Data obtained from the entry records of 4,116 adult men and women in Calgary,Alberta previously enrolled in the Pure North S'Energy Foundation's health program. Individuals with a clinical diagnosis of diabetes mellitus or a HOMA-IR value above 16 were excluded. Standard statistical analyses carried out. An association between 25(OH)D and the endpoints of interest was first established. The residual values after adjusting HOMA-IR, serum insulin, QUICKI, and blood pressure for BMI, age, sex, and waist circumference were then regressed against serum 25(OH)D using 40-percentile data blocks, stepping upward from the lowest 25(OH)D values by 5 percentile increments through the 7th decile. Regression coefficients are then computed for each data block and each outcome variable. The location (along the 25(OF)D continuum) of the largest and most statistically significant coefficients is noted, and regions of null response noted as well.
