Nocturnal asthma and the importance of - PowerPoint Presentation

1. Nocturnal asthma and the importance of race/ethnicity, genetic ancestry, and the early life microbiomeMethods and Application for Health Disparities ResearchHealth Disparities Research CollaborativeAlbert M. Levin, Ph.DDepartment of Public Health SciencesHenry Ford Health System

2. OutlineNocturnal asthmaEthnically diverse study of nocturnal asthma in adolescents and adults with asthmaMethods to study genetic admixtureEarly-life gastrointestinal microbiome and nocturnal asthmaMethods to study the microbiome

3. RAR-Related Orphan Receptor A (RORA) GeneGenome-wide association study of insomniaRORA is a core component gene controlling human circadian rhythms (Sato et al. Neuron 2004)NARG2RORA

4. Moffatt et al. NEJM 2010

5. Nocturnal AsthmaNight time awakening due to asthma symptomsDiurnal variation in lung function

6. Nocturnal AsthmaCommon (30-70%)Increased asthma morbidityExacerbationsUse of controller therapyPoor sleep -> sleepiness during the dayPoor academic performanceIncreased mortalityFew large scale epidemiologic studies performed

7. Study of Asthma Phenotypes and Pharmacogenetic Interactions by Race-Ethnicity Keoki Williams, MD, MPH> 6,000 asthmatics enrolled to dateEthnically diverseAfrican Americans and European AmericansAssess what factors are associated with nocturnal asthma in a ethnically/racially diverse and large sample of asthmatics

8. Definition of Nocturnal AsthmaOvernight pulmonary function testingAsthma Control Test question“How many times in the last four weeks did your asthma symptoms wake you up at night or earlier than usual in the morning”Not at allOnce or twiceOnce per week2-3 times per week≥ 4 times per weekAny vs. none

9. SAPPHIRE Subject Characteristics

10. Factors Associated with nocturnal asthma and differences by race-ethnicity

11. Factors Associated with nocturnal asthma and differences by race-ethnicity

12. Factors Associated with nocturnal asthma and differences by race-ethnicity

13. Factors Associated with nocturnal asthma and differences by race-ethnicity

14. Factors Associated with nocturnal asthma and differences by race-ethnicityAfrican American asthmatics were 2.56 (95%CI 2.24-2.93) more likely to report nocturnal symptoms than European Americans

15. Number of nights with sleep disturbancePtrend<0.001OR = 4.05 (95%CI 3.04-5.45)

16. Could genetic factors play a role?African American asthmatics were 2.56 (95%CI 2.24-2.93) more likely to report nocturnal symptoms than European AmericansAfrican Americans are an admixed populationGenomes are composed of genes from more than 1 ancestral populationsWest Africa and EuropeAmong African Americans, if increasing % West African ancestry (from 0% to 100%) is associated with increasing risk of nocturnal asthma, this would indicate a role of inherited genetic factors.

17. Genetic Admixture

18. Crossing over during meiosis produces admixed chromosomes

19. 1st Generation2nd GenerationAfter ManyGenerations

20. Genome-wide genetic ancestryPercent of ancestry contributed by one of the ancestral populations for an individual% W. African = Green/(Green +Blue)*100W. African European

21. Genetic Admixture

22. Ancestry informative markers (AIMs)Single nucleotide polymorphisms (SNPs; >20 million)ACGTCACTGT[C/T]GCCTTCGAGAIMs are markers that differ in allele frequency between ancestral populationsWest African: C=100% and T=0%Europe: C=0% and T=100%Each person can have 0, 1, or 2 C alleles (i.e. the W. African allele)

23. Ancestry informative markers (AIMs)AIMs are markers that differ in allele frequency between ancestral populationsWest African: C=100% and T=0%Europe: C=0% and T=100%2EuropeanW.Africanlocus ancestry Observed: #W. African allelesAIMS C/T

24. Ancestry informative markers (AIMs)AIMs are markers that differ in allele frequency between ancestral populationsWest African: C=100% and T=0%Europe: C=0% and T=100%2EuropeanW.Africanlocus ancestry Observed: #W. African allelesAIMS C/T

25. Ancestry informative markers (AIMs)AIMs are markers that differ in allele frequency between ancestral populationsWest African: A=100% and C=0%Europe: A=0% and C=100%21EuropeanW.Africanlocus ancestry Observed: #W. African allelesAIMS C/TA/C

26. Ancestry informative markers (AIMs)AIMs are markers that differ in allele frequency between ancestral populationsWest African: A=100% and C=0%Europe: A=0% and C=100%21EuropeanW.Africanlocus ancestry Observed: #W. African allelesAIMS C/TA/C

27. Ancestry informative markers (AIMs)AIMs are markers that differ in allele frequency between ancestral populationsWest African: A=100% and G=0%Europe: A=0% and G=100%212EuropeanW.Africanlocus ancestry Observed: #W. African allelesAIMS C/TA/CA/G

28. Ancestry informative markers (AIMs)AIMs are markers that differ in allele frequency between ancestral populationsWest African: A=100% and G=0%Europe: A=0% and G=100%212EuropeanW.Africanlocus ancestry Observed: #W. African allelesAIMS C/TA/CA/G

29. Ancestry informative markers (AIMs)AIMs are markers that differ in allele frequency between ancestral populationsWest African: G=100% and A=0%Europe: G=0% and A=100%2120EuropeanW.Africanlocus ancestry Observed: #W. African allelesAIMS C/TA/CA/GG/A

30. Ancestry informative markers (AIMs)AIMs are markers that differ in allele frequency between ancestral populationsWest African: G=100% and A=0%Europe: G=0% and A=100%2120EuropeanW.Africanlocus ancestry Observed: #W. African allelesAIMS C/TA/CA/GG/A

31. Ancestry informative markers (AIMs)AIMs are markers that differ in allele frequency between ancestral populationsWest African: G=100% and T=0%Europe: G=0% and T=100%21202EuropeanW.Africanlocus ancestry Observed: #W. African allelesAIMS C/TA/CA/GG/AG/T

32. Ancestry informative markers (AIMs)AIMs are markers that differ in allele frequency between ancestral populationsWest African: G=100% and T=0%Europe: G=0% and T=100%21202EuropeanW.Africanlocus ancestry Observed: #W. African allelesAIMS C/TA/CA/GG/AG/T

33. Genome-wide African Ancestry in SAPPHIRE African Americans (n=1,040)

34. Genome-wide European Admixture in African Americans Population% EuropeanAncestryNew Orleans22.5 ±1.6Pittsburgh20.2 ±1.6New York19.8 ±2.1Maywood, IL18.8 ±1.4Houston16.9 ±1.6Detroit16.3 ±2.7Baltimore15.5 ±2.6Philadelphia13.8 ±1.9Charleston, SC11.6 ±1.3Parra et al. AJHG 1998

35. African Ancestry and Nocturnal Asthma Among African AmericansIncreasing genome-wide % African ancestry is associated with increased risk of nocturnal asthma(OR, 3.47; 95% CI, 0.90–13.39; P = 0.069)%AfricanAncestryLung FunctionFEV1 and FVCNocturnalAsthmaLevin et al. (2014) AJRCCM

36. What does it mean? Racial disparity in prevalence of nocturnal asthma is in part explained by heritable genetic variationEvidence that there are unique genetic variation that differentiates non-nocturnal and nocturnal asthma

37. Human Microbiome ProjectThe first phase was to catalog the microoganisms that live on and in the body in healthy adults.Second phase studies are now exploring the role of the microbiome in both health and disease.The Human Microbiome: Our Second GenomeChristine C. Johnson, PhD

38. The Human Microbiome“In almost every measure you can think of, we're more microbial than human.” –Lita Proctor, Director of the HMPMicrobial cells outnumber human cells 10:1Microbial genetic material outnumber human 100:1 ~ 3 lbs of microbes in the human gut~ 60% of stool dry matter is microbial mass Chuanwu Xi, Biostatistics course presentationCostello et al. (2009)http://www.npr.org/blogs/health/2013/07/22/203659797/staying-healthy-may-mean-learning-to-love-our-microbiomes?ft=1&f=103537970

39. Wayne County Health, Environment, Allergy & Asthma Longitudinal Study (WHEALS) Birth CohortPregnant mothers recruited 2003-2007,Detroit Michigan USA (urban/suburban)Racially and socio-economically diverse1 month study visit: N=110, Median=35 days ,IQR=17 days6 month study visit: N=149, Median=201 days, IQR=37 daysN=1,258N=826N=2594-year Interviewasthma and nocturnalsymptoms assessed

40. Asthma and nocturnal symptomsAsthma – parental report of a physician diagnosis of asthmaNocturnal symptoms – parental report of waking due to nocturnal coughNot when sick with cold or chest infection

41. Asthma in WHEALS and MAAPAsthma prevalence 13% (111 of 826) in the cohortSimilar in MAAP sub-cohort (15%)Asthma1-Month N (%)6-Month N (%)Yes17 (15%)22 (15%)No93 (85%)127 (85%)Total110 (100%)149 (100%)

42. Asthma and Nocturnal SymptomsNocturnal asthma prevalence 52% (58 of 111) in the cohortSimilar in sub-cohort (51% , 20 of 39)AsthmaNocturnal Cough1-Month N (%)6-Month N (%)YesYes8 (47%)12 (55%)No9 (53%)10 (45%)

43. A Brief History of Technology to Study the MicrobiomeEarly studies were culture-dependent, meaning a large amount of bacteria had to be grown in a lab to study it. The vast majority of microbes have yet to be successfully culturedDNA-based culture-independent methods developed in the 1980’s that resolved both of these issuesEarly methods were quite expensive and time consumingNext-generation high-throughput sequencing in 2005 finally made culture-independent methods accessibleMicrobiome research is still in its infancy and is rapidly developing.Morgan & Huttenhower (2012)

44. Why?To determine what bacteria species, and in what abundances, are present in communitiesTo examine differences between several communitiesHow?The 16S rRNA gene is common to all bacteria. We count the number of 16S genes present in a sample as a surrogate measure of number of bacterial species present in a sample.“Tag sequencing”This saves us from having to sequence the entire genomeIntro to Bacterial SequencingMorgan & Huttenhower (2012)

45. High Rank = More GeneralLow Rank =More SpecificOperational Taxonomic Unit (OTU) Grouping at a taxonomic level, such as genus, species, etc.We typically group at the species level to get the most informationSometimes we only know higher-rank classificationsBiological Classification

46. Operational Taxonomic Unit” (OTU) TableSample 1Sample 2Sample 3Sample 4OTU 152010L. acetotoleransOTU 2220150B. acidifaciensOTU 388130A. baumanniiOTU 45211630L. camelliae

47. Alpha Diversity vs. Beta DiversityBoth act as summary statisticsAlpha Diversity: Diversity within a sampleRichness – how many unique species are present?Evenness – how evenly distributed are they?Diversity – how many unique species are present and how evenly distributed are they? Beta Diversity: Diversity between samplesPairwise metrics to build a square distance/dissimilarity matrix between all samplesScale is usually 0-1, where 1 = very dissimilar, 0 = very similarThere are dozens of metric choices

48. Alpha and Beta DiversitySubject ASubject BRichness = 3Richness = 3 Beta Diversity(Subject A, Subject B) = 0

49. Subject ASubject BRichness = 3 Richness = 3 Beta Diversity(Subject A, Subject B) = 1Alpha Diversity High Beta Diversity

50. Statistical MethodsComposition differencesPERMANOVA (Canberra distance)Bacterial Community Indices Richness, evenness, diversityKruskal-Wallis (non-parametric t-test)Operational Taxonomic Unit (OTU) testsZero-inflated negative binomialFalse discovery rate (FDR) “q-value” to account for multiple testingQ-value < 0.05 (i.e. FDR of 5%) KingdomPhylumClassOrderFamilyGenusSpecies

51. Community Indices – 6 monthp=0.024p=0.004p=0.019

52. Compositional Differences – 6-monthAsthma with andwithout nocturnal symptoms (p=0.012)

53. OTUs Associated with Nocturnal Cough: 6-Month Visit

54. Predicted Function of Differential OTUs27 biological pathways were predicted to be active and also associated with asthma with nocturnal symptoms (q<0.05) 21 (77.8%) enriched function in nocturnal asthmaticsExamplesCircadian RhythmSteroid BiosynthesisCaffeine metabolismPesticide degradation (Atrazine and DDT)

55. ConclusionsFirst studies to :Identify a racial disparity in nocturnal asthma, which is partially explained by genetic admixtureSuggest an association between the early life GI microbiome and nocturnal symptoms in children.Support the growing body of literature linking the microbiome to asthma:RiskSeverity

56. AcknowledgmentsHenry Ford Health System Kevin Bobbitt PhDAndrea Cassidy-Bushrow PhD Christine Cole Johnson PhD Suzanne Havstad MA Christine Joseph PhD Haejin Kim MD Kyra Jones MEd Alexandra Sitarik MS Ganesa Wegienka PhD Kim Woodcroft PhD Edward M. Zoratti MDUniversity of California-San FranciscoHomer Boushey MD Kei Fujimura PhDSusan Lynch PhDUniversity of Michigan Nicholas Lukacs PhDGeorgia Regents University Dennis R. Ownby MDWe thank the participants and families of those who have participated in both the SAPPHIRE and WHEALS cohorts.MAAP InvestigatorsFundingNational Institute of Allergy and Infectious DiseasesParticipantsSAPPHIRE InvestigatorsL. Keoki Williams MD Karen Wells MSYun Wang MS

57.

58. Your “11th Organ System”Though some microbes cause human disease, the vast majority (>99%) of them are beneficial and essential to our health.Trains the human immune system starting at birthAids in digestionInfluences metabolism and fat storageStimulates cell growthThe human microbiome is a delicate ecosystem, and imbalanced of this system has been linked to:Asthma, Allergies, Crohn's Disease, Celiac Disease, Obesity, Diabetes, AutismChuanwu Xi, Biostatistics course presentationCostello et al. (2009)http://www.npr.org/blogs/health/2013/07/22/203659797/staying-healthy-may-mean-learning-to-love-our-microbiomes?ft=1&f=103537970

59. How Do We Go From Raw Sequences to an OTU Table?QIIME: An open source software package for comparison and analysis of microbial communitiesQIIME takes users from their raw sequences to processed/cleaned/analyzable data, to downstream statistical analysis/visualization/graphics. Wraps many other software packagesCan create a pipeline of several steps in one placehttp://qiime.org/Sample 1Sample 2Sample 3Sample 4OTU 152010L. acetotoleransOTU 2220150B. acidifaciensOTU 388130A. baumanniiOTU 45211630L. camelliae?

60. Obtaining OTUs from SequencesSimilar sequences are clustered into groups, which form OTUsTwo main clustering strategies:Closed ReferenceThrow away clusters based on unknown dataLess bacteria, but “safer” dataWe have much more confidence that all taxa clusters are real bacteria, not irregularities in the data Open ReferenceKeep clusters based on unknown data“Novel” data: we have the ability to detect previously undiscovered or unclassified bacteriaTradeoff: more bacteria = more computation timeDo you hypothesize that differences are due to a small number of rare taxa or not?

61. Compositional DifferencesNo evidence of differences in the 1-month visit samples (p=0.59)Evidence of differences in the 6-month visit samples (p=0.004)

62. Compositional DifferencesEvidence of differences in the 6-month visit samples (p=0.004)Group 1Group 2P-value*R2Nocturnal asthmatic Non-nocturnal asthmatic 0.016%Nocturnal non-asthmatic0.105%Non-nocturnal non-asthmatic0.031%Non-nocturnal asthmaticsNocturnal non-asthmatic1.005%Non-nocturnal non-asthmatic0.361%Nocturnal non-asthmaticNon-nocturnal non-asthmatic1.001%*P-values corrected for multiple testing using a Bonferroni correction.

63. Asthma and Nocturnal SymptomsAmong non-asthmatics, nocturnal symptom prevalence 10% (69 of 715) in the cohortSimilar in sub-cohort (9% , 19 of 220)AsthmaNocturnal Cough1-Month N (%)6-Month N (%)YesYes8 (47%)12 (55%)No9 (53%)10 (45%)NoYes8 (9%)11 (9%)No85 (91%)116 (91%)Total110 (100%)149 (100%)

64. Asthma and Nocturnal SymptomsAsthmaNocturnal Cough1-Month N (%)6-Month N (%)YesYes8 (7%)12 (8%)No9 (8%)10 (7%)NoYes8 (7%)11 (7%)No85 (78%)116 (78%)Total110 (100%)149 (100%)