Page 1
HispanicCommunityHealthStudy/StudyofLatinos(HCHS- SOL)
NationalHeart,Lung,andBloodInstituteNationalInstituteofDiabetesandDigestiveandKidneyDiseases
NationalInstituteofNeurologicalDisordersandStrokeNationalInstituteofDeafnessandOtherCommunicationDisorders
NationalInstituteofDentalandCraniofacialResearchNationalInstituteforMinorityHealthandHealthDisparities
OfficeofDietarySupplements
Page 2
ChallengestoHealthDisparitiesResearch
• Barriers– LegalStatus– PastTraumas–Discrimination–Marginalization–DistrustofGovernment– CulturalorLinguisticDifferences
2(USC, Center for Health Journalism)
Page 3
Hispanic/LatinoDefined
Jaimes, Londono, Halpern, JAMA Dermatol, 2013
A- Latino refers to persons whose origin or ancestries are from countries of Latin America.
B-The US Office of Management and Budget uses the terms Hispanic and Latino interchangeably to refer to persons who indicated that their origin is Mexican, Puerto Rican, Cuban, Central and South American, or other Spanishculture or Spanish-speaking country or origin, regardless of race.
C-Hispanic include individuals whose origin or ancestry comes from Hispania, or D-Spanish-speaking persons of Latin American descent living in the United States.
3
Page 4
GrowthanddiversificationofUSLatinopopulation
Page 5
USHispanic/Latinos
1980 1990
2000 2010Pew Hispanic Center (www.pewhispanic.org)
5
Page 6
HCHS- SOLcohortoverview• Multicenterprospectivecohortrecruitedin2008-2011
• N=16,415participantsaged18to74yearsold1/3ofparticipantswere18-45yearsatbaseline (undersampled)2/3ofparticipantswere>45yrs atbaseline(oversampled)AreabasedhouseholdsampleinfourUScities
• TOPMed contactPIs:RobertKaplanandKariNorth• GeneticSIGchairs:K.North,E.Boerwinkle,T.Sofer• GeneticAnalysisCenter:BruceWeirandCathyLaurie• OthersitePIs:M.Daviglus (Chicago),N.Schneiderman (Miami),G.Talavera(SanDiego),J.Cai(CC)
• SpirometryreadingcenterPIs:GrahamBarr,PaulEnright,JohnHankinson
Page 7
RecruitmentN=16,415,ages
18-74
Annualfollow-upinterviewstodetermineoutcomesandchangesinkeyexposures2009-
2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019
Secondclinicvisit2014-2017
Baselineclinicvisit2008-2011
StudyTimeline
Page 8
Diseaseandbiometricphenotypes• Prevalent and incident pulmonary diagnoses + exacerbations• Pulmonary Function Testing (basal and post-bronchodilator)• Mortality (total and cause-specific)• Prevalent and incident CVD (MI, stroke, heart failure)•Anthropometry/Weight loss/gain • Lab values: lipids, glucose, OGTT, insulin, inflammation, CBC, Hepatitis A, B C, Total and HDL cholesterol, LFT, renal, etc• Oral/dental health• Pregnancy complications•Hearing • Blood pressure, ECG, echocardiography• Sleep quality and disorders • Cognitive Function• Substance Abuse • Diabetes• Health behaviors (smoking, diet, physical activity, etc)
Page 9
StudyPopulation
Miami,FL
Chicago,IL
SanDiego,CA
Bronx,NY
9
• HispanicCommunityHealthStudy/StudyofLatinos(HCHS/SOL)– Community-basedstudyof16,415menandwomen,18-76yearsofageatbaselineexamination(2008-2011)
– Complexsamplingdesign– Self-identifiedas
• CentralAmerican• Cuban• Dominican• Mexican• PuertoRican• SouthAmerican• Other/Multiple
Pew Hispanic Center (www.pewhispanic.org)
Page 10
4outof5cohortmemberswerebornoutsideofthe50states
Lengthoftimelivinginthe50statesondateofenrollment
Page 11
DemographicCharacteristics
All Cuba Domin.Republic
Mexico PuertoRico
Cent.Amer.
So.Amer.
N 16,415 2,201 1,400 6,232 2,590 1,634 1,022
Men,% 40 46 34 37 41 39 40
College,% 15 20 15 12 14 14 22
Income>$50K,% 11 8 7 14 14 7 11
PreferSpanish,% 77 91 80 81 42 89 89
USresidence<10years,% 31 55 27 27 8 38 47
Daviglus ML JAMA 2012
Page 12
Asthmaprevalence
Mexican PuertoRican
Cuban CentralAmerican
Dominican SouthAmerican
Physician-diagnosedasthma,ever
7.8(6.8–8.9)
30.9(27.8–34)
23.3(21.1–25.7)
12.8(10.7–15)
15.7(13.0–18.7)
10.4(7.9–13.5)
Physician-diagnosedasthma,current
3.4(2.7–4.1)
15.3(13.2–17)
8.6(7.1–10.4)
4.5(3.4–6.0)
6.7(5.3–8.5)
4.3(2.9–6.2)
Page 13
AsthmaprevalenceishighestinLatinoswhocametothe50statesduringearlychildhood
●
●
●●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●●●
●
●
●●●
●●
●
●
●
●●
●●●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
Estimated prevalence of asthma by age immigrated to US
Age immigrated to US
Prev
alenc
e of
ever
asth
ma
diagn
osed
by M
D
●
0
10
20
30
40
50
60%
1 5 10 15 20 25 30 35 40 45 50 55 60 65 70US
born
Note: Prevalence estimates and 95% confidence intervals are from a survey data logistic regression model with effects for age(continuous), sex, Hispanic background, and age at immigration (grouped into 65 categories)
GrahamBarrAmJRespir Crit CareMed 2015
+
Page 14
ElinaJerschow AmericanJournalofPreventiveMedicine2017
Page 15
ElinaJerschow AmericanJournalofPreventiveMedicine2017
Page 16
SOLGeneticProjects
• Omics inLatinos(OLa)GWAS
• PAGEconsortium– Metabochip– MEGAchip
• Wholegenomesequencing
16
Page 17
SOLGeneticProjects
• Omics inLatinos(OLa)GWAS
• PAGEconsortium– Metabochip– MEGAchip
• Wholegenomesequencing
17
Page 18
OLa GWASProjectSNPMicroarrayData
12,803studyparticipantsprovidedconsentforgeneticstudies
Illumina Omni2.5M+~150kcustomcontent‘SoL_HCHS_Custom_15041502_B’
Arraydesign– Papanicolaou(NHLBI),RotterandTaylor(LABiomed)
GenotypingperformedbyIllumina &QCbyLABiomed
QAbySOLGeneticAnalysisCenter(Univ.Washington)
Imputationto1000Gphase3
18
Page 19
AnkleBrachialIndex Dental
Anthropometrics Diabetes
Anxiety/Depression Electrocardiography
BloodCellCount PulmonaryDisease
BloodPressure Sleep
ChronicKidneyDisease mtDNA/YChr
Lipids Reproductivetraits
Smoking
19
SOLGeneticAnalysisWorkingGroups
Page 20
www.olgastudy.org
20ScrolldownforallworkinggroupsandtheirmembersDirectoryhascontactinformation
Page 21
1.EstimaterelatednessusingKING-robust,whichisrobusttodiscretepopulationstructurebutnottoadmixtureordeparturesfromHWEwithinsub-populations.
2.Partitionthesampleintoamutuallyunrelatedsetandtheremaining(relativesoftheunrelatedsetandpossiblyeachother)
3.Performstandardprincipalcomponentsanalysis(PCA)onthesetofunrelatedindividualsandprojectontorelatedindividuals
4.Re-estimaterelatednessusing‘PC-Relate’,whichprovidesunbiasedkinshipcoefficientsinthepresenceofpopulationstructure,admixtureandHWEdepartures,usingindividual-specificallelefrequenciesestimatedfromsampleeigenvectors.
5.Repeatsteps2-5togetfinalsetsofeigenvectorsandkinshipcoefficients
De-convolutionofancestryandrelatednessinadmixedpopulationsMattConomos andTimThornton(UniversityofWashington)
21
Page 22
GeneticAnalysisIssues1. Tieredconsentallowsvariablelevelsofdatasharing.Reconsentingat
eachinpersonvisitmaynecessitatewithdrawalofsubjects
1. Samplesurveydesign,2-stageprobabilitysamplinga. Primarysamplingunit=UScensusblockgroupb. Secondarysamplingunit=householdc. Householdbasedsampling(average1.8enrolledperHH)
2. Relatedness,populationstructureandadmixturea. ~85%ofsubjectsaremutuallyunrelatedb. ~15%areeachrelatedtosomeoneintheunrelatedsetc. SOLparticipantsareverydiverseethnicallyandgenetically
i. Caribbeangroups:Cuban,Dominican,PuertoRicanii. Mainlandgroups:Mexico,CentralAmerica,SouthAmerica
d. Novelmethodusedtode-convoluteancestryandrelatedness(Conomos &Thornton)
MixedmodelandGEEapproacheshavebeendevelopedtohandletheseissuesinassociationtests(GWAS).BothcontrolTypeIerrorwell,butmixedmodelhasmorepower.(Mixedmodel– Conomos +Thornton;GEE- Lin+Tao)
Page 23
Populationstructure
Page 24
Principalcomponentsanalysis(PCA)
Conomos 2016.
PCAplots(EV1-3)
Page 25
Proportionofancestralbackgrounds
Conomos etal.2016.
Page 26
ContinentalancestryproportionsestimatedusingADMIXTUREsoftwareAutosomal
Xchromosome
Page 27
SeparatePCAforeachofthe3ancestries
HCHS-SOLgroups
Referencepopulations
SharonBrowningG32016
Page 28
SeparateclusteringofAfricancomponentofHondurans/Guatemalans
• Garífuna population.
• MayhaveancestryfromspecificpartsofAfrica.
• Ordriftduetosmallpopulationsize.
HCHS-SOLgroups
Referencepopulations
Africa
SharonBrowningG32016
Page 29
ThreedistinctAmerindianclusters
• Mexico,CentralAmerica,SouthAmerica.
HCHS-SOLgroups
Referencepopulations
Americas
SharonBrowningG32016
Page 30
SeparateclusteringofEuropeancomponentofPuertoRicans
• MayincludeancestryfromNorthAfricanorMiddleEast.
• Ordriftduetosmallpopulationsize.
HCHS-SOLgroups
Referencepopulations
Europe
SharonBrowningG32016
Page 31
Eachofthefirst5Eigenvectorsdifferentiatesamongself-identifiedgroups
• Usethese5eigenvectorstoadjustforancestryinassociationtestsandtodefinegeneticanalysisgroupsforstratifiedanalysis• Comparedwithself-identifiedbackground(personalorfamilyplaceoforigin),thegroupsaremorehomogeneousgeneticallyandincludeindividualswithmissingor“other”self-identification 31
Page 32
SOLGeneticProjects
• Omics inLatinos(OLa)GWAS
• PAGEconsortium– Metabochip– MEGAchip
• Wholegenomesequencing
32
Page 33
KariNorthUNC-CH
CALiCOConsortium
PopulationArchitectureusingGenomicsandEpidemiology(PAGE)Study
RuthLoosMountSinaiSchoolofMedicine
MSSMBiobank
TaraMatiseRutgersUniversity
CoordinatingCenterCharlesKooperbergFHCRC
Women’sHealthInitiative
ChrisHaimanUSC
MultiethnicCohortStudy
33
Page 34
ThePAGEStudyPopulationArchitectureusingGenomicsandEpidemiologyIINetwork
• Goal:Investigateancestrallydiversepopulationstogainabetterunderstandingofhowgeneticfactorsinfluencesusceptibilitytodisease.
• FocusonUSminoritypopulations.
17,328
4,696
22,250
653
3,944
1,056
African-American
Asian
Hispanic/Latino
NativeAmerican
NativeHawaiian
Others
ILLUMINA ARRAY SEMINAR, SEPTEMBER 2016
Page 35
35
●
●
●●●
●●●●●
●
●
●●●●
●●●●
●
●
●●●
●●●●●
●
●
●●●
●●●●●
Common Low Frequency Rare
AAC
●
●●
●●●
●●●●
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.4, 0
.5
0.3, 0
.4
0.2, 0
.3
0.1, 0
.2
0.5, 0
.1
0.04,
0.05
0.03,
0.04
0.02,
0.03
0.01,
0.02
0.005
, 0.01
Variant Frequency
Impu
tatio
n Ac
cura
cy, M
ean
r2
Array●
●
●
●
Omni1MOmniExpressAxiom.AFRMEGA
●
●
●●
●●
●●●●
●
●
●●
●●
●●●●
●
●
●●
●●
●●●●
●
●
●●
●●
●●●●
Common Low Frequency Rare
AFR
●
●●
●●●
●●●●
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.4, 0
.5
0.3, 0
.4
0.2, 0
.3
0.1, 0
.2
0.5, 0
.1
0.04,
0.05
0.03,
0.04
0.02,
0.03
0.01,
0.02
0.005
, 0.01
Variant FrequencyIm
puta
tion
Accu
racy
, Mea
n r2
Array●
●
●
●
Omni1MOmniExpressAxiom.AFRMEGA
●
●●
●●●
●●●●
●
●●
●●
●●
●●●
●
●
●●
●
●
●●●●
●
●●
●●
●●
●●●
Common Low Frequency Rare
AMR
●
●●
●●●
●●●●
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.4, 0
.5
0.3, 0
.4
0.2, 0
.3
0.1, 0
.2
0.5, 0
.1
0.04,
0.05
0.03,
0.04
0.02,
0.03
0.01,
0.02
0.005
, 0.01
Variant Frequency
Impu
tatio
n Ac
cura
cy, M
ean
r2
Array●
●
●
●
Omni1MOmniExpressAxiom.LATMEGA
●
●
●●
●●
●●●●
●
●
●
●●
●●●●●
●
●
●
●●
●
●●●●
●
●
●
●●
●●●●●
Common Low Frequency Rare
SAS
●
●
●●
●●
●●●●
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.4, 0
.5
0.3, 0
.4
0.2, 0
.3
0.1, 0
.2
0.5, 0
.1
0.04,
0.05
0.03,
0.04
0.02,
0.03
0.01,
0.02
0.005
, 0.01
Variant Frequency
Impu
tatio
n Ac
cura
cy, M
ean
r2
Array●
●
●
●
Omni1MOmniExpressAxiom.EASMEGA
●
●
●
●●
●●●●●
●
●
●
●●
●●●●●
●
●
●
●
●
●
●●●●
●
●
●
●
●
●●●●●
Common Low Frequency Rare
EAS
●
●
●●
●
●●●●●
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.4, 0
.5
0.3, 0
.4
0.2, 0
.3
0.1, 0
.2
0.5, 0
.1
0.04,
0.05
0.03,
0.04
0.02,
0.03
0.01,
0.02
0.005
, 0.01
Variant Frequency
Impu
tatio
n Ac
cura
cy, M
ean
r2
Array●
●
●
●
Omni1MOmniExpressAxiom.EASMEGA
●
●
●●
●●
●●●●
●
●
●
●●
●●●●●
●
●
●
●●
●●
●●●
●
●
●
●●
●●●●●
Common Low Frequency Rare
EUR
●
●
●●●
●●●●●
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.4, 0
.5
0.3, 0
.4
0.2, 0
.3
0.1, 0
.2
0.5, 0
.1
0.04,
0.05
0.03,
0.04
0.02,
0.03
0.01,
0.02
0.005
, 0.01
Variant Frequency
Impu
tatio
n Ac
cura
cy, M
ean
r2
Array●
●
●
●
Omni1MOmniExpressAxiom.EURMEGA
ComparisonofMEGAImputationAccuracytoOtherCommercialChips(SupplementaryFigure)
Page 36
SOLGeneticProjects
• Omics inLatinos(OLa)GWAS
• PAGEconsortium– Metabochip– MEGAchip
• Wholegenomesequencing
36
Page 37
HCHS/SOLsampleselectionforwholegenomesequencingN=1917asthmacases(everdiagnosedbyaphysician)N=4503controls(neverdiagnosedbyaphysician)
NHGRI/PAGE ~270samplesselectedforhighAmerindianancestryPrimarilyCentralandSouthAmerican(onlyonePuertoRican)WashingtonU.sequencingcenter
NHGRI/CCDG ~4000samplesselectedatrandom Baylorsequencingcenter
TOPMed Totalapprovedallocationof2150 Baylorsequencingcenter1. Allever-asthmacasesnotpreviouslyselectedbyPAGEorCCDG,N=1277.(Bringing
totalofever-asthmacasesto1917acrossallthreesampleselections)
2. SixsamplespreviouslyselectedbyPAGE,whichwillserveascross-sequencingcentercontrols(onepergeneticanalysisgroupe.g.Cuban,Mexican,Dominican,etc),N=6
3. RemainingPuertoRicansnotincludedinPAGEorCCDGselections,N=867.
Page 38
WethanktheparticipantsandstaffoftheHispanicCommunityHealthStudy/StudyofLatinos(HCHS/SOL)fortheircontributionstothisstudy.ThebaselineexaminationofHCHS/SOLwascarriedoutasacollaborativestudysupportedbycontractsfromtheNHLBItotheUniversityofNorthCarolina(N01-HC65233),UniversityofMiami(N01-HC65234),AlbertEinsteinCollegeofMedicine(N01-HC65235),NorthwesternUniversity(N01-HC65236),andSanDiegoStateUniversity(N01-HC65237).Thefollowinginstitutes,centers,andofficescontributedtothefirstphaseofHCHS/SOLthroughatransferoffundstotheNHLBI:NationalInstituteonMinorityHealthandHealthDisparities,NationalInstituteonDeafnessandOtherCommunicationDisorders,NationalInstituteofDentalandCraniofacialResearch(NIDCR),NationalInstituteofDiabetesandDigestiveandKidneyDiseases(NIDDK),NationalInstituteofNeurologicalDisordersandStroke,andNIHOfficeof DietarySupplements.TheGeneticAnalysisCenterattheUniversityofWashingtonwassupportedbyNHLBI andNIDCRcontracts(HHSN268201300005CAM03and MOD03).Additional analysissupportwasprovidedby1R01DK101855-01and13GRNT16490017.GenotypingeffortsweresupportedbytheNIHDepartmentofHealthandHumanServices(HSN26220/20054C),NationalCenterforAdvancingTranslationalScienceClinicalTranslationalScienceInstitute(UL1TR000124),andNIDDKDiabetesResearchCenter(DK063491).