Wednesday, October 29, 2008

I Reject Your False Null Hypothesis!....at least I think I maybe do.

Do two negatives make a positive? In math they do, but how about in the world of science? If you attack a hypothesis that used bad science with bad science does that make it null?

A long time ago I took a logic class in college and one of the arguments had something to do with fish, gills and whales. If I remember it went something like this:

Fish swim in the sea, whales swim in the sea, therefore all Whales are fish

In the world of toxicology we have to make a lot of conjecture simply because no one wants to volunteer to be the “case” side for the dose. There are a number of scientists that think endocrine disruptors - small amounts of natural and synthetic hormones or chemicals that act like hormones - are causing a number of health issues, primarily male reproductive problems and breast cancer. There is evidence to show in laboratory studies that this is plausible, however, in humans the link is not adequately supported according to some scientists.

Dr. Stephen Safe with Texas A&M University lectured our class the other day discussing his paper called “Endocrine disruptors and human health: is there a problem” (note: there is no question mark at the end so I am not sure if it is a statement or a question)

Good science needs it skeptics, and when it comes to endocrine disruptors there is no bigger naysayer than Dr. Safe. For the record, I tend to be skeptical about a lot of conclusions about cause and effect, but I try to have an open mind so that I don’t miss the truth, at least the truth that can be supported by sound science.

And therein lies the problem with Dr. Safe's lecture and paper on endocrine disruptors. He believes, and he has published support for his point of view, that their hypothesis is based on incorrect or inaccurate data. The foundation for their belief is that endocrine-active chemicals may be responsible for the increased incidence of breast cancer and disorders of the male reproductive tract. If this foundation is not true, as Dr. Safe suggests, then their hypothesis is not correct.

Dr. Safe makes a convincing argument and backs it up with his own meta-analysis. So what is a lowly grad student like me supposed to take away from this? Well for me, if I am not convinced or have concerns with the premise then I will have a heck of a time convincing someone else in the general public that knows less than I do.

Where he lost me, both in his lecture and in his paper, was on “Xenoestrogens and breast cancer” A xenoestrogen is defined as a by-products of industrial or chemical processing that have estrogen-like effects. There are three major naturally occurring estrogens in women, estradiol, estriol, and estrone. Each one does something different but they are all called estrogens. From his paper:
"The endocrine disruptor hypothesis regarding decreased male reproductive capacity suggest that inappropriate in utero exposure to estrogens plays a role in the testicular dysgenesis syndrome (TDS) Swerdlow et al. (1997) tested this hypothesis by investigating the rates of testicular cancer [in] dizygotic and monozygotic twins since studies have shown that free estrogen levels are [naturally] higher in the former group. Their results showed that there was a 50% increase in the risk for testicular cancer in dizygotic compared to monozygotic twins."

To counter this, Dr. Safe then writes, in the same paragraph:

"A recent twin study from Denmark directly tested the role of in utero exposure to estrogens on sperm counts/quality of singletons mono- and dizygotic twins [showed] no significant differences in sperm quality in any of the three groups of men and concluded that “higher prenatal concentrations of oestrogen are not related to reduced sperm counts in adulthood.” Offspring of women who were prescribed high pharmacologic doses of DES or estrogens during pregnancy are among the highest in utero exposed individuals to hormones. Studies in the United States and Finland showed that fertility in these high exposure groups was not different from a control population."

So here is where I scratch my head and say "say what?!?":
  • Does the higher concentration of oestrogen mean the same effect would be seen for any of the other estorgens, single or combined?
  • Is there an additive, synergistic, or potentiating effect that low levels of Xenoestrogens could play that are different than what the natural production of hormones would show, or the consumption of a single hormone, like DES would show. I am reading this as: Because oestrogen and DES are estrogen-like, any estrogen-like material would show similar results.
  • Swerdlow showed an “increase in the risk for testicular cancer” to support the hypothesis whereas Dr. Safe rebuts this with research that showed “no significant differences in sperm quality.” An apples and oranges problem here.
Like I said I am just a lowly grad student trying to figure out how all this stuff fits together. The fact that I pose these questions may prove that I am ignorant as all get out. But because I do, because they are there and I can see them, they can be asked by anyone.

If we has a profession are going to tell people to be afraid we better have something real to show them, and if we tell them not to worry, we had better be beyond reproach.

No comments: