An international team of researchers has identified key networks within the brain which they say interact to increase the risk that an individual will think about – or attempt – suicide.
Writing today in Molecular Psychiatry, the researchers say that their review of existing literature highlights how little research has been done into one of the world’s major killers, particularly among the most vulnerable groups.
The facts in relation to suicide are stark: 800,000 people die globally by suicide every year, the equivalent of one every 40 seconds.
Suicide is the second leading cause of death globally among 15-29 year olds.
More adolescents die by suicide than from cancer, heart disease, AIDS, birth defects, stroke, pneumonia, influenza, and chronic lung disease combined.
As many as one in three adolescents think about ending their lives and one in three of these will attempt suicide.
“Imagine having a disease that we knew killed almost a million people a year, a quarter of them before the age of thirty, and yet we knew nothing about why some individuals are more vulnerable to this disease,” said Dr. Anne-Laura van Harmelen, co-first author from the University of Cambridge.
“This is where we are with suicide. We know very little about what’s happening in the brain, why there are sex differences, and what makes young people especially vulnerable to suicide.”
A team of researchers, including Hilary Blumberg, MD, John and Hope Furth Professor of Psychiatric Neuroscience at Yale, carried out a review of two decades’ worth of scientific literature relating to brain imaging studies of suicidal thoughts and behaviour. In total, they looked at 131 studies, which covered more than 12,000 individuals, looking at alterations in brain structure and function that might increase an individual’s suicide risk.
Combining the results from all of the brain imaging studies available, the researchers looked for evidence of structural, functional, and molecular alterations in the brain that could increase risk of suicide.
They identified two brain networks – and the connections between them – that appear to play an important role.
The first of these networks involves areas towards the front of the brain known as the medial and lateral ventral prefrontal cortex and their connections to other brain regions involved in emotion.
Alterations in this network may lead to excessive negative thoughts and difficulties regulating emotions, stimulating thoughts of suicide.
The second network involves regions known as the dorsal prefrontal cortex and inferior frontal gyrus system.
Alterations in this network may influence suicide attempt, in part, due to its role in decision making, generating alternative solutions to problems, and controlling behaviour.
The researchers suggest that if both networks are altered in terms of their structure, function or biochemistry, this might lead to situations where an individual thinks negatively about the future and is unable to control their thoughts, which might lead to situations where an individual is at higher risk for suicide.
“The review provides evidence to support a very hopeful future in which we will find new and improved ways to reduce risk of suicide,” said Professor Hilary Blumberg.
“The brain circuitry differences found to converge across the many studies provide important targets for the generation of more effective suicide prevention strategies.
“It is especially hopeful that scientists, such as my co-authors on this paper, are coming together in larger collaborative efforts that hold terrific promise.”
The majority of studies so far have been cross-sectional, meaning that they take a ‘snapshot’ of the brain, rather than looking over a period of time, and so can only relate to suicidal thoughts or behaviours in the past.
The researchers say there is an urgent need for more research that looks at whether their proposed model relates to future suicide attempts and at whether any therapies are able to change the structure or function of these brain networks and thereby perhaps reduce suicide risk.
The review highlighted the paucity of research into suicide, particularly into sex differences and among vulnerable groups.
Despite suicidal thoughts often first occurring as early as during adolescence, the majority of studies focused on adults.
“The biggest predictor of death by suicide is previous suicide attempt, so it’s essential that we can intervene as early as possible to reduce an individual’s risk,” said co-first author Dr. Lianne Schmaal from the University of Melbourne.
“For many individuals, this will be during adolescence. If we can work out a way to identify those young people at greatest risk, then we will have a chance to step in and help them at this important stage in their lives.”
Even more striking, despite the fact that transgender individuals are at increased risk for suicide, just one individual in the 131 samples included for the review was identified to be transgender.
“There are very vulnerable groups who are clearly not being served by research for a number of reasons, including the need to prioritise treatment, and reduce stigma,” said van Harmelen. “We urgently need to study these groups and find ways to help and support them.”
In 2018, the researchers launched the HOPES (Help Overcome and Prevent the Emergence of Suicide) study, supported by the mental health research charity MQ. HOPES brings together data from around 4,000 young people across 15 different countries in order to develop a model to predict who is at risk of suicide.
Over the course of the project, the team will analyse brain scans, information on young people’s environment, psychological states and traits in relation to suicidal behaviour from young people from across the world, to identify specific, universal risk-factors.
Suicide is a worldwide public health problem with approximately 800,000 victims per year and a leading cause of death in most societies1.
In addition, 10 to 20 times more attempt suicide, and a history of such attempts is considered a major risk factor of future suicide death2.
However, only a small minority of depressed persons will die from suicide5. Identifying depressed patients at risk of suicide is therefore crucial for developing sustainable and efficient preventive interventions. Unfortunately, the only clinical risk factor assessment available has poor predictive power6.
Both structural and functional neuroimaging studies have been conducted in individuals with a history of suicide attempts, shedding light on a potential role of the ventral and dorsal prefrontal regions, the anterior cingulate cortex, the temporal and parietal cortices, as well as selected subcortical nuclei, among others.
Moreover, recent studies have suggested that some deficits observed may be heritable, being found in close relatives of suicide victims who never attempted suicide13, which is in agreement with the known heritability of suicidal acts14.
However, knowledge about alterations affecting the organization and functioning of brain networks in relation to suicidal behavior is much more limited.
In the field of functional neuroimaging, functional connectivity (FC) analyses of resting-state functional Magnetic Resonance Imaging (rs-fMRI) datasets is an established method for studying the FC of a specific brain region or the related network architecture15–17.
Rs-fMRI allows examination of the tonic rather than phasic activation level underlying functional connectivity, which might be a more powerful way to identify intrinsic network abnormalities in a specific population. In contrast to task-based fMRI studies, rs-fMRI studies are not confounded by a subject’s motivation, present cognitive state or by specific task-related effects, such as the impact of practice or applied strategy, thus increasing the inter-subject and intra-subject reproducibility.
Furthermore, as shown by Fox and Greicius18, resting state studies likely have a better signal to noise ratio than task-based fMRI studies. Thus, rs-fMRI is more apt to identify specific and reproducible markers of neural dysfunction associated with suicidal behavior.
An increasing number of pathological conditions have been associated with abnormal FC between particular brain regions or in network organizations19, providing potentially valuable information for understanding the pathophysiology of these disorders.
Despite the relevance of FC analyses, so far, few rs-fMRI studies have been conducted in suicide attempters, and none in relatives.
Two studies used the amplitude of low-frequency fluctuation (ALFF) method to explore abnormal resting-state brain activity20,21, and showed changes in ALFF values in the middle and superior temporal, ventromedial prefrontal, and occipital regions in suicide attempters compared to depressed controls.
Using independent component analysis (ICA), Zhang et al.22 found increased FC in the cerebellum and the occipital cortex as well as decreased FC in the precuneus in adolescent depressed suicide attempters compared to depressed controls.
Very recently, Kang et al.23 demonstrated abnormally increased FC between the amygdala (used as a seed region) and the insula, orbitofrontal cortex and middle temporal gyrus in adult suicide attempters with major depressive disorder (MDD) compared to MDD controls.
With the growing use of connectomics, we are now able to shift the view from a local connectivity level towards a global network perspective.
Advanced mathematical approaches such as graph theory provide comprehensive insights into the key organizational principles of brain networks (e.g. small-worldness) that support efficient neural processing.
Graph theory has especially proven to be useful in the analysis of such data, providing multiple metrics to assess the topological properties of the underlying brain graphs26.
Recent studies have shown that healthy brain functioning is characterized by higher clustering and smaller shortest path lengths (i.e. higher global efficiencies) compared to a random network27, pointing toward both central features of a small-world configuration, i.e. segregated and integrated information processing.
Assortativity29 is a topological measure of network resilience and defined as a correlation coefficient between the degrees of all nodes on two opposite ends of a link. Moreover, the so-called rich-club coefficient describes the density of connectivity only between high-degree nodes (“hubs”) and has been assumed to indicate overall brain communication and resilience30.
For example, abnormal rich-club organization has been reported in patients with schizophrenia31.
Thus, these graph metrics provide information about the global organizational properties of functional and structural brain networks of a given group, as well as differences between groups32.
A recently developed statistical method called the network-based statistic (NBS) yields additional information about differences in local connectivity, offering an effective way to deal with the multiple comparison problems arising in the analysis of seed-based connectivity33.
To the best of our knowledge, there are currently no studies investigating alterations in global network functioning and local functional connectivity (using the NBS approach) in relation to suicidal behavior.
Thus, since suicidal behavior is characterized by a complex set of affective, cognitive and interpersonal dysfunctions7, the current study first aims at examining differences in the above-described global properties of all functional connections between brain regions, i.e. the connectome, using rs-fMRI data.
We combined data from two independent samples of depressed patients with MDD, with and without a personal history of suicidal behavior, and healthy controls to increase the statistical power in order to investigate global connectome alterations.
We further investigated a sample of first-degree biological relatives of suicide victims and relatives of depressed patients without a family history of suicide to uncover the heritable components of potential differences in the global network organization.
We only included patients with MDD in the present study, because previous studies directly comparing patients with MDD, bipolar disorder and schizophrenia showed considerable heterogeneity with regard to structural abnormalities34,35 as well as with regard to alterations in functional connectivity36. The second aim of the study was to investigate differences in local FC to identify specific brain regions associated with suicidal behavior.
The two analytic approaches described above were employed: graph theory analyses were conducted to identify potential changes in global topologic properties. NBS was applied to test for connectome-wide differences in network connectivity (while controlling for multiple comparisons).
Due to the multifaceted nature of suicidal behavior, we hypothesized that the functional disorganization of resting-state brain networks would exist at both the global and local scales. We especially hypothesized that global parameters of the functional connectome could similarly differentiate MDD patients with and without a personal history of suicidal behavior, as well as healthy relatives of suicide victims from healthy relatives of patients with MDD without a family history of suicidal behavior.
Based on previous studies in suicidal behavior, we expected to find differences in the brain networks implicated in impulsivity, social and emotional processing and decision-making, notably fronto-temporo-parieto-striatal structures.
More information: Schmaal, L, van Harmelen, A.-L. et al. Imaging suicidal thoughts and behaviors: a comprehensive review of 2 decades of neuroimaging studies. Molecular Psychiatry; 2 Dec 2019; DOI: 10.1038/s41380-019-0587-x