Excessive Exercise, Eating Pathology and Suicidality

Eating disorders (EDs) are associated with increased risk for suicide. Recent research suggests that excessive exercise (EE), an ED compensatory behavior, may increase suicide risk via increasing one's capability for suicide which is comprised of an elevated pain tolerance and a diminished fear of death. However, additional research is needed to determine the exact mechanisms by which EE may influence capability for suicide. Specifically, no studies to date have examined the influence of EE and eating pathology on behavioral indices of capability for suicide. The present study had two aims: 1) this study sought to examine how EE engagement and ED status influences pain avoidance behaviors in comparison to controls; and 2) the second aim of the present study was to determine how EE engagement and ED status influence fear of death and avoidance of death-related stimuli. We recruited 87 participants based on ED and EE status (22 controls, 22 EE only, 21 ED only, and 22 ED+EE). Hierarchical regression with dummy coding was used to compare our groups on the following outcomes: 1) breakpoint and total key presses on a pain avoidance progressive ratio task (CPT); and 2) breakpoint and total key presses on a death avoidance progressive ratio task (viewing and rating of suicide images). The results of the present study provide preliminary, albeit mixed support for our hypotheses. Specifically, we found that the EE only (β = -0.31, p = .017) and ED only (β = -0.30, p = 0.02) groups differed significantly from controls on one measure of pain avoidance (CPT PR breakpoint); the ED only group (β = -0.31, p = .017) differed significantly from controls on one measure of death avoidance (suicide-related images PR total presses). However, our ED group characterized by EE did not differ significantly from controls on any of our behavioral measures. See supplementary file for Tables 1 and 2 which provide group comparisons on main variables of interest (Table 1) and a correlation matrix for continuous variables (Table 2). Exploration of these findings, limitations, and areas for future research are included in our general discussion.