Healthy people -- the "controls" used by researchers for comparison with the sick -- often make studies difficult by refusing to give blood to determine their genetic make-up. While sick people give blood almost as a matter of routine, the healthy controls may be scared of the needle, unwilling to be inconvenienced -- or concerned about the privacy of the results.
When this happens, research on the interactions between people’s genetic susceptibility and their exposures to chemicals, radiation or other environmental factors is frustrated.
But now, two research statisticians at the National Institute of Environmental Health Sciences, National Institutes of Health, say that taking blood from those healthy volunteers is often unnecessary.
NIEHS’ David M. Umbach, Ph.D., and Clarice R. Weinberg, Ph.D., report in the journal Statistics in Medicine that they have found that in many cases you can assume the random nature of the healthy volunteers’ genetic make-up and not bother testing for it. The volunteers can just be interviewed about exposures -- and face no needles.
Drs. Umbach and Weinberg report that when the disease or condition is rare, scientists can get as accurate results by comparing the genotype, or genetic make-up, of the patients with the problem to an assumed random occurrence of genetic susceptibility in the healthy controls, who need only be interviewed about exposures.
"This," Dr. Weinberg said, "can reduce the costs of some studies and make possible other studies that previously would have been thought too invasive to be practical -- not to mention reducing privacy problems and the ‘ouch’ factor."
These problems can be serious. While the blood from diseased patients is readily available for testing, in most cases, healthy volunteers often balk at the inconvenience of going to a clinic, or at the anticipated discomfort of the needle -- or fear that, if a genetic susceptibility is found, it might become public, hurting their employability and insurability. Their reluctance has made certain studies difficult or too costly to pursue.
If even 20 percent of the volunteer controls drop out because they fear giving blood or find it inconvenient, a study may become invalid.
Even when blood samples may still be required for a study, the Umbach-Weinberg statistical proofs show that, if the exposure and the genetic susceptibility factor occur independently in the population, valid conclusions can be reached without linking the results, even by a number or code, to the volunteer who gave the blood sample. This de-linking can reassure volunteers that a particular gene they have cannot become public -- and possibly make them undesirable to employers or insurers.
Umbach and Weinberg say their time- and money-saving short-cuts can only fail when individuals with the genetic susceptibility factor have a different chance of experiencing the environmental exposure than individuals without the genetic factor. This could happen when a gene codes for an enzyme that may make a person more likely or less likely to be exposed to a chemical or other factor. (For example, the gene that codes for an enzyme that results in flushing, headache and speeded heart rate from drinking alcohol makes people with that genetic variant avoid alcohol. Thus, a study of how this gene and alcohol consumption jointly affect a disease risk could not be done under this new method. )
The researchers illustrated their new methods by re-analyzing data from research in which all the data were easily available: A study of infants’ cleft palate and their mothers’ smoking during pregnancy, as well as a genetic susceptibility in the infant. This is a variant of the gene for transforming growth factor alpha. The study found that when mothers smoked and the infant was genetically susceptible, the risk of cleft palate increased synergistically.
In the actual study, blood samples from control infants without cleft palate were readily available at the hospital. But Umbach and Weinberg’s subsequent work showed that the synergism between smoking and the genetic susceptibility factor could have been demonstrated without the controls’ blood.
NIEHS is located in Research Triangle Park, N.C. Most other parts of NIH are in the Maryland suburbs of Washington, D.C.
Reporters may obtain a copy of the study by calling (919) 541-7860 or -3665.