Polycythemia Vera and Age

I heard from Dr. Samuel Lesko, medical director of the Northeast Regional Cancer Institute in Scranton, Pa., regarding yesterday's post on polycythemia vera and waste-coal-fired power plants:

This is a potentially interesting observation. However, the incidence figures are for unadjusted rates (crude rates); age-adjusted rates would be more appropriate. The risk of p. vera increases dramatically with age and many of the high risk counties on the map are also home to many older residents. So, age cannot be ruled out as a possible explanation for the high rates in these counties. If a similar pattern was observed for a map showing age-adjusted incidence rates, it would make a more compelling case that something is going on. There may be other confounders as well, but age can't be ignored.

Indeed, epidemiologists commonly adjust for age in their disease incidence analyses. After all, counties with a higher proportion of older residents will typically show a higher incidence of diseases related to aging. And while polycythemia vera usually occurs within the age range of 20 to 80, its mean age of onset is 60, so it is generally a disease affecting older people. That certainly needs to be taken into consideration before any firm conclusions are to be drawn about the disease's distribution.

Unfortunately, the Pennsylvania Department of Health does not make it easy for an ordinary citizen like me to calculate age-adjusted incidence rates for polycythemia vera. When I requested numbers on the disease from PADOH, staff informed me that they would provide me with only the crudest data -- the count of cases by county. They declined even to break down the count by gender. (Polycythemia vera is slightly more common in men than in women.) Anything above and beyond that, they said, would require significant work on their part -- and significant cost on my part. Since I do not receive any funding for this Web site, and since I am not a person of independent means, I took what they would give me for free and did the best I could with it.

It's true that Pennsylvania has a higher proportion of elderly residents than the nation as a whole. And the four counties where I identified the very highest polycythemia incidence rates -- at least double the state’s rate -- are also among the state's oldest counties. To consider just one measure, 15.3 percent of Pennsylvania’s population is over 65 compared to 12.4 percent of the U.S. population, according to the latest figures from the U.S. Census Bureau. And the populations of the four counties with the most dramatically elevated polycythemia rates are even older than the state in general. The percent of the population over 65 is 19.1 percent in Cambria County, 18.9 percent in Schuylkill County, 18.7 in Luzerne County and 17.3 in Blair County.

Perhaps age could partly explain the dramatically elevated polycythemia vera incident rates in those counties and in the state as a whole. But could it completely account for a state rate that's somewhere between nine and 23 times the national rate? (The U.S. incidence rate is between 0.6 and 1.6 cases per million people, while in Pennsylvania for the years 2001 and 2002 there were 14 cases per million people, according to PADOH and the federal Agency for Toxic Substances and Disease Registry.) Could age completely account for county rates that are more than double the state's already elevated rate?

I'm skeptical. But perhaps those agencies with the means and ability to conduct more refined analyses than I -- PADOH? ATSDR? NRCI? -- would be willing to take a look at this.