Mercury, creatinine, outliers and autism

There are certain areas of the autism research landscape which have created real divisions among some professionals, scientists and people with autism and their families. One of them surrounds the relative contribution (or not) of heavy metals such as mercury to the onset and presentation of autism in some people. The theory being that environmental exposure to mercury (in whatever form) may somehow interfere with things like neuronal development and contribute to some of the symptoms seen in autism at least in some people. Quite a good overview of the whole debate is presented by Garrecht & Austin* (full-text). The obvious background support to this theory are the very real effects that have been documented as being associated with environmental mercury exposure, particularly in children. The other side of the coin is the epidemiological evidence suggestive of no overall connection between certain mercury-related exposure factors and events and the rising numbers of cases of autism.

A recent study by Barry Wright and colleagues* (full-text) contributes some more to this area of research with their findings of no overall difference in urinary mercury excretion, but (and it is a big but) a small number of outlier results perhaps warranting further research attention.

So, their study summarised:

• From an initial participant group (N=251) comprising 54 UK children with autism, 42 of their siblings and 155 children without autism (121 going to mainstream school and 34 children attending a special school), urinary mercury levels were attained in 230 children via ICP-MS. The analysing laboratory was blind to who was who.
• Based on analysis of mercury concentration either correcting or not correcting for creatinine (an important urinary ratio marker), no significant group differences were noted across the participants. Even allowing for age, gender and the number of amalgam fillings, there were no significant differences and including analysis of various other metals such as lead, cadmium and copper, no significant group differences were noted.
• That being said, there were a small number of 'outliers' whose results appeared to indicate elevated levels of mercury to be present. These results appear in the autism and special school group (assuming that this group included children with other learning or developmental issues).

Depending on your viewpoint, you could interpret these results a number of ways. You could say, no overall group differences so mercury is not going to be an issue in relation to autism and is certainly not driving the increase in cases. On the other hand you could say, OK no group effect but given that autism is a heterogeneous condition driven by the phrase 'if you've met one person with autism, you've met one person with autism' the results don't preclude a small subgroup whose presentation might be tied into a factor like mercury. It all depends on your point of view. Also as per the study discussion, levels of urinary mercury is one thing, levels of mercury in blood, plasma, brain, etc. is another.

I also note another interesting detail from the study. Let me run a sentence past you: "A lower creatinine in the ASD group (not statistically significant) raises the mercury to creatinine ratio in this subgroup, but not to statistically significant levels". I was interested to read this in view of my own research interest in creatinine in relation to autism (see here). The logic goes that a lower level of urinary creatinine might artificially inflate whatever other metabolite you are looking at as per this previous post. Granted in this study, they looked at both corrected and uncorrected concentrations and found no group difference on either occasion but still the question of why creatinine levels might have been lower is an interesting one.

* Garrecht M. & Austin DW. The plausibility of a role for mercury in the etiology of autism: a cellular perspective. Toxicological & Environmental Chemistry. 2011; 93: 1251-1273
DOI:  10.1080/02772248.2011.580588

** Wright B. et al. A Comparison of Urinary Mercury between Children with Autism Spectrum Disorders and Control Children. PLoS ONE. February 2012.
DOI: 10.1371/journal.pone.0029547