Genetic Risk Score for Serum 25-Hydroxyvitamin D Concentration Helps to Guide Personalized Vitamin D Supplementation in Healthy Finnish Adults

Riitta J. Sallinen, Olga Dethelsen, Sanni Ruotsalainen, Robert D. Mills, Timo Miettinen, Tuija E. Jääskeläinen, Annamari Lundqvist, Eero Kyllönen, Heikki Kröger, Jaro Karppinen, Christel Lamberg-Allardt, Heli Viljakainen, Mari A. Kaunisto, Olli Kallioniemi

Research output: Contribution to journalArticleScientificpeer-review


Background Genetic factors modify serum 25-hydroxyvitamin D [25(OH)D] concentration and can affect the optimal intake of vitamin D.

Objectives We aimed to personalize vitamin D supplementation by applying knowledge of genetic factors affecting serum 25(OH)D concentration.

Methods We performed a genome-wide association study of serum 25(OH)D concentration in the Finnish Health 2011 cohort (n = 3339) using linear regression and applied the results to develop a population-matched genetic risk score (GRS) for serum 25(OH)D. This GRS was used to tailor vitamin D supplementation for 96 participants of a longitudinal Digital Health Revolution (DHR) Study. The GRS, serum 25(OH)D concentrations, and personalized supplementation and dietary advice were electronically returned to participants. Serum 25(OH)D concentrations were assessed using immunoassays and vitamin D intake using FFQs. In data analyses, cross-sectional and repeated-measures statistical tests and models were applied as described in detail elsewhere. Results GC vitamin D-binding protein and cytochrome P450 family 2 subfamily R polypeptide 1 genes showed genome-wide significant associations with serum 25(OH)D concentration. One single nucleotide polymorphism from each locus (rs4588 and rs10741657) was used to develop the GRS. After returning data to the DHR Study participants, daily vitamin D supplement users increased from 32.6% to 60.2% (P = 6.5 x 10(-6)) and serum 25(OH)D concentration from 64.4 +/- 20.9 nmol/L to 68.5 +/- 19.2 nmol/L (P = 0.006) between August and November. Notably, the difference in serum 25(OH)D concentrations between participants with no risk alleles and those with 3 or 4 risk alleles decreased from 20.7 nmol/L to 8.0 nmol/L (P = 0.0063).

Conclusions We developed and applied a population-matched GRS to identify individuals genetically predisposed to low serum 25(OH)D concentration. We show how the electronic return of individual genetic risk, serum 25(OH)D concentrations, and factors affecting vitamin D status can be used to tailor vitamin D supplementation. This model could be applied to other populations and countries.

Original languageEnglish
JournalJournal of Nutrition
Issue number2
Pages (from-to)281–292
Number of pages12
Publication statusPublished - Feb 2021
MoE publication typeA1 Journal article-refereed

Fields of Science

  • 1184 Genetics, developmental biology, physiology
  • 3143 Nutrition
  • vitamin D
  • genetic risk score
  • personalized nutrition
  • DHR Study
  • Health 2011 survey
  • CYP2R1

Cite this