Development and validation of a predictive model of hypovitaminosis d in general adult population: SCOPYD study

Abstract

The worldwide global increase in serum 25‐hydroxyvitamin D (25(OH)D) measurements has led some countries to restrict reimbursement for certain clinical situations only. Another approach could consist in providing physicians with screening tools in order to better target blood test prescription. The objective of the SCOPYD study was to identify the best combination of predictors of serum VitD concentration among adults aged 18–70 years. Potential risk factors for VitD deficiency were collected using a comprehensive self‐administered questionnaire. A multivariable linear regression was used to build a predictive model of serum 25(OH)D concentration. Among 2488 participants, 1080 (43.4%) had VitD deficiency (<50 nmol/L) and 195 (7.8%) had severe deficiency (<25 nmol/L). The final model included sunlight exposure in the preceding week and during the last holidays, month of blood sampling, age, sex, body mass index, skin phototype, employment, smoking, sport practice, latitude, and VitD supplementation in preceding year. The area under the curve was 0.82 (95%CI (0.78; 0.85)) for severe deficiency. The model predicted severe deficiency with a sensitivity of 77.9% (95%CI (69.1; 85.7)) and a specificity of 68.3% (95%CI (64.8; 71.9)). We identified a set of predictors of severe VitD deficiency that are easy to collect in routine that may help to better target patients for serum 25(OH)D concentration determination.