Prostate cancer risk and prediagnostic serum 25-hydroxyvitamin D levels (Finland) Merja H. Ahonen 1 , Leena Tenkanen 2 , Lyly Teppo 3 , Matti Hakama 4 & Pentti Tuohimaa 1, * 1 University of Tampere, Medical School, PO Box 607, 33101 Tampere, Finland. Fax: +358-3-215 6170, E-mail: Pentti.Tuohimaa@uta.®; 2 Helsinki Heart Study, Helsinki, Finland; 3 Finnish Cancer Registry, Helsinki, Finland; 4 Tampere School of Public Health, Tampere, Finland (*Author for correspondence) Received 18 October 1999; accepted in revised form 22 May 2000 Key words: andropause, epidemiology, prostate cancer, risk factor, vitamin D. Abstract Objectives: The aim was to evaluate the association between serum vitamin D (25-hydroxyvitamin D) level and risk of prostate cancer. Methods: The nested case±control study was based on a 13-year follow-up of about 19,000 middle-aged men who attended the ®rst screening visit within the Helsinki Heart Study and were free of clinically veri®ed prostate cancer at baseline. Through record linkage with the ®les of the Finnish Cancer Registry, 149 prostate cancer cases were identi®ed in the cohort. They were matched (1:4) to probability density sampled controls for age, time of sample retrieval, and residence. Serum levels of 25-hydroxyvitamin D (25-VD) at entry were measured for cases and controls. The relative risks of prostate cancer were derived using conditional logistic regression analysis. Results: Prostate cancer risk, analyzed by quartiles of the 25-VD levels, was inversely related to 25-VD. Men with 25-VD concentration below the median had an adjusted relative risk (OR) of 1.7 compared to men with 25-VD level above the median. The prostate cancer risk was highest among younger men (<52 years) at entry and low serum 25- VD (OR 3.1 nonadjusted and 3.5 adjusted). Among those younger men (<52 years), low 25-VD entailed a higher risk of non-localized cancers (OR 6.3). The mean age at diagnosis of the patients with 25-VD concentration above the median was 1.8 years higher than that of patients with vitamin D below the median (63.1 vs 61.3 years). Conclusions: We conclude that low levels of 25-VD associated with an increased risk for subsequent earlier exposure and more aggressive development of prostate cancer, especially before the andropause. Introduction It has been hypothesized that vitamin D de®ciency increases the risk of initiation and progression of prostate cancer [1]. Results from geographical studies suggest an inverse relationship between the level of solar radiation and prostate cancer mortality [2, 3]. Since the major source of vitamin D is through sunlight-induced photobiosynthesis in the skin, it was hypothesized that the protective eects observed may be mediated by vitamin D levels. However, analytical epidemiological studies on the role of vitamin D in the etiology of prostate cancer have been controversial [4±7]. Experimental and in-vitro studies suggest that vitamin D inhibits cell proliferation and induces cell dierenti- ation in several target organs including prostate, and the eect is mediated by vitamin D receptor (VDR) [8]. Both normal and cancerous prostate cells have VDR [9±11], indicating that they are target cells that can respond to vitamin D. The active form of vitamin D, 1,25- dihydroxyvitamin D (1,25-VD), inhibits proliferation and induces dierentiation in human prostate cancer cell lines [9, 12]. Furthermore, the invasiveness of human prostate cancer cells is inhibited by 1,25-VD [13]. In an experimental rat model, a vitamin D analogue reduced prostate cancer incidence [14]. Low serum concentration of 25-hydroxyvitamin D (25-VD) is considered to be the best indicator of vitamin D de®ciency, since the active form of vitamin D, 1,25- VD, is controlled tightly to maintain serum calcium levels and has a short half-life (4±6 h compared to 3 weeks for 25-VD) [15]. Because serum 1,25-VD re¯ects neither vitamin D de®ciency nor intraprostatic 1,25-VD, we measured serum 25-VD, which is a precursor for prostatic 1,25-VD synthesis [16]. The availability of stored serum samples collected from men with prostate Cancer Causes and Control 11: 847±852, 2000. 847 Ó 2000 Kluwer Academic Publishers. Printed in the Netherlands.