Genome-wide interaction study of physical activity and genetic susceptibility on colorectal cancer using UK biobank data

Colorectal cancer (CRC) risk is influenced by a complex interplay between genetic predisposition and lifestyle factors, such as physical activity (PA). We aimed to conduct a genome-wide interaction study (GWIS) to explore single nucleotide polymorphisms (SNPs), and genes modulated by PA on CRC risk using data from the UK Biobank. Among 272,270 eligible participants, 2,979 CRC cases were matched with 11,435 controls using a incidence density matching approach to avoid potential biases that may arise when using excessively large unmatched control groups, and to preserve comparability in the timing and distribution of exposure. PA was defined as whether individuals met the international criteria. We used conditional logistic regression models to assess the significance for the SNP x PA interaction on CRC, and we also performed gene-level analysis by aggregating the results of SNP-level analysis. Several SNPs showed nominal interaction signals with p < 5 × 10⁻⁶, including loci mapped to ABI3, ZBTB16, and GABRB3, though none reached significance after FDR correction. Interaction and main effects were often in opposite directions. At the gene and pathway levels, RNASEL, NSD1, and efferocytosis showed nominal signals, although none reached statistical significance after correction. Although we could not find associations that met the significance threshold after adjusting for multiple testing, these preliminary findings help us to understand the interplay between genes and lifestyle in CRC.