Search

Background The thyroid cancer (TC) incidence in China has increased dramatically during the last three decades. Typical in this respect is the case of Hangzhou city (China), where 7147 new TC cases were diagnosed during the period 2008–2012. Hence, the assessment of the TC incidence risk increase due to environmental exposure is an important public health matter. Methods Correlation analysis, Analysis of Variance (ANOVA) and Poisson regression were first used to evaluate the statistical association between TC and key risk factors (industrial density and socioeconomic status). Then, the Bayesian maximum entropy (BME) theory and the integrative disease predictability (IDP) criterion were combined to quantitatively assess both the overall and the spatially distributed strength of the “exposure-disease” association. Results Overall, higher socioeconomic status was positively correlated with higher TC risk (Pearson correlation coefficient = 0.687, P < 0.01). Compared to people of low socioeconomic status, people of median and high socioeconomic status showed higher TC risk: the Relative Risk (RR) and associated 95% confidence interval (CI) were found to be, respectively, RR = 2.29 with 95% CI = 1.99 to 2.63, and RR = 3.67 with 95% CI = 3.22 to 4.19. The “industrial density-TC incidence” correlation, however, was non-significant. Spatially, the “socioeconomic status-TC” association measured by the corresponding IDP coefficient was significant throughout the study area: the mean IDP value was − 0.12 and the spatial IDP values were consistently negative at the township level. It was found that stronger associations were distributed among residents mainly on a stripe of land from northeast to southwest (consisting mainly of sub-district areas). The “industrial density-TC” association measured by its IDP coefficient was spatially non-consistent. Conclusions Socioeconomic status is an important indicator of TC risk factor in Hangzhou (China) whose effect varies across space. Hence, socioeconomic status shows the highest TC risk effect in sub-district areas. [ABSTRACT FROM AUTHOR]