We frequently get texts and emails from friends, colleagues, and readers who come across an article or a study, recognize it as a Random Act of Medicine, and send it over. “This seems like it’s up your alley.” Multiple readers sent over a recent intriguing article in the New York Times titled “Surprising New Research Links Infant Mortality to Crashing Bat Populations.”
The article covered a very clever study by University of Chicago environmental economist Eyal Frank, published in the journal Science. The study—which we’re going to dive into more deeply—illuminates one of the ways our environment and policies surrounding it can impact our health.
The only thing we learned about bats in medical school was that they can carry rabies, so if a person is exposed to a bat that has, say, snuck into their home, they might need to receive post-exposure treatments against rabies (a very burdensome ordeal; our condolences to anyone who has gone through this).
But as kids learn in science class, bats play an important role in our ecosystem. This includes ecosystems that support agriculture in the U.S., as Frank describes in the paper:
Existing research in ecology documents that bats provide biological pest control through their high population size and predation rates on a variety of insects, many of which are crop pests. Insectivorous bats consume 40% and above of their body weight in insects each night.
Damages from crop pests can substantially reduce agricultural productivity. In the US, about 13% of crops are estimated to be destroyed by insects each year, which represents a loss of $27.6 billion a year.
So if something were to happen to insect-eating bat populations, it could result in increases in crop pest insect populations, which in turn could result in loss of crops. Without bats to control the pests, farmers would have to use more insecticides to prevent the crop loss. And if those pesticides have negative effects on human health, then we have a connection between the bat population and the health of the human population.
Starting in about 2006, something did happen to bat populations: a disease called “white nose syndrome,” or WNS, began to spread across insect-eating bat populations in the eastern half of the U.S.
Research indicates the fungus that causes WNS, Pseudogymnoascus destructans, is likely exotic, introduced from Europe. What started in New York in 2006 has spread to more than half of the United States and five Canadian provinces by August 2016, leaving millions of dead bats in its path. WNS causes high death rates and fast population declines in the species affected by it…
Researchers call the disease “white-nose syndrome” (WNS) because of the visible white fungal growth on infected bats’ muzzles and wings. This cold-loving fungus infects bats during hibernation, when the bats reduce their metabolic rate and lower their body temperature to save energy over winter. Hibernating bats affected by WNS wake up to warm temperatures more frequently, which results in using up fat reserves and then starvation before spring arrives.
Assuming that the timing of the emergence of WNS in a local bat population is as good as random with respect to human health—a quite reasonable assumption—then we have a natural experiment on our hands. Changes in human health occurring shortly after WNS infects a bat population can reasonably be attributed to WNS.
But since the fungus that causes WNS has never been shown to directly infect or harm humans, it leaves the other changes brought on by WNS—namely, increased use of toxic insecticides because there are fewer bats around to prevent crop loss—as the best explanation for human health problems, as Frank illustrates:
Question 1: Was there more insecticide use when WNS reduced local bat populations?
WNS first emerged in the state of New York in 2006, and spread southward and westward over the subsequent decade. Of course, we would only expect to see increases in insecticide use after WNS started affecting insect-eating bat populations in an area. Using data from the U.S. Fish & Wildlife Service, Frank measured the amount of insecticide used per square kilometer of cropland in the years before and after WNS was detected in a given U.S. county. Combining all counties into a single analysis centered around the time WNS was first detected locally, here’s what he found:
In the year WNS was first detected in a county, insecticide use increased, and continued to increase over the subsequent 6 years. On average, this represented an increase of 2.7 kg of insecticide per square kilometer of cropland. His results were similar in additional analyses where he accounted for things like weather and larger trends in pesticide use. There was also no similar trend in fungicide or herbicide use associated with WNS entering an area, further confirming the notion that insecticides were being used to address the loss of natural pest control from the bats. Indeed, it appeared that as WNS wipes out bat populations in a county, farmers compensate with insecticides to prevent crop loss.
Question 2: What happened to human health after WNS emerged in an area?
If insecticide use increased as a result of WNS, and if insecticides can cause human health problems, then we should expect to see an increase in those health problems. But are insecticides used to prevent crop loss actually toxic to humans?
There are a number of different insecticide chemicals that are used for agricultural purposes that can pose serious risks to the central nervous system (among other organ systems when humans are exposed to sufficient quantities—after all, the purpose of these chemicals is to kill living organisms). Federal regulations on pesticide use for crops focus on ensuring that the amount of chemicals people are exposed to is sufficiently small as to not harm humans. That amount will be smallest for infants, our smallest humans.
Considering there were not widespread reports of insecticide toxicity among humans, it made sense to look at humans where a small increase in chemical exposure could make a difference: among the infants. So using the same process as before, but this time looking at infant mortality (specifically, “internal” infant mortality that excluded external causes like accidents and homicide), Frank looked at infant mortality rates in counties before and after WNS was first detected, again combining them into a single analysis:
Internal infant mortality rates rose in the years following WNS first detection in a county—an average of about 0.5 deaths per 1,000 live births. Notably, there was no increase in external causes of infant mortality (which we would expect since insecticide use shouldn’t impact these types of deaths), nor were there changes in outcomes like infant birth weight (evidence against less healthy babies being born during this period).
Taken together with the insecticide findings, Frank was able to estimate the impact of insecticide use on infant mortality: every 1% increase in insecticide use resulted in an estimated 0.25% increase in internal infant mortality rates.
Could something else have explained the increases in infant mortality?
Frank’s analyses make it clear that the introduction of WNS to an area resulted in both increases in insecticide use and increases in infant mortality rates. To attribute the increase in infant mortality to the insecticides, we have to assume that WNS cannot lead to infant deaths in other ways.
For example, if there were weather changes that led to both the introduction of WNS to an area and led to increased infant mortality, we couldn’t blame all the infant deaths on the insecticides. But in additional analyses, Frank found that the findings held even when adjusting for weather changes.
Alternatively, if decreasing bat populations led to increases in the mosquito population (perhaps outside of the areas where insecticide was sprayed), which in turn led to more mosquito-born illnesses, we couldn’t blame the insecticides. But no such rise in mosquito-born illness in infants has been reported.
Another possibility might be related to crop destruction itself. If crop destruction affects the supply, and therefore the prices of crops, this might ultimately find its way to families in the form of higher food prices. One would expect this effect to not be localized, however, since the markets for crops are not localized. Relatedly, local economic impacts arising from crop destruction may have impacts on infant mortality, but those economic impacts would need to be large.
So in the absence of any other plausible explanation for increased infant mortality that would have coincided with the emergence of white nose syndrome—and there doesn’t seem to be one—insecticide use remains as the most likely explanation.
What do we do with this information?
One of the reasons this study is remarkable is that it answered two completely different research questions: one about medicine, and one about the environment.
The natural experiment created by WNS provided an opportunity to understand the impact of agricultural insecticides on infants, a very important issue that would otherwise be very difficult to study. This information can inform further development of safe practices in their agricultural use.
But the study also provided solid evidence about our ecosystem—an invasive fungus can have far reaching effects on bat populations, insect populations, our food supply, and infant health. Of course, this is only one example of a single issue on a single health outcome. Surely there are many other connections like these, and taking advantage of natural experiments is a powerful way to study them.
We’ll leave you with some of Eyal Frank’s conclusions from his study: “White-nose syndrome is but one of many threats that bats face, including those that are shared with multiple other species, such as habitat loss and climate change. Improving our understanding of how changes in biodiversity affect human well-being will be important when designing and implementing conservation policies.”
_(33741700644).jpg){kind=link}
