یک تحقیق تجربی از تمایل عاطفی و تحمل درد فیزیکی در خودآسیبی عمدی: نقش تعدیل کننده پریشانی فردی

Abstract Although theoretical and clinical literature emphasize the role of both an unwillingness to experience emotional distress and physical pain tolerance in deliberate self-harm (DSH), research on their associations with DSH remains limited. This study sought to examine the relationships between DSH and the willingness to experience emotional distress and tolerate physical pain, including the moderating role of interpersonal distress in these relationships. To this end, young adults with recent DSH (n = 43) and controls without any DSH (n = 52) were randomly assigned to 1 of 2 emotion-induction conditions (distressing or neutral), after which behavioral measures of both the willingness to experience distress and physical pain tolerance were obtained. Consistent with hypotheses, findings indicated heightened physical pain tolerance among self-harming individuals only under conditions of interpersonal distress. Furthermore, findings provided some support for the hypothesized association between DSH and the unwillingness to experience emotional distress, suggesting that self-harming women evidence less willingness to experience emotional distress only under conditions of depleted regulatory capacity (eg, following an interpersonal stressor).

1. Introduction Clinical researchers have become increasingly interested in behaviors involving the deliberate, direct destruction of body tissue without conscious suicidal intent, but resulting in injury severe enough for tissue damage to occur, referred to here as deliberate self-harm (DSH) [1], [2] and [3]. Deliberate self-harm is a serious clinical concern. Although this behavior is, by definition, distinguished from suicidal behaviors involving an intent to die, individuals who engage in DSH are at heightened risk for suicide [4]. Furthermore, DSH is associated with a wide range of negative interpersonal and intrapersonal consequences, including shame, guilt, and social isolation [5] and [6]. Although originally studied primarily within the context of borderline personality disorder (BPD) [5], a growing body of evidence suggests that DSH is much more common among nonclinical populations than previously thought. In particular, evidence suggests that community young adults are at especially high risk for DSH, with rates of DSH among various nonclinical young adult populations ranging from 17% to 41% [7], [8], [9] and [10]. Further, although the vast majority of research on DSH has focused exclusively on the factors associated with this behavior among female samples, recent findings indicating comparable rates of DSH among female and male college students [11], adolescents [12], and military recruits [2] highlight the importance of examining the development and maintenance of DSH among male subjects as well.

3. Results 3.1. Preliminary analyses A series of t tests and χ2 analyses were conducted on demographic characteristics to determine equivalence across group (DSH vs non-DSH) and condition (distressing vs neutral). Results indicated no significant differences in racial background, gender, education, marital status, or income across group or condition (Ps > .10), and all effect sizes were small (ηp2s <.02, contingency coefficients <.20). Moreover, although DSH participants were significantly younger than non-DSH participants (19.30 ± 1.73 vs 20.04 ± 1.73, respectively; P <.05), the effect size associated with this difference was small (ηp2 < .05) and none of the dependent variables was significantly associated with age. Thus, this variable was not included as a covariate in the primary analyses [79]. With regard to the dependent variables, the willingness to experience emotional distress (as indexed by MTPT-C latency to termination scores) was not significantly associated with any demographic (including age, gender, racial background, and income; Ps > .10) or clinical (including depressive symptoms, BPD symptoms, and lifetime DSH frequency; Ps > .10) characteristics. In addition, MTPT-C latency to termination scores were not significantly associated with negative affect at baseline (P > .05), in response to the interpersonal script (P > .10), or in response to the MTPT-C (P > .10), nor were they significantly associated with changes in negative affect from baseline to post-manipulation (r =.05, P > .10). Findings that MTPT-C latency to termination scores were not significantly associated with negative affect in response to the interpersonal script or MTPT-C itself provide further support for the construct validity of this measure, suggesting that latency to termination scores assess the willingness to experience emotional distress, rather than the amount of distress experienced. Finally, although MTPT-C latency to termination scores were not associated with pain threshold during the CPT (r = .05, P > .10), they were associated with pain tolerance during this task (r = .20, P < .05). Thus, pain tolerance during the CPT was included as a covariate in analyses of the willingness to experience emotional distress on the MTPT-C. Likewise, physical pain tolerance (as indexed by latency to terminate the algometer task) was not significantly associated with age, racial background, or income (Ps > .09), nor was it significantly associated with depressive symptoms, BPD symptoms, or lifetime DSH frequency (Ps > .10). However, latency to terminate the algometer task was significantly associated with gender (r = .55, P < .001), with men demonstrating greater pain tolerance than women (t = 6.33, P < .001). Further, although latency to terminate the algometer task was not significantly associated with negative affect at baseline or in response to the MTPT-C (Ps > .10), it was significantly associated with negative affect in response to the interpersonal script (r = .21, P < .05), as well as pain threshold and tolerance on the CPT and pain threshold on the algometer task (rs > .21, Ps < .05). Therefore, negative affect in response to the interpersonal script, pain threshold and pain tolerance on the CPT, and pain threshold on the algometer were included as covariates in analyses of physical pain tolerance on the algometer task. 3.2. Manipulation check Providing support for the experimental manipulation, results of a 2 (distressing vs neutral condition) × 2 (baseline vs post-manipulation) repeated measures analysis of variance (ANOVA) for negative affect revealed a significant condition × time interaction (F1,93 = 33.85, ηp2 = .27, P < .001), with participants in the neutral condition reporting a significant decrease in negative affect following presentation of the script (t = −3.43, P < .01; due specifically to decreases in ratings of feeling nervous and scared) and participants in the distressing condition reporting a significant increase in negative affect following the script (t = 4.74, P < .001; due to increases in ratings of feeling distressed, upset, guilty, hostile, irritable, and ashamed). Moreover, the group × time interaction was not significant within either condition (Fs < 3.60, Ps > .05), indicating that the change in levels of negative affect in response to both the distressing and neutral scripts did not differ across groups, and suggesting that DSH participants did not evidence greater reactivity to the distressing script. Indeed, when controlling for baseline negative affect, negative affect in response to the distressing script did not differ significantly between DSH and control participants (F1,45 = 2.11, ηp2 = .04, P > .10). Likewise, providing support for the use of MTPT-C latency to termination scores as a measure of the willingness to experience emotional distress, results of a 2 (DSH vs. control group) × 2 (post-manipulation vs. post-MTPT) repeated measures ANOVA for negative affect revealed a significant main effect of time (F1,93 = 28.24, ηp2 = .23, P < .001), with participants reporting an increase in negative affect in response to the MTPT-C. Further, the group × time interaction was not significant (F1,93 = 1.04, ηp2 = .01, P > .10), indicating that the MTPT-C resulted in a comparable increase in levels of distress for the DSH and control groups. Furthermore, providing evidence that all participants completed the algometer task under some level of distress, results of a 2 (DSH vs control group) × 2 (distressing vs neutral condition) × 2 (baseline vs post-MTPT) repeated measures ANOVA for negative affect revealed a significant main effect of time (F1,91 = 36.92, ηp2 = .29, P < .001), with participants reporting a significant increase in negative affect following the MTPT-C. Further, neither the group × time interaction (F1,91 = 2.44, ηp2 = .03, P > .10) nor the 3-way (group × condition × time) interaction (F1,91 = 2.70, ηp2 = .03, P > .10) was significant, indicating that the increase in the level of distress experienced by the DSH and control groups before starting the algometer was comparable. 1 Importantly, however, allowing us to examine differences in physical pain tolerance as a function of emotional distress, the condition × time interaction was significant (F1,91 = 7.50, ηp2 = .08, P < .01) (suggesting that participants in the distressing condition were experiencing more distress at the start of the algometer task than those in the neutral condition). Finally, providing support for the use of latency to terminate the algometer task as a measure of physical pain tolerance, findings suggest that the algometer was effective in inducing physical pain. Specifically, participants reported experiencing pain as a result of the algometer after an average of 5.17 ± 1.81 seconds, and all participants reported feeling pain within 12 seconds. Further, pain tolerance on this task ranged from 4.3 to 24.5 seconds, with all participants continuing with the task after the point at which they first experienced pain. Notably, however, participants ranged in their willingness to persist beyond the initial report of pain from less than 1 to 16 seconds. 3.3. Primary analyses A 2 (group) × 2 (condition) × 2 (gender) analysis of covariance (ANCOVA) (controlling for physical pain tolerance during the CPT) was conducted on MTPT-C latency to termination scores. Consistent with hypotheses, no significant main effects were found for group (F1,86 = 0.05, ηp2 = .00, P > .10), condition (F1,86 = 0.00, ηp2 = .00, P > .10), or gender (F1,86 = 0.03, ηp2 = .00, P > .10). Contrary to expectations, however, results also did not reveal a significant group × condition interaction (F1,86 = 2.48, ηp2 = .03, P > .10), although a significant 3-way (group × condition × gender) interaction emerged (F1,86 = 7.65, ηp2 = .08, P < .01) ( Fig. 2). Specifically, findings were consistent with hypotheses among women, as DSH participants in the distressing condition demonstrated less willingness to experience emotional distress than control participants and terminated the MTPT-C more quickly than all other groups, whereas DSH participants in the neutral condition demonstrated more willingness to tolerate emotional distress than controls. Interestingly, however, the opposite pattern of results was found for men, with DSH participants in the distressing condition demonstrating greater willingness to experience emotional distress than controls and DSH participants in the neutral condition demonstrating less willingness to experience emotional distress than controls. 2 Interactive effect of group (DSH vs non-DSH), condition (distressing vs ... Fig. 2. Interactive effect of group (DSH vs non-DSH), condition (distressing vs neutral), and gender (women vs man) on the willingness to experience emotional distress, as indexed by latency in seconds to terminate the MTPT-C (N = 95). Note. Means presented in this figure are adjusted means. Figure options A 2 (group) × 2 (condition) × 2 (gender) ANCOVA (controlling for physical pain threshold and tolerance during the CPT, negative affect in response to the interpersonal script, and pain threshold on the algometer task) was conducted on latency to terminate the algometer task. Consistent with past literature, findings indicated a significant main effect of gender (F1,81 = 7.00, ηp2 = .08, P < .05), with men demonstrating higher physical pain tolerance than women. No significant main effects were found for condition (F1,81 = 2.67, ηp2 = .03, P > .10) or group (F1,81 = 1.57, ηp2 = .02, P > .10). However, results did reveal a significant group × condition interaction (F1,81 = 5.13, ηp2 = .06, P < .05), with DSH participants in the distressing condition evidencing the highest pain tolerance ( Fig. 3). 3 Contrary to the findings for MTPT-C latency to termination scores, the 3-way (group × condition × gender) interaction was not significant (F1,81 = 3.29, ηp2 = .04, P > .05). Group (DSH vs. non-DSH) by condition (distressing vs. neutral) interaction for ... Fig. 3. Group (DSH vs. non-DSH) by condition (distressing vs. neutral) interaction for physical pain tolerance, as indexed by latency in seconds to terminate the algometer task (N = 95). Note. Means presented in this figure are adjusted means.