A cross-sectional and follow-up voxel-based morphometric MRI study in adolescent anorexia nervosa

https://doi.org/10.1016/j.jpsychires.2008.03.013Get rights and content

Abstract

The objective was to examine whether cerebral volumes are reduced, and in what regions, in adolescents with anorexia nervosa and to study changes after nutritional recovery. Twelve anorexia nervosa (DSM-IV) patients aged 11–17 consecutively admitted to an Eating Disorders Unit were assessed by means of psychopathological scales, neuropsychological battery and voxel-based morphometric (VBM) magnetic resonance imaging at admission and after 7 months’ follow-up. Nine control subjects of similar age, gender and estimated intelligence level were also studied. The two groups showed differences in gray matter (F = 22.2; p < 0.001) and cerebrospinal fluid (CSF) (F = 21.2; p < 0.001) but not in white matter volumes. In anorexic patients, gray matter volume correlated negatively with the copy time from the Rey Complex Figure Test. In the regional VBM study several temporal and parietal gray matter regions were reduced. During follow-up there was a greater global increase in gray matter (F = 10.7; p = 0.004) and decrease in CSF (F = 22.1; p = 0.001) in anorexic patients. The increase in gray matter correlated with a decrease in cortisol (Spearman correlation = −0.73; p = 0.017). At follow-up there were no differences in global gray matter (F = 2.1; p = 0.165), white matter (F = 0.02, p = 0.965) or CSF (F = 1.8; p = 0.113) volumes between both groups. There were still some smaller areas, in the right temporal and both supplementary motor area, showing differences between them in the regional VBM study. In conclusion, in adolescent anorexic patients gray matter is more affected than white matter and mainly involves the posterior regions of the brain. Overall gray matter alterations are reversible after nutritional recovery.

Introduction

Anorexia nervosa is a disorder that tend to begin in adolescence and has severe physical morbidity and psychosocial disability as well as the highest mortality of any psychiatric disorder (Sullivan, 1995). The disorder can cause severe medical complications even in adolescent patients with a relatively short duration of disorder such as heart abnormalities (Mont et al., 2003) or osteopenia (Castro et al., 2001). Endocrine alterations are also common in malnourished anorexia nervosa patients, specially low level of thyroid hormones and insulin-like growth factor-I (IGF-I) or high levels of cortisol (Stoving et al., 1999, Castro et al., 2004). There is also evidence of brain abnormalities in these patients. Studies using computer tomography and magnetic resonance imaging (MRI) show that, while underweight, anorexia nervosa patients have enlarged sulci and ventricles and decreased brain mass (Nussbaum et al., 1980, Datlof et al., 1986, Dolan et al., 1988, Kornreich et al., 1986). With weight restoration these abnormalities tend to improve, but it is still not clear whether they normalize completely. Some studies have found a complete reversibility of cerebral changes with weight restoration (Swayze et al., 1996, Golden et al., 1996), but others have not (Kingston et al., 1996). Swayze et al. (1996) studied anorexic patients using MRI and found that their ventricular volume significantly decreased with weight restoration, but they did not have a control group. Kingston et al. (1996) also reported enlarged ventricular volumes in the anorexia nervosa group using MRI. At follow-up, many patients with a 10% weight gain had a decrease in measures of ventricular and sulcal volume. With adolescents, Golden et al. (1996) used MRI to find that 12 female anorexia nervosa patients had enlarged lateral ventricles while underweight that then normalized after weight recovery.

Moreover, from these studies it was not clear whether the changes occurring while underweight are related to gray matter or white matter. Katzman et al. (1996) studied gray and white matter volumes rather than ventricular volumes alone, using global volumetric MRI in 13 underweight adolescents with anorexia nervosa. They found increased ventricular and sulcal cerebrospinal fluid (CSF) volumes and significant decrease in gray and white matter volumes in comparison with 8 healthy controls. In another study of the same group (Lambe et al., 1997) with 12 weight recovered anorexic patients (mean age 18.9 years; mean of weight recovered period = 2.4 years), 18 healthy subjects and 13 anorexic patients still underweight, found that recovered patients have significant deficits in gray matter volumes but were not different from controls in white-matter volumes. The same authors followed 6 patients of the 13 from the first study, and carried out a second MRI 2–3 years after the first one and found significant increases in white matter volume and decreases in CSF volume, but significant deficits in gray matter volume and enlarged sulci persisted (Katzman et al., 1997). Swayze et al. (2003) found total white matter and several regional white matter volumes significantly reduced in 18 adult anorexia nervosa patients. Total gray matter was not significantly reduced but some gray matter regions, such as the frontal and parietal regions, were. White and gray matter volumes increased at follow-up after weight recovery. In the only study with anorexic patients using a voxel-based morphometric analysis (VBM) on magnetic resonance images, Wagner et al. (2006) found that long-term recovered anorexia and bulimia nervosa adult patients were similar to controls in total and regional gray and white matter volumes. Nevertheless they did not show results of the regional VBM study and only overall values were given. Moreover, evaluation was done only once in already recovered patients in a cross-sectional design. At present, it remains uncertain whether reductions in gray and white matter are generalized brain abnormalities or are specific to concrete regions; whether they affect gray matter more, as some studies seem suggest; and, above all, whether, at least in young patients, these reductions recover completely with nutritional normalization. Using longitudinal MRI and each individual’s first examination as the reference, changes can be assessed over time. Moreover, novel automatic methods for image processing such as VBM can provide more reliable results than methods that manually trace regions of interest, because the latter are observer-dependent and imply a decision, a priori, concerning what structures to assess (Connan et al., 2006). Regional voxel-based morphometry gives a more rapid and extensive survey of different gray and white matter abnormalities than manually drawn region of interest analysis the most common method of analysis in structural MRI.

To our knowledge, no longitudinal studies have been carried out using VBM on low-weight and short-term weight recovered adolescent anorexic patients. The objective of the present study was therefore to evaluate the cerebral structural abnormalities – overall as well as regional – analyzed with the use of optimized voxel-based morphometric magnetic resonance imaging, in children and adolescents who had anorexia nervosa, and to assess their changes longitudinally after weight recovery.

Section snippets

Subjects and procedures

The group of patients comprised 12 children and adolescents, 11 girls and 1 boy, aged 11–17, who fulfilled the DSM-IV-TR diagnostic criteria for anorexia nervosa. Diagnosis was made by clinical interview. All patients were consecutively admitted to the Eating Disorders Unit of the Child and Adolescent Psychiatry and Psychology Department at the Hospital Clinic, University of Barcelona (Barcelona, Spain). Clinical variables were recorded and a range of scales were administered. Hormonal data

General characteristics of anorexic patients

Mean body mass index was 14.8 (SD = 2.0) and mean weight loss was 26.8% (SD = 8.4). Nine patients (75%) were restrictive and 3 (25%) were purging. Mean period between the onset of disorder and first evaluation was 8.3 (SD = 3.1) months. One (8.3%) patient at first evaluation and 3 (25%) patients at follow-up were under psychopharmacological treatment with fluoxetine or fluvoxamine.

Comparison between anorexic patients and controls at first assessment

Comparison between anorexic patients and control group at first evaluation in demographic and clinical variables is shown

Discussion

To our knowledge, this is the first VBM study to analyze differences between adolescent anorexia nervosa patients and controls, and changes after treatment. Moreover, it is the first time that regional VBM has been carried out in anorexic patients. The main finding in the present study is that adolescent anorexic patients, even after relatively short duration of the disorder, have lower overall gray matter and higher CSF cerebral volumes than controls, and that this global difference disappears

Conflict of interest

The authors reported no biomedical financial interests or potential conflicts of interest.

Contributors

The authors of this paper are the contributors of this study.

Role of funding source

Study supported by the Spanish Ministry of Health, Instituto de Salud Carlos III (ISCIII), CIBERSAM Research Network. The ISCIII had no further role in study design; in the collection, analysis and interpretation of data; in the writing the report; and in the decision to submit the paper for publication.

Acknowledgements

Study supported by the Spanish Ministry of Health, Instituto de Salud Carlos III, CIBERSAM Research Network.

References (31)

  • M. Nussbaum et al.

    Cerebral atrophy in anorexia nervosa

    Journal of Pediatrics

    (1980)
  • R.K. Stoving et al.

    A review of endocrine changes in anorexia nervosa

    Journal of Psychiatric Research

    (1999)
  • N. Tzourio-Mazoyer et al.

    Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain

    Neuroimage

    (2002)
  • A. Wagner et al.

    Normal brain tissue volumes after long-term recovery in anorexia nervosa and bulimia nervosa

    Biological Psychiatry

    (2006)
  • J. Castro et al.

    The Eating Attitudes Test: validation of the Spanish version

    Evaluacion Psicologica

    (1991)
  • Cited by (144)

    View all citing articles on Scopus
    View full text