Lack of Insight Linked to Brain Changes in Anorexia Nervosa

Liam Davenport

August 04, 2016

Abnormalities in brain regions involved in forming insight may explain why some patients with anorexia nervosa (AN) have trouble recognizing their dysfunctional eating habits, a new imaging study shows.

Investigators found that patients with AN who have poor insight are significantly more likely to have connective abnormalities in regions of the brain associated with error detection, conflict monitoring, and self-reflection.

Led by Alex Leow, MD, PhD, University of Illinois-Chicago College of Medicine, the investigators found that AN patients with low scores on a measure of insight and delusionality had significant abnormalities in the caudal anterior cingulate cortex (ACC) and the posterior cingulate cortex (PCC), in addition to other brain regions. Similar, albeit nonspecific, changes were seen in individuals with body dysmorphic disorder (BDD).

In a release, Dr Leow noted that AN patients with poor insight "may not generate an 'error message' when told, for example, that they are putting themselves at serious risk for death by severe restricting.

"Thus, it is plausible that their brains literally don't believe that they are severely underweight and their behavior is dangerous, even when objective evidence suggests otherwise," she added.

The study was published online July 19 in Psychological Medicine.

Potential for Novel Therapies

To examine the modular organization of brain structural connectivity, the team studied 29 individuals with BDD, 24 weight-restored AN patients, and 31 healthy control individuals, all of whom had been free of psychoactive medications for 8 weeks prior to study entry.

Participants underwent diffusion-weighted MRI. Whole-brain white matter tractography was used to construct connectivity matrices to allow the comparison of modular structures across the groups.

Employing a technique for reconstructing the hierarchical modularity of brain connectomes called Path Length Associated Community Estimation, the team found that one brain model or community was significantly different between AN patients and healthy control persons (P < .01).

The module included the right caudate, the right pallidum, the right accumbens, the right caudal ACC, the right PCC, and the right rostral ACC in healthy persons. In AN patients, it comprised the right caudate, the right accumbens, the right rostral ACC, the right lateral orbitofrontal cortex (OFC), the right medial OFC, and the right frontal pole.

In individuals with BDD, there was a trend for abnormality in the same module, although this did not reach significance (P = .051) and included the right caudate, the right pallidum, the right rostral ACC, the right posterior cingulate, and the right medial OFC.

There were significant differences in mean normalized path length between AN patients and healthy control persons (P = .038) and between AN patients and individuals with BDD (P = .002), at 1.52 vs 1.41 and 1.43, respectively.

Further analysis revealed that there was a significant association between scores on the Brown Assessment of Beliefs Scale, which measures insight and delusionality, and path length in AN patients in the right caudal ACC (r = 0.51; P = .044) and the right PCC (r = 0.53; P = .035).

The researchers explain that this association suggests that individuals with AN "with worse insight have longer nodal path length in these regions, which indicates less efficient information transfer with the rest of the brain."

However, after taking into account body mass index (BMI) and duration of illness, the associations were reduced.

The team notes that because of the cross-sectional nature of the study, the results cannot prove causality. The investigators point out, however, that the modulating effect of BMI and disease duration suggest that "secondary effects of the illness, which could accumulate over time, could have resulted in the development of a relationship between insight and efficiency of information transfer between the right caudal ACC and the rest of the brain."

Further studies, they suggest, could include individuals with obsessive-compulsive disorder (OCD), inasmuch as the condition shares several clinical features with BDD and AN.

"In addition, the rate of co-morbid OCD is approximately 30% in AN and approximately 30% in BDD," they add.

The investigators suggest that the findings may open avenues for novel treatments for AN by improving patients' ability to detect a mismatch between their perceptions of self and reality.

Speaking to Medscape Medical News, Dr Leow said that she and her colleagues "are trying to explore novel virtual reality technologies that we think can potentially 'amplify' this error signal, and hopefully, by doing this, we can help these patients develop better insight."

Seeing this sort of intervention as potentially "transdiagnostic," Dr Leow said she believes that, for the brain to learn, "we need to be able to appreciate what we think we are doing and how that is different from what we are actually doing.

"For me, this has huge implications, not just in anorexia but in terms of how the brain can learn most efficiently.... This is the most exciting part of our research," she added.

"Compelling, Intriguing"

Commenting on the findings for Medscape Medical News, Joanna E. Steinglass, MD, associate professor of clinical psychiatry, Columbia University Medical Center, New York City, said that the findings are "compelling and intriguing."

Noting, however, that AN is "inherently a behavioral disorder," she said that "the insight component of this study seems somewhat less persuasive, but I would be interested to see if their findings apply also if you are looking at behaviors more central to the disorder."

As for insight being targeted for intervention, Dr Steinglass said that that is "a broad term."

"There's a long history of attempts to use psychoanalysis and psychodynamic treatments for eating disorders and for anorexia nervosa in particular, and one of the goals of psychodynamic psychotherapy is to increase insight.

"It's pretty clear that kind of therapy, while it may help people feel better, it does not improve outcomes; it doesn’t change weight, it doesn’t change eating behavior."

However, Dr Steinglass believes using virtual reality to try to improve insight "is a really interesting direction to go, and it would be interesting to see what happens if you apply it to behaviors."

The research was funded by a grant from the National Institute of Mental Health. The authors have disclosed no relevant financial relationships.

Psychol Med. Published online July 19, 2016. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.