Serum 25-hydroxy vitamin D and prostate cancer risk in a large nested case-control study

Abstract

Background: Vitamin D compounds inhibit prostate tumorigenesis experimentally, but epidemiologic data are inconsistent with respect to prostate cancer risk, with some studies suggesting nonsignificant positive associations.

Methods: The 25-hydroxy vitamin D [25(OH)D]-prostate cancer relation was examined in a nested case-control study within the Alpha-Tocopherol, Beta-Carotene Cancer Prevention Study of 50- to 69-year-old Finnish men. We matched 1,000 controls to 1,000 cases diagnosed during up to 20 years of follow-up on the basis of age (±1 year) and fasting blood collection date (±30 days). Conditional multivariate logistic regression models estimated ORs and 95% CIs. All statistical significance testing was 2-sided.

Results: Cases had nonsignificantly 3% higher serum 25(OH)D levels (P = 0.19). ORs (95% CIs) for increasing season-specific quintiles of 25(OH)D concentrations were 1.00 (reference), 1.29 (0.95-1.74), 1.34 (1.00-1.80), 1.26 (0.93-1.72), and 1.56 (1.15-2.12), with P(trend) = 0.01. Analyses based on prespecified clinical categories and season-adjusted values yielded similar results. These findings seemed stronger for aggressive disease [OR (95% CI) for fifth quintile of serum 25(OH)D [1.70 (1.05-2.76), P(trend) = 0.02], among men with greater physical activity [1.85 (1.26-2.72), P(trend) = 0.002], higher concentrations of serum total cholesterol [2.09 (1.36-3.21), P(trend) = 0.003] or α-tocopherol [2.00 (1.30-3.07), P(trend) = 0.01] and higher intakes of total calcium [1.82 (1.20-2.76), P(trend) = 0.01] or vitamin D [1.69 (1.04-2.75), P(trend) = 0.08], or among those who had received the trial α-tocopherol supplements [1.74 (1.15-2.64), P(trend) = 0.006].

Conclusion: Our findings indicate that men with higher vitamin D blood levels are at increased risk of developing prostate cancer.

Impact: Greater caution is warranted with respect to recommendations for high-dose vitamin D supplementation and higher population target blood levels.

Figure 1

Individual serum 25(OH)D concentration plotted by week of blood collection in 1,000 controls in the ATBC Study (1985–2005). Smoothed line represents predicted 25(OH)D values calculated using a locally-weighted polynomial regression.

Figure 2

Published nested case-control studies of serum 25(OH)D and risk of prostate cancer