BACKGROUND: Our purpose was to evaluate the relationship between valproic acid (VPA)-induced hyperammonemia (HA) and the prevalence of minimal hepatic encephalopathy (MHE) cognitive impairment among psychiatric inpatients.
METHODS: Fifty-two psychiatric inpatients prescribed VPA were tested for MHE impairment after achieving steady-state VPA and ammonia concentrations during hospitalization between December 2013 and June 2014. The relationship between steady-state VPA and ammonia concentration was tested by correlation coefficient. Patients completed a battery of 5 psychometric tests that determined a Psychometric Hepatic Encephalopathy Score (PHES), which was used to test the association between a PHES <−4 cutoff for MHE impairment and HA exposure (ammonia >50 μmol/L) by chi-square testing.
RESULTS: Steady-state VPA plasma concentration was not correlated significantly with ammonia concentration (r = 0.24, P = .093). The patients with HA did not have a higher proportion of MHE cognitive impairment than patients with normal ammonia exposure (43.8% vs 66.7% respectively, P = .806).
CONCLUSIONS: Steady-state VPA concentration was not correlated with ammonia concentration and VPA-induced HA was not associated with a greater prevalence of MHE impairment. This suggests that the MHE impairment PHES cutoff might not detect VHE in psychiatric inpatients without cirrhosis, especially because inpatients could experience cognitive impairment related to acute mental illness.