Objective

To determine the effect of statistical correction for intra-individual variation on estimated urinary iodine concentration (UIC) by sampling on 3 consecutive days in four seasons in children.

Setting

School-aged children from urban and rural primary schools in Harbin, Heilongjiang, China.

Participants

748 and 640 children aged 8–11 years were recruited from urban and rural schools, respectively, in Harbin.

Primary and secondary outcome measures

The spot urine samples were collected once a day for 3 consecutive days in each season over 1 year. The UIC of the first day was corrected by two statistical correction methods: the average correction method (average of days 1, 2; average of days 1, 2 and 3) and the variance correction method (UIC of day 1 corrected by two replicates and by three replicates). The variance correction method determined the SD between subjects (S_b) and within subjects (S_w), and calculated the correction coefficient (F_i), F_i=S_b/(S_b+S_w/d_i), where d_i was the number of observations. The UIC of day 1 was then corrected using the following equation:

Results

The variance correction methods showed the overall F_i was 0.742 for 2 days’ correction and 0.829 for 3 days’ correction; the values for the seasons spring, summer, autumn and winter were 0.730, 0.684, 0.706 and 0.703 for 2 days’ correction and 0.809, 0.742, 0.796 and 0.804 for 3 days’ correction, respectively. After removal of the individual effect, the correlation coefficient between consecutive days was 0.224, and between non-consecutive days 0.050.

Conclusions

The variance correction method is effective for correcting intra-individual variation in estimated UIC following sampling on 3 consecutive days in four seasons in children. The method varies little between ages, sexes and urban or rural setting, but does vary between seasons.