Everyone knows the pleasures of having his prejudices confirmed by the evidence. The pleasures of changing one’s mind because of the evidence are somewhat less frequently experienced, though none the less real. Among those pleasures is that of self-congratulation on one’s own open-mindedness and rationality. It would therefore delight me to learn that my prejudice about obesity — that it is a natural consequence of overeating, which is to say of human weakness and self-indulgence — was false.
I therefore read with interest and anticipation a recent article in the New England Journal of Medicine with the title “Microbiota, Antibiotics, and Obesity.” The connection of antibiotics with obesity had not previously occurred to me; perhaps the real reason why so many people now have the appearance of beached whales was about to be revealed to me.
The Danish authors of the article start by telling us that the worldwide problem of growing obesity is more complex than commonly assumed, and that “dietary and genetic factors have but partial roles in the development of obesity.” But how partial is partial?
The authors go on to tell us that very early life is crucial in the establishment of the intestinal flora upon which our nutrition so strongly depends. It can be disturbed by the very early exposure to antibiotics, possibly with permanent effects.
Experiments on mice have shown that young mice exposed to penicillin during the mother’s pregnancy and during the weaning period had long–lasting changes in their body composition and biochemistry. They lay down more fat, their liver enzymes alter so that they turn more food into fat, and their bone is less dense, than in mice not exposed to penicillin. Young mice given penicillin after they had weaned did not show the same changes. It is an age-specific effect.
Mice can be raised so that their intestines are completely germ-free. In an experiment on such mice, it was shown that if their sterile large intestines were colonized by bacteria from mice that had been exposed to penicillin they laid down more fat than mice with sterile large intestines colonized by the bacteria from untreated mice. Moreover, mice colonized by penicillin-treated bacteria had reduced intestinal immune responses. Whether this later led to food allergy the authors did not say.
As the authors say with commendable understatement, “In humans, similar studies are difficult to conduct.” One would certainly hope so. But epidemiological studies have shown that children born by caesarean section or treated very early in life with antibiotics are more likely to become fat later in life. These are only statistical associations, however, and no causal relationship between antibiotics early in life and later obesity has been proven in humans. It would be wrong to withhold life-saving antibiotics just because of a hypothetical risk of subsequent obesity.
What do the studies on mice prove?
As is often the case, the authors are more enthusiastic about their pet theory than the evidence they provide in its favor warrants. In the first place, man is not a mouse. In the second, the authors give no indication of the size of effect exerted by antibiotics on the fatness of mice. They use the word “significant” without specifying in what sense they are using it. Statistical significance is not the same as significance in any other sense. It seems to me likely that if the effect was very large the authors would have said so, indeed trumpeted it. Here is a statistic that did not bark.
This article, then, does not force me to give up my prejudices as I had hoped it might. As yet, I do not have to give up the idea that obesity is principally the natural consequence of overeating, and therefore of human weakness. And is there anything more gratifying to contemplate than human weakness?
image illustration via shutterstock / Tom Burlison