Connectivity findings associated with suicidality in young people with non-suicidal self-injury

• In Event: Functional brain predictors and correlates of suicidal ideation and non-suicidal self-injury in adolescents

Abstract

Non-suicidal self-injury (NSSI) is a strong predictor of future suicide, possibly even more so than a prior history of suicide crises (Mars et al., 2019). An important area of inquiry is to develop a more nuanced understanding of risk factors for suicide within this population. Adolescents and young adults with NSSI show aberrant connectivity at rest and during tasks. As many of these differences may be driven by shared features of psychopathology, within-NSSI group analyses are warranted as a means to differentiate individuals at highest risk. Exploring neurobiological mechanisms may be fruitful as we continue to develop interventions that most effectively target both NSSI and suicide. Here we examine the association between suicidality and resting state functional connectivity (RSFC) between regions of interest that capture both the salience and emotion (SEN) and default mode networks (DMN).

Method: Adolescents and young adults aged 13-21 completed a diagnostic evaluation. We used the Inventory of Statements about Self-Injury (Glenn & Klonsky, 2011) to gather information about NSSI. Participants completed either the Personality Assessment Inventory-Adolescent (PAI-A; Morey, 2007a) or the adult PAI (Morey, 2007b). We used t-scores from the suicide subscale of the PAI and PAI-A as a dimensional measure of potential suicide risk and controlled for depressive symptoms using total scores on the BDI-II (Beck et al., 1996).

We acquired RSFC data using a Siemens 3T scanner. We centered on the amygdala and the rostral anterior cingulate cortex (rACC) as regions of interest for the SEN and DMN respectively. We examined connectivity between these ROIs with other SEN and DMN regions including the insula, precuneus, posterior cingulate cortex (PCC), nucleus accumbens (NAc), and orbitofrontal cortex (OFC).

Results: Twenty-four adolescents and young adults (Mage=17.53 years) had usable data. Suicidality was positively associated with RSFC between the right rACC and amygdala (r=.649, p=.001) and NAc (r=.525, p=.010); between the right amygdala and insula (r=.634, p=.001), OFC (r=.659, p=.001), precuneus, (r=.636, p=.001), PCC (r=.629, p=.001), and NAc (r=.674, p<.001); between left rACC and amygdala (r=.477, p=.021) and NAc (r=.712, p<.001); and between left amygdala and insula (r=.565, p=.005), orbitofrontal cortex (r=.524, p=.010), precuneus (r=.475, p=.022), PCC (r=.467, p=.025), and NAc (r=.533, p=.009).

Conclusions: Higher suicidality in young people with NSSI was associated with greater RSFC between and within specific ROIs of the DMN and SEN. While some of our findings contrast from prior research (Du et al., 2017; Johnston et al., 2017), the present study departs from this prior work as we focus on a dimensional measure of suicidality in a diagnosis-independent sample. Our findings suggest that high-risk individuals with NSSI demonstrate increased connectivity between regions implicated in motivation, interoceptive awareness, and self-referential and reward processing. This provides preliminary information regarding the neurobiological profile of adolescents and young adults at highest risk for suicide. Further investigation of the relationships between connectivity and suicidality among larger samples is needed to fully elucidate the meaning of these findings. Future work will also benefit from using intervention as a means to examine the mechanisms of change associated with suicidality.

Authors

• Mindy Westlund Schreiner, University of Utah
  Presenting Author

• Bonnie Klimes-Dougan, University of Minnesota - Twin Cities
  Non-Presenting Author

• Bryon Mueller, University of Minnesota
  Non-Presenting Author

• Kathryn Cullen, University of Minnesota
  Non-Presenting Author