Statistical Analysis
Statistical analyses were performed using IBM SPSS (version 18, Somers, NY, USA). The statistical significance standard was set at 5%. Data normality was tested using the Shapiro Wilk test. If data points were not normally distributed, statistical analysis was attempted on the natural logarithm of the values to improve the symmetry and homoscedasticity of the distribution. Log-transformed values of 25(OH)D3, PTH, glucose, insulin, c-peptide, C-reactive protein, HbA1c, calcium and phosphorus were therefore used for comparative, univariate and multivariate analyses. For homogeneity of presentation, both normally distributed and non-normally distributed data are presented as medians with interquartile ranges (IQR). Differences in categorical variables were analyzed by χ test, and two-tailed, unpaired T- test or analysis of covariance were used for continuous variables depending on their distribution. Regression analyses and analysis of variance for multiple dependent variables by one or more factor variables or covariates were calculated using the general linear model multivariate procedure, comprising of the effects of covariate interactions with the variables of interest. To determine the impact of vitamin D status on metabolic variables, different models were attempted and the following covariates of interest, as well as their interaction, were used if not mutually excluded by the collinearity test: age, gender, BMI, percent fat body mass, fat body mass in weight, lean body mass in weight, and WHR. Multivariate linear and logistic regression analysis were used to estimate the coefficients of the linear equation, involving one or more independent variables, that best predict the value of the dependent variable.