260. A 1H-MRSI hippocampal study in bipolar patients with a history of alcohol abuse

Abstract

Bipolar illness has the highest lifetime prevalence rate for co-morbid substance abuse, specifically alcohol, than other Axis I diagnoses. MRI studies have shown the hippocampus to be susceptible to the neurotoxicity of alcohol. The purpose of this study was to evaluate the aggregate neuropathology of this co-morbidity by proton magnetic resonance spectroscopic imaging (1H-MRSI). This study is a post-hoc analysis of our previous 1H-MRSI study which found decreased NAA in bilateral hippocampi in bipolar patients compared to controls (Bertolino et al., 1999). 17 patients with bipolar I disorder were scanned on a 1.5T GE magnet (pulse sequence; TR 2200 msec, TE 272 msec, voxel resolution 1.4 ml). Metabolite levels of NAA, creatine + phosphocreatine (CRE), and choline compounds (CHO) were measured as ratios of the areas under each peak. Blind to the 1H-MRSI data, the medical record and SKID were reviewed to identify a history of alcohol abuse (x sobriety: 55 months). All patients were symptomatic only of their mood disorder and were not currently abusing alcohol. Unpaired t-tests were utilized to evaluate diagnostic group differences. The mean NAA/CHO ratio in 1 & r hippocampus was higher in patients with a history of alcohol abuse in comparison to patients without such history (1: 1.28 ± 0.21 vs. 1.04 ± 0.27; n = 17, t = 1.84, p = 0.08; r: 1.26 ± 0.15 vs. 1.07 ± 0.18; t = 2.35, p = 0.03). Similar findings were noted for NAA/CRE ratio in 1 hippocampus (1: 1.64 ± 0.24 vs. 1.34 ± 0.27; t = 2.29, p = 0.036 r: 1.63 ± .17 vs. 1.35 ± 0.3, t = 2.09, p = 0.053). These data were particularly striking when analyzed by gender for men. (NAA/CHO: 1; 1.29 ± 0.24 vs. 0.93 ± 0.05, n = 10, t = 3.46, p = 0.009, r; 1.29 ± 0.15 vs. 1.02 ± 0.13, t = 2.98, p = 0.018) (NAA/CRE: 1; 1.66 ± 0.27 vs. 1.21 ± 0.09, t = 3.49, p = 0.008, r; 1.67 ± 0.14 vs. 1.31 ± 0.30, t = 2.43, p = 0.04). These data suggest differential 1H-MRS pathology in bipolar patients with and without a history of alcohol abuse. As previous studies have reported a decrease in NAA in bipolar patients, these ratios, in patients with a history of alcohol abuse may suggest a greater decrease in CHO and CRE signals or a reactive increase in NAA. To what extent these abnormalities suggest differential degrees of neuronal loss (NAA) and/or subsequent glial hyperplasia (CHO) and the persistence of these abnormalities despite abstinence require further controlled investigation.