A new Opportunity for Joint Health

SCIENTIFIC COMPENDIUM ON ROSE HIP POWDER

Index

03-06 PhytotheraPy research 20 (2006): Theevidenceforclinicalefficacyofrosehipandseed: asysTemaTicreview 06-13 scand. J rheumatol (2005): apowdermadefromseedsandshellsofarose-hip subspecies(rosacanina)reducessympTomsofknee andhiposTeoarThriTis: arandomized,double-blind,placebo-conTrolled clinicalTrial

13-21 PhytomedIcIne 11 (2004): aherbalremedy,hybenviTal(sTand.powderofa subspeciesofrosacaninafruiTs),reducespainand improvesgeneralwellbeinginpaTienTswiTh osTeoarThriTis—adouble-blind,placebo-conTrolled, randomisedTrial22-27 current theraPeutIc research Vol. 64 (2003): TheeffecTsofasTandardizedherbalremedymade fromasubTypeofrosacaninainpaTienTswiTh osTeoarThriTis:adouble-blind,randomized, placebo-conTrolledclinicalTrial

28-31 InflammoPharmacology Vol. 7 (1999): TheanTi-inflammaToryproperTiesofrose-hip 31-36 InflammoPharmacology Vol. 7 (1999): rosehipinhibiTschemoTaxisandchemiluminescenceof humanperipheralbloodneuTrophilsinviTroand reducescerTaininflammaToryparameTersinvivo

37 oVerVIeW on aBstracts

03

TheeVIdenceforclInIcal effIcacy ofrose hIPandseed: a systematIc reVIeWc. chrubasik1, r. K. duke2,3 and s. chrubasik1,3

1instituteofforensicmedicine,universityoffreiburgi.br.,albertstr.9,79104freiburgi.br.,germany2pharmaceuticalchemistry,facultyofpharmacy,universityofsydney,nsw2006,australia3herbalmedicinesresearchandeducationcentre,facultyofpharmacy,universityofsydney,nsw2006australia

Background: Theobjectiveofthisreviewistoevaluatewhetherclinicalresearchhasgainedanyevidenceofeffectivenessofrosacaninapreparations.

methods: severaldatabasesandothersourcesweresearchedtoidentifyrandomizedcontrolledtrialsofrosacaninapreparations.

results:Trialsweredescribedinanarrativeway,takingintoconsiderationmethodologicalqualityscores.fourtrialswereincludedinthisreviewandtwowereidentifiedassubgroupanalyses.

conclusion:moderateevidenceexistsfortheuseofapowderoftheseedsandhusksofarosacaninasubspeciesinpatientssufferingfromosteoarthritis.copyright©2006Johnwiley&sons,ltd.keywords:osteoarthritis;pain;rosehipandseed;nutraceutical.

study populationThe german commission e monograph (blu-menthal, 1998) summarizes the indications forrosehipandseedintraditionalmedicinewhichincludethepreventionandtreatmentofcoldsandinfluenza-likeinfections,infectiousdiseases,prophylaxisandtherapyofvitamincdeficienci-es,fever,forincreaseintheimmunemechanismduringgeneralexhaustion,gastricspasms,gas-tricaciddeficiency,preventionofinflammationofthegastricmucosaandgastriculcers,as‘sto-machtonic’,forintestinaldiseases,fordiarrhea,asprophylaxisof intestinalcatarrhs,asa laxa-tive, for gallstones, gall- and discomforts andailments,diseasesanddiscomfortsofthelowerurinarytract,dropsy,asa‘tonicforkidneys’,asadiuretic,forgout,disordersofuricacidmeta-bolism,arthritis,sciatica,diabetes,inadequateperipheralcirculation,asanastringent,forlungailments,andasaneyerinse.

The monograph stated that the effectivenessoftheherbformostofitsclaimedapplicationswasnotdocumented.investigationsinratsandrabbitsfailedtodemonstrateanincreaseddiu-resisandahypoglycaemiceffect,respectively(anon.,1998).however,apotentantioxidativeeffectwasseeninvitro(anon.,1998).

Theaimofthisstudywastoevaluatewhetherinthemeantimeclinicalresearchhasgainedanyevidenceofefficacyforrosehipandseed.

methodescomputerized literature searches were car-ried out by the authors (medline, pubmed,cochranecollaboraTionlibrary,embase(ovidtechnologies)backto1985andalsoma-nuallytoidentifyrandomizedcontrolledstudies(rcT)investigatingpreparationsofrosacanina(‘or’rosehip‘or’rosehip‘or’rosehipandseed,hagebutte (medline rosa ‘or’ fruit; drug ef-fects)).Thefollowingdatawereextractedfromeach study: authors’ names; date of publica-tion;countryoforigin; typeofstudy, includingnumber of study centres; participants (num-bers, disease(s), characteristics of the studypopulation(age,size,weight,gender));dura-tionofacuteexacerbationorchronicdisease;baselinevalueswithdetailsonpainandprevi-oustreatments;additionaltreatments;typesofoutcome measures; summary statistics; timingofoutcomeassessment;withdrawalsanddrop-outs; and adverse events. methodologicalquality and level of evidence were assessedasdescribed inaprevious review(gagnieretal., 2004): Quality items: (a) eligibility criteriaspecified, (b) randomization appropriate, (c)treatment allocation concealed, (e) similarityatbaseline,(f)outcomemeasuresandcontrolinterventions explicitly described, (g) co-inter-ventions comparable, (h) outcome measuresrelevant, (i) adverse events and (J) drop-outsfullydescribed,(k)samplesizebasedonaprioripowercalculation,(l) intention-to-treatanaly-

table 1. (a) eligibility criteria specified, (B) randomization appropriate, (c) treatment allocation conce-aled, (e) similarity at baseline, (f) outcome measures and control interventions explicitly described, (g) co-interventions comparable, (h) outcome measures relevant, (I) adverse events and (J) drop-outs fully described, (K) sample size based on a priori power calculation, (l) intention-to-treat analysis, (n) point estimates and measures of variability presented for the primary outcome measure, (o) appropriate timing giving a total score (ts) of 13

Phytomedicine2004; 11: 383–391

abcefghiJklnoTs

n=1125g/dayvsplacebo

cross-overover3months

oamultiplesitesyesyesyesyesyesyesyesyesyesyesnoyes11

80 (subgroup) 145 (subgroup)

osteoarthritis cartilage 2004; 12 suppl 2curr ther res clin exp

2003; 64: 21–31

n=1005g/dayvsplacebo

parallelover4months

hip,kneeyesyes

don‘tknowyesyesyesyesyesyesyesnoyes10

n=94(112)5g/dayvsplacebo

cross-overover3months

hip,kneeyesyesyesyesyesyesyesyesnoyesnoyes11

n=32(112)5g/dayvsplacebo

cross-overover3months

handyesyesyesyesyesyesyesyesnoyesnoyes11

resultsatotalof88(30pubmed,24medline),citationswerescreenedand4rcTsidentified(warholmetal.,2003;reinetal.,2004a,b;wintherandkharazmi, 2004),however, twowere identifiedas subgroupanalyses (reinetal., 2004a;win-therandkharazmi,2004).alltrialswerecarriedout with a powder of the seeds and husks ofarosacaninasubspecies inpatientssufferingfromosteoarthritis.afulldescriptionofthestu-dies isplacedon thewebpagehttp://remed-chrubasik.uniklinikfreiburg. de. The two mainstudieswereofhighquality(Ts10,11,Table1),but not confirmatory. relief of joint pain wasgreaterafter3and4monthsoftreatmentwith5gpowder/daycomparedwithplacebo, re-spectively(n=112,p<0.01;n=100,p<0.05).likewise,activitiesofdailylivingweremoreim-provedandconsumptionofrescuemedicationwassignificantlyless.

discussionour systematic review shows that clinical evi-denceofeffectivenesshasonlybeengainedin

thefieldofosteoarthritis.Thereisevidencethatnutritionalsupplementationwithadrypowderof a rosa canina subspecies may decreaseboth osteoarthritic pain and the consumptionof additional synthetic pain medications. Theproprietarypowderhasapotentantioxidativeeffect (daels-rakotoarison et al., 2002), inhi-bited chemotaxis and chemiluminescenceofhumanperipheralbloodneutrophils invitroandreducedcertaininflammatoryparametersin vivo (kharazmi and winther, 1999; wintheret al., 1999). a galactolipid contributes to theantiinflammatoryprinciple(larsenetal.,2003).painfularthritisisusuallytreatedwithnonsteroi-dalantiinflammatorydrugs(nsaids)(pincusetal.,2000),althoughtheiruseisoftenassociatedwith adverse gastrointestinal events that maybe life threatening in some patients (smalleyetal.,1995).Thecostofhealthcareresourcesspentonpreventingandmanagingtheseside-effectswascalculatedtobearoundoneca-nadiandollarforeverydayofnsaidtreatment(rahmeetal.,2001).safertherapiesarethere-forerequiredandhaveledtotheintroduction

sis,(n)pointestimatesandmeasuresofvariabi-litypresentedfortheprimaryoutcomemeasu-re,(o)appropriatetiminggivingaTotalscore(Ts)of13;levelsofevidencestrong–poolingofdata or at least 2 confirmatory studies (consi-deringitemskandn)demonstratingaclinical

relevanteffect,moderate–consistentfindingsamong one confirmatory study with a clinicalrelevant effect and/ or multiple exploratoryrcTs,insufficient–onelowqualityrcT,conflic-ting–inconsistentfindingsamongmultipletrials,noevidencefromtrials–norcTs

of selectivecox-2 inhibitors for the treatmentofchronicpain(graingerandcicuttini,2004).however,recently,rofecoxib(vioxx®)althoughassociatedwithastatisticallysignificantlylowerincidence of upper gastrointestinal bleedings(watsonetal.,2004)wasvoluntarilywithdrawnfrom the market due to increased risk of car-diovascular events (davies and Jamali, 2004).somenutraceuticalsmaybepromisingalterna-tivestosyntheticmedicationsinthetreatmentof musculoskeletal pain, although conclusivestudiesarerequiredforallofthem:proprietarypreparationsfromdevil’sclaw(gagnieretal.,2004),willowbark (chrubasiketal.,2000a,b),ginger(chrubasiketal.,2005),avocado-soybe-an(ernst,2003)andglucosaminesupplementedwithandwithoutsharkchondroitin(mcalindonetal.,2000).confirmatorystudies(consideringitemskandn)arealsorequiredforthepropri-etaryrosehipandseedpreparationinordertoprove the effectiveness beyond any doubt inthetreatmentofosteoarthritis.moreresearchisneededtoclarifytheothersupposedrosehipandseedeffects.

referencesanon.1998.rosacaninal.inhagerhand- buchderpharmazeutischenpraxis,5thedn. band4,hänselr,kellerk,rimplerh, schneiderg(eds).springer-verlag:berlin, heidelberg,newyork,3:447–452.blumenthalm.1998.Thecompletegerman commissionemonographs.Theamerican botanicalcouncil:austin,Texas,368–369.chrubasiks,eisenberge,balane,weinberger T,luzzatir,conradtc.2000a.Treatmentof lowbackpainexacerbationswithwillow barkextract:arandomizeddouble-blind study.amJmed109:19–14.chrubasiks,kunzelo,modela,conradtc, blacka.2000b.Treatmentoflowbackpain withaherbalorsyntheticanti-rheumatic: arandomizedcontrolledstudy.willowbark extractforlowbackpain.rheumatology 40:1388–1393.chrubasiks,pittlermh,roufogalisbd.2005. zingiberisrhizome.acomprehensivereview onthegingereffectandefficacyprofiles. phytomedicineinpressdaels-rakotoarisonda,gressierb,Trotinfetal. 2002.effectsofrosacaninafruitextracton neutrophilrespiratoryburst.phytotherres 16:157–161.daviesnm,Jamalif.2004.cox-2selective inhibitorscardiactoxicity:gettingtothe heartofthematter.Jpharmpharm sci7:332–336.

ernste.2003.avocado-soybeanunsaponifiab- les(asu)forosteoarthritis–asystematic review.clinrheumatol22:285–288.gagnierJJ,chrubasiks,manheimere.2004. harpagophytumprocumbensforosteo- arthritisandlowbackpain:asystematic review.bmccomplementalternmed4:13.graingerr,cicuttinifm.2004.medical managementofosteoarthritisoftheknee andhipjoints.medJaust180:232–236.kharazmia,wintherk.1999.rosehipinhibits chemotaxisandchemiluminescenceof humanperipheralbloodneurophils invitroandreducescertaininflammatory parametersinvivo.inflammopharma- cology7:377–386.larsene,kharazmia,christensenlp, christensensb.2003.anantiinflammatory galactolipidfromrosehip(rosacanina) thatinhibitschemotaxisofhuman peripheralbloodneutrophilsinvitro.Jnat prod66:994–995.mcalindonTe,lavalleymp,gulinJp,felsondT. 2000.glucosamineandchondroitinfor treatmentofosteoarthritis:asystematic qualityassessmentandmeta-analysis.Jam medassoc283:1469–1475.pincusT,swearingenc,cumminsp,callahan lf.2000.preferencefornonsteroidalanti-in flammatorydrugsversusacetaminophen andconcomitantuseofbothtypesof drugsinpatientswithosteoarthritis.Jrheu- matol27:1020–1027.rahmee,Josephl,kongsx,watsondJ, lelorierJ.2001.costofprescribednsaid- relatedgastrointestinaladverseeventsin elderlypatients.brJclinpharmacol52: 185–192.reine,kharazmia,wintherk.2004a.aherbal remedy,hybenvital(stand.powderofa subspeciesofrosacaninafruits),reduces painandimprovesgeneralwellbeingin patientswithosteoarthritis--adouble-blind, placebo-controlled,randomisedtrial. phytomedicine11:383–391.reine,kharazmia,Thamsborgg,wintherk. 2004b.aherbalremedymadefroma subspeciesofrose-hiprosacanina, reducessymptomsofkneeandhip osteoarthritis.osteoarthrcartil12(suppl2): 80.smalleywe,raywa,daughertyJr,griffinmr. 1995.nonsteroidalanti-inflammatorydrugs andtheincidenceofhospitalizationsfor pepticulcerdiseaseinelderlypatients. amJepidemiol141:539–545.

0504

warholmo,skaars,hedmane,molmenhm, eikl.2003.Theeffectsofastandardized herbalremedymadefromasubtypeof rosacaninainpatientswithosteoarthritis: adouble-blind,randomized,placebo- controlledclinicaltrial. currTherres64:21–31.watsondJ,yuQ,bologneseJa,reicinas, simonTJ.2004.Theuppergastrointestinal safetyofrofecoxibvs.nsaids:anupdated combinedanalysis.currmedresopin20: 1539–1548.

wintherk,kharazmia.2004.apowder preparedfromseedsandshellsofsubtype ofrose-hiprosacaninareducespain inpatientswithosteoarthritisofthehand –adoubleblind,placebo-controlledstudy. osteoarthrcartil12(suppl2):145.wintherk,reine,kharazmia.1999.Theanti- inflammatorypropertiesofrose-hip.inflam- mopharmacology7:63–68.

0706

aPoWdermadefromseedsandshellsofarose-hIP suBsPecIes(rosacanina)reduces symPtomsofKneeandhIP oste-oarthrItIs:arandomIzed, douBle-BlInd, PlaceBo-controlled clInIcal trIalK Winther1, K apel2, g thamsborg2

1departmentofclinicalbiochemistry,copenhagencountyhospitalgentofte,and2departmentofrheumatology,copenhagencountyhospitalglostrup,universityofcopenhagen,denmark

objective: Theaimofthisstudywastodeterminewhetheraherbalremedymadefromasubspe-ciesofrose-hip(rosacanina)mightreducesymptomsofosteoarthritisandconsumptionofrescuemedicationinpatientssufferingfromosteoarthritis.

methods: ninety-fourpatientswithosteoarthritisofthehiporkneewereenrolledinarandomized,placebocontrolled,double-blindcrossovertrial.forty-sevenpatientsweregiven5goftheherbalremedydailyforaperiodof3monthsandtheremainingpatientsweregivenasimilaramountofplacebo.Thegroupinitiallytreatedwithplacebowasthenchangedtorose-hipandviceversaforanother3-monthperiod.uponinclusionandafter3weeksand3monthsofeachtreatmentperi-od,pain,stiffness,disability,andglobalseverityofthediseasewerescoredonawesternontarioandmcmasteruniversities(womac)questionnaire.after3weeksoftreatment,patients,ifpossib-le,wereallowedtoreducetheirconsumptionof‘rescuemedication’.datawereanalysedonthebasisofintentiontotreat.

results:rose-hipresultedinasignificantreductioninwomacpain(p<0.014)ascomparedtoplacebo,whentestingafter3weeksoftreatment.Theconsumptionof‘rescuemedication’signi-ficantlydeclinedasaresultofactivetreatment(p<0.027).womacdisability,stiffness,andglobalassessmentofseverityofthediseasewerenotalteredby3weeksbutdecreasedsignificantly(p<0.018,p<0.038,andpv0.035,respectively)after3monthsoftreatment.

conclusion:Thedatasuggestthatthepresentherbalremedycanalleviatesymptomsofoste-oarthritisandreducetheconsumptionof‘rescuemedication’.

osteoarthritisisadiseasethatreachesyoungersportspersonsofbothsexes,manymiddle-agedpeople,andthemajorityoftheolderpopulati-on.ithasrecentlybeenclaimedthatlong-termtreatment with glucosamine sulfate can re-pairthedestroyedcartilage,whichisnormallythoughttobethemainelementofthedisease(1). however, most treatment is still directedagainstsymptomsofthedisease,suchaspainandstiffness,whichareresponsibleforthemainreductionindailyactivitiesoftenreportedinos-teoarthritis.

non-steroidalanti-inflammatorydrugs(nsaids),acetylsalicylic acid, and glucocorticoids areoften used for treatment of such symptoms,although treatments can result in serious sideeffectssuchasbleeding,gastricerosions,andliverandkidneydamage(2,3).cyclooxygena-se-2 inhibitors,which selectively inhibit theen-zymecyclooxygenase,havealsoexertedunfa-vourable effects (4) and the daily cost of thetreatmentisstillveryhigh.paracetamol,whichforadecadewasregardedasasafedrug,wasrecentlyreportedtoenhancetheriskofuppergastrointestinalproblems(5).forthesereasonstherehasbeena search fornewcompoundsthat could minimize pain and stiffness withouttheserioussideeffectsmentionedabove.vari-ousherbalremedies,especiallyextractsofgin-ger and avo cado/soybean unsaponifiables,have shown promising results in patients withosteoarthritis(6,7).morefocusonremediesofa herbal origin might therefore, in the future,change the treatment of patients with oste-oarthritisbyaconsumptionpatternwithfewersideeffects.

inflammatorycellssuchaspolymorphonuclea-tedleucocytesparticipateininflammationandtissuedamagebyliberatingproteolyticandhy-drophilic enzymes as well as oxygen radicals.wehave found thata standardizeddrypow-dermade from seedsand shellsofa subtypeofrose-hip(rosacanina)reducesthemigrationrateofpolymorphonucleatedleucocytesinvit-roandtheserumconcentrationofc-reactiveprotein in humans (8), an effect unrelated tothehighvitaminccontentofrose-hip(9).

moreover,someoftheosteoarthriticvolunteerswho participated in these preliminary studiesclaimedthattheirpainsymptomsweredrama-ticallyreducedafteraperiodoftreatment(8).This encouraged us to investigate whether astandardizedpowdermadefromthesamewildtypeofrose-hip(rosacanina)wouldalleviatesymptomssuchaspainandstiffnessandimpro-

vedailyfunctionsinosteoarthriticpatients.wealsowantedtoevaluatewhetheraneffect, ifpresent, was of sufficient magnitude to influ-ence the daily consumption of pain relievingmedicine.

Patients and methodsstudy populationpatients were recruited from the outpatientclinicsof thedepartmentofrheumatologyofcopenhagenuniversityhospitalinglostrupandof the instituteforclinicalresearch.Thestudywasapprovedbytheethicscommitteeofvejleandcopenhagencounties(no.9980042pmc).patientswererecruitedafterannouncementsinlocalnewspapers.Theprimaryinclusioncriteriawereageover35yearsandsymptomatickneeor hip osteoarthritis. osteoarthritis of the kneeorhipwasdiagnosedaccordingtotheclinicaland radiologicalcriteriaof theamericancol-legeofrheumatology(10,11).majorexclusioncriteria were inflammatory arthritis, fibromyal-gia, depression, and substantial abnormalitiesinhaematological,hepatic,renal,ormetabolicfunctions. furthermore, we excluded patientswhoreceivedglucosaminesulfate,chondroitinsulfate, intra-articular hyaluronate, or systemicorintra-articularglucocorticoidsinthe6weeksprecedingenrolment.

design and treatmentThestudywasarandomized,double-blind,pla-cebocontrolled, crossover trial with three suc-cessiveperiods:a14-dayrun-inperiodandtwosubsequenttreatmentperiodsof3months.af-tertherun-inperiod,patientswereallocatedtoreceiveactivemedicationandplaceboinran-domorderinthetwotreatmentperiods(figure1).allocationwascarriedoutinblocksoffourbyacomputerprogram.activemedicationcom-prised biologically standardized rosehip pow-der(litozin).allcapsuleswereproducedfromthesamebatch.identicalcapsulescontainingan inactivepowderofsimilartaste,smell,andcolourwereproducedforplacebo.Thedosa-ge was a total of 5 g of rose-hip powder ad-ministereddailyasfivecapsuleseachof0.5gof the rose-hiporplacebo, tobetaken in themorningandagain in theeveningalongwithameal.compliancewithstudytreatmentwasestablishedbyaskingthepatientaboutmisseddosesandbycountingthenumberofreturnedcapsules.

The rose-hip powder used has been on themarketasaherbalremedyinthescandinaviancountries for almost a decade. it is producedfromfruitsofaselectedsubtypeofrosacani-

na.Theplantsarealwaysgrowninstandardizedfieldsaccording togoodagriculturalpracticeandharvestingtakesplaceonlywhenthefruitsare mature. immediately after harvesting, thefruitsarefrozen.whenthefruitsarethawedla-teron,aspeciallasertechniqueisusedtoen-sureoptimalfruitsfortheproductionofpowder.a computerized technique ensures that thedryingprocessneverexceeds40°candthedrypowder,whichcontainselementsoftheseedsaswellastheshellsoftherose-hip,iscontrolledregardingvitaminandmineralcontent.patientsusing nsaids regularly were advised to conti-nue using the same dosage during the entirestudy.however,patientswereadvisedtoredu-ceintakeofotheranalgesicsifpossible,suchasparacetamolorsyntheticopioidsafterthefirst3weeksofeachofthetwotreatmentperiods.duringthestudyperiod,thepatientswere ins-tructednottochangetoanothergenerictypeof thesameanalgesicor tousesimilar tabletscontaining a different quantity of the samepainkiller.neitherwaspatientsallowedtostartupanynewtypeofpainrelievingmedication.The consumption of analgesics was recordeddailybythepatients inadiary.Thechangeinconsumptionofanalgesics,ineachofthetwotreatment periods, was estimated by subtrac-tingtheconsumptionofmedicationinthepast2weeksfromthatoftheinitial2weeks.noothercointerventions for osteoarthritis were allowedduringtheentirestudyperiod.

outcome measuressymptoms of osteoarthritis were assessed bythewesternontarioandmcmasteruniversities(womac)osteoarthritis index,avalidated,di-sease-specific questionnaire addressing seve-rity of joint pain (five questions), stiffness (twoquestions), limitation of physical function (17questions), and patients’ global assessmentofdiseaseseverity referringtothe48hbeforeassessment (11). The visual analogue scaleversionof the indexwasused, that iswith thepatientassessingeachquestionbya100mmvisualanaloguescale.ahigherwomacscorerepresents worse symptom severity, with 2500mm being the worst possible total score (12).womac scores were assessed at the begin-ning,after3weeks,andattheendofeachofthe two treatment periods. womac score ofjoint pain was the primary outcome measuretogetherwiththeconsumptionofanalgesicsta-kenduringthetwodifferenttreatmentperiods.womac scores of stiffness, limitation of physi-cal function, and patients’ global assessmentofdiseaseseverityandoccurrenceofadverseeventsweresecondaryoutcomemeasures.

randomIsatIon(n=94)

1 Personal reasons1 started Prednisolon1 difficulty to swallov capsules

1 Personal reasons1 did not feel that treatment worked

1 Personal reasons1 acid regurgitation

2 Personal reasons1 acid regurgitation1 Insisted on knowing sequense of treatment

1 Intercurrent surgery

1 Personal reason 1 With drew as treatment did not work any longer

actIVe treatment(n=47)

PlaceBo treatment(n=47)

3 W

ee

ks

3 W

ee

ks

3 m

ont

h

3 m

ont

h

3 W

ee

ks

3 W

ee

ks

3 m

ont

h

3 m

ont

h

all withdrawals are given in the boxes

figure 1. flow diagram.

statistical analysisbasedonawithin-patientssdof10%,wecal-culated thata sample sizeof90patients inacrossoverdesignwouldgiveapowerof90%indetecting more than a 15% difference in thewomacscoreof jointpainat the5% levelofsignificance. statistical analysis was based onthe intention-to-treat principle with last obser-vation carried forward. The wilcoxon test formatchedpairswasused throughout. subana-lysiscomparingparallelgroupswasperformedusingthemann–whitneytest.dataaregivenasmeanvalues±sd.

resultsPatientsa total of 94 patients, comprising 54 women(meanage66years;range38–92)and40men(meanage65years;range48–85)wereenrol-led in the study and randomized to either re-ceiveplacebofirstandthenactivetreatment(groupa,n=547)oractive treatment firstandthenplacebo(groupb,n=547).Therewerenosignificantdifferencesingenderorageoncom-

paringtheaandbgroups(datanotgiven).inthe entire group the mean body mass index(bmi) was 27 kg/m2 (range 19–41). in group athe bmi was 27.3 kg/m2 (range 19–39) and ingroupb,26.6kg/m2(range22–41),anon-signi-ficantdifference.ingroupa13ofthepatientsweretakingnsaids,18paracetamol,10synthe-ticopioidssuchastramadolandcodeine,and19didnotuseanyrescuemedicationatall.ingroup b the corresponding numbers of pati-entswere:nsaid15,paracetamol21,synthetic

opioids 6, and no medication at all 17. Thesevalueswerenotsignificantlydifferentfromthevaluesreportedingroupa.Therewasnosignifi-cantdifferenceinthenumberofpatientsdrop-pingoutofthestudywhencomparingthetwodifferenttreatmentsortheaandbgroups(fordetails seefigure1).Therewerenosignificantdifferencesinosteoarthriticcharacterizationoncomparingtheaandbgroups,asdetailed inTable1.compliancewas92.5%withhyben-vi-taland90.5%withplacebo.

table 1. characterization of osteoarthritis.

all patients (n=594)

kneeosteoarthritishiposteoarthritishipandkneeosteoarthritis

Initial Womac scorespainstiffnessadlpgad

582115

33.7(19.4)39.2(19.4)35.3(21.6)43.9(24.4)

Placebo–active (n=547)

29117

30.4(18.1)35.6(22.0)34.0(21.1)43.6(22.6)

active–Placebo (n=547)

29108

37.0(20.4)42.5(26.2)36.7(22.2)44.3(26.8)

p-value Pa vs. aP

29108

37.0(20.4)42.5(26.2)36.7(22.2)44.3(26.8)

table 2. Womac scores for pain, stiffness, daily activities (adl), and patients’ evaluation of disease severity (Pgad) in all the included patients (n594). data given are mean values with sd in parentheses.

start

pain

stiffness

adl

pgad

33.7(19.4)33.7(19.4)39.2(24.4)39.2(24.4)35.3(21.6)35.3(21.6)43.9(24.4)43.9(24.4)

3 weeks

35.3(21.5)29.4(18.3)40.0(24.2)34.0(20.5)39.7(25.3)35.9(27.7)42.3(21.2)39.2(22.4)

delta value

2.1(16.8)7.4(14.9)2

3.3(19.0)7.5(16.7)4

-20.7(221.4)2.2(22.7)6

8.2(25.1)8

8.2(22.6)10

placeboactiveplaceboactiveplaceboactiveplaceboactive

3 months

2.1(16.8)7.4(14.9)2

3.3(19.0)7.5(16.7)4

-20.7(221.4)2.2(22.7)6

8.2(25.1)8

8.2(22.6)10

delta value

5.1(18.3)1

7.0(19.7)3

5.0(23.2)8.0(21.6)5

-20.2(25.5)6.4(17.5)7

7.8(28.8)9

14.1(28.1)11

3 weeks

0.014

0.198

0.165

0.682

3 months

0.125

0.038

0.018

0.035

p-value placebo vs. active

1pv0.005, 2pv0.001, 3pv0.003, 4pv0.001, 5pv0.006, 6pv0.002, 7pv0.003, 8pv0.004, 9pv0.031, 10pv0.002, 11pv0.001. the p-values givenare relative to pretreatment values (initial values).

Primary outcome measurewomacscoresforjointpain,fortheentirestudypopulation,aregiveninTable2.after3weeksofactivetreatment,womacscoresforjointpaindeclined from 33.7±19.4 to 29.4±18.3, a deltareductionof7.4±14.9mm(pv0.001),comparedtoachangefrom33.7±19.4to35.3±21.5,adel-tachangeof2.1±16.8(pv0.299),whenplacebotreatmentwasgiven(Table2).Thechangecom-paringthetwodifferentgroupswasstatisticallysignificantatthep<0.014 level.after3monthsoftreatment,thesamepatternwasobserved,althoughthechangeswerenotstatisticallysig-nificant (p<0.125). Thepercentageofpatientsexperiencingareductioninthewomacscorefor joint pain after the initial 3 weeks of treat-

mentwassignificantlyhigherwhenactivetreat-mentwasgiven(82%)thanwhenplacebowasgiven(49%)(p<0.004)(figure2).after3monthsoftreatment,thepercentagesofrespondersinthetwogroups,althoughstillinfavourofactivetreatment,didnotdiffersignificantly.

diariesoftheconsumptionof‘rescuemedica-tion’ indicated, in accordance with the studydesign,thattheintakeofnsaidswasunchan-gedduringthetwodifferenttreatmentperiods(p<0.803)(datanotgiven).adeclineof40%intheconsumptionofparacetamol(dataavaila-ble in21patients)wasobservedasaresultofactivetreatment(p<0.052).

0908

1110

figure 2. Percentage of patients experiencing a reduction in the Womac score for joint pain after 3 weeks of treatment in the group initially given placebo and in the group initially given activetreatment.

whenamann–whitneysubanalysiswasappliedontheconsumptionofparacetamolineachofthetwogroups,duringthefirst3monthoftreat-ment, rosehip powder resulted in a significantreductioninthenumberoftabletstakenduringa2-weekperiod(14.0±24.0;p<0.031)comparedto an insignificant increase of 7.9±15.5 tabletsobservedasaresultofplacebotreatment.Thebetween-groupdifferencewas51%(p<0.027).Theconsumptionofweakopioids(dataavaila-bleinonlysevenpatients)showedasimilarre-ductionintheconsumptionduringactivetreat-ment(p<0.0313)(datanotgiven).asrelativelyfewerpatientsweretakingweakopioids,asua-nalysiswasnotperformedonweakopioids.

49 % 82 %

50 –

100 –

0

(p<0.035) when active treatment was given,ascomparedtoplacebo.Thewerenosignifi-cantdifference inthealleviationofsymptomscomparing patients with osteoarthritis of thehip to patients with osteoarthritis of the knee.asacarry-overeffectcanblunttheimpactoftreatmentinacrossoverdesign,wealsoanaly-sed,separately,thegroupinitiallytreatedwithplacebo and then actively treated (group a)and the group initially given active treatmentandthenplacebo(groupb).groupashowedasignificantimprovementinactivitiesofdailyli-ving(adl)functionandareductioninpatients’overall feelingofdiscomfortfromtheirdiseasepatients global assessment of disease severity(pgad)after3weeksand3monthsofactivetreatment.Theimpactonpainandstiffness,alt-houghpresent,didnotattain statistical signifi-cance(Table3).

patients in group b showed a statistically sig-nificant reduction inpain, stiffness,andpgadasaresultofactivetreatment.Thesechanges,however,didnot returntopretreatment levelsduringthefollowingplacebotreatmentperiod,suggestingcarryover(Table3).acomparisonoftheaandbgroupsregardingpainandstiffnessyielded mann–whitney p-values of 0.001 and0.016, respectively, when evaluating after theinitial3weeksoftreatment.althoughthiscom-parisonbetweengroupswasstillalsoinfavourofactivetreatmentafterthefirst3monthsoftreat-ment,statisticalsignificancewasnotobtainedand further statistically significant changes inwomacparameterswerenotobservedwhencomparing the initial 3-month periods of thetwo different treatments. an identical patternas described for womac data was also ob-servedforrescuemedication(datanotgiven).Therewasno significantdifference indropoutrateormilderunwantedsideeffects reportedduringtreatment(Table4).

secondary outcome measureswomacscoresforstiffness, limitationofphysi-cal function, and patients’ global assessmentof disease severity for the entire study popu-lation are given in Table 2. after 3 months oftreatment,therewasasignificantreduction inwomacsymptomscoresforstiffness(p<0.037),womacscores for limitationofphysical func-tion improved (p<0.018), whereas patients’globalassessmentofdiseaseseveritydeclined

table 3. Womac scores for pain, stiffness, daily activity (adl), and patients’ evaluation of disease severity (Pgad) in group a (placebo first, then active treatment) and in group B (active treatment first, then placebo). data given are mean values with sd in parentheses.

Initial value

groupapainstiffnessadlpgad

group Bpainstiffnessadlpgad

(n=547)30.4(18.1)35.6(22.0)34.0(21.1)43.6(22.6)

(n=547)37.0(20.4)42.5(26.2)36.7(22.2)44.3(26.8)

3 weeks

34.5(23.1)37.1(25.9)36.1(22.3)40.2(22.3)

28.9(19.0)36.2(21.7)38.0(34.2)37.6(23.3)

delta value

-22.5(13.6)-21.4(19.3)-20.3(10.5)

8.1(23.5)

9.6(13.9)9

8.9(14.4)13

-20.4(28.2)6.8(19.4)18

3 months

36.3(20.4)38.0(23.6)38.3(20.3)

48.9(25.5)6

33.8(17.6)39.8(21.6)37.0(18.1)44.4(39.3)

delta value

2.3(14.9)3.2(22.8)2.5(14.7)5.2(29.6)

8.1(17.4)10

9.2(21.4)14

5.3(15.0)(17)

12.6(28.0)19

3 weeks

29.9(17.7)31.4(19.0)33.5(17.6)41.1(21.5)

36.0(20.0)42.8(22.4)43.6(27.9)44.5(20.2)

delta value

5.1(15.6)1

5.9(19.2)(2)

5.3(13.8)4

9.6(25.5)7

7.0(18.5)11

7.8(18.0)15

-21.2(29.1)9.3(27.4)

3 months

31.9(23.4)33.8(25.5)33.0(23.0)38.1(22.9)

34.9(20.6)44.0(24.5)45.3(32.7)41.5(19.3)

delta value

5.9(21.9)6.8(21.9)(3)

7.6(19.9)5

15.3(28.6)8

7.8(20.9)127.8(18.0)15

-21.2(29.1)9.3(27.4)

1p<0.042, 2p<0.076, 3p<0.095, 4p<0.002, 5p<0.025, 6p<0.018, 7pv0.022, 8p<0.001, 9p<0.001, 10p<0.011, 11p<0.031, 12p<0.012, 13p<0.001, 14p<0.022, 15p<0.037, 16p<0.084, 17p<0.068, 18p<0.044, 19p<0.010, 20p<0.049. P-values given are relative to pretreat-ment values. P-values given in parantheses indicate borderline significance.

table 4. dropout rate and unwanted effects in patients after 3 months while on placebo or ac-tive treatment.

Pla-cebo

Dropped out during treatment

Reasons for dropoutfeltthattreatmentdidnotworkforpersonalreasonsacidregurgitationdifficultytoswallowcapsulesstartedprednisolonetreatmentintercurrentsurgeryinsistedonknowingkindoftreatment

Milder unwanted effects reported duringtreatment that did not cause withdrawalfrequentvoidingdiarrhoeaconstipationshortepisodeofmildurticaria

ac-tive

p-value

7

2310001

1210

7

0311110

3221

ns

nsnsnsnsnsnsns

nsnsnsns

discussionThis study shows that a standardized rose-hippowder,madefromasubtypeofrosacanina,hasabeneficialsymptomaticeffectinpatientswith knee and hip osteoarthritis. The percen-tage of patients who reported at least somereductioninwomacpainafter3weeksofac-tivetreatmentwas82%comparedtoa49%re-duction in thegroup treatedwithplacebo.aplaceboeffectof thesamemagnitudeas re-portedherewasalsoreportedinarecentstudyevaluating the impact of a ginger extract onpainfromosteoarthritisoftheknee(6). inthatstudy,whichreporteda50%reductioninpainduring placebo compared to a 66% reduc-tionduringactivetreatment,earlytestingwasalso performed. The placebo impact, in bothstudies,mighthavedeclinedifthestudieshadbeenrunningforalongerperiodoftime.

we used a validated, disease-specific, andvery sensitive questionnaire and were able todemonstrateareductioninjointpainandstiff-nessaswellasan improvedphysical functionin thesepatientsafter treatmentwith thepre-sentrose-hippowder.painthatwassignificant-ly reducedafter3weeksoftreatmentdidnotattainstatisticalsignificancewhentestedafter3 months. during the course of the 3-monthtreatment period in which the patients recei-ved active treatment, there was, however, asignificantreductionintheconsumptionoftra-ditional painkillers such as paracetamol andsyntheticopioidsascomparedtothegroupre-

ceivingplacebo.wesuggestthatthischangeinconsumptionofadditionalpainkillers,whichpatientswereallowedtoreduceafterthefirst3weeksoftreatment,mayexplainthelackofsignificancewhenpainwasevaluatedafter3monthsoftreatment.furthermore,thepowderwaswelltoleratedanddidnotgiverisetoanyseriousadverseeffects;infact,stiffnessandglo-balassessmentofdisease severity significantlydeclinedanddailyactivitiessignificantlyimpro-vedafter3monthsofactivetreatment.ourre-sultsaresupportedbythefindings ina recentnorwegianstudyinwhichtreatmentwithpow-derfromthesamesubtypeofrosehipresultedin improvedjointmobilityandless jointpain inpatientsonawaiting list foreitherhiporkneesurgeryduetoosteoarthritis(13).

Thereare,however,reservationstoourconclu-sion.Thedosewaspossiblynotoptimalandalongterm study is needed to confirm that thereduction in symptoms is persistent, and thatlong-termtreatmentdoesnot result in sideef-fects different from what was observed withplacebo.

Thepresentdata,however,seemtofitwellintoearlier,morebasic,reportsfromourlaboratoryindicating that the present version of rose-hippowder, when used in higher doses, reducespain in osteoarthritis and affects mechanismsof importance to joint disease (8, 9). it is alsoencouraging tonote that inanother studyai-mingtotestpatientswithosteoarthritis,onthewaitinglistforhiporkneereplacement,thepre-sentpowder,given ina similardose, reducedpain and improved mobility, suggesting thatthepowdermayworkinboththeearlyandatestagesofosteoarthritis(13).

when responders to treatment were askedaboutthetimebeforesomealleviationofpainoccurred, the earliest response reported waswithin2weeks.moreover,acertaincarry-overeffectwasdemonstratedinthepresentstudy,andcarry-overwasalsodemonstratedinano-therstudyusingthesamerosehippowderandanidenticalstudydesign(14).Thismayindicatethatthepresentpowderdoesnotworklikethetraditionalpainkillersnormallyused in the treament of osteoarthritis. as reported earlier, onemodeofactionmightbeananti-inflammatoryaction mediated by leucocyte neutrophils (8,9). indeed, we were able to show that c-reative protein and also the chemotaxis of neutr

1312

philleucocytesweredecreasedinvitroaswellas invivo,usingconcentrationsof thepresentsubtype of rose-hip, comparable to the doseused in this study (8, 9). chemotaxis of leuco-cyte neutrophils also significantly declinedwhenmeasuredinasubfractionofthepresentpatients(15). itseemslikelythereforethatonemechanismofthepresentpowdermightbeofan anti-inflammatory origin. indeed, the anti-inflammatoryhypothesisseemstobereceivingincreased attention. we have shown that theanti-inflammatory impact of the present sub-type of rose-hip was not related to vitamin candsuggestedthatanotherpossiblyunknownactiveingredientmightbefoundintherosehippowder(8).anactiveingredientthatcaninhibitthechemotaxisofhumanneutrophilleucocyteshasbeenisolatedrecentlyfromthepresentsub-typeofrose-hip,makingtheanti-inflammatoryhypothesismore likely(16).aframinghamstu-dyand,morerecently,adanishstudyindicatethat patients with osteoarthritis might benefitfromvitaminc(17,18).asthepresentpowderis rich in natural vitamin c, an additional me-chanismmightbeofvitamincorigin.

The powder does not seem to be involved inthearachidonicacidpathwayasplateletag-gregationwasnotaffectedwhenthepowderwas tested in healthy volunteers and patientsonwarfarintreatment(19).Thisisdifferentfromtheanti-inflammatoryagentsreferredtointheintroductionofthispaperandwesuggestthatthismighthelptoexplainwhyside effects in this study were comparable tothatoftheplacebo.

insummary,wesuggestthatthepresentstan-dardizedpowder,madefromasubtypeofrosacanina,canalleviatepaintoanextentthatcaninfluence theconsumptionof rescuemedica-tion. itshouldbeemphasizedthatthepresentdatamaynotapplytoanytypeofrose-hip,asspecies can be different regarding biologicalactivity(20).furtherresearchshouldaimtofindtheoptimaldose,testtheimpactoflong-termtreatmentandcompare thatwith the impactofnsaids,andevaluatethebiologicalactivityofdifferentsubtypesofrosehip.

acknowledgements

we are grateful to hyben-vital international,langeland, denmark for providing capsulescontainingplaceboandrosehippowder.

references1. reginsterJy,deroisyl,rovatilc,leerl, lejeunee,bruyereo,etal.long-term effectofglucosaminesulphateonoste- oarthritisprogression:arandomised, placebo-controlledclinicaltrial.lancet 2001;357:251–6.2. hochbergmc,altmanrd,brandtkd, clarkbm,dieppepa,griffinmr,etal. guidelinesforthemanagementof osteoarthrosis(parts1and2).arthritis rheum1985;38:1535–46.3. vaneJr,bottingrm.anti-inflammatory drugsandtheirmechanismofaction. inflammres1998;47:578–87.4. mukherjeed,nissense,TopoleJ.riskof cardiovasculareventsassociatedwith selectivecox-2inhibitors.Jammedassoc 2001;286:954–9.5. rodriqueslac,hernandes-diazs.relative riskofuppergastrointestinalcomplications amongusersofacetaminophenandnon- steroidalanti-inflammatorydrugs.epide- miology2001;12:570–6.6. altmanrd,markussenkc.effectofginger extractonkneepaininpatientswithoste- oarthritis.arthritisrheum2001;44:2531–8.7. appelboomT,schuermansJ,verbruggen g,henrotiny,reginsterJy.symptoms modifyingeffectofavocado/soybean unsaponifiables(asu)inkneeosteoarth- ritis.scandJrheumatol2001;30:242–7.8. wintherk,reine,kharazmia.Theanti-in flammatorypropertiesofrose-hip.inflammo pharmacology1999;7:63–8.9. kharazmia,wintherk.rose-hipinhibitsche- motaxisandchemiluminescenceofhuman bloodneutrophilsinvitroandreduces certaininflammatoryparametersinvivo. inflammopharmacology1999;7:377–86.10.altmanr,asche,blochd,borensteind, brandtk,etal.developmentofcriteria fortheclassificationandreportingof osteoarthritis.classificationofosteoarthritis oftheknee.arthritisrheum 1986;29:1039–49.11.altmanr,alarcong,appelrouthd.The americancollegeofrheumatologycriteria fortheclassificationandreportingof osteoarthritisofthehip.arthritisrheum 1991;34:505–14.12.bellamyn,buchananwb,goldschmidtch, campbellJ,stittlw.validationstudyof womac:ahealthstatusinstrumentfor measuringclinicallyimportantpatient relevantoutcomestoantirheumaticdrug therapyinpatientswithosteoarthritis

ofthehipandknee.Jrheumatol 1988;15:1833–40.13.warholmo,skaars,hedmane,mølmen hm,eikl.Theeffectsofastandardised herbalremedymadefromasubtypeof rosacaninainpatientswithosteoarthritis: adouble-blind,randomised,placebo-con- trolledclinicaltrial.currTherapres 2003;64:21–31.14.reine,kharazmia,wintherk.aherbal remedy,hybenvital,reducespainand improvesgeneralwellbeinginpatients withosteoarthritis–adouble-blind, placebo-controlled,randomisedtrial. phytomedicine2004;11:383–91.15.Thamsborggm,apelk,reine,wintherk.a randomized,doubleblind,placebo- controlled,crossovertrialoftheherbal remedyhybenvitalinpatientswith osteoarthritis.Thu0214eular,stockholm, sweden,10–15June2002.16.larsene,kharazmia,christensenlp, christensensb.anantiinflammatory galactolipidfromrose-hip(rosacanina)

thatinhibitschemotaxisofhumanblood neutrophilsinvitro. Jnatprod2003;66:994–5.17.mcalindonTe,Jacquesp,zhangy,hannan mT,aliabadip,weissmannb,etal.do anti-oxidantmicronutrientsprotectagainst thedevelopmentandprogressionofknee osteoarthrosis?arthritisrheum 1996;39:648–56.18.Jensennh.reducedpainfrom osteoarthritisinhipjointorkneejointduring treatmentwithcalciumascorbate.aran- domised,placebo-controlled,crossover trial.ugeskrlaeger2003;165:2563–6.19.reine,kharazmia,wintherk.rose-hip givenasastandardiseddrypowderexerts anti-inflammatoryproperties,withoutinflu- encingplateletaggregationandthecoa- gulationcascade.1stinternational congressonheartdisease,washington, may1999.20.brandtk,aakesonb.healthpromoting compoundsinvegetablesandfruit.plant prod2002;29:43–44.

aherBal remedy, hyBen VItal (stand. PoWderofasuBsPecIes of rosacanIna fruIts), reduces PaInandImProVes general WellBeIng In PatIentswiThosteoarthrItIs—a douBle-BlInd, PlaceBo-controlled, randomIsed trIale. reina, a. Kharazmib, K. Wintherc,*ainstituteforclinicalresearch,kolding,denmarkbdepartmentofclinicalmicrobiology,universityhospital,copenhagen,denmarkcdepartmentofclinicalbiochemistry,copenhagencountyhospitalingentofte,universityofcopenhagen,

abstractThetreatmentofosteoarthritis,adiseasethateventuallyaffectsthemajorityoftheolderpopulation,involvesthealleviationofsymptomssuchaspainandstiffness,andthereductionofinflammation.Thedouble-blind,placebocontrolled,crossoverstudyreportedhereexaminedtheeffectofhybenvital,aherbalremedymadefromasubtypeofrosacaninaandrecentlyreportedtohaveanti-inflammatoryproperties,onthesymptomsofosteoarthritis.onehundredandtwelvepatientswithosteoarthritiswererandomlyallocatedtotreatmentwitheitherhybenvital5gdailyoranidenticalplacebofor3months,followedimmediatelybythealternativetreatment.Thepatientsassessedchangesinjointpainandstiffnessaftereachtreatmentperiodona5-pointcategoricalscale.generalwellbeing,includingmood,sleepqualityandenergywerealsoassessedandrecordedinapersonaldiary.

1514

Theresultsinthetwoarmsofthecrossoverdifferedmarkedly.groupa(placebofirst)showedsignificantlymoreimprovementfromhybenvitalthanfromplacebo,po0:0078forpainando0:0025forstiffness.butgroupb(hybenvitalfirst)revealedapositiveeffectofthesameorderasforhybenvitalingroupa,notonlyfromtheactivedrug,butalsofromplacebo(differencenotsignificant).anidenticalpatternwasobservedwhenweevaluatedgeneralwellbeingfromthediaryrecords.whenpatients,onthebasisofreductioninjointpain,weredividedintorespondersandnonresponders,thefirst3monthsofactivetreatment(groupa)showedaresponserateof31/47(66%)comparedtothatofplacebo(groupb)18/50(36%),po0:0185:nomajorsideeffectsoccurredineithergroup.Thedataindicatethathybenvitalreducesthesymptomsofosteoarthritis.weinterpretthemarkeddifferencesintheresponsesofthetwogroupsasindicatingastrong‘‘carryover’’effectofhybenvital.©2004elseviergmbh.allrightsreserved.

Introductioninflammatorycellssuchaspolymorphonuclearleukocytesareknown tobecausally involvedin inflammation, in pain and tissue damage.The damage is caused by release of proteo-lyticandhydrophilicenzymesaswellas toxic,reactive oxygen radicals derived from cellsactivatedinthetissuesandjoints(harris,1988).Thenon-surgicaltherapyofosteoarthritis,adi-sease that attacks many of the middle-agedandthemajorityoftheolderpopulation,invol-vesalleviationofthesymptomsassociatedwiththedisease,suchaspainandstiffness,andthereductionof inflammation.acetylsalicylicacidanda rangeofnon-steroidalanti-inflammato-ry drugs including ibuprophen, indomethacinandnaproxen,aswellasglucocorticoids,havebeenusedforthetreatmentofarthritis(hoch-bergetal.,1995;vaneandbotting,1998).Thesedrugshaveavarietyoftoxicandunwantedef-fects, including interferencewithhaemostasis,gastricerosionandadverseeffectsontheliverandkidneys(hochbergetal.,1995;vaneandbotting, 1998). selective inhibitors of the cyc-looxygenase- 2 system have recently shownpromising analgesic and anti-inflammatoryproperties,withoutthesideeffectsmentioned.Theyare,however,stillexpensiveandanega-tiveeffectonthecirculatorysystemcannotbeexcluded (mukherjee et al., 2001). recently,acetaminophenwasshowntoworsentheriskofuppergastrointestinalcomplications(rodriguesandhernandes-diaz,2001).Thereis,therefore,stillaneedforasafe, low-costremedyforthelong-termtreatmentofsymptomsinosteoarth-ritis.wehaveearliershownthatastandardiseddrypowder,hybenvital,madefromthehipsofaparticular subtypeofrosacanina, reducedbothchemotaxisandthegenerationofoxygenradicalsinpolymorphonuclearcells(wintheretal.,1999;kharazmiandwinther,1999).

Theplantsused for thecurrentpreparationofhybenvitalpowderaregrowninstandardisedfieldsaccordingtogoodagriculturalpractice.

harvestingtakesplacewhenthefruitsarema-tureandallfruitsarebroughttofreezingfacili-tieswithoutdelay. later, the selectionofopti-malfruitsforproductionofthepowderismadeby a laser technique and the temperature ofthesubsequentcontrolleddryingprocessneverexceeds 40 c. The powder contains seeds aswell as husks (rosae pseudofructus cum fruc-tibus) fromacertain subtypeofr.canina,bynameliTo,andisfinallystandardisedtocontainatleast500mgvitamincper100ghybenvitalpowder. further constituents are: pectins (to-tal)58.0mg/g,b-carotene57.9mg/kg,b-sitos-terol0.5mg/g,folicacid1.6mg/kg,vitamine4.6mg/100g,mg170mg/100g,zn1.0mg/100g, copper 10.9 mg/100 g and non-quantifiedbut reported flavonoids. The antiinflammato-ry effect of the powder is not related to thewell-knownhighvitaminccontentofrosehipextracts (kharazmiandwinther, 1999). but wehave earlier shown that, hyben vital modifiesinflammation by reducing both chemotaxisandthegenerationofoxygenradicalsinpoly-morphonuclearwhitecells(wintheretal.,1999;kharazmiandwinther, 1999).moreover,manyvolunteers have claimed that pain from oste-oarthritis was diminished after a few weeks oftreatmentwiththepowder(wintheretal.,1999;kharazmiandwinther,1999).all thesefindingsencouraged us to investigate whether hybenvital,inalargercontrolledtrialasnowreported,wouldaffectpain,stiffnessandgeneralwellb-eing and the consumption of pain-reducingmedicines, inparticularparacetamoland thesynthetic opioid Tramadol, in patients with os-teoarthritis.

methodsPatientsafter we had obtained approval for the trialfromthelocalethicalcommittee,125cauca-sianoutpatientswereenrolledthroughadver-tisements in local newspapers. The study wasperformedaccordingtogoodclinicalpractice

anddesignedtoaccord,asfaraspossible,withthe guidelines on conduct of clinical trials onosteoarthritis devised by the osteoarthritis re-search society international. The only notableexceptionwasthatthestudyincludedpatientswitharthritisofvariousjointsinsteadofconfiningittoasinglejoint(altmanetal.,1996).Thevo-lunteersallgavetheiroralandwritteninformedconsent. They had all been earlier diagnosedbytheirowngeneralpractitionerorlocalrheu-matologist as suffering from osteoarthritis, andwere reported to have an x-ray verified dia-gnosisandsymptomsofprimaryosteoarthritisinthehip,knee,hand,shoulderorneck,orsomecombination of these, for at least the last 12months.allreportedpainoftheaffectedjointsofatleastmildtomoderateseverity.weexclu-ded patients with liver or kidney disease andthoseknown to suffer fromallergyorahistoryof drug or alcohol abuse. we also excludedpatientswithcancer,rheumatoidarthritis,fibro-myalgia, gout, serious cardiovascular disease,asthma requiring treatment with steroids, andany other disease which would substantiallyinfluence thepatients’qualityof life. likewise,we excluded those who had received intra-articular hyaluronate, glucosamine sulphate,immunosuppressivedrugssuchasgoldorpeni-cillamineorinjectionsofglucocorticoidswithinthe6weekspriortothestudy,andpatientswhowere found tobeunable toco-operateafterthefirstevaluation.

trial designThe trial was of a double-blind, placebo-con-trolled,crossoverdesign,andrandomisationoftreatmentallocationwasperformed inblocksof four with the block size unknown to the in-vestigators. The design had three immediate-ly successive periods: a 14 days run-in periodfollowedby randomisedallocationof thetwotreatment periods of 3 months each. The twoprimary efficacy parameters were: change injointpainandthealterationofconsumptionofconcomitant ‘‘rescue’’ medication for allevi-ating pain, evaluated after each of the two,blinded,3-monthstreatmentperiods.Thethreesecondaryefficacyparameterswere:jointstiff-ness,generalwellbeingincludingmood,ener-gyandsleepquality,andasubjectiveoverallevaluation of preference for one or other ofthe study medications. The run-in period wasintendedprimarily forpatients tobecomeac-customed to the ideas of the trial, and to beinstructed in and practise the daily subjectiveassessment/ record-keeping required, ratherthanasaformal‘‘baseline’’.however,wetookthe opportunity during this period to measure

blood pressure and removed a routine bloodsampleformeasurementofhaemoglobin,cre-atinine,sodiumandpotassium,bloodglucoseand cholesterol. The patients were then ran-domlyallocated,inblocksoffour,byacompu-ter-generated allocation schedule, to receivecapsulescontainingeitherabiologically stan-dardised rose hip powder (hyben vital) or anidentical placebo. The capsules were kept innumbered containers. The daily dosage wasfive0.5gcapsulesa.m.andp.m.oneorotheroftheresponsibleinvestigatorsenrolledallpati-ents.Thepatientsaswellastheresearchteamwere kept blind throughout the study. after 3months, the groups switched immediately tothealternativetreatmentforafurther3months.immediately after each of the two treatmentperiods,afurtherroutinebloodsamplewasta-kenandbloodpressurewasmeasured.whenthe trial had been completed, all data wereentered onto the spreadsheet, after whichthe treatment code was broken and patientswereseparated intotwogroupsaccordingtothetreatmentsequencetheyhadreceived.ittranspiredthatgroupastartedoutwith56ran-domisedpatientswhotookplacebofirst,follo-wedbyhybenvital,whilegroupbcomprisedthesamenumberofrandomisedpatientswhotookhybenvitalfirst,followedbyplacebo(fig.1).Thedataforthetwogroupsseparatelywerealso entered on the spreadsheet, which wasthen mailed to the statistician, who was alsokeptblindastothetreatmentcode.

methods of assessing clinical effectPrimary efficacy parametersThecardinalitemofinformationobtainedwastheend-of-treatmentsubjectiveassessmentsofanychangesinpainthathadoccurredduringeachofthetreatments.Thesewereestimatedby the patients on a 5-step categorical scalerangingfrom0(nochange)to4(almost totalrelief of pain). here, the higher the score thegreatertheclinicalbenefit,ariseof1categoryrepresenting25%improvement.Thistechniquealsoallowedustocalculatethenumberofre-spondersandnon-respondersineachgroup.

each type of ‘‘rescue’’ analgesic consumedwasnoteddailybythevolunteersinadiary.allpatientstakingnsaidsregularlyonprescriptionfromtheirgeneralpractitionerswereadvisedtocontinuesuchtreatment,withoutanychangeindosage,throughoutthestudy.Threeweeksintoeachofthetwotreatmentperiodswerecom-mendedthepatientstoreducetheirconsump-tionofconcomitantpain-relievingmedicine, ifatallpossible.consumptionofsuchmedication

1716

wasrecordeddailyinadiary,andattheendofeach3-monthstreatmentperiodwecalculatedtheconsumptionofeachtypeofnon-trialpain-relievingmedicine.aspatientsnormallyuseawide range of rescue medications, we simpli-fiedaccountingofthembytransformationintoparacetamol equivalents, as devised by thedanish health authorities (lægemiddelkata-loget,2002).Thus25mgofTramadoland25mgofcodeinewouldbeconsideredasequivalentto1000mgparacetamolandaspirinwouldbeconsideredequaltoparacetamol.

secondary efficacy measuresThepatientsmadeasubjectiveassessmentofjointstiffnessattheendofeachtreatmentperi-od,ona5-stepcategoricalscalerangingfrom0 (no change) to 4 (almost total relief of thesymptom),asdevisedforpain.inaddition,thepatientsmadeadailysubjectiveassessmentoftheseverityofjointpain(inthemorningandla-terintheday),stiffness(inthemorningandlaterin the day) and the state of wellbeing, sleep,energy,andmoodwas recordedbythepati-ent inadiary.eachaspectwasassessedandrecorded on a separate 10-point categoricalscale, where an increasing score denoted in-creasingdisability.anaverageofeachkindofmeasurementwastakenforstatisticalcompari-sonoftreatments.

Patients’ overall evaluation of the study medicationon the final day of the trial, before the treat-ment code had been broken, the supervisingphysician asked this question of the patient:Taking all aspects into consideration, did youdevelop a definite preference for one of thetreatments,ornot?

statistical techniqueswe based the sample size on results from anearlierclinicaltrialusingthesamedrypowder.data from all the randomised patients wereenteredonthespreadsheet.statisticalevalua-tion was based on the intention to treat (iTT),withthelastvaluecarriedforward.weappliedwilcoxon’stestformatchedpairswhenevalu-atingthestudyasasimplecrossover trialandwhen we compared effects occurring withinthesamegroupofpatients.Themann–whitneytestwasappliedtocomparisonofgroupsaandbafter3-monthstreatment.Theonlyexceptionsweresimpleyes/noquestions,towhichfisher’stest was applied. data given are mean ±sd.anypvalueequaltoor<0.05wasregardedasstatisticallysignificant.

resultsdescription of patientsofthe125eligiblepatientswhorespondedtoouradvertisement,weeventuallyenrolled112,including71women,meanage68years(range33–93)and41men,meanage64years(range35–89)(seeflowdiagramoffig.1).

matching of groupsdetailsaregiveninTable1.Thetwogroupswerevirtuallyidenticalintheirdemographicdata,intheseverityanddistributionofosteoarthritisandin theirconsumptionof rescuemedication; in-

deed,therewasnosignificantbetween-groupdifferenceinanyofthe16itemsofTable2.Themean body mass index for the included pati-ents was 26.9, range 18–42 kg/m2. althoughonly85patientscompletedthetrial,thetwofi-nalgroupsofper-protocolpatientswerestillnotsignificantly different. we consider the groupsthereforetohavebeenverywellmatched(Ta-ble1).

fifteenpatientsdroppedoutbefore thefirst3monthsperiodwasfinished,leaving97patientsfor thesecondpartof thestudyand85com-pleted both treatment periods (fig. 1). beforethecodewasbroken,a further5wereexclu-dedbecauseofprotocolviolationdetectedonevaluationofthepatient’srecordformbeforethedatawereenteredonthespreadsheet.Thisleft 80patients, 46womenand34men, foraper-protocolanalysis.of the randomisedpati-

ents,59hadarthritisoftheknee,46ofthehip,40hadinvolvementofthehands,18oftheneckand14oftheshoulderoracombinationofthe-sedifferentjoints.Thedropoutswerecorrespon-dinglyrepresentedbyallthedifferentjointsmen-tionedandtherewerenomajordisagreementsbetween the iTT and the per-protocol analy-sis—hencewereferonlytotheiTTanalysisifnototherwise stated.of the includedpatients, 40weretakingnsaidsregularly,40paracetamol,12Tramadol,3codeine,2aspirin,2morphine,and1dextropropoxyphen.Thirtyofthepatientstooknorescuemedicationwhatever.whenasubanalysisoftheinitialvaluesoftheplacebo-firstgroup(n¼56)versustheactivetreatmentfirstgroup(n¼56)weremade,therewerenosignificantdifferencesinbodymassindex,age,sex,jointsinvolved,consumptionofnsaidandrescuemedication(Table1).

125 patients eligible

1 entered an other study3 chose other therapy9 declined to take part

112 randomised

9 withdrawals3 personal reasons3 protocol violation1 high dose steroids1 diagnosed cancer1 acid regurgitation

6 withdrawals3 personal reasons1 protocol violation1 high dose steroids1 acid regurgitation

groupa

n=47

groupB

n=50

outcome measurements

outcome measurements

7 withdrawals2 personal reasons2 protocol violation2 intercurrent surg.1 diarhoea

5 withdrawals1 personal reasons3 protocol violation1 high dose steroids

groupa

n=43

groupB

n=42

n=56 n=56

group a: active first

group B: Placebo

first

fig. 1. flow-chart showing the dropout rate of the different time points of the study.

table 1. Baseline demographic and osteoarthritic characteristics of the study population

age(years)sex women men

bmi(kg/m2)no.ofpatientswithoaofthehipno.ofpatientswithoaofthekneeno.ofpatientswithoaoftheneckno.ofpatientswithoaoftheshoulderno.ofpatientswithoaofthehandno.ofpatientsonnsaidsnoofpatientsonparacetamolno.ofpatientsontramadolno.ofpatientsoncodeinnoofpatientsonaspirinnoofpatientsonmorphinenoofpatientsondext.ppox.phen.no.ofpatientswithnomedication

Intention-to-treat population

Placebo treatment(n = 56)

66.8±11.8

3421

26.8±5.02030117

17202171111

26

Per-protocol-population

active treatment(n = 56)

67.1±11.6

3720

27.7±4.9262977

23201952110

24

Placebo treatment(n = 39)

active treatment(n = 56)

67.5±10.6

2118

27.7±4.91524106

15171860110

19

67.0±10.8

2615

27.5±5.5212466

19191640110

18

compliancecompliance, as calculated from the proporti-on of study medication (number of capsules)returned by the patients, was 92.8 ± 11% forhybenvitaland90.6±11%forplacebo(non-significantdifference).compliance in thepla-cebo-first group was 92.3 ± 10.0% and for ac-tive treatment first 90.5 ± 8.0% (nonsignificantdifference).

Primary efficacy measures: paindetailsaregiveninTable2.Themostimportantitemofclinical information—thepatients’finalevaluations of change in pain—showed a re-markable difference between the groups. ingroupa(placebofirst),therewasahighlysig-nificantdifference in favourofhybenvital—a

mean rise from 1.02 ± 1.45 after placebo (animprovement of 25%), to 1.91 ± 1.43 (an im-provement close to 50% of the improvementscale)observedafter3monthsofhybenvitaltreatment,p<0.0078.butgroupb(startingwithactive treatment) showed no significant dif-ference between the two treatments: 1.45 ±1.28 units for active treatment, as comparedwith1.72±1.37forplacebo,p=0.6084.Table2,upperpanel,andthehistogramsoffigs.2aand b illustrate the large between-treatmentdifferences, when groups a and b are com-pared.groupapatientsshowedamarkeddif-ferencebetweenthetwotreatmentsateverydegreeofresponse,whilebshowednoconsis-tentpatternofdifferencebetweentreatments.The carryover effect that we postulate as re-

1918

table 2. Pain given on a scale from 0 (no reduc-tion) to 4 (almost total relief of pain), consumpti-on of rescue medication given as paracetamol equivalents (g)

sponsibleforthisbetween-groupsdiscrepancy(seealsodiscussion) likewiseblunted the levelofsignificancewhenthetwotreatmentgroupswerelumpedtogether:therewasagainnosig-nificantdifferencebetweentheeffectsof thetwotreatments(p<0.0991),datanotshown.anevaluationofbetween-groupdifferencesafteronly3-monthstreatmentdidnotattainstatisti-cal significance,althoughan improvementof50%wasobservedinfavourofactivetreatment(p<0.101)datanotshown.we also made an alternative analysis of thedatabyidentifyingtwocategoriesofsubject—‘‘responders’’ who by definition showed atleastonecategoryofimprovementand‘‘non-responders’’, who showed less improvementthan this. if we compare the a and b groupsafterthefirst3monthsoftreatment,theoveralloutcomeof theanalysis is that31/47 (66%)ofsubjectsrespondedtohybenvital,while18/50(36%)respondedtoplaceboandthiswassigni-ficantatp<0.0128).Thecorrespondingper-pro-tocolevaluationyieldedapvalueof0.0428.

Primary efficacy measures: painTwenty-threepatientshandedinmedicaldiari-esadequateforiTTanalysisoftheiruseofnsaidsinaccordancewiththeprotocol.consumptionduringthetwotreatmentperiodswasfoundtobe identical (data not shown). paracetamolandacetylsalicylicacidwereadministeredas500mgtabletsandTramadolandcodeineas50and25mgtablets,respectively.Twentyfivepatients handed in medical diaries adequatefor iTT analysis of their daily use of paraceta-

Placebo

Group A: Placebo first, then active treatment

painrescuemedicationstiffness

1.02±1.45227.40±249.50

0.91±1.38

active treatment

1.91±1.43127.90±143.30

1.91±1.25

p-value

0.00780.00240.0025

Group B: Active treatment, then placebo

painrescuemedicationstiffness

1.45±1.28127.50±94.00

1.28±1.35

1.72±1.3777.70±51.1

1.71±1.47

0.60840.14520.3850

Placeboactive treatment p-value

stiffness estimated on a scale from 0 (no reduction) to 4 (almost total relief of stiffness) is given for groups a and B. data given are mean ±sd.

molandsevenandfourandtwopatients, re-spectively, handed in diaries adequate for iTTanalysesoftheirdailyuseofTramadol,codeineandaspirin.apatternverymuchlikethatpre-viouslydescribedforpain,occurred.groupa,placebofirst,dataavailable from12patients,showed after 3 months a mean consumptionof 227.4 ± 249.5 g. however, this consumptionwasreducedto127.9±143.3gafter3monthsofactivedrugtreatment.Thisdeclineof99.4±163.9g(p<0.0024)compriseda44%reduction.The b group, active treatment first, with dataavailablefrom15volunteers,showedafterthefirst3monthsofactive treatmentameanva-lueof129.50±91.00g,avalueclose towhatwasobserved in thesecondactive treatmentphaseofthegroupapatients(seeTable2).afurther 3 months placebo treatment, in the bgroup, resulted in a non-significant decline to77.70±51.1g(Table2).nosignificantchangewaspresentwhenthetwogroupswerelumpedtogether(p<0.1420),datanotshown.aneva-luationofthetwogroupsafter3-monthstreat-mentshowedplacebovaluesof227.4±249.5gandactivevaluesof128.4±94.3g.Thereduc-tion,infavourofactivetreatment,was44%,butwasnotstatisticallysignificant.when,however,asubanalysiswasmadeonthedeltachangeinconsumptionofrescuemedicationfromthebe-ginningofeachofthetwo3-monthstreatmentperiods (the two initialweeksof treatment) totheendofeachoftherespectiveperiods(thefinal twoweeksof the3-month treatmentpe-riod),therewasasignificantreduction incon-sumption of rescue medication from activetreatment,whencomparingplaceboandac-tivetreatment(p<0.006),datanotshown.

25 –

50 –

75 –

100 –

0 0 1 2 3 4

25 –

50 –

75 –

100 –

0 0 1 2 3 4

Perc

ent

ag

e ü

f pa

tient

s

non-responders respondersgraded

Placebo active treatment

n=50p=0.0078placebovsactivetreatment

reliefofpain

a

25 –

50 –

75 –

100 –

0 0 1 2 3 4

25 –

50 –

75 –

100 –

0 0 1 2 3 4

Perc

ent

ag

e ü

f pa

tient

s

non-responders respondersgraded

Placeboactive treatment

n=47p=0.6084activetreatmentvsplacebo

reliefofpain

B

fig. 2. (a) histograms comparing, in group a subjects (placebo first then hyben Vital), the degree of improvement in pain relief from pla-cebo (white columns) as compared with hyben Vital (shaded columns). the height of each co-lumn indicates the percentage of patients who experienced pain relief of category 0, 1, 2, 3 or 4, corresponding to 0%, 25%, 50%, 75% or 100% relief of pain (p value refers to the added scores comparing the two different treatments). (B) his-tograms comparing in group B subjects (hyben Vital first then placebo), the degree of impro-vement in pain relief from hyben Vital (shaded columns) as compared with placebo (white columns). the height of each column indicates the percentage of patients who experienced a given pain relief of category 0, 1, 2, 3 or 4, corresponding to 0%, 25%, 50%, 75% or 100% relief of pain (p value refers to the added scores comparing the two different treatments).

secondary efficacy measuresJointstiffness,testedonascalefrom0(noim-provementatall)to4(almosttotalreliefofstiff-ness) revealed an almost identical pattern tothat found forpain. ingroupa, the initialpla-cebovaluewas0.91±1.38(animprovementof23%onthescale)ascomparedto1.91±1.25(animprovementof48%)whileonhybenvitaltherapy),p<0.0025.groupb,however,showedno significant difference between treatments:hyben vital 1.28 ± 1.35 versus placebo 1.71 ±1.47,p<0.3850(Table2).norwasthereanysig-nificantdifferencewhenthe twogroupsweretakentogether (p<0.1612),datanot shown.acomparison of the two groups after 3 monthsoftreatment,althoughinfavourofactivetreat-ment, did not attain statistical significance(p<0.153),datanotshown.Thediaryrecordsofjointpainandstiffnessinthemorningandlaterintheday,wellbeing,mood,energyandsleep,availableindiariesfrom47patients,showedthe

samesharpdistinctionbetweengroupsasfortheprimaryparameters.Theplacebo-firstgroupa(n=26)showed,inallmeasurementsadistinctdifferenceinfavourofhybenvital.Thechangewashighly significant, stiffnessandpain in themorninggivingpvaluesof0.0016and0.0127,respectively,andsleepquality,moodandge-neralwellbeing,0.0096,0.0124and0.0164, re-spectively. but in the hyben-first group b, thetwo sets of results appeared indistinguishable,andtherewasnotasingleinstanceofanythingapproachingastatisticallysignificantdifferencebetween the two treatmentgroups,as shownbyameanplevelofmorethan0.50(detailsnotshown).Themajorityofthesignificantchangesobserved in favour of active treatment in theplacebo-firstgroup,wereconfirmed,whensub-analysiscomparingtheaandbgroupafter3-months treatment was made: stiffness in themorning,p<0.054;paininthemorning,p<0.036;general wellbeing, p<0.012; mood, p<0.017;andsleepquality,p<0.005.

Patients’ preference for treatmentThe separate groups again showed a largedifference,similar inpatterntothatdescribedabove. in group a, 24 patients reported thatthey felt most improvement from hyben vital,while8patientspreferredplaceboand9werenotsure(p<0.0070).ingroupb,12patientspre-ferredthefirsttreatment(hybenvital)whereas20votedforplacebotreatmentand8didnothave any preference (p<0.2153). comparisonof theaandbgroups(fisher’s test)gaveapvalueof<0.0040infavourofhybenvital.

routine screening testshaemoglobin,bloodglucose,creatinineandso-diumandpotassiumlevelswereunaffectedbyeithertreatment.norwerethereanychangeswhenthosepatientswithbloodglucose levelsabove 5.5 mmol/l were analysed separately.anunexpectedfindingwasthathybenvitalre-sultedinasmallbutsignificant8.5%falloftotalcholesterol.

unwanted effectsalthough27oftheoriginal112subjectsrecrui-teddroppedoutduringthe6-monthstreatmentperiod, only 3 of these defaulted because ofadverseeffects:acidregurgitationoccurredinonepatientduringplacebotherapyandinoneduringactivetreatment,andoneotherpatientwithdiarrhoeadroppedoutwhileonplacebo;fordetailsseefig.1.intheremaininggroupthe-rewere12whoreportedmilderunwantedef-fects.Thesewereasfollows:frequencyofmic-turition4(threewhileonactivetreatmentand

2120

onewhileonplacebo);waterbrash3(presentinbothtreatments);diarrhoea2(present inbothtreatments); constipation 2 (1 during placeboand1duringbothtreatments);urticaria1(whileonplacebo).Therewerenomajorsideeffectsofanykindinthewholegroup.

discussionInterpretation of trial resultsThe chief advantage of a crossover trial, asused here, is that in comparing the effects oftwosuccessivetreatmentsonthesame‘‘arm’’of the trial, each patient acts as his/her per-fectcontrol,soconcernaboutmismatchingofthegroups—animportantsourceoferror—canbe forgotten.awhollyuncomplicatedcrosso-vertrialwithapositiveresultcanbeexpectedto yield three pieces of information: a within-group significant comparison of the two testsubstances—onefromeachofthetwoarmsofthe trial (andmoreor less identicalwitheachother),andasignificantbetween-groupscom-parison at the crossover point, provided thatthegroupshavebeenwellmatched, since inthiscase thepatientsdonotactas theirowncontrols.

lookingattheresultsofthetrialdescribedhere,itisobviousthattheyarefarfromthisidealisedpattern.Thatarmofthetrialgivenplacebofirstdoes show a significant, clear-cut differencebetweentheeffectsofthetwotestsubstances.so far so good, but the other arm— activesubstancefirst,placebosecond—showsnosig-nificantdifferencebetweenthetwo.webelie-vethatbyfarthemostlikelyexplanationofthisdiscrepancybetweenthetwoarmsofthetrialis a strong ‘‘carryover’’ effect of hyben vital.Thisisacommon,majorcomplicationofcross-overtrialsandthereasonfortheinclusionofa‘‘washout’’periodaftercrossover.

Theusualtacticalresponseistowriteoffalldataafterthecrossoverpointandtosupplementthesinglewithingroupresultobtainedintheplace-bo-firstarm,withabetween-groupscomparisonatthecrossoverpoint.butthis,usingtheprimaryefficacydataofTable2,alsogaveanon-signi-ficantresult.Thisraisesthepossibilitythatacar-ryover effect is not the whole explanation—aslowonsetof theactivedrugeffectcouldbeanother factor. The strength and significanceofthedifferencebetweenplaceboandactivedrug seen in group a is supported by severalancillaryaspects.ifthereductioninpainsensa-tionwasevaluatedafter3-monthtreatmentonayes/nobasis,therewasasignificantreductionofpainfromactivetreatmentwhencompared

toplacebo.inagreementwiththisfinding,pre-ferencefortreatmentaorbwasalsoinfavourof active treatment and the diary recordingsonpain,generalwellbeing,moodandsleepingqualitywereallstatisticallysignificantinfavourof active treatment. Taken together these fin-dings seem to fully justify confirmation of theactionofhybenvitalbyalarge-scale,parallel,placebo-controlled,blindstudy,andthis isourintention.

r.canina(the‘‘dogrose’’,thecommonwild-briarroseofenglishhedgerows)issaidtohavebeen so named because the ancient greeksbelieveditsroottobeeffectiveagainstthebiteofamaddog(brewer,1981).inthiscontext,plinytheelderusedtheplant’sclassicalgreekname‘‘cynorrhodos’’,combiningtheverbalrootsof‘‘dog’’ and ‘‘red’’ (pliny, 1966). although hy-benvitalhasbeenmarketedinscandinaviaforseveralyears,moderneuropeaninterestintheplanthasbeenconcentratedonpreparationsmadefromthehipsratherthantheroot,mainlybecauseoftheirhighcontentofvitaminc,andherbal tea infusionsof ‘‘cynorrhodon’’are stillusedtoday.

it iswidelyknownthatrosehipscontainsignifi-cantamountsofvitaminc,butitseemshighlyunlikelythatthisaccountsformuch,orindeedany,oftheactivityofhybenvitalinthistrial.alarge-scale study in 1996 on the framinghampopulationgroupshowedthatthemiddleandhighesttertilesofdailydietaryvitamincintakedid protect against the long-term progressionofkneeosteoarthritis(especiallyagainstlossofcartilage).butthelowestintaketertile—adailymeanof 81mg formenand94 forwomen—had no such protective effect (mcalindon etal.,1996).Thevitaminccontentofahybenvi-taldosageof5gdaily,asusedinthistrial,isonly26mg,i.e.onlyone-thirdoftheframinghamlo-westtertileandthereforeveryunlikelytocontri-butesignificantlytotheactionofhybenvital.

we have earlier shown that hyben vital signi-ficantly reduces the migration of neutrophils,whenestimatedafter1monthoftreatment(win-theretal., 1999;kharazmiandwinther, 1999).oneexplanationforthelesseningofsymptomsduring hyben vital treatment could thereforebe a reduction of the inflammation that is anintegralpartofthepathogenesisofosteoarth-ritis(harris,1988).Thishypothesishasgainedin-creasing interest, as an active ingredient thatinhibits neutrophil chemotaxis, has now beenisolated from the present subtype of rose hip(larsenetal.,2003).ifthepresentsuggestionis

correct, it could also explain the pronouncedcarryovereffect;onceinflammationhassubsi-ded,itrequiresacertainintervaloftimebeforethe process can be reactivated. as rose hipshavebeenusedindailyhouseholduseforcen-turies,itissurprisingthattheiranti-inflammatorypropertyhasnotbeendetectedbeforenow.apossibleexplanationisthatdifferentspeciesofrosehipvaryintheiranti-inflammatoryproper-ties(brandtand(akesson,2002).inanotherstu-dytestingapossibleinteractionbetweenrosehip and warfarin, we could not show any ef-fectoncoagulationandplateletaggregability(winther,2000).Thissuggeststhatrosehip,un-likensaid,aspirinandginger—anothernaturalremedyalsousedforsymptomsofosteoarthritis(altman and markussen, 2001)—does not af-fectthearachidonicacidandcyclo-oxygena-sesystem.Thiscouldexplainwhytheincidenceofsideeffects is lowerforhybenvitalthanforthetherapiesmentionedabove.

conclusionwehavefoundthattheherbalremedyhybenvitalhasamoderatealleviatingeffectonjointpain and improves general wellbeing, sleepqualityandmoodinpatientswithosteoarthritis,withoutproducingany sideeffects.weconsi-der that the results warrant a largescale dou-ble-blind, long-term, placebo-controlled andparallelstudyofhybenvital.

acknowledgementshybenvitalinternational,langeland,denmarksupportedthestudy.

referencesaltman,r.d.,markussen,k.c.,2001.effect ofagingerextractonkneepainin patientswithosteoarthritis.arthritisrheum. 44(11),2531–2538.altman,r.d.,brandt,k.,hochenberg,m., moskowitz,r.d.,1996.conductofclinical trialsinpatientswithosteoarthritis. osteoarthr.cartilage4,217–243.brandt,k.,a(kesson,b.,2002.healthpromo- tingcompoundsinvegetablesandfruit. plantprod.29,43–44.brewer’sdictionaryofphraseandfable (revisededition,1981)346.in:evans,i.h. (ed.),cassell,london.p.1213.harrisJr.,e.d.,1988.pathogenesisof rheumatoidarthritis:adisorderassociated withdysfunctionalimmunoregulation. in:gallin,J.h.,goldstein,i.m.(eds.), snyderman,r.(eds),inflammation.basic principlesandclinicalcorrelates. ravenpress,newyork,pp.751–773.

hochberg,m.c.,altman,r.d.,brandt,k.d., clarck,b.m.,dieppe,p.a.,griffin,m.r., moskowitz,w.,schnitzer,T.J.,1995. guidelinesforthemedicalmanagementof osteoarthritis(part1and2).arthritisrheum. 38(11),1535–1546.kharazmi,a.,winther,k.,1999.rose-hipinhibits chemotaxisandchemiluminescenceof humanbloodneutrophilsinvitroand reducescertaininflammatoryparameters invivo.inflammopharmacology7(4), 377–386.larsen,e.,kharazmi,a.,christensen,l.p., christensen,s.b.,2003.anantiinflammatory galactolipidfromrose-hip(rosacanina) thatinhibitschemotaxisofhuman peripheralbloodneutrophilsinvitro.J.nat. prod.66,994–995.lægemiddelkataloget,2002.dansklægemid- delinformationa/s:issn0105-287x.mcalindon,T.e.,Jacques,p.,zhang,y., hannan,m.T.,aliabadi,p.,weissman,b., rush,d.,levy,d.,felson,d.T.,1996.do antioxidantmicronutrientsprotectagainst thedevelopmentandprogressionofknee osteoarthritis?arthritisrheum.39(4),648– 656.mukherjee,d.,nissen,s.e.,Topol,e.J.,2001.risk ofcardiovasculareventsassociatedwith selectivecox-2inhibitors.Jama286(8), 954–959.Theolderpliny,1966.naturalhistoryvll:books xxlvxxvll,149.williamheinemann,london (isbn:0-434-99419-7).rodrigues,l.a.c.,hernandes-diaz,s.,2001. relativeriskofuppergastrointestinal complicationsamongusersof acetaminophenandnon-steroidalanti- inflammatorydrugs.epidemiology12, 570–576.vane,J.r.,botting,r.m.,1998.anti- inflammatorydrugsandtheirmechanismof action.inflamm.res.47(2)578–587.winther,k.,2000.rose-hip,intheformof hybenvital,hasnoimpactoncoagulation, plateletfunctionandfibrinolysis.in:procee- dingsoftheThirdinternationalexhibition andconferenceonnutraceuticalsand foodforvitality,palexpoexhibitionand conferencecentre,geneva,switzerland, may2–3.winther,k.,rein,e.,kharazmi,a.,1999.The antiinflammatorypropertiesofrose-hip. immunopharmacology7,63–68.

2322

TheEffEctsofastandardIzed herBal remedymade fromasuBtyPeofrosa canIna In PatIentswiThosteoarthrItIs: a douBle-BlInd, randomIzed,PlaceBo-controlled clInIcal trIalodd Warholm, md, sigrun skaar, md, ewa hedman, rn, hanna maria mølmen, rn, and liv eik, rndepartmentoforthopaedicsurgery,vestfoldcentralhospital,Tønsberg,norway

Background: astandardizedrose-hippowderproducedfromtheseedsandhusksoffruitfromasubtypeofrosacaninahasbeenreportedtoinhibitleukocytefunctionsthatcausecellinjuryinosteoarthritis.

objective: Theaimofthisstudywastoassesstheimpactofstandardizedrose-hippowderonmobilityofthehipandkneejoints,activitiesofdailyliving,qualityoflife,andpaininpatientswithosteoarthritis.

methods:patientswithadiagnosisofosteoarthritisofeitherthehiporknee,verifiedonradiogra-phy,participatedinthisrandomized,placebo-controlled,double-blindstudy.halfofthepatientsweregivenfive0.5-gcapsulesofstandardizedrose-hippowdertwicedailyfor4months,andtheotherhalfreceivedidenticalplacebocapsulestwicedailyforthesameperiod.mobilityofthehiporkneewasmeasuredinbothgroupsaftertheinitialscreeningandagainafter4monthsoftherapy.

results:onehundredpatients(65women,35men;mean[sd]age,65.2[11.1]years)weredividedinto2treatmentgroupsof50patientseach.hipjointmobilityimprovedsignificantlyinthetreat-mentgroupcomparedwiththeplacebogroup(p=0.033).similarly,paindecreasedsignificantlyinthetreatmentgroupcomparedwiththeplacebogroup(p=0.035).Twopatients(4%)fromeachgroupwithdrewduringtheearlystagesofthetrialforreasonsnotrelatedtotreatment.

conclusions:inthisstudypopulation,standardizedrose-hippowderreducedsymptomsofosteoarthritis,as64.6%ofpatientsreportedatleastsomereductionofpainwhilereceivingtreat-ment.standardizedrose-hippowdermayimprovehipflexionandreducepaininpatientswithosteoarthritis.(currTherresclinexp.2003;64:21–31)copyright©2003excerptamedica,inc.

keywords:osteoarthritis,stiffness,pain,rose-hippowder,rosacanina.

Introductionduring the past decade, the commonly useddrugsforosteoarthriticpainwereaspirin,othernonsteroidal anti-inflammatory drugs (nsaids),and corticosteroids. 1 however, side effectshave been associated with prolonged use ofthesedrugs.during thepast5years, selectiveinhibitors of cyclooxygenase-2 (an enzyme in-volved in the synthesis of proinflammatory cy-tokines) have shown promising analgesic andanti-inflammatory actions without serious ad-verse effects. 2 however, these drugs are ex-pensive,andtheneedremainsforalowcost,safe remedy for long-term treatment of oste-

oarthritis. as a possible alternative, a standar-dized rose-hip powder* made from the seedsandhusksoffruitfromasubtypeofrosacaninaisavailable.Thispowderinhibitsleukocytefunc-tions thatcausecell injury inosteoarthritis. Theplantsaregrownaccordingtogoodagricultu-ral practice in standardized fields in denmarkandsweden.whenthefruitsaremature,theyare harvested and frozen immediately. selec-tionofoptimalfruitsforlaterproductionofpow-derismadebyalasertechnique,andthecom-puterizeddryingprocessdoesnotexceed40c.Thevitaminandmineralcontentofthepowderiscontrolled.uncontrolledexploratorytrials3,4of

this standardizeddry rose-hippowder showedanalgesicaction inpatientswithosteoarthritis.Thisfindingwasevidencedbyamean(sd)de-creaseintheserumconcentrationofc-reactiveproteinfrom8.25(4.9)mg/lbeforetreatmentto6.67(2.6)mg/laftertreatment,andinhibitionofpolymorphonuclearchemotaxis.Thesefindingswere sufficient toencourage thepresent trial.Theaimofthisstudywastoassessthe impactofthestandardizedrose-hippowderonmobi-lityofthehipandkneejoints,activitiesofdailyliving(adls),qualityoflife,andpaininpatientswithosteoarthritis.

Patient and methodsThis was a single-center, double-blind, rando-mized, placebo-controlled study. all patientshad a diagnosis of osteoarthritis of the hip orknee,verifiedonradiography,within12monthsbeforethestudy.patientswithpainfor6monthsandwhowereonawaitinglistforeitherhiporkneesurgery,oronalistforfinalevaluationforsurgery,were included.patientswhoreportedallergy to plant products or who had severeasthmaorliverdiseasewereexcluded.allpa-tients provided written informed consent toparticipate,andapprovalfromtheethicscom-mitteeofthestudysite(anoutpatientclinicinnorway)wasobtained.

patientswererandomizedingroupsof10usingan independent computerized system. onegroup was randomized to treatment with five0.5-g capsules of standardized rose-hip pow-der twice daily for 4 months. The other groupreceived the same quantity of placebo cap-sules(identicalinappearance,taste,andsmelltotherose-hippowdercapsules)forthesametimeperiodastheactivetreatmentgroup.

Primary outcome measuresmobilityofthehiporkneewasmeasuredinbothgroupsaftertheinitialscreeningandagainaf-ter4monthsoftherapy.mobilitymeasurementsincludedthefullrangeofexternalandinternalrotationofthehip;maximumflexionandexten-sionofthehipandkneemeasuredusingago-niometer (gallusplesner,oslo,norway)duringpassivemovement;andactivevoluntaryrotati-on,flexion,andextensionbythepatient.goni-ometrycanresultinsomevariationifthetestisnotconductedbythesameresearcherateachvisit.forthatreason,allmeasurementswereta-ken by the same investigator and data givenareexpressedasthemeanof3testepisodes.Themeasurementsofjointmovementarepre-sentedin2ways:asthenumericmeasurementstakenandalsoasadegreeofrestriction,calcu-

latedbysubtractingthesemeasurementsfromastandardvalueof125°forhipflexion,140°forkneeflexion,and45°forexternalandinternalhiprotation.5

secondary outcome measuresatthestartofthetrialandagainafter1,2,and4monthsof treatment,patients recordedanydifficultiesinperformingadls,suchaswalking,getting into and out of a car, shopping, andgettingupanddownfromthelavatory.Thedif-ficultywasestimatedonavisualanalogscaleranging from0 (nodifficulty) to10 (greatdiffi-culty).after4monthsoftherapy,patientsgavetheiroverallassessmentoftheeffectivenessofthestudymedicationonreliefofjointpainusingacategoric scaleof0 (no improvement) to4(almost total relief of pain). The patients alsowereaskedaboutreliefofpainonasimpleyes-or-noquestionnaireafter1,2,and4monthsoftreatment.

duringthetrial,patientswereaskedtomaintaintheirdailydosageofnsaids.anychangesthatdidoccurweretoberecordedinadiary.com-pliance was estimated by counting the num-berofcapsules returnedbypatients.adverseeventswererecordedonthecase-reportformscompletedateachvisit.

statistical analysisstatisticalanalysiswasperformedonanintent-to-treatbasis.resultsinthe2groupswerecom-paredusingthemann-whitneytestforparalleldata.Thewilcoxonsignedranktestformatchedpairs was used to compare baseline findingswiththoseafter1,2,and4monthsoftreatmentin each group separately. The chi-square testwas used for the questionnaires. all data arepresentedasmean(sd).statisticalsignificancewassetatp<0.05.

resultsone hundred patients (65 women, 35 men;mean [sd] age, 65.2 [11.1] years) were enrol-led.Thetreatmentgroupcomprised34womenand16men(mean[sd]age,65.1[12.2]years).Theplacebogroupcomprised31womenand19men(mean[sd]age,65.3 [9.9]years). Thedemographic and osteoarthritic characteris-tics of the 100 patients entering the study (in-tent-to-treatpopulation)andofthe96patientswhocompletedthestudy(per-protocolpopu-lation)areshown inTable i. Thedemographiccharacteristics and consumption of medicinewere similar in the intent-to-treat and per-protocolpopulations.atbaseline,activeflexionofthe hip, however, was significantly different in

theactive-treatmentgroupversustheplacebogroup in both the intent-to-treat and the per-protocol populations. active external rotationof thehipwassignificantlydifferent in theac-tive-treatment group compared with the pla-cebogrouponly inthe intent-to-treatpopula-tion.allpassivemovementswerecomparablebetweengroups.

amongthe100patients,therewere44hipjoints(25 in the treatmentgroup,19 in theplacebogroup)and56kneejoints(25inthetreatmentgroup,31intheplacebogroup)involvedinthetrial. all patients had experienced osteoarth-ritic pain for 2 to 12 years. four patients (4%)

withdrewduringtheearlystagesof thetrial:1womanand1man in theplacebogroupbe-causeofcardiacproblemsandasorethroat,respectively,and1womanand1man in thetreatmentgroupduetothepossibilityofhipsur-geryearlierthanexpectedandbecauseofthedesirenottocontinue,respectively.These4pa-tientscomprised3hipjoints(1inthetreatmentgroupand2intheplacebogroup)and1kneejoint(inthetreatmentgroup).Thebaselinede-mographiccharacteristics,medication,andos-teoarthriticcharacteristicsofthe2groupsweresimilar,exceptforrangeofmotionforactivehipflexionandactiveexternalhiprotation(p=0.041fortreatmentgroupvsplacebogroup).

the following improvements were observed inthe placebo group: walking down the street(p<0.05),gettingintoandoutofacar(p<0.001),putting on/taking off stockings (p<0.001),and getting up and down from the lavatory(p=0.274).Theseimprovementswerenotfoundat4monthsoftreatmentintheplacebogroup.in contrast, the group treated with the stan-dardized rose-hip powder showed significantchanges in themajorityofadl functionsafter1monthoftreatment,asfollows:walkingdownthe street (p<0.001),getting intoandoutofacar(p<0.05),shopping(p<0.001),puttingon/ta-kingoffstockings(<0.001),andgettingupanddownfromthelavatory(p<0.05).after2monthsof treatment, improvement was found in allof these adls (p<0.001 for all), and this groupcontinued to show significant improvement inthemajorityofadlperformancesatmonth4compared with baseline, as follows: walkingdownthestreet (=0.038),getting intoandoutofacar(p=0.054,borderlinesignificant),shop-ping (p=0.024), putting on/taking off stockings(p=0.019),andgettingupanddown fromthelavatory(p=0.016).

effects of 4 months’ treatment on Joint movementpatientsreceivingstandardizedrose-hippowdershowedsignificantimprovementsat4monthsinpassive hip flexion (p=0.003), external rotation(p=0.006), and internal rotation (p<0.001) (Ta-bleii).Theplacebogroupshowedasignificantimprovement in passive hip internal rotation

table I. Baseline demographic and osteoarthritic characteristics of the study population.

Placebo(n=50)

age,y*sex,no.(%) woman manno.(%)ofpatientswithoaofthehip

no.(%)ofpatientswithoaoftheknee

hip joint movement, deg*passiveflexionactiveflexionpassiveexternalrotationactiveexternalrotationpassiveinternalrotationactiveinternalrotationKnee joint movement, deg*passiveflexionactiveflexion

no. (%) of patientstaking concomitantmedicationnonensaidsparacetamolopioidsasthmamedicationantihypertensiveheartdiseasemedication

65.3(9.9)

31(62.0)19(38.0)

19(38.0)

31(62.0)

111.05(12.76)97.63(15.49)19.72(11.56)13.06(10.17)28.61(11.61)21.39(10.68)

128.71(14.37)123.55(14.73)

15(30.0)20(40.0)12(24.0)

2(4.0)2(4.0)2(4.0)

5(10.0)

srhP(n=50)

65.1(12.2)

34(68.0)16(32.0)

25(50.0)

25(50.0)

116.00(13.92)105.60(13.10)†

26.40(9.74)20.00(9.79)§

28.80(13.17)21.20(12.61)

132.40(9.14)124.80(11.77)

11(22.0)24(48.0)14(28.0)

0(0.0)0(0.0)2(4.0)3(6.0)

Intent-to-treat Population

srhP=standardized rose-hip powder; oa=osteoarthritis; nsaIds=nonsteroidal anti-inflammatory drugs.*Values are expressed as mean (sd). †P=0.020 versus placebo. ‡P=0.039 versus placebo. §P=0.041 versus placebo.

Placebo(n=48)

srhP(n=48)

Per-Protocol Population

65.8(14.7)

29(60.4)19(39.6)

17(35.4)

31(64.6)

111.47(13.20)97.94(16.01)20.00(16.01)13.44(10.76)29.37(12.09)22.19(11.10)

128.71(14.37)120.35(24.95)

14(29.2)19(39.6)12(25.0)

2(4.2)2(4.2)2(4.2)

5(10.4)

65.5(14.2)

33(68.8)15(31.3)

24(50.0)

24(50.0)

115.62(14.09)105.42(13.34)‡

25.62(9.13)19.17(9.05)

28.75(13.45)21.04(12.85)

132.08(9.20)124.58(11.97)

10(20.8)23(47.9)14(29.2)

0(0.0)0(0.0)2(4.2)3(6.3)

(p=0.031),butnotinflexionorexternalrotation.The between-group comparison at 4 monthsshowedasignificantdifferenceinimprovementinpassivehipflexion(p=0.033),butnotininter-nalorexternalrotation.

Thesamepatternsofchangeinjointmovement(and inpvalues)werefoundwhenhipflexion

and rotation were actively performed by thepatients(Tableiii).however,itshouldbenotedthat the baseline values for active hip flexionandactiveexternalhiprotationwerenotiden-ticalinthe2groups(Tablei),whichmakestheinterpretationoftheseresultsdifficult.

changesinpassiveflexionofthekneedidnotdiffersignificantlybetweenthe2groups(datanotshown).activetreatmentresulted iname-an(sd)improvementof2.71(4.42°)(p=0.012);thisvalue improvementwas3.75° (5.32°) in theplacebo group (p=0.005). asimilar pattern oc-curredwhenflexionwasperformedactivelybythepatientsattherequestoftheresearcher.

activities of daily livingchanges in difficulty performing adls did notdiffer significantly between the 2 groups. sig-nificant improvementwasobserved in the fol-lowingadlsintheplacebogroupafter1monthoftreatment:walkingdownthestreet(p<0.05),getting intoandoutofacar (p=0.258), shop-ping (p<0.001), putting on/taking off stockings(p=0.251),andgettingupanddown fromthelavatory(p=0.154).after2monthsoftreatment,

table II. Passive hip joint movements before therapy and standardized rose-hip powder (srhP) and placebo.

flexionsrhpplaceboexternal rotationsrhpplaceboInternal rotationsrhpplacebo

116.00(13.92)111.05(12.76)

26.40(9.74)19.72(11.56)

28.80(13.17)28.68(11.61)

9.013.9

18.625.3

16.216.4

Baseline, deg*

*Values are expressed as mean (sd). †P 0.003 versus pretreatment. ‡P 0.033 versus placebo. §P 0.006 versus pretreatment. ||P 0.001 versus pretreatment. ¶P 0.031 versus pretreatment.

119.37(14.09)†‡

112.38(14.27)

28.96(8.84)§

22.50(11.40)

34.38(13.41)||

33.13(12.09)¶

40.06.7

17.110.0

35.024.0

type ofmovement

restriction of movement,

deg

at 4 monthsof therapy, deg*

Improvement, %

Joint Painsignificantly greater relief of joint pain wasfoundinthegroupreceivingstandardizedrose-hippowderthanintheplacebogroupafter4monthsoftreatment(p=0.035;figure).atmonth4, 31 of 48 (64.6%) patients in the activetreat-mentgroupreportedsomeeffect, ranginguptoalmosttotal reliefofpain,whereas17of48(35.4%)patientsreportednoeffect.inthepla-cebo group, 27 (56.3%) patients reported noeffect of treatment, whereas 21 (43.8%) pati-entsreportedvariousdegreesofimprovement.when pain relief was assessed on a yes-or-nobasis,significantlymorepatientsinthetreatmentgroupcomparedwiththeplacebogroupindi-

catedthattheyhadpainreliefatboth1month(p=0.014)and4months(p=0.046)oftreatment,butnotat2monthsoftreatment.

compliance, concomitant medication, and tolerabilitycompliance was 98% in the treatment groupand97%intheplacebogroup.althoughpati-entswereaskedtomaintaintheirdailydosesofanalgesictherapythroughoutthestudy,inthegroupreceivingthestandardizedrose-hippow-der,7(14.6%)patientsreducedtheirconsumpti-onofnsaids,andnoneincreasedit.incontrast,4(8.3%)patientsintheplacebogroupdecrea-sed their consumption ofnsaids, and 4 (8.3%)

2524

patientsincreasedit.Thedecreaseinnsaidusein the treatment group was statistically signifi-cant (p< 0.016); however, the between-groupdifferencewasnot.Three(6.3%)patientsinthetreatment group and 2 (4.2%) in the placebogroupdecreasedtheirconsumptionofparaceamol. in the placebo group, 1 of the 2 (50%)patients takinganopioiddrug(tramadol) re-ducedtheirconsumptionofthatdrug.

The only adverse event reported was mildgastrointestinaldiscomfort (2 [4.2%]patients ineachgroup).

20 –

40 –

60 –

0 0 1 2 3 4

figure. degree of joint pain relief on a scale from 0 (no impact) to 4 (almost total relief of pain) after 4 months of treatment with standardized rose-hip powder (srhP) or placebo. P=0.035 for srhP versus placebo (mann-Whitney test).

discussionTheaimofthiscontrolledstudywastoanswerthefollowingquestions:doesthestandardizedrose-hip powder improve mobility of the hipandkneejoints?doesitreducethefunctionaldisabilityinperformingadlsthatgoeswiththerestrictedhipandkneejointmovements?doesitrelievepain?wefoundthat,inthegrouptre-ated with standardized rose-hip powder, (1)functionalcapacityof thehip,asassessedbyanobjectivemethod,wasimproved;(2)theim-pact on functional capacity and adls, whenmeasured subjectively, was less pronounced;and(3)painwasreducedinapproximatelytwothirdsof thesepatients.This response ratewascomparabletothatreportedforginger,6,7ano-thernaturalremedyoftenusedbypatientswithosteoarthritis.

Thedifferencebetweentheeffectsonobjectivemeasuresofhipandkneeflexion isdifficult toexplain.Thelarge-scale,controlledtrial8ofavo-

cado/soybeanunsaponifiablesin101instancesofosteoarthritisof thehipand62of thekneeshowedasimilar,sharpdifferencebetweenthetherapeutic response of the 2 joints. The factthatthehipjointisaballandsocket,whereasthekneedifference,andthepossibilitythatthepainisdifferentlymediatedinthe2jointsisba-sedonunsupportedconjecture.

pain is the cardinal symptom of osteoarthritis.duetodegenerationofthecartilageandlackofjointstability,smallintra-articulartraumasdooccur. injuries of this kind are reflected in bio-chemical responses, some of which involvecytokines.9cytokineshaveproinflammatoryef-fectsthataremanifestedasepisodesofpain,jointswelling,andredness.ourinterestinthesemechanisms lies in the fact that the standar-dized rose-hip powder used here inhibits thepolymorphonuclear chemotaxis that is a stepintheproinflammatoryactionofvariouscytoki-nes.Thiscouldbethebasisoftheeffectsofthestandardized rose-hip powder on joint pain.4

furthersupportforananti-inflammatoryactionof this compound is that the serum concen-tration of c-reactive protein, a marker for in-flammation, decreases significantly duringtreatment with the compound, as shown byamean(sd)decreasefrom8.25(4.9)mg/lto6.67(2.6)mg/l.3,4Thebasicmechanismoftheanti-inflammatoryactionofstandardizedrose-hip powder does not reside in a blockade ofthe cyclooxygenase pathway, as is known tobethecasefortheanti-inflammatorydrugs(as-pirinandothernsaids)andtheherbalremedyginger.10,11Thiswasshowninastudy12measuringplateletaggregationduringtreatmentwiththesamestandardizedrose-hippowderindosesfarhigher than that used in the present study. incontrast todrugs inhibiting thecyclooxygena-sepathway,plateletaggregationwasnotaf-fectedbythesehighdoses.infact,thepowderseemstostabilizecellmembranes,asshownbythe finding that erythrocytes from individualstreatedwiththepowder,whenroutinelystoredinabloodbank,leaklesshemoglobinthanex-pected.13

natural vitamins c and e are present in stan-dardizedrose-hippowder.however,itdoesnotseemlikelythatthesevitaminscanexplainthepresentfindingsbecausevitamincwasnotin-volvedintheanti-inflammatoryactionreportedfor rose-hip powder,4 and vitamin e has beenreportedtobeineffectiveforsymptomaticre-lief of osteoarthritis.14 also, the prevalence ofgastrointestinaladverseeventswas low in thepresenttrialandsimilartothatofplacebo.mo-

reover,severalyearsofuseofthepowderinthescandinaviancountrieshasnotdisclosedsigni-ficantdataonanyadverseevents.

although a significant increase was found inmobilityofthehipjointandasignificantdecre-aseinpainwasfoundinthemajorityofpatientswho received the standardized rose-hip pow-der,theclinicalbenefitof4monthsoftreatmentshould not be overestimated. future researchshould include long-term studies to evaluatejointmobility,clinical improvements,andcon-sumptionofnsaidsandothertypesofconco-mitantpain-reducingmedicine.itisalsoimpor-tant tofindtheactive ingredient(s) in rose-hipandclarifywhetherthecontentofsuchactiveingredient(s)(aswellasthecontentofvitaminsand minerals) differ among subtypes, as spe-ciesofrose-hipcanbeverydifferentfromeachotherregardingbiologicalactivity.15

conclusionsin this study population, standardized rose-hippowderreducedsymptomsofosteoarthritis,as64.6%ofpatientsreportedatleastsomereduc-tionofpainwhilereceivingtreatment.standar-dizedrose-hippowdermayimprovehipflexionandreducepaininpatientswithosteoarthritis.

acknowledgementsTheauthorsthankhyben-vitalinternational(Tul-leboelle, langeland, denmark) for supplyingthecapsulesof standardized rose-hippowderandplacebo.

references1. hochbergmc,altmanrd,brandtkd,etal. guidelinesforthemedicalmanagement ofosteoarthritis.part1.americancollege ofrheumatology.arthritisrheum. 1995;38:1535–1540.2. simonls,lanzafl,lipskype,etal.prelimi- narystudyofthesafetyandefficacyof sc-58635,anovelcyclooxygenase2 inhibitor:efficacyandsafetyintwoplace- bocontrolledtrialsinosteoarthritisand rheumatoidarthritis,andstudiesof gastrointestinalandplateleteffects.arthritis rheum.1998;41:1591–1602.3. wintherk,reine,kharazmia.Theanti-in- flammatorypropertiesofrose-hip.inflam- mopharmacology.1999;7:63–68.4. kharazmia,wintherk.rosehipinhibits chemotaxisandchemiluminescenceof humanperipheralbloodneutrophilsinvitro andreducescertaininflammatoryparame- tersinvivo.inflammopharmacology. 1999;7:377–386.

5. kendallfp,mccrearyek.muscles:Testing andfunction.3rded.baltimore:williams& wilkins;1983.6. bliddalh,rosetzskya,schlichtingp,etal.a randomized,placebo-controlled,crossover studyofgingerextractsandibuprofenin osteoarthritis.osteoarthritiscartilage. 2000;8:9–12.7. altmanrd,markussenkc.effectsofa gingerextractonkneepaininpatientswith osteoarthritis.arthritisrheum.2001;44:2531– 2538.8. blotmanf,maheue,wulwika,etal. efficacyandsafetyofavocado/soybean unsaponifiablesinthetreatmentof symptomaticosteoarthritisoftheknee andhip.aprogressive,multicenter,three- month,randomised,double-blind, placebo-controlledtrial.revrhumengled. 1997;64:825–834.9. endress,ghorbanir,kelleyve,etal.The effectofdietarysupplementationwith n-3polyunsaturatedfattyacidsonthe synthesisofinterleukin-landtumornecrosis factorbymononuclearcells.nenglJmed. 1989;320:265–271.10.mustafaT,srivastavakc,Jensenkb.drug developmentreport(9):pharmacology ofginger,zingiberofficinale.Jdrugdev. 1993;6:25–39.11.kiuchif,iwakamis,shibuyam,etal. inhibitionofprostaglandinandleukotriene biosynthesisbygingerolsanddiarylhepta- noids.chempharmbull(Tokyo).1992; 40:387–391.12.wintherk.impactoncoagulation,platelet functionandfibrinolysis.in:proceedings ofthe3rdinternationalexhibitionand conferenceonnutraceuticalsandfood forvitality;may3–5,2000;geneva,switzer- land.abstract.13.wintherk,kharazmia,reine.rose-hip, givenasastandardiseddrypowder,exerts anti-inflammatoryandcellpreserving propertiesinhumans.presentedatthe2nd internationalcongressoncoronaryartery disease;october18–21,1998;florence, italy.14.brandc,snaddonJ,baileym,cicuttinif. vitamineisineffectiveforsymptomatic reliefofkneeosteoarthritis:asixmonth doubleblind,randomised,placebo controlledstudy.annrheumdis. 2001;60:946–949.15.brandtk,a°kessonb.healthpromoting compoundsinvegetablesandfruit.plant production.2002;29:43–44.

2726%

of P

atie

nts

degree of Joint Pain relief

srhP (n=48)

Placebo (n=48)

2828

TheantI-Inflammatory ProPertIesofrose-hIPK. WInther 1,*, e. reIn1 and a. KharazmI2

ldepartmentofclinicalchemistry,koldinghospital,kolding,denmark2departmentofclinicalmicrobiology,universityhospital(rigshospitalet),copenhagen,denmark

received2december1998;revised4february1999;accepted5february1999

abstract – Theanti-inflammatorypropertiesofrose-hiparedescribedinthisshortreport.rose-hipextract reducedchemotaxisofperipheralbloodneutrophilsandmonocytesofhealthy subjectsinvitro.dailyintakeofrose-hippowderforfourweeksbyhealthyvolunteersandpatientssufferingfromosteoarthritis,resultedinreducedserumc-reactiveprotein(crP)levelsandreducedchemo-taxis ofperipheralbloodneutrophils.Theresultsindicatethat rose-hippossessesanti-inflammatorypropertiesandmightbeusedasareplacementorsupplementforconventionaldrugtherapiesinpatientswithosteoarthritis.

1. IntroductIonTherehavebeenundocumentedlayclaimsthatrose-hip,normallyknownfor itshighvitaminccontent,mayreducethepaininpatientssuffe-ringfromosteoarthritis.wehaverecentlyshownthat rose-hip extract reduced the chemotaxisofperipheralbloodpolymorphonuclearleuko-cytes(pmns)andmonocytesinvitro[1].Thisac-tivitywasindependentofthevitaminccontentof rose-hip.furthermore, the levelofcrpandthechemotaxisofneutrophilswerereducedinhealthysubjectsunderrose-hiptreatment.Thepurposeofthisstudywastoinvestigatewhetherthe natural product rose-hip, administered asdrypowdertovolunteersofwhichfourweresuf-feringfromclinicalosteoarthritis,hadanyeffectontheclinicalsignsandsymptomsandcertaininflammatoryparameters.

2. suBJects and methods2.1. subjectseightmalevolunteers,freefromanyknownall-ergic,hepatic,cardiovascularor infectiousdi-seases,meanage52years(range47-62),wereenteredintothestudy.fourofthemhadneverexperiencedanypainofmuscularorjointorigin.The other four had all been engaged in hardphysicalworkindifferentareasofconstructionformostoftheiradultlife.onehadsufferedfromclinical osteoarthritis for more than 20 years,withpainespeciallyinthekneeandelbow.Thepainhadbeenalleviatedbyinjectionsofstero-iddirectly intothe jointsandbyacetylsalicylicacid and nonsteroid anti-inflammatory drugs(nsaids).Thesecondpatienthadosteoarthritisandmoderatepaininthekneeandtheankle,periodicallyrelievedbyacetylsalicylicacid.The

thirdpatienthadpainfromosteoarthritisoftheankleandhadbeenperiodically treatedwithnsaidsandacetylsalicylicacid.Thefourthpa-tienthadosteoarthritisoftheelbowandshoul-derof10yearsduration,normallytreatedwithaspirin or paracetamol. The volunteers weretreatedwith45grams(highdose)ofhybenvi-talrose-hipdailyforfourweeks.Thetreatmentwaswithdrawnforatleastonemonth,thenfol-lowedbyanother treatment for fourweeksatadailydoseof10grams (lowdose).rose-hipwastakentogetherwithamainmeal.afterfourweeksof thehighdoserose-hip intake,at theendoftreatment-freeintervalsandattheendofthelowdoseintake,thevolunteerswereas-kedaboutthepossibleside-effects,andbloodsamples were collected for clinical chemistryandpmnchemotaxis studies.allbloodsamp-leswere takenbetween8 : 30and9 : 00ambythesamelaboratorytechnicianafter30mi-nutes of rest, and analyzed immediately. forchemotaxis,heparinizedbloodwastakenusingvacuotainers. The time-lapse between bloodsamplingandchemotaxiswasthesameforthepatientsandcontrolsubjects.

2.2. rose-hiprose-hippowderofrosacaninawaskindlypro-videdbyhybenvital,langeland,denmark. The rose-hip powder used in thesestudieswasawellcharacterizedandstandardizedbatchcontainingbothseedsandshell.duringthedryingprocedureoftherose-hippowder,thetemperatureneverex-ceeded40°cfortheinvitrostudies,awaterextractofrose-hipwasprepared.Theextractiontookplaceat4°c

2.3. crP determinationserum crp was estimated by a turbidometricmethodusingahitachi717turbidometer.crpantiserumwasfromoriondiagnostica,helsinki,finland.crpdilutionbufferandhumancrpca-librator was purchased from daco a/s, glos-trup,denmark.Thenormalrangeinourlabora-toryis≥10mg/1.

2.4. chemotaxischemotaxis was carried out using a modifiedboydenchamberassay[2].fortheinvitrostu-dies, pmns isolated from peripheral blood ofthe subjects were preincubated with variousdilutionsof rose-hipextract for30 rainat37°cfollowingpreincubation,chemotaxisofthecel-lstowardsthechemotacticpeptidef-met-leu-phe(tmlp)ataconcentrationof10-5morzymo-sanactivatedserum(zas)atadilutionof1:200wastested.fortheinvivostudies,thechemota-xisofperipheralbloodneutrophilsfromhealthycontrolsubjectsandpatientstowardsfmlpandzas was determined. The migrated ceils werecountedbyacomputer-assistedimageanaly-sissystem.

2.5. statistical analysisstatistical analysis of the data was performedby using the wilcoxon test for matched pairs.alldataaregivenasmean±sem.pvaluesof≤0.05wereconsideredsignificant.

2 –

4 –

6 –

8 –

10 –

0High dose

No treatment Low dose

figure 1. c-reactive protein levels in the serum of eight volunteers given high dose, low dose or no treatment. the values are given as mean ± sem in rag/1. there were sta-tistically significant differences between crP levels from the high dose group and no treatment (p ≤ 0.02) and low dose group and no treatment (p ≤ 0.05).treatment.

3. results and discussionrose-hipextractatconcentrationsaslowas100µg/mlinhibitedthechemotaxisofpmnsinvitro(datanotshown).cellviabilityafterincubationwithrose-hipextractwasgreaterthan98%.asshowninfig.1,serumcrplevels,althoughwi-thin normal range, declined significantly bothinthehigh-dose(p≤0.02)andinthelow-dosegroups(p≤0.05)ascomparedtotheno-thera-pygroup.Thecrp levels (mean±sem) in thepatientgroupwere5.75±2.95,6.67±2.67and8.25 ± 4.98 with high dose, low dose and notherapy, respectively. in thecontrolgroupthecrplevelswere4.75±0.75,4.00±0.0and7.25±1.03withhighdose, lowdoseandnothera-py, respectively. The neutrophil chemotaxisdataareshowninTable1.chemotaxistowardsfmlpdeclinedbyapproximately60%and50%,inthehighdoseand lowdosegroup, respec-tively,withpvaluesof0.01and0.02.chemota-xis towardszasalsodeclined inboththehighdose (p ≤ 0.01) and in the low dose groups(p ≤ 0.02). The decline in chemotaxis of cellsfromthepatientsandthecontrolsundertreat-ment with rose-hip was similar. The decline inchemotaxiswasobservedinallthe8subjects.Themean±semvaluesforfmlpwere187±60compared to 370 ± 39 and for zas 414 ± 136comparedto673±27inthehighdosepatientgroupcomparedwithno therapy. in thehighdosecontrolgroupthemean±semresponsetofmlpwas101±45ascomparedto308±22andtozas272±125ascomparedto6004-49ascomparedtonotherapy.

The salient finding of the present study is thatrose-hip,givenasdrypowder,loweredcrple-vels significantly and inhibited chemotaxis ofperipheral blood neutrophils in human malevolunteers. To our knowledge, this finding hasnotbeen reportedbefore.Therearevery fewreports in the literature on other properties ofrose-hip.rose-hiphasbeenusedassourceofvitamin c in tea and other products [3]. cellssuchaspolymorphonuclearleukocytes(pmns)andmonocytesareinvolvedintheinflammato-ryprocessandtissuedamageininflammatorydiseasessuchasarthritisandatherosclerosis[4].Thedamageiscausedbythereleaseofprote-olytic and hydrolytic enzymes as well as toxicoxygen radicals [5]. acetylsalicylic acid, non-steroidanti-inflammatorydrugsandglucocorti-coidshavebeenusedforthetreatmentofthe-sediseases[6,7].Thesedrugshaveavarietyofsideeffectssuchasgastricerosionandkidneydisturbances. The present study demonstratesthatadministrationof rose-hip topatientswithosteoarthritis, diagnosed on a clinical basis,

3130

inconclusion,theanti-inflammatoryandpain-relieving properties of the natural product ro-sehip,combinedwith its safety, lowpriceandease of administration, provide an attractivestrategytouserose-hipaspartofasupplementtoatherapeuticregimenfor osteoarthritis. a large scale placebo-con-trolledclinicalstudywillberequiredtoextendconfirmationoftheanti-inflammatoryeffectofrose-hip.

acknowledgements

Technical assistance of kirsten mossin, hanneTamstorf and anne asanovski and support ofthedanishrheumatismassociationisacknow-ledged.

references1. k.winther,a.kharazmiandb.rangaard (1997).cellpreservingandantiinflammato- rypropertyofrose-hip(hybenvital).pos- sibleclinicalimplication,1st.int.congress oncoronaryarterydiseases:frompreven- tiontointervention,p.68.prague,czech republic.2. eJensenanda.kharazmi(1991). computer-assistedimageanalysisassay ofhumanneutrophilchemotaxisinvitro, J.immunol.method144,43-48.3. a.leungands.foster(1996). encyclopediaofcommonnatural ingredients.Johnwiley,newyork.4. em.ridker,m.cushman,m.J.stampfer, r.eTraceyandc.h.hennekens(1997). inflammation,aspirinandtheriskof cardiovasculardiseasesinapparently healthymen,neweng.J.med.336,973-979.5. e.d.harris,Jr.(1988).pathogenesisof rheumatoidarthritis:adisorderassociated withdysfunctionalimmunoregulation,in: inflammation:basicprinciplesandclinical correlates,J.h.gallin,i.m.goldsteinand r.snyderman(eds),pp.751-773.raven press,newyork.6. m.c.hochberger,r.d.altman,k.d. brandt,b.m.clarck,ra.dieppe,m.r. griffin,r.w.moskowitzandT.J.schnitzer (1995).guidelinesforthemedical managementofosteoarthritis(partone), arthritisrheum.38,1535-1540.7. m.c.hochberger,r.d.altman,k.d. brandt,b.m.clarck,ea.dieppe,m.r. griffin,r.w.moskowitzandT.J.schnitzer (1995).guidelinesforthemedicalmanage- mentofosteoarthritis(parttwo),arthritis rheum.38,1541-1546.

8. y.matzner,r.drexlerandm.levy(1984). effectofdipyrone,acetylsalicylicacidand acetaminophenonhumanneutrophilche- motaxis,eur.J.clin.invest.14,440-443.9. a.kempandJ.smith(1987).Theeffectof salicylateonhumanleukocytemigration, clin.exp.immunoi.49,233-238.10.i.rivkin,v.foschiandc.h.rosen(1976). inhibitionofinvitroneutrophilchemotaxis andspontaneousmotilitybyanti-inflamm- atoryagents(39518),ptvc.soe.exp.biol. med.153,236-240.

11.h.b.kaplan,h.s.edelson,h.m.korchak, w.rgiven,s.abramsonandg.weismann (1984).effectofnon-steroidalanti-inflamm- atoryagentsonhumanneutrophilfunctions invitroandinvivo,biochem.pharmacol. 33,371-378.12.e.g.maderazo,s.ebreauxandc.l. woronick(1984).inhibitionofhuman polymorphonuclearleukocytecellresponse byibuprofen,J.pharm.sci.73,1403-1406.

reducedthelevelsoftheacutephaseproteincrpandperipheralbloodneutropbilchemota-xis.similarresultswerefoundinthefourhealthysubjects who had never experienced pain ofosteoarthritisorigin.symptomswereassessedaspainseverityonascaleof1-10andchangeinlimitationofjointmovement.alleviationofphy-sicalsymptomsbyrose-hipinthepatientscorre-latedverywellwiththereducedchemotaxisofperipheralbloodneutrophilsandreducedlevelofcreafterthevolunteersstoppedtakingrose-hip,thechemotaxisofneutrophilsandthelevelsofcrprosetotheuntreatedvalues.itisinteres-tingtonotethattheinitialcrpvalueswerehig-herinthepatientthanthecontrolgroup.Thein-hibitionofchemotaxisobservedinourstudywascomparabletothatobservedwithacetylsalicy-licacidasreportedbymatzneretal.[8].ontheotherhandkempandsmith[9]showedthatin-cubationofneutrophilsinvitrowithsodiumsali-cylateincreasedthechemotaxisofthesecells.a similar increased response was observed innormalindividualsafteringestionofsodiumsa-licylate[9].somenon-steroidanti-inflammatorydrugs such as ibuprofen at in vivo obtainableconcentrationsinhibitedneutrophillocomotionby50%,similartoourfindingswithrose-hip[10-12].Thepatientswhocomplainedofmildpainofosteoarthritisorigin, reported that theirpaindeclined after 14 days of rose-hip intake. Thepainrelievingeffectofrose-hipinthesepatientswascomparabletothatofnsaidandacetylsa-licylicacid.inallcasesthepainreturned12-14daysafterstoppingintake.noallergicreactionsorgastrointestinaldisturbanceswereobservedduringtherapy.Therewasnomajordifferencebetweenthepainalleviatingeffectofrose-hipgivenatthetwodifferentdoses.Threepatientshadtotalpainrelieffromrose-hipandwereun-abletodistinguishthedifferencebetweenthehighdoseandthelowdose.however,onepati-entfeltthathighdosegavehimtotalreliefwhe-reaslowdosedecreasedthepaindramaticallybutnotcompletely.

high dose

chemotaxis(fmlp)chemotaxis(zas)

144±38.7a

343±89.7a

table 1. chemotaxis of peripheral blood neutro-phils from the eight volunteers at the end of high dose intake, 28 days after cessation of intake and at the end of low dose intake of rose-hip powder. the results are given as mean ± sem

no therapy

339±24.0637±29.3

low dose

172±18.0b

432±39.9b

a comparison of high dose with no therapy p ≤ 0.01.b comparison of low dose with no therapy p ≤ 0.02.

rose hIP InhIBIts chemotaxIsandchemIlumInescenceofhuman PerIPhe-ral Blood neutroPhIlsinVItroandre-duces certaIn Inflammatory Parameters In VIVoarsalan KharazmI1* and KaJ WInther 2

1departmentofclinicalmicrobiology,rigshospitaletafsnit7806,Tagensvej20,dk-2200copenhagen,denmark2departmentofclinicalchemistry,koldinghospital,kolding,denmark

received25may1999;revised12July1999;accepted15July1999

abstract – objectiveanddesign:Theobjectiveofthisstudywastoinvestigatetheleucocyte-rela-tedantiinflammatorypropertiesofrosehip.

materials and methods:Theeffectofrosehiponanumberofinflammatoryparameterswaseva-luatedusingthefollowingmodels:(1)Theeffectofrosehipextractonchemotaxisandchemilu-minescenceofperipheralbloodpolymorphonuclearleucocytes(pmns)fromhealthysubjectsinvitro;(2)Theeffectofrosehipadministeredtohealthysubjectsonserumlevelsofcreatinineandc-reactiveproteinandonchemotaxisandchemiluminescenceofperipheralbloodpmns.

results: rosehipextractatconcentrationshigherthan500µg/mlinhibitedthechemotaxisandchemiluminescenceofperipheralbloodpolymorphonuclearleucocytesinvitro.dailyintakeofrosehippowderatdosesof45gramsorlowerbyhealthysubjectsresultedinreducedchemota-xisofperipheralbloodpmnsandreducedthelevelofserumcreatinineandacutephaseproteincrp.

conclusions:rTheseresultsindicatethatrosehippossessesantiinflammatorypropertiesandmightbeusedasareplacementorsupplementforconventionaldrugtherapiesinsomeinflammatorydiseasessuchasarthritis.

keywords:rosehip;rosacanina;neutrophil;chemotaxis;crp;antiinflammatory.

3332

1. Introductioninflammatory diseases such as arthritis involveabroadspectrumofdifferentclinicalmanifes-tations.inflammatorycellssuchaspolymorpho-nuclearleucocyteshavebeenshowntobein-volved in the inflammatoryprocessand tissuedamage. inflammatory cytokines such as Tnfappear to be involved in the amplification ofthediseaseprocess.Thedamageiscausedbythereleaseofproteolyticandhydrolyticenzy-mes as well as toxic reactive oxygen radicalsfromthesecellsactivatedinthetissueandjoints(harris,1988).Therapyofinflammatorydiseasesinvolvesalleviationofthesymptomsassociatedwiththedisease,suchas,reliefofpain,reduc-tion of inflammation and increase of motion.acetylsalicylicacid(aspirin)andothernon-ste-roidanti-inflammatorydrugssuchasibuprofenmethotrexate and naproxen, and glucocorti-coidshavebeenusedforthetreatmentofar-thritis(hochbergeretal.,1995a,b;ridker,etal.,1997).controlofthesymptomswiththesedrugsrequireslongtermdailytreatment.Thesedrugshaveavarietyof toxicandothersideeffects,suchasgastricerosionandadverseeffectsonkidneysandliver.someofthesedrugs,particu-larlytheglucocorticoids, inhibittheimuunere-sponsetoinfections.Therefore,thereisagreatneedforalternativetherapiesforthemanage-mentofarthritiswhichcaneliminatetheneedfor conventional drugs and their side-effects,particularly for prolonged daily use. in a shortcommunicationwehavereportedontheanti-inflammatoryactivityofrosehipinfoursubjectssufferingfrommildosteoarthritis(wintheretal.,1999).Thepurposeof thisstudywasto investi-gate inmoredetail theanti-inflammatorypro-perty of the natural product rose hip, utilizinginvitromethods ina largernumberofhealthysubjects.

2. materials and methodes2.1. rose hipTheextractwaspreparedbyincubating80mgofhybenvital rosehip (langeland,denmark)drypowderfromrosacaninawith4mlofmi-nimal essential medium (mem) containing 50units/mlofpenicillinand0.05mg/mlofstrepto-mycin, for19hat4°c. Theextractswerepre-pared from either the whole fruit powder, theshells or the seeds. The shells and the seedswere separated from each other by splittingthe dried fruit and separating the shells fromtheseedsmanually.Theywerethengroundinamortar.chemicalanalysesofhybenvitalrosehipwasperformedbysteinslaboratoriuma/s,holstebro,denmark.followingincubationofthepowdersinmem,themixtureswerecentrifuged

at4000rpmfor10min.Thesupernatantswerecollected,sterilefilteredanddilutedfurther.Thephofextractpreparationswasadjustedtoph7.2beforeuse.

2.2. chemotaxisThe chemotaxis assay was performed using amodifiedboydenchambertechniqueasprevi-ously described (Jensen and kharazmi, 1991).pmns isolated from peripheral blood of heal-thy subjects were preincubated with differentdilutionsofrosehipextractfor30minat37°c.following preincubation, the chemotaxis ofthe cells towards the chemotactic peptide f-met-leu-phe (fmlp) or zymosan activated se-rum(zas)weretested.Themigratedcellswerecountedbyacomputer-assistedimageanaly-sissystem.

2.3 chemiluminescencechemiluminescence assay was used as ameasureofoxygen radicalgenerationbyac-tivated pmns. The method was performed aspreviously described (kharazmi et al., 1984).pmns were preincubated with different dilu-tions of rose hip extract and then stimulatedwith either fmlp or opsonized zymosan. Theoxidativeburst responseof theactivatedcellswasmeasuredbyaluminometer(1250-lkbwal-lace).

2.4. subjectsThirteen healthy volunteers represented bybothsexeswithameanageof47years(range30-59years)wereincludedinthisstudy.allthesubjects includedwerewithoutknowncardio-vascular, immunological, kidney, liver, allergic,rheumatological or haematological disorders.Thevolunteersweretreatedwith45gofhybenvital rosehipdaily for28days(highdose), fol-lowedforanother28daysduringwhichhybenvitalrosehipwasnottaken.attheendofthisperiod, thevolunteers receivedanother treat-mentofrosehipatadoseof10gdailyfor28days(lowdose).beforeinclusion,allvolunteerswent through a screening procedure to assu-rethatnoneoftheabovementioneddiseaseswerepresent.moreover,beforeinclusionbloodsamples were taken for c-reactive protein(crp)measurementtoassurethatnoneoftheincludedvolunteerssufferedfromunknown in-fectiousdiseases.

2.5. Blood chemistrybloodpotassium,sodium,serumcreatinine,ala-nine aminotransferase, alkaline phosphatase,lactate dehydrogenase, bilirubin, hemoglo-bin and total cholesterol were also measured

beforeinitiationofthetreatment.allmeasure-mentswereperformedaccordingtothecon-ventional laboratory routine. all the abovementionedparametersexceptserumcreatini-neandcrpwererepeatedafter5,10,21and28 days of treatment, 28 days after stoppingrosehiptherapyandagainat theendof lowdosetreatment.serumcreatinineandc-reac-tiveprotein(crp)weretestedbeforetherapy,after10and28daysoftreatment,28daysfol-lowingcessationofthetreatmentandfinallyatthe end of low dose treatment. rose hip wastakentogetherwithamealat12.00noon.onthedaysofbloodsamplingrosehipwastakentogether with a light meal at 09.00 a.m., twohours before blood sampling. blood samplingwasalwaysperformedafter15minat rest sit-tinginachair.

2.6. statistical analysisstatistical analysis of the data was performedbyusingwilcoxontestformatchedpairs,pva-luesof<0.05wereconsideredsignificant.

3. results3.1. analysis of rose hipTable 1 shows the chemical analyses of thecommercially available hyben vital rose hip.rosehippowdercontainsproteins,carbohyd-rates,alowamountoffatandseveralvitaminssuchasvitamina,vitaminb,vitaminc,vitamineandvitamink.Thepowderalsocontainsseve-ralminerals.uptakeofvitamincpresentinhy-benvitalpowderwasasgoodas,orevenbet-terthan,thatofvitamincwhengivenintabletform.Theconcentrationsandkineticsofuptakethroughthegastrointestinaltractoftheequiva-lentof250mgvitamincinrosehippowderwassimilarto500mgvitaminc intabletform.Thebetterabsorptionofvitamincinrosehippow-dermaybeduetoalargersurfaceareaofrosehippowderascomparedtovitaminctablets.

3.2. chemotaxisinitial dose-response experiments were perfor-medanditwasfoundthattheextractofrosehip at concentrations equivalent to 500µg/mlandhigherinhibitedchemotaxisofpmnsinvit-ro.asshown infig.1, rosehipextractatcon-centrations of 500 µg/ml and higher inhibitedchemotaxis of human peripheral blood neu-trophils; ph-adjusted rose hip extract at theseconcentrationswasasstronglyinhibitoryasthenon-ph-adjustedrosehipextract.Thetwoma-jorpartsof rosehip– shellsand seeds --weretestedseparatelyfortheiractivityonpmnche-motaxis. it was shown that by far most of theinhibitory activity resided in the shells (fig. 1).

Theinhibitionofchemotaxisbyrosehipshellsatboth the1000µg/mland500µg/ml levelswassignificantlyhigherthanthatofseeds(p≤0.01and p ≤ 0.04, respectively). when comparingrosehipshellswithwholepowder,therewassig-nificantly higher inhibition by rose hip shells at500µg/ml(p≤0.03)butnotat1000µg/ml.

3.3. chemiluminescenceas shown in Table2, rosehipextract inhibitedthechemiluminescenceofpmnsactivatedbyopsonizedzymosan.adjustmentofphtophysi-ologicalvaluesintheextractdidnotinfluencetheinhibitoryeffectmarkedly.vitamincincrys-tallineformusedascontroluptoaconcentrati-onof5000µg/mlhadalmostnoeffectonpmnchemiluminescencewhenthephofvitamincsolution was adjusted to the physiological ph7.2. vitamin e (alpha-tochopherol) was alsoused as a known antioxidant control. vitamineataconcentrationofover 1µg/ml inhibitedchemiluminescence.

figure 1. effect of rose hip extract on polymorpho-nuclear leukocytes (Pmn) chemotaxis in vitro. cells were preincubated with various concentra-tions of rose hip powder as given in the x-axis for 30 min. the data are presented as percent inhibi-tion of Pmn chemotaxis for each subject tested.

shell wholepowder

20 –

40 –

80 –

60 –

100 –

0

1000 1000 1000500 500 500

%in

hib

itio

n

seeds

µg/ml

proteincarbohydratefatvitamincenergy

table 1. chemical analyses of the commercially available hyben Vital rose hip powder. the va-lues are given for 100 g of dry powder

6.2g39.0g4.0g

560mg916kJ

4. ex vivo studies 4.1. Blood chemistryno significant changes occured in potassium,sodium, alanine aminotransferase, alkalinephosphatase,lactatedehydrogenase,bilirubin,haemoglobinortotalcholesterolcomparingva-luesfrombeforeintaketovaluesobtainedafter5,10,21and28daysofhighdosetherapy,va-luesobtained28daysafterstoppingintakeandthoseobtainedattheendoflowdosetherapy(datanotshown).

serumcreatinine,however,declinedsignificant-ly compared with initial value given as mean4±sem(90.0±2.1µmol/1)tovaluesobtainedafter10days:(87.4±1.8µmol/1)and28days(84.94±1.9µmol/1)ofintake,respectively3(p<0.001).whentreatmenthadbeenstoppedfor28daystheserumcreatininelevelssignificantlyincreased (93.2 ± 1.9µmol/1) (p < 0.001) andweresimilartovaluesobtainedbeforeintake.

Thedataonc-reactiveproteinaregiveninfig.2. similar to the findings on serum creatinine,crpvalueswerealsodecreasedduringintakeof rosehip. The initialmean4±semvaluesofcrpwere5.38±0.4mg/landdeclinedto3.31±0.49mg/and4.31±0.47mg/l,after10and28daysofintakerespectively(p<0.05).afterstop-ping therapy for28days, the levels increasedto5.75±0.54mg/l(p<0.051ascomparedwiththatpreviously.

2 –

4 –

6 –

7 –

0 Day 0 Day 10 Day 23 Day 63

50 –

100 –

150 –

200 –

250 –

300 –

350 –

0 No treatment Low dose High dose

cr

p(

mg

/l)

figure 2. levels of serum c-reactive protein (crP) as given in mg/l from subjects on the start (day 0), 10 days and 28 days during intake of rose hip and 35 days after cessation of treatment (day 63) with 45 g daily intake of rose hip. the results are given as mean 4 ± sem values from 13 subjects. the mean value on day 10 was significantly lower than that on day 0 and day 63 (p ≤ 0.05).

2 –

4 –

6 –

7 –

0 Day 0 Day 10 Day 23 Day 63

50 –

100 –

150 –

200 –

250 –

300 –

350 –

0 No treatment Low dose High dose

pm

nc

he

mo

taxi

s(f

mlp

)

4.2. chemotaxispmnchemotaxisintheperiodduringwhichthevolunteershadnottakenanyrosehippowderwas compared with values obtained in thepreceeding28days(figs3and4).Themean±semvalueofpmnchemotaxistowardstheche-motacticpeptidefmlpwas103.6±60.0whentestedonday28oftreatmentwithrosehipascomparedwith298.9±-26.2whenbloodsam-plesweretaken28daysaftercessationofrosehip intake(p<0.001).Themean±semvaluesfor pmn chemotaxis towards zymosan activa-tedserum(zas)whichcontainsthebiologicallyactivechemotacticfactorc5awas218±60.0as compared to 529.9 ± 39.9 when tested 28daysaftercessationoftreatmentwithrosehip(p<0.001).Thedeclineinchemotaxisresponseto fmlp was 65% in 12 out of 13 volunteers: aconsiderabledeclineofchemotaxis response.The decline in chemotactic response to zaswas59%,alsoaconsiderabledeclinein12outof13volunteers. itwas thesamesubjectwhodidnotrespondtotherapyinbothassays.

figure 3. chemotaxis of peripheral blood poly-morphonuclear leukocytes (Pmn) from subjects during and 28 days after cessation of treatment with 45 g (high dose) or 10 g (low dose) daily in-take of rose hip for 28 days or no treatment. the chemotaxis is determined towards the chemo-tactic petide fmlP. the results are given as mean ± sem number of ceils migrated from 13 subjects. the mean chemotaxis values for both low dose and high dose were significantly lower than that for no treatment group (p ≤ 0.01 and p ≤ 0.001, respectively).

4.3. clinical findings

noallergic reactionsoranyother side-effectswereobservedduringtherapy.onlytwovolun-teers complained of mild gastrointestinal gasdisturbancesattheendofthestudy,whileonthehighdose.

2 –

4 –

6 –

7 –

0 Day 0 Day 10 Day 23 Day 63

50 –

100 –

150 –

200 –

250 –

300 –

350 –

0 No treatment Low dose High dose

pm

nc

he

mo

taxi

s(f

mlp

)

figure 4. chemotaxis of peripheral blood poly-morphonuclear leukocytes (Pmn) from subjects during and 28 days after cessation of treatment with 45 g (high dose) or 10 g (low dose) daily in-take of rose hip for 28 days or no treatment. the chemotaxis is determined towards zymosan activated serurm (zas). the results are given as mean ± sem number of cells migrated from 13 subjects. the mear chemotaxis value for high dose was significantly lower than that for no treat-ment group (p ≤ 0.00l),

5. dIscussIon

The studies described in this communicationdemonstrate that theextract fromrosehip in-hibited, invitro, thechemotaxisandoxidativeburst responseof thehumanperipheralbloodpolymorphonuclearleucocytes,importantandabundantinflammatorycellsinvolvedinthepa-thogenesisofarthritis.furthermore,administrati-onofrosehiptohealthyvolunteersforaperiodof28daysinhibitedthechemotacticresponseofneutrophilsbyapproximately60%orhigher.moreover, rosehip lowered the levelof serumcreatinine and the acute phase protein c-re-active protein in volunteers with values withinnormal range, which is below 10 mg/l. serumcreatininelevelswerewithinthenormalrangeinallthevolunteers(males55-125andfemales45-100µmol/l).however,thedeclinewasstatis-ticallysignificantandmightindicateenhancedglomerularfiltration.Thebloodchemistrydatapresented in this study showed that intake ofrose hip had no harmful effect on any of theliveifunctionsdeterminedinthisstudy.

studiesontheinhibitionofneutrophiloxidativeburstresponsebyrosehipextractshowedthatthiseffectwasnotduetovitaminccontentoftheextract.Thisisshownbytheinabilityofph-ad-justedvitamincto inhibitchemiluminescencewhereasphadjustedrosehipextarctwasstillasinhibitory as non-ph-adjusted extract. in orderto determine which part of rose hip exhibitedtheinhibitoryeffectonchemotaxistheextractfromshells,seedsandthewholepowderwerepreparedandtestedinpmnchemotaxisassay.as shown in fig. 1 the major inhibitory activitywasfoundtoresideintheshells.itwillbeinteres-tingtoidentifythecompound(s)responsiblefortheanti-inflammatoryactivityofrosehip.The inhibition of chemotaxis observed in ourstudy was comparable to that observed withacetylsalicylicacidas reportedbymatzneretal.(1984).ontheotherhandkempetal.(1982)showed that incubation of neutrophils in vitrowithsodiumsalicylateincreasedthechemota-xisofthesecells.similarincreasedresponsewasobserved in normal individuals after ingestionofsodiumsalicylate(kempetal.,1982).somenon-steroid anti-inflammatory drugs such asibuprofenatattainableconcentrationsduringtherapy has been shown to inhibit neutrophillocomotion by 50%; a finding which is similartoourfindingswithrosehip(rivkinetal.,1976;kaplanetal.,1984;maderazoetal.,1984).

3534

rose hip extract concs(µg/ml)

25001000500

3780

1 subject 2 subject

2780

3 subject

5712

nd

table 2. effect of rose hip extract on human periphe-ral blood polymorphonuclear leucocyte (Pmn) chemiluminescence. the data are presented as percent inhibition as compared with control

nd: not determined.

6. conclusIon

rosehippossessesanti-inflammatoryandanti-oxidantproperties.Thesepropertiesareimpor-tant in alleviation of tissue damage in the in-flammatorysites.asanaturalproduct,rosehiphasnosideeffects,issafeandcanbeadminis-teredeasily. itcanbedesignedfordailycon-sumptionassupplementalpartofatherapeuticregimenforsomeinflammatorydiseases,orasaprophylacticregimenforindividualshavingageneticorenvironmentalpredispositiontothe-se diseases. a large scale placebo-controlledclinical study will be required to extend con-firmationoftheantiinflammatoryeffectofrosehipdescribedinthisreport.

acknowledgementsexpert technicalassistanceofanneasanovskiandmounirainouzisacknowledged.ThisworkreceivedpartialfinancialsupportfromTheda-nishrheumatismassociation.

referencesharris,Jr.,e.d.(1988).pathogenesisof rheumatoidarthritis:adisorderassocia- tedwithdysfunctionalimmunoregulation. in:inflammation.basicprinciplesand clinicalcorrelates,gallin,J.h.,goldstein, i.m.andsnyderman,r.(eds),pp.751- 773.ravenpress,newyork.hochberger,m.c.,altman,r.d.,brandt, k.d.,clarck,b.m.,dieppe,ea.,griffin, m.r.,moskowitz,r.w.andschnitzer,T.J. (1995).guidelinesforthemedicalmana- gementofosteoarthritis(part1),arthritis rheum.38,1535-1540.hochberger,m.c.,altman,r.d.,brandt,k.d., clarck,b.m.,dieppe,ea.,griffin,m.r., moskowitz,r.w.andschnitzer,T.J.(1995). guidelinesforthemedicalmanagementof oseteoarthritis(part2),arthritisrheum.38, 1541-1546.Jensen,eandkharazmi,a.(1991).computer- assistedimageanalysisassayofhuman neutrophilchemotaxisinvitro,J.lmmunol. methods144,43-48.kaplan,h.b.,edelson,h.s.,korchak,h.m., given,w.e,abramson,s.andweismann, g.(1984)effectofnon-steroidalanti-in- flammatoryagentsonhumanneutrophil functionsinvitroandinvivobiochem. pharmacol.33,371-378.kemp,a.andsmith,J.(1982).Theeffectof salicylateonhumanleukocytemigration, clin.exp.lmmunol.49,233-238.kharazmi,a.,hØiby,n.,döing,g.andvalerius, n.h.(1986).pseudomonasaeruginosa exoproteasesinhibithumanneutrophil

chemiluminescence,infect.immun.44, 587-591.maderazo,e.g.,breaux,s.p.andworonick, c.l.(1986).inhibitionofhumanpolymor- phonuclearleukocytecellresponseby ibuprofen,J.pharm.sci.73,1403-1406.matzner,y.,drexler,r.andlevy,m.(1984). effectofdipyrone,acetylsalicylicacidand acetaminophenonhumanneutropbal chemotaxis,eur.J.clin.invest.14,440-443.ridker,p.m.,cushman,m.,stampfer,m.J., Tracey,r.p.andhennekens,c.h.(1997) inflammation,aspirinandtheriskof cardiovasculardiseasesinapparently healthymen,n.eng.J.med.336,973-979.rivkin,i.,foschi,v.androsen,c.h.(1976). inhibitionofinvitroneutrophilchemotaxis andspontaneousmotilitybyanti-inflamm- atoryagents(39518),proc.soc.exp.biol. med.153236-240.winther,k.,rein,e.andkharazmi,a.(1999). Theanti-inflammatorypropertiesofrose-hip inflammopharmacol.7,63-68.

36

OVERVIEW ON ABSTRACTSreine.,kharazmia.,Thamsborgg.,wintherk.(2004):aherbalremedy,madefromasubspeciesofrose-hiprosacanina,reducessymptomsofkneeandhiposteoarthritis.abstract.9thworldcon-gressoftheosteoarthritisresearchsocietyinternational,chicago.

warmholmo.,skaars.,hedmane.,molmerh.,eikl.(2001):alternativeTherapiesinhealthandmedicine.firstinternationalconferenceoncomplementaryalternative&integrativemedicineresearch,sanfrancisco.

wintherk.,kharazmia.,reine.(2005):apowdermadefromasubspeciesofrosehip(rosacani-na),reduceswomacsymptomsscoreaswellascholesterollevelinpatientssufferingfromoste-oarthritis.abstract.10thworldcongressonosteoarthritis,boston.

wintherk.,kharazmia.(2004):apowderpreparedfromseedsandshellsofasubtypeofrose-hiprosecaninareducespaininpatientswithosteoarthritisofthehand–adouble-blind,place-con-trolled,randomizedstudy.abstract.9thworldcongressoftheosteoarthritisresearchsocietyinternational,chicago.

wintherk.,kharazmia.(2004):asubtypeofrosacaninaexertsanti-inflammatoryproperties.abs-tract.Thexxixnordiccongressinclinicalchemistry,malmö.

wintherk.(2006):breakthroughsinreducinginflammation.abstract.nuTraconconference,anaheim.

winther,k.(2000):impactoncoagulation,plateletfunctionandfibrinolysis.abstract.copenha-gen.

reine.,wintherk.(2001):ldl-cholesterolandc-reactiveproteinisinfluencedbyrose-hip,aran-domized,doubleblind,placebocontrolledtrial.abstract.xiv.internationalsymposiumondrugsaffectingmetabolism,newyork.

kharazmia.,songJ.,wintherk.,hoibyn.(2000):naturalmedicineandtheimmunesystem:immu-nomodulatorypropertiesofginsengandantiinflammatorypropertiesofrose-hip.abstract.vita-foodsinternationalconference2000,geneva.

wintherk.,kharazmia.,reine.(1998):rose-hip,givenasastandardiseddrypowder,exertsanti-inflammatoryandcellpreservingpropertiesinhumans.abstract.2ndinternationalcongressoncoronaryarterydisease,florence.

reine.,kharazmia.,wintherk.(1999):rose-hipgivenasastandardiseddrypowderexertsanti-inflammatoryproperties,withoutinfluencingplateletaggregationandthecoagulationcascade.1stinternationalcongressonheartdisease,washington.

bütners.,Jacobseno.,andersenJ.r.,wintherk.(2004):Theimpactofastandardizedpowdermadefromasubtypeofrose-hiponibdQsymptomscoreinpatientswithcrohn’sdiseaseorul-cerativecolitis.abstract.scandinavianJournalofgastroenterology,oslo.

37

NOTICE

A new Opportunity for Joint Health

SCIENTIFIC COMPENDIUM ON ROSE HIP POWDER

0710

-01-

100