Low blood levels of selenium, selenoprotein P and GPx3 are associated with accelerated biological aging: results from the Berlin Aging Study II (BASE-II)

Background: Biological age reflects inter-individual differences in biological function and capacity beyond chronological age. DNA methylation age (DNAmA) and its deviation from chronological age, DNAmA acceleration (DNAmAA), which was calculated as residuals of leukocyte cell count adjusted linear regression of DNAmA on chronological age, were used to estimate biological age in this study. Low levels of serum selenium, selenoprotein P (SELENOP), and the selenocysteine-containing glutathione peroxidase 3 (GPx3) are associated with adverse health outcomes and selenium supplementation is discussed as an anti-aging intervention. Methods: In this study, we cross-sectionally analyzed 1568 older participants from the observational Berlin Aging Study II (mean age ± SD: 68.8 ± 3.7 years, 51% women). Serum selenium was measured by total reflection X-ray fluorescence (TXRF) spectroscopy and SELENOP was determined by sandwich ELISA. GPx3 was assessed as part of a proteomics dataset using liquid chromatography–mass spectrometry (LC–MS). The relationship between selenium biomarkers and epigenetic clock measures was analyzed using linear regression analyses. P values and 95% confidence intervals (not adjusted for multiple testing) are stated for each analysis. Results: Participants with deficient serum selenium levels (< 90 μg/L) had a higher rate of biological aging (DunedinPACE, β = − 0.02, SE = 0.01, 95% CI − 0.033 to − 0.004, p = 0.010, n = 865). This association remained statistically significant after adjustment for age, sex, BMI, smoking, and first four genetic principal components (β = − 0.02, SE = 0.01, 95% CI − 0.034 to − 0.004, p = 0.012, n = 757). Compared to the highest quartile, participants in the lowest quartile of SELENOP levels showed an accelerated biological aging rate (DunedinPACE, β = − 0.03, SE = 0.01, 95% CI − 0.051 to − 0.008, p = 0.007, n = 740, fully adjusted model). Similarly, after adjustment for confounders, accelerated biological age was found in participants within the lowest GPx3 quartile compared to participants in the fourth quartile (DunedinPACE, β = − 0.04, SE = 0.01, 95% CI − 0.06 to − 0.02, p = 0.001, n = 674 and GrimAge, β = − 0.98, SE = 0.32, 95% CI − 1.6 to − 0.4, p = 0.002, n = 608). Only the association with GPx3 remained statistically significant after multiple testing correction. Conclusion: Our study suggests that low levels of selenium biomarkers are associated with accelerated biological aging measured through epigenetic clocks. This effect was not substantially changed after adjustment for known confounders.