<p>Difficulty with reversal learning (RL)—appropriately shifting behavior following outcome contingency changes—may represent a shared or distinct mechanism across eating disorder (ED) diagnoses. We tested whether RL-related neural correlates differ among adults without and with EDs (anorexia nervosa, restricting subtype [AN-R], AN binge-eating/purging subtype [AN-BP], and bulimia nervosa [BN]) and whether these correlates correspond to ED-symptom severity and frequency. We hypothesized that individuals with EDs would demonstrate differential neural activation during RL relative to individuals without, that activation would differentiate AN-BP and BN versus AN-R, and that activation would predict ED-severity metrics. Medically stable participants with AN-R (<i>n</i> = 22), AN-BP (<i>n</i> = 20), and BN (<i>n</i> = 29) comprised the ED group (<i>N</i> = 71), contrasted with non-ED controls (<i>N</i> = 27). Participants (91% female; M<sub>age</sub> = 25.9; 80% white, 14.5% Asian) completed clinical interviews and, in a separate session, a probabilistic RL task during functional magnetic resonance imaging. We examined differences in neural activation during RL in the ventral striatum and ventrolateral prefrontal cortex (vlPFC) between the ED and non-ED groups and between diagnostic groups, and conducted exploratory whole-brain analyses. Relations between neural activation and ED symptoms were examined. Lower right vlPFC and ventral striatum activation during RL characterized EDs. No between-ED diagnosis differences emerged. Lower right vlPFC activation predicted more frequent binge eating and purging but not global ED psychopathology. Individuals with EDs may experience difficulty recruiting certain RL-related brain regions, which may relate to difficulty changing ED behaviors. Future directions include investigation of how RL-associated neural networks maintain ED symptoms and influence treatment outcomes.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Prefrontal and ventral striatal neural correlates of reversal learning in anorexia nervosa and bulimia nervosa

  • Matthew F. Murray,
  • Brianne N. Richson,
  • Glen Forester,
  • Neil P. Jones,
  • Elizabeth N. Dougherty,
  • Angeline R. Bottera,
  • Lisa M. Anderson,
  • Lauren M. Schaefer,
  • Erika E. Forbes,
  • Jennifer E. Wildes

摘要

Difficulty with reversal learning (RL)—appropriately shifting behavior following outcome contingency changes—may represent a shared or distinct mechanism across eating disorder (ED) diagnoses. We tested whether RL-related neural correlates differ among adults without and with EDs (anorexia nervosa, restricting subtype [AN-R], AN binge-eating/purging subtype [AN-BP], and bulimia nervosa [BN]) and whether these correlates correspond to ED-symptom severity and frequency. We hypothesized that individuals with EDs would demonstrate differential neural activation during RL relative to individuals without, that activation would differentiate AN-BP and BN versus AN-R, and that activation would predict ED-severity metrics. Medically stable participants with AN-R (n = 22), AN-BP (n = 20), and BN (n = 29) comprised the ED group (N = 71), contrasted with non-ED controls (N = 27). Participants (91% female; Mage = 25.9; 80% white, 14.5% Asian) completed clinical interviews and, in a separate session, a probabilistic RL task during functional magnetic resonance imaging. We examined differences in neural activation during RL in the ventral striatum and ventrolateral prefrontal cortex (vlPFC) between the ED and non-ED groups and between diagnostic groups, and conducted exploratory whole-brain analyses. Relations between neural activation and ED symptoms were examined. Lower right vlPFC and ventral striatum activation during RL characterized EDs. No between-ED diagnosis differences emerged. Lower right vlPFC activation predicted more frequent binge eating and purging but not global ED psychopathology. Individuals with EDs may experience difficulty recruiting certain RL-related brain regions, which may relate to difficulty changing ED behaviors. Future directions include investigation of how RL-associated neural networks maintain ED symptoms and influence treatment outcomes.