Urine Se concentration poorly predicts plasma Se concentration at sub-district scales in Zimbabwe, limiting its value as a biomarker of population Se status

Introduction The current study investigated the value of urine selenium (Se) concentration as a biomarker of population Se status in rural sub-Saharan Africa. Method Urine and plasma Se concentrations were measured among children aged 6-59 months (n = 608) and women of reproductive age (WRA, n = 781) living in rural Zimbabwe (Murehwa, Shamva, and Mutasa districts) and participating in a pilot national micronutrient survey. Selenium concentrations were measured by inductively coupled plasma-mass spectrometry (ICP-MS), and urine concentrations were corrected for hydration status. Results The median (Q1, Q3) urine Se concentrations were 8.4 mu g/L (5.3, 13.5) and 10.5 mu g/L (6.5, 15.2) in children and WRA, respectively. There was moderate evidence for a relationship between urine Se concentration and plasma Se concentration in children (p = 0.0236) and WRA (p = < 0.0001), but the relationship had poor predictive value. Using previously defined thresholds for optimal activity of iodothyronine deiodinase (IDI), there was an association between deficiency when indicated by plasma Se concentrations and urine Se concentrations among WRA, but not among children. Discussion Urine Se concentration poorly predicted plasma Se concentration at sub-district scales in Zimbabwe, limiting its value as a biomarker of population Se status in this context. Further research is warranted at wider spatial scales to determine the value of urine Se as a biomarker when there is greater heterogeneity in Se exposure.