An exploratory magnetic resonance imaging study of suicidal ideation in individuals at clinical high-risk for psychosis

Highlights

• Suicide is a major cause of death in psychosis.

• The neurobiology of suicide in psychosis is understudied, especially in early psychosis.

• We used magnetic resonance imaging to examine the neurobiology of suicidal ideation in individuals at clinical high-risk for psychosis (CHR).

• CHR individuals with SI had thicker middle temporal and right insular cortices than CHR individuals without SI and healthy control subjects.

• These pilot findings underscore the potential for the use of brain imaging biomarkers of suicide risk in CHR individuals.

Abstract

Suicide is a major cause of death in psychosis and associated with significant morbidity. Suicidal ideation (SI) is very common in those at clinical high-risk for psychosis (CHR) and predicts later suicide. Despite substantial work on the pathobiology of suicide in schizophrenia, little is known of its neurobiological underpinnings in the CHR or putatively prodromal state. Therefore, in this pilot study, we examined the neurobiology of SI in CHR individuals using structural MRI. Subjects were aged 14–30 and met criteria for the Attenuated Positive Symptom Psychosis-Risk Syndrome (APSS) delineated in the Structured Interview for Psychosis-Risk Syndromes (SIPS). Suicidality was assessed using the Columbia Suicide Severity Rating Scale (C-SSRS). Volumetric MRI scans were obtained on a 3T Phillips scanner. MRI data were available for 69 individuals (19 CHR without SI, 31 CHR with SI and 19 healthy control subjects). CHR individuals with SI had thicker middle temporal and right insular cortices than CHR individuals without SI and healthy control subjects. The location of these findings is consistent with neurobiological findings regarding suicide in syndromal psychosis. These findings underscore the potential for the use of brain imaging biomarkers of suicide risk in CHR individuals.

Introduction

Individuals with schizophrenia die almost 15 years earlier than individuals without schizophrenia (Hjorthoj et al., 2017). Much of this excess risk is related to suicidality (Bushe et al., 2010). Upwards of half of individuals with psychotic disorders experience suicidal ideation (SI) during their lifetimes (Fenton et al., 1997; Skodlar et al., 2008) and approximately 30% attempt suicide at some point in their lives (Barrett et al., 2011). The lifetime risk of dying by suicide in schizophrenia and related disorders has been reported to be between 5 and 10% (Carlborg et al., 2008; Kasckow et al., 2011; Palmer et al., 2005; Siris, 2001). Therefore, suicidality constitutes a significant problem among individuals with psychotic disorders.

Suicide and suicidal ideation are especially problematic in the early stages of psychosis when patients are at an even greater risk of dying by suicide (Dutta et al., 2010; Mortensen and Juel, 1993; Nordentoft et al., 2004; Palmer et al., 2005). Neurobiological work on suicide in psychosis is limited but has revealed abnormalities in white matter (Lee et al., 2016; Long et al., 2018; Rusch et al., 2008) and brain function (Minzenberg et al., 2015; Potvin et al., 2018). Some of the most consistent work has demonstrated cortical thinning or lower gray matter volumes in several regions including in temporal and insular cortices (Besteher et al., 2016; Giakoumatos et al., 2013) as well as other abnormalities (Aguilar et al., 2008; Spoletini et al., 2011). The importance of these brain regions to suicide is highlighted by a recent, comprehensive review by Schmaal and colleagues (Schmaal et al., 2020). In this review, Schmaal and colleagues provide a model of suicidal thoughts and behaviors in which an extended ventral and posterior network, including temporal and insular cortices, may be responsible for negative and blunted emotional and mood states, while a more dorsal network would be more responsible for suicidal behavior due to its role in cognitive control and flexibility (Schmaal et al., 2020).

There is substantially less work on suicidality in the pre-syndromal or clinical high-risk (CHR) period. However, the available evidence suggests that SI and suicide attempts are very common – upwards of 50% or more in some studies (Andriopoulos et al., 2011; DeVylder et al., 2012; Gill et al., 2015; Hutton et al., 2011). These data are consistent with evidence that the mere presence of positive symptoms, either frank or attenuated, is strongly associated with suicidality (Kelleher et al., 2013, 2012; Yates et al., 2019).

Despite these important findings which suggest that identifying suicidality, and biomarkers of suicidality, in the early stages of psychosis could potentially identify those who will go on to attempt suicide, there remain no studies of the neurobiology of suicide in the CHR phase of the illness.

Aleman and Denys have written that despite the prevalence of suicidality, its societal impact, and its financial burden, as well as that suicidality is preventable and treatable, not enough progress has been made towards understanding and treating suicidal behavior (Aleman and Denys, 2014). They suggest that psychiatry has neglected research on suicidality for reasons such as cultural taboo, its complicated and multi-factorial nature, and potentially diverse etiologies between fatal and non-fatal suicide attempts. To address these issues, they advise a number of solutions, including funding further research on the psychological and neurobiological roots of suicide.

Therefore, the goals of this exploratory, pilot study were to examine the neurobiology of SI, the precursor to suicidal behavior, among CHR individuals. We focused on cortical thickness given that the most robust and replicated findings on the neurobiology of suicide in psychosis suggest abnormalities in cortical thickness and gray matter volumes in schizophrenia (Girgis, 2020). In particular, while we examined the full cortical mantle, our primary areas of interest were insula and temporal cortex given that these were the primary areas found to be abnormal in two prior studies of cortical thickness and gray matter volume in individuals with schizophrenia and related psychotic disorders who had previously attempted suicide (Besteher et al., 2016; Giakoumatos et al., 2013). We hypothesized that temporal and insular cortices among CHR individuals with SI would be thinner than in CHR individuals without SI and healthy control subjects.