A Neuropsychological Analysis of Self-Injurious Behavior

Read this article as part of a course!

Attempting a neuropsychological conceptualization of self-harm is admittedly a tentative endeavor, given the variety of behaviors that could be considered and the psychological interpretations of such varied activities. When clinicians first hear about or directly experience patients that engage in self-harm, the initial response can be a mixture of alarm and confusion. Contemplating the putative neurological underpinnings of such behavior can potentially increase our ability to conceptualize self-harm as an intelligible element of human experience, albeit one that is certainly both paradoxical and troubling. In so doing, the hope is that such conceptualizations can be shared in a manner that opens up a productive dialogue between patient and therapist. Enabling patients to grasp the neuropsychological foundation and functional aspects of self-harm could help provide greater clarity and insight into the broader reasons for their actions. This, in turn, could provide a clear basis for the focus and direction of treatment. While it may initially be difficult to comprehend why someone would voluntarily cause bodily injury to themselves, consideration of the possible psychological functions of the behavior in these populations, as well as the developmental neuropsychology underlying such behavior, may provide an explanatory framework that links development, functional neuroanatomy, affect regulation, and therapeutic focus.

Why do people self-injure? From a psychological perspective in general (and from psychodynamic and behavioral perspectives respectively), self-injurious behavior is assumed to have a rationale, in that the behavior serves a purpose and has a definite function or functions. There are elements of the behavior that, either in the act itself, or in the preparation for the act, or in the aftermath of the act, have distinctly reinforcing properties or purposeful elements for the individual. From both psychodynamic and behavioral perspectives, there may also be reinforcing aspects of the behavior that the individual isn’t explicitly aware of and yet contribute to the performance and persistence of the behavior. As such, a single self-harm behavior may be over-determined, having multiple interconnected lines of causation leading to it as well as multiple consequences leading from it, making it a go-to behavior (or behavior of choice) on the part of the individual (Suyemoto, 1998). While a myriad of possible reasons for self-injury have been posited, a recent meta-analysis of patient responses and interviews by Edmondson, Brennan, and House (2016) revealed a pattern that broadly supports several main psychological functions, perhaps most importantly the capacity of the behavior to assist in regulating emotions. Other reasons for the behavior include exertion of influence on others by engaging in the behavior, punishment of the self or punishment of someone else, coping with dissociative feelings, enjoying the sensations derived from the experience, and averting an actual suicide attempt by engaging in self-injurious actions. Arguably, many of the aforementioned reasons may overlap and interact with one another. For example, the feelings of pain perceived as punishment may be tolerated or enjoyed if the individual is self-critical and view themselves worthy of punishment (Fox, O’Sullivan, Wang, & Hooley, in press); equally, averting a suicide attempt may be directly related to the ability of the behavior to manage strong negative affect that would otherwise be done away with by suicide. A subsequent meta-analysis carried out by Taylor et al. (2018) found comparable reasons reported for self-injury, with the intrapersonal function of regulating distressing or negative emotional states being most prevalent, thus supporting an affect-regulatory function of the behavior (Bresin, Kling, & Verona, 2018). Given the consensus that self-injury is utilized expressly to regulate negative affect, and that it appears to possess the capacity to do so, it is this function that will be examined in more detail from a neuropsychological perspective.

Emotional Pain and Physical Pain: Little Difference to the Brain

First and foremost, pain serves a highly adaptive role in the life of an organism: it alerts the individual to external and internal situations that are likely to endanger the safety and longevity of the organism. The subjective experience of pain is brought about by neurophysiological activity—all pain is quite literally “in your head” even if its origins occur outside of the brain. When an individual experiences pain, the fight-or-flight response from the sympathetic nervous system is often invoked, mobilizing the organism to take action to stop whatever situation is producing the painful experience. A rather basic example of a painful experience is that of placing your hand on a hot stove or sharp object: tissue damage detected by free nerve endings is relayed via A-delta fibers to the spine (producing a reflexive withdrawal of the hand) and then to the thalamus and various regions of the brain, including subcortical limbic systems and areas of the cortex devoted to bodily representation and motivation and emotion. Quite instinctively and naturally pain tells us which behaviors to stop and which to avoid in order to keep us safe and alive. Self-injury is therefore quite counterintuitive.

Emotional pain can be confusing from a psychological and behavioral standpoint, given that the situations producing the pain are no longer readily identifiable and therefore more difficult to avoid, escape, or control. For many who self-injure, the pain caused by self-inflicted damage to body tissue is apparently seen as preferable to the pain that comes from uncertainty in relationships (Will this person leave me?/Does this person love me?) or from a real or perceived loss in relationships (This person has left me/This person does not love me). Adolescents appear to be at a greater risk of self-harming compared to adults, with adolescent females having the greatest risk (Morgan et al., 2017; Sornberger, Heath, Toste, & McLouth, 2012). Being bullied and/or being a sexual minority also increases the likelihood of self-harm during adolescence (DeCamp & Bakken, 2016). Within the adult population, self-harm behaviors are most prevalent among those meeting the diagnostic criteria for depression or borderline personality disorder. In both adolescent and adult populations, chief among the reasons for self-harm are feelings of low self-worth and fragile self-identity, both of which can be heightened by social exclusion and social rejection. In adolescents this is understandable from a developmental perspective, with this period arguably being quite tumultuous both in terms of the subjective experience of emotion and emotion regulation.

Adolescence is a time of role exploration that may be accompanied by role confusion and an evolving sense of self within a social environment; being bullied during this developmental period is likely to exacerbate any feelings of doubt and self-judgement. Low self-worth may be tied to self-injury as a form of self-punishment due to feeling that in some instances such pain is deserved. While self-punishment may be functionally linked to feelings of low self-worth, a poor sense of self may also be directly related to an inability to cope with strong emotional states induced by social rejection or social exclusion. Particularly among those diagnosed with borderline personality disorder, perceived rejection or abandonment by others can throw an already precarious sense of self into more pronounced disarray due to emotional pain that is experienced as unmanageable. The desire to stop overwhelming emotional pain of this sort (“psychache”) is often cited as a reason for attempting suicide (Shneidman, 2005). A possible window into the neuropsychological rationale behind self-injury has emerged from studies into the neurological basis of pain itself, with neuroimaging studies showing that emotional pain and pain with a basis in tissue damage have similar neuroanatomical meeting points within subcortical structures and the cerebral cortex. In an fMRI study, Eisenberger, Lieberman, and Williams (2003) examined the effects of simulated social exclusion (via interaction in a group video game) while the participants’ brain activity was monitored. When intentionally excluded from the game, the anterior cingulate cortex was significantly more active, with self-reported levels of distress correlated highly with activation of the anterior cingulate as well. Perhaps a more telling and provocative finding is that revealed in a study by Kross, Berman, Mischel, Smith, and Wager (2010) where individuals who had suffered recent romantic breakups, when viewing photos of their ex-partner, exhibited increased activity in regions of their brain that were also activated by physical pain induced by painful temperature increases. Noteworthy is that the participants had experienced recent interpersonal rejection and that the stimuli were visual representations of the individuals who had rejected them. The authors noted that brain areas activated by their recollection of and reflection upon that rejection correlated significantly with brain regions activated by physically painful heat and included the thalamus, the insula, the dorsal anterior cingulate cortex, and the secondary somatosensory cortex:

The brain systems that underlie social rejection developed by co-opting brain circuits that support the affective component of physical pain. The current findings substantively extend these views by demonstrating that social rejection and physical pain are similar not only in that they are both distressing [but in that] they share a common representation in somatosensory brain systems as well. (Kross et al., 2010, p. 6273)

This strengthens the view that emotional pain, because it shares a very similar neurological architecture and occupies much of the same neurological real estate as physical pain, could be perceived as life-threatening to an organism and therefore needs to be taken just as seriously as physical pain (Eisenberger, 2012).

In Fonagy, Gergely, Jurist, and Target’s (2002) model of the development of mentalization and Schore’s (1994) developmental neurobiological model of human cognitive and affective functioning, emotion regulation is intricately tied to and dependent upon the development of the cognitive and neurological representation of the self. In early life, the infant experiences the strong emotions brought about by hunger, cold, and physical pain. Through these experiences a child learns that an external human force, in the form of the primary caregiver, will predictably and reliably assist them in interpreting and regulating these strong emotional states, especially negative affect states. Physical, emotional, and verbal reassurance will be provided for negative affective states, and nonverbal and verbal encouragement, and perhaps appropriate containment, will be provided for positive affective states such as joy and excitement. Through these interactions, attachment patterns are formed with the primary caregiver and feelings of security are established (Bowlby, 1988; Horney, 1950). In time, and through experience-dependent learning and development, the child is able to internalize the abilities and capacities modeled by the adult caregiver in regulating emotional states and responses. This internalization of the affect-regulating capacity of the caregiver allows for the development of the self as an affect-experiencing and affect-regulating entity separate from the primary caregiver. This model is echoed in the model of borderline personality traits presented by Meares (2012) linking the disorder to an insufficiently developed sense of self, with related problems in the regulation of negative affect, where stressful events and interpersonal challenges are viewed apart from a personal and interpersonal narrative, being seen as arbitrary and uncontrollable occurrences. In several neurobiological models of self-concept and affect-regulation (particularly those of Schore and Meares), there is an emphasis on the regulatory role of the frontal cortical structures linked to the limbic system structures implicated in emotional responses to external and internal stimuli.

When the neurological representation or conceptualization of the self has not been developed to the point where significant emotional reactions can be regulated, withstood, and constructively utilized, the experience of these strong emotions may be overwhelming and likely threatening to the very nature of the self and the well-being of the individual. Essentially, there is no overriding self-structure in which to contextualize these painful emotional experiences and few mature psychological defenses that allow these emotional states to be experienced as palatable and surmountable. Neurologically, the ability to internally tolerate or regulate emotional responses would then be tied to the strength or stability of the individual’s sense of self. Indeed, in the neurodevelopmental model proposed by Schore (1994), the individual’s sense of self is tied intricately to the internalization of the emotion-regulating capacities of the primary caregiver. Caregivers who negate or invalidate the emotional displays of their children fail to provide any point of reference for the internalization of this capacity for affect regulation, leaving the child with the confusing dual perspective that their emotions are both illegitimate and unmanageable: they can’t have the emotion, and if they do have the emotion they are told they shouldn’t have it. Such responses to emotional displays leave the child with little or no idea as to what to do with emotions when they inevitably occur.

When working with patients who display an inability to manage, contain, and otherwise make sense of strong negative emotions, a common refrain often heard is: My goal is to not have any emotions at all. For human beings this is an unattainable and unrealistic goal, however. Affective neuroscience and psychodynamic theory are in general agreement in asserting that, arguably, emotions are the foundational core of human experience and consciousness (see Damasio, 1999; Panksepp, 1998; Schore, 1994; Solms, 2013). Furthermore, paraphrasing Freud, suppression of emotion is not an adaptive strategy in the long term, given that emotions always re-emerge in some other form. In the individual who has never learned to tolerate, regulate, or constructively utilize emotions, and perhaps has even learned either from experience or directly via spoken rules that emotions are dangerous things to have, the experiencing of any form of emotional pain can understandably be nearly unbearable. The emotional pain, the psychache, is perceived as a very real, oppressive, and quite threatening phenomenon because it is harnessing the same neurological structures that encode physical damage and existential threat, and there is no conceivable way for the individual to defend against it given the current limitations in processing of emotions. A fairly logical—perhaps even reasonable—solution to this problem, from the stance of the one experiencing this uncontrollable and persistent pain, would be to introduce a painful experience that was fully understood and fully controllable. This is partially in line with the classical gate theory of pain proposed by Melzack and Wall (1965), wherein other sensory signals (such as touch and pressure) occurring at the site of tissue injury can alter the experience of the pain caused by the injury. Indeed, if this concrete and predictable physical pain were to overlap neurologically with the unpredictable emotional pain, it could perhaps have the ability to overwrite the emotional pain, supplanting it at least temporarily. One of the perceived benefits of inducing physical pain is that the physical pain may not only unseat the emotional pain temporarily, but also that the physical pain has a discernible beginning and end, providing not only a clear sense of control but also a strong sense of relief when the pain subsides. In line with the respective theorizing of Fonagy et al. (2002), Meares (2012), and Schore (1994) linking deficiencies in self-concept with deficiencies in the self-regulation of affect, limbic system structures such as the anterior cingulate cortex and the amygdala tend to be hyperactive in individuals diagnosed with borderline personality disorder (Niedtfeld, Schulze, et al., 2010; Plener, Bubalo, Fladung, Ludolph, & Lulé, 2012), and activation of these brain regions has been found to decrease after the experience of painful physical stimuli (Niedtfeld, Kirsch, et al., 2012; Niedtfeld, Schulze, et al., 2010; Schmahl et al., 2006), meaning that “pain might result in increased inhibitory interactions (i.e., negative coupling) between neural areas associated with the processing of emotions and brain regions supporting the regulation of negative affect” (Niedtfeld, Kirsch, et al., 2012, p. 6).