Cancer Study Surveyed the Wrong People
A Missouri Department of Health’s report fell short along many dimensions
This article originally appeared in St. Louis Beacon on March 26, 2013.
"Then why aren't you looking over there?" asks the bystander.
The man looks up and replies, "Because the light is better over here."
The Department of Health looked where the light was shining, not where we think the keys were lost.
We are part of a group of Florissant natives and McCluer North High School graduates who two years ago started to ask questions about why an unusually large number of our childhood friends – with whom we were connected via Facebook – were being diagnosed with cancers in their 30s and early 40s. Our group contains auditors, lawyers, and social science professors, so we are sophisticated enough to understand that sometimes patterns happen randomly.
Lived near Coldwater Creek
For example, if you flip a coin six times, sometimes by random chance you get all six heads (1.6 percent of the time, actually). Only after we noticed the pattern in cancer rates did we learn about the piles of the nuclear waste dumped starting in the 1940s that were located near the source of Coldwater Creek, a body of water that runs through our old neighborhoods and often backed up and flooded our basements.
Prior to our inquiries into the impacts of this nuclear waste, the last major press coverage occurred when we were children – worried about football games and dates for homecoming dances, not the potential long-term impacts of ionizing radiation. Imagine our horror when we learned that back in 1989, testing of the banks of the creek in our neighborhoods revealed elevated levels of toxins. Could this explain the high cancer rates we were noticing?
Before long, the story went viral. Today our Facebook group “Coldwater Creek – Just the Facts Please” has more than 7,000 members, and more than 2,000 reports of illnesses like cancer, autoimmune diseases and thyroid problems. Members have posted photos of smiling Little League teams from the 1970s playing baseball on fields right by the waste piles (the fields have since been closed, and are being cleaned up). There have been chilling reminiscences of kids playing in the creek catching two-headed crawdads. And story after heartbreaking story of lives impacted at too young an age by cancers.
No accounting for mobility
The Missouri Department of Health recently released a study looking at rates of cancer from 1996 to 2004 among then-current residents of a subset of the ZIP codes of concern. They concluded that there is nothing to worry about. While there are elevated rates of some types cancers, including breast, colon, kidney and prostate, the report claims are more likely the fault of the residents’ lifestyles.
The report is completely uninformative, because like the man looking for his keys under the street light, they were looking for cancers among current residents. And the current residents are generally not those who were potentially exposed to the radiation prior to the cleanup efforts.
According to the 2007-2011 American Community Survey, more than 75 percent of residents of our old ZIP code – 63033, one of the ZIPs considered in the study – have moved to their current residence since 1990. Fifty six percent moved since 2000. The waste piles were placed on the Environmental Protection Agency’s National Priorities List in 1989, and cleanup has been ongoing.
The rate of neighborhood turnover in the area today stands in sharp contrast to our experience growing up there in the 1970s and 1980s. The vast majority of our kindergarten classes still lived in the area through high school. But almost none of us still lives in the ZIP codes that were studied. Many live in other parts of the St. Louis area. Some of us are unlucky enough to currently live out of state.
From a public health perspective, the most important question is whether those of us with prolonged childhood exposure to the waste are suffering ill effects. Cancer rates among current residents are also important, but this misses the point.
As if studying the wrong population was not enough, the study fell far short along other dimensions. The data analysis was limited to the years 1996-2004 (apparently because of the department’s lack of access to updated Census data that appear to be available here). Approximately half of the illnesses reported on the Facebook page have been diagnosed after 2004. The study also excluded several important ZIP codes in its analysis.
More important, the report claims that the elevated cancers are likely due to residents’ lifestyles. But three of the four elevated cancer types – breast, colon, and kidney – are known to be diseases associated with ionizing radiation exposure.
We believe the Missouri Department of Health takes its mission to promote and protect the public’s health seriously. But its study sheds no light where the real problem might be. The nature of the issues at play here – with high geographic mobility and a substantial latency period between potential exposure and diagnosis – highlights the limitations of standard epidemiological techniques.
In the 21st century, social media is an important way for us to connect. It is through social media that we uncovered this troubling pattern in the first place. The department of health should go back to the drawing board and tap some of Missouri’s best minds to devise a study to really answer the question. It will not be easy and will require a better data collection effort than has been undertaken in the past. We would look forward to working together to uncover the truth.
- Diane Whitmore Schanzenbach is an associate professor in the School of Education and Social Policy.