When I was trained as a microbiologist in the 1980s the bacterial universe contained three kinds of microbes: commensals, doing no harm but occupying spaces like the oral cavity and the gut; opportunists, causing harm only in immuno-compromised individuals and pathogens, that cause harm in immune-competent individuals. In the 21th century a series of studies showed a relationship between particular bacteria in the gut and faeces and cancer, colorectal cancer in particular. Subsequently the focus shifted slowly but surely from the identification of a particular microbe in association with a disease to microbial population composition in association with disease . Such studies focusing on cancers and autoimmune diseases relied particularly on representational and thorough sequencing of genes and less on culturing of microbes and were labeled ” microbiome “analyses.
The microbial geneticist Joshua Lederberg, who received the Nobelprize in Physiology or Medicine in 1958, coined the term “Microbiome ” for the ecological community of commensal and pathogenic microorganisms that share our body. Following this original designation in 2001 by Lederberg, confusion took hold on what exactly means what. Four terms are relevant: microbiota, metagenome, microbiome and metabolome.
Generally the microbiota is considered the collective of micro-organisms that normally inhabits an environment, the metagenome is defined as the genes and genomes of the microbiota, including plasmids, the microbiome is generally defined as genes and genomes of the microbiota ( as well as the products of the microbiota and the host environment in some definitions) and the metabolome is defined as the collective metabolites of the microbiome. The key issue on the table is since the discovery of the enormously variable population of microbes in the gut, what the impact is of the microbiome or microbiota on health and disease.
This in turn point to the issue of causality: is a single microbe or a family of microbes or a composite of microbes in any way related to either health in the broadest sense or to disease or its course or its outcome. In the old days these last elements were directly linked to virulence factors, often coded on plasmids and therefore a bacterium like vibrio cholera epidemic is taxonomically defined by its genome and its virulence by in this case two plasmids, one coding a phage receptor and another one coding for the phage carrying the virulence factor. My point is that only sequencing the genome ignores the plasmids that are transferable factors and by not culturing one ignores the potential of outgrowth potential carrying a plasmid or not. In the Science issue of January 15 Allyson Byrd and Julia Segre of the Microbial Genomics Section of the National Human Genome Research Institute at NIH address this issue somewhat. Their contribution is entitled: Adapting Koch’s postulates, criteria for disease causation must take microbial interactions into account.
My view is , that criteria for causation should be adapted, if a clear case of causation invalidates the criteria. The risk of adapting, to be precise increasing the sensitivity of the criteria, is that specificity of the criteria goes down and as a consequence the precision, the accuracy and the predictive value goes down with it. The Science article of Byrd and Segre really is not about causation but about an alternative form of protection ,distinct from the classical immune system consisting of innate and adaptive responses. Byrd and Segre point to either a single bacterium or a population of bacteria neutralizing the virulence factors of a given pathogen solidly identified as the cause of a disease. I agree that there is proof of mechanisms suggesting such protective effects. However this does not negate the potential of a bacterium to be the cause of a disease in the absence of such protective bacteria, therefor it is no argument to change the criteria for causation. Their provocative article ,however does raise the question: what is a healthy microbiome and if, and how, the absence of a healthy microbiome contributes to disease? And does a microbiota or microbiome as a whole exists that either causes or contributes to disease and/or disease progression?
The philosopher David Hume was the first to address rules of causation in 1738 in his book: A Treatise of Human Nature. The next giant of the philosophy of causation is Sir
Austin Bradford Hill, who published more than two hundred years later, in 1965, his groundbreaking article: The Environment and Disease: Association or Causation? The criteria for causation of Hume and Bradford Hill are very close to each other. Hill’s criteria for causation were: temporality , dose-response, consistency, strength of association, analogy, specificity , biological plausibility and experimental proof. Only on the last two Hume missed out in 1738. Again a decade later, in 1976 Alfred Evans directly extended Koch’s postulates for proof that a microbe is the cause of an infection and extended the concept to chronic diseases. By doing so Evans ends up virtually with the same criteria as Hill, however in more clinical and pathogenic terms and linking those to epidemiology. Evans, in order to come to causation requires exposure to an environmental agent and/ or a defect in host response. However, Evans has somewhat more difficulty with a concept as aging as the ” gradual” cause of accumulation of diseases.
The definitions Evans put forward, are really worthwhile to remember when studying or trying to proof causality:
- Prevalence of the disease should be significantly higher in those exposed to the putative cause than in case controls not so exposed.
- Exposure to the putative cause should be present more commonly in those with the disease than in controls without the disease when all risk factors are held constant.
- Incidence of the disease should be significantly higher in those exposed to the putative cause than in those not so exposed as shown in prospective studies.
- Temporarily , the disease should follow exposure to the putative agent with a distribution of incubation periods on a bell shape curve.
- A spectrum of host responses should follow exposure to the putative agent along a logical biological gradient from mild to severe.
- A measurable host response following exposure to the putative agent should regularly appear in those lacking this before exposure (i.e., antibody, cancer cells) or should increase in magnitude if present before exposure.
- Experimental reproduction of the disease should occur in higher incidence in animals or man appropriately exposed to the putative cause than in those not so exposed.
- Elimination or modification of the putative cause or of the vector carrying it should decrease the incidence of the disease.
- Prevention or modification of the host’s response on exposure to the putative agent should decrease or eliminate the disease.
- The whole thing should make biological and epidemiological sense.
I really like the last one, derived more or less from all that is stated above.
The latest fine-tuning of causality criteria was done by Mervyn Susser in his article:
What is a Cause and How Do We Know One? A Grammar for Pragmatic Epidemiology, published in the American Journal of Epidemiology in 1991. Susser published in 1973 the first ‘true’ textbook ( Oxford University Press) on Causal Thinking in the Health Sciences. I consider his criteria combined with the ones put forward by Evans the most complete.
Susser states that the following criteria and definitions ” seem the most useful and the least tautologic” :
- Strength is defined by the size of estimated risk within the constraints of probability levels, confidence levels, or other measurements of likelihood.
- Specificity is defined by the precision with which one variable, to the exclusion of others, will predict the occurrence of another, again to the exclusion of others. Specificity of cause and specificity of effect are subclasses of specificity. 2.1 Specificity in the cause implies, in the ideal, that a given effect has a unique cause. 2.2 Specificity in the effect implies, in the ideal, that a given cause has a unique effect.
- Consistency is defined (inductively) by the persistence of an association upon repeated test ( any of which may be deductive). Survivability and Replicability are subclasses of consistency. 3.1 Survivability is defined by the number and, specifically, the rigor and severity of tests of association. 3.2 Replicability is defined by the number and ,specifically, the diversity of tests of association.
- Predictive performance is defined deductively by the ability of a causal hypothesis drawn from an observed association to predict an unknown fact that is consequent on the initial association.
- Coherence is defined by the extend to which a hypothesized causal association is compatible with pre-existing theory and knowledge. Coherence can be considered in terms of many subclasses. 5.1 Theoretical coherence requires compatibility with pre-existing theory. 5.2 Factual coherence requires compatibility with pre-existing knowledge. 5.3 Biological coherence requires compatibility with current biological knowledge that is drawn from species other than human or ,in humans, from levels of organization other than the unit of observation, especially those less complex than the person. 5.4 Statistical coherence requires compatibility with a comprehensible or, at the least, conceivable model of the distribution of cause and effect ( it is enhanced by simple distributions readily comprehended- for instance a dose-response relation- and is obscured by those that are nonlinear and complex).
All we now need to ask ourselves is: is there such a thing as a healthy microbiome and is diversity the key. Causality criteria are not very helpful in pinpointing a particular microbiome as the cause of health of the human host. And related to that: can we apply the above criteria to define an unhealthy microbiome ( as in : defining the contribution of particular bacteria groups and families in the gut to getting ill or the lack of a diverse healthy microbiome contributing to disease progression). A vaccine or a drug targeting the unhealthy bacteria and those exclusively, may finally proof the importance of that particular target bacterium in causing a particular disease and help define a healthy microbiome as a bacterial community in the gut that reflects the absence of disease or reversely defines an unhealthy microbiome as a bacterial community in the gut that thrives on and sustains an unhealthy, inflamed, may be even , a tumor-ridden gut.