Last week on Science-Based Medicine I wrote an article about the embrace by PETA (people for the ethical treatment of animals) of pseudoscience – in this case they are engaging in a fearmongering campaign linking dairy products to the risk of autism or increasing the severity of autism.

Yesterday I received an e-mail from PETA, essentially doubling down on their prior embrace of pseudoscience:

Dear Steven,

I want to follow up on your story last week about PETA’s campaign that points out how a dairy-free diet may help children with autism. PETA’s website and campaign serve to provide parents with potentially valuable information, albeit mostly anecdotal, from families’ findings—for example, just this week, the editor of Autism Eye magazine, Gillian Loughran, who has a 14-year-old son with autism, contacted us in support of our campaign and wrote a letter to the editor on our behalf (see below). Until such time as there is a large study into whether there is a dairy-autism link (and one we hope will not be funded by the dairy industry), it seems unwise to overlook a growing body of anecdotal information supporting that removing dairy and gluten from the diet of a child with autism may improve the child’s sleep, behavior, and concentration. We hope this letter will change your mind about PETA’s campaign—or, at least, that you will share this letter with your readers so that they can arrive at an informed opinion.

Thank you for your time.

Best regards,

Heather 

There are multiple problems with this position.

I reviewed the evidence in my original SBM article. A 2014 systematic review of studies looking at the association of casein (dairy protein) exposure and autism risk or symptoms concluded:

We observed that the evidence on this topic is currently limited and weak.

Further, the best studies to date are all negative, showing no correlation when exposure to casein is blinded, including this 2010 study, and this 2014 study.

PETA’s position is that it is appropriate to spread fear and recommend major lifestyle changes based upon anecdotal evidence, when the best scientific evidence we have shows the fear is not valid. I would also add that the fear has a low (although non-zero) plausibility. Evidence always needs to be put into the context of plausibility.

I do agree that a large definitive trial would be valuable. I don’t think it is scientifically necessary, but the popular belief in a connection within the autism community would make a definitive clinical trial valuable (although if history is any guide, it won’t end the popular belief).

PETA, therefore, is defending spreading fear about a low probability risk when the current scientific evidence is negative. The problem with this, of course, is that there are countless potential risks out there with low probability and no supporting scientific evidence.

PETA, apparently, wants to spread those that support its ideological agenda.

Next is PETA’s support for anecdotal evidence. This is highly problematic, as I have discussed numerous times before. Anecdotal evidence, meaning uncontrolled subjective observations, are worse than worthless as scientific evidence. They are misleading. They are subject to confirmation bias and are more likely to lead to conclusions that support our prior belief, rather than conclusions that actually are true.

Here is the anecdote from the letter referred to by PETA:

My son, now 14, is a strapping lad who is taller than his father. He is growing, maturing and learning quite well. But when he consumes dairy products, his symptoms come back. He can’t concentrate, focus or sleep well.

At best anecdotes are a starting point for researchers. They are fine as a source of hypotheses to be tested, but they are not sufficient to test those hypotheses.

Should we base our behavior and recommendations on anecdotal evidence? That depends on the nature of the evidence. If it is reported that eating a certain product is causing people to drop dead, then it is prudent to recall the product until the truth is sorted out.

In other words, for objective, dramatic, and immediate consequences, anecdotal evidence carries more weight. (We don’t need a study telling us that jumping out of a plane with a parachute is preferred over jumping out of a plane without one.) For subjective, variable, or delayed outcomes, anecdotal evidence becomes progressively unreliable.

What about the association of dairy consumption and autism symptoms? This is exactly the kind of evidence for which anecdotal observation is most subject to confirmation bias and is therefore least reliable.

As an example – many parents believe, based on subjective experience, that sugar makes kids hyperactive. Blinded evidence, however, clearly shows no correlation.

This is the exact kind of anecdotal evidence now being cited by PETA to defend their unscientific fearmongering of dairy products.

Perhaps the biggest problem with the PETA position, however, is the assumption that avoiding casein is risk or cost free. This is far from the truth.

Parents of some children with autism, depending on the nature and severity of their symptoms, already face an increased challenge. Keeping a child on a strict diet is no easy task, and just adds further burden to the parents.

Further, avoiding major food categories carries with it the risk of poor nutrition. It is not impossible to have good nutrition on a Vegan diet, but it is difficult. When making any public health recommendations you have to consider the impact of those recommendations in the real world, not just the effect in an idealized situation.

If parents are convinced by PETA’s fear campaign to avoid dairy in their young children, what will be the net effect? How many children will this save from autism (the evidence so far suggests the answer is none), vs how many children will have compromised nutrition?

Further, human breast milk protein is 40% casein. In effect, PETA is telling parents not to breast feed, when the evidence suggests that breastfeeding is the best nutrition for infants.

Conclusion

It is interesting that PETA thinks the addition of one anecdote about dairy and autism symptoms would have any effect on my position. I suspect they are just desperately trying to save face in response to the criticism of their blatantly unscientific fearmongering campaign.

There is at present no reason to advise parents to avoid dairy in all children (obviously some children will have problems with dairy, such as intolerance or allergy, but those are special cases).

There is very low plausibility to the notion that dairy, or specifically casein, causes or worsens autism. The scientific evidence we do have shows no connection, and most compellingly, the best data with blinded observations are completely negative.  Anecdotal reports in this case are of the most unreliable nature, and are exactly those that have not held up to scientific evidence in the past.

Further, when considered from the point of view of a full cost and risk-benefit analysis, the cost and the risk of telling parents to avoid casein is far greater than the established risk from casein. Therefore, even a “just in case” approach is not justified.

PETA seems intent, however, on digging itself deeper into this pseudoscientific hole, which will likely just do more damage to their reputation.