We read with interest the article by Alvarez et al, which aimed to investigate the relations of circulating 25-hydroxyvitamin D [25(OH)D] and parathyroid hormone (PTH) concentrations with direct measurements of insulin sensitivity, after robust measures of body composition and fat distribution were accounted for. We would like to express our opinion and a different interpretation of the data provided by authors, with the hope that other points for discussion are brought up.

In a very recent publication, Alvarez et al provided novel findings suggesting that dietary vitamin D is independently associated with insulin sensitivity in African Americans (AAs) but not in European Americans (EAs). Interestingly, the 2 groups were identical for hepatic insulin sensitivity [homeostatic model assessment (HOMA)], whereas S_i, a method for measuring insulin sensitivity that encompasses both hepatic and peripheral tissues, was lower in AAs, therefore suggesting a pivotal role for insulin resistance in skeletal muscle [especially in the presence of identical body mass index (BMI)] in correlation with 25(OH)D. In the present article, the authors suggest that 25(OH)D and PTH concentrations are independently associated with whole-body insulin sensitivity and suggest that these variables may influence insulin sensitivity through independent mechanisms. In fact, multiple linear regression analysis indicated that 25(OH)D and PTH concentrations were independently related to S_i after adjustment for age, race, and intraabdominal adipose tissue. It is well known, however, that adipose tissue is the natural reservoir for lipo-soluble 25(OH)D. The higher BMI and the higher subcutaneous fat content found in AAs (although the latter difference was not statistically significant) could therefore explain the differences in 25(OH)D concentration, as well as in HOMA index, found by the authors.