Abstract
The literature in children and adults regarding thyroid function tests in obesity continues to grow. Most reports indicate that serum thyrotropin (TSH) concentrations are high normal or mildly elevated, or that the prevalence of elevated TSH is greater in obese than in nonobese persons (7,8,12,13,16,21 –23). Moreover, most (11,13,18,19,23) but not all (8) indicate that serum TSH concentrations decline or the prevalence of elevated serum TSH decreases after diet-induced weight loss. In some of these studies the decline in serum TSH after weight loss correlated better with changes in fasting insulin or triglycerides. Some, but not all reports found a correlation between changes in TSH with changes in body mass index (BMI) and circulating leptin (16,23) after diet-induced weight loss.
In the context of the TSH data, what are serum thyroid hormone concentrations in obesity? In the reports just cited, serum free thyroxine (fT4) concentrations were reported to be elevated (8,16), normal (7,12,13,21), or low (22,24) in obesity. Similarly in some reports serum free triiodothyronine (fT3) concentrations were elevated (8,13,16,19), in others they were normal (12,21), and in some they were low (22,24). In general, in reports in which serum T3 or fT3 concentrations were elevated, serum T4 or fT4 concentrations were normal (13) or elevated (8,16). In reports in which basal serum T3 or fT3 concentrations were normal, serum fT4 concentrations were also normal (12). When serum T3 or fT3 concentrations were reported to be low, serum fT4 concentrations were also low (22,24). Further, most (8,13,19,23) but not all (11) reports indicate that serum fT3 concentrations decline during diet-related weight loss, findings entirely consistent with the older literature. When studied, fT4 concentrations in obesity either declined (8,11) or did not change (18,23) with diet-induced weight loss.
In summary, obesity is associated with an upward shift in serum TSH, and in severe obesity serum TSH is above the normal range in about 20% of individuals. In contrast, serum thyroid hormone concentrations are only mildly affected in obesity without a consistent trend. Diet-induced weight loss is accompanied by a decline in serum TSH with no consistent trend in thyroid hormone levels except that during periods of fasting serum T3 declines. It is likely that the part of the reason for the variability data relates to age and other differences in the populations studied as well as in the design of weight loss study protocols.
Is thyroid status disturbed in obesity and, if so, what is the disturbance? Three possible states of thyroid status, the first in which local concentrations of thyroid hormone are inadequate, the second in which local concentrations of thyroid hormone are optimal, and the third in which local concentrations of thyroid hormone are excessive, can be postulated. “Thyrotoxic” is an optimal word for the third state. Unfortunately, “hypothyroid” and “euthyroid” are not because they are also used to express how the thyroid hormone secretion rates compare with the reference range. If obesity induces thyroid hormone resistance and there is a compensatory increase in thyroid hormone production, for example, then these patients might be considered to be euthyroid or near-euthyroid with respect to thyroid status but hyperthyroid with respect to their thyroid hormone secretion. Because the pattern of thyroid function tests in obesity is one in which TSH is high normal or mildly elevated, without consistent values for circulating thyroid hormone levels, some consider obesity to be a state of so-called “subclinical hypothyroidism,” a term generally accepted to mean mild primary thyroid failure. Others do not hold this view, some suggesting a state of thyroid hormone resistance in obesity that would lead to “inappropriate” TSH hypersecretion. It should be noted that current serum TSH assays do not measure its bioactivity; it is doubtful that satisfactory studies of TSH bioactivity in obesity have been performed, but the possibility that TSH bioactivity is altered in obesity should be considered.
There are fewer studies regarding the relationship of obesity to other thyroid disorders. Marzullo et al. (24) reported that the prevalence of anti-thyroid peroxidase antibodies was higher in obesity. In contrast, Rotondi et al. (22) noted that the prevalence of anti-thyroid antibodies in obese subjects with elevated serum TSH and normal thyroid hormone concentrations was lower than in similar normal weight subjects. Another observation is that a hypoechoic pattern on thyroid ultrasound may be less specific for autoimmune thyroid disease in obese than in normal weight subjects (25).
Perhaps the most notable association between obesity and thyroid disorders is that obesity is associated with an increased risk of differentiated thyroid carcinoma (2 –6). Elsewhere in this issue Clero et al. (28) report an analysis of two studies from New Caledonia and French Polynesia that found a positive association between an elevated BMI and thyroid cancer. Their analysis indicated that thyroid cancer risk was associated with height, weight, BMI, percent body fat (BF%), and body surface area (BSA). Remarkably, when adjustment was made for BSA, the association with height, weight, BMI, and BF% did not remain. The authors indicate that thyroid volume is positively related to BSA and hypothesize that increased thyroid volume, in turn, raises the risk of thyroid cancer. Whether the latter is the case is not certain, but the study is of great interest. It should motivate investigators to consider and power future studies to test the relative importance of BSA in thyroid-related obesity studies.
Is there a link between the thyroid status changes and the heightened risk of thyroid cancer in obesity? It is intriguing that TSH tends to be elevated in obesity and that serum TSH concentrations are relatively elevated in thyroid cancer (29 –32). It is easy to envision a causal relationship whereby obesity leads to elevated serum TSH and this in turn to thyroid cancer. In support of the initial step are the well-established positive effects of the fat cell product leptin on TSH and thyroid hormone secretion (33,34).
To date most of the literature in this field does not distinguish between the various types of obesity. More studies in this area are needed. Additionally, the study of Clero et al. (28) is a call to consider anthropomorphic features other than obesity that might relate to thyroid pathophysiology, realizing that the metabolic perturbations underlying the various phenotypes must ultimately be the focus of study.