GATT DEPARTMENT OF C TG AT T G C C G T A T C G C A C A G C A T G A A G AT G TG ATAGAG UNIVERSITY COLLEGE CORK
Gut Microbiota, Diet and Health in the Elderly Population
Marcus Claesson 1st October 2012 ISAPP
http://apc.ucc.ie “Gut microbiota as an indicator and agent for nutritional health in elderly Irish subjects”
Why elderly? • Increasing proportion in population • Changes in microbiota composition and activity • Increased infection rates • Increased inflammatory diseases • Prospects for dietary intervention/modulation How does the gut microbiota composition change with age? Infancy • Simple • Temporal instability • Dominated by Bifidobacteria
Adulthood • More complex • Temporal stability • Dominated by Firmicutes & Bacteroidetes
Physiological changes of the aging intestine: • Reduced motility => altered nutrition dynamics • Reduced dentition & taste => altered diet • “Inflamma-aging” => chronic low-level inflammation How does the gut microbiota composition change with age?
•Elderly gut microbiota in a state of flux (Mitsuoka et al. 1978)
•Conflicting age-related changes of the major phylogenetic groups: • Bacteroidetes up; Actinobacteria down (Hopkins et al. 2001) • Bacteroidetes up; Firmicutes down in Irish (Claesson et al. 2011) • No Bacteroidetes change for 100yr old Italians compared to 30 & 70yr olds (Biagi et al. 2010)
• Composition does not vary between European countries (Lay et al. 2005) • Composition does varies between European countries (Mueller et al. 2006) Is community location of the elderly associated with microbiota? • Location: proxy for general health condition • 178 elderly (≥65yrs) Irish subjects • 83 Community-dwelling • 20 Day hospital (out-patient) • 15 Rehabilitation (≤6 weeks) • 60 Long-stay (>6 weeks) • (13 Young healthy controls) • No antibiotics treatment ≤1 month prior sampling • Collected stools samples • 16S rDNA amplicons (454) & shotgun (Illumina) sequencing • Metabolomics (NMR) • Food Frequency Questionnaire => long-term diet • BMI, frailty, malnourishment, depression, cognitive function & dementia 5.4mio 16S rDNA reads => 47,500 OTUs Subjects separated by community location Unweighted UniFrac OTU PCoA Weighted UniFrac OTU PCoA
Community Long-stay Young control Subjects separated by community location
Unweighted UniFrac OTU PCoA
Hierarchical Ward-linkage clustering based on Spearman correlation coefficients of the proportion of OTUs for each subject Genus abundance across locations
Enriched in Community (p<0.005) Enriched in Long-stay (p<0.005)
stay
Roseburia Parabacteroides - Firmicutes/Clostridia/Clostridiales/Lachnospiraceae Bacteroidetes/Bacteroidia/Bacteroidales/Porphyromona Coprococcus daceae Firmicutes/Clostridia/Clostridiales/Lachnospiraceae Eubacterium Hydrogenoanaerobacterium Firmicutes/Clostridia/Clostridiales/Eubacteriaceae Firmicutes/Clostridia/Clostridiales/Ruminococcaceae Subdoligranulum Anaerosporobacter Firmicutes/Clostridia/Clostridiales/Ruminococcaceae Firmicutes/Clostridia/Clostridiales/Lachnospiraceae Dorea Barnesiella Firmicutes/Clostridia/Clostridiales/Lachnospiraceae Bacteroidetes/Bacteroidia/Bacteroidales/Porphyromonada Papillibacter ceae Firmicutes/Clostridia/Clostridiales/Ruminococcaceae
Relative abundanceCommunityin Relative Butyricicoccus Anaerotruncus Firmicutes/Clostridia/Clostridiales/Ruminococcaceae Firmicutes/Clostridia/Clostridiales/Ruminococcaceae abundanceLong in Relative … … … … 13 genera in total 17 genera in total What impact has diet on microbiota?
Food Frequency Questionnaire (FFQ) • Long-term dietary habits • FFQ data for 96% elderly subjects • 147 food types (beef/apples/white rice/potatoes/milk/porridge etc) • Healthy Food Diversity (HFD): how diverse AND healthy a diet is FFQ multivariate analysis
Correspondence analysis Complete-linkage clustering based on
Euclidean distances to PC1
FFQ CoA FFQ
Driving food food Drivingtypes
DG1: “low fat / high fibre” DG3: “moderate fat / low fibre” DG2: “moderate fat / high fibre” DG4: “high fat / moderate fibre” Diversity of microbiota and diet Microbiota & diet by community location
Unweighted UniFrac PCoA vs. FFQ PCA Weighted UniFrac PCoA vs. FFQ PCA
Diet Microbiota Community Long-stay Microbiota & diet by duration in long-stay care
Diet Microbiota N/A (C+DH) Week0to6 (Rehab) Week6toYear1 Year1+ Enterotypes
• Arumugam et al., 2011 Nature: • 39 individuals from 6 countries (+239 US/DK individuals) • “All people can be classified into 3 enterotypes” • Dominant genera: • Bacteroides • Prevotella • Ruminococcus (Blautia / Lachnospiraceae) • Wu et al., 2011 Science: • 98 individuals from the US • Only Bacteroides & Prevotella enterotypes stable • Stable over time • Associated with long-term diet • Bacteroides: high-fat/low-fibre • Prevotella: low-fat/high-fibre Wu et al. Wu et al. Arumagam et al. (weighted) (unweighted) Enterotype clustering in the elderly
Microbiota function: Metabolomics and Metagenomics • 29 subjects representative of C/R/LS • Metabolomics (n = 29) – NMR spectroscopy of faecal water – Spectra -> bins -> metabolites (PCA) • Shotgun metagenomics (27 of 29) – Total extracted bacterial DNA sequenced – 51mio 2x91bp Illumina reads/sample – 126 Gb of DNA sequenced – 2.51mio predicted genes Separation of location by faecal water metabolome Pairwise PLS-DA of NMR spectra
Community Rehab Long-stay
Dr. Martina Wallace and Dr. Lorraine Brennan, Univ. College Dublin Integrating metabolome & microbiota
Co-inertia of microbiota & NMR spectrum metabolome metabolite PCA coloured by location
Associated microbiota at genus level Shotgun metagenome: differentially abundant SCFA genes
Butyrate Acetate Propionate
BCoAt: Butyryl-CoA transferase / Acetyl-CoA hydrolase ACS: Acetate-formyltetrahydrofolate synthetase / Formate-tetrahydrofolate ligase PCoAt: Propionyl-CoA:succinate-CoA transferase / Propionate CoA-transferase Inflammatory markers vary by community location Microbiota-health correlations
Health/clinical markers Possible confounders • BMI: Body Mass Index – Antibiotics: • CC: Calf Circumference • Exclude <1mo • MAC: Mid-Arm Circumference • >1mo had no sign. effect on • SBP: Systolic Blood Pressure µ-biota (α- or ß-diversity) • DBP: Diastolic Blood Pressure – Quantile regression • CCI: Charlson Index of Comorbidity model adjusted for: • Barthel Index of Activities of Daily Living •Age • FIM: Functional Independence Measure •Gender • MMSE: Mini-Mental State Exam •Location • MNA: Mini-Nutritional Assessment •Medication
Microbiota separation correlates with health measures Location-specific unweighted UniFrac PCoAs
All four location subjects Community-only subjects Community Long-stay Long-stay-only subjects
Following adjustment for age/gender/location/medication, microbiota correlates significantly with e.g. frailty and inflammation. Prospective studies needed to establish causality. How to reduce complexity of microbiota composition?
Co-abundance groups (CAGs): groups of genera that are positively correlated with each other Microbiota changes across location
is mirrored by changes in health
STAY, UNHEALTHYSTAY,
-
COMMUNITY, HEALTHY COMMUNITY, LONG Summary (Claesson, Jeffery et al., 2012 Nature)
• Microbiota in elderly is different depending on community location • Driven by habitual diet • Microbiota alterations correlate with health changes especially in long-stay
Diet shapes gut microbiota, which might impact on health in elderly people May lead to carefully designed dietary supplements to promote healthier aging Acknowledgements
Paul O’Toole Ian Jeffery Anthony Fitzgerald Eibhlis O’Connor Denis O’Mahony Siobhán Cusack Paul Ross Hugh Harris Catherine Stanton Susana Conde Gerald Fitzgerald Jennifer Deane Fergus Shanahan Orla O’Sullivan Ted Dinan Mary Rea Martina Wallace Colm Henry Julian Marchesi Mairead Coakley Lorraine Brennan Patricia Egan Michael O’Connor Susan Power Douwe van Sinderen Karen O’Donovan Colin Hill Ann O’Neill Cillian Twomey Norma Harnedy Kieran O’Connor Bhuna Laks Lorraine Brennan Martina Wallace
The Cork City Geriatricians Group