STUDY PROTOCOL

Intermittent Fasting (Alternate Day Fasting)in Healthy, Non-obese Adults: Protocol for a CohortTrial with an Embedded Randomized Controlled PilotTrial

Norbert J. Tripolt . Slaven Stekovic . Felix Aberer . Jasmin Url . Peter N. Pferschy . Sabrina Schroder .

Nicolas Verheyen . Albrecht Schmidt . Ewald Kolesnik . Sophie H. Narath . Regina Riedl .

Barbara Obermayer-Pietsch . Thomas R. Pieber . Frank Madeo . Harald Sourij

Received: May 17, 2018 / Published online: July 25, 2018� The Author(s) 2018

ABSTRACT

Background/objectives: Alternate day fasting(ADF) is a subtype of intermittent fasting and isdefined as a continuous sequence of a fast day(100% energy restriction, zero calories) and afeed day (ad libitum food consumption),resulting in roughly 36-h fasting periods. Pre-vious studies demonstrated weight reductionsand improvements of cardiovascular risk factorswith ADF in obese subjects. However, rigorousdata on potential endocrine, metabolic and

cardiovascular effects, besides weight loss, arelacking. Therefore we aim to investigate theshort- and mid- to long-term clinical andmolecular effects of ADF in healthy non-obesesubjects.Methods: We will perform a prospective cohortstudy with an embedded randomized controlledtrial (RCT) including 90 healthy subjects. Thirtyof them will have performed ADF for at least6 months (mid-term group). Sixty healthy sub-jects without a particular diet before enrolmentwill serve as the control group. These subjectswill be 1:1 randomized to either continuingtheir current diet or performing ADF for4 weeks. All subjects will undergo study proce-dures that will be repeated in RCT participantsafter 4 weeks. These procedures will includeassessment of outcome parameters, dual-energyX-ray absorptiometry, measurement of

Norbert J. Tripolt and Slaven Stekovic have contributedequally to this work.

Enhanced digital features To view enhanced digitalfeatures for this article go to https://doi.org/10.6084/m9.figshare.6683597.

N. J. Tripolt � F. Aberer � J. Url � P. N. Pferschy �B. Obermayer-Pietsch � T. R. Pieber � H. Sourij (&)Division of Endocrinology and Diabetology,Department of Internal Medicine, MedicalUniversity of Graz, Graz, Austriae-mail: ha.sourij@medunigraz.at

S. Stekovic � S. Schroder � F. Madeo (&)Institute of Molecular Biosciences, University ofGraz, Graz, Austriae-mail: frank.madeo@uni-graz.at

J. Url � P. N. Pferschy � S. H. Narath �B. Obermayer-Pietsch � T. R. Pieber � H. SourijCenter for Biomarker Research in Medicine, CBMed,Graz, Austria

N. Verheyen � A. Schmidt � E. KolesnikDivision of Cardiology, Department of InternalMedicine, Medical University of Graz, Graz, Austria

R. RiedlInstitute for Medical Informatics, Statistics andDocumentation, Medical University of Graz, Graz,Austria

F. MadeoBioTechMed Graz, Graz, Austria

Adv Ther (2018) 35:1265–1283

https://doi.org/10.1007/s12325-018-0746-5

endothelial function, an oral glucose tolerancetest, 24-h blood pressure measurement, retinalvessel analysis, echocardiography and physicalactivity measurement by an accelerometer.Blood, sputum, buccal mucosa and faeces willbe collected for laboratory analyses. Participantsin the RCT will wear a continuous glucosemonitor to verify adherence to the studyintervention.Planned outcomes: The aim of this project is toinvestigate the effects of ADF on human physi-ology and molecular cellular processes. Thisinvestigation should gain in-depth mechanisticinsights into the concept of ADF and form thebasis for larger subsequent cohort recruitmentand consecutive intervention studies.Trial registration: NCT02673515; registered 24November 2015. Current protocol date/version:7 February 2017/version 1.8.

Keywords: Alternate day fasting; Bloodpressure; Cell death; Cell signalling; Cohortstudy; Endothelial function; Glucosemetabolism; Healthy subjects; Insulinsensitivity; Intermittent fasting

INTRODUCTION

Background

Intermittent fasting, alternate day fasting (ADF)and further forms of periodic caloric restrictionare not really methods of energy deprivation,because—at least theoretically—persons couldeat double on the second day. Intermittentfasting in heterogeneous forms has been prac-ticed for religious and health reasons for hun-dreds of years [1]. ADF regimens are defined as‘feast days’ on which food is consumed ad libi-tum, which alternate with ‘fast days’ on whichfood is withheld or reduced. The feast and fastperiods are typically 24 h each, but may vary[4]. This is the most frequently used protocol forintermittent fasting studies in mice and rats [2].In everyday life, this protocol is often modifiedto allow a small caloric intake (e.g. about 25% ofthe individual’s energy needs).

Over the past decade, ADF has gained greatpopularity in a weight management context[3–5], leading to a widely used lifestyle trendthat is said to have effects on general healthstatus and prolongation of life expectancy.However, scientific evidence for the effects ofADF in humans is rare and metabolic data areoften extrapolated from animal studies only.

Several trials found beneficial metabolic andcardiovascular effects, such as reduced circulat-ing glucose concentrations [6–8] as well asreductions in heart rate or blood pressure [9]and cholesterol or triglyceride concentrations[10] after 8–24 weeks of ADF in rodents.

Additionally, a study in which mice wereallocated to either intermittent fasting with ahigh-fat diet in the feeding periods or regularlyfeeding with isocaloric, carbohydrate and fatintake in both groups demonstrated that ADFwas associated with prevention of obesity,hyperinsulinaemia, hepatic steatosis andinflammation [11].

Besides caloric restriction and effects onestablished cardiovascular risk factors, inter-mittent fasting seems to activate cell autophagyprocesses that increase cellular stress resistanceand remove accumulated molecules and orga-nelles that can lead to various pathologies[12–14]. Further research suggested a life-pro-longing effect of ADF in rodents [15–17].

Findings from ADF studies in humansshowed significant weight reductions [18–23],reduction in waist-to-hip ratio [18, 20–22, 24],fat mass [18, 20–23, 25–27] and also fat-freemass [22, 25–27]. Also the effects seen in theanimal studies on cardiovascular risk factors,such as reductions in triglycerides [28–31], low-density lipoprotein (LDL) cholesterol [28–30]and systolic blood pressure [30] and increases inLDL particle size [21, 32, 33] and high-densitylipoprotein (HDL) cholesterol [21, 28, 31] wereobserved. Many studies [18, 21–23, 25,27, 28, 34] showed no change in fasting glucoselevels, though some of these trials[21, 22, 25, 27] reported a decrease in fastingcirculating insulin concentrations.

A major limitation of the majority of thesetrials [35] is that they either compared theeffectiveness of intermittent fasting with con-tinuous energy restriction without a

1266 Adv Ther (2018) 35:1265–1283

Table 1 Visit overview

Randomized Controlled TrialRCT Study period

Baseline

(Visit 1)

Visit

2 3 4

Follow-up

(Visit 5)

Time point 0

9 days ± 4 days after

baseline

21 days ± 2 days after

visit 2

7 days ± 2 days after

visit 3

2 years aftervisit 3

Enrolment

Eligibility screen X

Informed consent X

Randomization X

Allocation X

Interventions

Alternate Day Fasting (ADF rand.)

Ingestion ad libitum (Controls)

Assessments

CRF

Demographic data X

Anthropometric measurements X X

Vital signs (Blood pressure, heart rate) X X X X

Medical History X X

Concomitant medication X X X X

Physical examination X X

Laboratory

Oral glucose tolerance test (OGTT) X X

Saliva sampling X X

Urine pregnancy test (females) X X

Buccal mucosa sampling X X

Adv Ther (2018) 35:1265–1283 1267

Table 1 continued

Issue of devices

24h ambulatory blood pressure monitoring

X X

Physical activity measurement X X

Continuous glucose monitoring XX (ADF only)

Glucometer for calibration of glucose monitoring

X X

Investigations

Endothelial function (EndoPAT) X X

Bone densitometry and body composition (Dual-energy X-ray absorp�ometry)

X X

Hand grip test X X

Echocardiogram X X

Electrocardiogram (ECG) X X

Retina vessel analysis X X

Resting energy expenditure (REE) X X

Questionnaires

Food frequency questionnaire (FFQ) X X X

International physical activity questionnaire (IPAQ)

X X X

Health survey short form (SF36) X X X

Cohort studyLong term ADF cohort Study period

Baseline Visit 2 Visit 3

Time point 09 days ± 4 days after

baseline 2 years (± 2 months)

after visit 3

Enrolment

Eligibility screen X

Urine sampling X X

Faeces sampling X X

Laboratory measurements including routine safety lab

X X X

1268 Adv Ther (2018) 35:1265–1283

Table 1 continued

Demographic data X X

Anthropometric measurements X X

Vital signs (blood pressure, heart rate) X X X

Medical history X X

Concomitant medication X X X

Physical examination X X

Laboratory

Oral glucose tolerance test (OGTT) X X X

Saliva sampling X X

Urine pregnancy test (females) X X

Buccal mucosa sampling X

Laboratory measurements including routine safety lab

X X X

Urine sampling X X X

Interventions

Alternate day fasting

Assessments

CRF

Faeces sampling X

Issue of Devices

24h ambulatory blood pressure monitoring X

Physical activity measurement X

Continuous glucose monitoring X

Glucometer for calibration of glucose monitoring

X

Investigations

Endothelial function X

Allocation X

Informed consent X

Randomization

Adv Ther (2018) 35:1265–1283 1269

randomized controlled study design or exam-ined the effect of intermittent fasting alone inan uncontrolled way. The majority of these tri-als were performed in overweight subjects withor without type 2 diabetes mellitus.

Our trial is the first rigorously randomizedcontrolled human study comparing metabolicand molecular effects of ADF with 100% caloricrestriction on the fast day to a control groupwith daily food intake over a period of 4 weeksin healthy subjects, controlled for study groupadherence by continuous glucose monitoring.In addition, we will include a cohort of 30subjects performing ADF for already more than6 months.

The primary objective is to elucidate towhich extent ADF influences human physiol-ogy in healthy individuals both short- and mid-to long term. The secondary objective of thisstudy is to investigate novel molecular markersand cellular pathways of ageing and age-relateddiseases and whether they are influenced byADF. The information gained through thisproject will be used to develop a strategy forfollow-up studies in model organisms andhumans. As a long-term goal, we want to iden-tify new targets (metabolites, macromoleculesand pathways) that can be used for more precise

clinical diagnosis and in the treatment andprevention of ageing and age-related diseases.

The overarching aim of this research projectis to elucidate whether and to what extentalternate day fasting influences physiology inhealthy individuals in short- and mid- to long-term ADF. The secondary objective is to definenovel molecular markers and cellular pathwaysof ageing and age-related diseases and whetherthey are influenced by ADF.

METHODS

Study Design

The InterFast trial is a cohort study with anintegrated randomized controlled pilot trial.The aim of our cohort study with embeddedpilot randomized control trial (RCT) is toinvestigate the effects of repeating fasting peri-ods on human physiology, the ageing processand molecular cellular processes in humans. Byperforming this trial we are able to study mid-to long-term (subjects in the cohort study whohad already practiced ADF for at least 6 months)and short-term effects (subjects randomized toADF for 4 weeks or control group) of this

Table 1 continued

Echocardiogram X

Electrocardiogram (ECG) X

Retina vessel analysis X

Resting energy expenditure (REE) X

Questionnaires

Food frequency questionnaire (FFQ) X X

International physical activity questionnaire (IPAQ)

X X

Health survey short form (SF36) X X

Bone densitometry and body composition (Dual-energy X-ray absorp�ometry)

X

Hand grip test X

1270 Adv Ther (2018) 35:1265–1283

nutritional intervention. All study participantsin the RCT and cohort trial attend our studycentre for a 2-year follow-up visit. In this studyvisit vital signs, anthropometric measures andblood samples are collected (see Table 1). Fur-thermore, patients are asked to answer ques-tionnaires about nutritional behaviours andquality of life.

Planned Outcomes

The following outcome measures include boththe cohort study and the RCT. In the cohorttrial the group comparison will be performed atbaseline, whereas in the embedded RCT, thechanges over 4 weeks between the groups willbe compared.

Primary Outcome Measures (DifferencesBetween Groups)

• Differences in body composition (fat massand lean body mass).

• Differences in resting energy expenditure.• Differences in insulin sensitivity [Matsuda

index, insulin sensitivity index (ISI), home-ostatic model assessment of insulin resis-tance (HOMA-IR) index, quantitative insulinsensitivity check index (QUICKI)].

• Difference in glycaemic pattern (continuousglucose measurement).

• Differences in endothelial function (reactivehyperaemia index, stiffness index, aorticaugmentation index).

• Differences in mean 24-h blood pressure.• Differences in hormonal regulation (repro-

duction hormones, appetite hormones andothers).

Secondary Outcome Measures (DifferencesBetween Groups)

• Levels of various cell signalling and cell cycleproteins.

• Post-translational changes of signallingproteins.

• Differences in parameters of cell death(apoptosis, stress resistance, mitochondrialdamage).

• Differences in various markers of ageing[telomere length, insulin growth factor-1(IGF-1), etc.]

• Difference in epigenetic markers andmetabolomics.

• Markers of inflammation and cardiovascularhealth.

• Differences in echocardiographic measures.

Study Registration

The study is performed according to the prin-ciples of the International Declaration of Hel-sinki and of good clinical practice (GCP), isapproved by the ethics committee of the Med-ical University of Graz (27–166 ex 14/15) and isconducted at the Division of Endocrinology andDiabetology at the Medical University of Graz(academic hospital), Austria. The study is regis-tered in the Database for Clinical Trials of theUS National Institutes of Health (ClinicalTri-als.gov Identifier: NCT02673515). Importantprotocol modifications will be reported to thelocal Ethics Committee of the Medical Univer-sity of Graz. This study follows the internationalrecommendations for interventional trials.

Sample Selection

The study population consists of 90 healthysubjects. Thirty participants already have prac-ticed alternate day fasting for at least 6 months.Participants in the RCT are healthy subjectswithout recent AFD experience and are matchedat least by sex and age. A separate informedconsent was created for the interventional trial.

Due to the pilot nature of the study, no for-mal power calculation was performed.

Recruitment

Patients without previous ADF history who areassigned for the 4-week short-term fasting trialare identified via Primary Care and advertise-ments (Facebook, newspaper, trial-specificwebsite). ADF subjects are recruited by adver-tisements and leaflets at Alternate Day FasterEvents and online via different blogs. Subjectsare recruited by the investigators and the study

Adv Ther (2018) 35:1265–1283 1271

nurse in charge. Detailed information about thestudy background and protocol are given, andany potential questions brought forward by thesubjects are answered before signing theinformed consent. Each eligible subject who iswilling to participate in the study signs a writ-ten informed consent before any particularstudy procedure.

Randomization Procedures

As part of the pilot RCT, the study participantswithout previous ADF history are randomlyassigned to one of the two groups, ADF orcontrol group, in a 1:1 ratio, using the web-based randomization tool ‘‘Randomizer forClinical Trials’’ (www.randomizer.at), providedby the Institute of Medical Informatics, Statis-tics and Documentation of the MedicalUniversity Graz, Austria. The randomization isstratified by gender.

Inclusion Criteria

• Age between 35 and 65 years, both inclusive.• Body mass index (BMI) between 22.0 and

30.0 kg/m2, both inclusive.• Fasting blood glucose \110 mg/dl without

glucose-lowering medication.• LDL cholesterol \180 mg/dl without lipid-

lowering medication.• Blood pressure \140/90 mmHg without

blood pressure-lowering medication.• Stable weight (change \± 10%) for

3 months immediately prior to the study.

Exclusion Criteria

• History of metabolic disorder.• History of cardiovascular disease.• Acute or chronic inflammatory disorder.• Known malignancy.• Tobacco consumption within 5 years.• Abuse of recreational drugs within 5 years.• Alcohol abuse (more than 15 drinks/week).

• Dietary restrictions (e.g. vegetarianism andveganism).

• Women and men on hormonalsupplementation.

• Women or men on hormone-based contra-ceptive agents within the past 2 months.

• Therapy with antidepressants within thepast 6 months.

• Regular therapy with acetylsalicylic acid orcurrent medication to regulate blood sugar,blood pressure or lipids.

• Women who are pregnant, breast-feeding oraiming to become pregnant during thecourse of the trial.

Data Collection

Once informed consent has been signed,physicians complete a paper-based case reportform (CRF) for each participant to documentsocio-demographic information, anthropomet-ric measurements, vital signs, informationabout medical history, concomitant medica-tion, physical examination and dietary habits.Source documents comprise the CRF and hos-pital records as well as laboratory records. Alldocuments are stored safely in a confidentialmanner. The subjects are referred to by a uniquestudy subject number/code; their initials anddate of birth will be reported on all study-specific documentations. Personal identifiersare on the signed informed consent forms,subject identification log and subject clinicalfile, all of which will be stored securely by theclinical site. Source data are prepared accord-ingly for internal monitoring and possibleexternal audits or inspections by regulatoryauthorities (Figs. 1, 2).

Interventions

ADF group: Subjects are asked to refrain fromeating on the fast days and to eat ad libitum onthe days between the fast days. On fasting days,subjects are only allowed to consume water,flavoured carbonated water, unsweetened tea or

1272 Adv Ther (2018) 35:1265–1283

coffee. Furthermore, diet sodas or so called‘‘calorie-free’’ meals or beverages are restricted.

Measurements

Oral Glucose Tolerance Test (OGTT)Various insulin sensitivity indices derived froma frequently sampled, extended OGTT are cal-culated accounting for the various weighting ofhepatic and peripheral insulin sensitivityexpressed by the different indices. Insulin,C-peptide and glucose are measured before and15, 30, 60 and 120 min after a rapid ingestion of75 g glucose, dissolved in 300 ml of faucet water(�Glucoral 75 citron, Germania Pharmazeutika,Vienna). A standard gauge cannula is placedinto a subcutaneous vein for blood sampling. Toprevent blood clotting in the cannula and tokeep it working, it is flushed with sterile normalsaline after each blood drawing. The blood ateach time point is placed into a fluoride oxalatetube (1 ml) for plasma glucose and into a serumtube for analyses of insulin and c-peptide levels.

The insulin sensitivity index is calculatedfrom the OGTT according to four different

equations: the Matsuda index (ISOGTT):

ISoGTT ¼ 10000ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

glucose 0�insulin 0ð Þ�ðmeanp glucose �

mean insulinÞ [36], the HOMA-IR: HOMA - IR ¼FPG mmol=lð Þ�FSI U=lð Þ

22:5 [37] and the QUICKI ¼1

log insulin 0ð Þþlogðglucose 0Þ [38]. The last equation was

proposed by Stumvoll et al.: ISI ¼ 0:222�0:00333 � BMI � 0:0000779 � Ins 120 � 0:000422�age [39].

Beta cell function is estimated in the fasting

state with HOMA - b ¼ 20�Insulin 0Glucose 0�3:5 [37] and

during the OGTT with the Stumvoll index:1st phase ¼ 1283 þ 1:829 � Ins30 � 138:7 � Glc30 þ 3:772 � Ins0 plus 2nd phase ¼ 286 þ 0:416�Ins 30 � 25:94 � Gluc 30 þ 0:926 � Ins 0 [39] andthe ratio of the incremental insulin to glucoseresponse over the first 30 min during the OGTTDInsulin ð30ÞDGlucose ð30Þ.

Endothelial FunctionEndo-PAT 2000 (Itamar Medical Ltd., Caesarea,Israel) is used to measure endothelium-depen-dent vaso-reactivity [40]. In brief, before mea-surements, the subjects are in supine position

Fig. 1 Study design

Adv Ther (2018) 35:1265–1283 1273

for a minimum of 10 min in a quiet, tempera-ture-controlled room with dimmed lights.Probes are placed on both index fingers andpulse wave amplitudes are detected and recor-ded during the investigation. After a 5-min

baseline measurement, arterial flow is occludedusing a cuff on the non-dominant arm. The cuffis inflated to 60 mmHg above systolic pressure.After 5 min of occlusion, the cuff is rapidlydeflated to allow reactive hyperaemia. Pulse

Fig. 2 Flow chart enrolment

1274 Adv Ther (2018) 35:1265–1283

wave amplitudes are recorded again for at least5 min [41]. The ratio of the peripheral arterialtone (PAT) signal after cuff release comparedwith baseline is calculated through a computeralgorithm automatically normalizing for base-line signal and indexed to the contralateral arm.The calculated ratio represents the reactivehyperaemia index (RHI) [42].

Bone Densitometry and Body CompositionBone density scanning, also called dual-energyX-ray absorptiometry (DXA) or bone densito-metry, is an enhanced form of X-ray technologyused to measure bone mass. Whole body DXAmeasurement is performed with a GE LunariDEXA (GE Healthcare, Waukesha, WI, USA) forthe purpose of estimating body fat percentageaccording to the departmental standard oper-ating procedure. Body regions are defined usingstandard anatomical partitions. Scan areas areanalysed to determine lean mass, fat mass, bonemineral density and total body composition.

Laboratory Measurements (Clinicaland Hormonal)Insulin and c-peptide are measured by chemi-luminescence on an ADVIA Centaur system(Siemens Healthcare Diagnostics, Eschborn,Germany). Hormones such as anti-muellerianhormone (AMH), testosterone, cortisol, thyreo-tropin, triiodothyronine and thyroxine, oestro-gen, sexual hormone-binding globulin (SHBG),luteotropic hormone (LH), follicle-stimulatinghormone (FSH) and 25(OH)vitamin D are mea-sured using automated analysers: AMH byBeckmann-Coulter, Krefeld, Germany; testos-terone, cortisol, thyreotropin, triiodothyronineand thyroxine by Siemens ADVIA Centaur,Eschborn, Germany; SHBG by Roche Diagnos-tics, Mannheim, Germany; oestrogen, LH andFSH by Triturus, Biomedical Diagnostics,Antwerp, Belgium, 25(OH)vitamin D by iSYS,IDS, Boldon, UK, respectively. Samples forappetite hormones (leptin, ghrelin, PYY andothers) are stored at -80 �C until analysis aftercentrifugation. Free fatty acids are measuredenzymatically (Wako Chemical, Neuss, Ger-many) on an Olympus AU640 (Olympus Diag-nostica, Hamburg, Germany). Routine

parameters are determined using a cobas�

analyser (Roche Diagnostics, Mannheim,Germany).

Laboratory Measurements (SpecializedResearch—Ageing Parameters)Molecular parameters of ageing are determinedby a variety of already established and previ-ously described methods. Total levels as well aslevels of phosphorylated proteins within dif-ferent cell signalling pathways are measuredusing the Western blot system. Inflammatorymarkers, insulin signalling and precise charac-teristics of molecular pathways are analysedusing certified and commercially availableV-PLEX Assays (Meso Scale Discovery, Rockville,MD, USA). Epigenetic changes (methylationand acetylation) are analysed via traditionalWestern blotting using antibodies targetingpost-translationally modified histones. Allmethods based on antigen–antibody interactionare followed using commercially available anti-bodies to assure better reproduction and opti-mal processes and procedures. Metabolomics,proteomics and acetylproteomics are measuredusing high-performance liquid chromatography(HPLC)-coupled mass spectrometry (MS) orHPLC-coupled MS/MS methods. Differentparameters of cell death (apoptosis, stress resis-tance, mitochondrial damage) and age-relatedparameters are analysed using BD BioscienceLSRFortessa and FACSAria IIu flow cytometrysystems.

Physical Activity MeasurementThe MoviSens device consists of a5.3 9 3 9 2 cm-sized body and can be fixed witha clip to the hip area. During the measurements,the sensor is placed on the right side of thesubjects’ hip. The sensor is a three-axial accel-eration sensor with a range of ± 8 g, 12-bit res-olution and 128-Hz sampling rate. The recordeddata from the sensor, including raw data fromthe acceleration sensor, can be displayed on acomputer via USB cable. Energy expenditure isdisplayed in 1-s steps. Short time intervals allowmonitoring spontaneous activities. The recog-nition of different activities is based on theextraction of mathematical and statistical

Adv Ther (2018) 35:1265–1283 1275

features of the raw acceleration signal. The fea-tures are calculated for each 4-s segment. Cal-culated features are the maximum frequency,step count and number of mean crossings.These features are the input information ofquantifying the activity of the subject. Activitiesthat can be detected are rest (combination oflying, sitting and standing), bicycle or ergome-ter, going upstairs, walking (sensor-based dif-ferentiation of jogging, going downstairs,walking slowly, normally and fast) and unde-fined activity. The accuracy of the activityrecognition algorithm is discussed by Jatobaet al. [43]. This device measures the physicalactivity of study participants over the 7-dayperiod.

Non-invasive 24-h Ambulatory Blood PressureMonitoringBefore measurement, arm circumference ismeasured to allow correct choice of cuff size.Subjects wear the blood pressure device for asingle 24-h period. During this time, the deviceis programmed to inflate and record bloodpressure at pre-specified intervals (usually every15 min during daytime hours and every 30 minduring nighttime hours), which provides 50–75approximately evaluable blood pressure record-ings during the 24-h period. Study participantsare asked to keep an activity log throughout the24-h period so activities can be connect ontothe blood pressure recordings. The process foreducating subjects about the blood pressuredevice and setting up the test takes approxi-mately 10 min. Patients should undergo testingon a regular workday [44]. The devices used arestandardized brachial, oscillometric, automatedMobil-O-Graph� 24-h PWA (Stolberg, Ger-many) devices, which are used in the course ofmany studies.

Resting Energy Expenditure (REE)REE is a component of energy expendituremeasured by indirect calorimetry (IC). Subjectshave to rest at least 30 min after not less than8 h of sleep and after at least 3 h of fasting.During measurements, subjects must be keptfully awake and completely relaxed at normalbreath. The measurements have to be carried

out for at least 30 min and are performed atstandard neutral hospital room temperature. Abreath mask covers the subject’s face. Oxygenconsumption and carbon dioxide productionwill be measured and energy expenditure cal-culated by the Weir formula [45].

Hand Grip TestOne of the most common methods of measur-ing muscle strength is the isometric gripstrength test. We will measure isometric gripstrength using a handgrip dynamometer(JAMAR�, Nottinghamshire, UK). The test isperformed in the sitting position with the upperarm of tested extremity adducted, the elbowflexed at 90�; the forearm and wrist are set inneutral position [46]. The testing protocol con-sists of three maximal isometric contractions for5 s on both hands with a rest period of at least60 s. The mean value of the three contractions isused for determination of grip strength. Thesubjects are instructed to squeeze thedynamometer as hard as possible [47].

Electrocardiogram (ECG)A 12-lead ECG is performed by the study nurse,who is familiar with the used ECG device (Mac1200, Marquette Hellige GmbH, Freiburg, Ger-many). The ECG printout is interpreted, signedand dated by a co-investigator.

Saliva SamplingSaliva is collected with Salitubes (DRG Instru-ments Marburg, Germany) using a routinemethod for saliva collection and dedicatedtubes as well as standardized instructions for thecollection. The hormones and salivary micro-biome are analysed (adapted ELISAs by BSM,Vienna, Austria).

Buccal Mucosa SamplingThe study participants perform a mouth washwith distilled water (three replicates) to removetraces of saliva, contamination and surfacemucous. Then exfoliation of the two cheeks tomaximize the cell sample and remove anyunknown trace that could be caused by sam-pling only one cheek is performed. The exfoli-ation should use a disposable brush similar to

1276 Adv Ther (2018) 35:1265–1283

that used for a Papanicolaou smear, performingten rotations of the brush against the inner wallof the cheeks starting in the centre and gradu-ally increasing the circumference, producing aspiral effect to increase the sampling of a largerarea and prevent erosion of a single region.After exfoliation, the brushhead must be placedin a container (test tube) with 0.4 ml of buffersolution containing 0.01 M Tris hydrochloride(Tris HCl), 0.1 M acid ethylenediaminete-traacetic (EDTA) and 0.02 M sodium chloride(NaCl) with pH 6.8, turned in the way that thecells are released and displaced on the inneredge of the container [48].

The micronucleus (MN) assay in exfoliatedbuccal cells is a useful and minimally invasivemethod for monitoring genetic damage inhumans [49]. The buccal micronucleus cytome(BMCyt) assay is a minimally invasive methodfor studying DNA damage, chromosomal insta-bility, cell death and the regenerative potentialof human buccal mucosal tissue. This method isincreasingly used in molecular epidemiologicalstudies for investigating the impact of nutrition,lifestyle factors, genotoxin exposure and geno-type on DNA damage, chromosome malsegre-gation and cell death. The biomarkers measuredin this assay have been associated withincreased risk of accelerated ageing, cancer andneurodegenerative diseases [50].

Blood Sampling and StorageBlood sampling is performed during the morn-ing (7:00 a.m.–8:30 a.m.) after overnight fast-ing. Before blood sampling, subjects remain inthe seated position for at least 10 min.

Over the study in those participating in thecohort study, approximately 150 ml of blood intotal is taken to determine the parameters out-lined above; in those also participating in theRCT (cohort study and RCT) overall approxi-mately 200 ml of blood is drawn. A potentialremainder of serum and plasma is stored pseu-donymised (patient number and visit numberonly) at -80 �C at the Biobank of the MedicalUniversity of Graz for potential future analyses.Therefore, subjects are asked to sign a separateinformed consent for the Biobank.

Faeces SamplingFaeces sampling is performed using stool col-lection tubes by Sarstedt, Numbrecht, Germany,with a stool collector (Susse Stuhlfanger,Gudensberg, Germany). One tube is used fornative stool, the other prefilled with stool DNA-stabilizer for collection of DNA-stabilized stoolspecimens. Directions for safe and hygienicfaecal collection are delivered to the subjects,explaining in words as well as in pictograms.Bacterial DNA are extracted from saliva andstool samples using the MagNA Pure LC DNAIsolation Kit III (Roche). The 16S rRNA gene isamplified in a PCR reaction, sequenced withnext-generation sequencing technology (RocheGenome Sequencer FLX or Illumina MiSeq) andinterpreted by the respective software analysers.

Continuous Glucose Monitoring (CGM)Minimally invasive CGM is performed using theAbbott FreeStyle Navigator II glucose-monitor-ing device (Abbott Diabetes care, Alameda, CA,USA) and the Medtronic Ipro2 (Minneapolis,USA). A sensor is introduced painlessly throughthe skin into the subcutaneous fat tissue. Then,a small transmitter is connected to the sensor.This construction is fixed to the skin with anadhesive tape for better adherence. The abdo-men and upper arm are the two main bodylocations for applying the sensors, which areworn for 6–8 days by the study participants.Capillary blood sugar measurement performedwith the commercially available Accu-Chek�

Performa glucometer (Mannheim, Germany) isconducted to collect reference values for therequired sensor calibration for the MedtronicSystem (at least all 12 h). The Abbott sensor isfactory calibrated and therefore does not requirecalibration by capillary prick blood. Howeversubjects are instructed to scan the sensor withthe corresponding recorder at least every 8 h.Each scan displays a real-time glucose result, ahistorical trend and the direction the glucose isheading indicated by an arrow.

Physical ExaminationPhysical examination is done by the trialphysician, who assesses the health status ofeach study participant (e.g. general appearance,

Adv Ther (2018) 35:1265–1283 1277

neck/thyroid, chest/lung., etc.) and evaluateseach section as normal/abnormal; if an abnor-mality is clinically significant he or she willdefine the findings in writing.

Anthropometric MeasurementsBody weight is measured with the patientstanding and then registered after rounding tothe nearest 500 g. Height is measured using ametric tape with the patient standing againstthe wall and the value marked by a ruler placedhorizontally to the head of the patient. BMI isestimated using the weight in kilograms dividedby the second power of the height expressed inmeters.

Waist circumference is measured by the samestudy nurse in the morning in the fasting state.According to the World Health Organization(WHO) guidelines, waist circumference is mea-sured at the midpoint between the top of theiliac crest (upper edge of the main pelvic bone)and lower margin of the last palpable rib in themid axillary line (lowest point of the rib cagethat can be located by touch along the side ofthe body). Hip circumference is measured in asimilar manner, with the tape being passedaround the hips at the widest circumference ofthe buttocks.

Vital SignsBlood pressure is taken using an automatedsphygmomanometer Boso Medicus Uno (Bosch& Sohn GmbH, Juningen, Germany) after a5-min rest. The blood pressure should be mea-sured in a sitting position, with legs uncrossed,the back and arms in an upright position. Sub-jects should not talk during the measurement.Pulse as beats per minute is recorded after rest-ing for 5 min in a sitting position by palpatingthe radial artery.

EchocardiographyTransthoracic echocardiography is performedwith a Vivid E9 (GE Healthcare, Chalfont StGiles, UK) at the Department of CardiologyMedical University of Graz, Graz, Austria.Examinations are performed by investigatorswith vast experience in echocardiography (EKor NV) who are blinded to individual

participant data, except name and birth date.Recorded images and loops are digitally savedfor later analyses. All measurements are per-formed by a single investigator (NV) underblinded conditions, and AS serves as supervisorfor data measurement. For acquisition of imagesin the parasternal view the participant is placedin the steep left-lateral decubitus position. Theparticipant’s left arm is raised above the head.Images are acquired during quiet respiration.The image rate is at least 30 frames per second.Target measurements include, among others,the left ventricular mass index, end-diastolicand end-systolic volume, left ventricular ejec-tion fraction, parameters of left ventriculardiastolic function and left atrial volume index.All data are assessed as recommended in inter-national guidelines [29, 51].

International Physical Activity Questionnaire(IPAQ)IPAQ is used in the current study. This tool wasdeveloped in 1998 for measuring physicalactivity and was followed by extensive reliabil-ity and validity testing undertaken across 12countries (14 sites) during 2000 [52, 53]. Thefinal results suggested that this tool hadacceptable measurement properties for use inmany settings and in different languages andwas suitable for national population-basedprevalence studies of participation in physicalactivity. The short version of IPAQ tool is usedin the current study. The tool includes a set offour questionnaires. The questions focus on thetime spent being physically active in the last7 days.

SF 36 Health QuestionnaireThe short form 36 (SF 36) health questionnairepresents the short version of a battery of 149health questions used in the RAND Corp. studyof health insurance in the US [54]. The aspectsof the questions regarding functional status,general wellbeing and overall evaluation ofhealth can be completed in less than 10 minwhile retaining the validity and reliability of thelonger questionnaire [55]. Because the study isperformed in Austria, the German version of thequestionnaire is used.

1278 Adv Ther (2018) 35:1265–1283

Food Frequency Questionnaire (FFQ)A self-administered, semi-quantitative FFQ wasdeveloped to assess usual food consumptionwithin the German Health Examination Surveyfor Adults 2008–2011 (DEGS) [56]. The relativevalidity of this questionnaire was studiedamong participants of another nationwide sur-vey, the German National Nutrition Monitoring(NEMONIT) [57]. The FFQ includes questionsabout the frequency and amount of 53 fooditems consumed during the past 4 weeks. Fre-quency of consumption of food behaviours isrequested according to specified categories. Inaddition, the respondents have to indicate theportion size of the food items consumed inpredefined answering categories. Pictures areused to aid the estimation of portion size for 33food items. In total, 29 food groups are pre-sented by the food frequency questionnaire.Because the study is performed in Austria, theGerman version of the questionnaire is used.

Data Analysis

Statistical analyses for primary and secondaryoutcomes as well as for baseline values are donein a descriptive and exploratory way. Primaryoutcomes are defined by various parameters,whereas each primary outcome has to be oper-ationalized for the statistical analysis to inves-tigate the influence of ADF, as the physiologicalchanges are represented in various parameters.

Insulin sensitivity is here stated as a repre-sentative example. Insulin sensitivity is calcu-lated according to four different equations(detailed information in the methods section)that are influenced by BMI and age, amongstothers. Insulin sensitivity is also related tophysical activity. The physiological relations arevery complex. To understand various effectsand multiple relations statistical multivariatemodels are built to have comprehensible sce-narios: GLMs (generalized linear mixed models)with age, BMI and physical activity as addi-tional variables with interactions are used to seesignificant and main covariates.

Further, for exploratory analysis continuousvariables are summarized with standarddescriptive statistics tables and represented

graphically with displays such as box plots andhistograms, and categorical variables are descri-bed by frequency tables and standard graphicaldisplays such as bar plots. Scatterplots are used toillustrate possible correlations between the vari-ables. Student’s t tests or Mann-Whitney U testsand chi square tests are used as appropriate.Comparisons within groups are performed usinga paired t test or Wilcoxon rank-sum test.p\0.05 is considered to indicate statistical sig-nificance, whereas appropriate correction formultiple testing such as the Bonferroni correc-tion or false discovery rate (Benjamini andHochberg) is applied. For high-dimensional‘‘omics’’ data appropriate multivariate modellingsuch as random forests and partial least squaresregression are used to elucidate differencesbetween groups concerning potential molecularmarkers regarding ageing effects. If necessary, astatistical analysis plan is prepared in addition tothe study protocol. Statistical software R [58] andSAS v.9.4 are used for statistical analysis.

STRENGTHS AND LIMITATIONSOF THIS STUDY

• This is the first prospective cohort study withan embedded randomized intervention trialthat will be conducted to investigate short-and mid- to long-term effects of ADF onhuman physiology and molecular cellularprocesses in healthy subjects.

• The methods for analysis of the outcomesare precise and reliable.

• Glucose levels are assessed using CGM.• In our study, subjects are only allowed to

consume water, flavoured carbonated water,tea or sugarless coffee on their fast days(complete caloric restriction) compared withprevious trials (partially reduction of dailycalorie intake).

• Besides routine laboratory, hormonal andmolecular measurements, we collect manyclinical data and biological biobank materialand perform a large set of investigationsincluding calorimetry, endothelial functionmeasurement and echocardiography, whichwill allow us look into the mechanisms ofADF and to generate hypothesis.

Adv Ther (2018) 35:1265–1283 1279

ETHICS AND DISSEMINATION

Ethics Approval and Consentto Participate

The study is performed according to the prin-ciples of the International Declaration of Hel-sinki and to the principles of good clinicalpractice (GCP) and was approved by the ethicscommittee of the Medical University of Graz(27–166 ex 14/15) on 27 February 2016. Thetrial is conducted at the Division ofEndocrinology and Diabetology at the MedicalUniversity of Graz (academic hospital), Austria.The protocol is registered in ClinicalTrials.govusing the following identifier: NCT02673515.

Important protocol modifications are repor-ted to the local ethics committee of the MedicalUniversity of Graz.

Participants are asked to sign informed con-sent before being enrolled into the trial, incompliance with the Declaration of Helsinkiand the WHO standards. Subjects are informedof the study objectives and the risks and benefitsof the examinations they will undergo. Allsubjects are given enough time to decide whe-ther they wish to participate in the trial.

The trial includes collection of biologicalsamples (blood, urine, stool, sputum), of whichsubjects are informed carefully. The samples arestored at -80 �C for use in future research at theBiobank of the Medical University of Graz.

Dissemination Plan

We anticipate that the results of this study will bedisseminated through peer-reviewed journalsand national and international academic con-ferences. Additionally, all trial participants willbe invited to an informational event, in whichstudy results will be presented by the study team.

DISCUSSION

Clinical research studies of ADF with robustdesigns and high levels of clinical evidence aresparse. The primary aim of this trial is to eluci-date whether and to what extent alternate day

fasting influences human physiology in healthyindividuals in both short- and mid- to long-term investigations. Besides routine laboratory,hormonal and molecular measurements, wecollect many clinical data as well as biologicalbiobank material and perform a large set ofinvestigations including calorimetry, endothe-lial function measurement and echocardiogra-phy, which will allow us look into mechanismsof ADF and generate hypotheses that can befurther tested in larger clinical trials.

Trial Status

The first patient was included on 8 April 2015. Thestudy is ongoing; close out is scheduled for April2019. Participants are no longer being recruited.

ACKNOWLEDGEMENTS

We are very grateful to our study participantsfor their cooperation.

Funding. The study is funded by the Austri-anScience Fund FWF (Austria) for grantsP29262-B28, P29203-B27 and P27893-B28(Frank Madeo). Frank Madeo acknowledgessupport from NAWI Graz and the BioTechMed-Graz flagship project ‘‘EPIAge’’. No funding orsponsorship was received for the publication ofthis article.

Authorship. All named authors meet theInternational Committee of Medical JournalEditors (ICMJE) criteria for authorship for thisarticle, take responsibility for the integrity ofthe work as a whole and have given theirapproval for this version to be published.

Authorship Contributions. Harald Sourij,Norbert J. Tripolt, Slaven Stekovic, SabrinaSchroder, Frank Madeo, Barbara Obermayer-Pi-etsch, Thomas R. Pieber designed the studyprotocol; Jasmin Url, Peter N. Pferschy and FelixAberer have made substantial contribution toacquisition of data. Norbert J. Tripolt and Har-ald Sourij wrote the clinical protocol andobtained authorization from the Ethics

1280 Adv Ther (2018) 35:1265–1283

Committee and wrote the manuscript. AlbrechtSchmidt, Ewald Kolesnik, Nicolas Verheyenperformed and evaluated the echocardiographicexaminations. Sophie H. Narath and ReginaRiedl are responsible for statistical analyses.Harald Sourij is the principal investigator of thetrial. Thomas R. Pieber and Frank Madeo are thechief investigators of this trial. All authors readand approved the final manuscript.

Disclosures. Norbert J. Tripolt, Slaven Ste-kovic, Felix Aberer, Jasmin Url, Peter N. Pfer-schy, Sabrina Schroder, Nicolas Verheyen,Albrecht Schmidt, Ewald Kolesnik, Sophie H.Narath, Regina Riedl, Barbara Obermayer-Pi-etsch, Thomas R. Pieber, Frank Madeo andHarald Sourij have nothing to disclose.

Compliance with Ethics Guidelines. Thetrial is conducted in accordance with the ethicalprinciples of the Declaration of Helsinki, Inter-national Council for Harmonisation GoodClinical Practice (GCP) and the applicablecountry-specific regulatory requirements.Informed consent was obtained from all indi-vidual participants included in the study.

Open Access. This article is distributedunder the terms of the Creative CommonsAttribution-NonCommercial 4.0 InternationalLicense (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommer-cial use, distribution, and reproduction in anymedium, provided you give appropriate creditto the original author(s) and the source, providea link to the Creative Commons license, andindicate if changes were made.

REFERENCES

1. Faris MA, Kacimi S, Al-Kurd RA, Fararjeh MA, Bus-tanji YK, Mohammad MK, et al. Intermittent fastingduring Ramadan attenuates proinflammatorycytokines and immune cells in healthy subjects.Nutr Res. 2012;32(12):947–55.

2. Joslin PM, Bell RK, Swoap SJ. Obese mice on a high-fat alternate-day fasting regimen lose weight andimprove glucose tolerance. J Anim Physiol AnimNutr (Berl). 2017;101(5):1036–1045.

3. Catenacci VA, Pan Z, Ostendorf D, Brannon S,Gozansky WS, Mattson MP, et al. A randomizedpilot study comparing zero-calorie alternate-dayfasting to daily caloric restriction in adults withobesity. Obesity (Silver Spring).2016;24(9):1874–83.

4. Varady KA, Bhutani S, Klempel MC, Kroeger CM,Trepanowski JF, Haus JM, et al. Alternate day fastingfor weight loss in normal weight and overweightsubjects: a randomized controlled trial. Nutr J.2013;12(1):146.

5. Varady KA, Hoddy KK, Kroeger CM, Trepanowski JF,Klempel MC, Barnosky A, et al. Determinants ofweight loss success with alternate day fasting. ObesRes Clin Pract. 2016;10(4):476–80.

6. Pedersen CR, Hagemann I, Bock T, Buschard K.Intermittent feeding and fasting reduces diabetesincidence in BB rats. Autoimmunity.1999;30(4):243–50.

7. Anson RM, Guo Z, de Cabo R, Iyun T, Rios M,Hagepanos A, et al. Intermittent fasting dissociatesbeneficial effects of dietary restriction on glucosemetabolism and neuronal resistance to injury fromcalorie intake. Proc Natl Acad Sci USA.2003;100(10):6216–20.

8. Wan R, Camandola S, Mattson MP. Intermittentfasting and dietary supplementation with 2-deoxy-D-glucose improve functional and metabolic car-diovascular risk factors in rats. FASEB J.2003;17(9):1133–4.

9. Mager DE, Wan R, Brown M, Cheng A, Wareski P,Abernethy DR, et al. Caloric restriction and inter-mittent fasting alter spectral measures of heart rateand blood pressure variability in rats. FASEB J.2006;20(6):631–7.

10. Krizova E, Simek V. Influence of intermittent fast-ing and high-fat diet on morphological changes ofthe digestive system and on changes of lipidmetabolism in the laboratory mouse. Physiol ResAcad Sci Bohemosl. 1996;45(2):145–51.

11. Hatori M, Vollmers C, Zarrinpar A, DiTacchio L,Bushong EA, Gill S, et al. Time-restricted feedingwithout reducing caloric intake prevents metabolicdiseases in mice fed a high-fat diet. Cell Metab.2012;15(6):848–60.

12. Duan W, Lee J, Guo Z, Mattson MP. Dietaryrestriction stimulates BDNF production in the brainand thereby protects neurons against excitotoxicinjury. J Mol Neurosci MN. 2001;16(1):1–12.

13. Duan W, Mattson MP. Dietary restriction and2-deoxyglucose administration improve behavioraloutcome and reduce degeneration of dopaminergic

Adv Ther (2018) 35:1265–1283 1281

neurons in models of Parkinson’s disease. J Neu-rosci Res. 1999;57(2):195–206.

14. Yu ZF, Mattson MP. Dietary restriction and2-deoxyglucose administration reduce focalischemic brain damage and improve behavioraloutcome: evidence for a preconditioning mecha-nism. J Neurosci Res. 1999;57(6):830–9.

15. Carlson AJ, Hoelzel F. Apparent prolongation of thelife span of rats by intermittent fasting. J Nutr.1946;31:363–75.

16. Goodrick CL, Ingram DK, Reynolds MA, FreemanJR, Cider N. Effects of intermittent feeding uponbody weight and lifespan in inbred mice: interac-tion of genotype and age. Mech Ageing Dev.1990;55(1):69–87.

17. Johnson JB, Laub DR, John S. The effect on healthof alternate day calorie restriction: eating less andmore than needed on alternate days prolongs life.Med Hypotheses. 2006;67(2):209–11.

18. Varady KA, Dam VT, Klempel MC, Horne M, CruzR, Kroeger CM, et al. Effects of weight loss via highfat vs. low fat alternate day fasting diets on freefatty acid profiles. Sci Rep. 2015;5:7561.

19. Teng NI, Shahar S, Manaf ZA, Das SK, Taha CS,Ngah WZ. Efficacy of fasting calorie restriction onquality of life among aging men. Physiol Behav.2011;104(5):1059–64.

20. Klempel MC, Kroeger CM, Varady KA. Alternate dayfasting (ADF) with a high-fat diet produces similarweight loss and cardio-protection as ADF with alow-fat diet. Metab Clin Exp. 2013;62(1):137–43.

21. Bhutani S, Klempel MC, Kroeger CM, TrepanowskiJF, Varady KA. Alternate day fasting and enduranceexercise combine to reduce body weight andfavorably alter plasma lipids in obese humans.Obesity. 2013;21(7):1370–9.

22. Belza A, Toubro S, Stender S, Astrup A. Effect of diet-induced energy deficit and body fat reduction onhigh-sensitive CRP and other inflammatory mark-ers in obese subjects. Int J Obes. 2009;33(4):456–64.

23. Eshghinia S, Mohammadzadeh F. The effects ofmodified alternate-day fasting diet on weight lossand CAD risk factors in overweight and obesewomen. J Diabet Metab Disord. 2013;12(1):4.

24. Kroeger CM, Klempel MC, Bhutani S, TrepanowskiJF, Tangney CC, Varady KA. Improvement incoronary heart disease risk factors during an inter-mittent fasting/calorie restriction regimen: rela-tionship to adipokine modulations. Nutr Metab.2012;9(1):98.

25. Heilbronn LK, Smith SR, Martin CK, Anton SD,Ravussin E. Alternate-day fasting in nonobese sub-jects: effects on body weight, body composition,and energy metabolism. Am J Clin Nutr.2005;81(1):69–73.

26. Harvie MN, Pegington M, Mattson MP, Frystyk J,Dillon B, Evans G, et al. The effects of intermittentor continuous energy restriction on weight loss andmetabolic disease risk markers: a randomized trialin young overweight women. Int J Obes.2011;35(5):714–27.

27. Harvie M, Wright C, Pegington M, McMullan D,Mitchell E, Martin B, et al. The effect of intermit-tent energy and carbohydrate restriction v. dailyenergy restriction on weight loss and metabolicdisease risk markers in overweight women. Br JNutr. 2013;110(8):1534–47.

28. Johnson JB, Summer W, Cutler RG, Martin B, HyunDH, Dixit VD, et al. Alternate day calorie restrictionimproves clinical findings and reduces markers ofoxidative stress and inflammation in overweightadults with moderate asthma. Free Radic Biol Med.2007;42(5):665–74.

29. Lang RM, Bierig M, Devereux RB, Flachskampf FA,Foster E, Pellikka PA, et al. Recommendations forchamber quantification. Eur J Echocardiogr.2006;7(2):79–108.

30. Varady KA, Bhutani S, Church EC, Klempel MC.Short-term modified alternate-day fasting: a noveldietary strategy for weight loss and cardioprotec-tion in obese adults. Am J Clin Nutr.2009;90(5):1138–43.

31. Horne BD, Muhlestein JB, Lappe DL, May HT, Car-lquist JF, Galenko O, et al. Randomized cross-overtrial of short-term water-only fasting: metabolicand cardiovascular consequences. Nutr Metab Car-diovasc Dis. 2013;23(11):1050–7.

32. Klempel MC, Kroeger CM, Varady KA. Alternate dayfasting increases LDL particle size independently ofdietary fat content in obese humans. Eur J ClinNutr. 2013;67(7):783–5.

33. Varady KA, Bhutani S, Klempel MC, Lamarche B.Improvements in LDL particle size and distributionby short-term alternate day modified fasting inobese adults. Br J Nutr. 2011;105(4):580–3.

34. Halberg N, Henriksen M, Soderhamn N, StallknechtB, Ploug T, Schjerling P, et al. Effect of intermittentfasting and refeeding on insulin action in healthymen. J Appl Physiol (1985). 2005;99(6):2128–36.

35. Seimon RV, Roekenes JA, Zibellini J, Zhu B, GibsonAA, Hills AP, et al. Do intermittent diets providephysiological benefits over continuous diets for

1282 Adv Ther (2018) 35:1265–1283

weight loss? A systematic review of clinical trials.Amsterdam: Molecular and Cellular Endocrinology;2015.

36. Matsuda M, DeFronzo RA. Insulin sensitivity indi-ces obtained from oral glucose tolerance testing:comparison with the euglycemic insulin clamp.Diabetes Care. 1999;22(9):1462–70.

37. Matthews DR, Hosker JP, Rudenski AS, Naylor BA,Treacher DF, Turner RC. Homeostasis model assess-ment: insulin resistance and beta-cell function fromfasting plasma glucose and insulin concentrations inman. Diabetologia. 1985;28(7):412–9.

38. Katz A, Nambi SS, Mather K, Baron AD, FollmannDA, Sullivan G, et al. Quantitative insulin sensitiv-ity check index: a simple, accurate method forassessing insulin sensitivity in humans. J ClinEndocrinol Metab. 2000;85(7):2402–10.

39. Stumvoll M, Van Haeften T, Fritsche A, Gerich J.Oral glucose tolerance test indexes for insulin sen-sitivity and secretion based on various availabilitiesof sampling times. Diabetes Care.2001;24(4):796–7.

40. Bonetti PO, Pumper GM, Higano ST, Holmes DR Jr,Kuvin JT, Lerman A. Noninvasive identification ofpatients with early coronary atherosclerosis byassessment of digital reactive hyperemia. J Am CollCardiol. 2004;44(11):2137–41.

41. Korhonen P, Syvanen K, Aarnio P. Surrogates oflarge artery versus small artery stiffness and ankle-brachial index. Int J Angiol. 2011;20(3):167–72.

42. Bonetti PO, Lerman LO, Lerman A. Endothelialdysfunction: a marker of atherosclerotic risk. Arte-rioscler Thromb Vasc Biol. 2003;23(2):168–75.

43. Jatoba LC, Grossmann U, Kunze C, Ottenbacher J,Stork W. Context-aware mobile health monitoring:evaluation of different pattern recognition methodsfor classification of physical activity. In: Conferenceproceedings: annual international conference ofthe IEEE engineering in medicine and biologysociety IEEE engineering in medicine and biologysociety conference, 2008;2008:5250–3

44. Ernst ME, Bergus GR. Noninvasive 24-hour ambu-latory blood pressure monitoring: overview oftechnology and clinical applications. Pharma-cotherapy. 2002;22(5):597–612.

45. Weir JB. New methods for calculating metabolicrate with special reference to protein metabolism.J Physiol. 1949;109(1–2):1–9.

46. Mathiowetz V, Weber K, Volland G, Kashman N.Reliability and validity of grip and pinch strengthevaluations. J Hand Surg. 1984;9(2):222–6.

47. White C, Dixon K, Samuel D, Stokes M. Handgripand quadriceps muscle endurance testing in youngadults. Springerplus. 2013;2:451.

48. Meneguetti DU, da Silva AF, Bosso R, Zan R, RamosL. New method for detection of mutagenicity inoral mucosa the through of micronucleus test.HOAJ Biol. 2012;1:8.

49. Holland N, Bolognesi C, Kirsch-Volders M, BonassiS, Zeiger E, Knasmueller S, et al. The micronucleusassay in human buccal cells as a tool for biomoni-toring DNA damage: the HUMN project perspectiveon current status and knowledge gaps. Mutat Res.2008;659(1–2):93–108.

50. Thomas P, Holland N, Bolognesi C, Kirsch-VoldersM, Bonassi S, Zeiger E, et al. Buccal micronucleuscytome assay. Nat Protoc. 2009;4(6):825–37.

51. Nagueh SF, Smiseth OA, Appleton CP, Byrd BF 3rd,Dokainish H, Edvardsen T, et al. Recommendationsfor the evaluation of left ventricular diastolic functionby echocardiography: an update from the americansociety of echocardiography and the European Asso-ciation of Cardiovascular Imaging. Eur Heart J Car-diovasc Imaging. 2016;17(12):1321–60.

52. Ainsworth BE, Bassett DR Jr, Strath SJ, Swartz AM,O’Brien WL, Thompson RW, et al. Comparison ofthree methods for measuring the time spent inphysical activity. Med Sci Sports Exerc. 2000;32(9Suppl):S457–64.

53. Craig CL, Marshall AL, Sjostrom M, Bauman AE,Booth ML, Ainsworth BE, et al. International physi-cal activity questionnaire: 12-country reliability andvalidity. Med Sci Sports Exerc. 2003;35(8):1381–95.

54. Ruta DA, Hurst NP, Kind P, Hunter M, Stubbings A.Measuring health status in British patients withrheumatoid arthritis: reliability, validity andresponsiveness of the short form 36-item healthsurvey (SF-36). Br J Rheumatol. 1998;37(4):425–36.

55. Ruta D,Garratt A, AbdallaM,Buckingham K,Russell I.The SF 36 health survey questionnaire. A valid mea-sure of health status. BMJ. 1993;307(6901):448–9.

56. Robert-Koch-Institut. DEGS - Studie zur GesundheitErwachsener in Deutschland. Berlin: Robert KochInstitut; 2009.

57. Haftenberger M, Heuer T, Heidemann C, Kube F,Krems C, Mensink GB. Relative validation of a foodfrequency questionnaire for national health andnutrition monitoring. Nutr J. 2010;9:36.

58. R Foundation for Statistical Computing. R: a lan-guage and environment for statistical computing.Vienna: R Foundation for Statistical Computing;2015.

Adv Ther (2018) 35:1265–1283 1283