No one has the answer yet!
Everyone’s been talking about the health situation in Brazil lately, with the recent Zika virus outbreak and reports of increases in babies born with microcephaly and other birth defects. If you are unfamiliar with what’s going on, I suggest you read this detailed guide to the situation from the New York Times. (I have also included several links to more information about the topics I discuss at the bottom of this article.)
But what I want to talk about is how, in the quest to find answers for the apparent increase in babies with microcephaly, so many scientists, journalists, and public figures have gotten the concepts of correlation and causation all botched up. As scientists continue to search for a causal link between the Zika virus and the apparent increase in babies born with microcephaly in Brazil, competing theories have been popping up like daisies. The World Health Organization (WHO) released a new report several days ago attempting to squelch the rumors circulating through the media, but the rumors won’t stop until people understand what it actually means to prove cause and effect.
Here’s what I have to say: At this point in the process of investigating the microcephaly situation, anyone who claims to have a definitive answer is either lying or doing bad science. Real, good scientists don’t go throwing causation around all willy-nilly.
Recently, a group of doctors in Argentina, called Doctors in Pesticide-Sprayed Towns, has published a report directly blaming a widely-used mosquito pesticide, pyriproxyfen, as the cause of the apparent microcephaly outbreak. They base their claim primarily on the fact that many regions reporting upticks of microcephaly cases have added pyriproxyfen to their drinking water to control mosquito populations since 2014.
I have a problem with their claim. Not because it’s alternative, but because it’s deliberately presented by Doctors in Pesticide-Sprayed Towns as the only plausible explanation when it is not supported by any hard science. It’s based only on circumstantial evidence. Yet, it has been repeated, uncritically, over and over again.
Considering strength of evidence
There is nothing wrong with alternative theories. It’s important to consider many plausible options when investigating an emerging health issue. But the caveat is this: it’s important to qualify the evidence. It’s not enough to say, “I have this theory and there’s lots of evidence to support it.” First, it’s important to qualify how good the evidence is and to what degree it supports the conclusion.
The relationships between human health, viruses, and environmental toxins are very complex, and absolutes are rare. That’s why it’s important to put evidence into perspective; certain types of evidence are more compelling than others. For instance, a correlation between mothers reporting Zika symptoms during pregnancy and then having babies with microcephaly is an OK piece of evidence. But finding evidence of the Zika virus in the placentas and brains of microencephalic babies is even better evidence. Still, neither of these is enough to establish a cause-effect relationship.
Taking all factors into consideration
In the midst of a global health concern, it is important to take a big-picture perspective and consider all the evidence available. Complex problems require complex analysis, and no one bit of data can tell the whole story. It doesn’t mean much to say that one piece of evidence suggests a very strong correlation if you ignore an equally important piece of evidence that suggests the opposite. To put forth an honest theory, it’s just as important to discuss the evidence working against your claim as the evidence working for it. Even if the evidence for one theory is strong, the evidence for a different theory might be even stronger. In this case, the evidence for a Zika-microcephaly link isn’t overwhelming, but it is still much stronger than the evidence for a pyriproxyfen-microcephaly connection.
The problem is that many of the theories surrounding microcephaly are little more than skepticism. They put forth a possibility based on a limited amount of evidence, while overstating the importance of the evidence and ignoring data to the contrary. This is what the supposed pesticide-microcephaly link seems to be; there is a little bit of evidence to suggest a link, but only on a superficial level. It only seems probable if you ignore all the evidence that is inconsistent with that link.
The Zika-Microcephaly Claims
So far, scientists think the most plausible cause for the apparent increase in microcephaly cases in Brazil is the Zika virus. They think the virus might cause developmental defects in the fetus when a pregnant woman is infected with the virus. The WHO and other researchers are still researching the connection, but the data is still somewhat ambiguous.
The Doctors in Crop-Sprayed Towns suggest the Brazilian Ministry of Health, the WHO, and the Pan American Health Organization (PAHO) have been quick to blame the Zika virus in order to cover up a pesticide poisoning problem. In reality, all but the Ministry of Health have been careful to say that they don’t have enough evidence yet to prove any kind of causal link. In their reports the WHO, PAHO, and the Center for Disease Control and Prevention (CDC) in the US freely admit the limitations of their evidence, discuss alternative interpretations, and include evidence that suggests there’s no link at all.
The organizations leading the research efforts admit that the evidence correlating microcephaly and Zika isn’t at all definitive, and that there are many limitations to the data they have to work with. One such limitation is the difficulty researchers have had determining with any certainty how serious the “outbreak” even is. Records reporting how often babies were born with microcephaly before the outbreak seem unreliable, and thus no one’s sure how big the increase in babies born with the condition is. Moreover, it seems that doctors have grossly overestimated the number of microcephaly cases since the beginning of the outbreak, as the vast majority of potential microcephaly cases have been reviewed and determined not to be microcephaly.
What’s most important, though, is that the researchers are aware of the limitations of their evidence and that they are communicating this ambiguity to the public.
Here’s a look at how the WHO and the Center for Disease Control (CDC) have presented their evidence for the Zika-microcephaly connection:
- “The increased occurrence of microcephaly associated with cerebral damage characteristically seen in congenital infections in Zika virus-affected areas is suggestive of a possible relationship. Additional studies are warranted to confirm the association and to more fully characterize the phenotype.” –CDC
- “A causal relationship between Zika virus infection and birth malformations and neurological syndromes has not yet been established, but is strongly suspected. ” –WHO
- “Six countries/territories (Brazil, French Polynesia, El Salvador, Venezuela, Colombia and Suriname) have reported an increase in the incidence of cases of microcephaly and/or Guillain-Barré syndrome (GBS) following a Zika virus outbreak. Microcephaly has so far been reported only from Brazil and French Polynesia. Puerto Rico and Martinique have also reported cases of GBS associated with Zika virus infection, but without evidence of an overall increase in the incidence of GBS.” –WHO
- “…before the November MoH alert…infant head circumference was not routinely recorded. Hence, it is possible that mild cases of microcephaly might not have been reported.” –CDC
- “Because Zika virus infection was not laboratory-confirmed in infants or their mothers, the history of a nonspecific rash illness during pregnancy is subject to recall bias and might have resulted in misclassification of potential Zika virus exposure.” –CDC
- “…special studies, including case-control studies, are needed to confirm the association, determine the magnitude of the potential risk, and identify other possible risk factors.” –CDC
As you can see, the WHO and CDC are not making any strong claims about a link until they can gather more reliable data and have a solid enough scientific basis to strongly support a causal link. They report on the limitations of their evidence and research as well as contradictory evidence that challenges the link between Zika and microcephaly. Let’s contrast this dialogue with the way that the Doctors in Crop-Sprayed towns present their theory that the pesticide pyriproxyfen is the cause of microcephaly in Brazil.
The Pyriproxyfen-Microcephaly Claims
The Doctors in Crop-Sprayed Towns present their evidence in a very different way than the WHO and CDC. Despite the fact that they are proposing a never-before-heard-of link between microcephaly and the pyriproxyfen pesticide, they seem overly confident in their claims, despite having weaker evidence for this link than the WHO does for the Zika link. Some of their claims sound plausible on the surface, but don’t take into consideration equally compelling evidence to the contrary. What they’re doing is jumping to conclusions—specifically, jumping to the conclusion that a weak correlation between this specific pesticide and microcephaly means that the pesticide is a cause.
Here’s a look at how the Doctors in Crop-Sprayed Towns present their evidence for the pesticide-microcephaly connection in their report :
- “A dramatic increase of congenital malformations, especially microcephaly in newborns, was detected and quickly linked to the Zika virus by the Brazilian Ministry of Health. However, they fail to recognize that in the area where most sick persons live, a chemical larvicide producing malformations in mosquitos has been applied for 18 months, and that this poison (pyroproxyfen) (sic) is applied by the State on drinking water used by the affected population.”
- “Malformations detected in thousands of children from pregnant women living in areas where the Brazilian state added pyriproxyfen to drinking water are not a coincidence, even though the Ministry of Health places a direct blame on the Zika virus for this damage.”
- “The evolution from zygote to embryo, from embryo to foetus and from foetus to newborn, is not far from the development process of the mosquito affected by pyriproxyfen. They also very easily try to ignore that in humans, 60% of our active genes are identical to those of insects such as the Aedes mosquito,”
- “Brazilian doctors (Abrasco) are claiming that the strategy of chemical control is… a commercial manoeuvre from the chemical poisons industry, deeply integrated into Latin American ministries of health as well as WHO and PAHO.”
Note how much stronger the language the Doctors from Crop-Sprayed Towns use is in comparison to the WHO or CDC. They describe pyroproxyfen as a “poison” and insist that the increase in microcephaly and the somewhat recent introduction of the pesticide “are not a coincidence.” In addition, they reveal their political bias when they directly imply a global conspiracy involving the WHO and PAHO (and, perhaps, Monsanto), to poison Brazilians for profit, which clearly is not a scientific claim. It’s also worth mentioning that Doctors from Crop-Sprayed Towns is an advocacy group that has campaigned against using all kinds of pesticides and GMOs.
The biggest problem I see here is that the authors insist that their results cannot mean anything but causality, which immediately calls into question their skills and integrity as scientists. Good scientists would require vigorous investigation and a preponderance of evidence before making such a claim—which is exactly the kind of rigorous treatment scientists are giving the Zika-microcephaly theory before putting too much confidence in the connection. The truth is, the fact that the pesticide happens to be present in the same place as some of the microcephaly cases is only a mild correlation supported by circumstantial evidence. And circumstantial evidence is all they have. Something perhaps worth looking into eventually, but nothing to draw any firm conclusions from.
In addition, they are ignoring some pretty compelling evidence to the contrary:
- Although the authors say that the WHO and the Brazilian ministry of health was quick to place “direct blame” for the developmental disorders on Zika, this is not entirely true. The Brazilian health minister did show a great deal more confidence than researchers have when he said that, “For us, there’s no doubt that the microcephaly epidemic is a consequence of the epidemic of Zika virus…” However, as I’ve said, the WHO and CDC have all been careful not to presume a causal link between Zika and microcephaly until they have more evidence. But even so, the evidence for a Zika link is much stronger than a pyriproxyfen link, and confidence in the link has been growing steadily among scientists.
- Other viruses that are similar to Zika have been known to cause microcephaly, and there is a scientific basis for understanding how these viruses cause birth defects like microcephaly. However, there is no known biological basis for a pesticide like pyiproxyfen to cause microcephaly and other developmental defects in humans or other mammals.
- There is no reason to think that pyriproxyfen would cause developmental abnormalities in humans; it is a mild larvicide (kills insects in the larval stage) with no plausible mechanism for causing microcephaly, even in very large doses, and has been determined safe for drinking water. Their claim that pyriproxyfen could affect human development because it can stunt insect development doesn’t make sense if you understand how mammalian development is entirely different from invertebrate (insect) development. As pharmacologist and toxicologist Ian Musgrave from the School of Medicine Sciences at the University of Adelaide in Australia says, “Pyriproxyfen acts by interfering with the hormonal control growth cycle of insects from hatching, to larvae, to pupa. This hormone control system does not exist in organisms with backbones, such as humans, and pyriproxyfen has very low toxicity in mammals as a result.”
- Some cities that experienced an apparent increase in microcephaly did not use pyriproxyfen. According to the director of Disease Control and Diseases at the Health Department of Pernambuco, a region hard-hit by both microcephaly and the Zika virus, called attention to the fact that one of the cities reporting the most microcephaly cases, Recife, doesn’t even use pyriproxyfen. If pyriproxyfen is really behind the microcephaly increase, you would not expect to see microcephaly increasing in places that don’t use pyriproxyfen.
- The Pyriproxyfen pesticide was approved by the US Environmental Protection Agency in 2001, the WHO in 2004 and is widely used in many countries. It is an ingredient in hundreds of pesticide products and has been used in the US and in Europe for many years (although not in drinking water, even though it is considered safe for this use). However, the apparent increase in microcephaly cases only appeared in the last few years, and has not appeared at all in many other countries that regularly use the pesticide.
- Although the authors claim that other countries with Zika virus outbreaks haven’t experienced increased microcephaly cases, there’s no way to know if this is true. Doctors in French Polynesia, where there has been a recent Zika outbreak, are investigating a possible increase in developmental defects.
Here’s what the Brazilian Ministry of Health had to say (English translation via Google Translate): “Unlike the relationship between the Zika virus and microcephaly, which has had its confirmation attested in tests that indicated the presence of the virus in samples of blood, tissue and amniotic fluid, the association between the use of pyriproxifen and microcephaly has no scientific basis. Importantly, some localities that do not use pyriproxifen were also reported cases of microcephaly.”
While the Zika-microcephaly link does have a scientific basis, it’s certainly not beyond doubt, even though the Brazilian Health Minister seems to think so. It’s actually very difficult, and requires extensive research, to prove that a correlation means what you think it means, and this is why the World Health Organization has been so cautious not to claim a cause-effect relationship between Zika and microcephaly. Despite many pieces of good evidence pointing in that direction, there’s still too many unknowns and too much contradictory evidence. Scientists need much more than that to make any firm conclusions.
Mere correlation: the case of the storks and the babies
It’s important to realize that a correlation can be very, very strong and still have nothing to do with cause and effect. The classic example from a 1911 textbook uses the example of a strong correlation between storks and baby births to illustrate how two things that are correlated can have no real connection to one another. The number of human babies babies born in an Alsatian village was directly correlated with the number of storks that nested in the city. If you confuse correlation with causation you could come up with this conclusion: storks actually deliver human babies, just like the nursery tale. But, in actuality, there was a third, completely unrelated factor that explained them both: the size of the village. Obviously, a larger village is going to have more people, and more people means more babies born. A larger village also means more houses, with more chimneys, which were a preferred nesting site for storks. So instead of more storks causing more babies, the correlation could be simply explained by the fact that larger villages provide more nesting sites for storks, and also more people to give birth to more babies.
In the stork example, it’s easy to see the disconnect between the correlation and the cause, because common sense tells us that it’s impossible for storks to have any real effect on baby births, or vice versa. But when it comes to topics we have little expertise in, or that are poorly understood, it’s harder to see that disconnect, and we fall for cause-effect theories that are actually just based on correlation. “More storks = more babies born” is an easy fallacy to see, because most people know something about how baby births work. On the other hand, more pesticides = more babies born with microcephaly doesn’t immediately stand out as incorrect, because most people, and even most scientists, don’t understand the complex effects of pesticides. But ignorance shouldn’t lower our standard of evidence, and we should never draw conclusions that aren’t actually there in the data.
Making tough decisions with imperfect science
There’s one important consideration that I have not mentioned yet. In the case of a debilitating human condition, we usually don’t have the luxury to wait for a clear understanding of the condition or its causes—we’re forced to act on the information we have, however imperfect it may be. That’s why the WHO has declared a Public Health Emergency despite not having any definitive answers yet, and that’s why one southern state in Brazil has preemptively discontinued using the pesticide pyriproxifen.
But it is in these situations that we need to be even more diligent about qualifying the evidence we have. A grab-bag full of individual pieces of data, some supporting one cause and some supporting another, doesn’t help us figure out the best way to move forward and minimize human risk. All evidence is not created equal; only by carefully analyzing each piece of evidence and qualifying it based on its strength, in comparison to competing evidence, can we find the best working theory—or at least the one that’s the least imperfect.
I am not saying that the pesticides are not worth looking into. I’m not saying it’s impossible that they could be a cause or contributing factor to microcephaly. What I am saying is that, right now, there’s no good reason to think that it’s a cause, at least right now. Let’s let a suspicion remain just that—a suspicion—until there’s some real, compelling evidence to take this theory seriously. In the meantime, there’s no reason to cause unnecessary alarm, and there’s no reason to take funds and focus away from research into microcephaly causes that actually have a scientific basis.
If you are interested in looking at the evidence in more depth, here are some links to more information: