Iodoprophylaxis and thyroid autoimmunity: an update

Autoimmune thyroid diseases (AITD) are the most common autoimmune disorders [1, 2]. Autoimmune thyroiditis (AT) is the paradigm of “destructive” organ-specific autoimmunity, and Graves’ disease (GD) is the unique clinically relevant case of organ-specific autoimmune disease resulting in hyperfunction of the target organ and the most common cause of hyperthyroidism with the possible exception of areas with severe iodine deficiency. The multi-faceted clinical presentation is due to multiple pathogenetic mechanisms involving (cell-mediated in autoimmune hypothyroidism due to thyroid destruction and autoantibody-mediated in autoimmune hyperthyroidism) [3‐5]. In any case, the presence of thyroid autoantibodies has been regarded for decades as a hallmark of AITD: notably, antibodies directed against thyroid peroxidase (TPOAb) or thyroglobulin (TgAb) [6]. Such autoantibodies are present in both AT and GD, although more frequently and at higher levels in the former. The prevalence of thyroid antibodies varies between populations under the influence of genetic [7‐10] and environmental [7, 11] factors.

The prevalence of AT is far higher than that of GD (about 5- to tenfold) [1]. Therefore, the presence of such autoantibodies is predominantly associated with decreased thyroid function. It has to be underlined that the presence of circulating TPOAb or TgAb does not always coincide with the presence of disease: clinical disease requires significant organ damage, in which cell-mediated mechanisms play a major role, as above mentioned. Such thyroid autoantibodies can be predictive of disease development over time [12] but also be a transient phenomenon; thus, their prevalence is higher than clinical disease [13]. On the other hand, TSH-receptor autoantibodies (TRAb) are the only autoantibodies displaying a direct pathogenetic role, most often causing hyperthyroidism due to their thyroid-stimulating activity: in fact, since the development of highly sensitive assays [14], their prevalence is closely similar to that of clinical GD.

Normal thyroid activity is dependent on iodine intake. In fact, iodine is the most important environmental factor in upholding thyroid function [15].

Iodine deficiency has been since centuries a major epidemiological problem. In fact, since the very beginning of life outside the oceans, some 400 million years ago, the availability of iodine in the environment became very poor: there is circumstantial very early evidence for iodine deficiency disorders in Homo sapiens as well as in Neanderthals, and iodine availability and the ability to use it has probably been an important factor in human evolution [16]. Endemic goiter is the most visible sign of iodine deficiency, but its most devastating consequence is brain damage causing mental retardation in children.

In Table 1, the optimal levels of iodine intake by age and population group according to the World Health Organization (WHO) and the United States Institute of Medicine (IOM) are reported. Iodine supplementation through salt iodization, implemented by the WHO and the International Council for the Control of Iodine Deficiency Disorders (ICCIDD, now Iodine Global Network), is a worldwide, effective strategy for preventing iodine deficiency-related problems. Its safety and efficacy profile has been extensively investigated, and benefits far outweigh the potential iodine-induced risks [17, 18]. Iodine prophylaxis through iodized salt has been shown to modulate the pattern of thyroid diseases and also to exert a pivotal role in abating iodine deficiency disorders (IDD) [19, 20].

Iodine excess also exerts important effects on thyroid function, being able to transiently block it (Wolff-Chaikoff effect), or producing long-lasting hypothyroidism; moreover, iodine excess may cause transient destructive thyrotoxicosis via toxic damage, or sustained hyperthyroidism in the presence of autonomously functioning tissue [21]. Finally, and more relevant to this review, thyroid autoimmunity can be driven by exposure to iodine excess.

For years, several mechanisms have been suggested to explain the immunomodulatory effects of iodine and specifically the relationship between the level of iodine intake and thyroid autoimmunity [22, 23]. It is still not clear whether the observed immunological changes are due to a direct iodine effect on immune effector cells, or whether they represent a secondary response to a metabolic and/or toxic effect of iodine on thyroid tissue. It has been hypothesized that a sudden shift from very low to high iodine intake may induce damage to thyroid tissue by free radicals [24, 25] whose importance in triggering/exacerbating thyroid autoimmunity is presently widely recognized [26‐28].

The role of iodine in inducing thyroid autoimmunity is strongly supported by animal models. Excess of iodine can anticipate and exacerbate the occurrence of spontaneous thyroiditis in genetically predisposed animals, by increasing the immunogenicity of thyroglobulin (Tg) [23, 29‐31]. Thyroglobulin is an iodinated protein and is an integral part of the hormone synthesizing mechanisms of the thyroid follicle.

Some authors described iodinated thyroglobulin eliciting a greater immune response compared with less iodinated thyroglobulin molecules in in vitro experiments [29, 32, 33]. It has been demonstrated that the enhanced iodination of thyroglobulin induces the expression of new cryptic epitopes on the molecule that could be responsible for triggering the autoimmune process [30]. The rise of TgAb associated with iodine intake in humans, similarly to what is observed in animals, is likely due to an increased immunogenicity of Tg [30, 34‐36]. This may be explained by the fact that Tg is the only self-antigen that undergoes post-translational modification as a consequence of the environmental supply of iodine, with the exposure of previously hidden epitopes. Direct evidence supporting the latter mechanism was recently provided by Latrofa et al. [37] who carried out a detailed study of the fine specificity of TgAb detected in sera of an endemic population before and after implementation of iodine prophylaxis.

However, while the role of the iodine intake level in the etiology of non-autoimmune thyroid disorders is well established, the association between population iodine intake level and AITD is more devious [6]. Taking into account the global health benefit of iodine prophylaxis implementation, the question arises as to whether the implementation of universal salt iodization has led to an increased incidence of AITD and, if so, what is its clinical relevance/impact.

As above mentioned, it is well known that variable iodine intake may be accompanied by thyroid disease: adverse consequences of low iodine intake are a major epidemiological problem, but thyroid dysfunction may also occur at higher intake levels [38‐40].

In subjects with an underlying thyroid disorder (underlying autoimmune thyroiditis, previous treatment with radioactive iodine or history of external thyroid irradiation, previous subtotal thyroidectomy, postpartum or subacute thyroiditis and intake of some medications, such as lithium, which interferes with iodine organification and thyroid hormones release), acute excessive iodine intakes may lead to temporary overt or subclinical hypothyroidism that resolves when iodine intakes decrease [41].

Moreover, many studies indicate that even relatively small changes in the level of iodine intake of a population may result in changes in the spectrum of thyroid diseases [42]. Iodine-induced hyperthyroidism has been observed at daily iodine intakes of less than 300 µg [43, 44]. An increased incidence of hyperthyroidism in previously iodine-deficient areas has been described, although transiently, even during careful implementation of universal salt iodization [15, 45‐47]. Therefore, the World Health Organization (WHO) recommends that iodine fortification of a population should be initiated cautiously and be followed by a monitoring program to register its effects and counteract any side events.

It has been proposed the term “iodine memory” [15] when evaluating the association between current iodine intake and the epidemiology of thyroid disease in a population, because the degree of a previous exposure to different level of iodine intake (low or high) may have influenced the current occurrences of diseases. An increase from low to normal iodine intake is associated with a reduction in thyroid size in the population within a few years [48], while thyroid nodularity is more or less irreversible [49]. It remains to be elucidated if the high frequency of hypothyroidism in areas with excessive iodine intake is a reversible phenomenon. Studies of hypothyroid patients living in Japan and Korea with high iodine intake have indicated that a reduction in iodine intake may normalize thyroid function in some of these patients [50, 51]. Individuals with preexisting thyroid disease or those previously exposed to iodine deficiency may be more susceptible to thyroid disorders due to an increase in iodine intake, in some cases at intakes only slightly above physiological needs. Thyroid dysfunction due to excess iodine intake is usually mild and transient, but iodine-induced hyperthyroidism can be life-threatening in some individuals [21].

The question arises as to such “side pathological effects” are related to the development of thyroid autoimmunity, although there is no doubt that other factors play a role (specifically, the development of hyperthyroidism in iodine-deficient areas where the prevalence of nodular goiter, and consequently the likelihood of nodular functional autonomy, is high).

The effect of public iodization programs on the development of thyroid antibodies is yet uncertain. As previously described, high iodine intake may trigger thyroid autoimmunity, and it might be expected that the prevalence of harboring circulating thyroid antibodies would be low in populations with a low iodine intake and high in populations with a high iodine intake. However, the association between iodine intake and the presence of circulating thyroid antibodies is considerably more complex, taking also into consideration that multiple genetic and environmental factors are involved in the pathogenesis of autoimmune thyroiditis.

The most common autoantibodies against the thyroid gland associated with autoimmune thyroiditis are thyroid peroxidase antibody (TPOAb) and thyroglobulin antibody (TgAb) [52]. Antibody levels seem to correlate well with the severity of histological thyroiditis [53], but in some cases, histological evidence of autoimmune thyroiditis is not accompanied by detectable TPOAb or TgAb in blood [54].

As already noted, it is difficult to interpret and compare the many studies where both thyroid antibodies and iodine status have been measured in one or more populations [55].

Iodine deficiency has been associated with goiter and people with nodular goiter relatively often have circulating thyroid antibodies [56, 57], probably caused by enhanced release of thyroid antigens from the abnormal gland. Thus, circulating thyroid antibodies are probably more common in populations where goiter is common because of iodine deficiency. Consistently with such a scenario, recent findings indicate that iodine deficiency in early pregnancy is associated with higher risk of TPOAb and TgAb positivity and that the presence of both autoantibodies at high titers is associated with increased risk of subclinical or overt hypothyroidism [58], and optimal iodine supply results in the lowest risk for thyroid autoantibody positivity [59].

It has also been reported that the prevalence of lymphocytic thyroid infiltration becomes high after an increase in population iodine intake [60]. Moreover, a sudden increase in iodine intake in an iodine-deficient population may induce enhanced thyroid autoimmunity [61], but this may be a transient phenomenon [62].

Two groups of Sri Lankan schoolgirls were studied in 1998 and 2001 in relation to the evolution of thyroid autoimmunity, with the change in goiter prevalence, during 3 years of iodine prophylaxis [63]. This study demonstrates that in 2001, goiter prevalence and thyroid autoimmunity rates were significantly lower than in 1998; the pattern of thyroid Ab was different in the two groups (more TPOAb in 2001 group); in 2001 alone, the occurrence of hypothyroidism was correlated with the presence of thyroid autoimmunity. These results indicated, for the first time, an evolution of thyroid autoimmunity markers during the course of iodine prophylaxis: when iodine supplementation is first introduced it may have induced the greater prevalence of TgAb in 1998 in comparison with 2001. It is noteworthy that the reversibility of thyroid autoimmunity occurred mainly in those individuals who showed low titers of TgAb alone in 1998: this fact suggests that the immune system’s response to iodine intake could be heterogeneous, possibly reflecting different immunological background.

In Greece, thyroid autoimmunity has been positively associated with increased iodine intake, as well as with the female gender [64]. Moreover, mean TSH values were increased in females and decreased with age. The latter is probably due to the presence of autonomous goiter in older subjects, as a result of long-term status of iodine deficiency in the past.

As well, an increased prevalence of thyroid autoantibody positivity and subclinical hypothyroidism was observed in Denmark, after 4–5 years from the implementation of a cautious iodoprophylaxis program in 3570 members from the original cohort [65]. Similar results were obtained in the Pescopagano’s survey, a small rural community followed up for as long as 15 years from iodoprophylaxis with iodized salt implementation, which in turn resulted in marked reduction of the prevalence of goiter and thyroid autonomy [66].

However, other studies report no increase or a reduction in thyroid autoimmunity with time following iodine salt fortification. A large (more than 350,000 people) retrospective study conducted in Tasmania between 1995 (median population UIC: 75 μg/L) and 2013 (median population UIC: 108 μg/L) and aimed at analyzing the relationship between trends in TSH and TPOAb testing and different phases of iodine nutrition failed to detect any increasing trend over time of TPOAb following iodine salt fortification, whereas a decreasing trend of overt thyroid dysfunction was apparent [67]. Another study, which analyzed the impact of iodine nutrition on thyroid diseases few years after the introduction of the iodine prophylaxis program (1997–2001, median population UIC: 123 μg/L) and more recently in Northeast Germany (2008–2012, median population UIC: 112 μg/L), also found a stable prevalence of markers of thyroid autoimmune disorders and a decreased prevalence of goiter [68]. More recently, in a study conducted in China after two decades of universal salt iodination program, a stable prevalence of thyroid antibody positivity in the population has been reported. This study also found an inverse relationship between iodine intake and thyroid antibodies, suggesting that UIC between 100 and 300 μg/L is optimal and safe for thyroid autoimmunity [69]. More specifically, a number of data suggest that only iodine supplementation regimens attaining relatively high median UIC levels (i.e., 200 µg/L or more) are able to induce relevant long-term alteration of thyroid function due to autoimmune disease, and the relationship between population iodine supply and occurrence of AITD could be depicted by a “U-shaped” function [70, 71]: in other words, the likelihood of developing thyroid autoimmune diseases would increase when median population UIC exceeds approximately 300 µg/L, but as well when is less than 100 µg/L (see Fig. 1). The majority of these results come from Chinese populations, in whom marked differences in iodine supply in single region have been observed due to different iodine supply strategies [69, 71‐75]. In line with such a scenario, according to at least two recent reports cited above [58, 59] both iodine deficiency and iodine excess are associated with increased risk of thyroid autoantibody positivity in pregnancy.

Moreover, circulating thyroid antibodies are not extraordinarily common in populations with stable high iodine intake [38], although thyroid dysfunction is more frequent when the iodine intake is high [76].

Importantly, most recent data from the Danish survey in North Jutland, involving an open cohort of over 300,000 individuals, demonstrated no increase of overt hypothyroidism (related, as a rule, to autoimmune thyroiditis) and a decrease of hyperthyroidism (due to thyroid autonomy, but also due to autoimmunity, i.e., Graves’ disease) over a period of more than 15 years following implementation of mandatory iodine fortification [77].

Finally, and on the same line of evidence, the last data derived from the Sri Lanka iodoprophylaxis survey showed that more than 20 years after its implementation the prevalence of TgAb consistently declined, and no increase of the prevalence of hypothyroidism (even subclinical) was recorded [78].

Thus, although several reports suggest a somehow promoting effect of iodine salt fortification on thyroid autoimmunity, its role in causing clinically relevant autoimmune thyroid disease in long term is not clearly proven.