Response to Yüksel Güler and Haymana re: “Influence of Dietary Habits on Oxidative Stress Markers in Hashimoto's Thyroiditis”

Dear Editor:

I thank Bağdagül Yüksel Güler and Cem Haymana for commenting on “Influence of dietary habits on oxidative stress markers in patients with Hashimoto's Thyroiditis (1).” The aim of the study was to evaluate how diet can influence the oxidative stress balance of the human body in the setting of autoimmune thyroiditis (2). Our study findings indicate that a diet rich in vegetables and raw foods like the Mediterranean diet is able to counterbalance the altered redox homeostasis, while meat consumption appears to be linked to a general increase in inflammation and oxidative stress. We provided evidence that this dietary pattern may represent an efficacious tool to prevent/counteract thyroid autoimmunity (2).

I really appreciate Drs. Bağdagül's and Cem's interest in our study. First, the authors state that “the ultrasound results of patients were not reported in the manuscript,” and that it is particularly important to know the ultrasonographic patterns of patients with normal serum antithyroid antibody levels. I agree with them on the relevance of ultrasonography (US) in the diagnostic work-up of thyroid autoimmune diseases and we are aware that a hypoechoic pattern on thyroid US can predict/detect thyroid autoimmunity in subjects without serum thyroid autoantibodies positivity (3). For this reason, all subjects who were included in the study underwent thyroid US, as stated in the material and methods section. In the results section it was reported that the subjects who were diagnosed with Hashimoto's thyroiditis displayed the typical US features of autoimmune thyroiditis, that is, heterogeneous echostructure with diffuse or patchy hypoechogenicity. On the contrary, those who were not diagnosed with Hashimoto's thyroiditis, and served as control subjects, had no ultrasound alterations. The US features of subjects with negative serum antithyroid antibody levels were reported and were normal (no ultrasound alterations, page 3, right column, “results” section). This allowed us to surely exclude an autoimmune disorder in these individuals.

Second, I respectfully disagree with them that it is important to numerically report anti-thyroperoxidase antibodies (TPOAb) levels in the control group “for the objectivity of the study.” This does not seem to me a relevant issue. The crucial point is that all control subjects tested negative for TPOAb, while those who were diagnosed with Hashimoto's thyroiditis tested positive. For the sake of simplicity, we reported the numerical values of TPOAb only from patients, avoiding displaying normal values from healthy subjects, as is commonly done in other studies. Moreover, the normal range was very narrow (0–10 IU/L) in our laboratory.

Lastly, in Table 3 of the article (2), the values of confidence interval (CI) related to sex and body mass index (BMI) were incorrectly reported, whereas odds ratio (OR) and p-values were correct. The right values were sex OR 0.859, 95% CI [CI 0.770–0.958], p = 0.006; BMI OR 0.842, [CI 0.499–1.421], p = 0.081. It was an error in transcribing the data in Table 3, and an erratum is added to the present letter.

In conclusion, I appreciate the authors' interest and their careful evaluation of the article. However, I disagree with them concerning the methodological issues they raise. The concerns they had raised are not well founded and do not affect the correct interpretation of our data. The study cohort is well characterized, and autoimmune thyroiditis has been correctly diagnosed or ruled out according to the currently accepted guidelines (4). All relevant data were clearly reported in the article and no important data were omitted.