the field to stop collapsing ideation and attempts into a single category. If future longitudinal studies confirm that weakening parahippocampal connectivity with lateral temporal and sensorimotor networks heralds the emergence or escalation of suicidal thoughts, clinicians may one day have an objective, biologically grounded tool for identifying which depressed adolescents need the most intensive intervention, and for monitoring whether that intervention is rewiring the circuits that matter.

Subject of Research: Parahippocampal network dysfunction in adolescents with major depressive disorder who have suicidal ideation versus suicide attempts, studied with resting-state fMRI

Article Title: Parahippocampal network dysfunction distinguishes suicidal ideation from suicide attempts in adolescents with major depressive disorder: a cross-sectional resting-state fMRI study

Article References: Lin, Z., Zheng, H., Zha, R., Liu, S., Chai, K., Li, J., Lv, X., & Zhong, H. (2026). Parahippocampal network dysfunction distinguishes suicidal ideation from suicide attempts in adolescents with major depressive disorder: a cross-sectional resting-state fMRI study. BMC Psychiatry. https://doi.org/10.1186/s12888-026-08609-6

Image Credits: AI Generated

DOI: 10.1186/s12888-026-08609-6

Keywords: adolescent depression, suicidal ideation, suicide attempts, parahippocampal gyrus, resting-state fMRI, functional connectivity, graph theory, major depressive disorder, sensorimotor network, lateral temporal cortex, suicide risk assessment, brain networks