Missing Microbes: How the Overuse of Antibiotics Is Fueling Our Modern Plagues (19 page)

But people who never acquired
H. pylori
in childhood or who had their bacteria knocked out by antibiotics have high levels of acid well past the fortieth year of life. Thus, for possibly the first time in human history and prehistory, large numbers of people are reaching middle age with fully intact acid secretion. For them, the stomach contents moving up into the esophagus are highly acidic and have more digestive enzymes—and are more damaging. And with the markedly decreased prevalence of
H. pylori
in childhood, most of today’s children are growing up with different acid regulation than the children of prior generations had, with
H. pylori
out of the physiologic picture. Reflux in children, once ultra-rare, keeps increasing, and many children today are being treated with medications to reduce gastric acid levels. Could these events be connected?

We were finding that
H. pylori
, discovered as a pathogen, is really a double-edged sword: as you age, it increases your risk for ulcers and then later for stomach cancer, but it is good for the esophagus, protecting you against GERD and its consequences, including a different cancer. As
H. pylori
is disappearing, stomach cancer is falling, but esophageal adenocarcinoma is rising. It is a classic case of amphibiosis. The facts are consistent.

11.

TROUBLE BREATHING

Most people are aware that asthma, a disease recognized since antiquity, has become a massive health problem. Statistics from developed countries, where records have been kept for the past seventy years or more, show that rates have been doubling and tripling. The graphs look like what you would want for your pension fund but instead indicate an increase in terrible suffering and sometimes early death.

Physicians have known for many years that GERD and asthma are somehow linked. Many GERD patients develop the wheezing, coughing, and constricted airways emblematic of asthmatic attacks. And when asthma patients receive treatment for GERD to reduce stomach acidity, their breathing often improves. Despite the connection, most physicians believe that GERD accounts for only a small fraction of asthma cases.

One theory that explains how the two disorders are related is purely mechanical. When stomach acid travels up the esophagus, it can spill down the windpipe, creating irritation. But that explanation fails to account for the allergies and hay fever often associated with asthma. Asthma is the leading edge of a group of related disorders that involve too much sensitivity to foreign substances.

After our studies showed that
H. pylori
can protect us from GERD, I began to wonder whether it protects us from asthma as well. Maybe the rising incidence of asthma is related to fewer children acquiring
H. pylori
early in life and more children having it eliminated inadvertently with antibiotics. Could subclinical unrecognized cases of GERD caused by the lack of
H. pylori
be driving the asthma epidemic?

Although it made sense and was consistent with what we were learning at the time, the mid-1990s, the connection presented a major and controversial leap. The fall of
H. pylori
and the rise of asthma could both be true but unrelated, just as the rise of asthma parallels the rise in home TV sets or the number of Volkswagens on the road.

I attempted to get several colleagues who worked in lung diseases interested in studying this potential relationship with me, but it was too far-fetched, and besides the medical community at large was focused on
H. pylori
’s dangers. To investigate my hypothesis, I needed to study a population of patients with asthma, but without a collaboration with a clinical scientist working in that area, it would be impossible.

Then, in 2000, I moved from Vanderbilt in Tennessee to New York to become the chair of medicine at New York University. It was a great opportunity to return to my alma mater and to help build a strong department. But despite the administrative tasks and pressure, I didn’t want to give up research. At a new place, I had new chances. I asked my colleagues, “Who works on asthma here?”

Everyone pointed to Dr. Joan Reibman, a specialist in lung diseases, who in 1991 had established a clinic at Bellevue Hospital for adult asthma. Joan listened politely to my ideas, but with little enthusiasm. One of Joan’s great intellectual strengths is her skepticism. Wild ideas come up all the time. She wasn’t buying mine unless I could show her evidence. Fair enough.

Joan agreed to use the clinic she started at Bellevue Hospital to enroll patients in the type of study we wanted. Their friends and relatives who did not have asthma would serve as the controls. She performed a battery of tests to characterize their lung function and their allergies. Fortunately for me, beginning in 2002, she also collected and froze blood specimens from the participants that we could use to assess their
H. pylori
status. Joan’s support was critical to test the hypothesis; she was and still is dedicated to finding answers for how to improve care for patients with asthma.

By 2004, Joan’s team had collected blood from more than five hundred people. We agreed that she would send my team the serum specimens under code to blind us from knowing who had asthma and who was a control. This removed chances of unintentional bias in the analysis. Guillermo ran the blood tests, and then we divided the results into positive, negative, and uncertain, and through repetition we resolved essentially all the ambiguities. Later that year, we sent the results to Joan and her team, which included Michael Marmor, a seasoned epidemiologist accustomed to the type of statistical analyses we needed. Joan called a few weeks later and told me that, much to her surprise, she and Mike had found an inverse association between
H. pylori
and asthma. Still she was dubious. After all, how could a stomach microbe protect against asthma?

We agreed to meet and review the results. A week later, Joan, Mike, and the rest of their team came to the VA hospital where my lab is and where I have a small office. Joan described the participants—318 patients with asthma and 208 healthy controls—and announced that the statistical analysis showed that people positive for
H. pylori
were 30 percent less likely to have asthma than people without the organism. This was true even when they took into account variables that could otherwise explain the propensity for asthma.

This was the first, early support for my theory. Even so, there were many possible ways to interpret the data.

“What about
cagA
?” I asked. We had run blood tests for both
H. pylori
status, to see who was carrying the organisms, and also for
cagA
-positive strains, just as we had done for ulcer disease, stomach cancer, and esophageal diseases.

“We haven’t analyzed that yet,” Joan said.

I was disappointed because
cagA
is really the key marker. The
cagA
-positive strains are the ones that are the worst to have in relation to ulcers and the best to have in relation to a healthy esophagus. If I had to predict where the strongest story for asthma would be, I would be betting on
cagA
as the best predictor for protection against asthma.

“Well,” Joan said. “We’ll just have to look at that later.”

Then Mike interrupted. “Wait a minute!” he said. “I should be able to get that.”

With that, he began to type on his laptop. We all watched in silence. After about thirty seconds of typing, with a flourish Mike pushed a final button. And a few seconds later he read from the screen: “
cagA
+
; odds ratio of 0.6.”

Eureka. That meant that people carrying
cagA-
positive strains of
H. pylori
were 40 percent less likely to have asthma than people without
H. pylori
. This was stunning.

The strains most related to stomach cancer and ulcers turned out to be the most beneficial for GERD and now for asthma. It seems like a paradox, but the finding now can be explained by the fact that
cagA-
positive strains are the most highly interactive with their hosts. By then, we understood how these strains operate: they are constantly injecting their own materials into human stomach cells. It is as if there are two different populations of
H. pylori
strains. Some are vigorous and highly interactive—the
cagA
-positive strains. The others, the
cagA
-negative strains, may be regarded as more sluggish; they have much less contact with the cells of their human host.

The
cagA
-positive strains probably are living a little closer to our cells, while the others are further offshore in the lumen. Thus, not surprisingly, the
cagA
-positive strains are the most damaging to the stomach wall. But because they are the most interactive, they
also
have the potential to be the most beneficial in helping regulate our physiology.

Next, Joan examined the clinical records of the asthma patients to learn how old they were when their disease was diagnosed. Were they children or adults when their symptoms first appeared? We found that
H. pylori–
positive subjects were, on average, twenty-one years old when their asthma started. For those without
H. pylori,
the average age of onset was eleven. This was a striking difference. It showed that lack of
H. pylori
was more commonly associated with childhood-onset asthma and suggested that in people who were bound to get asthma, the presence of
H. pylori
might delay the process. A couple of years later, a large study was done among children in Manitoba, Canada. The investigators found that antibiotic use in the first year of life was associated with a significantly greater chance of having asthma at the age of seven. They weren’t looking for
H. pylori,
but their findings were consistent with my general hypothesis.

In Joan’s study, the blood samples also were tested for antibodies to allergens, so that we could examine whether having
H. pylori
correlated with allergic responses. Here again we saw a connection: the presence of
H. pylori
was associated with fewer reactions to the allergens, suggesting that
H. pylori
can protect against allergy.

An abstract of our findings was submitted to the 2005 annual meeting of the American Thoracic Society and presented in May of that year. Unfortunately, it was received with a huge yawn. Our work was outside the mainstream of asthma studies, and even the pulmonary disease specialists on Joan’s team were not as impressed by the results as I was.

*   *   *

I pressed on. Could we reproduce our findings in another population? If the findings were real, we should see them again. I thought about using a large study, called NHANES III, in which the United States selected twenty thousand people as a representation of the overall population and between 1988 and 1994 gave them a series of health exams.

The results of blood tests were still available, including the subject’s
H. pylori
status. Sitting in the same little office at the VA where I had met Joan’s team some months earlier, in March 2006 I suggested to Dr. Yu Chen, a young epidemiologist new to NYU, that we use these data to test the hypothesis of an inverse association of
H. pylori
and asthma. Yu agreed and was able to find records from NHANES III with information on the asthma status and the
H. pylori
status of more than 7,600 people. Joan’s study, involving 500 people, was large as studies go, but this one was about fifteen times bigger.

On May 5, 2006, Yu e-mailed me: “I have done some analysis using the NHANES data…” she wrote. “… It’s kind of strange.”

Rushing to catch a plane to Chicago, I stuffed the tables she sent me into my briefcase. A few hours later in the quiet of the cabin I took them out. The results were clear: Yu had done a terrific job, and her analyses showed that there was an inverse association of
H. pylori
and asthma in NHANES III. It was especially pronounced for
cagA
-positive strains. In fact, the percentage margin was just about 40 percent, almost identical to that of Joan’s study.

Here was a second large, independent, blinded study that had shown almost exactly the same result as the first. This could not be explained by random chance. Although other issues needed to be considered, and these data did not indicate whether lack of
H. pylori
predisposed to asthma or vice versa, sitting in the plane with the hum of the engines outside and my neighbor asleep next to me I knew at that moment that my hypothesis was correct. I felt I had hiked up a very long trail and finally, gasping for breath and soaked in sweat, I reached the summit of the peak I had been climbing—it was a moment of exhilaration.

Yu’s study showed more nuance. All of the inverse association was observed in children under the age of fifteen. The effect was specific for childhood-onset asthma but there was no effect for adult-onset asthma. Although incidence of asthma has increased since World War II, it has risen most dramatically among children. Childhood-onset asthma affects kids in cities and in the country, all over the developed world, but the poor are especially susceptible. There are many theories to explain this, but a popular one is that poor children are more heavily exposed to cockroaches and other insects, which may be important triggers of asthma. But not every child in a cockroach-infested home becomes asthmatic, and plenty of children get asthma with nary a cockroach in sight. To me, the mystery is not why someone becomes allergic and wheezes after being exposed to a cockroach. That part I understand. Rather the question is why after the exposure, when most children stop wheezing, other children do not.

The NHANES III records also contained data on hay fever, or allergic rhinitis. Again we found an inverse effect, again in children but not in adults, again stronger for
cagA-
positive strains. This work provided the first indication that the presence of
H. pylori
in a child’s stomach might protect against hay fever. As with asthma, hay fever has been getting more common in children, as
H. pylori
has been disappearing.