Response to comments on urine iodine excretion in patients with euthyroid nodular disease.

To the Editor: In the September–October/2011 issue of your journal I commented1 on the article by Cakir and colleagues on urinary iodine excretion in patients with euthyroid nodular disease.2 I am glad that Cakir responded to my concerns.3 Cakir made the point that there are controversies about the cut-off values on urinary iodine excretion and that measurement of iodine levels in spot urine sample is found to be as sensitive as other screening modalities. As well, Cakir indicated that he confirmed his results by a dietary iodine questionnaire. I believe these points need further clarification.

First, as I stated in my letter,1 the subject of cut-off points for urinary iodine level makes no difference as the basic concept did not prevail. Iodine levels in casual urine samples (absolute or iodine creatinine ratio) or in 24-hour urine samples do not evaluate the iodine status of individuals.4 When correction of iodine deficiency in a community is contemplated, the community rather than those individuals with values lower than any specific cutoff point is targeted. The Korean study5 quoted by Cakir in support of his work evaluated patients prescribed a low iodine diet in preparation for radio-iodine therapy and that is an entirely different issue. On the other hand, confirmation of iodine status with dietary questionnaire may not be completely valid. Dietary assessments of the iodine status have many more sources of error than measurement of the element in a biological fluid like urine. Although, weighted records and measurement of iodine in duplicate portions are superior to semi-quantitative method used by Cakir and colleagues in their original study, the iodine intake estimated by direct analysis of a duplicate portion gave values 3 to 6 times those values given by the weighted record method. Furthermore, iodine-rich foods like fish have been shown to contribute only 5% of daily iodine intake.6 Indeed, the iodine content of the same type of diet may vary a lot; in the same plant iodine content of old leaves is higher than of younger ones.

Second, Cakir and colleagues started their original study2 with three groups significantly different in the proportions of smokers and those with additional diseases. Any difference in outcome after one year of follow up could be attributed to one of these two variables. In truth, these two variables are confounding variables that could seriously compromise the internal validity of the study and ideally care of them should have been taken while designing the study. Patients showing these two variables should have been excluded or made equally represented in the different study groups. The later could have been the case if the sample size was large enough. Alternatively, an attempt should have been undertaken to correct for their effects by considering them as co-variables. A statistician could have been of great help. It might be of interest to look back at the original data and use smoking status as a grouping variable, i.e., compare the change in volume of thyroid lobes and nodules in the 20 smokers and the 42 nonsmokers. This seems logical as the iodine status of Cakir’s population is not known while smoking status and additional diseases were documented. To conclude, I believe that Cakir and colleagues’ study raised some questions regarding the relationship between thyroid nodular disease on one side and smoking and other diseases on the other side. If the relationship between iodine deficiency and thyroid nodular disease needs to be assessed, a different approach may be required.