Houston Can Breath Easy over UT Study

Related articles

This article first appeared in the Houston Chronicle.

Our nation's air is getting cleaner -- including Houston's. There's no need to get rid of a system that's not broken, certainly not based on the results of a recent study from the University of Texas School of Public Health. Although the press reports have scared many, in fact this report falls far short of showing a causal relation between 1,3 butadiene (BD) and human illness.

The media outcry over the preliminary report of an association between air pollutants and leukemia is unjustified. Even the researchers themselves repeatedly note the uncertainties in the data and the approximations they had to make to generate even their soft findings.

Government and industry have long worked together to reduce air pollution, with notable improvements nationwide. Of course, some areas remain cleaner than others. But, based on this new preliminary report, there is no reason to campaign for excessively stringent standards for a particular chemical, BD, which has been detected at elevated levels in the area.

Researchers at the UT Health Science Center used 1995-2003 medical reports to determine the rate of various cancers, particularly leukemia. They estimated BD exposure for various regions near Houston by assuming that air concentration correlated with the distance from the Houston Ship Channel and the distance from state air pollution monitors.

Since they had no individual measurements of anyone's specific chemical burden with which to work, they had to use this vague approximation.

Air measurements over the period 1992-2003 were used as a substitute for exposure to BD.

Even using these crude approximations of dose and effect, their report indicated only an estimated 56% elevated risk for leukemia in children living within two miles of the ship channel, as compared with those living more than 10 miles away.

While 56% at first sounds like a big number, remember how small the normal risk of leukemia is and how few cases it takes to randomly skew its apparent frequency.

By contrast, for example, smoking increases risk of lung cancer not by 50% or 60% but by something like 20 times; that is, 2,000%. That is an example of a causal relationship that is real.

The "gold standard" for attempting to find cause and effect in scientific studies, though, is a randomized, controlled trial, or RCT.

In that type of study, two highly similar groups are compared after some intervention is given to one group and not the other. Both groups are evaluated for a specific outcome over a period of time. In large enough RCTs, even a small detected difference can be a real one: "significant," in scientific terminology.

The UT report is a good example of a bad study -- almost the opposite of a controlled experiment. The authors used outdated information on cancer rates, and they made only crude estimates of chemical exposure.

Why did they choose BD as the toxic chemical to accuse of causing increased risk?

With estimates this crude, they could have picked as a culprit almost anything in the air -- or water or food -- since there were so many other variables in this huge swath of the population.

Further, their information concerning amounts of BD in the air was admittedly unreliable, and they failed to control for exposure to other possible carcinogens, such as gasoline fumes or even cigarette smoke, which contains BD.

In fact, while articles written about this study refer, invariably, to BD as a "carcinogen," the EPA refers to it only as "reasonably likely" to be a human carcinogen. The studies supporting even that weak designation arose from long-term occupational exposure to the chemical, as in petroleum facilities -- much higher than the minuscule doses in Houston's air, even close to the ship channel (and those minuscule doses have fallen by more than 50% in the past three years, as technology has improved).

Our goal should be the advancement of public health -- but based on solid evidence, not agenda-driven conjecture and illusory endpoints.

The "study" purporting to find a cancer link to BD is more junk than science, and will never see the light of day in a peer-reviewed journal.

In toxicology, it is an axiom that "the dose makes the poison."

Even if BD were a human carcinogen at high doses, it does not mean that it will cause cancer at the parts-per-billion level in Houston's air.