Serum calcium and incident and fatal prostate cancer in the Swedish AMORIS study

• 36 Background: Many observational studies have shown a positive association between intake of dairy products and prostate cancer (PCa) risk. From a biological point of view it is of interest to study this association as bone was recently shown to be a positive regulator of male fertility which suggests that regulation of bone remodeling and reproduction are linked. Since androgens promote cell proliferation and inhibit prostate cell death, it is possible that calcium (Ca) is linked to PCa risk via its link with the reproductive system. We studied the association between serum Ca and PCa while also accounting for levels of albumin, a protein to which Ca is bound. Methods: A cohort based on 192,183 men with baseline information on Ca (mmol/L) and albumin (g/L) was selected from the Swedish Apolipoprotein MOrtality RISk (AMORIS) study. Age-stratified multivariable Cox proportional hazard models were used to analyze associations between Ca and incident and fatal PCa risk. All models were adjusted for fasting status, glucose levels, socio-economic status, season at time of Ca measurement, Charlson comorbidity index, and history of fractures. Results: A 6,202 men were diagnosed with PCa and 672 died of PCa during mean follow-up of 12 years. A weak negative association was found between PCa risk and Ca (HR per SD: 0.97 (95%CI: 0.95-1.00)). A similar association was also found between albumin-corrected Ca and PCa risk (HR: 0.96 (0.89-1.03), 0.94 (0.87-1.01), and 0.92 (0.86-0.99) for the 2nd, 3rd, and 4th quartile compared to the 1st; P for trend: 0.02). No association was found with fatal PCa, nor was there effect-modification by overweight. A strong positive association between Ca and death was observed when censoring for PCa (HR per SD: 1.13 (95%CI: 1.12-1.15)). Conclusions: Serum levels of Ca were weakly negatively associated with PCa risk in our study when adjusted for age and history of comorbidities and fractures. A negative association between Ca and PCa risk is likely explained by the strong relation between Ca and non-PCa death. These competing risks need to be handled in order to define whether Ca is causally involved in PCa aetiology or whether it only acts a marker of other metabolic events in the causal pathway.

Publikations- och innehållstyp

vet (ämneskategori)

kon (ämneskategori)