Purpose: Higher homocysteine levels were found in actively drinking patients with alcohol dependence. Recent studies have shown that high homocysteine levels are associated with alcohol-withdrawal seizures. The aim of the present study was to calculate the best predictive cutoff value of plasma homocysteine levels in actively drinking alcoholics (n = 88) with first-onset alcohol-withdrawal seizures.
Methods: The present study included 88 alcohol-dependent patients of whom 18 patients had a first-onset withdrawal seizure. All patients were active drinkers and had an established diagnosis of alcohol dependence, according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). Sensitivity and specificity were calculated by using every homocysteine plasma level found in the study population as cut-off value. A Bayes theorem was used to calculate positive (PPV) and negative (NPV) predictive values for all cutoff values used.
Results: The highest combined sensitivity and specificity was reached at a homocysteine plasma cutoff value of 23.9 microM. Positive predictive values ranged from 0.23 to 0.745; the maximum was reached at a homocysteine plasma level of 41.7 microM. Negative predictive values ranged from 0.50 to 0.935, with a maximum at a homocysteine plasma level of 15.8 microM.
Conclusions: Homocysteine levels above this cutoff value on admission are a useful screening tool to identify actively drinking patients at higher risk of alcohol-withdrawal seizures. This pilot study gives further hints that biologic markers may be helpful to predict patients at risk for first-onset alcohol-withdrawal seizures.