Christian Bromberg, Sharon Green-Hennessy, Ph.D., Theresa DiDonato, Ph.D., Cara Jacobson, Psy.D.

Attachment Style, Dissociation, and Substance Use: An Exploration of Parallel Affect Regulation Strategies

View the poster >>

Many psychological disorders are characterized by affective disturbances (Gross & Jazaieri, 2014). Despite the increasing clinical and research interest in affect disturbance, the ways in which persons with and without psychiatric disorders regulate affect is complex and remains somewhat unclear.

One group that seeks to shield themselves from distressing affect are individuals classified as having an avoidant attachment style (Bowlby, 1969; Kobak & Sceery, 1988). Such individuals employ both pre- and postemptive “deactivating” strategies that distance the individual from strong affect (Fraley et al., 2000; Mikulincer & Orbach, 1995). These strategies bear a striking resemblance to detachment dissociation, often manifested through depersonalization and derealization, such that these general affect regulation strategies may be variants of the same underlying mechanism.

There are reasons to suppose that substance use functions similarly, including the induction of dissociative-like states and repression of affectively charged internal stimuli, following drug/alcohol consumption (Somer et al., 2010; Steele & Josephs, 1990). The aim of this proposed study is to examine the association among avoidant attachment style, detachment dissociation, and substance use, as they are proposed to operate in a significantly similar manner in regulating affect. More specifically, I am hypothesizing that individuals who self-classify as avoidantly attached will endorse significantly more dissociative symptoms than individuals who self-classify as securely attached. I am also hypothesizing that individuals who self-classify as avoidant and endorse high to very high levels of psychological distress over the past 12 months will endorse significantly more symptoms of substance abuse and dependence compared to individuals who self-classify as avoidant and endorse relatively low levels of psychological distress over the past 12 months.

Additionally, since the proposed mechanisms rely on detachment from affect, and given that distress is a common motivator for entry into treatment, the help seeking behavior of those who are avoidantly-attached will be further examined. Both their tendency to minimize perceived distress and their reluctance to reach out to others for support could inhibit avoidantly-attached individuals from accessing treatment services when needed (Diener & Monroe, 2011; Vogel & Wei, 2005). Thus, these individuals may engage in treatment primarily due to coercion from others, given that this is a common occurrence among individuals who mitigate negative affect through chemical dissociation (Emiliussen et al., 2017; Zeldman & Ryan, 2004). Consequently, this would necessitate an alternative therapeutic approach tailored specifically to individuals who may have low internal motivation to participate. Hence, the degree to which coercion from others, as opposed to self-reported distress, will be examined among avoidantly-attached versus securely-attached. Consequently, this would necessitate an alternative therapeutic approach tailored specifically to individuals who may have low internal motivation to participate. Hence, the degree to which coercion from others, as opposed to self-reported distress, will be examined among avoidantly-attached versus securely-attached.

More specifically, I am hypothesizing that among individuals self-classified as avoidantly-attached, when self-reported psychological distress and presence of coercion to seek mental health care are entered as predictors of reported service use of the past 12 months, presence of coercion will significantly predict mental health service use. Additionally, among individuals self-classified as securely-attached, when self-reported psychological distress and presence of coercion to seek mental health care are entered as predictors of reported service use of the past 12 months, self-reported distress and coercion will significantly predict mental health service use, with self-reported distress increasing the likelihood and coercion decreasing the likelihood.

The results of this study will shed light on the ways in which avoidantly-attached individuals manage strong, negative affect. This could provide information regarding tailored treatment approaches that may assist these individuals with managing and tolerating strong emotions without resorting to maladaptive affect regulation strategies. Additionally, if clinicians have more information regarding the mechanisms by which these strategies work, we can better understand their psychological impact and correlates. Furthermore, if this research can provide information regarding the motivation for avoidantly attached individuals, clinicians may be able to increase treatment attrition.

To examine these hypotheses, I will be using The National Comorbidity Survey -- Replication (NCS-R; Kessler et al., 2004), a survey intended to yield prevalence and correlate estimates of Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV; American Psychiatric Association, 1994) psychopathology. The National Institute of Mental Health fielded the survey to a nationally representative sample of adult, civilian, non-institutionalized, household members living in the continental U.S. (including students living in campus group housing) between February 2001 and April 2003. A total of 10,843 primary participants were approached to be interviewed.  The response rate was 70.9% for primary respondents and 80.4% for secondary respondents. Face-to-face interviews were carried out with 9,282 participants.

Regarding measures, Attachment style was measured using a modified version of the Hazan and Shaver’s (1987) self-report attachment style measure. Traditionally, the Hazan and Shaver measure presents each participant with a vignette portraying one of the three attachment categories (secure, avoidant, anxious-ambivalent) and asks the participant to identify which description is most like them. The measure was modified for the current study to instead ask participants to rate how similar each vignette was to themselves on a 4-point Likert scale, with 1 = not at all like me to 4 = a lot like me. Originally created by Stein et al. (2012), a depersonalization/ derealization scale was used in this study to assess for detachment dissociation. This scale is a 3-item self-report questionnaire. Each scale item was derived from the Dissociative Experiences Scale (Bernstein & Putnam, 1986). Level of distress was assessed using the Kessler Psychological Distress Scale (K10; Kessler et al., 2002). The K10 is a 10-item self-report questionnaire that asks respondents how frequently they have experienced symptoms of psychological distress during the past 30 days. Domains covered on the K10 correspond to nervousness, hopelessness, restlessness or fidgeting, depression, and worthlessness. Drug and alcohol use modules from the World Health Organization diagnostic interview (Kessler et al., 2004b) were administered following an endorsement to any screening questions inquiring whether the respondent had ever used (1) tobacco (cigarettes, cigar or pipe); (2) alcohol; (3) cannabis, hashish; (4) cocaine; (5) tranquilizers, stimulants, painkillers or other prescription drugs; or (6) any other illicit drug including heroin, opium, glue, LSD or peyote. For each substance endorsed, including alcohol, a separate assessment was made and respondents were asked about DSM-IV symptoms of abuse and dependence for that substance. Regarding service use, respondents were asked a single item measure about any lifetime use of professional health services specifically for problems with their emotions, nerves, and/or substance use. Respondents who endorsed having seen any professional regarding their emotions, nerves, or use of alcohol or drugs in the last 12 months were asked about their motivation for having seen that professional. This 1-item self-report question is as follows: “When you went to see a professional about your emotions, nerves or substance use in the past year, was this something you wanted to do, or did you go only because someone else was putting pressure on you?” “Both” is included as a possible response option. Respondents who endorsed their service use as being due to both factors will be included in the category of individuals who have utilized services due to being pressured by someone else, as research has shown that social desirability bias may influence individuals to endorse research scale items that will present the individual in the most favorable manner that is consistent with prevailing social norms (King & Bruner, 2000; van de Mortel, 2008).

Regarding statistical analyses, analyses will be conducted on the secure and avoidant attachment groups to determine if they differ significantly on any demographic variable. If such differences exist, and they are significantly correlated with the dependent variable of interest, they will be controlled for in subsequent analyses.

Assuming no covariates need to be controlled for, hypothesis one, which examined the difference between the securely and avoidantly-attached individuals on detachment dissociation will be assessed using an independent-samples t-test. Similarly, hypothesis two will also be tested with a t-test comparing avoidantly-attached individuals who score above the clinical range on the K10, with those who score below that mark in terms of number of substance items endorsed.

Hypotheses three and four will each be tested via logistic regression with having sought any mental health or substance treatment (yes/no) as the dependent variable and coercion (yes/no) and amount of distress (continuous K10 scores) as the predictors.  With hypothesis three, this logistic regression will be used only to test the avoidantly-attached subjects and with hypothesis four only to test the securely-attached subjects.