Regional patterns of brain metabolites in AIDS dementia complex

Neuroimage. 2004 Nov;23(3):928-35. doi: 10.1016/j.neuroimage.2004.07.033.

Abstract

The relationship of the cellular changes in the HIV-infected brain to the onset and progression of AIDS dementia complex (ADC) remains uncertain. We undertook an in vivo proton magnetic resonance spectroscopy (MRS) study and used factor analysis to identify specific cellular and regional brain changes that may serve as metabolic markers of ADC. The ratio of N-acetyl aspartate (NAA), choline (Cho), and myoinositol (MI) over creatine (Cr), markers of neuronal and glial cell metabolism, were measured in the basal ganglia, centrum semiovale, and parietal cortex from 100 subjects with and without ADC. Three metabolic patterns were identified, which we termed "inflammatory" (mainly MI/Cr elevations in all three regions plus Cho/Cr increases in the centrum semiovale and parietal cortex), "basal ganglia" (mostly NAA/Cr and Cho/Cr elevations in the basal ganglia), and "neuronal" (primarily NAA/Cr reductions in the centrum semiovale and the parietal cortex). Logistic regression analysis revealed that, adjusted for age, basal ganglia and neuronal pattern scores were strongly associated with ADC but inflammatory levels were not. We conclude that by using factor analysis, we are able to combine multiple metabolites across brain regions in a biologically plausible manner and construct a predictive model of ADC adjusting for relevant factors such as age.

Publication types

  • Clinical Trial
  • Multicenter Study
  • Research Support, Non-U.S. Gov't
  • Research Support, U.S. Gov't, P.H.S.

MeSH terms

  • AIDS Dementia Complex / metabolism*
  • AIDS Dementia Complex / pathology
  • Adult
  • Algorithms
  • Basal Ganglia / metabolism
  • Basal Ganglia / pathology
  • Brain / pathology
  • Brain Chemistry / physiology*
  • Encephalitis / metabolism
  • Encephalitis / pathology
  • Factor Analysis, Statistical
  • Female
  • Humans
  • Image Processing, Computer-Assisted
  • Logistic Models
  • Magnetic Resonance Imaging
  • Male
  • Neurons / metabolism
  • Neurons / pathology