Dissertations, Theses, and Capstone Projects

Date of Degree


Document Type


Degree Name



Public Health


Lorna E. Thorpe

Subject Categories



Age-Period-Cohort; Mediation; Thyroid cancer


Background: Thyroid cancer (TC) is a common cancer diagnosis in the United States, whose incidence is increasing. Disaster and radiation treatment studies show high doses of radiation can cause TC. Some diagnostic procedures, whose use is increasing, expose individuals to low-dose radiation but can also incidentally detect subclinical TC. Evidence regarding low-dose radiation risk is limited.

Objectives: To investigate the possible association of greater use of diagnostic procedures with TC, either causally through radiation exposure or via incidental detection.

Methods: Two data sources were used: National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) data and Fire Department of the City of New York (FDNY) data. Using age-period-cohort analyses of SEER data (1973-2011), trends in TC incidence in the United States were decomposed into the effects of age, period and cohort in the context of changes in diagnostic procedures. Using FDNY data, the association of radiation exposure from diagnostic procedures with TC risk was assessed. Whether heightened medical attention (screening using diagnostic procedures and number of medical exams) mediated the previously-observed association between World Trade Center (WTC) exposure and TC was also investigated.

Results: Between 1973 and 2011, TC incidence has increased and mortality has remained stable. Independent age, period and cohort effects were observed for TC incidence in the United States. Risk increased around the 1993-1997 period, consistent with hypotheses of increased detection via diagnostic procedures, and decreased around the 1970s birth cohort. Analyses by cancer subtypes and tumor size showed mostly similar results.

Among firefighters, radiation exposure from diagnostic procedures and TC was not statistically significant (adjusted hazard ratio [HR]=1.02 per mSv), although after adjusting for confounders, the highest radiation exposure category >10 mSv, had a HR of 1.91, 95% confidence interval (CI) 0.55-5.18. TC risk was higher among firefighters with WTC exposure on 9/11/2001 but not after age-adjustment (HR=1.75, 95% CI 0.80-3.82). The association was partially mediated by heightened medical attention; 21.9% of the association was due to screening diagnostic procedures and medical exams.

Conclusions: Incidental detection of TC may be contributing to the increasing incidence and possibly exposure to low-dose radiation from diagnostic procedures as well.

Included in

Epidemiology Commons