Plasma homocysteine levels and cognitive status in long-term stay geriatric patients: a cross-sectional study.

Many studies have established an association of total plasma homocysteine (Hcys) levels and the risk for dementia and Alzheimer disease. However, little is known on the relation between Hcys and cognitive status in long-term stay geriatric patients. This cross-sectional study was conducted in a long-stay geriatric center and included 186 eligible patients, staying for more than 6 months in the facility. We looked for patients' clinico-demographic data, as well as for laboratory data and cognitive status, evaluated by the mini-mental state examination (MMSE). Patients were divided into three groups according to Hcys levels and analyzed in order to explore potential association with cognitive levels, considering clinically important cofounders. Cognitive state of patients of the highest Hcys tertile was significantly impaired, compared with other groups (p < 0.0001) and characterized by a worse nutritional status. Pearson's correlation coefficient for Hcys and MMSE resulted -0.251 (p = 0.0005). The linear regression model for MMSE showed that homocysteine is significantly associated with MMSE: mean MMSE score of patients with homocysteine >13.7 was significantly lower than the mean MMSE of patients with homocysteine < or = 8.5, after controlling for all other parameters in the regression equation (beta = -2.7685, p = 0.007). We conclude that in our sample of patients, Hcys was associated with cognitive impairment. Hcys levels >13.7 micromol/l remained a significant independent parameter associated with the severity of cognitive impairment, after controlling for other confounders. A better understanding of the mechanisms underlying such interrelations may have research and interventional implications.