Toggle Health Problems and D

Vitamin D deficiency predicted with 80 percent accuracy with just 11 parameters – June 2013

Can Vitamin D Deficiency Be Predicted by Simple Patient Characteristics?

Endocr Rev, Vol. 34 (03_MeetingAbstracts): OR31-2
Presented at Encrinology Conference in SF June 2013
Evelien Sohl1, Natasja M. van Schoor1, Renate T. de Jongh1 and Paul Lips, MDPhD1
1 VU University Medical Center, Amsterdam , Netherlands

Introduction: Vitamin D deficiency is common in older individuals and the current method to determine a deficiency is by measuring serum 25-hydroxyvitamin D (25(OH)D). This study aimed to develop a risk profile that can be used to easily identify older individuals at risk for vitamin D deficiency.

Methods: This study was performed within the Longitudinal Aging Study Amsterdam, which is an ongoing cohort study of a representative sample of the Dutch older population. During the second measurement cycle (1995/1996) serum 25(OH)D was determined (n = 1320) and many aspects of physical, cognitive, emotional and social functioning were assessed. Vitamin D deficiency was defined as serum 25(OH)D < 50 nmol/L. A backwards logistic regression procedure was used to select predictors using Akaike’s Information Criterion (p < 0.157). A total risk score was calculated by, firstly, dividing the individual regression coefficients by the lowest regression coefficient in order to create simple scores and, secondly, by adding up these simple scores.

Results: Vitamin D deficiency was present in 46.2 %.
The following predictors for vitamin D deficiency were identified:

  1. gender,
  2. age,
  3. bicycling,
  4. gardening,
  5. sporting,
  6. BMI ,
  7. smoking,
  8. alcohol use,
  9. season of blood collection,
  10. the presence of appetite, and
  11. having a partner.

The AUC is 0.76, and Hosmer-Lemeshow goodness-of-fit test was not significant (p=0.65), which indicates that the model fits the data well.
With a total risk score cut-off point of 40 (range 2-77), the model predicted vitamin D deficiency with a sensitivity of 70% and specificity of 69%.

With a cut-off point of 54, the sensitivity was 34% and specificity 94%.
Persons with a score ≤ 20 (9% of the participants) had 14% chance to be vitamin D deficient, whereas this chance was 81% for a score > 60 (10% of the participants).

Conclusion: A total risk score, including eleven predictors that can easily be assessed, was developed and able to predict serum 25(OH)D below 50 nmol/L accurately. This risk score may be useful in clinical practice to identify persons at risk for vitamin D deficiency.

See also VitaminDWiki

See any problem with this page? Report it (WORKS NOV 2021)