Intraindividual variation in urinary iodine concentrations: effect of adjustment on population distribution using two and three repeated spot urine collections

Objectives To determine the effect of adjustment for intraindividual variation on estimations of urinary iodine concentrations (UIC), prevalence of iodine deficiency and population distribution of iodine status. Setting Community-dwelling older adults from New South Wales, Australia. Participants 84 healthy men and women aged 60-95years were recruited prior to introduction of the mandatory iodine fortification programme. Primary and secondary outcome measures UIC data were collected from three spot urine samples, each 1week apart. Repeated measures analysis of variance were determined between-person (s(b)) and total (s(obs)) SDs. Adjusted UIC values were calculated as ((person's UIC-group mean)x(s(b)/s(obs)))+group mean, and a corrected UIC distribution was calculated. Results The s(b)/s(obs) for using three samples and two samples were 0.83 and 0.79, respectively. Following adjustment for intraindividual variation, the proportion with UIC <50g/L reduced from 33% to 19%, while the proportion with UIC 100g/L changed from 21% to 17%. The 95th centile for UIC decreased from 176 to 136g/L. Adjustment by taking averages yielded a lesser degree of contraction in the distribution than the analysis of variance method. Conclusions The addition of information about intraindividual variability has potential for increasing the interpretability of UIC data collected to monitor the iodine status of a population.