henhao
Sign in
Genetic & Genomic MedicineNature Genetics
Peer-reviewed journal article

Combined effects of host genetics and diet on human gut microbiota and incident disease in a single population cohort

Youwen QiniD, Aki Samuli HavulinnaiD, Yang LiuiD, Pekka JousilahtiiD, Scott C. RitchieiD, Alex TokolyiiD, Jon G. SandersiD, Liisa M ValstaiD, Marta BrożyńskaiD, Qiyun ZhuiD, Anupriya TripathiiD, Yoshiki Vázquez‐BaezaiD, Rohit Seth LoombaiD, Susan C. ChengiD, Mohit JainiD, Teemu NiiraneniD, Leo LahtiiD, Rob D. KnightiD, Veikko V SalomaaiD, Michael InouyeiD, Guillaume MériciD

Abstract

Co-evolution between humans and the microbial communities colonizing them has resulted in an intimate assembly of thousands of microbial species mutualistically living on and in their body and impacting multiple aspects of host physiology and health. Several studies examining whether human genetic variation can affect gut microbiota suggest a complex combination of environmental and host factors. Here, we leverage a single large-scale population-based cohort of 5,959 genotyped individuals with matched gut microbial shotgun metagenomes, dietary information and health records up to 16 years post-sampling, to characterize human genetic variations associated with microbial abundances, and predict possible causal links with various diseases using Mendelian randomization (MR). Genome-wide association study (GWAS) identified 583 independent SNP-taxon associations at genome-wide significance (p<5.0x10-8), which included notable strong associations with LCT (p=5.02x10-35), ABO (p=1.1x10-12), and MED13L (p=1.84x10-12). A combination of genetics and dietary habits was shown to strongly shape the abundances of certain key bacterial members of the gut microbiota, and explain their genetic association. Genetic effects from the LCT locus on Bifidobacterium and three other associated taxa significantly differed according to dairy intake. Variation in mucin-degrading Faecalicatena lactaris abundances were associated with ABO, highlighting a preferential utilization of secreted A/B/AB-antigens as energy source in the gut, irrespectively of fibre intake. Enterococcus faecalis levels showed a robust association with a variant in MED13L, with putative links to colorectal cancer. Finally, we identified putative causal relationships between gut microbes and complex diseases using MR, with a predicted effect of Morganella on major depressive disorder that was consistent with observational incident disease analysis. Overall, we present striking examples of the intricate relationship between humans and their gut microbial communities, and highlight important health implications.

PDFDOIData
Full-text digest ↓

02 /Specification & Classification

Click any tag to browse matching papers
Citations
585
FWCI
39.50
Field percentile
top 1%
OpenAlex Topics

In-depth summary

✦ Multi-agent verified full-text synthesis

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 1 (13 Sept 2020) · CC BY 4.0

In 5,959 Finns with genotypes and stool metagenomes, GWAS found 583 independent SNP-taxon associations; dairy intake modified LCT effects on Bifidobacterium, and ABO-linked Faecalicatena lactaris tracked secretor status.

Earlier host-genetics studies of the gut microbiome replicated poorly across cohorts with differing sampling and methods, and most hits did not survive multiple-testing correction. Here one homogeneous Finnish cohort with matched genotypes, shotgun metagenomes, diet questionnaires and up to 16 years of registry follow-up yields an LCT-Bifidobacterium signal stronger than a 28-cohort, 18,473-person meta-analysis and permits diet-by-genotype and incident-disease tests.

Key results

  1. ShownA genome-wide scan of 2,801 prevalent gut taxa found widespread host-genetic associations. Three loci (LCT, ABO, MED13L) passed a study-wide threshold that corrects for 1,328 effective tests.

    5,959 individuals, 7,979,834 variants; 478 GTDB taxa (17% of those tested) associated at p<5.0×10^-8; 583 independent SNP-taxon associations after GCTA-COJO; study-wide threshold 3.8×10^-11; rs3940549 (LCT region, r^2=0.87 with rs4988235) with Actinobacteria p=5.0×10^-35; rs545971 (ABO) with F. lactaris p=1.1×10^-12; MED13L rs187309577 with Enterococcus p=1.8×10^-12; median λGC 1.0051. · Genome-wide association analysis of gut…; Genome-wide association analysis

  2. ShownDairy intake modified the LCT association. Lactose-intolerant (rs4988235:CC) people who reported regular dairy had more Bifidobacterium, and lactase-persistent (TT) people showed no dairy effect. Negativibacillus and CAG-81 sp000435795 showed the inverse pattern, and UBA3855 sp900316885 was lower in TT dairy consumers.

    Among CC individuals, Bifidobacterium was higher with regular dairy (n=736) than without (n=245), p=1.75×10^-13 (Wilcoxon); CT was intermediate; age did not alter the trend. Negativibacillus and CAG-81 sp000435795 were lower in CC than TT dairy consumers (p=0.049, p=0.041); UBA3855 sp900316885 was lower in TT dairy consumers (p=8.23×10^-5). · Combined effect of host genetics and dietary…; Figure legends

  3. ShownThe ABO effect on Faecalicatena lactaris, a mucin degrader, was largely driven by secretor status (non-secretors showed no variation across ABO genotypes) and stayed higher in A/B/AB-antigen secretors at both low and high fibre intake. Collinsella lacked its blood-antigen CAZyme families.

    Secretors vs non-secretors: F. lactaris higher within rs545971:CT (p=3.6×10^-4), TT (p=9×10^-4), and blood types A and AB (each p=1.24×10^-5), but not within CC (p=0.4339), B or O; non-secretors flat across genotypes. A/B/AB-secretors above others in both low- and high-fibre quartiles (p=1.15×10^-9, p=4.4×10^-3). GH110, GH136 and CBM32 found in F. lactaris, none in 9 Collinsella species. · Functionally distinct ABO-associated bacteria…; Prediction of ABO blood groups and secretor…; Profiling of carbohydrate-active enzymes…

  4. SuggestedBidirectional Mendelian randomization (MR) nominated many microbe-disease links. One of them, higher Morganella raising the risk of major depressive disorder (MDD), agreed with incident MDD over up to 16 years of registry follow-up.

    96 MR predictions over 46 diseases (34 microbe-to-disease), e.g. Faecalicoccus on anorexia nervosa OR=1.8 per SD (95% CI 1.3-2.5, p=2.0×10^-4, IVW). Only Morganella-MDD held observationally: Cox HR=1.11 (95% CI 1.01-1.22) per SD, 181 incident events, adjusted for age, sex, BMI. Morganella's lead variant rs192436108 in PDE1A had p=6.16×10^-8, just short of genome-wide significance. · Causal inference predictions between microbes…; Bidirectional two-sample Mendelian randomization…; Cox proportional hazards regression

  5. SuggestedA low-frequency MED13L variant (MAF=0.0111) was associated with Enterococcus faecalis, and E. faecalis was modestly higher in people who already had colorectal cancer (CRC) at sampling. The authors could not test the proposed CRC link robustly.

    rs143507801 (MAF=0.0111; 130 GA heterozygotes, 1 GG homozygote) associated with E. faecalis at p=7.26×10^-11; rs187309577 in MED13L with genus Enterococcus at p=1.8×10^-12. E. faecalis slightly higher in 14 individuals with CRC at sampling than in 5,941 without (p=0.014, Wilcoxon); genotype-by-CRC groups were too small to compare. · MED13L association with Enterococcus faecalis as…; Genome-wide association analysis of gut…; Figure legends; Unexplored links with disease and the nervous…

Methods, data and reproducibility

Linear mixed-model GWAS (BOLT-LMM; covariates 10 genetic PCs, age, sex, genotyping batch) of centre-log-ratio-transformed abundances of 2,801 GTDB taxa from shallow shotgun stool metagenomes, with GCTA-COJO conditional analysis. Follow-ups: genotype-by-diet stratified Wilcoxon tests, CAZyme profiling of reference genomes, bidirectional two-sample MR against 46 diseases, and Cox models of incident MDD.

  • FINRISK 2002 (FR02) Finnish population cohort: imputed array genotypes, shallow shotgun stool metagenomes, diet questionnaires, national register follow-up to 31/12/2018 · human · n = 5,959 individuals in GWAS (7,231 stool samples sequenced)
  • Finnish-population-specific imputation reference panels · human · n = 2,690 whole-genome and 5,092 whole-exome sequencing samples
  • GTDB release 89 reference genomes (custom Centrifuge index; CAZyme scans) · bacteria (Bifidobacterium, F. lactaris, Collinsella genomes scanned) · n = 327 Bifidobacterium, 2 Faecalicatena lactaris, 15 Collinsella genomes scanned
  • MR-Base GWAS summary statistics (autoimmune/inflammatory, cardiometabolic, psychiatric/neurological, bone, cancer and other diseases) · n = 46 diseases
  • 1000 Genomes EUR panel used as the LD reference for clumping · human (1000 Genomes)
Methods & tools
BOLT-LMM, GCTA-COJO, plink v2.0, Eagle2 v2.3, IMPUTE2 v2.3.0, FlashPCA2, Centrifuge, matSpDlite, TwoSampleMR, MR-PRESSO, run_dbCAN2 v2.0.11, R packages compositions, vegan, survival
Data & accessions
https://thl.fi/en/web/thl-biobank/for-researchers
Reagents & environment
Versions given for Eagle2 v2.3, IMPUTE2 v2.3.0, plink v2.0, run_dbCAN2 v2.0.11 and GreenAlgorithms v1.0; references were GTDB release 89 and CAZy (31 July 2019 update). No container or requirements file stated.
Assay conditions & compute
GreenAlgorithms v1.0 puts the main computation at no less than 531.94 kg CO2e; hardware, runtime and memory not stated.
Reproducibility gaps
Individual-level genotype, metagenome and health data are available only by written application to the THL Biobank. No study analysis code is linked; the only code URL is the third-party run_dbCAN2 repository.
Limitations (6)
  • Diet was self-reported, which the authors acknowledge is not fully reliable; they argue that cohort size and effect magnitude still make the dairy result robust. · stated by the authors
  • There was no information on lactose malabsorption symptoms in genetically lactose-intolerant people who reported regular dairy intake. · stated by the authors
  • The MED13L, E. faecalis and CRC link could not be validated: there were too few CRC cases (n=14) and variant carriers, and genotype-by-CRC groups could not be compared. · stated by the authors
  • MR predictions that use microbial traits may be distorted by multiple unaccounted confounders; the authors present them as leads for future research. · stated by the authors
  • All associations come from one cohort of European-ancestry Finns (non-European samples excluded) with no independent replication cohort. The Morganella-MDD hazard ratio's lower confidence bound is 1.01, after 96 MR predictions were screened. · noted in this summary
  • The mechanistic readings (competition for lactose, F. lactaris use of secreted antigens) rest on SNP-predicted blood group and secretor status and on CAZyme gene content from few reference genomes (2 for F. lactaris). No functional assays were done. · noted in this summary

Who should read:Read this if you study host genetic effects on the gut microbiome, gene-by-diet interactions such as LCT and dairy, or Mendelian randomization that uses microbial abundances as exposures.

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISKurilshikov '21Lopera-Maya '22Xu '19Liu '21Wang '21Liu '22
73% semantic matchGenetics
Alexander Kurilshikov, Carolina Medina‐Gómez, Rodrigo Bacigalupe+97 more·Nature Genetics·Jan 18, 2021· 1,785 citations