Thursday, May 22, 2014

Causation, Correlation, and H. pylori

Thanks to the end of the spring semester, I finally have time to sit and get some thoughts down. This post was specifically motivated by a couple of recent items that have been making their rounds on the internet. The first of these is a new book and associated book tour by Dr. Martin Blaser entitled "Missing Microbes: How Overuse of Antibiotics is Fueling our Modern Plagues". While I haven't read the book yet (but plan too on a couple of upcoming cross country flights), I have seen a couple of interviews with Marty. Without saying much else, to my microbiologist ears some of these interviews have been filled with a little bit too much hyperbole. I say this carefully as I have an immense amount of respect and admiration for Marty as a scientist. I also say this knowing that some believe the only way to draw public attention to the problems of antibiotic overuse is to stand up confidently and surely and overemphasize solidity of the scientific basis for these arguments.  It's quite OK if that is your viewpoint, although on the other hand this same logic has fueled a campaign of mistrust of climate scientists over global warming. Since I worked with H. pylori (HP) during graduate school, have been in the HP literature, and since I've had some conversations with Marty, I've been thinking about the topics of his book for a while. Hopefully I can provide a slightly different yet still enlightening viewpoint. In my mind microbes rule the world and overuse of antibiotics is a horrible problem of modernity, but some nuanced, yet key, points have been glossed over in the media coverage. Won't say too much since they're pretty self explanatory, but the second motivation for this post was a brilliant set of spurious correlations.

So let's dive in to the nougat of some of Blaser's arguments using published articles. Helicobacter pylori is a well known human pathogen and causative agent of ulcers, chronic gastritis, and stomach cancer. A Nobel prize was even awarded to Drs. Marshall and Warren for demonstrating this link to gastric disease. The estimated number of infected worldwide is 50%, which I 'm pretty sure is required citation in the introduction of every HP article. This is a pretty old estimate, however, and I'm not sure it's still that high (not going to go there in this post). Of that 50%, only a small percentage (~10% is a number that seems to pop up a lot) actually develop gastric disease due to HP. By these numbers, there are a significant amount of non-disease causing infections, and this has partly contributed to the speculation that HP might have some benefits to humans. It's likely that every mammal has some kind of Helicobacter if you look closely enough, but HP seems to only naturally reside within humans. There are also other species of Helicobacter that infect humans. There are few if any natural reservoirs of HP other than humans (it really doesn't survive well outside of hosts, although researchers are always looking and occasionally find interesting leads), and HP has been associated with humans for pretty much as long as there have been humans. Given this information, there is a lot of fodder for thinking that HP might have co-evolved with humans and to potentially help humans. The data, and our understanding of evolution, doesn't necessarily back this idea up.

I'm going to focus on asthma, but for the most part any other "benefit" of HP you may stumble across falls into the same explanatory ballpark. There does seem to be a negative correlation between HP and asthma (more HP, less asthma) as well as links to immune responses. HP also appears to protect neonatal mice from asthma in relevant model systems.  I'm not going to dispute these pieces of data, the work is pretty solid, but it's unclear what else correlates with these observations. Those that grow up in first world countries grow up in a much cleaner environment than their ancestors. There is a growing body of evidence that animal and plant (unpublished but it's coming) immune systems need to be trained by some microbes while developing. Therefore, growing up in more "sterile" environments could lead to immune disfunction. This line of thought has been crystallized in the "hygiene hypothesis". HP is likely disappearing in frequency due to modern living, but so is our exposure to other microbes. It's tempting to focus on associations between HP and asthma, but such correlations could be explained though the loss of other microbes (singly or as flora). The article I linked to above didn't test other members of the microbiome or gastric pathogens for the ability to tone down asthma, and it's unclear whether living or dead HP would suffice for this effect. I also want to point out that the main pathogenesis factor identified for HP in the gastric environment (CagA toxin) is not required for asthma protection. There's some smoke there, but there are a lot of other possibilities that could explain correlations between HP and asthma. Moreover, and this is total handwaving speculation, but if it's just the antigenic effects of HP that matter immune responses we may be able to just synthesize an antigen pill or vaccinate to replace the effect.

I also want to touch a bit on the idea that we are driving "good" microbes extinct with antibiotics. While it's clear that antibiotic doses certainly alter our microbiomes temporarily, it's unclear whether any microbes are truly going extinct. Current data supports the idea that the human microbiome is fairly resilent to change. That is, if you perturb the microbiome (say with antibiotic doses) it tends to bounce back into shape given enough time. Since very few of us are on constant antibiotic doses, it's still unclear whether there is a huge change in your microbiome over the course of your life or through multiple random antibiotic treatments. Of course, while acute infections by pathogens like Clostridium difficile can occur if you clear out most of your resident microbes, these situations seem to be cured through exchange of microbes from other sources. It's also important to keep in mind that it appears to be very difficult to drive microbes to extinction in any instance. We've only eliminated two pathogens in nature through vaccination regimes during the course of modern medicine, rinderpest and smallpox. There is no reason to believe overall selection pressures to develop antibiotic resistance are any different across bacteria, and we worry (rightly so) about multi-drug resistant bacterial pathogens. Why would antibiotics only selectively kill the good guys?

Lastly, a point on Human-H. pylori co-evolution. HP might very well affect the incidence of asthma (or other chronic disorders like Crohn's disease) and such chronic situations can make our lives very, very difficult. When researcher's speak of "benefits" of microbiota, they are implicitly speaking about co-evolution. For instance, since X bacteria does Ygood,  this relationship has evolved to be beneficial to humans over time. This equation is leaving out discussion of an intrinsically important metric, human fitness. While evolution can discriminate between very small fitness differences in human populations, I'm going to go out on a limb and guess that asthma has not been a huge selective force throughout the course of human history. Fitness is all about your offspring surviving, and while asthma may make like a little more difficult, there is no evidence that chronic conditions such as asthma really and truly affect human fitness. In the absence of a fitness effect, it's impossible to have co-evolution. I may be completely wrong about this (and if I am, please point me towards the data!), but in the absence of changes to human fitness a lot of these microbiome links become just-so stories.

One story to finish this piece off....I started working with H. pylori in 2001 in graduate school in Oregon. It wasn't until Spring of 2004 I had my first asthma attack. Was it HP that gave me asthma, or was it the constant flood of pollen from very frisky trees? It's an N of 1, but at least in my case there is a positive correlation between HP and asthma.

Disqus for