Serum levels of selenium and smoking habits at age 50 influence long term prostate cancer risk; a 34 year ULSAM follow-up

Abstract

Background: Serum selenium level (s-Se) has been associated with prostate cancer (PrCa) risk. We investigated the relation between s-Se, smoking and non-screening detected PrCa and explored if polymorphisms in two DNA repair genes: OGG1 and MnSOD, influenced any effect of s-Se.

Methods: ULSAM, a population based Swedish male cohort (n = 2322) investigated at age 50 for s-Se and s-Se influencing factors: serum cholesterol, erythrocyte sedimentation rate and smoking habits. At age 71 a subcohort, (n = 1005) was genotyped for OGG1 and MnSOD polymorphisms.

Results: In a 34-year-follow-up, national registries identified 208 PrCa cases further confirmed in medical records. Participants with s-Se in the upper tertile had a non-significantly lower risk of PrCa. Smokers with s-Se in the two lower tertiles (≤80 μg/L) experienced a higher cumulative incidence of PrCa than smokers in the high selenium tertile (Hazard Ratio 2.39; 95% CI: 1.09-5.25). A high tertile selenium level in combination with non-wt rs125701 of the OGG1 gene in combination with smoking status or rs4880 related variation of MnSOD gene appeared to protect from PrCa.

Conclusions: S-Se levels and smoking habits influence long-term risk of PrCa. Smoking as a risk factor for PrCa in men with low s-Se is relevant to explore further. Exploratory analyses of variations in OGG1 and MnSOD genes indicate that hypotheses about patterns of exposure to selenium and smoking combined with data on genetic variation in genes involved in DNA repair can be valuable to pursue.

Figure 1

Participant flow chart. Flow chart of the investigated population of the study within the ULSAM cohort.

Figure 2

a, b: Cumulative incidences of outcomes. Cumulative incidences of death without prostate cancer for smokers (black line) and non smokers (grey line) by tertiles 1 and 2 combined (panel a) and tertile 3 (panel b) of serum selenium at baseline, 50 years of age in the genotyped subcohort. Figure 2 c, d: Cumulative incidences of prostate cancer for smokers (black line) and non-smokers (grey line) by tertiles 1 and 2 combined (panel c) and tertile 3 (panel d) of serum selenium at baseline, 50 years of age in the genotyped subcohort. Tertile serum selenium definitions; tertile 1 and 2 combined: ≤81 μg/L, and tertile 3: > 81 μg/L.

Figure 3

a, b: Hazard ratios for prostate cancer by OGG1 genotype. Hazard ratios (HR), with 95% confidence interval, for later diagnosis of prostate cancer in tertile 3 versus tertile 1(black line) and tertile 2 vs. tertile 1(gray line) of serum selenium levels at baseline, 50 years of age, by OGG1 genotypes. To the left (a) from top are results for SNPs rs2304277 and rs125701 and at the bottom for all genotyped men and the full cohort. To the right (b) from top are results for SNPs rs293796 and rs2472037. Tertile serum selenium definitions: tertile 1: ≤ 70 μg/L, tertile 2: 70.1- 81 μg/L and tertile 3: > 81 μg/L. "n" denotes number of participants in analysis.

Figure 4

a, b: Hazard ratios for prostate cancer by MnSOD genotype. Hazard ratios (HR), with 95% confidence interval, for later diagnosis of prostate cancer in tertile 3 versus tertile 1(black line) and tertile 2 vs. tertile 1(gray line) of serum selenium levels at baseline, 50 years of age, by MnSOD genotypes. To the left (a) from top are shown results for SNPs rs2758331, rs2842980, rs7855. To the right (b) from top are shown results for SNPs, rs5746141, rs5746136 and rs4342445. Tertile serum selenium definitions: tertile 1: ≤ 70 μg/L, tertile 2: 70.1- 81 μg/L and tertile 3: > 81 μg/L. "n" denotes number of participants in analysis.