Elsevier

Biological Psychiatry

Volume 50, Issue 3, 1 August 2001, Pages 171-178
Biological Psychiatry

Brain metabolic changes associated with symptom factor improvement in major depressive disorder

https://doi.org/10.1016/S0006-3223(01)01117-9Get rights and content

Abstract

Background: Symptoms of major depressive disorder (MDD) have been linked to regional brain function through imaging studies of symptom provocation in normal control subjects and baseline studies of subjects with MDD. We examined associations between change in depressive symptom factors and change in regional brain metabolism from before to after treatment of MDD.

Methods: Thirty-nine outpatients with MDD underwent 18F-fluorodeoxyglucose positron emission tomography scanning before and after treatment with either paroxetine or interpersonal psychotherapy. Associations were determined between changes in regional brain metabolism and changes in four Hamilton Depression Rating Scale factors (anxiety/somatization [ANX], psychomotor retardation [PR], cognitive disturbance [COGN], and sleep disturbance) and two corresponding Profile of Mood States subscales (tension [TENS] and fatigue [FATIG]).

Results: Improvement in ANX, PR, TENS, and FATIG factors was associated with decreasing ventral frontal lobe metabolism. Improvement in ANX and TENS was also associated with decreasing ventral anterior cingulate gyrus (AC) and anterior insula activity, whereas improvement in PR was associated with increasing dorsal AC activity. COGN improvement was associated with increasing dorsolateral prefrontal cortex metabolism.

Conclusions: Brain regions that show significant relationships with symptom provocation in normal control subjects have similar relationships with MDD symptoms as they improve with treatment.

Introduction

Recent reviews of the functional brain imaging literature in major depressive disorder (MDD) have concluded that frontal-cortical and limbic structures mediate the symptomatic expression of MDD Drevets 1998, Goodwin 1996, Grasby 1999, Kennedy et al 1997, Ketter et al 1996, Mayberg 1997, Videbech 2000. All of these reviews, however, note inconsistencies in the literature, leading to speculation that the clinical heterogeneity of MDD may account for variable imaging findings Drevets 1998, Goodwin 1996, Kennedy et al 1997, Ketter et al 1996, Videbech 2000, and that the examination of subtypes or symptom factors of MDD may be more informative than the examination of the syndrome as a whole (Videbech 2000). This approach has already been reported in baseline studies of MDD (see below) and obsessive-compulsive disorder (Rauch et al 1998).

The majority of work examining depressive symptoms comes from studies of normal control subjects, in which depressive symptoms are induced during scanning. The most common finding has been an association between anxiety (induced by medication, visual imagery, or recall of anxiety-provoking situations) and increased activity in the anterior cingulate gyrus (AC) Benkelfat et al 1995, Chua et al 1999, Gottschalk et al 1992, Javanmard et al 1999, Kimbrell et al 1999, Servan-Schreiber et al 1998, Wik and Wiesel 1991, particularly on the left Benkelfat et al 1995, Chua et al 1999, Gottschalk et al 1992, Kimbrell et al 1999, Wik and Wiesel 1991. Other brain areas commonly showing increased activity with induced anxiety are the inferior frontal gyrus (IFG)/anterior insula region Benkelfat et al 1995, Chua et al 1999, Kimbrell et al 1999, Liotti et al 2000, Servan-Schreiber et al 1998, ventral prefrontal cortex (VPFC) Benkelfat et al 1995, Chua et al 1999, Liotti et al 2000, Wik and Wiesel 1991, and anterior temporal lobe Benkelfat et al 1995, Chua et al 1999, Kimbrell et al 1999, Liotti et al 2000. These same brain regions (AC, IFG/anterior insula, VPFC, and anterior temporal lobe) have also been found to have increased regional cerebral blood flow in patients with anxiety disorders (obsessive-compulsive disorder, simple phobia, and posttraumatic stress disorder) when their symptoms are provoked Rauch et al 1994, Rauch et al 1995, Rauch et al 1996.

In addition to anxiety, transient sadness has been induced in normal control subjects during scanning using visual images or recollections of unhappy events. Such induced sadness has been associated with increased activity in the AC George et al 1995, Liotti et al 2000, Mayberg et al 1999, insula George et al 1995, George et al 1996, Liotti et al 2000, Mayberg et al 1999, and VPFC (Pardo et al 1993). One group Liotti et al 2000, Mayberg et al 1999 also found negative associations between induced sadness and activity in the dorsolateral prefrontal cortex (DLPFC).

Few studies have examined associations between regional brain activity and MDD symptoms directly. One group (Bench et al 1993) found positive correlations between anxiety and regional cerebral blood flow (rCBF) in the posterior cingulate cortex and inferior parietal lobule bilaterally, and between cognitive performance and rCBF in the left medial prefrontal cortex (some of these subjects were elderly and some had cognitive impairment). Negative correlations between psychomotor retardation and rCBF in the left DLPFC and left angular gyrus were also found. This group (Dolan et al 1994) also reported positive correlations between measures of cognitive performance (including memory, digit span testing, and serial sevens) and rCBF in the medial prefrontal cortex and AC bilaterally. Another group (Abercrombie et al 1998) found a positive correlation between right amygdala metabolism and negative affect in MDD patients.

In the classic factor analysis of the Hamilton Depression Rating Scale (HAM-D) used for the present study, sadness and psychomotor retardation (items 1 and 8, respectively) were found to cluster together in MDD subjects (Cleary and Guy 1975). Therefore, in the current study, sadness and psychomotor retardation were grouped together, and the following hypotheses were drawn based on this grouping of symptoms. It is important to note, however, that transient sadness as studied in normal control subjects may not be identical to the more sustained sadness and depressed mood experienced by patients with MDD studied here.

In synthesizing the above studies of MDD symptoms, several similarities and one difference between the brain mediation of measures of anxiety and psychomotor retardation/sadness were found. Both anxiety and psychomotor retardation/sadness were associated with increased activity in the VPFC, AC, insula, and anterior temporal lobe, whereas only psychomotor retardation/sadness has been associated with decreased activity in the DLPFC.

Although baseline associations between brain activity and MDD symptom factors have been examined, no one has yet reported associations between changes in brain activity and changes in MDD symptom factors with treatment. We have previously reported overall regional brain metabolic changes in subjects with MDD from before to after treatment with either paroxetine or interpersonal psychotherapy (IPT) (Brody et al, in press). In the present study, we undertook a preliminary investigation into the relationship between improvement in specific MDD symptom clusters and change in regional brain metabolism in these subjects, with the goal of establishing relationships between symptomatic change in MDD and regional metabolic change.

Based on literature cited above, we hypothesized that glucose metabolic changes in the VPFC, AC, insula, and anterior temporal lobe would have positive associations with changes in anxiety and psychomotor retardation/sadness factors, whereas change in DLPFC metabolism would have a negative association with change in the psychomotor retardation/sadness factor.

Section snippets

Subjects and brain scanning

Detailed methodology concerning subject characteristics, brain imaging techniques, and brain image processing has been reported previously (Brody et al, in press). Subjects were recruited from a general psychiatry screening phone service at the University of California, Los Angeles (UCLA) Neuropsychiatric Institute, and from newspaper advertisements. Eligible subjects were given a description of the study, and written consent was obtained using a form approved by the UCLA Office for Protection

Region of interest analysis

In the ROI analysis, changes in factors measuring anxiety (ANX and TENS) were significantly, positively associated with changes in the left VLPFC (t = 2.8, p = .008 and t = 2.5, p = .02, respectively), indicating that decreases in left VLPFC metabolism were associated with improvement in ANX and TENS (Table 1). Changes in both ANX (t = 3.0, p = .03) and TENS (t = 2.3, p = .03) were also positively associated with change in left ventral AC metabolism, indicating that decreases in left ventral

Discussion

In the present study, anxiety and sadness/psychomotor retardation were the depressive symptom factors most strongly linked to brain metabolic change with treatment. Decreases in both objective (ANX) and subjective (TENS) measures of anxiety were associated with decreases in ventral frontal lobe metabolism (in both the ROI and SPM analyses), with the strongest association on the left (in the ROI analysis). Similarly, improvement in both objective (PR) and subjective (FATIG) measures of

Acknowledgements

Supported in part by the National Alliance for Research in Schizophrenia and Depression (A.L.B. and L.R.B), a Veterans Administration Advanced Research Career Development Award (A.L.B.), the Charles A. Dana Foundation Consortium on Neuroimaging Leadership (S.S.), and R01 MH-53565 (L.R.B.).

References (48)

  • D.L Ackerman et al.

    Characteristics of fluoxetine versus placebo responders in a randomized trial of geriatric depression

    Psychopharmacol Bull

    (1997)
  • A Baddeley

    The fractionation of working memory

    Proc Natl Acad Sci U S A

    (1996)
  • C.J Bench et al.

    Regional cerebral blood flow in depression measured by positron emission tomographyThe relationship with clinical dimensions

    Psychol Med

    (1993)
  • C Benkelfat et al.

    Functional neuroanatomy of CCK-sub-4-induced anxiety in normal healthy volunteers

    Am J Psychiatry

    (1995)
  • A.L Brody et al.

    Regional brain metabolic changes in patients with major depression treated with either paroxetine or interpersonal psychotherapy

    Arch Gen Psychiatry

    (2001)
  • M Cleary et al.

    Factor analysis of the Hamilton Depression Scale

    Drugs Exp Clin Res

    (1975)
  • O Devinsky et al.

    Contributions of the anterior cingulate gyrus to behavior

    Brain

    (1995)
  • R.J Dolan et al.

    Dorsolateral prefrontal cortex dysfunction in the major psychosesSymptom or disease specificity?

    J Neurol Neurosurg Psychiatry

    (1993)
  • R.J Dolan et al.

    Neuropsychological dysfunction in depressionThe relationship to regional cerebral blood flow

    Psychological Med

    (1994)
  • W.C Drevets

    Functional neuroimaging studies of depressionThe anatomy of melancholia

    Annu Rev Med

    (1998)
  • L.F Fabre

    A 6-week, double-blind trial of paroxetine, imipramine, and placebo in depressed outpatients

    J Clin Psychiatry

    (1992)
  • J.P Feighner et al.

    A double-blind study of bupropion and placebo in depression

    Am J Psychiatry

    (1984)
  • K.J Friston et al.

    Spatial registration and normalization of images

    Hum Brain Mapp

    (1995)
  • K.J Friston et al.

    Statistical parametric maps in functional imagingA general linear approach

    Hum Brain Mapp

    (1995)
  • Cited by (0)

    View full text