Relative proportions of participants with serum 25(OH)D concentrations (nmol/L) below different clinical thresholds.¹

1Clinical thresholds were chosen to encompass the recommendations in the current literature [2][4][7][30][6]. <25 nmol/L = deficient, 25–50 nmol/L = insufficient, 50–75 nmol/L = sufficient, >75 nmol/L = optimal. 2Measurements for each time period were made over three weeks at the end of each season. “Winter 2011” represented the follow-up period, nine months after completion of the primary study, and after the participants were released from restricting vitamin D supplementation (study exclusion criterion>800 IU/day). 3Number of subjects assessed at each time period are shown, and mean serum 25(OH)D concentration (±standard deviation) were estimated by mixed methods linear regression, adjusted for participant age and time from beginning of study. Large variation in participant numbers (Winter 2009, Spring 2010) is because a small group of participants commenced the study at the end of Winter (September) 2009 and completed at the end of Winter 2010; the majority commenced at the end of Spring (December) 2009 and completed at the end of Spring 2010. 4The relative proportion of participants having serum 25(OH)D concentration below different clinical thresholds shown was compared to the relative proportion at the end of Winter (September) 2010 (chosen as the references because the end of Winter 2009 represented the pilot group only), estimated using repeated-measures negative binomial regression and expressed as an incidence rate ratio (IRR; 95% confidence intervals; p-values).