<?xml version="1.0" encoding="UTF-8"?>
<opml version="2.0">
  <head>
    <title>095-03 Dysbiosis, microbial metabolites, microbiome therapeutics, and causal inference</title>
    <ownerName>Integrated Medical Foundations</ownerName>
  </head>
  <body>
    <outline text="Dysbiosis, metabolites, and causal inference">
      <outline text="What dysbiosis does and does not mean">
        <outline text="Altered community associated with disease"/>
        <outline text="No universal taxonomic definition"/>
        <outline text="Loss of function or expansion of harmful activity"/>
        <outline text="Illness, diet, drugs, transit, inflammation all change it">
          <outline text="Association must not be read as cause"/>
        </outline>
      </outline>
      <outline text="Sampling and measurement limits">
        <outline text="Stool is accessible but sheds luminal material"/>
        <outline text="Time, storage, oxygen, extraction, transit shift results"/>
        <outline text="Relative abundance can rise while counts are unchanged"/>
      </outline>
      <outline text="Sequencing and complementary methods">
        <outline text="Sixteen-S amplifies marker regions">
          <outline text="Usually genus or approximate species only"/>
        </outline>
        <outline text="Shotgun metagenomics resolves genes and strains">
          <outline text="Detects dead organisms, cannot prove expression"/>
        </outline>
        <outline text="Culture proves viability and allows phenotype testing"/>
        <outline text="Quantitative PCR is sensitive for sought targets only"/>
        <outline text="Absolute quantification prevents compositional misreading"/>
      </outline>
      <outline text="Short-chain fatty acids and bile acids">
        <outline text="Anaerobic fermentation produces short-chain fatty acids">
          <outline text="Butyrate fuels colonocytes, shapes barrier and immunity"/>
          <outline text="Stool level is production minus uptake"/>
        </outline>
        <outline text="Bile acids conjugated by liver, modified by microbes">
          <outline text="Antibiotics reduce conversion, altering C. difficile"/>
          <outline text="Protective or injurious by species and site"/>
        </outline>
      </outline>
      <outline text="Other metabolites and drug interactions">
        <outline text="Tryptophan metabolism yields indoles"/>
        <outline text="Choline and carnitine become trimethylamine, then the oxide">
          <outline text="Diet and kidney clearance confound the association"/>
        </outline>
        <outline text="Protein fermentation gives ammonia, phenols, sulfide"/>
        <outline text="Cross-feeding: one species&#x27; waste is another&#x27;s substrate"/>
        <outline text="Microbes activate prodrugs and inactivate active drugs">
          <outline text="Response association may reflect severity or diet"/>
          <outline text="Mechanism needs pathway isolation and exposure change"/>
        </outline>
      </outline>
      <outline text="Inflammation and Clostridioides difficile">
        <outline text="Inflammation raises oxygen and electron acceptors">
          <outline text="Facultative organisms exploit the new conditions"/>
          <outline text="They are markers and amplifiers, not original cause"/>
        </outline>
        <outline text="C. difficile shows ecological causation">
          <outline text="Antibiotics disrupt colonisation resistance"/>
          <outline text="Spores germinate, vegetative cells expand, toxins injure"/>
          <outline text="Toxigenic detection without diarrhoea may be colonisation"/>
        </outline>
      </outline>
      <outline text="Microbiome therapeutics">
        <outline text="Faecal transplantation restores community function">
          <outline text="Highly effective in selected recurrent infection"/>
          <outline text="Screening reduces but cannot remove unknown risk"/>
          <outline text="Success here does not justify unregulated broad use"/>
        </outline>
        <outline text="Probiotics are strain, dose, and host specific">
          <outline text="Evidence does not transfer between products"/>
        </outline>
        <outline text="Prebiotics are selectively used substrates">
          <outline text="Fibres may raise fatty acids but worsen bloating"/>
        </outline>
        <outline text="Diet acts on microbes and host directly">
          <outline text="Short-term taxonomic change is not an outcome"/>
        </outline>
        <outline text="Precision tools: narrow agents, phage, engineered bacteria">
          <outline text="Phage therapy faces range, immunity, manufacturing"/>
          <outline text="Ecological replacement can open unintended niches"/>
        </outline>
      </outline>
      <outline text="Study designs and causal inference">
        <outline text="Association studies confounded by drugs, diet, geography">
          <outline text="Regression cannot control unmeasured factors"/>
          <outline text="Behaviour change before diagnosis gives reverse causation"/>
        </outline>
        <outline text="Longitudinal studies give order, not causality">
          <outline text="Preclinical physiology may drive early change"/>
        </outline>
        <outline text="Randomised interventions are stronger if they hit the mediator">
          <outline text="Diet and antibiotics act on the host too"/>
        </outline>
        <outline text="Gnotobiotic animals show biological possibility">
          <outline text="Donor communities change in a new host"/>
          <outline text="Transfer is support, not proof for human disease"/>
        </outline>
        <outline text="Mendelian randomisation limited by noise and pleiotropy"/>
        <outline text="Classifiers may predict from treatment or geography"/>
        <outline text="Koch&#x27;s postulates fit polymicrobial disease poorly">
          <outline text="Use temporality, dose, manipulation, reversibility"/>
          <outline text="A community function can be causal across taxa"/>
        </outline>
      </outline>
      <outline text="Biomarkers, claims, and the hierarchy">
        <outline text="Biomarkers need locked protocols and external validation">
          <outline text="They must improve decisions beyond clinical data"/>
          <outline text="Report performance, comparator, population, indeterminates"/>
        </outline>
        <outline text="Personalised claims often exceed evidence">
          <outline text="No single optimal community exists"/>
          <outline text="Relative abundance advice medicalises normal variation"/>
          <outline text="Strongest use: pathogen detection and selected transplant"/>
        </outline>
        <outline text="Therapeutics must define the intended mechanism">
          <outline text="Judge by patient outcome and safety"/>
          <outline text="Benefit possible without permanent engraftment"/>
        </outline>
        <outline text="Causal hierarchy: measure, associate, order, reproduce">
          <outline text="Show mechanism, perturb it, improve an outcome"/>
          <outline text="Dysbiosis is a hypothesis, not a diagnosis"/>
        </outline>
      </outline>
    </outline>
  </body>
</opml>
