Lois Henry: "Trust me" science doesn't cut it
| Tuesday, Jul 21 2009 06:22 PM
Last Updated Tuesday, Jul 21 2009 06:22 PM
Oh brother.
Those were my first thoughts when I read yet another study designed to scare the breath out of us, literally.
This time around, it's air pollution making babies dumb in utero (that means before they're born).
The study, of 249 poor kids in New York City linked exposure to air pollution before birth with lower IQ scores later in life, which researchers said bolstered evidence that bad air may harm the developing brain.
Words like "link," "may," "suggest," "associated with" and "could" make me very nervous when used in frightening studies like this one that not only cause worry over the health of our children but could, might, may, probably will, be used to create even more restrictive regulations for benefits that are at best uncertain.
Basics from the study:
The moms wore air monitors for 48 hours during the last few months of pregnancy to gauge inner-city pollution exposure. Of the 249 kids studied, 140 were in the high exposure group.
At age 5, those kids scored lower than kids not in the high exposure group.
All the children lived in low-income neighborhoods and and fewer moms in the high exposure group had graduated from high school.
I haven't read the study, just the article about it, which we dutifully ran on the front page as many news organizations did because A) it's a study and B) it's scary.
But since I started delving into these kinds of studies more deeply, I've learned to be a more skeptical.
I contacted Stanley Young, assistant director for bioinformatics at the National Institute of Statistical Sciences in North Carolina, and asked for his thoughts.
Turns out he had already asked the authors of this study for their data set to see if he could replicate their results. They declined. (This has been an ongoing issue with the authors of other studies "linking" PM2.5 exposure with an increased death rate, but that's another rant.)
Without the data sets, Young and other scientists were left to read the study and ponder its findings with the rest of us.
Just a couple of points I pondered were that the kids came from low-income neighborhoods and at least some had uneducated parents, two factors long associated with academic underachievement.
Young, who has a Ph.D in statistics and genetics, had more scientific ponderings. He noted the researchers did vast numbers of statistical comparisons and it doesn't appear they adjusted their analysis to reflect the number of questions.
Asking loads and loads of questions means you increase your odds of getting a "hit" (something that is statistically significant) by chance alone.
"Statistically significant does not equal true," Young explained. "But you will have to trust them as they will not give up their data set."
He also noted that the lead author of the air-pollution-leads-to-dumb-kids study used at least some of the same children and others in previous research that linked (there's that word again!) prenatal air pollution exposure with genetic abnormalities that could increase risks for cancer; smaller newborn head size; reduced birth weight; developmental delays at age three; and children's asthma.
Whew!
Again, if you're talking about low-income communities, factors such as diet, alcohol consumption, smoking and a host of others could be responsible for all those findings.
And I can't help wondering, if you're looking at the same set of kids and see developmental delays early on and then low IQ scores later, is it the air? Or the kids?
As Young points out, if you don't let other scientists scratch around your data sets, we'll never really know for sure.
Despite my skepticism, I'm not saying we shouldn't do these kinds of experiments and studies, we should.
But the data should be open for inspection. In fact, considering the possible regulatory impacts on all our lives, if a dime of government money is used, it should be required by law .
Perhaps I have trust issues.
Opinions expressed in this column are those of Lois Henry, not The Bakersfield Californian. Her column appears Wednesdays and Sundays. Comment at people.bakersfield.com/home/Blog/noholdsbarred, call her at 395-7373 or e-mail lhenry@bakersfield.com
