Experimental Gerontology 58 (2014) 30–33

Contents lists available at ScienceDirect

Experimental Gerontology

j ourna l homepage: www.e lsev ie r .com/ locate /expgero

Lower lean mass and higher percent fat mass in patients withAlzheimer's disease

R. Buffa a, E. Mereu a, P. Putzu b, R.M. Mereu b, E. Marini a,⁎a Department of Environmental and Life Sciences, University of Cagliari, Cittadella Universitaria, 09042 Monserrato, Italyb Geriatric Division, SS Trinità Hospital, ASL 8, Cagliari, Italy

Abbreviations: R, resistance; Xc, reactance; Z, impedanspecific reactance; Zsp, specific impedance; BIVA, bioelectrFM, fatmass; DXA, dual-energy X-ray absorptiometry;MMtion; ADL, activities of daily living; GDS, geriatric depressiassessment; ICW, intracellular water; ECW, extracellular w⁎ Corresponding author at: Department of Environmen

of Cagliari, Cittadella Universitaria, 09042 Monserrato, CaE-mail address: emarini@unica.it (E. Marini).

http://dx.doi.org/10.1016/j.exger.2014.07.0050531-5565/© 2014 Elsevier Inc. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 5 April 2014Received in revised form 9 July 2014Accepted 11 July 2014Available online 11 July 2014

Section Editor: Holly M Brown-Borg

Keywords:Alzheimer's diseaseBody compositionNutritional statusBioelectrical impedance vector analysisCognitive impairmentFunctional decline

In this study we analyzed body composition in relation to cognitive and functional status, in a cross-sectionalsample of patients with Alzheimer's disease (AD).Seventy individuals (27 men, 78.1 ± 6.5 years; 43 women, 80.4 ± 5.6 years) with mild–moderate stages of AD(clinical dementia ratings 1 and 2)were selected from the Alzheimer Center, SS. Trinità Hospital, ASL 8 of Cagliari(Italy). Cognitive and psycho-functional status was evaluated using mini-mental state examination (MMSE), ac-tivities of daily living (ADL) scale, and geriatric depression scale (GDS). Mini-nutritional assessment (MNA) wasapplied. Anthropometric measurements were taken and body mass index (BMI) was calculated. Body composi-tionwas assessed bymeans of specific bioelectrical impedance vector analysis (BIVA), using the references for theelderly. In comparisonwith the reference group, patientswith AD showed similar BMI andMNA, but peculiar bio-electrical characteristics: lower phase angles and longer vectors (p b 0.05). According to specific BIVA, this bio-electrical pattern is indicative of a reduction of lean tissue mass and an increase of percent fat mass (FM%). Amore accentuated lean mass reduction (p b 0.05) was observed in women with worse cognitive status and aFM% increase (p b 0.01) in women with worse functional status.In conclusion, patients with AD had lower lean tissue mass and higher percent fat mass than healthy elderly in-dividuals. Inwomen, this patternwas associatedwith cognitive and functional decline, as indicated byMMSE andADL values. Specific BIVA showed to be a suitable technique in the elderly, that could enhance BMI and MNA in-formation in the evaluation of nutritional status.

ce; Rsp, specific resistance; Xcsp,ical impedance vector analysis;SE,mini-mental state examina-on scale; MNA, mini-nutritionalater.

tal and Life Sciences, Universitygliari, Italy.

© 2014 Elsevier Inc. All rights reserved.

1. Introduction

Epidemiological studies show that body composition variations arelinked to both the onset and the progression of Alzheimer's disease(AD). The nature of the relationship and the underlying causal factorsare not completely defined because of the phenomenological complex-ity of the disease, the age-related body composition variations, and therisk factor changes across the life course.

According to some authors, high levels of fat mass during mid-lifeare associated with late-onset AD (Luchsinger and Gustafson, 2009).However, weight loss is a typical feature of the disease, and is associatedwith disease severity and faster clinical progression, and can appear be-fore the recognition of clinical symptoms of AD (Gillette-Guyonnetet al., 2000). Various causal factors, such as anorexia or neuropsychiatric

disturbances, metabolic alterations, and medial temporal cortex atro-phy, have been proposed for weight loss (Gillette-Guyonnet et al.,2000). On the other hand, undernutrition itself could accelerate cogni-tive and functional decline, via the reduction of leptin levels, due tothe reduced fat mass (Soto et al., 2012), or via sarcopenia and frailty,due to the reduced lean mass (Lee, 2011).

The comprehension of such relationships requires an accurate defi-nition of body composition variations. However,most researches use in-direct techniques, such as anthropometry. Body mass index (BMI) is ananthropometric indicator of undernutrition, overweight and obesity,largely used in epidemiologic studies. However, it is sensitive to bothmuscular and fat mass and may be unable to disentangle their relativecontribution. This limit is particularly relevant in the elderly, wherefrailty, sarcopenia or sarcopenic obesity syndromes are linked to bodycomposition variations, in which the physiological decline of leanmass can be associated to an increase of fat mass, without weightchanges.

Specific bioelectrical impedance vector analysis (BIVA sp) is a recent-ly validated technique for assessing body composition (Buffa et al.,2013; Marini et al., 2013). It is accurate, safe, time-saving and cost-effective, and hence represents a promising tool both in routine clinicalpractice and in epidemiologic studies.

31R. Buffa et al. / Experimental Gerontology 58 (2014) 30–33

The aim of this study was to analyze the body composition charac-teristics of elderly patients with mild to moderate Alzheimer's diseasein relation to their cognitive and functional status, using specific BIVA.

2. Subjects and methods

2.1. Subjects

The study group consisted of 70 individuals (27men and 43women)with mild–moderate stages of Alzheimer's disease (CDR 1 and 2), se-lected from the Alzheimer Center, SS. Trinità Hospital, ASL 8 of Cagliari(Italy). The mean age was 78.1 ± 6.5 years for men and 80.4 ±5.6 years for women.

The Alzheimer's disease was diagnosed according to the NINCDS-ADRDA criteria (National Institute of Neurological and CommunicativeDisorders and Stroke, and the Alzheimer's Disease and Related Disor-ders Association) and the stage of dementia was assessed by the CDRscale (clinical dementia rating).

All patients had been treatedwith theAchE-inhibitor donepezil (5 to10 mg per day). The mean length of therapy was 4.4 (±2.2) years inmen and 4.6 (±1.8) years in women.

In accordance with the Helsinki Declaration of 1975, as revised in2013, all patients or their tutorswere informed about the researchproto-col and they consented to take part in the study. Exclusion criteria were:a) physical handicaps; b) uncompensated chronic diseases (uncontrolledtumoral pathology, III–IV class cardiac decompensation, chronic gastro-intestinal disease, renal or hepatic insufficiency); c) nutritional support.

The reference population consisted of 560 healthy individuals(265 men and 295 women) aged 65 to 100 years (women: 76.0 ±7.1 years; men: 77.0 ± 7.2 years), all born in Italy and recruited on avoluntary basis. The specific bioelectrical characteristics of the reference,and their interpretation, have been described elsewhere (Saragat et al.,2014). Detailed information on cognitive (based on mini-mental stateexamination, MMSE), psychological (geriatric depression scale, GDS),functional (activities of daily living, ADL) and nutritional (mini-nutri-tional assessment, MNA) status is available for a sub-sample of the ref-erence population (collectively, 130 men and 120 women) (Buffa et al.,2010a; Saragat et al., 2012). This sub-sample shows, in average, a nor-mal functional (ADL; men: 5.3± 0.5; women: 5.8 ± 0.6), psychological(GDS; men: 2.4 ± 1.9; women: 3.6 ± 3.3), and nutritional status (nearthe threshold for malnutrition risk, in women) (MNA; men: 25.5 ± 2.6;women: 24.2 ± 2.8), and a quite normal cognitive condition (MMSE;men: 25.3 ± 2.8; women: 23.0 ± 4.1).

2.2. Measurements

The cognitive and psycho-functional status was assessed at the mo-ment of the survey, using mini-mental state examination (Folstein et al.,1975), activities of daily living scale (Katz et al., 1963), and 15 item geri-atric depression scale (Sheikh and Yesavage, 1986). The mini-nutritionalassessment (Guigoz et al., 1994) was applied. Anthropometric (stature,weight, upper arm, waist, and calf circumferences) and bioelectrical im-pedance measurements (50 kHz and 0.8 A, with a single-frequency im-pedance analyzer; BIA 101, Akern, Firenze, Italy) were taken accordingto standard procedures by a single experienced operator (RB).

2.3. Statistical analyses

According to the procedure described by Marini et al. (2013) and byBuffa et al. (2013), the specific bioelectrical impedance vector analysiswas applied. The method has been recently proposed as an extensionof the classic BIVA, conceived by Piccoli et al. (1994), and its efficacyhas been verified by comparison with dual-energy X-ray absorptiome-try (DXA).

In order to eliminate the effect of “conductor” length (L) and cross-sectional area (A) on bioelectrical values, R and Xc are multiplied

by a correction factor A/L (in meters), where A (m2) = 0.45 armarea + 0.10 waist area + 0.45 calf area, and L (m) = 1.1 height.Cross-sectional areas are estimated as C2/4Л (m2), where C (m) refersto segmental circumference. Specific bioelectrical values, rescaled multi-plying by 100 (resistivity, Rsp, and reactivity, Xcsp, Ohm·cm), can be an-alyzed with the same statistical procedure of classic BIVA. The phaseangle (degrees) is calculated as arctan (Xc/R); the impedivity vector(Zsp) as (Rsp2 + Xcsp2)0.5.

The major axis of specific tolerance ellipses refers to bioelectricalvariations associated with changes of the relative quantity of fat mass(individuals with a higher FM% toward the upper pole). The minoraxis refers to changes of the lean tissue mass (left side: more leanmass; right side: less lean mass) and of extracellular/intracellularwater ratio (ECW/ICW; with higher values on the right side).

Anthropometric and bioelectric variables of patients were comparedwith the reference by means of two-way ANOVA, considering sex andhealth status effects.

For each psycho-functional variable, and for length of therapy, thesample of patients with AD was divided into two groups (below andabove the median) and bioelectrical and anthropometric variableswere compared by means of the Student t-test. The difference betweenthe mean specific impedance vectors was assessed by means ofHotelling's T2 test.

Statistical analysis was performed bymeans of Systat package (13.1)and of the newly-assembled specific BIVA software (open access versionunder a Creative Commons Attribution-NonCommercial-ShareAlike Li-cense, on the website http://specificbiva.unica.it/).

3. Results

Patients with Alzheimer's disease showed mild-to-moderate cogni-tive impairment (MMSE, men: 19.4 ± 5.6; women: 19.0 ± 4.9), andpsychological (GDS, men: 3.3 ± 2.9; women: 4.2 ± 3.5) and functionaldeclines (ADL, men: 4.3 ± 1.3; women: 3.7 ± 1.6). MNA mean valuewas equal to 25.5 (±3.1) in men (81.5% of well-nourished individuals)and to 24.3 (±4.1) in women (64.4% of well-nourished individuals), al-most the same levels as the reference.

A significant sexual dimorphism was observed for height, weight,and bioelectrical variables, with higher phase angle and lower specificresistance and impedance in men (Table 1).

In comparison with the Italian reference sample, patients of bothsexes showed lower height and weight values, but a similar BMI(Table 1). Bioelectrical vectors of patients (both sexes) were character-ized by low phase angles and high specific impedance values, and werealmost entirely (95.7%) located on the right side of the specific toleranceellipses of the Italian elderly (Saragat et al., 2014) (Table 1).

In women, bioelectrical characteristics changed according to cogni-tive and functional status, but not to the GDS scale of depression.Women with worse conditions, i.e. with MMSE and ADL values belowthemedian, showed lower phase (p= 0.010) and higher specific vectorlength (p= 0.024), respectively (Fig. 1). These bioelectrical characteris-tics can be related to a low fat mass and high ECW/ICW (as indicated bythe lower phase), and high FM% (as indicated by the longer vector). BMIvalues were higher in the groups with worse cognitive and functionalstatus, significantly in the case of ADL (p = 0.002). In men the patternwas similar but the differences were not significant.

MNA values were not significantly different in patients with differ-ent cognitive and functional statuses.

In both sexes, no bioelectrical difference was observed between pa-tients with a different length of therapy (T2 = 3.5; p = 0.202).

4. Discussion

In this study, patients with mild to moderate stage Alzheimer's dis-ease had lower body weight, different body compositions, but similarBMI and MNA, with respect to the reference. According to specific

Table 1Two-way ANOVA for the comparison between patients with Alzheimer's disease and the reference population.

Men Women Fcondition Fgender Fcondition × gender

Patients Reference Patients Reference

Mean ± s.d.f Mean ± s.d.f Mean ± s.d.f Mean ± s.d.f

Height (cm) 160.3 ± 7.2 162.0 ± 8.5 145.5 ± 5.5 150.2 ± 8.0 11.614⁎⁎⁎ 191.433⁎⁎⁎ 2.080Weight (kg) 67.6 ± 9.0 69.5 ± 11.1 56.5 ± 11.8 60.1 ± 11.0 4.416⁎ 56.774⁎⁎⁎ 0.349BMIa (kg/m2) 26.4 ± 3.6 26.4 ± 3.3 26.7 ± 5.5 26.6 ± 4.1 0.004 0.241 0.023Rbspc (Ohm·cm) 454.0 ± 48.9 391.8 ± 57.9 509.0 ± 90.7 462.0 ± 80.1 34.800⁎⁎⁎ 45.782⁎⁎⁎ 0.676Xcdspc (Ohm·cm) 42.8 ± 6.3 42.6 ± 9.9 44.8 ± 9.6 47.9 ± 11.2 1.115 7.193⁎⁎ 1.550Zespc (Ohm·cm) 456.1 ± 49.1 394.2 ± 58.2 511.0 ± 91.0 464.6 ± 80.5 33.904⁎⁎⁎ 45.398⁎⁎⁎ 0.408Phase (°) 5.4 ± 0.7 6.2 ± 1.2 4.9 ± 0.7 5.9 ± 1.1 52.130⁎⁎⁎ 8.013⁎⁎ 0.306

⁎ = p b 0.05.⁎⁎ = p b 0.01.⁎⁎⁎ = p b 0.001.

a Body mass index.b Resistance.c Specific.d Reactance.e Impedance.f Standard deviation.

32 R. Buffa et al. / Experimental Gerontology 58 (2014) 30–33

BIVA, patients of both sexes showed a reduction of the leanmass and anincrease of ECW/ICWandof the percent fatmass. In fact, specific bioelec-trical vectors of patients were characterized by low phase angles andhigh impedance values, and were almost entirely located on the rightside of the specific tolerance ellipses of the Italian elderly (Saragatet al., 2014). This side corresponds to high values of ECW/ICW (Buffaet al., 2013) and to low values of fat-free mass (Marini et al., 2013), ofappendicular muscle mass in particular (Buffa et al., 2013; Saragatet al., 2014). The shift of the bioelectrical pattern toward the upper-right area is indicative of high relative quantities of fat mass (Buffaet al., 2013; Marini et al., 2013) and of sarcopenic obesity (Mariniet al., 2012).

As to hydration status, a previous analysis on a partially overlappingsample of patients of AD showed that a mean tendency to dehydrationwas present in the mild–moderate stage, becoming more apparent inthe severe stage (Buffa et al., 2010b).

The analysis of intra-sample variability confirmed the above de-scribed bioelectrical pattern, showing that body composition peculiari-ties of patients with AD accentuate with the progression of the disease.To be noted that these changeswere not related to the length of therapy

Fig. 1. Distribution of specific impedance vectors of female patients with different levels of cogvalues above the median; black dots: MMSE or ADL values below the median). MMSE, mini-mreactance, T2, Hotelling's statistic; D,Mahalanobis distance. Themajor axis of specific tolerancethe upper pole); the minor axis (mainly related to phase angle) refers to variations of lean tiss

with donepezil. This lack of relationship matches with the results of arecent study (Droogsma et al., 2013) on the effect of another AchE-inhibitor (galantamine), showing that weight loss in patients with ADwas not attributable to long-term treatment.

The cognitive impairment (as assessed by MMSE) was associatedwith a lower lean tissuemass (lower phase) and theworsening of func-tional status (assessed by ADL) with higher FM% (longer specificvectors). The few studies on body composition that have been carriedout in patients with AD consistently show that patients are character-ized by a reduction of the lean mass, as assessed by DXA (Burns et al.,2010; Coin et al., 2012), or classic BIVA (Saragat et al., 2012). Some au-thors (Auyeung et al., 2008; Nourhashémi et al., 2002) found a similardecline in patients with other types of dementia or cognitive impair-ment, while others observed a reduction in fat-free mass only in sub-jects with severe dysfunction (Wirth et al., 2011). Abellan van Kanet al. (2013) found that gait speed and handgrip strength, but not mus-cle mass, were associated with cognitive impairment.

The FM% increase described in the present sample, especially in indi-viduals with worse functional status, suggests a condition of sarcopenicobesity. A similar condition in patients with AD was hypothesized by

nitive and functional impairment within the tolerance ellipses (white dots: MMSE or ADLental state examination; ADL, activities of daily living; Rsp, specific resistance, Xcsp, specificellipses (mainly related to vector length) refers to variations of FM% (higher values towardue mass (more mass toward the left side).

33R. Buffa et al. / Experimental Gerontology 58 (2014) 30–33

Saragat et al. (2012) on the basis of classic BIVA and anthropometry, buthas not been observed by other authors (Burns et al., 2010; Coin et al.,2012; Wirth et al., 2011). However, a recent review by Kohara (2014)showed an age-related association between sarcopenic obesity andphysical function decline. Furthermore, Levine and Crimmins (2012)found an association between sarcopenic obesity and low cognitivefunctioning.

In spite of the disease-related changes of body composition, BMI andMNA were similar in patients and in the reference and did not show adecrease with cognitive impairment. The lack of relationships betweenBMI (i.e., a nonspecific measure of body composition) and cognitive de-cline was previously observed by Burns et al. (2010). However, it couldbe possible that the association between cognitive status and BMI coulddepend on the level of cognitive deterioration, becoming detectableonly in the severe stage of dementia, as observed in other investigations(Coin et al., 2012; Wirth et al., 2011). As a matter of fact, the sample ofpatients under study is in the mild–moderate stage and characterizedby low levels of malnutrition. This could also be the cause of the lackof association between cognitive impairment and MNA, instead ob-served by other authors in patients with moderately severe to very se-vere stages of dementia, using the same specific BIVA (Camina Martínet al., 2014).

The cross-sectional design and small sample size are the main limi-tations of this study. In spite of that, our data indicate that patientswith mild–moderate stages of AD have lower lean tissue mass andhigher percent fatmass than the reference. This pattern, possibly relatedto a condition of sarcopenic obesity, is associated with functional de-cline. Moreover, this study shows the usefulness of specific BIVA forassessing body composition in patients with Alzheimer's disease. Thissimple technique could be applied to monitor programs on physical ac-tivity or nutritional interventions, enhancing BMI or MNA informationin the evaluation of nutritional status.

Acknowledgments

The authors acknowledge the patients who participated in thisstudy. Furthermore, the authors thank the organizers and participantsof the Healthy Aging Research Centre conference (HARC) “Nutritionand diet for age-related cognitive decline and dementia” (March, 6–7,2014; Łódź, Poland) for stimulating and inspiring discussions.

This researchwas financially supported by the University of Cagliari.EM acknowledges financial support from “Regione Autonoma dellaSardegna” (CRP-59903, 2012) through a research grant on fundings ofthe Project PO Sardegna FSE 2007–2013, L.R.7/2007 Promozione dellaricerca scientifica e dell'innovazione tecnologica in Sardegna.

References

Abellan van Kan, G., Cesari, M., Gillette-Guyonnet, S., Dupuy, C., Nourhashémi, F., Schott,A.M., Beauchet, O., Annweiler, C., Vellas, B., Rolland, Y., 2013. Sarcopenia and cognitiveimpairment in elderly women: results from the EPIDOS cohort. Age Ageing 42,196–202.

Auyeung, T.W., Kwok, T., Lee, J., Leung, P.C., Leung, J., Woo, J., 2008. Functional decline incognitive impairment—the relationship between physical and cognitive function.Neuroepidemiology 31, 167–173.

Buffa, R., Floris, G., Lodde, M., Cotza, M., Marini, E., 2010a. Nutritional status in the healthylongeval population from Sardinia (Italy). J. Nutr. Health Aging 14, 97–102.

Buffa, R., Mereu, R.M., Putzu, P.F., Floris, G., Marini, E., 2010b. Bioelectrical impedance vec-tor analysis detects low body cell mass and dehydration in patients with Alzheimer'sdisease. J. Nutr. Health Aging 14, 823–827.

Buffa, R., Saragat, B., Cabras, S., Rinaldi, A.C., Marini, E., 2013. Accuracy of specific BIVA forthe assessment of body composition in the United States population. PLoS One 8,e58533.

Burns, J.M., Johnson, D.K., Watts, A., Swerdlow, R.H., Brooks, W.M., 2010. Reduced leanmass in early Alzheimer disease and its association with brain atrophy. Arch. Neurol.67, 428–433.

Camina Martín, A., de Mateo Silleras, B., Barrera Ortega, S., Domínguez Rodríguez, L.,Redondo del Río, P., 2014. Specific bioelectrical impedance vector analysis (BIVA) ismore accurate than classic BIVA to detect changes in body composition and in nutri-tional status in institutionalised elderly with dementia. Exp. Gerontol. 57, 264–271.http://dx.doi.org/10.1016/j.exger.2014.06.016.

Coin, A., Veronese, N., De Rui, M., Mosele, M., Bolzetta, F., Girardi, A., Manzato, E., Sergi, G.,2012. Nutritional predictors of cognitive impairment severity in demented elderlypatients: the key role of BMI. J. Nutr. Health Aging 16, 553–556.

Droogsma, E., van Asselt, D.Z., van Steijn, J.H., Schuur, T., Huinink, E.J., 2013. Effect of long-term treatment with galantamine on weight of patients with Alzheimer's dementia.J. Nutr. Health Aging 17, 461–465.

Folstein, M.F., Folstein, S.E., McHugh, P.R., 1975. “Mini-mental state”. A practical methodfor grading the cognitive state of patients for the clinician. J. Psychiatr. Res. 12,189–198.

Gillette-Guyonnet, S., Nourhashemi, F., Andrieu, S., de Glisezinski, I., Ousset, P.J., Riviere, D.,Albarede, J.L., Vellas, B., 2000. Weight loss in Alzheimer disease. Am. J. Clin. Nutr. 71,637S–642S.

Guigoz, Y., Vellas, B., Garry, P.J., 1994. Mini Nutritional Assessment: a practical assessmenttool for grading the nutritional state of elderly patients. Facts Res. Gerontol. 2, 15–59.

Katz, S., Ford, A.B., Moskowitz, R.W., Jackson, B.A., Jaffe, M.W., 1963. Studies of illness inthe aged. The index of ADL: a standardized measure of biological and psychosocialfunction. JAMA 185, 914–919.

Kohara, K., 2014. Sarcopenic obesity in aging population: current status and future direc-tions for research. Endocrine 45, 15–25.

Lee, E.B., 2011. Obesity, leptin, and Alzheimer's disease. Ann. N. Y. Acad. Sci. 1243, 15–29.Levine, M.E., Crimmins, E.M., 2012. Sarcopenic obesity and cognitive functioning: the me-

diating roles of insulin resistance and inflammation? Curr. Gerontol. Geriatr. Res.2012, 826398.

Luchsinger, J.A., Gustafson, D.R., 2009. Adiposity and Alzheimer's disease. Curr. Opin. Clin.Nutr. Metab. Care 12, 15–21.

Marini, E., Buffa, R., Saragat, B., Coin, A., Toffanello, E.D., Berton, L., Manzato, E., Sergi, G.,2012. The potential of classic and specific bioelectrical impedance vector analysisfor the assessment of sarcopenia and sarcopenic obesity. Clin. Interv. Aging 7,585–591.

Marini, E., Sergi, G., Succa, V., Saragat, B., Sarti, S., Coin, A., Manzato, E., Buffa, R., 2013. Ef-ficacy of specific bioelectrical impedance vector analysis (BIVA) for assessing bodycomposition in the elderly. J. Nutr. Health Aging 17, 515–521.

Nourhashémi, F., Andrieu, S., Gillette-Guyonnet, S., Reynish, E., Albarède, J.L., Grandjean,H., Vellas, B., 2002. Is there a relationship between fat-free soft tissue mass and lowcognitive function? Results from a study of 7,105 women. J. Am. Geriatr. Soc. 50,1796–1801.

Piccoli, A., Rossi, B., Pillon, L., Bucciante, G., 1994. A newmethod formonitoring body fluidvariation by bioimpedance analysis: the RXc graph. Kidney Int. 46, 534–539.

Saragat, B., Buffa, R., Mereu, E., Succa, V., Cabras, S., Mereu, R.M., Viale, D., Putzu, P.F.,Marini, E., 2012. Nutritional and psycho-functional status in elderly patients withAlzheimer's disease. J. Nutr. Health Aging 16, 231–236.

Saragat, B., Buffa, R., Mereu, E., De Rui, M., Coin, A., Sergi, G., Marini, E., 2014. Specific bio-electrical impedance vector reference values for assessing body composition in theelderly. Exp. Gerontol. 50, 52–56.

Sheikh, J.L., Yesavage, J.A., 1986. Geriatric depression scale (GDS): recent evidence and de-velopment of a shorter version. In: Brink, T.L. (Ed.), Clinical Gerontology: A Guide toAssessment and Intervention. Hawthorne Press, New York.

Soto, M., Secher, M., Gillette-Guyonnet, S., Abellan van Kan, G., Andrieu, S., Nourhashemi,F., Rolland, Y., Vellas, B., 2012. Weight loss and rapid cognitive decline in community-dwelling patients with Alzheimer's disease. J. Alzheimers Dis. 28, 647–654.

Wirth, R., Smoliner, C., Sieber, C.C., Volkert, D., 2011. Cognitive function is associated withbody composition and nutritional risk of geriatric patients. J. Nutr. Health Aging 15,706–710.