Abstract
BACKGROUND: Observational studies have reported an association between thyroid function and colorectal cancer (CRC), with conflicting results. Elucidating the causal relationship between thyroid function and CRC facilitates the development of new preventive strategies to reduce CRC incidence. METHOD: We applied a two-sample Mendelian randomization (MR) method to evaluate the causal relationship between five thyroid-related indexes, including hyperthyroidism, hypothyroidism, thyroid stimulating hormone (TSH), free thyroxine (FT4) and basal metabolic rate (BMR), and CRC. Genome-wide association study statistics for thyroid-related phenotypes were obtained from the ThyroidOmics consortium, and summary statistics for genetic associations with CRC were obtained from the FinnGen consortium. We set a series of criteria to screen single nucleotide polymorphisms (SNPs) as instrumental variables and then performed bidirectional MR analysis, stratified analysis and extensive sensitivity analysis. Multiplicative random-effects inverse variance weighted was the primary analysis method, supplemented by weighted median and MR-Egger. RESULT: We identified 12 SNPs for hyperthyroidism, 10 SNPs for hypothyroidism, 41 SNPs for TSH, 18 SNPs for FT4, and 556 SNPs for BMR. Genetically predicted hyperthyroidism, hypothyroidism, TSH, and FT4 were not associated with CRC risk (all P > 0.05). Sensitivity analysis revealed no heterogeneity or pleiotropy. Genetically predicted BMR was significantly associated with increased CRC risk after removing outlier (OR = 1.30, P = 0.0029). Stratified analysis showed that BMR was significantly associated with colon cancer (OR = 1.33, P = 0.0074) but not rectal cancer. In the reverse analysis, there was no evidence of an effect of CRC on thyroid function (all P > 0.05). CONCLUSION: Our bidirectional MR analysis provides new insights into the relationship between thyroid function and CRC. CRC prevention may benefit from enhanced screening of high BMR populations.