For the anti-vaccinationists out there: The results of a real “vaxed versus unvaxed” study

ResearchBlogging.orgIn many ways, the anti-vaccine movement is highly mutable. However, this mutability is firmly based around keeping one thing utterly constant, and that one thing is vaccines. No matter what the evidence, no matter what the science, no matter how much observational, scientific, and epidemiological evidence is arrayed against them, to the relentlessly self-confident members of the anti-vaccine movement, it’s always about the vaccines. Always. Vaccines are always the root many human health problems, be they asthma, autoimmune diseases, autism, and chronic diseases of all types. Everything else is negotiable. For instance, back when Andrew Wakefield ignited a scare about the MMR vaccine by publishing a fraudulent case series linking the MMR to a new syndrome consisting of regressive autism and enterocolitis, it was the measles vaccine that caused this syndrome. Here in the U.S., it was the mercury in the thimerosal preservative that used to be used in many childhood vaccines until 2001 that was the cause of all evil. However, as scientists did more and more studies, testing vaccines to see if they were associated with an increased risk of autism and found zero, nada, zip association with either vaccines or the thimerosal preservative in vaccines, the anti-vaccine movement was nothing if not mutable. Before long, Jenny McCarthy was declaring that it was the “toxins” in vaccines that were causing autism, and Generation Rescue was asserting that children were getting “too many too soon.” Of course, as far as the anti-vaccine movement was concerned, the beauty of these ideas was (and is) that they are much harder to falsify scientifically because they are so much more vague.

Of late, the anti-vaccine movement has hit upon a new strategy. Specifically, they are demanding what they like to call a “vaxed versus unvaxed” study. Basically, their claim is that unvaccinated children are so much healthier than vaccinated children, and they think that such a study would prove it. Of course, they only hit on this message after making some rather embarrassing missteps. In particular, they didn’t seem to realize that a randomized, double blind study of vaccination according to the currently recommended schedule versus unvaccinated children was totally unethical. So, they figured out another angle. They acknowledge that a randomized trial of unvaccinated versus vaccinated children would not be feasible (although they appear not to be able to admit just how unethical it would be), and blithely suggest instead an epidemiological study of the vaccinated versus the unvaccinated without realizing just how horrendously difficult it would be to overcome the confounders that would plague such a study or that ethical considerations still require sound scientific justification for such a study. That’s why it’s so cute to see anti-vaccine loons trying to justify such a study.

All of which is why it’s pretty amusing that just such a study was recently reported in Germany. Can you guess what it found? Let’s just say that, to those of us who accept the science showing that vaccines do not cause autism, autoimmune diseases, asthma, and the like, the results were utterly unsurprising:

In their study, the authors compare the occurrence of infections and allergies in vaccinated and unvaccinated children and adolescents. These include bronchitis, eczema, colds, and gastrointestinal infections.

The evaluation showed that unvaccinated children and adolescents differ from their vaccinated peers merely in terms of the frequency of vaccine preventable diseases. These include pertussis, mumps, or measles. As expected, the risk of contracting these diseases is substantially lower in vaccinated children and adolescents.

Surprise, surprise! Unvaccinated kids differ from vaccinated kids only in having a much higher risk of vaccine-preventable disease! Who’da thunk it? But, as is my wont, I need to see the actual study, rather than a news report. So, as they say, let’s go to the tape. Unfortunately, the authors feel obligated to lay out exactly why they did their study:

The benefits, efficacy, and safety of protective vaccinations are widely scientifically proven. Furthermore, modern vaccines are well tolerated (4, 5). In spite of all this, some parents and doctors have reservations against vaccinations. The fear is that vaccinations overburden, stress, or weaken a child’s immune system and may therefore cause harm. As a result they think that vaccinated children are more prone to falling ill than non-vaccinated children. In addition, vaccinations are deemed to be responsible for the occurrence or increased incidence of other diseases, including chronic diseases (6, 7).

In other words, Generation Rescue, Dr. Jay Gordon, Jenny McCarthy, and others. They spread this misinformation.

Basically, what the German investigators did was to examine data collected aas part of the German Health Interview and Examination Survey for Children and Adolescents (KiGGS), which was conducted from May 2003 to May 2006 by the Robert Koch Institute. The data collected were primarily from a standardized, computer-assisted personal interview of the parents, but diagnoses reported and vaccination status were verified through medical records. Results were compared between vaccinated and unvaccinated children, defined thusly:

Children and adolescents were defined as unvaccinated if at the time of the KiGGS survey no documentation existed for any vaccination against diphtheria, pertussis, tetanus, Haemophilus influenzae type b, hepatitis B, poliomyelitis, measles, mumps, or rubella. By contrast, children who had by then received at least one vaccination according to their vaccination card were categorized as vaccinated. Immigrant families were excluded from the analyses in order to avoid misclassification because of the often missing or incomplete vaccination documents.

On the basis of the KIGGS data, investigators concluded that approximately 0.7% of nonimmigrant children and adolescents in Germany are completely unvaccinated. This is in line with estimates I have seen for the percentage of children in the U.S. who are completely unvaccinated, although this population is more difficult to ascertain in the U.S., at least on a national level. In any case, there was a vaccination card available for 13,453 children aged 1 through 17. Out of this group, for 94 children no evidence of vaccination could be found. Consistent with results in the U.S., the unvaccinated children tended to be from families of higher socioeconomic status. Less relevant, unvaccinated children tended to live in older German states than in the new German states.

Otherwise, the children in both groups were very similar, except in one way. This graph in Figure 1 tells the tale.


Basically, unvaccinated children had a much higher prevalence of vaccine-preventable infectious disease. The results were significant for pertussis, measles, and mumps, as described in the text:

The proportion of children and adolescents who had had pertussis, measles, mumps, and/or rubella was much higher in unvaccinated children than in those who had been vaccinated against the respective disease to a sufficient extent. For pertussis, the lifetime prevalence in unvaccinated subjects was 15.8% (95% CI 8.5 to 27.6, n = 11), in those with sufficient vaccination cover only 2.3% (95% CI 2.0 to 2.8), n = 184). For measles, the lifetime prevalence was 15.0 (95% CI 7.7 to 27.4, n = 10) in unvaccinated subjects and 5.2% (95% CI 4.7 to 5.8, n = 431) in sufficiently vaccinated ones. For mumps, the lifetime prevalence in unvaccinated subjects was 9.6% (95% CI 4.2 to 20.2, n = 7) and in vaccinated ones, 3.1% (95% CI 2.6 to 3.7, n = 305). For rubella, the lifetime prevalence was 17.0% (95% CI 9.4 to 29.0, n = 11) for unvaccinated subjects and 6.8% (95% CI 6.0 to 7.6, n = 642) for vaccinated ones. Differentiated stratified analyses confirmed the described higher proportions of subjects with the respective disease among unvaccinated subjects for boys and girls and for different age groups (data not shown).

It was at this point that the authors included an interesting control. One might imagine that if, for some reason, there were differences in infectious disease prevalence in general between the two groups, there might be differences in diseases against which we do not or cannot vaccinate, such as colds. What the investigators found was that there was no difference in the number of infections between vaccinated and unvaccinated children in any of the age groups studied:


In other words, as expected, the only difference in prevalence observed was in diseases for which vaccines make a difference, and those diseases were much less prevalent among vaccinated children than among unvaccinated children.

Finally, the investigators examined the hypothesis that vaccines increase the risk of allergies and immune system dysfunction by looking at atopic disorders. The authors examined the prevalence of three conditions, bronchial asthma, atopic eczema, and allergic rhinoconjunctivitis. Just like the case for diseases not currently vaccinated against, there was no difference in prevalence found between vaccinated and unvaccinated children, strongly suggesting that vaccines do not cause atopic disorders. This evidence is, of course, consistent with previous studies that have similarly failed to find evidence that vaccines increase the risk of asthma and other atopic disorders. Overall, this study is a near slam dunk against the idea that vaccines are in any way associated with atopic diseases, including this part:

In addition to atopic disorders, we further compared diseases–such as obstructive bronchitis, pneumonia and otitis media, heart disease, anemia, epilepsy, and attention deficit hyperactivity disorder (ADHD)–in unvaccinated and vaccinated subjects. No relevant differences in the lifetime prevalences were found, neither for different age groups nor between girls and boys. Schneeweiß et al. conducted a comprehensive literature review of vaccine safety, the central part of which was the evaluation of vaccine critical arguments on the basis of the current state of scientific knowledge. None of the hypotheses were found to be valid (5).

Let me repeat this again. None of the hypotheses were found to be valid.

It is true that, as large as it is, this study is relatively small given the low percentage of completely unvaccinated children. This made subgroup analysis impossible for some measures and limited the power of the analysis due to low numbers of unvaccinated children in several of the categories. After all, there were only 94 unvaccinated children, which makes it impossible to compare autism and ASD prevalence with the vaccinated cohort, because the expected prevalence of autism is only approximately 1 in 100 anyway. That means that on average roughly one autistic child would be expected in such a cohort. To look at differences in autism prevalence between the groups, many times more than 13,000 subjects would be needed, which would be a hugely expensive study, particularly if medical records and vaccination cards were reviewed for every child. This study also bolsters what I said before by showing how difficult it would be to do such a “vaxed versus unvaxed” study with sufficient subjects to provide adequate statistical power to detect even fairly large differences in prevalence. For diseases and conditions sufficiently prevalent to allow the detection of differences in the sample size the investigators had, though, there was no evidence of differences in conditions commonly blamed on vaccines between vaccinated and unvaccinated children. Certainly, this study provides no support for doing a larger study. For it to do that, there would have to be a hint of a whiff of an association between vaccinated status and the atopic diseases examined or the risk of infectious diseases not prevented by vaccines. There wasn’t. If I were on an IRB, and someone proposed doing a larger study, this German study would be an argument that there is inadequate scientific justification to do a larger study, particularly in light of other studies like this, this, and this, which suggest that early vaccination might actually be protective against atopic diseases.

The wandering band of anti-vaccine propagandists over at, for example, Age of Autism say again and again that they want a “vaxed versus unvaxed” study. Indeed, after the recent Supreme Court decision in Bruesewitz v. Wyeth upholding the primacy of the Vaccine Court in disbursing compensation to children injured by vaccines, AoA seems to have lost what’s left of its mind, posting multiple times about it as though it were some grave injustice and their rights had been completely stripped away. Add to that the picture of Barbara Loe Fisher comparing the anti-vaccine movement to Winston Churchill facing down Hitler, and the picture remains truly nauseating.

Be that as it may, I’d like to finish by offering our “friends” in the anti-vaccine movement a bit of advice. You keep saying you want the government to fund a “vaxed versus unvaxed” study. You keep repeating it until you’re blue in the face. Based on this German study, you might want to be a bit more careful. In other words, be careful what you wish for. You just might get it. And, based on this study and the vast preponderance of evidence, if that study you so fervently wish for is well-designed and conducted properly, you probably won’t like the results.

But why worry? If a “vaxed versus unvaxed” study is negative, the anti-vaccine movements will just ignore it. It’s what they do.


Schmitz, R; Poethko-Müller, C; Reiter, S; Schlaud, M (2011). Vaccination Status and Health in Children and Adolescents: Findings of the German Health Interview and Examination Survey for Children and Adolescents (KiGGS) Dtsch Arztebl Int, 108 (7), 99-104