Multiplicative, additive, and interactive associations of 25-hydroxyvitamin D with lung and prostate cancer

Abstract

Abstract. Results regarding the epidemiological association of vitamin D with lung (LCA) and prostate cancer (PCA) are controversial. This study tested whether serum 25-hydroxyvitamin D [25(OH)D] concentrations have interactive epidemiological associations with smoking, the number-one risk factor for LCA, and age, the number-one risk factor for PCA. Also, this study investigated whether the associations of 25(OH)D, smoking, age, alcohol consumption, body mass index, diet (the healthy Nordic diet score), and physical activity with incident LCA and PCA are multiplicative or additive. The study of association types makes it easier to select appropriate statistical methods. The Kuopio Ischaemic Heart Disease Risk Factor Study provided the data of 2578 men with 112 LCA and 300 PCA cases over 35 years by the end of 2019. Serum 25(OH)D did not associate with LCA and PCA or interact with smoking and age. The association of smoking with LCA was additive; 13 extra cases per 1000 men every 10 years. Age and alcohol consumption multiplicatively increased the hazard of LCA (hazard ratio, 95% confidence interval for age >50: 3.56, 1.82–6.17; drink per week: 1.01, 1.00–1.03), whereas adherence to healthy Nordic diet decreased it (per score point: 0.95, 0.89–1.00). The association of age >50 with PCA was additive; 2.5 extra cases per 1000 men every 10 years. To conclude, there was no epidemiological relationship of pre-diagnostic 25(OH)D concentrations with the incidence of LCA and PCA. The respective associations of smoking and age >50 with LCA and PCA were additive rather than multiplicative.