A Tale of Two Studies
“It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness…”
In March of 2007, two studies on different aspects of autism have been published. These two studies, although they are both published in peer-reviewed journals, could not be more different in quality.
The first study is “Sulfhydryl-reactive metals in autism”, by JK Kern, et al. This study attempts to correct the shortcomings of the dreadful Holmes et al study, which I have discussed in depth some time ago.
My mother used to tell me, “If you can’t say something nice, don’t say anything at all” – so, to follow the letter of that maxim (if not the spirit), I will begin with the positive aspects of this study:
[1] They did do their statistics correctly, using non-parametric statistics when it was clear that their data did not follow a normal distribution. Bully for them!
[2] They explained their reasons for picking the statistical analytical method they did (the Kruskal-Wallis test)
[3] The values they obtained for hair mercury for both subjects and controls were within the range found in the larger NHANES study, unlike the Holmes et al study, which had control values that were completely off the scale.
[4] Their use of the case-control method reduced concerns that the controls might be different from the subjects in some way unrelated to autism.
Now, unfortunately, we come to the major shortcomings of this study - shortcomings that are as fatal to its conclusions as those of the Holmes et al study. In fact, the major shortcoming is the same in both studies:
Hair Doesn’t Excrete Mercury
I’d like you to keep that sentence in mind for the next few paragraphs.
You see, despite their perseveration on the idea that mercury is excreted into the hair – which would, in turn, make a low hair mercury level indicative of “poor excretors” – mercury uptake by hair is completely passive. There is no excretion mechanism that can be “impaired”. Mercury in the blood flowing through the hair follicles is passively bound to the sulfhydryl groups of the amino acid cysteine, which is especially abundant in hair.
It’s as simple as that. No excretion, just passive absorption and binding to the growing hair.
But don’t take my word for it.
[1] Akira Yasutake and Noriyuki Hachiya “Accumulation of Inorganic Mercury in Hair of Rats Exposed to Methylmercury or Mercuric Chloride”. Tohoku J. Exp. Med., Vol. 210, 301-306 (2006) .
“These findings suggest that the inorganic mercury is also taken up by rat hair from the blood circulation, as is the MeHg, irrespective of the consequences of the biotransformation of MeHg or exposure to inorganic mercury itself.”
[2] Shi CY, Lane AT, Clarkson TW. Uptake of mercury by the hair of methylmercury-treated newborn mice. Environ Res. 1990 Apr;51(2):170-81.
“Distribution of mercury in pelt and other tissues was measured. The level of mercury in pelt was found to correlate with hair growth. The amount of mercury in pelt peaked when hair growth was most rapid and the total amount of mercury in pelt was significantly higher than that in other tissues, constituting 40% of the whole body burden. [Note: the hair of mice contains a larger percentage of body mercury than it does in humans. This is because the hair (pelt) of mice is a larger fraction of the mouse body mass than the hair of humans, especially human children]
However, when the hair ceased growing, the amount of mercury in pelt dramatically dropped to 4% of whole body burden and mercury concentrations in other tissues except brain were elevated.
Autoradiographic studies with tritium-labeled methylmercury demonstrated that methylmercury concentrated in hair follicles in the skin. Within hair follicles and hairs, methylmercury accumulated in regions that are rich in high-sulfur proteins.”
[3] Mottet NK, Body RL, Wilkens V, Burbacher TM. Biologic variables in the hair uptake of methylmercury from blood in the macaquemonkey. Environ Res. 1987 Apr;42(2):509-23.
“The total mercury (Hg) in hair and blood of 45 young healthy adult female Macaque fascicularis given 0, 50, 70, or 90 micrograms MeHg/kg body wt orally in apple juice daily revealed a close and constant ratio between blood Hg and hair.
The amount of hair Hg does not increase with time (maximum period of observation 490 days) at a given dose level. Also the ratio was unchanged between background and subtoxic dose levels. Individuals at a given dose level with a higher-than-average blood level had a proportionately higher hair level.”
”This was the most unkindest cut of all…”
The last article, you will notice, was authored by none other than Thomas Burbacher, who is currently on the chelationistas “Most Favored Scientists” list.
Bottom line: hair mercury (and presumably other sulfhydryl-reactive metals) reflects the blood mercury (etc.) at the time that portion of the hair was forming. If the hair mercury (etc.) is low, then the blood mercury (etc.) was low.
And the only way mercury (etc.) can get into the brain is via the blood, unless you inject it directly into the brain.
So, low hair mercury (etc.) equals low blood mercury (etc) equals low brain mercury (etc.).
Any questions?
Now that we have rather thoroughly demolished the idea of mercury being excreted into the hair, what else does the Kern et al study have to say?
The only graph in the entire article (Figure 1) is rather puzzling – it lists the mean ranks of the arsenic, cadmium, lead and mercury levels of the autistic and control groups. For those not familiar with non-parametric statistics (and for many that are), this may seem odd. Frankly, it is odd to show the mean ranks instead of the data themselves. It suggests that the data, seen naked and unadorned, might not be as convincing as the authors would like.
Table 3 gives the mean (average) and standard deviation of the various hair metal levels of the two groups. This, too, is odd, since they went to great lengths to explain that their data did not follow a normal distribution and yet give mean and standard deviation – two measures that imply a normal distribution. Better would have been to give a median (the middle value of the ranking) and range. But they didn’t.
And when you look at the numbers they did give, you begin to understand why they didn’t show a graph of their data.
Zooming in on the mercury levels, the mean (SD) of the autistic subjects was 0.14 mcg/g (0.11) and that of the controls was 0.16 mcg/g (0.10). You’d be right if you thought that these levels were awfully close, although the Kruskal-Wallis test showed them to be statistically different.
Even if there are statistically significant differences between the two groups, the question remains if they were clinically significant.
I might also point out that not only are these values far lower than those measured by Holmes et al, they are also a bit lower than the NHANES levels (mean 0.22 mcg/g). And, curiously enough, they used the same lab as the Holmes et al study.
Curious.
It was especially curious to see them cite Holmes et al as supporting their hypothesis without once mentioning how very different their numbers were.
So, what can we make of their data, then? Clearly, their hypothesis that autistic children are “poor excretors” based on hair metals is unsupportable in the face of data showing (over and over) that mercury and other metals are passively taken up by hair.
Perhaps, the “take home message” from this study is that autistic children absorb less mercury. That wouldn’t sit well with Kern et al, but there is data to support that hypothesis.
[4] Ballatori N, Wang W, Lieberman MW. 1998. Accelerated methylmercury elimination in gamma-glutamyl transpeptidase-deficient mice. Am J Pathol 152:1049-1055.
“There were no differences in methylmercury excretion between the wild-type and heterozygous mice; however, the GGT-deficient mice excreted methylmercury more rapidly at both dose levels. Wild-type and heterozygous mice excreted from 11 to 24% of the dose in the first 48 hours, whereas the GGT-deficient mice excreted 55 to 66% of the dose, with most of the methylmercury being excreted in urine.”
Since gamma-glutamyl transpeptidase also has a profound effect on the redox state of the cell and gluthathione, (both of which have been linked to autism by another of the chelationistas Most Favored Scientists, Jill James) what Kern et al are seeing may be nothing more than the effects of altered gamma-glutamyl transferase activity, which may be related or unrelated to autism.
So, by doggedly perseverating on a hypothesis (autistic children are "poor excretors"), two sets of researchers have missed a possible implication of their findings: that autistic children may actually be "over-excretors" of mercury.
After all, that's what the data seems to be saying.
And it's a lot more likely than hair being a significant excretory organ for mercury.
End of Part One.
Next:
Strong Association of De Novo Copy Number Mutations with Autism.
Sebat J, Lakshmi B, Malhotra D, Troge J, Lese-Martin C, Walsh T, Yamrom B, Yamrom B, Yoon S, Krasnitz A, Kendall J, Leotta A, Pai D, Zhang R, Lee YH, Hicks J, Spence SJ, Lee AT, Puura K, Lehtimaki T, Ledbetter D, Gregersen PK, Bregman J, Sutcliffe JS, Jobanputra V, Chung W, Warburton D, King MC, Skuse D, Geschwind DH, Gilliam TC, Ye K, Wigler M.
Science. 2007 Mar 15; [Epub ahead of print]
Prometheus
Note: Prometheus will be attending a conference of minor mythological figures next week and will not be able to moderate comments. Rest assured, when he returns, all pending comments will be dealt with in a firm but fair manner.
In March of 2007, two studies on different aspects of autism have been published. These two studies, although they are both published in peer-reviewed journals, could not be more different in quality.
The first study is “Sulfhydryl-reactive metals in autism”, by JK Kern, et al. This study attempts to correct the shortcomings of the dreadful Holmes et al study, which I have discussed in depth some time ago.
My mother used to tell me, “If you can’t say something nice, don’t say anything at all” – so, to follow the letter of that maxim (if not the spirit), I will begin with the positive aspects of this study:
[1] They did do their statistics correctly, using non-parametric statistics when it was clear that their data did not follow a normal distribution. Bully for them!
[2] They explained their reasons for picking the statistical analytical method they did (the Kruskal-Wallis test)
[3] The values they obtained for hair mercury for both subjects and controls were within the range found in the larger NHANES study, unlike the Holmes et al study, which had control values that were completely off the scale.
[4] Their use of the case-control method reduced concerns that the controls might be different from the subjects in some way unrelated to autism.
Now, unfortunately, we come to the major shortcomings of this study - shortcomings that are as fatal to its conclusions as those of the Holmes et al study. In fact, the major shortcoming is the same in both studies:
Hair Doesn’t Excrete Mercury
I’d like you to keep that sentence in mind for the next few paragraphs.
You see, despite their perseveration on the idea that mercury is excreted into the hair – which would, in turn, make a low hair mercury level indicative of “poor excretors” – mercury uptake by hair is completely passive. There is no excretion mechanism that can be “impaired”. Mercury in the blood flowing through the hair follicles is passively bound to the sulfhydryl groups of the amino acid cysteine, which is especially abundant in hair.
It’s as simple as that. No excretion, just passive absorption and binding to the growing hair.
But don’t take my word for it.
[1] Akira Yasutake and Noriyuki Hachiya “Accumulation of Inorganic Mercury in Hair of Rats Exposed to Methylmercury or Mercuric Chloride”. Tohoku J. Exp. Med., Vol. 210, 301-306 (2006) .
“These findings suggest that the inorganic mercury is also taken up by rat hair from the blood circulation, as is the MeHg, irrespective of the consequences of the biotransformation of MeHg or exposure to inorganic mercury itself.”
[2] Shi CY, Lane AT, Clarkson TW. Uptake of mercury by the hair of methylmercury-treated newborn mice. Environ Res. 1990 Apr;51(2):170-81.
“Distribution of mercury in pelt and other tissues was measured. The level of mercury in pelt was found to correlate with hair growth. The amount of mercury in pelt peaked when hair growth was most rapid and the total amount of mercury in pelt was significantly higher than that in other tissues, constituting 40% of the whole body burden. [Note: the hair of mice contains a larger percentage of body mercury than it does in humans. This is because the hair (pelt) of mice is a larger fraction of the mouse body mass than the hair of humans, especially human children]
However, when the hair ceased growing, the amount of mercury in pelt dramatically dropped to 4% of whole body burden and mercury concentrations in other tissues except brain were elevated.
Autoradiographic studies with tritium-labeled methylmercury demonstrated that methylmercury concentrated in hair follicles in the skin. Within hair follicles and hairs, methylmercury accumulated in regions that are rich in high-sulfur proteins.”
[3] Mottet NK, Body RL, Wilkens V, Burbacher TM. Biologic variables in the hair uptake of methylmercury from blood in the macaquemonkey. Environ Res. 1987 Apr;42(2):509-23.
“The total mercury (Hg) in hair and blood of 45 young healthy adult female Macaque fascicularis given 0, 50, 70, or 90 micrograms MeHg/kg body wt orally in apple juice daily revealed a close and constant ratio between blood Hg and hair.
The amount of hair Hg does not increase with time (maximum period of observation 490 days) at a given dose level. Also the ratio was unchanged between background and subtoxic dose levels. Individuals at a given dose level with a higher-than-average blood level had a proportionately higher hair level.”
”This was the most unkindest cut of all…”
The last article, you will notice, was authored by none other than Thomas Burbacher, who is currently on the chelationistas “Most Favored Scientists” list.
Bottom line: hair mercury (and presumably other sulfhydryl-reactive metals) reflects the blood mercury (etc.) at the time that portion of the hair was forming. If the hair mercury (etc.) is low, then the blood mercury (etc.) was low.
And the only way mercury (etc.) can get into the brain is via the blood, unless you inject it directly into the brain.
So, low hair mercury (etc.) equals low blood mercury (etc) equals low brain mercury (etc.).
Any questions?
Now that we have rather thoroughly demolished the idea of mercury being excreted into the hair, what else does the Kern et al study have to say?
The only graph in the entire article (Figure 1) is rather puzzling – it lists the mean ranks of the arsenic, cadmium, lead and mercury levels of the autistic and control groups. For those not familiar with non-parametric statistics (and for many that are), this may seem odd. Frankly, it is odd to show the mean ranks instead of the data themselves. It suggests that the data, seen naked and unadorned, might not be as convincing as the authors would like.
Table 3 gives the mean (average) and standard deviation of the various hair metal levels of the two groups. This, too, is odd, since they went to great lengths to explain that their data did not follow a normal distribution and yet give mean and standard deviation – two measures that imply a normal distribution. Better would have been to give a median (the middle value of the ranking) and range. But they didn’t.
And when you look at the numbers they did give, you begin to understand why they didn’t show a graph of their data.
Zooming in on the mercury levels, the mean (SD) of the autistic subjects was 0.14 mcg/g (0.11) and that of the controls was 0.16 mcg/g (0.10). You’d be right if you thought that these levels were awfully close, although the Kruskal-Wallis test showed them to be statistically different.
Even if there are statistically significant differences between the two groups, the question remains if they were clinically significant.
I might also point out that not only are these values far lower than those measured by Holmes et al, they are also a bit lower than the NHANES levels (mean 0.22 mcg/g). And, curiously enough, they used the same lab as the Holmes et al study.
Curious.
It was especially curious to see them cite Holmes et al as supporting their hypothesis without once mentioning how very different their numbers were.
So, what can we make of their data, then? Clearly, their hypothesis that autistic children are “poor excretors” based on hair metals is unsupportable in the face of data showing (over and over) that mercury and other metals are passively taken up by hair.
Perhaps, the “take home message” from this study is that autistic children absorb less mercury. That wouldn’t sit well with Kern et al, but there is data to support that hypothesis.
[4] Ballatori N, Wang W, Lieberman MW. 1998. Accelerated methylmercury elimination in gamma-glutamyl transpeptidase-deficient mice. Am J Pathol 152:1049-1055.
“There were no differences in methylmercury excretion between the wild-type and heterozygous mice; however, the GGT-deficient mice excreted methylmercury more rapidly at both dose levels. Wild-type and heterozygous mice excreted from 11 to 24% of the dose in the first 48 hours, whereas the GGT-deficient mice excreted 55 to 66% of the dose, with most of the methylmercury being excreted in urine.”
Since gamma-glutamyl transpeptidase also has a profound effect on the redox state of the cell and gluthathione, (both of which have been linked to autism by another of the chelationistas Most Favored Scientists, Jill James) what Kern et al are seeing may be nothing more than the effects of altered gamma-glutamyl transferase activity, which may be related or unrelated to autism.
So, by doggedly perseverating on a hypothesis (autistic children are "poor excretors"), two sets of researchers have missed a possible implication of their findings: that autistic children may actually be "over-excretors" of mercury.
After all, that's what the data seems to be saying.
And it's a lot more likely than hair being a significant excretory organ for mercury.
End of Part One.
Next:
Strong Association of De Novo Copy Number Mutations with Autism.
Sebat J, Lakshmi B, Malhotra D, Troge J, Lese-Martin C, Walsh T, Yamrom B, Yamrom B, Yoon S, Krasnitz A, Kendall J, Leotta A, Pai D, Zhang R, Lee YH, Hicks J, Spence SJ, Lee AT, Puura K, Lehtimaki T, Ledbetter D, Gregersen PK, Bregman J, Sutcliffe JS, Jobanputra V, Chung W, Warburton D, King MC, Skuse D, Geschwind DH, Gilliam TC, Ye K, Wigler M.
Science. 2007 Mar 15; [Epub ahead of print]
Prometheus
Note: Prometheus will be attending a conference of minor mythological figures next week and will not be able to moderate comments. Rest assured, when he returns, all pending comments will be dealt with in a firm but fair manner.
posted by Prometheus at 6:56 PM
4 Comments:
Say "Hi" to sisyphus for us.
It's also possible that autistic children may be consuming significantly less mercury from tuna, etc., because their parents are familiar with the mercury poisoning hypothesis and thus are more careful than parents of non-autistic children to avoid buying potentially contaminated foods.
Bonnie,
Absolutely correct! Kern et al and Holmes have blinded themselves to a number of alternative (and much more plausible) explanations for their data.
It's painful to see the lengths to which the authors of both studies go to ignore the simpler conclusions that their data point to.
Prometheus
As far as I remember from Biology in school, excretions are done through urinating, sweating, and breathing (out). I don't suppose our head sweat stays imprinted in the hair forever, as long as we're washing it every now and then.
<< Home