Association of Polygenetic Risk Scores Related to Cell Differentiation and Inflammation with Thyroid Cancer Risk and Genetic Interaction with Dietary Intake

Abstract

The incidence of thyroid cancer continues to increase steadily, and this increasing incidence cannot be attributed solely to the overdiagnosis of microcarcinoma or technical advancements in detection methods and may also depend on environmental and genetic factors. However, the impacts and interactions of genetic and environmental factors remain controversial, and they may differ in Eastern and Western countries. The study's purpose was to identify single nucleotide polymorphisms of genes related to cell differentiation and inflammation to influence thyroid cancer incidence and determine interactions with lifestyles in a large city hospital-based cohort. Genetic variants were selected by genome-wide association study with thyroid cancer participants (case; n = 495) and controls without cancers (n = 56,439). SNPs having gene-gene interactions were selected by generalized multifactor dimensionality reduction. Polygenic risk scores (PRSs) were generated by summing the number of selected SNP risk alleles. PRSs of the best model included 6 SNPs, that is, DIRC3_rs6759952, GAP43_rs13059137, NRG1_rs7834206, PROM1_rs72616195, LRP1B_rs1369535, and LOC100507065_rs11175834. Participants with a high-PRS had a higher thyroid cancer risk by 3.9-fold than those with a low-PRS. The following variables were related to an increased thyroid cancer risk; female (OR = 4.21), high white blood cell count (OR = 4.03), and high energy (OR = 7.00), low alcohol (OR = 4.11), and high seaweed (OR = 4.02) intakes. These variables also interacted with PRS to influence thyroid cancer risk. Meat/noodle diet patterns interacted with PRSs to increase thyroid cancer risk (p = 0.0023). In conclusion, women with a high-PRS associated with cell differentiation and inflammation were at an elevated thyroid cancer risk. Daily energy, seaweeds, and alcohol intake interacted with PRS for thyroid cancer risk. These results could be applied to personalized nutrition plans to reduce the risk of thyroid cancer.

Keywords: dietary patterns; gender; gene–gene interaction; polygenic risk scores; thyroid cancer; white blood cell.

Figure 1

The flow chart for the generation of polygenetic risk scores that influence thyroid cancer risk.

Figure 2

Adjusted odds ratio (OR) and 95% confidence intervals (CIs) of polygenetic risk score (PRS) with 5 SNPs and 6 SNPs generated by SNP–SNP interaction for the thyroid cancer risk. PRS with five SNPs and six SNPs, the best model of GMDR, was calculated by the summation of the number of risk alleles of five and six SNPs, and the calculated PRS were divided into three categories (0–3, 4–5, and ≥6) and (0–3, 4–6, and ≥7) by tertiles, respectively, as the low-PRS, medium-PRS, and high-PRS groups. Adjusted OR was analyzed by logistic regression with the covariates including age, gender, residence area, survey year, body mass index, education, income, menopause, initial menstruation, smoking, alcohol, energy, physical activity, fat percent intake, and carbohydrate percent intake. The reference group was the low-PRS in logistic regression. Red and blue boxes indicated the adjusted ORs for five SNPs and six SNPs, respectively, and the lines through red and blue boxes indicated 95% CI.

Figure 3

Prevalence of thyroid cancer among subjects in the low-, medium-, and high-PRS groups (determined using the 6 SNP genetic variant–genetic variant interaction model). PRS with 6 SNPs, the best model of GMDR, was divided into 3 categories (0–3, 4–6, and ≥7) by tertiles as the low-PRS, medium-PRS, and high-PRS groups. The nutrient and diet variables were categorized into two groups based on the specified cutoff values. The frequencies of thyroid cancer of PRS groups were calculated in low and high intake groups. A. In subjects categorized by gender. B. In subjects categorized by white blood cell count (WBC, cutoff value: 4 × 10 ⁹/L). C. In subjects categorized by daily energy intake (cutoff value: 100 percent of estimated energy intake). D. In subjects categorized by alcohol intake (cutoff value: 20 g/day). E. In subjects categorized by seaweed intake (cutoff value: 2.65 g/day). F. In subjects with a meats/noodle dietary pattern (cutoff value: 70th percentile). * Significantly different among the PRS groups p < 0.05, ** p < 0.01, *** at p < 0.001.