To the Editor:
I read with interest the article by Tsuda et al.1 Nevertheless, I was very disappointed that it failed to identify itself as a classic ecologic study and acknowledge that its findings were, therefore, vulnerable to the ecologic fallacy. Likewise, the accompanying commentary, by Scott Davis, failed to point out the ecologic study design.2
The flawed inferential logic, known as ecologic fallacy, threatens all studies that draw risk inferences based on community incidence rates without individual dose data, yet that is but one of problems with ecologic studies.3 Despite the well-known limitations of this study design, a bibliometric review of ecological studies published in major epidemiologic journals found that only 69% of articles clearly specified their study’s design by mentioning the word “ecological” or “ecologic,” and 49% failed to acknowledge susceptibility to ecologic fallacy as a major weakness.4
The reluctance of authors to label their studies as ecologic is understandable given the negative reception such studies often receive. But the Tsuda article goes beyond failing to acknowledge that it is ecologic. It actually hides its design by using “the residential address of the subjects in March 2011…as a surrogate for individual [dose],” and then reports measures of association with odds ratios and relative rates—risk metrics typically employed in case–control and cohort studies, respectively. These two alternative study designs are much more reliable because they are based on individual dose data and, therefore, not prone to be influenced by factors that vary between communities.
The thyroid cancer rates in the Tsuda study are not consistent with the risks found in earlier cohort and case–control studies and, therefore, do not seem to be credible based on our prior experience with radioactive iodine. This should be bluntly stated because the media do not appreciate that all study designs are not equally valid, and the public needlessly panics over studies of limited scientific value.
I must, therefore, concur with Dr. Davis that “these findings do not add anything new regarding radiation-induced thyroid cancer.” But I would further add that publishing studies that use ecologic study designs without acknowledging the issue of ecologic fallacy is a disservice to the people of Fukushima, who have already suffered greatly and do not need the added burden of groundless worry about their risk of thyroid cancer—a risk level that most epidemiologists would consider very small, notwithstanding the Tsuda study.
Timothy J. Jorgensen
Health Physics and Radiation Protection Program
Department of Radiation Medicine
Georgetown University Medical Center
1. Tsuda T, Tokinobu A, Yamamoto E, Suzuki E.. Thyroid cancer detection by ultrasound among residents ages 18 years and younger in Fukushima, Japan: 2011 to 2014. Epidemiology. 2016;27:316–322
2. Davis S.. Screening for thyroid cancer after the Fukushima disaster: what do we learn from such an effort? Epidemiology. 2016;27:323–325
3. Morgenstern H.. Ecologic studies in epidemiology: concepts, principles, and methods. Annu Rev Public Health. 1995;16:61–81
4. Dufault B, Klar N.. The quality of modern cross-sectional ecologic studies: a bibliometric review. Am J Epidemiol. 2011;174:1101–1107