AngryToxicologist is a scientist in the public health sector, knows plenty about toxicology, and is occasionally angry about it all. Drop me a line at That hed refers both to the fact that I promised to post on this last week and didn't as well as the fact that there have been way too few studies of Thimerosal given it's controversial status. you have probably all heard about the study. If fact, go over to Respectful Insolence and read all about it. People are still e-mailing me to comment on it so I will (only where I deviate from Orac or have something extra to mention).
1) It's interesting (and good) that they took into account mercury in the diet or from dental fillings to control for other exposures.
2) By my back of the envelope (I actually used an envelope!) calculation, the amount of methyl mercury in a can of albacore tuna is ~85 mcg. The highest cumulative exposure to the kids from birth to 7 months was 187.5 mcg and 37.5 mcg from birth to 1 month. Not that infants are eating mercury, but it puts things into perspective and begs the question: why doesn't the amount of mercury the mother gets make the amount of thimerosal meaningless? Answer: maybe it does. Caveats: Methyl doesn't equal ethyl mercury in exposure terms because ethyl is in the blood a shorter amount of time. Albacore tuna has more mercury in it than 'light' tuna.
3) The study found that tics were increased and there were very small decreases in speech articulation. Now, with a study with this many endpoints, you'll end up with some positive and negative associations by chance. However, these effects have been seen in a previous study of CDC's Vaccine Safety Datalink. Speech deficits also fit in with mercury toxicity.
What I think: I think that the conclusion was basically sound, however the authors talk about the replication of the speech/verbal and tic associations then conclude that
the overall pattern of results suggest that the significant associations may have been chance findings stemming form the large number of statistical tests that we performed.
I would have been a bit more careful with that wording since some of the positives were seen in a previous study. However, it could very well be due to chance (especially since the artic. wasn't seen in both sexes) and even if it was real, the difference is so slight over the range of thimerosal exposure, it's almost meaningless, clinically. So, If I had children who recieved thimerosal, I would find this study (and do find it) very reassuring.
Medicine & Health
Comments
Is this saying that ...
* The data analysis shows statistical significance
* When judgement is applied to the set of data and the analysis, it is realized that the significance may be false and only an artifact of the data set
... and that ...
* Your conclusion is that it is highly unlikely that any real-world effects could be detected?
Thanks ...
Posted by: Scott Belyea | October 4, 2007 10:31 AM
What do you think of yesterday's announcement encouraging mother's to eat fish and downplaying the impact of mercury?
www.foodkarmaalert.blogspot.com
Posted by: FoodSciYogi | October 5, 2007 7:55 AM
Scott,
Close what I mean to say is that the data analysis shows statistical significance for a few things but that statistical judgement makes it likely that it is an artifact. Scientific judgement makes it plausible that they are real (since they make biological sense and are repeatable). However, since they are so small, they may not be very meaningful, therefore coming to your conclusion that "it is highly unlikely that any real-world effects could be detected".
Posted by: AngryToxicologist | October 9, 2007 8:10 AM
thanks
Posted by: sohpet | March 31, 2009 3:16 PM
best sites
Posted by: kelebek | March 31, 2009 4:25 PM