The Relationship Between Total Serum 25-hydroxyvitamin D and Dietary and Supplemental Intake of Vitamin D in the Adult U.S. Population During 2003–2014

Abstract Objectives We sought to determine whether 25OHD increased during NHANES 2003–2014 and whether the effects of preselected covariates on 25OHD were similar within/across three 4-yr periods. Methods 25OHD was measured in adults (≥20 y) using RIA (2003–2006) and LC-MS/MS (2007–2010, 2011–2014). Descriptive and regression analyses were stratified by survey period to investigate the effect of demographic (age, race-Hispanic origin, season, sex) and lifestyle (diet and supplements containing vitamin D) factors. Using only LC-MS/MS data, a single model was used to assess changes in 25OHD between 2007–2010 and 2011–2014. Results Previous National Health and Nutrition Examination Survey (NHANES) data suggested that high-dose vitamin D supplements may explain increased serum concentrations of 25-hydroxyvitamin D (25OHD) from 1988–2006 to 2007–2010. Without adjustment, mean 25OHD steadily increased from 2003–2006 to 2011–2014 (P < 0.0001). However, the stratified regression analyses revealed the effect of age, sex, and supplement use with 25OHD was different in 2003–2006 (RIA) compared to 2007–2010 and 2011–2014 (LC-MS/MS). Therefore, to assess the presence of an increase in recent 25OHD, we limited the analysis to LC-MS/MS data (2007–2014). After covariate adjustments, the significant increase in unadjusted 25OHD from 2007–2010 to 2011–2014 (P = 0.048) was no longer statistically significant (P = 0.11). After adjustment, non-Hispanic blacks were on average 23 nmol/L lower than non-Hispanic whites; and supplement use of ≥1000 IU/d raised 25OHD by 32 nmol/L compared to none. Conclusions It was clear that we could not assess long-term trends in 25OHD because the 2003–2006 data were predicted, not measured. This led to limiting the final model to LC-MS/MS data. After adjusting for higher doses of supplements, along with other covariates in the final model, there was no statistically significant increase in 25OHD (2007–2014) confirming the role of supplement use as a driving factor in the unadjusted concentrations. Funding Sources None.