Hindawi BioMed Research International Volume 2017, Article ID 8468469, 14 pages https://doi.org/10.1155/2017/8468469 Research Article Declining Physical Performance Associates with Serum FasL, miR-21, and miR-146a in Aging Sprinters ReetaKangas,1TimoTรถrmรคkangas,1AriHeinonen,1MarkkuAlen,2,3HarriSuominen,1 VuokkoKovanen,1EijaK.Laakkonen,1andMarkoT.Korhonen1 1GerontologyResearchCenter,DepartmentofHealthSciences,UniversityofJyvaskyla,Jyvaskyla,Finland 2DepartmentofMedicalRehabilitation,OuluUniversityHospital,Oulu,Finland 3CenterforLifeCourseHealthResearch,UniversityofOulu,Oulu,Finland CorrespondenceshouldbeaddressedtoReetaKangas;[email protected] Received22April2016;Revised2November2016;Accepted21November2016;Published3January2017 AcademicEditor:DavidC.Hughes Copyrightยฉ2017ReetaKangasetal.ThisisanopenaccessarticledistributedundertheCreativeCommonsAttributionLicense, whichpermitsunrestricteduse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited. Agingisassociatedwithsystemicinflammationandcellularapoptosisacceleratingphysiologicaldysfunctions.Whetherphysically activewayoflifeaffectstheseassociationsisunclear.Thisstudymeasuredthelevelsofseruminflammatoryandapoptoticmolecules, theirchangeover10years,andtheirassociationswithphysicalperformanceinsprint-trainedmaleathletes.HsCRP,cellcounts, HGB,FasL,miR-21,andmiR-146aweremeasuredcross-sectionally(๐=67,18โ90yrs)andserumFasL,miR-21,andmiR-146aand theiraging-relatedassociationswithphysicalperformancewereassessedovera10-yearfollow-up(๐ = 49,50โ90yrs).Thecross- sectionalstudyshowedpositiveagecorrelationsforneutrophilsandnegativeforlymphocytes,redbloodcells,HGB,FasL,andmiR- 146a.Duringthe10-yearfollow-up,FasLdecreased(๐=0.017)andmiR-21(๐<0.001)andmiR-146a(๐=0.005)levelsincreased. Whencombiningthemoleculelevels,aging,andphysicalperformance,FasLassociatedwithcountermovementjumpandbench press(๐<0.001),miR-21andmiR-146awithkneeflexion(๐=0.023;๐<0.001),andbenchpress(๐=0.004;๐<0.001)andmiR- 146awithsprintperformance(๐<0.001).Thestudiedserummoleculeschangedinanage-dependentmannerandwereassociated withdecliningphysicalperformance.Theyhavepotentialasbiomarkersofaging-relatedprocessesinfluencingthedevelopmentof physiologicaldysfunctions.FurtherresearchisneededfocusingontheoriginsandtargetsofcirculatingmicroRNAstoclarifytheir functioninvarioustissueswithaging. 1.Introduction a deficit of important cells, both recognized in aging. One way to regulate immune cell homeostasis, and subsequent Physical exercise affects inflammatory state. The common immuneresponse,isthroughFasligand-Fasreceptor(FasL- understanding is that an acute bout of exercise results in Fas)interactiononthecellsurfaces,whichleadstotargetcell a temporal inflammatory response, while regular training destruction [6]. There are indications that habitual training has a protective anti-inflammatory effect [1โ4]. In addition influences the apoptotic processes of the immune system. to immune cells, skeletal muscle tissue contributes to the AccordingtoMoorenetal.[7],thebasallevelsofleukocyte inflammatoryresponsebyreleasinginflammatorymolecules, apoptosisaswellasthelevelsofexercise-inducedapoptosis such as TNF-๐ผ and IL-6, following of acute exercise [5]. aredistinctbetweenhighlytrainedandpoorlytrainedmen, During inflammation, whether induced by age, disease, or suggesting training-induced adaptation to leukocyte home- acuteexercise,controllingthecellularbalanceoftheinflam- ostasis. matory cells is important. Apoptosis, or programmed cell Traces from inflammation and cellular apoptosis can destruction,isacrucialmechanismformaintainingcellular be detected from the circulation by measuring specific balanceinalltissues.Disruptionsinthecellularhomeostasis moleculesfromtheblood.Inadditiontotheclassicalinflam- leadtoeitheranaccumulationofpoorlyfunctioningcellsor mation (e.g., CRP, IL6, and TNF-๐ผ) and apoptotic markers 2 BioMedResearchInternational Cross-sectional Group A Group B Group C Group D 2012 Age (yrs) 18to39 50to66 66to79 79to90 N 18 16 18 15 Follow-up Baseline End-point Baseline End-point Baseline End-point baseline in2002, end-point in2012 Age (yrs) 40to56 50to66 56to69 66to79 69to80 79to90 N 16 16 18 18 15 15 Figure1:Descriptionsofthestudydesignsneededinthecurrentstudy. (cytochrome c, Fas, and FasL), the small RNA molecules In addition, the highest levels of miR-146a during exercise called microRNAs (miRs) are novel and potentially even were found to correlate with peak VO2max [15], whereas more sensitive tools for screening these physiological pro- miR-21 levels have been shown to be upregulated in males cesses.miRsarenoncodingRNAsregulatinggeneexpression withlowVO2 max,indicatingthecontrastingrolesofthese by blocking the translation of specific target mRNA into two miRs [18]. Nielsen et al. [17] demonstrated that miR- proteins.miRsarereleased,eitheractivelyorpassively,from 146a decreased immediately after a bout of acute exercise, differentcelltypesintothebloodstream,wheretheyreflect whereasthebasallevelsofmiR-21weredownregulatedafter thegeneregulationchangesintheircellsoforigin.Theirrole 12weeksofendurancetraining.BothmiR-21andmiR-146a inintercellularcommunicationhasalsobeenrecognized[8]. havealsobeenshowntoexistatdifferentlevelsintheplasma Inhealthyconditions,circulatingmiRsmostlikelyoriginate ofyoungmaleenduranceandstrengthtrainedathletes[19]. from blood or epithelial cells or from organs with high Theabove-mentionedstudiesdemonstratethatthesespecific vascularization,includingskeletalmusclecells[9].Owingto miRs detected in the circulation are affected differently by trauma, cancer, cardiovascular disease, or other conditions exercise type and duration and have potential as indicators influencingmetabolism,themiRsignatureinthecirculation ofphysiologicalchanges[15,16,19,20]. changes. InordertoevaluateifcirculatingFasL,miR-21,andmiR- mir-21andmiR-146aareamongthemiRsthatseemtobe 146ahavethepotentialasbiomarkersoftrainingadaptations responsive to various physiological stimuli. These miRs are inagingathletes,longitudinalstudiesareneeded.Theaimof associated with aging-related processes such as senescence thisstudywastodeterminewhethercirculatingFasL,miR- andinflammation,miR-21havingproinflammatoryandmiR- 21, and miR-146a levels change over 10-year period among 146a having anti-inflammatory effects [10, 11]. Both miRs competitivemalemasterssprinterswithalong-termtraining induce apoptosis by targeting FasL-Fas signaling, which background and to assess their associations with physical strengthenstheinterplayofthesemoleculeswitheachother performancemeasuresandaging. and their role as regulators of immune cell homeostasis [12,13].CirculatingmiR-21hasalsobeenshowntopromote 2.Methods cachexia-relatedapoptosisinskeletalmusclecells[14].Phys- ical exercise creates a whole body adaptive responses that 2.1.StudyDesignandEthics. Thisstudyispartofanongoing also affect miR regulation and expression in different cell Athlete Aging Study (ATHLAS) [21โ25] on young adult types and subsequently in the circulation [15โ17]. miR-21 athletesandmastersathletesfromdifferentsportdisciplines. andmiR-146ahavealsobeenshowntobehighlyresponsive The participants were recruited from the memberships of to physical exercise. The pioneering work by Baggish et Finnishathleticorganizations.Thepresentstudycomprised al. [15] demonstrated that these miRs were associated with malesprinters.Thesprintershadalong-termbackgroundin cardiovascular/musculoskeletal adaptations and low-grade sprinttrainingandhadbeensuccessfulinnationalorinterna- inflammation and changed in response to acute exercise as tional100โ400msprintingevents.Bothcross-sectionaland well as sustained training. CirculatingmiR-146a levels were longitudinalstudydesignswereused(Figure1).Inthecross- upregulatedbyacutecyclingexercisebeforeandafter90days sectionalanalysis(usingdatafromfollow-upmeasurements ofsustainedrowtraining,whereasmiR-21wasonlyupregu- conductedin2012),thesprintersweredividedintofourage latedbyacuteexercisepriortothesustainedtrainingperiod. groups:(A)18to39years(๐=18),(B)50to66years(๐=16), BioMedResearchInternational 3 (C)66to79years(๐=18),and(D)79to90years(๐=15).In USA) according to the manufacturerโs protocol. Reactions addition, baseline measurements (conducted in 2002) were were performed as duplicates and a common sample was availablefor49oldermasterssprinters(aged50to90years addedtoeachplateinordertoobservewhethercontrolling at follow-up) belonging to age groups B, C, and D, thereby forinterassayvariationwasnecessary.HsCRPmeasurements allowing longitudinal analysis. Both the baseline and the atbaselineandfollow-upwereperformedwithanImmulite follow-up measurements consisted of similar standardized 1000ImmunoassaySystem.Leukocytecountswerepartofthe two-daymeasurements. standardmedicalcompletebloodcountanalysis. Prior to the measurements, training and health history wereelicitedandevaluatedusingquestionnaires.Theaverage 2.4.RNAExtractionandmiRAnalyses. TheRNAextraction number of training years among the masters athletes at methods and miR-21 and miR-146a analyses have been baseline was 34.1 ยฑ 15.1 years. More detailed self-reported described previously [26]. Briefly, total RNA was isolated traininghistories(trainingfrequency,sprint-specifictraining from100๐lofserumwithatotalRNApurificationkit(Nor- hours, and other training hours) and their changes over 10 gen Biotek Corporation, Thorold, ON, Canada) according years are presented in Tables 1 and 3. The sprint-specific to the manufacturerโs protocol. Synthetic C. elegans Cel- traininghoursperweekincludedsprints,jumps,andstrength ๓ธ ๓ธ miR-39 (5 -UCA CCG GGU GUA AAU CAG CUU G-3 , training, while the other training hours included all other Invitrogen) (25fM, concentration determined by dilution notably strenuous physical exercises. Subjects over age 55 series) was added at the lysis step to all of the samples as underwentamedicalexamination.Currentabilitytopartic- a spike-in control in order to monitor the efficiency and ipateinphysicallydemandingmeasurementwasassessedad uniformity of the RNA extraction and qPCR procedure. hoc,individuallybythestudyphysician.Thehealthofallthe participants was in general good with no acute conditions RNA was reverse-transcribed to cDNA (๐tot = 10๐l) by (infections,traumas)orfunctionallylimitingchronicneuro- a Taqman reverse transcription kit and the qPCR (๐tot = logical,cardiovascular,endocrinological,ormusculoskeletal 10๐l) was performed with a Taqman Universal Mastermix ๓ธ conditions. In a few cases, some physical tests were not II NO Ung using Taqman miR assays (hsa-miR-21: 5 - ๓ธ ๓ธ performedduetolocalmusculoskeletalpain. UAGCUUAUCAGACUGAUGUUGA-3 , hsa-miR-146a: 5 - ๓ธ Thestudywasconductedaccordingtotheguidelinesof UGAGAACUGAAUUCCAUGGGUU-3 ) (Device: Applied theDeclarationofHelsinki.Allparticipantswereinformeda Biosystems,ABI7300). priori about the possible risks and discomfort of the phys- Ct values less than 35 were accepted for the analysis. ical and clinical measurements. Written informed consent, All the samples were normalized to a reference sample of including permission for the use of the gathered data for an average sprint athlete across the plates. ฮCt values were researchpurposesonly,wasprovidedbythestudysubjects. calculated as ฮCt = mean CtmiR-X โ mean CtmiR(reference). The study protocol was approved by the ethics committees Each reaction was performed in duplicate and the relative oftheUniversityofJyvaยจskylaยจ(in2002forbaseline)andthe expressions were calculated by using the 2โฮCt method. To CentralFinlandHealthcareDistrict(in2012forfollow-up). observethepossiblecontaminationsandprimer-dimers,no templatecontrols(NTCs)wereusedineithertheRTorqPCR 2.2. Physical Performance and Body Composition Measure- reactions. ments. Participantswereinstructedtorest(noheavytraining orcompetition)twodaysbeforethemeasurements.Onthe 2.5. Statistical Analyses. In the age-correlation analyses of firstmeasurementday,theparticipantsperformedamaximal the cross-sectional design, Pearsonโs correlation coefficient 60m sprint twice on an indoor running track with spiked wasusedforcontinuousvariablesandSpearmanโscorrelation running shoes. Explosive force production of the lower coefficient for ordinal variables. The paired sample ๐ก-test limbs was measured by a vertical countermovement jump for parametric variables and Wilcoxon signed rank test (CMJ).Isometrickneeflexionforceandisometricupperlimb for nonparametric variables were used in the longitudinal force were measured by bench press performance using a comparison in the 10-year follow-up design. In addition, David 200 dynamometer (David Fitness and Medical Ltd., GeneralizedEstimatingEquations(GEE)modelswerecon- Outokumpu,Finland).Thebestof2or3trialswasrecorded structedto study theassociationofFasL,miR-21, andmiR- in the subsequent analyses. In addition, the assessment of 146a with the performance measures over age. The model total body fat and lean mass (LBM) was performed with was based on the two measurements points (2002, 2012) bioelectrical impedance (Spectrum II, RJL System, Detroit, used in the longitudinal study with unstructured working MI). More detailed descriptions of the measurements are givenelsewhere[22โ25].Participantsโmealsduringthemea- correlation matrix specification and athletesโ age as the surementdayswerearrangedbytheresearchorganizers. descriptive metric of time. As the effect of age is generally nonlinear we used polynomial terms (quadratic and cubic) 2.3. Serum Analyses. To obtain serum, venous blood was ofagetoincludecurvatureinmodelingtheage-relatedeffect collected on the second measurement day under standard of the FasL and the miRs on the outcomes. The results fasting conditions at least 12h after exercise. The serum were adjusted for LBM. More detailed prediction equations sampleswerestoredatโ80โCuntilanalyzed.SerumFasLcon- based on GEE-models are presented in the supplemen- centrationwasmeasuredusingaHumanFasLigand/TNFSF6 tary data (S1 in Supplementary Material available online at Quantikine(cid:2) ELISA Kit (R&D Systems, Minneapolis, MN, https://doi.org/10.1155/2017/8468469). 4 BioMedResearchInternational CI/coefficientofetermination โ0.4to0.10.031โ0.3to0.20.0070.0to0.60.096 0.7to0.90.707โโ0.9to0.60.612โโ0.7to0.20.220โโ0.9to0.60.602 etesand(2)masters %d hl 5 at 9 all nsofthevariableswithage. Correlationwithage๐(onlyB,C,D)/value โ(๐=46)0.176๐=0.241S(๐=46)โ0.085๐=0.573S(๐=46)0.310P=0.036 S(๐=36)0.841P<0.001โ(๐=35)0.782P<0.001โ(๐=41)0.469P=0.002โ(๐=38)0.776P<0.001 correlationsarepresented:(1) atio of Age el ntn D. oupsandcorr %CI/coefficiedeterminatio โโ0.8to0.50.434โโ0.6to0.10.144โ0.1to0.40.032 0.8to0.90.806โโ0.9to0.70.656โโ0.8to0.50.419โโ0.9to0.60.604ยฑdasmeansS gr 95 nte measuresindifferentage Correlationwithage๐(allgroups)/value โ(๐=61)0.659P<0.001S(๐=64)โ0.380P=0.002S(๐=64)0.179๐=0.157 S(๐=48)0.898P<0.001โ(๐=50)0.810P<0.001โ(๐=58)0.647P<0.001โ(๐=55)0.777P<0.001 18to90yrs.Resultsareprese ce es n g calperforma D79โ90yrs๐=15() 3.1ยฑ1.1(๐=14)2.7ยฑ2.3(๐=14)2.0ยฑ2.5(๐=14) 10.81ยฑ1.30(๐=10)18.9ยฑ4.7(๐=11)190ยฑ49(๐=12)565ยฑ126(๐=11) athletesfroma Table1:Self-reportedtraininghistoryandphysi ABC18โ39yrs50โ66yrs66โ79yrs๐=18๐=16๐=18()()()Self-reportedtraining3.6ยฑ1.66.8ยฑ2.33.5ยฑ1.3Frequency(times/wk)(๐=15)(๐=14)2.6ยฑ1.73.2ยฑ3.27.0ยฑ5.5Sprinttraining(h/wk)(๐=14)0.4ยฑ0.61.3ยฑ2.02.5ยฑ4.2Othertraining(h/wk)(๐=14) Physicalperformance7.52ยฑ0.368.43ยฑ0.639.29ยฑ0.61Sprint60m(s)(๐=12)(๐=13)(๐=13)45.7ยฑ11.832.0ยฑ4.225.8ยฑ5.0CMJ(cm)(๐=15)(๐=12)(๐=12)326ยฑ86ยฑ227ยฑ3425357Isometrickneeflexion(N)(๐=17)(๐=15)(๐=14)1307ยฑ347945ยฑ159717ยฑ175Isometricbenchpress(N)(๐=17)(๐=14)(๐=13) Tableisformedbasedonthecross-sectionalstudydesign(2012)includingalltheSSpearmanโscorrelationcoefficient.athletesonly.CMJ:countermovementjump. BioMedResearchInternational 5 3.Results ofthevariation,respectively.Alownegativecorrelationwas observed between age and HGB both when all the athletes 3.1.ParticipantCharacteristicsandSerumFasLandmiRLevels were included (๐ = โ0.383, 95% CI = โ0.6 to โ0.2, ๐ = in Different Age Groups. Participantsโ characteristics in the 0.002) and when only the masters athletes were included differentagegroupsandagecorrelationsofthevariablesare (๐ = โ0.370, 95% CI = โ0.6 to โ0.1, ๐ = 0.010), with age presentedinTables1and2.Theagecorrelationsarepresented accounting for 14.7% and 13.7% of the variation in HGB, in two different ways: (1) inclusive of all the participants respectively. Of the studied serum molecules, a modest (ages18to90)and(2)inclusiveonlyofthemastersathletes negativecorrelationbetweenageandFasLwasdetectedwhen (ages50to90).Amodestnegativecorrelationwasobserved all the athletes were included in the analyses (๐ = โ0.596, betweenageandtrainingfrequencywhenalltheathleteswere 95%CI=โ0.7toโ0.4,๐ < 0.001),withageaccountingfor includedintheanalyses(๐ = โ0.659,95%CI=โ0.8toโ0.5, 35.5%ofthevariationinFasLconcentration.Amodestposi- ๐<0.001),withageaccountingfor43.3%ofthevariationin tivecorrelationwasdetectedbetweenageandmiR-146awhen trainingfrequency.Thesamecorrelationwasverylowwhen alltheathleteswereincluded(๐ = โ0.611,95%CI=โ0.7to onlythemastersathleteswereincluded(๐ = โ0.176,95%CI โ0.4,๐ < 0.001),withageaccountingfor37.3%ofthemiR- =โ0.4to0.1,๐ = 0.241),withageaccountingonlyfor3.1% 146a variation. No significant correlations between age and ofthevariation.Amongthephysicalperformancemeasures, the serum molecules were detected when only masters ahighpositivecorrelationwasfoundbetweenageand60m athleteswereincludedintheanalyses. sprinttimeamongalltheathletes(๐=0.898,95%CI=0.8to 0.9,๐<0.001),withageaccountingfor80.6%ofthevariation insprinttime.Thepatternwassimilarwhenonlythemasters 3.2.ChangesinthePhysical PerformanceofMastersAthletes athletes were included in the analyses (๐ = 0.841, 95% CI over 10 Years. Changes in the training frequencies and = 0.7 to 0.9, ๐ < 0.001), with age accounting for 70.7% of the change percentage of the specific physical performance thevariationinsprinttime.Theotherphysicalperformance measures in 10 years are presented in Table 3. The results measuresshowedamodesttohighnegativecorrelationwith areshownformasterssprintersasawholeanddividedinto agewhenalltheathleteswereincluded(CMJ:๐=โ0.81,95% 3 age groups (B: 50โ66yrs, C: 66โ79yrs, and D: 79โ90yrs CI=โ0.9toโ0.7,๐<0.001,65.6%;kneeflexion:๐=โ0.647, presenting the end-point ages). The self-reported weekly 95% CI =โ0.8 to โ0.5,๐ < 0.001, 41.9%; andbench press: trainingfrequency,sprint,andothertraininghoursdecreased ๐ = โ0.777, 95% CI = โ0.9 to โ0.6, ๐ < 0.001, 60.4%). when all the masters athletes were involved in the analyses Whenonlythemastersathleteswereincluded,highnegative (๐ < 0.001,๐ < 0.001,and๐ = 0.003,resp.).Whengroup- correlations were observed between age and CMJ and age ingtheathletes,thedeclinesinalltheself-reportedtraining andbenchpress(CMJ:๐ = โ0.782,95%CI=โ0.9toโ0.6, activities remained significant for the group B (๐ = 0.011, ๐ < 0.001;benchpress:๐ = โ0.776,95%CI=โ0.9toโ0.6, ๐ = 0.009, and ๐ = 0.008, resp.). For group C, no signi- ๐ < 0.001),withageaccountingfor61.2%and60.2%ofthe ficant changes in the training activities over 10 years were variation,respectively.Betweenageandkneeflexion,onlya reported. For the group D, weekly training frequency and modestnegativecorrelationwasdetected(๐ = โ0.469,95% sprinttraininghoursdecreasedsignificantly(๐ = 0.004,๐ = CI=โ0.7toโ0.2,๐=0.002),withageaccountingfor22%of 0.023,resp.).Fromthephysicalperformancemeasures60m thevariationinkneeflexionstrength. sprint,CMJandisometricbenchpresschangedduringthe10 Amongthebodyanthropometricvariables,modestneg- yearsโ follow-up period when all the masters sprinters were ativecorrelationswereobservedbetweenageandheight,age included(๐<0.001forallmeasures).Whenthemastersspr- and weight, and age and LBM when all the athletes were intersweredividedinto3agegroupsthechangesremained included (height: ๐ = โ0.512, 95% CI = 0.7 to โ0.3, ๐ < for all groups B (๐ = 0.001, ๐ = 0.001, and ๐ = 0.036, 0.001;weight:๐ = โ0.419,95%CI=โ0.6toโ0.2,๐ < 0.001; resp.),C(๐ = 0.002,๐ = 0.001,and๐ = 0.008,resp.),andD LBM:๐ = โ0.525,95%CI=โ0.7toโ0.3,๐ < 0.001),with (๐=0.005,๐<0.001,and๐<0.001,resp.),withthehighest ageaccountingfor26.2%,17.6%,and27.6%ofthevariation, deteriorationseenamongtheoldestathletes(groupD).There respectively.Ofthebloodcellcounts,lowtomodestnegative werenosignificantchangesinthekneeflexionstrengthin10 correlations were detected between age and lymphocyte years. percentage and age and RBC when all the athletes were included (LYM%: ๐ = โ0.384, 95% CI = โ0.6 to โ0.2, ๐ = 3.3. Changes in the Serum hsCRP, FasL, miR-21, and miR- 0.002;RBC:๐ = โ0.433,95%CI=โ0.6toโ0.2,๐ < 0.001), 146a Levels of Masters Athletes over 10 Years. The change with age accounting for 14.7% and 18.7% of the variation, percentages of the blood parameters, hsCRP, FasL, miR-21, respectively. In addition, a modest positive correlation was and miR-146a over 10 years are presented in Table 3. The detectedbetweenageandneutrophilpercentage(๐ = 0.411, resultsareshownforthemasterssprintersbothasawholeand 95% CI = 0.2 to 0.6, ๐ = 0.001), with age accounting for dividedinto3agegroups(B:50โ66yrs,C:66โ79yrs,andD: 16.9%ofthevariationinneutrophils.Whenonlythemasters 79โ90yrspresentingtheend-pointages).SerumhsCRPdid athletes were included, the pattern was similar in slightly notchangeoverthe10-yearperiodamongtheathletes.The weakeragecorrelations(LYM%:๐ = โ0.315,95%CI=โ0.6 serumFasLconcentrationsdecreased(๐=0.017)andserum to0.0,๐ = 0.031;RBC:๐ = โ0.370,95%CI=โ0.6toโ0.1, miR-21andmiR-146alevelsincreased(๐<0.001,๐<0.005, ๐ = 0.010; and NEUT%: ๐ = 0.326, 95% CI = 0.0 to 0.6, resp.)amongthemasterssprintersasasinglegroupduring ๐ = 0.025),withageaccountingfor9.9%,13.7%,and10.6% the 10 years. At the age-group level, the changes were 6 BioMedResearchInternational 95%CI/coefficientofdetermination โโ0.7to0.20.215โโ0.7to0.30.323โโ0.7to0.30.318โ0.5to0.00.091 โ0.2to0.40.010โ0.6to0.00.099โ0.4to0.20.0130.0to0.60.106โโ0.6to0.10.137 โโ0.6to0.10.141 โโ0.6to0.20.187 โ0.4to0.20.010โ0.5to0.10.033โ0.4to0.20.014โ0.4to0.20.011 ways:(1)oneincludingallcells,PLT:platelet,HGB: woood eue ntbl variableswithage. Correlationwithag๐(onlyB,C,D)/val โ๐=490.464()P=0.001โ๐=490.568()P<0.001โ๐=490.564()P<0.001โ๐=490.301()P=0.035 ๐=470.099()๐=0.509โ๐=470.315()P=0.031โ๐=470.112()๐=0.455๐=470.326()P=0.025โ๐=470.370()P=0.010 โ๐=470.375()P=0.009 โ๐=470.443().002P=0 โ๐=470.102()๐=0.999โ๐=410.181()๐=0.258S๐=49โ0.118()๐=0.419S๐=49โ0.105()๐=0.473 orrelationsarepresentediUT:neutrophils,RBC:red rticipantcharacteristicsandbloodparametersindifferentagegroupsandcorrelationsofthe BCDCorrelationwithage95%CI/coefficientof50โ66yrs66โ79yrs79โ90yrs๐(allgroups)/valuedetermination(๐=16)(๐=18)(๐=15) โ๐=67โโ0.512()0.7to0.3ยฑยฑยฑ178.28.1172.84.7172.44.9P<0.0010.262โ๐=67โโ0.419()0.6to0.2ยฑยฑยฑ80.29.771.06.269.46.4P<0.0010.176โ๐=67โโ0.525()0.7to0.3ยฑยฑยฑ66.37.060.24.158.34.8P<0.0010.276โ๐=67โ0.028()0.3to0.2ยฑยฑยฑ13.96.111.13.711.23.6๐=0.8570.001 ยฑยฑโ๐=65โ5.72.25.40.80.057()0.3to0.2ยฑ5.61.0๐=17๐=14๐=0.652()()0.003ยฑยฑโ๐=65โโ0.234.25.631.09.40.384()0.6toยฑ35.96.2๐=17)๐=14P=0.0020.147(()ยฑยฑโ๐=65โ11.51.910.92.40.174()0.4to0.1ยฑ11.53.0๐=17)๐=14๐=0.165(()0.030ยฑยฑ๐=6554.36.658.19.50.411()0.2to0.6ยฑ52.66.9๐=17)๐=14P=0.001(()0.169ยฑยฑโ๐=65โโ4.90.44.40.30.443()0.6to0.2ยฑ4.90.4๐=17)๐=14P<0.000(()0.187ยฑยฑ210.0199.7ยฑโ๐=65โ242.30.100()0.3to0.183.642.9๐=0.42944.30.010(๐=17)๐=14()ยฑยฑยฑโ๐=65โโ152.9149.49.3140.38.00.383()0.6to0.2(๐=17)๐=14P=0.00212.4()0.147 ยฑยฑโ๐=65โ2.36.01.31.20.035()0.3to0.2ยฑ1.73.3๐=17)๐=14(()0.7840.001ยฑยฑยฑโ๐=59โโ56.821.956.018.252.323.40.596()0.7to0.4๐=14(๐=14)๐=13P<0.001()()0.355Sโ๐=67โ0.0950.3to0.1()ยฑยฑยฑ2.221.411.741.051.650.62๐=0.4440.009Sโโ๐=67โ0.6110.7to0.4()ยฑยฑยฑ1.831.161.500.711.510.96P<0.0010.373ยฑdesign(2012)includingalltheathletesfromages18to90yrs.ResultsarepresentedasmeansSD.Agechletes.LBM:leanbodymass,WBC:whitebloodcells,LYM:lymphocytes,MXD:mixedleukocytes,NESotein,FasL:Fas-ligand,andRE:relativeexpression.Spearmanโscorrelationcoefficient. Pa dyatpr Table2: A18โ39yrs(๐=18) odycomposition ยฑ180.15.0 ยฑ78.36.8 ยฑ67.25.7 ยฑ11.24.1 ยฑ5.61.4 ยฑ39.27.2 ยฑ12.34.8 ยฑ48.68.5 ยฑ5.10.4 ยฑ211.944.0 ยฑ154.39.3 ยฑ1.65.1 ยฑ92.324.7 ยฑ1.810.75 ยฑ5.822.66 hecross-sectionalstucludingonlymasterssensitivityc-reactive b tnh Anthropometricsand Height(cm) Weight(kg) LBM(kg) Bodyfatmass(kg) Bloodcellcount WBC LYM% MXD% NEUT% RBC PLT HGB Serummolecules hsCRP(mg/L) FasL(pg/ml) miR-21(RE) miR-146a(RE) Tableisformedbasedontheathletesand(2)oneihemoglobin,hsCRP:hig BioMedResearchInternational 7 Table3:Thechangeinphysicalperformancemeasures,self-reportedtrainingamounts,andserummoleculelevelsamongallandage-grouped masterssprintersinthe10-yearfollow-up. All B C D 50โ90yrs 50โ66yrs 66โ79yrs 79โ90yrs Trainingfrequency Change(times/wk) โ0.89ยฑ1.22(๐=42) โ0.91ยฑ1.15(๐=14) โ0.73ยฑ1.43(๐=15) โ1.06ยฑ1.09(๐=13) ๐value P<0.001 P=0.011 ๐=0.067 P=0.004 Sprinttraining Change(h/wk) โ1.84ยฑ3.00(๐=42) โ1.80ยฑ2.25(๐=14) โ1.94ยฑ4.08(๐=15) โ1.77ยฑ2.43(๐=13) ๐values P<0.001 P=0.009 ๐=0.078 P=0.023 Othertraining Change(h/wk) โ1.02ยฑ2.62(๐=32) โ1.53ยฑ1.61(๐=11) โ0.93ยฑ3.44(๐=11) โ0.56ยฑ2.49(๐=10) ๐value P=0.006 P=0.008 ๐=0.221 P=0.424 Sprint60m(s) 11.9(6.7โ16.5) 8.2(5.7โ10.0) 12.2(5.7โ16.4) 15.8(12.1โ22.8) Change% (๐=35) (n=13) (๐=12) (๐=10) ๐value P<0.001W P=0.001W P=0.002W P=0.005W CMJ(cm) Change% โ15.4ยฑ10.2(๐=32) โ9.2ยฑ6.3(๐=11) โ15.7ยฑ10.0(๐=11) โ22.0ยฑ10.4(๐=10) ๐value P<0.001 P=0.001 P=0.001 P<0.001 Kneeflexion(N) Change% โ1.8ยฑ23.6(๐=40) 5.5ยฑ29.0(๐=15) โ3.6ยฑ20.1(๐=13) โ9.07ยฑ18.0(๐=12) ๐value ๐=0.062 ๐=0.896 ๐=0.287 ๐=0.111 Isometricbenchpress(N) Change% โ10.9ยฑ11.9(๐=36) โ5.3ยฑ9.8(๐=14) โ14.0ยฑ14.7(๐=12) โ14.9ยฑ8.51(๐=10) ๐value P<0.001 P=0.036 P=0.008 P<0.001 hsCRP Change% 112ยฑ349(๐=47) 203ยฑ512(๐=16) 71.5ยฑ230(๐=17) 56.8ยฑ220(๐=14) ๐value ๐=0.841 ๐=0.404 ๐=0.868 ๐=0.220 FasL Change% โ5.3ยฑ26.3(๐=41) โ18.3ยฑ16.7(๐=14) 1.22ยฑ29.1(๐=14) 1.78ยฑ27.9(๐=13) ๐value P=0.017 P=0.001 ๐=0.746 ๐=0.587 miR-21 77.0(1.1โ689) 107(21.0โ662) 37.9(โ34.7โ834) 136(21.3โ995) Change% (๐=49) (๐=16) (๐=18) (๐=15) ๐value P<0.001W P=0.007W ๐=0.267W P=0.017W miR-146a 95.4(โ7.4โ631) 93.0(โ3.0โ1101) 18.6(โ26.0โ444) 143(53.6โ861) Change% (๐=49) (๐=16) (๐=18) (๐=15) ๐value P=0.005W ๐=0.079W ๐=0.500W P=0.011W Theagerangesofthegroups(B,C,andD)representtheagesatfollow-up.DataarepresentedasmeansยฑSDforparametricvariablesandasmedian(IQR)for nonparametricvariables.W๐valuewascalculatedwithpaired๐ก-testfornormallydistributedvariablesandwithWilcoxonsignedranktestfornonparametric W variables. Wilcoxonsignedranktestfornonparametricvariables.CMJ=countermovementjump,hsCRP=highsensitivityC-reactiveprotein,andFasL: Fas-ligand. significantforFasLandformiR-21amongtheyoungestmas- model was constructed to combine the effects of a specific terssprinters(groupB;๐ = 0.001;๐ = 0.007,resp.)andfor serummolecule(FasL,miR-21,ormiR-146a)andthephysical miR-146aamongtheoldestgroup(D;๐ = 0.011).Thecha- performance measures (60m sprint, CMJ, knee flexion, ngewasalsosignificantformiR-21intheoldestgroup(group or bench press) over time. The association curves for the D,๐ = 0.017).Theoriginalvaluesforthebloodparameters physical performance measures according to the different arepresentedinsupplementarydata(S2). measuredcirculatingmoleculelevelsarepresentedinFigures 2,3,and4.Onlysignificantortrendingcurvesarepresented. 3.4. FasL, miR-21, and miR-146a Associations with Physical The results for the models were adjusted with the LBM. Performance Measures in the 10-Year Follow-Up. A GEE Figure 2 shows that the combination of the effects of FasL 8 BioMedResearchInternational 0.8 Countermovement jump p (cm) Effect Estimate Std. Err. P value OPm vnaliubeuas m ment ju 0.6 FFaassLLรage โ00.0.01045 00..000025 00..000077 <0.001 move 0.4 FasLรage2 โ0.009 0.003 0.006 er Note. Age was divided by 75. Model included also main effects of age, unt age2, and lean body mass. o C 0.2 aOmnibusWald-test for the joint significance of FasL,FasLรage, and FasLรage2. 20 30 40 50 60 70 80 90 Age FasL (pg/ml) Lower quartile Mean Upper quartile (a) 2000 Bench press Omnibus Effect Estimate Std. Err. P value P valuea N) 1500 FasL 26.08 7.349 <0.001 <0.001 ess ( FasLรage โ11.63 3.313 <0.001 h pr FasLรage2 1.21 0.370 0.001 c en 1000 Note. Age was divided by 15. Model included also main effects of age, B age2, and lean body mass. aOmnibusWald-test for the joint significance of FasL,FasLร age, and 500 FasLรage2. 20 30 40 50 60 70 80 90 Age FasL (pg/ml) Lower quartile Mean Upper quartile (b) Figure2:TheassociationofserumFasLconcentrationwithphysicalperformanceovertime.Valuesfortheyoungerparticipants(<40years) withoutfollow-upmeasuresarepresentedontheleftsideoftheimages.Theassociationsarebasedonthefollow-updesign(๐=49,>40yrs). Cross(ร)indicatesthatcaseislocatedwithin0โ37.5%,circle(I)within37.5โ67.5%,andsquare(โป)within67.5โ100%ofthecumulativeshare oftheFasLdistribution.The3differentlinespresenttheassociationsofthedifferentserummarkerlevelswiththephysicalperformance measures over time. The tables next to the curves present the model used in forming the prediction curves, in greater detail, including statisticsonthemaineffectsofthestudiedserummarkerandthepossiblequadraticandcubiceffects. and its interactions with age was statistically significant for FasLlevels.Forbenchpress,higherFasLlevelsattheyounger the CMJ (๐ < 0.001) and bench press (๐ < 0.001). Also, agespredictedasteadierdeclineinperformancewhilelower allcoefficientestimatesfortheseoutcomeswerestatistically FasL values predicted better sustained performance at the significantindicatingbothasignificantlinearandquadratic olderages(Figure2(b)). curvature term. The model for the CMJ, when the serum SignificanteffectsofmiR-21andagecombined(Figure3) FasL concentration was taken into account (Figure 2(a)), weredetectedforkneeflexion(๐ = 0.023)andbenchpress predictedasteeperdeclineinperformanceafterageof70with strength(๐ = 0.004).Statisticallysignificantcoefficientesti- slightly higher FasL serum concentrations than with lower mates relate mainly to miR-21 interactionwith quadratic of BioMedResearchInternational 9 0 0 5 Knee ๏ฌexion strength h (N) 400 Effect Estimate Std. Err. P value OPm vnaliubeuas ngt miR-21 801 582 0.169 0.023 on stre 300 mmiiRR--2211รรaaggee2 โ166453 9420.55 00..017125 exi miR-21รage3 โ13.5 6.96 0.053 ๏ฌ ee 0 Note. Age was divided by 15. Model included also main effects of age, age2, and n 0 K 2 lean body mass. aOmnibusWald-test for the joint significance of miR-21, miR-21รage, 00 miR-21รage2, and miR-21รage3. 1 20 30 40 50 60 70 80 90 Age miR-21 Lower quartile Mean Upper quartile (a) 0 Bench press 0 0 2 Omnibus Effect Estimate Std. Err. P value P valuea ess (N) 1500 mmiiRR--2211รage โ667199 430111 00..009407 0.004 pr miR-21รage2 163 78.6 0.038 ch miR-21รage3 โ13.2 6.47 0.042 n 0 Be 100 Note. Age was divided by 15. Model included also main effects of age, age2, and lean body mass. aOmnibusWald-test for the joint significance of miR-21, miR-21รage, 00 miR-21รage2, and miR-21รage3. 5 20 30 40 50 60 70 80 90 Age miR-21 Lower quartile Mean Upper quartile (b) Figure3:TheassociationofserummiR-21levelwithphysicalperformanceovertime.Valuesfortheyoungerparticipants(<40yrs)without follow-upmeasuresarepresentedontheleftsideoftheimages.Thepredictionsarebasedonthefollow-updesign(๐=49,>40yrs).Cross(ร) indicatesthatcaseislocatedwithin0โ37.5%,circle(I)within37.5โ67.5%,andsquare(โป)within67.5โ100%ofthecumulativeshareofthemiR- 21-distribution.The3differentlinespresenttheassociationsofthedifferentserummarkerlevelswiththephysicalperformancemeasures overtime.Thetablesnexttothecurvespresentthemodelusedinformingthepredictioncurves,ingreaterdetail,includingstatisticsonthe maineffectsofthestudiedserummarkerandthepossiblequadraticandcubiceffects. cubic terms of age indicating that curvature has a stronger Significant combination effects of miR-146a and age roleintheprediction.Forkneeflexionstrength,whenserum (Figure4)weredetectedforsprint(๐ < 0.001),kneeflexion miR-21 levels were taken into account, low miR-21 levels (๐<0.001),andbenchpressstrength(๐<0.001).ThemiR- predicted best performance prior to age 60 and high levels 146a levels predicted the largest differences in 60m sprint bestperformancebetweenages60and80(Figure3(a)).Low performanceaftertheageof70,afterwhichthelowestvalues miR-21 levels predicted highest performance in the bench predicted the best sprint performance (Figure 4(a)). For pressuntilage65,afterwhichthepredictioncurveswerevery knee flexion strength, the lowest miR-146a levels predicted similartoeachother(Figure3(b)). the best performance until age 60, after which, until age 10 BioMedResearchInternational 8 1 Sprint Omnibus 16 Effect Estimate Std. Err. P value P valuea m (s) 14 mmiiRR--114466aaรage โ87.3.453 22..4249 00..000011 <0.001 0 2 miR-146aรage2 โ3.06 0.908 0.001 6 1 nt miR-146aรage3 0.367 0.108 0.001 Spri 10 Note. Age was divided by 20. Model included also main effects of age,age2, and lean body mass. 8 aOmnibusWald-test for the joint significance of miR-146a, miR-146aรage, 5 miR-146aรage2, and miR-146aรage3. 6. 20 30 40 50 60 70 80 90 Age miR-146a Lower quartile Mean Upper quartile (a) 0 0 5 Knee ๏ฌexion strength h (N) 400 Effect Estimate Std. Err. P value OPm vanliubeuas gt miR-146a 759 202 <0.001 <0.001 n xion stre 300 mmmiiiRRR---111444666aaaรรรaaagggeee23 โโ2276884.87 661.397.732 <<<000...000000111 e e ๏ฌ 0 Note. Age was divided by 20. Model included also main effects of age,age2, and ne 20 lean body mass. K aOmnibusWald-test for the joint significance of miR-146a, miR-146aรage, 0 miR-146aรage2, and miR-146aรage3. 0 1 20 30 40 50 60 70 80 90 Age miR-146a Lower quartile Mean Upper quartile (b) 0 Bench press 0 0 2 Omnibus Effect Estimate Std. Err. P value P valuea ess (N) 1500 mmiiRR--114466aaรage โ113175 4341..55 <00.0.00021 <0.001 h pr miR-146aรage2 21.7 5.39 <0.001 c en 00 Note. Age was divided by 20. Model included also main effects of age,age2, and B 0 1 lean body mass. aOmnibusWald-test for the joint significance of miR-146a, miR-146aรage, 00 miR-146aรage2, and miR-146aรage3. 5 20 30 40 50 60 70 80 90 Age miR-146a Lower quartile Mean Upper quartile (c) Figure4:TheassociationofserummiR-146alevelwithphysicalperformanceovertime.Valuesfortheyoungerparticipants(<40years) withoutfollow-upmeasuresarepresentedontheleftsideoftheimages.Thepredictionsarebasedonthefollow-updesign(๐=49,>40yrs). Cross(ร)indicatesthatcaseislocatedwithin0โ37.5%,circle(I)within37.5โ67.5%,andsquare(โป)within67.5โ100%ofthecumulativeshare ofthemiR-146adistribution.The3differentlinespresenttheassociationsofthedifferentserummarkerlevelswiththephysicalperformance measuresovertime.Thetablesnexttothecurvespresentthemodelusedinformingthepredictioncurves,ingreaterdetail,includingstatistics onthemaineffectsofthestudiedserummarkerandthepossiblequadraticandcubiceffects.
Description: