NUTRITIONAL STATUS AND BODY COMPOSITION BY BIOELECTRICAL IMPEDANCE VECTOR ANALYSIS: A CROSS-SECTIONAL AND LONGITUDINAL STUDY IN MILD COGNITIVE IMPAIRMENT AND ALZHEIMER¿S DEMENTIA.

Aim To analyze nutritional status and body composition in Alzheimer’s disease (AD) and Mild Cognitive Impairment (MCI) with respect to elderly healthy controls (HC), in order to find any biomarker of disease. Methods A cross-sectional and a longitudinal study was performed in a memory clinic of a University-Hospital, by recruiting patients with mild-moderate AD, subjects with MCI and HC. Nutritional status was assessed at baseline for all the subjects and repeated at follow-up in AD patients by anthropometric parameters (body mass index; calf, upper arm and waist circumferences), Mini Nutritional Assessment (MNA) and body composition by bioelectrical impedance vector analysis (BIVA). For the cross-sectional study variables were analyzed by analysis of variance and subjects were grouped by cognitive status and gender; for the cross-sectional study in AD variables were analyzed by dependent t-test for repeated measures and linear regression analysis within gender. MCI were follow-up by repetition of neuropsychological tests to detect a potential progression to dementia and AD. MCI subjects’ characteristics at baseline were compared by outcome using non-parametric statistics. Results Sociodemographic variables did not differ among the three groups (59 mild AD, 34 MCI and 58 HC), except for females’ age, which was therefore used as covariate in a general linear multivariate model. MNA score was significantly lower in AD patients than in HC; MCI subjects achieved intermediate scores. AD patients (both sexes) had significantly (p<0.05) higher height-normalized impedance values and lower phase angles (body cell mass) compared with HC; a higher ratio of impedance to height was found in men with MCI with respect to HC. With BIVA method, MCI subjects showed a significant displacement on the RXc graph on the right side indicating lower soft tissues (Hotelling’s T2 test: men=10.6; women=7.9;p < 0,05) just like AD patients (Hotelling’s T2 test: men=18.2; women=16.9; p<0,001). After 8.7 ± 3.6 months of follow-up, bioelectrical variables of 40 AD patients did not significantly change. Forty-three MCI (28 females, 15 males) were followed up for 14.4 ± 8.6 months; 8 (6 females, 2 males) of them progressed to AD. Due to the limited number of males progressed, only females MCI’s bioelectrical characteristics were analyzed and those who progressed to clinically evident AD showed lower phase angles than stable MCI with considerable trend toward significance (5.9 1.0 vs. 5.2 0.6, p 0.069). Conclusion Bioelectrical parameters significantly differ from MCI and AD to HC, but remain stable after approximately 9 months of AD patients’ follow-up; MCI who progressed to clinically evident AD had a lower PA which approached the borderline of significance. Analysis of body composition with BIVA could detect early changes in body composition which may perhaps reflect early systemic manifestation of the AD process at MCI stage of disease, before anthropometric change becomes evident. Increasing the cohort of MCI and their longitudinal observation will provide further information to understand if a BIVA pattern indicating a worse nutritional status could be an early and sensitive marker of progression to dementia or specifically to AD in MCI subjects.