Screening of subjects at risk of severe vitamin D deficiendy: a clustering approach

Introduction: Avoiding vitamin D deficiency is essential regarding several health outcomes. Systematic blood testing may represent an important economic burden and systematic supplementation could lead, in some subjects, to a very high status, with unknown long-term consequences. Therefore, it is crucial to implement easy-to-apply strategies for screening at-risk patients.\nObjectives: Our objective was thus to characterize individuals at high risk of severe vitamin D deficiency (25OHD<10ng/ml).\nMethod / Design: A combination of hierarchical and nonhierarchical cluster analysis was performed on 1528 French Caucasian adults (45-60y) from the SU.VI.MAX cohort. The following baseline variables (collected through self-administered questionnaires and anthropometric measurements) were included in the clustering procedure:\nsevere vitamin D deficiency (yes/no, Roche Cobas® electrochemoluminescent assay on baseline plasma samples), gender, age, BMI, physical activity, educational level, dietary intake of vitamin D, latitude, sun exposure, Fitzpatrick phototype, and month of blood draw.\nResults: Two clusters were identified. Cluster 1 consisted of all participants with severe vitamin D deficiency and was characterized by the overrepresentation of: very low sun exposure, obesity, female gender, blood draw at the end of winter or early spring, Northern latitudes, irregular or low physical activity and the fairest skin phototypes.\nCluster 2 consisted of all participants without severe vitamin D deficiency.\nConclusions: This study presented for the first time a clustering analysis to identify high-risk individuals for severe vitamin D deficiency. This approach, based on easy-to-assess phenotypic, sociodemographic and lifestyle characteristics, can help to improve clinical practice by better targeting patients at need for vitamin D supplementation and/or blood testing.