Relationship between serum levels of insulin-like growth factors and subsequent risk of cancer mortality: Findings from a nested case-control study within the Japan Collaborative Cohort Study

Background: We investigated the association between serum levels of IGF-I, IGF-II, and IGFBP-3 and the subsequent risk of cancer mortality. Methods: Our case control study examined samples from 914 cancer deaths and their 2739 matched controls within the Japan Collaborative Cohort Study. Blood samples were obtained at the baseline and stored at 80 C until analysis for IGF-I, IGF-II, and IGFBP-3 levels. The conditional logistic model was used to estimate odds ratios (ORs) for cancer mortality associated with these serum levels. Results: The adjusted ORs for IGF-I quartiles ranged from 0.81 to 0.96 but were not significant. The adjusted ORs and 95% confidence interval (Cl) for the second, third, and fourth IGF-II quartiles were 0.64 (95% CI: 0.52-0.79), 0.71 (95% CI: 0.58-0.88), and 0.73 (95% CI: 0.59-0.91), respectively, while those for the respective IGFBP-3 quartiles were 0.77 (95% CI: 0.63-0.96), 0.75 (95% CI: 0.60-0.94), and 0.71 (95% CI: 0.56-0.90). In the model of IGF-I, and IGF-II additionally adjusted for IGFBP-3, the associations of high IGF's levels were similar as observed in the above models, while the association of IGFBP-3 shifted into non-significance after adjusting for IGF-II. Conclusion: An increased level of IGF-II was significantly associated with decreased risk of cancer mortality, whereas the association between IGF-I and all cancer mortality was not significant. The inverse association of IGFBP-3 level with all cancer mortality was affected when adjusting for IGF-II levels, shifting from significant to non-significant. Confirmation of these results from further cohort studies may aid in identifying the potential association between these molecules and the risk of cancer among the general Japanese population. (c) 2010 Elsevier Ltd. All rights reserved.