Discussion
The studies about the effects of soy on the thyroid gland became available in the literature quite a long time ago [12-13]. The hormonal and non-hormonal effects of soy have been demonstrated to occur primarily via isoflavones [14-15] . Isoflavones constitute a subclass of a larger and more ubiquitous group of plant chemicals, the flavonoids. The three isoflavones found in soy are genistein, daidzein, and glycitein [16]. Divi et al. have shown with in vitro andin vivo studies [17-18] that isoflavones inhibit the activity of thyroid peroxidase (TPO). Soy has been shown to cause thyroid dysfunctions in susceptible individuals and prevent the absorption of synthetic thyroid hormones [19-20]. Genisteins are known to affect cardiomyocytes, breast cancer cells, and osteoblasts in a dose-dependent manner [21-22] . Tran et al. demonstrated that; while 1-10 µM genistein induced thyrocyte hyperfunctioning, 100 µM genistein reversed those changes [22]. In vitro studies have shown that genistein and daidzein block TPO-catalyzed tyrosine iodination by acting as alternate substrates for iodination [18]. The incubation of either genistein or daidzein with TPO in the presence of hydrogen peroxide inactivates TPO irreversibly but the addition of iodide to the medium abolishes this inactivation [23]. These in vitro data show that soy increases the requirement for iodine [24].
Several hypotheses have been suggested about the effects of soy on the thyroid gland. Soy is rich in phytate; which can bind T3 and increase its elimination via faeces [25]. Isoflavones can bind to thyroid hormone receptors, resulting from their structural similarity with T3 [26] ; reduce plasma protein binding to T4 , or reduce the rate of T4 deiodination, which is a reaction to produce T3 with the 5’-deiodinase type I acting as the catalyzing enzyme [27] .
Of three studies, investigating the changes in serum TSH levels in adult women , only one demonstrated statistically significant TSH level alterations. Persky et al. found out increased TSH levels in the third and sixth months compared to the control group fed with fat-free milk . Studies investigating only the serum levels of free T3 and free T4 reported increases in the range of 8 to 16% [28-31]. However, those values were not statistically significant. Dillingham et al. have carried out the most comprehensive study so far, in which they investigated the effects of soy on thyroid function in male individuals. In that study, 35 young adult men were fed with a diet containing approximately 32 g of isoflavones. The study did not find any statistically significant differences in the serum levels of total T3, free T3, total T4, free T4, TSH, and thyroxine-binding globulin [26]. As it is observed, the studies on adults report variable results.
Also, experimental and clinical studies on fetuses and newborns obtained variable results similar to the results obtained in adult studies. Chang et al. demonstrated that in utero feeding of the rats with a diet containing 500 ppm starting from the 20th weeks of gestation resulted in the inactivation of thyroid peroxidase by approximately 50%. However, they observed no changes in the serum levels of T3, T4, and TSH; in the weight of the thyroid gland, or in the histopathological examination findings [32-3] . Conversely, Son et al. demonstrated that soy-containing diet caused thyroid hyperplasia in female rats, especially when the iodide intake was inadequate [34]. Conrad et al. [10] showed a prolonged increase in the serum TSH levels in infants fed with soy-containing formulas in comparison to those fed with soy-free formula. In their retrospective study on infants with hypothyroidism; Conrad et al. reviewed 70 infants fed with non-soy formulas and 8 infants fed with soy-containing formulas. After 4 months of L-thyroxine treatment, they observed similar T4 levels in the two groups but the levels of TSH increased in 17% of the infants fed with soy-free formula and in 62.5% of the infants fed with soy formula.
In our study, we observed body weight increases in the subjects of both soy and non-soy formula fed groups compared to the subjects fed with standard pellet feed in the control group. However, these changes were not statistically significant. We think that these findings occurred because of the higher calorie counts of the infant formulas than those of the rat pellet feed.
We observed that serum free T3, free T4, and TSH levels ​​increased in a dose-dependent manner in soy-fed subjects and that those increases were statistically significant in subjects consuming high-dose (2.64g / 100ml) soy-containing foods. Furthermore; we found that the inflammation parameters of the thyroid gland, Anti TG and Anti TPO, did not increase in subjects receiving low-dose soy but these parameters increased significantly in subjects receiving high-dose soy. In the histopathological examination, we observed a mild increase in the subjects fed with the high-dose soy formula but this change was not statistically significant. When we make an overall assessment of our study findings; we observe that feeding high-dose soy formula induces inflammation in the thyroid tissue, resulting in increased thyroid hormone release independently from suppressing effects of TSH. The reason why this functional change is not reflected in the histopathological findings in our study might be related to the duration of the feeding with the formulas. A longer duration of high-dose soy intake may induce more significant inflammation in the thyroid tissue.
In conclusion; feeding infants with high-dose soy containing formulas may trigger inflammatory processes in the thyroid tissue, resulting in thyroid dysfunction. As it is known that thyroid hormones are involved as catalyzers in many chemical reactions in the body, infant formulas with high soy content may cause several metabolic problems in babies in the long term.