Mixed effects linear model fitting assessment of predictive associations of variables with peripheral blood micronuclei/cytome abnormality frequencies¹^,².

1The base model, which was determined by the study design, was: MN frequency = Visit + Chemotherapy (3 types) + Radiation therapy + Visit by Chemotherapy* + Visit by Radiation therapy* with the study subject being a random effect 2Variables having predictive p values greater than 0.25 in the initial stepwise model included: Age (p = 0.9599); BMI (p = 0.2719); Income (p = 0.3728); Menopausal status (p = 0.4964); Alcohol consumption (p = 0.0816); Tumor grade (p = 0.7563); Tumor stage (p = 0.2981); Luminal A (p = 0.9807); HER2 positive (p = 0.6360); and Surgery (p = 0.3092) *The visit by chemo and visit by radiation interaction variables allowed these values to differ across time points. Mixed effects linear model fitting assessment of predictive associations of variables with peripheral blood micronuclei/cytome abnormality frequencies1,2.