12. Blood levels of mercury are related to diagnosis of autism: a reanalysis of an important data set (2007)

Goes number 12; Taylor number 12

 

Desoto MC, Hitlan RT. Blood levels of mercury are related to diagnosis of autism: a reanalysis of an important data set. J Child Neurol. 2007 Nov;22(11):1308-11.

http://www.ncbi.nlm.nih.gov/pubmed/18006963

The infamous Desoto & Hitlan paper.  This was subject to a great deal of critical commentary among scientifically and statistically expert autism parents.

Interverbal & Do'C at Autism Street:

The core statistical argument in DeSoto & Hitlan (2007) centers on whether a one-tailed test or two-tailed test is appropriate. These tailed tests are a means of determining whether statistical significance has been achieved. DeSoto & Hitlan (2007) note that Ip et al. used a two-tailed test. DeSoto & Hitlan (2007) criticize this choice. They claim that the hypothesis made by Ip et al. (2004) was directional. In the opinion of DeSoto & Hitlan, the authors should have used a one-tailed test....

Later in their commentary:

DeSoto & Hitlan (2007) are correct that a one-tailed test would have been a better match to the way Ip et al. described their hypothesis. However, the two tailed test offered by Ip et al. was more cautious. Further still, as DeSoto & Hitlan (2007) describe clearly in their review of literature, the data concerning autistics and mercury can go either way. While most studies have not shown a statistically significant difference in the hair or blood levels of autistic children compared to non-autistic controls, the statistically significant lower levels of hair mercury found in autistic children in one study (Holmes et al.) compared to the control group in the same study, have led some to introduce the poor-excretion hypothesis for autism. This unpredictability in direction necessitates a two-tailed test.

We also should consider a well-known informal rule that most new research students are likely exposed to in their statistics textbooks. One picks the type of tailed test, before one reviews the data. This is to prevent changing the type of test one uses to get a certain result. DeSoto & Hitlan by their own description had already reviewed the data and afterwards selected their one-tailed test. This violates the rule described above.

Orac Knows, in a comment on Mercury, Autism, and A Note on Scientific Honesty

The Hitlan paper exists only to try to spin a positive result out of a negative result. The “investigators” probably saw a p value of around 0.11 in the original study and realized that if they used a one-tail t-test they could cut it in half, to near 0.05.

EpiWonk:

We can conclude absolutely nothing about the association of ethylmercury in vaccines to autism from these data.

Further Commentary:

• November 17, 2007 Do'C & Interverbal, Autism Street,  A Tale of Two Tails
• November 29, 2007, Notorious LTP, Pure Pedantry, Mercury, Autism, and a Note on Scientific Honesty
• December 10, 2007 Prometheus, A Photon in the Darkness Winter Potporri
• January 23, 2008 Do'C & Interverbal, Autism Street, Desoto & Hitlan, Revisited Part I
• January 23, 2008 Do'C & Interverbal, Autism Street, Desoto & Hitlan, Revisited Part II
• January 27, 2008 Do'C, Autism Street, Desoto & Hitlan Revisited, Part III
• July 19, 2008 EpiWonk at EpiWonk An Introduction to Data Analysis

Desoto and Hitlan published an updated FAQ page at http://www.uni.edu/desoto/desoto_hitlan_autism.html

So.  Does this paper "demonstrate that vaccines can cause autism?"  No, no matter how much Desoto and Hitlan tried to spin it, this paper doesn't make the case.

Comments

[Barbara @therextras]

Seems like you should insert 'Not.' into the title of the post, eh?

[Liz Ditz]

Thanks for the comment, Barbara. Each of these (will be 30) posts are titled with the title of the study under discussion.