Effects of heavy drinking, binge drinking, and family history of alcoholism on regional brain metabolites

Abstract

Background: The main goals are to investigate the effects of chronic active heavy drinking on N-acetylaspartate (NAA) and other metabolites throughout the brain and to determine whether they are affected by family history (FH) of alcoholism and long-term drinking pattern.

Methods: Forty-six chronic heavy drinkers (HD) and 52 light drinkers (LD) were recruited from the community and compared on measures of regional brain structure using magnetic resonance imaging and measures of common brain metabolites in gray matter (GM) and white matter (WM) of the major lobes, subcortical nuclei, brainstem, and cerebellum using short-echo time magnetic resonance spectroscopic imaging. Regional atrophy-corrected levels of NAA, myoinositol (mI), creatine, and choline-containing metabolites were compared as a function of group, FH of alcoholism, and bingeing.

Results: Frontal WM NAA was lower in FH-negative HD than FH-positive HD and tended to be lower in women than men. Creatine-containing metabolites in parietal GM were higher in HD than LD. FH-negative compared with FH-positive HD also had more mI in the brainstem and tended to have lower NAA and more mI in frontal GM. Although parietal GM NAA was not significantly lower in HD than LD, it was lower in non-binge drinkers than bingers. Frontal WM NAA was lower in HD than LD, with the difference driven by a small number of women, FH-negative HD, and older age. Lower frontal WM NAA in HD was associated with lower executive and working memory functions and with lower P3b amplitudes at frontal electrodes.

Conclusions: Community-dwelling HD who are not in alcoholism treatment have brain metabolite changes that are associated with lower brain function and are likely of behavioral significance. Age, FH, and binge drinking modulate brain metabolite abnormalities. Metabolite changes in active HD are less pronounced and present with a different spatial and metabolite pattern than reported in abstinent alcoholics.

Fig. 1

Example of one slice of a three-slice ¹H MRSI dataset acquired with TR and TE of 1800 and 25 msec, respectively: MRSI slice positions are overlaid on a midsagittal MRI, a T1-weighted MRI corresponding to the center of the top MRSI slice, corresponding metabolite maps with overlaid MRI outline (left), two examples of cortical spectra, their automatically determined baselines, and overlaid spectral fits.

Fig. 2

Regional brain metabolite concentrations, expressed in arbitrary units (a.u.), plotted against measures of drinking severity, domain summary scores of neuropsychological functioning, and auditory P3b amplitude (μV). Spearman correlations and significance levels are indicated for each plot.