Monday, June 16, 2008

DDT thoughts

I promised on crooked timber that I would post this, so to keep my word I will even though it is silly. It is my ddtophilic reading of the DDT wiki.

My point was to have been that, once a chemical gets a bad rep, there is no way for it to be cleared, as it is impossible to prove a negative. The practice is, and should be, that any significant correlation is reported and taken seriously. Of course with many groups checking many correlations each, there will be lots of correlations which are significant at the 5% level. Then for most purposes, the tentative results are forgotten if they are not reproduced. This is fine for most purposes, but if the correlation is in a test for toxicity there is always new tentative evidence of new possible tocicities so a suspect chemical is never ever cleared.

In the case of DDT, there appears to be a fairly robust dangerous effect on reproductino and development because it is similar to estrogen. I don't have a real opinion on that one (time will tell). I wrote my comment in haste when I thought it was just the tentative correlation of the decade.

On toxicity

My reading of the WIKI is that it is about what one would expect for a benign chemical.
The list of suspected problems caused by DDT is long and alarming, but each seems to be supported by one or two studies. In the case of extensive follow up (breast cancer) meta analysis strongly suggested an effect very close to zero.

There are effects of huge doses reported in animals. More importantly there are reports of statistically significant epidemiological correlations. However, there isn't any particular alleged problem which is confirmed by meta-analysis of many studies.

If there were no true effect, I would expect a large number of initial reports of significant correlations due to publication bias and multiple tests statistics calculated in each study. My sense (if the wiki is complete) is that every decade there is a new discovered very bad effect (from one or two studies) then non confirmation.

To me this is the pattern that one would expect if an innocuous chemical were studied intensively.

Some points which make me suspect that DDT might be totally inoccuous (remember "might" makes right)

1.In an early study, humans voluntarily ingested 35 mg of DDT daily for about two years, and were then tracked for several years afterward. Although there was "suggestive evidence of adverse liver effects", no other adverse effects were observed.

^ CUETO C Jr, DURHAM WF, HAYES WJ Jr (1956). "The effect of known repeated oral doses of chlorophenothane (DDT) in man". J Am Med Assoc 162 (9): 890–7. PMID 13366680.

35 mg per day seems to me to be huge compared to exposure in actual use.

Breast cancer

Several studies have looked for associations between breast cancer and DDT exposure. Almost all studies have measured DDT or DDE blood levels at the time of breast cancer diagnosis or after. While individual studies have yielded conflicting results, taken as a whole, the studies of this design "do not support the hypothesis that exposure to DDT is an important risk factor for breast cancer."[57] These types of studies have been extensively reviewed:

* In 2007, the journal Cancer published a review of all of the epidemiological studies on breast cancer and DDT and DDE published between 2000 and 2006. The authors state that "Positive findings for well-controlled studies in the early 1990s of associations between breast cancer risk and the insecticide DDT, its breakdown product DDE, and PCBs prompted additional study. Snedeker reviewed studies of DDT/DDE and dieldrin, concluding that existing research strategies provided conflicting and mostly negative evidence…Updating the picture to 2006 provides…essentially unchanged conclusions for DDT/DDE…[I]n light of these findings, additional study of incident breast cancer in association with biological measures of DDE/DDT levels near the time of diagnosis is not a promising avenue."[58]

* similar conclusion based on meta-analysis deleted
* similar conclusion based on meta-analysis deleted
A new study in Environmental Health Perspectives found a strong association between exposure to the p,p-isomer of DDT early in life and breast cancer later in life.

Here there seems to have been a candidate consequence of DDT exposure which was extensively studied. The result is that the initial significant correlation seems to be one of the one in 20 expected due to chance (which would mean dozens in the literature if many studies are conducted and each separately tests for different possible toxicities).

Finally a new literature starts based on a new initial result.

There is also a literature on

1 comment:

Ed Darrell said...

It's interesting to me that the poisoning of the kid on the farm seems to have fallen out of the literature.

It's also interesting to me that the relatively frequent use of DDT as a suicide tool in Asia is almost wholly unreported in Europe and America.

It's easy to miss evidence of toxicity if one doesn't look at actual toxic effects.

Look here:

Mind you, LD-50 numbers must be calculated from actual deaths. It would be useful to have an account of the deaths that went into the calculation, but it is probably not wise to say they didn't occur.