As discussed above, several groups of EDCs may have thyroid disrupting potential, but only perchlorate and PCBs have been studied in more detail in humans. Perchlorate reduced expectedly thyroid iodine uptake, but so far, no significant effects on circulating thyroid hormones have been found after exposure to environmental levels of either perchlorate, thiocyanate, or nitrite. Most of the other chemicals have still only been studied in animal models, sporadically, in high doses in volunteers or after occupational or accidental exposure, and results are conflicting. However, all the mentioned chemicals can theoretically have thyroid disrupting properties and consequently further studies are needed to clarify the mechanisms and the general consequences of constant environmental exposure to lower doses. Although thyroid disrupting properties were not documented for all chemicals, especially vulnerable groups like pregnant women, foetuses and children of all ages may be more sensitive because of pregnancy- and growth-related added stress on the thyroid gland, in particular for people living in iodine insufficient areas. Most human studies are performed in groups like healthy volunteers, occupationally exposed individuals, or persons living in certain areas and do not include all thyroid relevant factors as life style, preexisting thyroid disease, age groups, or exposure to other EDCs. However, exposure during the foetal and neonatal period is of much concern, as it is a very vulnerable point in central nervous system development, especially in preterm children. Only few studies of the chemicals in question have addressed the issue of health effects on the offspring of exposed subjects. Yet, many of the potential thyroid disrupting chemicals accumulate both in nature and in exposed individuals and may have a negative influence on maternal thyroid function during pregnancy with consequent risk of impaired neurodevelopment of the foetus. While significant exposure to all these chemicals are suspected to affect human thyroid homeostasis, the effects of environmental exposure still remain to be confirmed in humans and, in particular, in vulnerable groups.
Epidemiological studies have reported that pre- and perinatal exposure to PCBs is associated with poorer neurodevelopment in neonates, toddlers and school-age children [142
]. The influence of PCBs on thyroid function has been suggested as a reasonable explanation for the results although this was not evaluated in detail. PCB correlated negatively to fT4 in pregnant women [148
], and therefore, even exposure at background levels could possible disturb foetal development.
The subjects in human epidemiological studies have always been exposed to many different compounds through different time periods, and it is, therefore, difficult to isolate specific effects of chemicals and their metabolites on functions of the human organism, which is an obvious caveat in concluding from such studies [59
Some studies have been performed in people more intensively exposed due to either occupation, residency in/near contaminated areas [74
], accidents [151
], or fish consumption [78
], but other studies have focused on general population exposures [58
]. There may, thus, be several reasons for the divergence in findings. One explanation could be current low exposure after reduction of allowed limits and, therefore, current unmeasurable levels of a chemical that once was present and exerted an effect. Conflicting results may also reflect that findings depend on the choice of biomarkers, detection methods of the examined EDCs, and sample material, for example, in maternal blood, breast milk, cord blood, or child blood. Furthermore the sex of the foetus, comorbidities, and medication as well as a possible influence from combined effects of other EDCs may influence study outcomes [72
]. Even in adult populations, there are probably both age and gender differences in responses in an adult population [83
Given that most of the mentioned chemicals have subtle influences on the thyroid axis, in many cases within the normal reference interval, the question is whether or not such subtle changes in, for example, maternal thyroid function can eventually compromise foetal neurological development. The relationship between T4 and TSH is very unique to each human [154
], and the variations within each person are much smaller than the variation within a population [155
], which is also the case during pregnancy [157
]. Comparison with more or less well-defined population-based reference ranges is probably quite irrelevant considering the discrepancy between these large ranges compared to the much narrower intraindividual variations in thyroid hormone levels [155
]. In addition, no first-trimester-specific reference ranges for fT4 analog assays currently exist, available commercial analog fT4 assays are unreliable in pregnant women, and fT4 levels are often over- or under-estimated. In these cases, TT4 and free thyroid hormones indexes are more reliable [159
]. Consequently, minor, yet real, changes in thyroid hormone levels due to EDC exposure in small human studies may easily be camouflaged by the broad interindividual variation. As human exposure is lifelong, starting during pregnancy and cumulative for persistent chemicals, it is not possible to design human studies evaluating thyroid function within an individual before and after exposure. Even small intervention studies, like the study with triclosan [109
], are performed on a preexisting background of chemical exposure to many other compounds simultaneously.
Despite this individuality of the thyroid function variables, the levels of TSH and thyroid hormones vary greatly during the early stages of life. TSH increases dramatically immediately after birth peaking at 30 minutes, followed by an increase in T4 and T3, where after all hormone levels decrease. Thyroid hormones measured in newborns may be affected by intrapartum stress [67
] and even by use of iodine containing antiseptics [160
]. Thus, estimation of any influence of thyroid disrupting chemicals on TSH and thyroid hormones during pregnancy, neonatal period, or early childhood should, therefore, allow for exact age as a critical confounder.
A possible influence of thyroid hormone-induced metabolism and elimination processes of EDCs, such as detoxification in the liver and kidneys, has not been extensively investigated, and further studies should be performed. Other confounding factors in interpretation of the many results include population-specific level of selenium and iodine, since deficiency of these two substances may render the thyroid system more prone to be affected by EDCs. In addition, exposure to EDCs may cause only transient changes in thyroid hormone levels, which cannot be traced afterwards but, nevertheless, may leave permanent effects on the central nervous system if occurring during a developmentally critical phase. Furthermore, measurement of peripheral thyroid hormone concentrations may not reflect a chemical effect on the full thyroid homeostasis (). As outlined in this paper, various chemicals may have different and antagonistic or synergistic effects on the thyroid axis. Such effects have also been found in studies of chemicals disrupting reproduction [18
Finally, it is not possible in association studies to distinguish whether EDCs could act by direct toxic effects or by indirect mechanisms via disrupting the thyroid function. More mechanistic studies are, therefore, warranted in the future.