Association of Drug Application and Hydration Status in Elderly Patients

nutrients

Article Association of Drug Application and Hydration Status in Elderly Patients

Laura Hoen 1, Daniel Pfeffer 1,2, Rico Zapf 1,2, Andrea Raabe 3, Janosch Hildebrand 1, Johannes Kraft 2 and Stefan Kalkhof 1,4,*

1 Institute for Bioanalysis, Coburg University of Applied Sciences and Arts, Friedrich-Streib-Str. 2, D-96450 Coburg, Germany; [email protected] (L.H.); [email protected] (D.P.); [email protected] (R.Z.); [email protected] (J.H.) 2 Division of Geriatrics, Klinikum Coburg GmbH, Ketschendorfer Str. 33, D-96450 Coburg, Germany; [email protected] 3 Division of Nephrology, Klinikum Coburg GmbH, Ketschendorfer Str. 33, D-96450 Coburg, Germany; [email protected] 4 Protein Biomarker Unit, Fraunhofer Institute for Cell Therapy and Immunology, Perlickstr. 1, D-04103 Leipzig, Germany * Correspondence: [email protected]

Abstract: Due to multifactorial reasons, such as decreased thirst and decreased total body water, elderly patients are vulnerable to dehydration. Mild cognitive impairment (MCI) or dementia increase the risk of dehydration and, in turn, dehydration decreases cognitive performance. The study aims to identify and assess differences in hydration status, taking into account patients’ drug treatment and diseases, using bioelectrical impedance vector analysis (BIVA), thereby revealing unfavorable aspects of prognosis. 447 geriatric patients (241 women, 206 men) including information on medication and bioelectrical impedance analysis (BIA) were investigated, which allowed studying Citation: Hoen, L.; Pfeffer, D.; Zapf, the association between 40 drugs and the hydration status. First, patients were divided into disease R.; Raabe, A.; Hildebrand, J.; Kraft, J.; groups. Renal disease and diuretic treatment were signiﬁcantly different in both sexes, whereas Kalkhof, S. Association of Drug cardiovascular patients differed exclusively for females. Next, drug enrichment was examined in Application and Hydration Status in either hyperhydrated or dehydrated patients. Simvastatin, candesartan, bisoprolol, amlodipine, Elderly Patients. Nutrients 2021, 13, olmesartan, furosemide, torasemide, allopurinol, mirtazapine, pantoprazole, cholecalciferol, and 1929. https://doi.org/10.3390/ resveratrol showed enrichment depending on hydration status. This study demonstrated that patients nu13061929 can be differentiated and stratiﬁed by BIVA, taking into account medication and disease associated

Academic Editor: Ina Bergheim with hydration status. Although patients diagnosed with MCI and therefore treated with resveratrol, BIVA still showed evaluated dehydration. This is unfavorable in terms of prognosis and requires

Received: 20 April 2021 special attention. Accepted: 31 May 2021 Published: 4 June 2021 Keywords: drug-nutrient interactions; nutrition; body composition; geriatrics; nutrition assessment; malnutrition; clinical nutrition; prevention Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affil- iations. 1. Introduction The demographic change leads more and more to an older population (age > 65) [1]. Due to multifactorial reasons, like reduced thirst, smaller fluid reserve and decreased total body water, this population is susceptible to dehydration and electrolyte abnormalities [2]. Copyright: © 2021 by the authors. In addition to physical changes and inadequate fluid intake, different age-associated med- Licensee MDPI, Basel, Switzerland. ication leads to a higher dehydration risk [3]. A variety of drug-induced mechanisms This article is an open access article like diarrhea, increase of urine volume, decrease of thirst sensation, increase of sweat distributed under the terms and production or decrease of appetite alter the hydration status [4]. Diuretics, e.g., increase conditions of the Creative Commons the urine volume, which is indicated to treat edema, but unmonitored use can result Attribution (CC BY) license (https:// in dehydration [3,5]. Furthermore, mild cognitive impairment (MCI) or dementia result creativecommons.org/licenses/by/ in a higher dehydration risk [2,6,7], and in return dehydration decreases the cognitive 4.0/).

Nutrients 2021, 13, 1929. https://doi.org/10.3390/nu13061929 https://www.mdpi.com/journal/nutrients Nutrients 2021, 13, 1929 2 of 13

performance [8,9]. Woiszel [10] recently showed that there is a trend for higher odds for impending dehydration when patients take procognitive medications and the negative effect of this becomes signiﬁcant when adjusting for dementia. The symptoms of dehydration have a huge range between simple dizziness, confusion to seizures and deaths [2]. Nevertheless, diagnosing dehydration is very challenging because diagnostic signs like capillary reﬁll time, abnormal skin turgor or respiratory pattern, which are widely used in children, cannot be applied to the elderly because of the aging process [2]. The bioelectrical impedance analysis (BIA) is thereby a relatively inexpensive, easy-to-use and non-invasive method for measuring the body composition [11]. By applying alternating electrical current the bioelectrical impedance analyzers determine the two components of bioelectrical impedance (Z) in the human body; the resistance (R) and reactance (Xc) [12]. The bioelectrical impedance vector analysis (BIVA) uses these individual components, normalized by height (H) of the subject, and represents them in the RXc graph (abscissa, R/H; ordinate, Xc/H) [13]. As reactance is directly related to the amount of soft tissue structures and resistance is inversely related to the intra- and extracellular water [14]; BIVA is a very promising tool for assessing a patient’s hydration and nutritional status [7,12,15]. Besides to the impedance measurement error the biological variability of subjects like age, sex, race, and body mass index (BMI) affects this approach [7,16]. This study aimed to show differences in patient’s hydration status considering their medicament treatment or diseases by BIVAand thereby reveal unfavorable aspects of prognosis.

2. Material & Methods 2.1. Study Design In 2017, 2018, and 2019, 471 outpatients treated in the Department of Geriatrics of Prof. Dr. Kraft (Regiomed Klinikum Coburg) were studied retrospectively. All patient data were included in this study in a blinded fashion and will be referred to as the Coburg cohort in the following. Inclusion criteria for the patients were age (between 65 and 98 years old (79.8 ± 7.0 (SD)) and a bioelectrical impedance analysis received at least once. The following data were collected at the date of the clinical investigation: BIA measurements, laboratory values (hematocrit, urea, sodium, creatinine and glucose), Clock-Drawing Test, Mini-Mental State Examination and medication. Preparation of the patients for BIVA was according to their medication and treatment plan. Because of the retrospectivity physical activity, dietary or any other nutritional aspects than medication, which may also affected the result of the BIVA analyses, were not included in this study.

2.2. Bioelectrical Impedance Vector Analyses BIA examinations were performed as part of routine inspections of geriatric patients who were not conditioned in the course of dieting, exercise, etc. Body composition was assessed with an InBody 770 multi-frequency (1, 5, 50, 250, 500, 1000 kHz), four-point bioelectrical impedance device (InBody 770, Cerritos, CA, USA). On this basis, the parameters phase angle, total body water (TBW) and extracellular water (ECW) were subsequently calculated. Impedance and reactance values were utilized to calculate Bioelectrical Impedance Vector Analyzes (BIVA) values as been described by Piccoli et al. [17]. For the analysis the resistance is needed, therefore the following formula is used: Z2 = R2 + Xc2; R, Z and Xc values of the whole-body were calculated as the sum of those readings for the right arm, right leg and trunk at 50 kHz measured with an electric current intensity of 80 µA[18]. BIVA analyses are done with the formulas provided by Piccoli [19]. To investigate the population distribution, the BIVA values of the patients were plotted against the reference populations in the BIVA Software [19]. As BIA is inﬂuenced by sex [16], males and females were compared separately. Sex-speciﬁc age and BMI matched population of the study of Roubenoff et al. [20] were used as reference. NutrientsNutrients 2021 ,2021 13, x, 13FOR, x FORPEER PEER REVIEW REVIEW 3 of 314 of 14

Nutrients 2021, 13, 1929 3 of 13

age ageand andBMI BMI matched matched population population of the of studythe study of Roubenoff of Roubenoff et al. et [20] al. [20]were were used used as refer- as refer- enceence. . TheThe Z-ScoreTheZ-Score Z-Score is generally is generally calculated calculated calculated by by dividing by dividing dividing each; each; each; resistance resistance resistance and and andreactance, reactance,reactance, with with the height,the height,the following height, following following the transformationthe transformationthe transformation by meanby meanby andmean and standard andstandard standard deviation deviation deviation of of a suitablea ofsuitable a suitable reference population.referencereference population. Based population. on Based the Based Coburg on the on Coburgthe patient’s Coburg patient’s data patient’s a data new data a referencenew a new reference reference population population population was was generated was consideringgeneratedgenerated considering sex, considering age and sex, body sex,age ageand mass andbody indexbody mass mass (BMI). index index (BMI). This (BMI). generated This This generated generated reference reference reference population pop- pop- was usedulationulation for was building was used used for the buildingfor Z-Score building the Ellipse Zthe-Score Z- toScore Ellipse classify Ellipse to the classify to 447 classify patientsthe 447the 447patients (see patients Section (see (seeSection 2.4 Section). 2.4).2.4). 2.3. BIA Confidence Intervals of Cardiovascular, Renal, Diuretic or Dementia Patients 2.3. 2.3BIA. BIAConfidence Confidence Intervals Intervals of Cardiovascular of Cardiovascular, Renal, Renal, Diuretic, Diuretic or Dementia or Dementia Patients Patients The patients were divided into subgroups: cardiovascular, renal, diuretic or dementia The Thepatients patients were were divided divided into into subgroups: subgroups: cardiovascular, cardiovascular, renal, renal, diuretic diuretic or demen- or demen- according to their medication. All groups were chosen because of their relation with the tia accordingtia according to their to their medication. medication. All groupsAll groups were were chosen chosen because because of their of their relation relation with with hydrationthe hydrationthe hydration status status andstatus and their andtheir increasedtheir increased increased occurrenceoccurrence occurrence (see (see (seeSection Section Section 3.1). 3.1 3 The.1).). The Thefollowing followingfollowing med- med- medica- tionsicationsications and and laboratory andlaboratory laboratory values values values were were usedwere used toused to divide divide to divide the the patients patientsthe patients into into the intothe corresponding correspondingthe corresponding groups: Cardiovasculargroups:groups: Cardiovascula Cardiovascula= bisoprolol,r = brisoprolol, = b nebivolol,isoprolol, nebivolol, n metoprolol,ebivolol, metoprolol, metoprolol, carvedilol, carvedilol, carvedilol, candesartan, candesartan, candesartan, ramipril, rami- rami- lisinopril,pril,pril, enalapril,lisinopril, lisinopril, atorvastatin,enalapril, enalapril, atorvastatin, ezetimibe/atorvastatin,atorvastatin, ezetimibe/ ezetimibe/atorvastatin,atorvastatin, simvastatin, simvastatin, simvastatin, amlodipine; amlodipine; amlodipine;Renal = al- lopurinolRenaRenal = alllopurinol = and allopurinol an elevatedand andan elevated an level elevated level of creatininelevel of creatinine of creatinine (>97 (>9µ7 (>9 mol/Lµmol/L7 µmol/L ( ♂ )/>80(♂)/>80 µmol/Lµ mol/Lµmol/L (♀)); (♀ )); Di- ureticDiuretiDiureti=c torasemide, = tcorasemide, = torasemide, furosemide, furosemide, furosemide, eplerenone,eplerenone, eplerenone, s spironolactone,pironolacton spironolactone, chlortalidone,e, chlortalidone, chlortalidone, hydrochlo- hydrochloroth-hydrochlo- iaziderothiaziderothiazide (HCT); (HCT); (HCT);Dementia Dementi Dementia= =memantine, ma emantine,= memantine, d donepezil,onepezil, donepezil, Ginkgo Ginkgo Ginkgo biloba biloba biloba extract, extract, extract, rivastigmine, rivastigmine, rivastigmine, striking Mini-Mental State Examination (MMSE) (Score ≤ 23) and Clock-Drawing Test strikingstriking Mini-Mental Mini-Mental State State Examination Examination (MMSE) (MMSE) (Score (Score ≤ 23)≤ and23) Clock and- Clock-DrawingDrawing Test Test (Score(Score ≥ 3). ≥ As 3). referenceAs reference groups, groups, all patients all patients were were applied applied, who, who were were not beennot been classified classified (Score ≥ 3). As reference groups, all patients were applied, who were not been classified for thefor specificthe specific disease. disease. Using Using this thisclassification classification system system 89% 89% of all of patientsall patients (397 (397 of 447) of 447) for the specific disease. Using this classification system 89% of all patients (397 of 447) were werewere included included in at in least at least one oneof the of groups.the groups. However, However, group group assignments assignments were were not notdis- dis- includedjunctive,junctive, e.g. in at ,e.g. an least ,individual an oneindividual of which the which groups. was was included However, included in the in group cardiovascularthe cardiovascular assignments group group were can notcanalso disjunctive,also ad- ad- e.g.,ditionally anditionally individual be listed be listed in which the in renalthe was renal group included group (Supplementary ( inSupplementary the cardiovascular Figure Figure S1). SThe group1). Thegroups can groups alsowere were additionallysplit split beup listed byup sex. by in sex.For the eachFor renal each patient group patient subgroup (Supplementary subgroup the mean,the mean, standard Figure standard S1). deviation Thedeviation groups and andlinear were linear correlation split correlation up by sex. Forcoefficient eachcoefficient patient between between subgroup the thecomponents components the mean, resistance/height standard resistance/height deviation and andreactance/height and reactance/height linear correlation were were calcu- coefficientcalcu- betweenlated.lated. Differences the Differences components between between the resistance/height meanthe mean impedance impedance vectors and vectors reactance/height in the in describedthe described groups were groups calculated.were were as- as- Dif- ferencessessedsessed with between with Hotelling’s Hotelling’s the meanT2 test T2 impedancetestand andgraphically graphically vectors with with 95% in the95% probability described probability confidence groups confidence ellipses, were ellipses, assessed withwhichwhich Hotelling’s correspond correspond T to2 test a tostatistically anda statistically graphically significant significant with difference 95% difference probability between between mean confidence mean vector vector displace- ellipses, displace- which 2 correspondmentsments on the on to RXcthe a statistically RXc graph graph (p < (0p significant.05). < 0.05). A significant A significant difference Hot elling’sHot betweenelling’s T2 test meanT testis equivalent vector is equivalent displacements to a tosig- a sig- on thenificant RXcnificant difference graph difference (p in< 0.05).R, in Xc R, or Xc A both or significant both parameters. parameters. Hotelling’s T2 test is equivalent to a significant difference in R, Xc or both parameters. 2.4. 2.4.Classification Classification of Patients of Patients according according to the to Hydration the Hydration Status Status and andEnrichment Enrichment of Medicaments of Medicaments 2.4. ClassificationTo haveTo have a look ofa look Patientson individual on individual According medication medication to the Hydrationenriched enriched in Status different in different and hydration Enrichment hydration statuses, of statuses, Medicaments a a Z-ScoreZ-Score model model based based on Normanon Norman et al. et [21] al. [21]was was outlined outlined and andthe patientsthe patients were were divided divided To have a look on individual medication enriched in different hydration statuses, a intointo three three groups, groups, based based on the on patient’sthe patient’s position position in this in thisZ-Score Z-Score model model (Figure (Figure 1) A: 1 )(i) A: (i) Z-Scoredehydrateddehydrated model patients basedpatients located on located Norman in sector in sector etI and al. I and[M;21 ](ii) M; was normally(ii) outlined normally hydrate andhydrated the locatedd patients located in sector in were sector A, divided A, intoB, CB, threeand C andD; groups, (iii) D; (iii)hyperhydrated basedhyperhydrated on thelocated patient’slocated in sector in positionsector O and O andK. in All thisK. patientsAll Z-Score patients appearing model appearing (Figurein the in the1) A: (i)other dehydratedother sectors sectors (E, patients G, (E, F, G, H, F, J, locatedH, N J,L, N P) L, were in P) sectorwere excluded excluded I and from M; from the (ii) followingthe normally following analysis. hydrated analysis. The located Theabsolute absolute in sector A,frequencies B,frequencies C and of D; each (iii)of each prescribed hyperhydrated prescribed drug drug and located andthe numberthe in number sector of patients of O patients and in K. each in All each group patients group were were appearing com- com- in theputed. otherputed. Based sectors Based on these on (E, these G,values F,values H, the J, relativethe N relative L, P) frequencies were frequencies excluded within within these from these groups the groups following were were calculated. calculated. analysis. The absoluteThe Thefold fold frequencieschange change (FC) (FC) was of eachwasdefined defined prescribed as the as ratiothe drug ratio of relative andof relative the frequency number frequency in of one in patients oneof the of extremethe in eachextreme group weregroupsgroups computed. (hyperhydrated (hyperhydrated Based or on dehydrated) or these dehydrated) values and the andthe relative medianthe medianfrequencies of the of relativethe relative within frequencies frequencies these of groups the of the were calculated.remainingremaining other The other fold two changetwogroups groups (FC)(normal (normal was hydrated defined hydrated plus as the plus the ratio remainingthe ofremaining relative extreme extreme frequency group). group). in En- one En- of the extremerichedriched drugs groups drugs were (hyperhydrated were defined defined by anby or occurrencean dehydrated) occurrence in the in and thepopulation thepopulation median study study of ≥15 the ≥15for relative statisticalfor statistical frequencies relevancerelevance and anda binary a binary logarithm logarithm of log of2 (logFC)2( FC)≥ │ ≥0.5 ││0.5 as│ minimal as minimal difference difference between between the the of the remaining other two groups (normal hydrated plus the remaining extreme group). groupsgroups and anda significant a significant difference difference in the in RXcthe RXc graph graph (Hotelling’s (Hotelling’s T2 test) T2 test) in minimum in minimum one one Enriched drugs were defined by an occurrence in the population study ≥ 15 for statistical gender.gender. The Thesignificance significance was was checked checked by Fisher by Fisher’s exact’s exact test testin which in which each each extreme extreme group group ≥ relevancewaswas compared compared and ato binarythe to normalthe logarithmnormal hydra hydrated of group.ted log group.2(FC) Using Using |0.5|asthe Benjaminithe Benjamini minimal & Hochberg & difference Hochberg’s method’s between method the 2 groupsp-valuesp-values and were a were significantadjusted adjusted for difference multiplefor multiple testing. in testing. the RXc graph (Hotelling’s T test) in minimum one gender. The significance was checked by Fisher’s exact test in which each extreme group was compared to the normal hydrated group. Using the Benjamini & Hochberg’s method p-values were adjusted for multiple testing.

Nutrients 2021, 13, 1929 4 of 13 Nutrients 2021, 13, x FOR PEER REVIEW 4 of 14

FigureFigure 1. 1.Subdivision Subdivision Model Model of of Z-Scores Z-Scores based based on on 50%, 50% 75%, 75% and and 95% 95% tolerance tolerance ellipses ellipses A: A: Annotation Annotation ofof the the 50%, 50%, 75% 75% and and 95% 95% tolerance tolerance ellipses ellipses based based on on Norman Norman et et al. al. [21 [21]]. Sectors. Sectors M M and and I I (dehydrated), (dehydrated), KK andand OO (hydrated), (hydrated), andand A, A, B, B, C, C, D D (control) (control) were were deﬁned defined to to classify classify the the patients patients according according to to the the hydration status. hydration status.

3.3. ResultsResults 3.1.3.1. Coburg Coburg Patients Patients Are Are Leaner Leaner and and Hyperhydrated Hyperhydrated in in Comparison Comparison to to Reference Reference Population Population BIABIA analyses analyses of of 471 471 outpatients outpatients were were performed performed in 2017in 2017 to 2019to 2019 in the in the geriatric geriatric clinic clinic in Coburg.in Coburg. Twenty-four Twenty- offour these of contained these contained incomplete incomplete patientinformation patient information and were therefore and were excluded,therefore whileexcluded, the remaining while the remaining 447 (241 women, 447 (241 206 women, men) were 206 men) used were for further used f analysis.or further Theanalysis. average The age average of the 447 age outpatients, of the 447 includedoutpatients, in thisincluded study (calledin this Coburgstudy (called cohort), Coburg was 79.8cohort),± 7.0 was years, 79.8 and ± 7.0 54% years, (241) and of 54% them (241) were of them female. were Main female. diagnoses Main diagnoses included included cardio- vascularcardiovascular disease disease (61%, 277), (61%, dementia 277), dementia (53%, 237),(53%, renal 237), diseaserenal disea (33%,se (33%, 149), and149), diuretic and diu- treatmentretic treatment (33%, 149).(33%, Initially, 149). Initially, differences differences between between the study the of study Roubenoff of Roubenoff et al. [20 et], whichal. [20], includedwhich included 455 elderly 455 patients elderly (161patients men (161 average men age average of 78.2 age± of4.3 78. years2 ± 4and.3 years 294women and 294 ± 78.4women4.5 78. years,4 ± 99%4.5 years, of whom 99% were of whom White) were being White) acquired being in the acquired context in of the Framinghamcontext of the HeartFramingham Study (USA), Heart were Study discovered (USA), were (Figure discovered2a,b). The (Figure used reference2a,b). The populations, used reference which pop- buildulations, the tolerance which build ellipses, the tolerance were chosen ellipses, because were they chosen matched because most they in terms matched of age, most BMI in andterms sex. of Plotting age, BMI the and BIA sex. data Plotting of the the Coburg BIA data patients of the in Coburg the normalized patients tolerance in the normalized ellipses published by Roubenoff et al. [20] revealed a shift to lower Z(Xc/H) (male: −0.72 ± 0.43, tolerance ellipses published by Roubenoff et al. [20] revealed a shift to lower Z(Xc/H) female: −0.88 ± 0.43) and higher (R/H) (male: 0.31 ± 1.07, female: 0.14 ± 0.93) values. (male: −0.72 ± 0.43, female: −0.88 ± 0.43) and higher (R/H) (male: 0.31 ± 1.07, female: 0.14 ± According to Piccoli et al. [17] this means that the Coburg patients can be classified to be 0.93) values. According to Piccoli et al. [17] this means that the Coburg patients can be more hydrated, cachectic and lean because they were laying predominantly in the lower classified to be more hydrated, cachectic and lean because they were laying predomi- right part. To take these shifts between the Coburg and the Roubenoff et al. [20] cohort nantly in the lower right part. To take these shifts between the Coburg and the Roubenoff into account and to investigate differences within the Coburg group, a new reference et al. [20] cohort into account and to investigate differences within the Coburg group, a population out of the Coburg people was build and the required reference values were new reference population out of the Coburg people was build and the required reference calculated. The BMI for male and female population is between 25 and 28 kg/m2 and values were calculated. The BMI for male and female population is between 25 and 28 an age between 65 and 98. Male population defined by n = 130 (number), mR = 290.7 kg/m2 and an age between 65 and 98. Male population defined by n = 130 (number), mR = (R/H mean), sR = 35.8 (R/H standard deviation), mXc = 21.4 (Xc/H mean), sXc = 4.4 (Xc/H standard290.7 (R/H deviation), mean), rsR = = 0.5740 35.8 (Correlation(R/H standard (R/H, deviation), Xc/H)). ThemXc female= 21.4 (Xc/H population mean), is definedsXc = 4.4 (Xc/H standard deviation), r = 0.5740 (Correlation (R/H, Xc/H)). The female population is by n = 115, mR = 360.6, sR = 42.6, mXc = 25.1, sXc = 4.9, r = 0.6061. With these new build referencedefined populations,by n = 115, m theR = 447 360.6, patients sR = 4 were2.6, m locatedXc = 25.1, in the sXc Z-Score= 4.9, r Ellipses= 0.6061. and With as expected,these new nearlybuild 50%reference (225 of populations, 447) were lying the 447 in the patients middle were in the located sectors in A, the B, Z C-Score and D Ellipses (Figure 2andc,d). as expected, nearly 50% (225 of 447) were lying in the middle in the sectors A, B, C and D (Figure 2c,d).

Nutrients 2021, 13, 1929 5 of 13 Nutrients 2021, 13, x FOR PEER REVIEW 5 of 14

(a) (b)

Nutrients 2021, 13, x FOR PEER REVIEWNutrients 2021, 13, x FOR PEER REVIEW 3 of 14 3 of 14

age and BMI matched populationage of and the BMIstudy matched of Roubenoff population et al. [20]of the were study used of asRoubenoff refer- et al. [20] were used as reference. ence. The Z-Score is generally calculatedThe byZ- Scoredividing is generally each; resistance calculated and by reactance, dividing witheach; resistance and reactance, with the height, following the transformationthe height, by following mean and the standard transformation deviation by ofmean a suitable and standard deviation of a suitable reference population. Based on thereference Coburg p opulation.patient’s data Based a new on thereference Coburg population patient’s data was a new reference population was generated considering sex, age andgenerated body mass considering index (BMI). sex, age This and generated body mass reference index pop-(BMI). This generated reference population was used for building theulation Z-Score was Ellipse used forto classifybuilding the the 447 Z- Scorepatients Ellipse (see( cSectionto) classify the 447 patients (see Section (d) 2.4). 2.4). Figure 2 2.. (a(a––dd):): Distribution Distribution of ofthe the Coburg Coburg patients patients in a inBioelectrical a Bioelectrical Impedance Impedance Vector Vector Analyses Analyses (male (male (a,c), female (a,c), female (b,d)) plotted against reported tolerance intervals of the reference population reported by Roubenoff et al. (a + b) and against (b,d)) plotted against reported tolerance intervals of the reference population reported by Roubenoff et al. (a + b) and 2.3. BIA Confidence Intervals of Cardiovascular2.3. BIA Confidence, Renal, IntervalsDiuretictolerance ofor CardiovascularDementia ellipses Patients calculated, Renal , basedDiuretic on or all Dementia 447 patients Patients being included in the Coburg’s cohort (c + d). against tolerance ellipses calculated based on all 447 patients being included in the Coburg’s cohort (c + d). The patients were divided into subgroups:The patients cardiovascular, were divided intorenal, subgroups: diuretic or cardiovascular, demen- renal, diuretic or dementia according to their medication.tia Allaccording groups towere their chosen medication. because All of groupstheir relation were3.2. chosen with Differences Differences because in of the their Group relation of Cardiovascular Cardiovascular, with , Dementia,Dementia, Renal and Diuretic PatientsPatients the hydration status and their increasedthe hydration occurrence status (seeand Sectiontheir increased 3.1). The occurrence following med-(see Section 3.1). The following med- To investigate how the BIVA values were altered due to the presence of diseases, ications and laboratory values wereications used and to laboratorydivide the valuespatients were into used the correspondingto divide theTo patients investigate into the how corresponding the BIVA values were altered due to the presence of diseases, pa- groups: Cardiovascular = bisoprolol,groups: nebivolol, Cardiovascula metoprolol,r = bisoprolol, carvedilol, nebivolol, candesartan, metoprolol,patientstients rami- were c werearvedilol, divided divided candesartan, into into four four groups rami- groups (cardiovascular, (cardiovascular, dementia, dementia, diuretics diuretics and renal). and renal). There pril, lisinopril, enalapril, atorvastatin,pril, lisinopril, ezetimibe/ enalapril,atorvastatin, atorvastatin, simvastatin, ezetimibe/ amlodipine;Thereawastorvastatin, a significantwas a ssignificantimvastatin, variation variationamlodipine; in the RXcin the graph RXc forgraph female for female and male and between male between the renal the disease renal Renal = allopurinol and an elevatedRena levell = allopurinol of creatinine and (>9 an7 elevatedµmol/L (level♂)/>80 of creatinineµmol/Ldisease(p (♀ (>9));< 7 0.0001; (µmol/Lp ♀ < p 0.0001;(♂)/>80= 0.0274) p µmol/L ♂ = and0 .0274)(♀ their)); and control their groupcontrol and group between and between patients patients under diuretic under Diuretic = torasemide, furosemide,Diureti eplerenone,c = torasemide, spironolacton furosemide,e, chlortalidone, eplerenone, h sydrochlo-pironolactone, chlortalidone,♀ hydrochlo-♂ diuretictreatment treatment (p < 0.0001; (p < p0.0001;= 0.0106) p = 0 and.0106) their and control their control group group (Figure (Figure3a,b). Comparing 3a,b). Com- rothiazide (HCT); Dementia = memantine,rothiazide d (HCT);onepezil, Dementi Ginkgoa = bilobamemantine, extract, d onepezil,rivastigmine, Ginkgo biloba extract, rivastigmine, paringthe patients the patients with cardiovascular with cardiovascular disease disease and those and withoutthose without cardiovascular cardiovasc disease,ular disease, there striking Mini-Mental State Examinationstriking Mini (MMSE)-Mental (Score State ≤ Examination23) and Clock (MMSE)-Drawing (Score Test ≤ 23) and Clock-Drawing Test therewas a was significant a significant difference difference in the in female the female group group (p < 0.0001), (p < 0.0001), but not but in not the malein the group male (Score ≥ 3). As reference groups,(Score all patients ≥ 3). As were reference applied groups,, who wereall patients not been were classified groupapplied (,p who = 0.5452) were not(Figure been 3classifiedc). The dementia patients could be significantly discriminated for the specific disease. Using thisfor theclassification specific disease. system Using 89% of this all classificationpatients (397 systemof(p 447)= 0.5452) 89% of (Figure all patients3c). The(397dementia of 447) patients could be significantly discriminated in the in the male group (p = 0.0074) (Figure 3d), but not in the female group (p = 0.6145). Because were included in at least one ofwere the groups. included However, in at least group one ofassignments the groups. were However, notmale dis- group group assignments (p = 0.0074) were (Figure not dis-3d), but not in the female group ( p = 0.6145). Because junctive, e.g., an individual whichjunctive, was included e.g., an individualin the cardiovascular which was group included can inalsoa the clear ad- cardiovascular differentiation group between can also thead- subgroups could be seen, the enrichment of drugs in ditionally be listed in the renal groupditionally (Supplementary be listed in the Figure renal S group1). The ( Supplementarygroups wereextreme split Figure groups S1). The (hyperhydrated groups were split and dehydrated) was subsequently investigated. up by sex. For each patient subgroupup by the sex. mean, For each standard patient deviation subgroup and the linear mean, correlation standard deviation and linear correlation coefficient between the componentscoefficient resistance/height between the and components reactance/height resistance/height were calcu- and reactance/height were calculated. Differences between the meanlated. impedance Differences vectors between in tthehe meandescribed impedance groups vectorswere as- in the described groups were assessed with Hotelling’s T2 test andsessed graphically with Hotelling’s with 95% T 2probability test and graphically confidence with ellipses, 95% probability confidence ellipses, which correspond to a statisticallywhich significant correspond difference to a statistically between meansignificant vector difference displace- between mean vector displacements on the RXc graph (p < 0.05).ments A significant on the RXc Hot graphelling’s (p < T02.05). test Ais significantequivalent Hotto aelling’s sig- T2 test is equivalent to a significant difference in R, Xc or bothnificant parameters. difference in R, Xc or both parameters.

2.4. Classification of Patients according2.4. Classification to the Hydration of Patients Status andaccording Enrichment to the Hydrationof Medicaments Status and Enrichment of Medicaments To have a look on individual medicationTo have a enrichedlook on individualin different medication hydration enrichedstatuses, ain different hydration statuses, a Z-Score model based on NormanZ- Scoreet al. [21]model was based outlined on Norman and the etpatients al. [21] were was outlineddivided and the patients were divided into three groups, based on the intopatient’s three positiongroups, inbased this onZ- Scorethe patient’s model (positionFigure 1 )in A: this (i) Z-Score model (Figure 1) A: (i) dehydrated patients located in sectordehydrated I and M;patients (ii) normally located hydratein sectord Ilocated and M; in (ii) sector normally A, hydrated located in sector A, B, C and D; (iii) hyperhydratedB, located C and in D; sector (iii) hyperhydrated O and K. All patients located appearingin sector O in and the K. All patients appearing in the other sectors (E, G, F, H, J, N L, P)other were sectors excluded (E, G,from F, H, the J, following N L, P) were analysis. excluded The from absolute the following analysis. The absolute frequencies of each prescribed drugfrequencies and the of number each prescribed of patients drug in each and group the number were com- of patients in each group were computed. Based on these values theputed. relative Based frequencies on these within values these the relative groups frequencieswere calculated. within these groups were calculated. The fold change (FC) was definedThe as fold the changeratio of (FC)relative was frequency defined as in the one ratio of the of relativeextreme frequency in one of the extreme groups (hyperhydrated or dehydrated)groups (hyperhydratedand the median orof dehydrated)the relative frequencies and the median of the of the relative frequencies of the remaining other two groups (normalremaining hydrated other plustwo thegroups remaining (normal extreme hydrated group). plus theEn- remaining extreme group). En- riched drugs were defined by anriched occurrence drugs werein the defined population by an study occurrence ≥15 for in statistical the population study ≥15 for statistical relevance and a binary logarithmrelevance of log2(FC) and ≥ a│ binary0.5│ aslogarithm minimal of difference log2(FC) ≥between │0.5│ the as minimal difference between the groups and a significant differencegroups in the and RXc a significantgraph (Hotelling’s difference T2 intest) the in RXc minimum graph (Hotelling’s one T2 test) in minimum one gender. The significance was checkedgender. by The Fisher significance’s exact test was in checked which each by Fisher extreme’s exact group test in which each extreme group was compared to the normal hydrawasted compared group. Using to the the normal Benjamini hydra &ted Hochberg group. Using’s method the Benjamini & Hochberg’s method p-values were adjusted for multiplep-values testing. were adjusted for multiple testing.

Nutrients 2021, 13, 1929 6 of 13

Nutrients 2021, 13, x FOR PEER REVIEWa clear differentiation between the subgroups could be seen, the enrichment of drugs6 of in14 extreme groups (hyperhydrated and dehydrated) was subsequently investigated.

(a) (b)

Figure 3. ConfidenceConﬁdence Ellipses Ellipses in in the the RXc RXc graphs graphs for forthe thedifferent different subgroups subgroups and andeach eachgender: gender: renal renal(a), diuretic (a), diuretic (b), cardi- (b), ovascular (c) and dementia (d), * = p < 0.01, ** = p < 0.005, *** = p < 0.005. cardiovascular (c) and dementia (d), * = p < 0.01, ** = p < 0.005, *** = p < 0.005.

3.3. Enrichment of Medication according According to to Patient’s Patient’s Hydration Hydration Status Status However, duedue to to the the obvious obvious difference difference between between the the Coburg Coburg patients patients and theand provided the pro- videdreference, reference, a generation a generation of a reference of a reference population population out of the out Coburg of the patientsCoburg topatients improve to im- the provecategorization the categorization of the patients of the was patients necessary. was necessary. For the Coburg For the cohort, Coburg the cohort, reference the values refer- wereence values mentioned were in mentioned Section 3.2 in. TherebySection 3 the.2. Thereby patients couldthe patients be divided could into be divided dehydrated, into dehydrated,normal hydrated normal and hydrated hyperhydrated and hyperhydrated according to according their position to their in theposition vector in graph. the vector For graph.each group, For each the group, baseline the characteristics baseline characteristics and the main and impedance the main impedance values were values shown were in Tableshown1. in The Table ratio 1. betweenThe ratio male between and male female and was female nearly was equal nearly in allequal of thein all three of the groups. three Thegroups. hyperhydrated The hyperhydrated group was group smaller was than smaller the dehydrated than the dehydrated group. The group. biggest The group biggest was group was the normal hydrated group, in the defined sectors (A, B, C and D) and should include ca. 50% of all patients (here 225 out of 447, Figure 2c,d). Dehydrated patients had a lower BMI (24.1 ± 3.6) in comparison to the hyperhydrated patients (28.6 ± 4.8) (p = 0.0001), whereas the Sodium level was not significantly different (p = 0.1315). Dehydrated

Nutrients 2021, 13, 1929 7 of 13

the normal hydrated group, in the defined sectors (A, B, C and D) and should include ca. 50% of all patients (here 225 out of 447, Figure2c,d). Dehydrated patients had a lower BMI (24.1 ± 3.6) in comparison to the hyperhydrated patients (28.6 ± 4.8) (p = 0.0001), whereas the Sodium level was not significantly different (p = 0.1315). Dehydrated and hyperhydrated patients were slightly younger than the normal hydrated group (p < 0.05). As expected dehydrated patients had higher impedance, reactance and resistance values than the normal hydrated group. All bioelectrical impedance values were significantly different between these groups (p < 0.005). It was striking that the creatinine level was lower in the dehydrated group in contrast to the normal hydrated (p = 0.002) and hyperhydrated group (p = 0.0002).

Table 1. Baseline Characteristics and impedance values of the patients classiﬁed by hydration status (see also Figure1), average ± standard deviation, * = p ≤ 0.05, ** = p ≤ 0.005, *** = p ≤ 0.0001.

Characteristics Dehydrated Normal-Hydrated Hyperhydrated N 46 225 27 Male/Female 18/28 109/116 13/14 Height [cm] 165.4 ± 9.5 167.7 ± 9.3 173.6 ± 8.9 ** Age [yr] 78.3 ± 8.5 * 80.6 ± 6.4 77.9 ± 6.6 * Weight [kg] 66.1 ± 12.8 ** 73.1 ± 11.8 86.1 ± 15.7 *** BMI [kg/m2] 24.1 ± 3.6 ** 26.0 ± 3.4 28.6 ± 4.8 ** Sodium [mmol/L] 141.7 ± 2.6 141.0 ± 2.9 141.6 ± 2.8 Hematocrit [L/L] 0.417 ± 0.036 ** 0.398 ± 0.038 0.382 ± 0.050 * Urea [mmol/L] 9.1 ± 8.2 7.7 ± 3.5 7.6 ± 0.9 Creatinine [µmol/L] 82.5 ± 28.8 ** 99.2 ± 33.9 114.9 ± 42.1 * Glucose [mmol/L] 5.9 ± 1.4 * 6.6 ± 2.2 6.3 ± 1.4 Total Number of Medication 5.2 ± 3.6 ** 7.6 ± 4.3 9.3 ± 4.0 PhA [◦] 4.5 ± 0.7 ** 4.1 ± 0.4 3.8 ± 0.8 ** Z [Ohm] 647.8 ± 64.1 *** 543.1 ± 57.5 458.1 ± 57.4 *** Xc [Ohm] 49.8 ± 5.2 *** 39.1 ± 4.7 29.9 ± 5.8 *** R [Ohm] 645.8 ± 64.4 *** 541.7 ± 57.4 457.0 ± 57.5 *** R/H [Ohm/m] 392.3 ± 48.9 *** 324.6 ± 43.4 264.5 ± 39.5 *** Xc/H [Ohm/m] 30.2 ± 3.4 *** 23.4 ± 3.1 17.3 ± 3.6 ***

Using the different hydration subgroups, the enrichment of specific medication was investigated. An enrichment of different drugs in both extreme groups of dehydrated and hyperhydrated patients could be seen with a higher enrichment in the hyperhydrated group (Table2). According to their indication, drugs were classified into the groups (i) cardiac, (ii) diuretic, (iii) endocrinologic, (iv) mild cognitive impairment, (v) psychotic/neurological, (vi) gastrointestinal, and/or (vii) nutritional supplements, to get an overview of the enriched indications. In total of the 40 drugs given to at least 15 patients, 12 drugs were found to be significantly (log2(FC) > 0.5, p < 0.05 in minimum one gender) more prescribed in hyperhydrated patients, whereas five were more consumed by dehydrated patients compared to normal hydrated patients (Table2). Especially cardiac associated medication like candesartan and amlodipine, was enriched in the hyperhydrated group. This underlines the results showed before (see Section 3.2). There was one gastrointestinal medication, pantoprazole, enriched. Allopurinol, a renal medication, was enriched in the hyperhydrated group. In the hyperhydrated group also two diuretics were enriched, torasemide and furosemide. Furosemide with a 12-times higher enrichment. Furthermore, Olmesartan and amlodipine/olmesartan were found to be enriched. In both extreme groups, cardiac associated medication was the most common representative (54% of the enriched medication of each group). In the dehydrated group mirtazapine, an antidepressant, and resveratrol, a nutritional supplement used for MCI treatment, could be found more often. The RXc-graph of Resveratrol (Figure4) showed for both genders lower Xc values and minimal lower R values in the control group, which is not receiving resveratrol. NutrientsNutrients 2021 2021, 13,, 13x FOR, x FOR PEER PEER REVIEW REVIEW 7 of7 of14 14

andand hyperhydrated hyperhydrated patients patients were were slightly slightly younger younger than than the the normal normal hydrated hydrated group group (p (

TableTable 1. Baselin1. Baseline Characteristicse Characteristics and and impedance impedance values values of theof the patients patients classified classified by byhydration hydration sta- sta- tustus (see (see also also Figure Figure 1), 1averag), average ±e s ±tandard standard deviation, deviation, * = *p = ≤ p 0 ≤.05, 0.05, ** =** p = ≤ p 0 ≤.005, 0.005, *** ***= p = ≤ p 0 ≤.0001, 0.0001, Bold Bold highlightshighlights significant significant findings. findings.

CharacteristicsCharacteristics DehydratedDehydrated NormalNormal-Hydrated-Hydrated HyperhydratedHyperhydrated N N 46 46 225225 27 27 Male/FemaleMale/Female 18/2818/28 109/116109/116 13/1413/14 HeightHeight [cm] [cm] 165.165.4 ±4 9 ±.5 9 .5 167.167.7 ±7 9 ±.3 9 .3 173.173.6 ±6 8 ±.9 8 **.9 ** AgeAge [yr] [yr] 78.78.3 ±3 8 ±.5 8 *.5 * 80.80.6 ±6 6 ±.4 6 .4 77.77.9 ±9 6 ±.6 6 *.6 * WeightWeight [kg] [kg] 66.66.1 ±1 1 ±2.8 12.8 ** ** 73.73.1 ±1 1 ±1.8 11.8 86.86.1 ±1 1 ±5.7 15.7 *** *** BMIBMI [kg/m [kg/m2] 2] 24.24.1 ±1 3 ±.6 3 **.6 ** 26.26.0 ±0 3 ±.4 3 .4 28.28.6 ±6 4 ±.8 4 **.8 ** SodiumSodium [mmol/L] [mmol/L] 141.141.7 ±7 2 ±.6 2 .6 141.141.0 ±0 2 ±.9 2 .9 141.141.6 ±6 2 ±.8 2 .8 HematocritHematocrit [L/L] [L/L] 0.410.417 ±7 0 ±.036 0.036 ** ** 0.390.398 ±8 0 ±.038 0.038 0.380.382 ±2 0 ±.050 0.050 * * UreaUrea [mmol/L] [mmol/L] 9.19. ±1 8 ±.2 8 .2 7.77. ±7 3 ±.5 3 .5 7.67. ±6 0 ±.9 0 .9 CreatinineCreatinine [µmol/L] [µmol/L] 82.82.5 ±5 2 ±8.8 28.8 ** ** 99.99.2 ±2 3 ±3.9 33.9 114.114.9 ±9 4 ±2.1 42.1 * * GlucoseGlucose [mmol/L] [mmol/L] 5.95. ±9 1 ±.4 1 *.4 * 6.66. ±6 2 ±.2 2 .2 6.36. ±3 1 ±.4 1 .4 TotalTotal Number Number of ofMedication Medication 5.25. ±2 3 ±.6 3 **.6 ** 7.67. ±6 4 ±.3 4 .3 9.39. ±3 4 ±.0 4 .0 PhAPhA [°] [ °] 4.54. ±5 0 ±.7 0 **.7 ** 4.14. ±1 0 ±.4 0 .4 3.83. ±8 0 ±.8 0 **.8 ** Z [Ohm]Z [Ohm] 647.647.8 ±8 6 ±4.1 64.1 *** *** 543.543.1 ±1 5 ±7.5 57.5 458.458.1 ±1 5 ±7.4 57.4 *** *** XcXc [Ohm] [Ohm] 49.49.8 ±8 5 ±.2 5 ***.2 *** 39.39.1 ±1 4 ±.7 4 .7 29.29.9 ±9 5 ±.8 5 ***.8 *** Nutrients 2021, 13, x FOR PEER REVIEW 8 of 14 R [Ohm]R [Ohm] 645.645.8 ±8 6 ±4.4 64.4 *** *** 541.541.7 ±7 5 ±7.4 57.4 457.457.0 ±0 5 ±7.5 57.5 *** *** R/HR/H [Ohm/m] [Ohm/m] 392.392.3 ±3 4 ±8.9 48.9 *** *** 324.324.6 ±6 4 ±3.4 43.4 264.264.5 ±5 3 ±9.5 39.5 *** *** Xc/HXc/H [Ohm/m] [Ohm/m] 30.30.2 ±2 3 ±.4 3 ***.4 *** 23.23.4 ±4 3 ±.1 3 .1 17.17.3 ±3 3 ±.6 3 ***.6 *** 5 Mirtazapine 35 −0.97 0.88 0.0181 * 0.0384 * 0.1305 0.2061 6 Pantoprazole UsingUsing85 the the different different1.07 hydration hydration subgroups,−1 subgroups,.32 0.0014the the enrichment enrichment ** 0.0084 of ofspecific** specific0.5671 medication medication 0.6301 was was Nutrients7 2021, 13Cholecalciferol, 1929 investigated. investigated.220 An An enrichment enrichment0.52 of ofdifferent different−0.62 drugs drugs in0.0163 inboth both *extreme extreme0.0384 groups groups* 0.0001 of ofdehydrated dehydrated *** 0.0015 and 8and of** 13 Class shows the classificationhyperhydratedhyperhydrated of the compound patients patients (1 = couldCardiac, could be 2 be= seen Diuretic, seen with with 3 a= Endocrinology, higher a higher enrichment enrichment 4 = Dementia/Mild in in the the hyperhydrated hyperhydrated Cognitive Impairment, 5= psychotic/neurological,groupgroup (Tab (Table 2).le 6 2). = gastrointestinal, 7 = nutritional supplements), T = total amount, log2(FC) = binary logarithm, Fisher’s exact test: * = p < 0.05, ** = p < 0.005; Hotelling’s T test: p♀/p♂ with and without adjustment for TableTablemultiple 2. Enriched2. Enriched testing medication using medication BenjaminiTable in thein 2. theEnriched dehydrated & dehydratedHochberg’s medication and andmethod, hy perhydratedhy inperhydrated the gender dehydrated- specificgroup. group. and p- values hyperhydrated (referring group.to extreme group vs. normal hydrated group) and log2(FC) are highlighted in bold if being significant (p < 0.05). log (FC) p-Value loglog22(FC)2(FC) p-Valuep-Value ClassClassClass CompoundCompoundCompound T TT ♀ ♀ ♂♂ AccordingHyperhydratedHyperhydrated to their indication,Dehydrated Dehydrated drugs were classified into the groups (i) cardiac, (ii) Hyperhydrated Dehydrated Hotelling B&H Hotelling B&H diuretic, (iii) endocrinologic, (iv) mild cognitiveHotellingHotelling impairment, B&HB&H (v) Hotelling psychotic/neurological, B&H B&H 1 1 SimvastatinSimvastatin 43 43 1.631.63 −2.70−2.70 0.01780.0178 * * 0.03840.0384 * * 0.64720.6472 0.67660.6766 1 Simvastatin(vi) gastrointes 43 1.63tinal, and/or− (vii)2.70 nutritional0.0178 supplements, * 0.0384 *to get an0.6472 overview of 0.6766 the en- MetoprololMetoprololMetoprolol riched17 17 17 indications. 1.601.601.60 In total of- -the - 40 drugs 0.20660.20660.2066 given to 0.2794at 0.27940.2794 least 150.0295 patients,0.02950.0295 * * * 12 0.0553drugs0.05530.0553 were Candesartan 84 * 1.23 −0.81 0.0180 * 0.0384 * 0.6766 0.6766 CandesartanCandesartan found84 84 to be significantly* 1.23* 1.23 (log−02(FC).81−0.81 > 0.5, p0.0180 < 0.01800.05 *in *minimum 0.03840.0384 * one * 0.6766gender)0.6766 more 0.6766 prescribed0.6766 Apixaban 33 1.16 −1.45 0.0456 * 0.0805 0.1692 0.2417 ApixabanApixabanBisoprolol in 33hyperhydrated 131 33 1.16* 0.841.16 patients, −whereas−10.70.45−1.45 five0.00240.0456 were0.0456 ** *more * 0.0805 0.0120consumed0.0805 * 0.1692by0.16920.4214 dehydrated 0.24170.2417 0.4862patients BisoprololBisoprololAmlodipine compared131 61131 to* normal 0.840.52* 0.84 hydrated−−00.62 patients.70−0.70 (Table0.0024 0.15800.0024 2). ** Especially ** 0.0120 0.23700.0120 *cardiac * 0.42140.0192 0.4214associated * 0.4862 medication0.4862 0.0384 * Amlodipine/Olmesartanlike c 17andesartan - and amlodipine,1.29 was enriched0.2315 in the 0.2894hyperhydrated0.0067 group. ** This 0.0251 under- * AmlodipineAmlodipineOlmesartan 61 19 61 0.520.52 - 0.97−0.62−0.62 0.65790.15800.1580 0.2370 0.67660.2370 0.01920.00670.0192 * ** * 0.03840.0384 0.0251 * * Amlodipine/OlmesartanAmlodipine/Olmesartan lin 17es 17the results- showed- before1.291.29 (see Section0.23150.2315 3.2). There 0.2894 0.2894was one 0.0067 gastrointestinal0.0067 ** ** 0.02510.0251 medica- * * 2 Furosemide 19 ** 3.57 - 0.0011 ** 0.0082 ** 0.0840 0.1400 OlmesartanOlmesartanTorasemide tion,1993 19pantoprazole, 1.09- - enriched.* − 0.97Allopurinol,1.690.97 0.00070.6579 a 0.6579renal *** medication, 0.00700.67660.6766 ** was0.00670.0067 enriched0.2142 ** ** 0.0251in 0.0251the 0.2794 hyper-* * hydrated group. In the hyperhydrated group also two diuretics were enriched, torase- 2 2 3FurosemideFurosemide Allopurinol 19 46 19 ** **1.203.57 3.57 −0.92- - 0.00010.00110.0011 *** ** ** 0.0082 0.00150.0082 ** ** 0.08400.08400.3690 0.14000.1400 0.4428 4/7TorasemideTorasemide Resveratrol mide93 45 93 and furosemide.1.09−1.871.09 Furosemide*1.37 −1* .69−1.69 with0.0007 0.01240.0007 a 12* ***- times*** 0.0070 0.0384 0.0070 higher ** * ** enrichment. 0.2142 0.01580.2142 * Furthermore,0.27940.2794 0.0384 * 5 MirtazapineOlmesartan 35 and−0.97 amlodipine/0.88olmesartan 0.0181 were * found 0.0384 to be * enriched.0.1305 In both 0.2061 extreme 3 3 AllopurinolAllopurinol 46 46 1.201.071.20 −−0.92−0.92 0.00010.00140.0001 ** *** *** 0.0015 0.00840.0015 ** ** 0.36900.3690 0.44280.4428 6 Pantoprazolegroups, 85 cardiac associated medication1.32 was the most common representative0.5671 (54% 0.6301 of the 4/74/7 7 ResveratrolResveratrol Cholecalciferol 45 220 45 −10.52.87−1.87 −1.370.621.37 0.01630.01240.0124 * * * 0.0384 0.03840.0384 ** * 0.0158 0.00010.0158 * *** * 0.03840.0384 0.0015 * *** enriched medication of each group). In the dehydrated group mirtazapine, an antidepres- Class shows the classification of the compound (1 = Cardiac, 2 = Diuretic, 3 = Endocrinology, 4 = Dementia/Mild Cognitive Impairment, 5 = psychotic/neurological, 6 =sant, gastrointestinal, and resveratrol, 7 = nutritional a nutritional supplements), supplement T = total amount, used log2(FC) for MCI = binary treatment, logarithm, could Fisher’s be exact found test: * = p < 0.05, ** = p < 0.005;more Hotelling’s often. T test: Thep /RXcp with-graph and withoutof Resveratrol adjustment (Figure for multiple 4) showed testing using for Benjamini both genders & Hochberg’s lower Xc method, gender-specific p-valuesvalues (referring and tominimal extreme♀ ♂ group lower vs. R normal values hydrated in the group)control and group, log2(FC) which are highlighted is not receiving in bold if being resvera- significant (p < 0.05). trol.

(A) (B)

Figure 4. ConfidenceConﬁdence Ellipses Ellipses in in the the RXc RXc graph graph for for resveratrol: resveratrol: (A (A——female,female, B—B—male)male) Significant Signiﬁcant differentiation differentiation of ofmale male (p =(p 0=.0158) 0.0158) and and female female (p = (p 0=.0124) 0.0124) patients patients receiving receiving resveratrol. resveratrol.

4.4. Discussion Discussion 4.1. Diseases Associated with Hydration Status Are Differentiable via Bioelectrical Impedance Analyses BIA is an extensively used, noninvasive method to assess the body composition, which allows evaluating important body compartments such as fat mass, intra- and extracellular water content or the overall hydration status [11]. Furthermore, BIA has been reported to indicate the presence of diseases such as catabolism in Type 2 diabetes [22], gastrointestinal disease [23], Alzheimer/mild cognitive impairment [24], edema [25] and Nutrients 2021, 13, 1929 9 of 13

stress/inflammatory biomarkers [26]. The multifrequency device used provides data regarding intracellular and extracellular water content as well as phase angle. Based on these values, hydration classification can also be performed. However, compared with BIVA, a lower association was observed between hydration status, defined by these values, and drug or disease enrichment (data not shown). Therefore, this study focused on the association of hydration classification determined by the BIVA method. However, due to the fact that not only sex and BMI alter the BIVA assessment but also the ethnics [16], a new reference population of the Coburg cohort was defined. This was done to get a proper classification of the patients into the shown groups of normal hydrated, dehydrated and hyperhydrated patients and to redefine classification within the BIVA ellipses. For all of the four main categories (renal, cardiovascular diseases, diuretics and dementia) alteration of fluid status of a patient could be expected. Nevertheless, only renal disease and diuretic treatment were significantly differentiated in both genders. Diuretics increase the urine volume in order to treat e.g., edema [5] and thus affect the hydration status. It has been shown that bioelectrical impedance values are significantly decreased when edema is on the body side which is measured [27]. The group of patients with renal disease was defined over an elevated level of creatinine and the intake of Allopurinol. Allopurinol, a xanthine oxidase inhibitor, is used to treat elevated levels of uric acid in the blood [28]. There is a complex relationship between serum urate and kidney function [28,29]. Dehydration leads to increased reabsorption of water in the kidneys [2]. Thereby urea is also reabsorbed. Elevated levels of creatinine, a product of muscle metabolism excreted by the kidneys, are associated with acute kidney injury [30,31]. The kidney regulates body fluid volume [32] and thereby alters the hydration status. The patients group of cardiovascular vs. non-cardiovascular patients could be discriminated in the female group, however, for the males the differentiation did not reach significance. The hydration status influences vascular function and cardiovascular regulation. Dehydration impairs cutaneous vascular function and alters blood pressure regulation at rest, during exercise and orthostatic stress [33]. Also in combination with orthostatic dysregulation, dehydration leads to a higher risk of falling, which is an important risk factor for morbidity and mortality in geriatric patients [34]. On the other hand, hyperhydration causes an increased cardiovascular preload and blood pressure [35], which leads to a higher risk for heart failure and stroke [36]. Although the dehydration risk is known to increase with dementia [6], the dementia group only differed significantly in the male group (Figure3d). The smaller group size of the female patients may was a reason for this.

4.2. Several Medications Are Enriched in Hyperhydrated and Dehydrated Patients However, even if possible effects of polypharmacy, sociodemographic or dietary characteristics of patients were not taken into account or were corrected for, 12 medications were found to be significantly more frequently present in dehydrated and/or hyperhydrated patients. Furthermore, there were more enriched drugs in hyperhydrated patients (eight drugs) compared to the dehydrated patients (four drugs) (Table2). In comparison to the Clinical Dehydration Score BIA tends towards values indicating Hyperhydration [37]. Since the hyperhydrated group of the study was smaller than the dehydrated group, this aspect should not have any influence on the shown data. The largest class of active ingredi- ents was used in cardiovascular diseases in which excessive water retention, e.g., formation of edema, is typical [38]. Interestingly, two diuretics were enriched in the hyperhydrated group; even though their indications are the treatment of e.g., fluid build-ups. The absence of patients in the dehydrated group showed that the patients are not over treated. Also, Ohara [39] showed that furosemide reduces extracellular water but not intracellular water, which leads to a shift of extracellular water to total body water. Coodley [40] noticed a statistic relevant elevation of resistance and reactance due to furosemide intake as the equitation used for BIA seems not valid in people with congestive heart failure. This finding can explain the elevated level of furosemide in the hyperhydrated group as these values form the base of the calculations. Torasemide, also a loop diuretic like furosemide, Nutrients 2021, 13, 1929 10 of 13

may have the same effects but there is no research yet and further analysis is needed to verify this thesis. On the other hand, hyperhydration is the requirement for initiating a diuretic treatment, so maybe there is low treatment response or only moderate dosing to avoid dehydration in this valuable geriatric patients cohort measured with the BIA. In the dehydrated group olmesartan, an angiotensin receptor blocker, used for the treatment of high blood pressure [41], is enriched. It has the potential for inducing dehydration. Mari- etta [42] showed that this drug-induced diarrhea which also can lead to chronic diarrhea and thereby common complications like dehydration and acute kidney injury occur [42]. Also olmesartan, especially in combination with amlodipine, can cause diuresis [43]. As olmesartan was enriched in the dehydrated group, BIVA may be used as an early warning. Mirtazapine seemed to be not indicated of being found in the dehydrated group. It is usually used as an antidepressant but according to Prof. Kraft, his patients received it as an appetizer because increased appetite is one of the most common side effects of this drug [44]. But it is also known that Mirtazapine decreases the thirst sensation [45], which may be an additional explanation for the ﬁnding of this drug in the dehydrated group. The discussed effects of the individual drugs have no direct causality. So it is possible, but not necessary, that olmesartan, for example, causes diuresis.

4.3. Dehydration and Mild Cognitive Impairment Can Lead to Mutual Deterioration Bickel et al. [46] showed that in patients with dementia, dehydration and electrolyte imbalances, as well as urinary tract infections and lower respiratory tract infections, were more common. Interestingly, resveratrol is found in the dehydrated group. In a dose of 250 mg per day, it was given to improve mild cognitive impairment. MCI is defined as a syndrome of cognitive decline. It is a stage between normal aging and the more serious state of dementia [47]. Different studies showed that resveratrol improves mild cognitive impairment [48,49], which should have a positive influence on the hydration state. But the patients are still dehydrated according to their BIVA values. Additionally, dehydration decreases cognitive performance [8,9], which can lead to a further deterioration in the cognitive state. This is implicating that patients diagnosed with MCI need closer controls of their hydration status. A geriatric study by Wojszel used 14 predictors including taking procognitive medication in a multivariable logistic regression analysis, which showed a trend in higher odds for impending dehydration for this predictor. It even becomes significant when dementia as a variable was added to procognitive medications. Both models showed an overall prediction success rate of 65.7% [10]. This emphasizes the results for Resveratrol in this study. Despite the increased dehydration risk, several epi- demiologic studies reported an association between adiposity and developing Alzheimer or dementia [50,51]. But Camina et al. [52] showed that BIVA reflects dementia-related changes in body composition better than BIA and found that patients with dementia have lower Xc values. The RXc graph (Figure4) revealed higher Xc values for patients receiving resveratrol. Therefore, the Xc value can be used as a diagnosing factor that should be regarded to minimize the risk for further cognitive decline.

4.4. Limitations In contrast to the information on which drugs the patients were taking at the time of the study, there was only insufficient and, due to the still statistically too low number of patients, non-significant information on how long and with which comedication the patients had already taken the drugs. However, this could have a strong influence in part because, for example, thiazide diuretics may cause different acute and chronic reactions depending on the duration of use. Also, the possible effects of polypharmacy or sociodemographic or dietary characteristics of patients were not directly considered. However, this may well have a significant impact on hydration status, which is further indicated by the marked shift in Z-score models in the Coburg cohort compared with the American Framingham Heart Study [20] (see Section 4.1). Nevertheless, the aspect that the study was not decisively adjusted with respect to these effects and still showed significant associations underlines Nutrients 2021, 13, 1929 11 of 13

the relevance of the results. However, it should be considered in return that a correction of the Z-score model for an application to further cohorts should be checked and, if needed, also performed. Furthermore, an association with hydration status was shown and discussed for several drugs (see Section 4.2). However, it must be kept in mind that direct causality cannot be inferred from the associations shown for the drugs.

5. Conclusions This study showed that there are associations between the hydration status of elderly patients determined by bioelectrical impedance analysis and disorders such as cardiovascular, dementia, renal, and diuretic diseases. Furthermore specific medication, several not yet associated with the altered fluid status, were significantly enriched in either dehydrated or hyperhydrated patient groups. Resveratrol (antidementive), olmesartan (antihypertensive), and mirtazapine (antidepressant) were more frequently prescribed in dehydrated patients. Beyond dehydrated patients also in hyperhydrated patients drugs, such as furosemide and torasemide (diuretics), simvastatin, candesartan, bisoprolol, amlodipine (cardiac medication), pantoprazole (proton pump inhibitor), cholecalciferol, (vitamin D supplement) and mirtazapine are enriched. Remarkably, patients diagnosed with even mild cognitive impairment who received resveratrol as supportive therapy in this study group showed reduced hydration levels despite intensive geriatric care. This underscores the need for more attention to this prognostically unfavorable aspect.

Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/nu13061929/s1, Figure S1: Venn-Diagramm of the overlap between the groups cardiovascular (car, red), renal (ren, yellow), dementia (dem, green) and diuretics (diu, blue), 50 patients remain unclassified (white). Author Contributions: Data curation: R.Z.; Investigation: D.P.; Writing original draft: L.H. and S.K.; Writing review & editing: A.R., J.H. and J.K. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the Bavarian State Ministry of Education and Culture, Science and Art (BMUKK)—grant No. VII.2.F1116.CO/28/3 the Collaborative Research Project “Measuring health—methods for evidence-based health promotion”. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not required due to retrospective blinded data analysis. Data Availability Statement: Additional data is available upon request. Acknowledgments: The authors greatly appreciate the technical support by Bettina Völk, Melanie Grosch (both Regiomed Clinics, Coburg) and Markus Kreuz (Fraunhofer Institute for Cell Therapy and Immunology, Leipzig). Conflicts of Interest: The authors declare that there is no conflict of interest regarding the publication of this article.

References 1. Schmidt, S.; Hendricks, V.; Griebenow, R.; Riedel, R. Demographic change and its impact on the health-care budget for heart failure inpatients in Germany during 1995–2025. Herz 2013, 38, 862–867. [CrossRef][PubMed] 2. Hooper, L.; Bunn, D.; Jimoh, F.O.; Fairweather-Tait, S.J. Water-loss dehydration and aging. Mech. Ageing Dev. 2014, 136–137, 50–58. [CrossRef] 3. Allison, S.P.; Lobo, D.N. Fluid and electrolytes in the elderly. Curr. Opin. Clin. Nutr. Metab. Care 2004, 7, 27–33. [CrossRef] [PubMed] 4. Puga, A.M.; Lopez-Oliva, S.; Trives, C.; Partearroyo, T.; Varela-Moreiras, G. Effects of Drugs and Excipients on Hydration Status. Nutrients 2019, 11, 669. [CrossRef] 5. Khow, K.S.F.; Lau, S.Y.; Li, J.Y.; Yong, T.Y. Diuretic-associated electrolyte disorders in the elderly: Risk factors, impact, management and prevention. Curr. Drug Saf. 2014, 9, 2–15. [CrossRef] Nutrients 2021, 13, 1929 12 of 13

6. Masot, O.; Lavedán, A.; Nuin, C.; Escobar-Bravo, M.A.; Miranda, J.; Botigué, T. Corrigendum to ’Risk factors associated with dehydration in older people living in nursing homes: Scoping review’ International Journal of Nursing Studies, Volume 82 (2018) Pages 90–98. Int. J. Nurs. Stud. 2018, 83, 103. [CrossRef][PubMed] 7. Camina Martín, M.A.; de Mateo Silleras, B.; Redondo del Río, M.P. Body composition analysis in older adults with dementia. Anthropometry and bioelectrical impedance analysis: A critical review. Eur. J. Clin. Nutr. 2014, 68, 1228–1233. [CrossRef] 8. Suhr, J.A.; Hall, J.; Patterson, S.M.; Niinistö, R.T. The relation of hydration status to cognitive performance in healthy older adults. Int. J. Psychophysiol. 2004, 53, 121–125. [CrossRef] 9. Masento, N.A.; Golightly, M.; Field, D.T.; Butler, L.T.; van Reekum, C.M. Effects of hydration status on cognitive performance and mood. Br. J. Nutr. 2014, 111, 1841–1852. [CrossRef] 10. Wojszel, Z.B. Impending Low Intake Dehydration at Admission to A Geriatric Ward- Prevalence and Correlates in a Cross- Sectional Study. Nutrients 2020, 12, 398. [CrossRef] 11. Kyle, U.G.; Bosaeus, I.; de Lorenzo, A.D.; Deurenberg, P.; Elia, M.; Gómez, J.M.; Heitmann, B.L.; Kent-Smith, L.; Melchior, J.-C.; Pirlich, M.; et al. Bioelectrical impedance analysis-part II: Utilization in clinical practice. Clin. Nutr. 2004, 23, 1430–1453. [CrossRef] 12. Marini, E.; Sergi, G.; Succa, V.; Saragat, B.; Sarti, S.; Coin, A.; Manzato, E.; Buffa, R. Efficacy of specific bioelectrical impedance vector analysis (BIVA) for assessing body composition in the elderly. J. Nutr. Health Aging 2013, 17, 515–521. [CrossRef] 13. Piccoli, A.; Rossi, B.; Pillon, L.; Bucciante, G. A new method for monitoring body fluid variation by bioimpedance analysis: The RXc graph. Kidney Int. 1994, 46, 534–539. [CrossRef] 14. Kyle, U.G.; Bosaeus, I.; de Lorenzo, A.D.; Deurenberg, P.; Elia, M.; Gómez, J.M.; Heitmann, B.L.; Kent-Smith, L.; Melchior, J.-C.; Pirlich, M.; et al. Bioelectrical impedance analysis-part I: Review of principles and methods. Clin. Nutr. 2004, 23, 1226–1243. [CrossRef] 15. Schotman, J.M.; van Borren, M.M.G.J.; Kooistra, M.P.; Doorenbos, C.J.; de Boer, H. Towards personalized hydration assessment in patients, based on measurement of total body electrical resistance: Back to basics. Clin. Nutr. ESPEN 2020, 35, 116–122. [CrossRef] [PubMed] 16. Piccoli, A.; Pillon, L.; Dumler, F. Impedance vector distribution by sex, race, body mass index, and age in the United States: Standard reference intervals as bivariate Z scores. Nutrition 2002, 18, 153–167. [CrossRef] 17. Piccoli, A.; Nigrelli, S.; Caberlotto, A.; Bottazzo, S.; Rossi, B.; Pillon, L.; Maggiore, Q. Bivariate normal values of the bioelectrical impedance vector in adult and elderly populations. Am. J. Clin. Nutr. 1995, 61, 269–270. [CrossRef][PubMed] 18. Oh, J.H.; Song, S.; Rhee, H.; Lee, S.H.; Kim, D.Y.; Choe, J.C.; Ahn, J.; Park, J.S.; Shin, M.J.; Jeon, Y.K.; et al. Normal Reference Plots for the Bioelectrical Impedance Vector in Healthy Korean Adults. J. Korean Med. Sci. 2019, 34, e198. [CrossRef] 19. Piccoli, A.; Pastori, G. BIVASoftware; Department of Medical and Surgical Sciences, University of Padova: Padova, Italy, 2002. 20. Roubenoff, R.; Baumgartner, R.N.; Harris, T.B.; Dallal, G.E.; Hannan, M.T.; Economos, C.D.; Stauber, P.M.; Wilson, P.W.; Kiel, D.P. Application of bioelectrical impedance analysis to elderly populations. J. Gerontol. A Biol. Sci. Med. Sci. 1997, 52, M129–M136. [CrossRef] 21. Norman, K.; Stobäus, N.; Pirlich, M.; Bosy-Westphal, A. Bioelectrical phase angle and impedance vector analysis–clinical relevance and applicability of impedance parameters. Clin. Nutr. 2012, 31, 854–861. [CrossRef][PubMed] 22. Dittmar, M.; Reber, H.; Kahaly, G.J. Bioimpedance phase angle indicates catabolism in Type 2 diabetes. Diabet. Med. 2015, 32, 1177–1185. [CrossRef][PubMed] 23. Cox-Reijven, P.L.M.; van Kreel, B.; Soeters, P.B. Bioelectrical impedance measurements in patients with gastrointestinal disease: Validation of the spectrum approach and a comparison of different methods for screening for nutritional depletion. Am. J. Clin. Nutr. 2003, 78, 1111–1119. [CrossRef][PubMed] 24. Cova, I.; Pomati, S.; Maggiore, L.; Forcella, M.; Cucumo, V.; Ghiretti, R.; Grande, G.; Muzio, F.; Mariani, C. Nutritional status and body composition by bioelectrical impedance vector analysis: A cross sectional study in mild cognitive impairment and Alzheimer’s disease. PLoS ONE 2017, 12, e0171331. [CrossRef][PubMed] 25. Piccoli, A.; Brunani, A.; Savia, G.; Pillon, L.; Favaro, E.; Berselli, M.E.; Cavagnini, F. Discriminating between body fat and fluid changes in the obese adult using bioimpedance vector analysis. Int. J. Obes. Relat. Metab. Disord. 1998, 22, 97–104. [CrossRef] [PubMed] 26. Tsigos, C.; Stefanaki, C.; Lambrou, G.I.; Boschiero, D.; Chrousos, G.P. Stress and inflammatory biomarkers and symptoms are associated with bioimpedance measures. Eur. J. Clin. Investig. 2015, 45, 126–134. [CrossRef][PubMed] 27. Codognotto, M.; Piazza, M.; Frigatti, P.; Piccoli, A. Influence of localized edema on whole-body and segmental bioelectrical impedance. Nutrition 2008, 24, 569–574. [CrossRef] 28. Stamp, L.K.; Chapman, P.T.; Palmer, S.C. Allopurinol and kidney function: An update. Jt. Bone Spine 2016, 83, 19–24. [CrossRef] [PubMed] 29. Kuwabara, M.; Bjornstad, P.; Hisatome, I.; Niwa, K.; Roncal-Jimenez, C.A.; Andres-Hernando, A.; Jensen, T.; Milagres, T.; Sato, Y.; Garcia, G.; et al. Elevated Serum Uric Acid Level Predicts Rapid Decline in Kidney Function. Am. J. Nephrol. 2017, 45, 330–337. [CrossRef] 30. Mercado, M.G.; Smith, D.K.; Guard, E.L. Acute Kidney Injury: Diagnosis and Management. Am. Fam. Physician 2019, 100, 687–694. 31. Baum, N.; Dichoso, C.C.; Carlton, C.E. Blood urea nitrogen and serum creatinine. Urology 1975, 5, 583–588. [CrossRef] 32. Clark, W.F.; Sontrop, J.M.; Huang, S.-H.; Moist, L.; Bouby, N.; Bankir, L. Hydration and Chronic Kidney Disease Progression: A Critical Review of the Evidence. Am. J. Nephrol. 2016, 43, 281–292. [CrossRef][PubMed] Nutrients 2021, 13, 1929 13 of 13

33. Watso, J.C.; Farquhar, W.B. Hydration Status and Cardiovascular Function. Nutrients 2019, 11, 1866. [CrossRef] 34. Lee, J.; Choi, M.; Kim, C.O. Falls, a fear of falling and related factors in older adults with complex chronic disease. J. Clin. Nurs. 2017, 26, 4964–4972. [CrossRef][PubMed] 35. Zorbas, Y.G.; Federenko, Y.F.; Naexu, K.A. Effect of daily hyperhydration on fluid-electrolyte changes in endurance-trained volunteers during prolonged restriction of muscular activity. Biol. Trace Elem. Res. 1995, 50, 57–78. [CrossRef] 36. Lionakis, N.; Mendrinos, D.; Sanidas, E.; Favatas, G.; Georgopoulou, M. Hypertension in the elderly. World J. Cardiol. 2012, 4, 135–147. [CrossRef] 37. Rösler, A.; Lehmann, F.; Krause, T.; Wirth, R.; von Renteln-Kruse, W. Nutritional and hydration status in elderly subjects: Clinical rating versus bioimpedance analysis. Arch. Gerontol. Geriatr. 2010, 50, e81-5. [CrossRef][PubMed] 38. Odeme Schönermarck, U. Herz oder Nieren? Finden Sie heraus, woran es liegt! MMW Fortschr. Med. 2015, 157, 56–60, quiz 61. [CrossRef] 39. Ohara, K.; Masuda, T.; Murakami, T.; Imai, T.; Yoshizawa, H.; Nakagawa, S.; Okada, M.; Miki, A.; Myoga, A.; Sugase, T.; et al. Effects of the sodium-glucose cotransporter 2 inhibitor dapagliflozin on fluid distribution: A comparison study with furosemide and tolvaptan. Nephrology 2019, 24, 904–911. [CrossRef] 40. Coodley, E.L.; Segal, J.L.; Smith, D.H.; Neutel, J.M. Bioelectrical impedance analysis as an assessment of diuresis in congestive heart failure. Ann. Pharmacother. 1995, 29, 1091–1095. [CrossRef] 41. Al-Majed, A.A.; Bakheit, A.H.H.; Abdel Aziz, H.A.; Al-Jallal, A.A.M. Olmesartan. Profiles Drug Subst. Excip. Relat. Methodol. 2017, 42, 241–286. [CrossRef] 42. Marietta, E.V.; Cartee, A.; Rishi, A.; Murray, J.A. Drug-induced enteropathy. Dig. Dis. 2015, 33, 215–220. [CrossRef][PubMed] 43. Walter, A.N.; Lenz, T.L. Hydration and Medication Use. Am. J. Lifestyle Med. 2011, 5, 332–335. [CrossRef] 44. Fava, M. Weight gain and antidepressants. J. Clin. Psychiatry 2000, 61, 37–41. 45. Leth-Møller, K.B.; Hansen, A.H.; Torstensson, M.; Andersen, S.E.; Ødum, L.; Gislasson, G.; Torp-Pedersen, C.; Holm, E.A. Antidepressants and the risk of hyponatremia: A Danish register-based population study. BMJ Open 2016, 6, e011200. [CrossRef] [PubMed] 46. Bickel, H.; Hendlmeier, I.; Heßler, J.B.; Junge, M.N.; Leonhardt-Achilles, S.; Weber, J.; Schäufele, M. The Prevalence of Dementia and Cognitive Impairment in Hospitals. Dtsch. Arztebl. Int. 2018, 115, 733–740. [CrossRef][PubMed] 47. Gauthier, S.; Reisberg, B.; Zaudig, M.; Petersen, R.C.; Ritchie, K.; Broich, K.; Belleville, S.; Brodaty, H.; Bennett, D.; Chertkow, H.; et al. Mild cognitive impairment. Lancet 2006, 367, 1262–1270. [CrossRef] 48. Kennedy, D.O.; Wightman, E.L.; Reay, J.L.; Lietz, G.; Okello, E.J.; Wilde, A.; Haskell, C.F. Effects of resveratrol on cerebral blood flow variables and cognitive performance in humans: A double-blind, placebo-controlled, crossover investigation. Am. J. Clin. Nutr. 2010, 91, 1590–1597. [CrossRef] 49. Witte, A.V.; Kerti, L.; Margulies, D.S.; Flöel, A. Effects of resveratrol on memory performance, hippocampal functional connectivity, and glucose metabolism in healthy older adults. J. Neurosci. 2014, 34, 7862–7870. [CrossRef] 50. Whitmer, R.A.; Gunderson, E.P.; Quesenberry, C.P.; Zhou, J.; Yaffe, K. Body mass index in midlife and risk of Alzheimer disease and vascular dementia. Curr. Alzheimer Res. 2007, 4, 103–109. [CrossRef][PubMed] 51. Kivipelto, M.; Ngandu, T.; Fratiglioni, L.; Viitanen, M.; Kåreholt, I.; Winblad, B.; Helkala, E.-L.; Tuomilehto, J.; Soininen, H.; Nissinen, A. Obesity and vascular risk factors at midlife and the risk of dementia and Alzheimer disease. Arch. Neurol. 2005, 62, 1556–1560. [CrossRef][PubMed] 52. Camina Martín, M.A.; de Mateo Silleras, B.; Nescolarde Selva, L.; Barrera Ortega, S.; Domínguez Rodríguez, L.; Redondo Del Río, M.P. Bioimpedance vector analysis and conventional bioimpedance to assess body composition in older adults with dementia. Nutrition 2015, 31, 155–159. [CrossRef][PubMed]