The Effects of Anabolic Steroids on Thyroid Function


by Karl Hoffman

One of the more commonly encountered assertions in the bodybuilding literature is that anabolic steroids (AAS) suppress thyroid function. What is the origin of this claim? Is it supported by evidence published in the scientific literature? And perhaps most important is it of any clinical significance, meaning if it does occur is it serious enough to worry about? Before reviewing the evidence for and against AAS induced thyroid impairment a short review of thyroid physiology is probably in order.

The thyroid gland secretes principally the hormone thyroxine (T4) along with small amounts of triiodothyronine (T3). The majority of the T3 in the circulation (about 80%) is formed from the conversion of T4 to T3 by the enzyme deiodinase. Most of this transformation takes place in the liver and kidney. T3 is considered the physiologically active hormone; in this sense T4 can be thought of as a prohormone.

The production within and secretion from the thyroid gland of T4 is under the control of Thyroid Stimulating Hormone (TSH), which is secreted from the pituitary gland. Higher levels of TSH lead to higher rates of hormone production and secretion from the thyroid. TSH in turn is regulated by another hormone secreted from the hypothalamus, thyrotropin-releasing hormone (TRH). TSH levels are also regulated in a negative feedback manner by the levels of circulating thyroid hormone. If extra hormone is administered, transcription of the TSH gene is decreased and less TSH is produced by the pituitary, leading to suppression of natural thyroid hormone production. Similarly, a decrease in the rate of thyroid hormone secretion leads to enhanced TSH production in an attempt to return to homeostasis.

Just as the bulk of circulating androgens and estrogens are bound to sex hormone binding globulin (SHBG), most of the thyroid hormone in circulation is bound to thyroid binding globulin (TBG). And as with SHBG and sex hormones, the levels of TBG influence the levels of total thyroid hormone in circulation. If TBG is depressed, total T4 and T3 levels will go down. An increase in TBG leads to higher values of total thyroid hormone. Note however that the small percentage of T3 and T4 that remain unbound to TBG (0.05% of T4 and 0.5% of T3), the so-called free fraction, is the portion considered physiologically active. So it is quite possible to have lowered total T3 if TBG is low, but still have normal levels of free T3. This condition is not indicative of thyroidal impairment since the bioactive free T3 is normal. Similarly when TBG is elevated total T4 and T3 are high, again with the possibility that physiologically active free hormone levels remain normal.

A number of drugs and medical conditions are capable of elevating TBG and hence total T4 and T3 levels. These include estrogen, oral contraceptives (OC), pregnancy, acute infectious hepatitis, and cirrhosis. Likewise there are drugs and medical conditions that lower TBG. These include cortisol, growth hormone, and very important to this discussion, anabolic steroids. So to recap, if a person were using AAS and had their total T4 and T3 measured, because TBG is low, those total values would register as low, but that would not necessarily mean that the bioactive (free) levels of T3 and T4 are low. This observation will be critical to our discussion of the effects of AAS on the thyroid gland. In a similar vein, a woman using OC might have elevated T4 and T3 because oral contraceptives raise TBG. This would not necessarily warrant a diagnosis of hyperthyroidism, as her free thyroid levels could be perfectly normal.

So we see that in order to assess thyroid function, measuring only the total T4 and/or T3 is inadequate because these values are strongly influenced by TBG levels. Other laboratory tests are required to determine whether low T4, say, is being caused by actual hypothyroidism, or reflects the use of a drug that is simply lowering TBG. One such test is the thyroid hormone binding ratio (THBR; T3 resin uptake). This test is essentially a measure of the number of TBG sites that are occupied by thyroid hormone. In a person who is hyperthyroid (high T4) there are fewer unbound TBG sites/more occupied sites (since obviously there is more thyroid hormone available to bind to them). In this case T3 resin uptake is high. Conversely, in hypothyroidism there are fewer occupied TBG binding sites, and T3 resin uptake is low. In the case where thyroid function is normal but TBG is elevated (oral contraceptives) it turns out T3 resin uptake registers LOW; conversely when TBG is lowered (AAS use), a lab report would show T3 resin uptake reading HIGH.

We can tabulate these various possible outcomes to give a clearer picture of how these two tests can be used to distinguish thyroid dysfunction from mere altered levels of TBG:

Hyperthyroidism High High High
Hypothyroidism Low Low Low
Normal, on OC High High Low
Normal, on AAS Low Low High


Perhaps the test most commonly performed test to determine thyroid status measures Thyroid Stimulating Hormone (TSH) levels. Typically in hypothyroidism, the thyroid is not secreting adequate levels of T4, and in an attempt to stimulate the thyroid, the pituitary secretes excess TSH. So in hypothyroidism, TSH is HIGH. The opposite is observed in hyperthyroidism: the excess thyroid hormone in circulation acts back on the pituitary to suppress TSH production. TSH is LOW in hyperthyroidism.

The above applies to so-called primary hypothyroidism/hyperthyroidism where the thyroid gland itself is malfunctioning. In secondary hyper/hypothyroidism the problem lies at the levels of the pituitary or hypothalamus. In this case the pituitary secretes insufficient TSH to stimulate the thyroid, resulting in hypothyroidism with low TSH. In secondary hyperthyroidism the pituitary secretes excess TSH, resulting in hyperthyroidism associated with elevated TSH.

Before the advent of highly sensitive TSH assays, it was common to perform a TRH challenge test. Recall the hypothalamus secretes TRH, in turn stimulating the pituitary to secrete TSH. In the TRH test, a bolus injection of synthetic TRH is administered. The body’s normal response is to secrete increased levels of TSH up to a peak at 20 minutes and then to decrease TSH secretion. In hyperthyroidism, TSH is being suppressed by circulating thyroid hormones so there is a suppressed response to TRH. In primary hypothyroidism, which is due to thyroid dysfunction with normal pituitary function, levels of thyroid hormones are very low and TSH levels are ordinarily raised; however, TSH increases greatly on TRH stimulation yielding an exaggerated response - it reaches a higher peak and does not decline for over an hour. In secondary hypothyroidism (where the pituitary is malfunctioning, not the thyroid) it doesn't matter how much TRH there is, the pituitary cannot make TSH so there is an absent response to TRH stimulation.

Now that we have reviewed the elements of thyroid physiology and gone over the basic tests to determine thyroid function, we are ready to review the literature regarding the effects of anabolic steroids on the thyroid. At the beginning of this review, we asked the question “what is the origin of the claim that AAS impair the thyroid?” The answer perhaps lies in a 1993 paper by Deyssig & Weissel (1). The authors looked at the effects of self-administered AAS use in an admittedly small group of five bodybuilders. Eight additional subjects served as controls. In the AAS using group, total T4, Total T3, and TBG were depressed relative to the control group. Recall this is consistent with the widespread observation that by lowering TBG, AAS lower total T4 and total T3 with no effect on the free hormone levels and hence no effect on TSH. Indeed in this study there was no difference in free T4 and TSH between the AAS group and the controls. Basal free T3 was not measured. So far everything is consistent with normal thyroid function accompanied by AAS induced depression of TBG.

The authors next performed a TRH test. Upon administration of THR, TSH values climbed significantly higher in the AAS group, and the T3 response was significantly lower in the AAS group. Recall that in hypothyroidism there is an exaggerated TSH response to TRH. Quoting from the study, “These T3 and TSH reactions to TRH point to a mild impairment of thyroid function as a consequence of the use of anabolic steroids.” However, stressing the fact that all unstimulated parameters were consistent with the simple suppression of TBG by AAS, the authors conclude that “the results of our cross-sectional study show that high doses of androgenic-anabolic steroids, as are used by some athletes, may impair thyroid function to an extent that is not clinically detectable and probably not relevant.” (Italics added.)

Moreover, when one scrutinizes the data, one sees that out of the five AAS-using subjects, only two had stimulated TSH values higher than the controls, one was only marginally but not significantly higher than in controls, and two had stimulated TSH values in the control range. In addition, no pre-study baseline stimulated TSH values were measured in any subjects. One could argue these facts call into question the authors’ conclusions of “mild impairment of thyroid function”.

How high were the “high doses” used by the participants? The subjects were using a number of different AAS including testosterone, nandrolone, stanozolol, and Dianabol, stacking them as bodybuilders typically do. The average total dose for all drugs combined was 1.26 grams/week, with a range of 740 – 1950 mg/week.

Alen et. al. (2) conducted a study along similar lines. Seven power athletes stacking testosterone, Dianabol, stanozolol, and nandrolone were monitored for (among other things) thyroid function during a 12 week period. In this study, total T3, total T4 and TBG were all depressed during the study period, while T3 resin uptake was elevated. All of these changes are again consistent with AAS induced suppression of TBG, with no direct effect on thyroid function. Interestingly, TSH dropped during the first 8 weeks of the study, and then began to climb. Free T4 dropped marginally but stayed within the normal range. The authors interpreted the data thusly:

It is tempting to suggest that decreases in serum TBG led to decreased protein binding of the thyroid hormones, T4 and T3, which is reflected in the elevated T3U-values [T3 resin uptake]. Increased availability of T4 and T3 would then lead to a compensatory decrease in serum TSH, and this, via decreased thyroid stimulation, would further decrease total concentrations of circulating T4 and T3. The measurements of thyroid function parameters performed support this reasoning. In general our findings suggest that thyroid hormones at the cellular level were not disturbed in our athletes.

While the approximate 20% drop in free T4 observed by Alen et al is suggestive of some degree of thyroid impairment, the consequences of this need to be interpreted carefully. First, free T4 stayed well within the normal range. Second, since free T3 was not measured, we do not know if there was any change in free T3, the metabolically active hormone. Lum et al. have shown that when serum T4 levels drop, the body upregulates the peripheral conversion of T4 to T3, maintaining metabolic homeostasis (3). So it is possible that any drop in free T3 could have been significantly smaller than the observed 20% drop in free T4.

We see here contradictory findings between the two studies discussed so far as regards TSH levels: Deyssig & Weissel observed no change in unstimulated TSH levels, while Alen et al observed a decline in TSH, although the values remained well inside the normal range. The finding consistent between the two studies is the AAS suppressed TBG and the consequent decline in total T4 and T3, and an increase in T3 resin uptake observed by Alen but not measured by Deyssig. Again, quoting from (2): “In relation to the changes in thyroid function parameters measured, we suggest that the primary target of androgen action was TBG biosynthesis.”

In a third study, this time performed by Malarkey et al in AAS-using females, the authors looked at Total T4, FreeT4, TSH, and TBG (4). The authors observed that

Thyroxine-binding proteins also were decreased in the steroid users, as reflected by the low thyroxine binding index and the decrease in total serum thyroxine levels. These latter changes had no significant influence on the biological activity of thyroid hormone, however, because the free thyroxine concentration and the thyroid stimulating hormone level were within normal limits. These findings are similar to those of a previous report of decreased thyroxine-binding globulin in men who were using anabolic steroids [2].

Note here though that while TSH was “within normal limits” it was nevertheless elevated significantly compared to controls (2.5 mU/L vs 0.8 mU/L).

The difference between this study and the previous one by Alen is that in (4) free T4 was unchanged, while in (2) there was a drop in free T4. Also here TSH was elevated in the AAS users while in Alen et al it was depressed. In the current study the combination of normal free T4 but elevated TSH is suggestive of subclinical hypothyroidism. However, to truly meet the criteria required for that diagnosis TSH would have to be elevated above 5.0 mU/L (although some physicians have argued that that threshold should be lowered). Technically, these subjects would be considered euthyroid (normal).

One criticism of the studies examined thus far is that in each case the subjects used a ****tail of anabolic steroids, including ones that aromatize and others that do not. Might there be a difference in the thyroidal effects of the two classes of drugs? A study by Lovejoy et al (5) addressed that question as part of research looking at the broader differences between the metabolic effects of oral (oxandrolone) and parenteral (testosterone) steroids. Lovejoy’s group administered each drug separately to groups of subjects. Testosterone aromatizes to estrogen, and estrogen has an opposite effect on TBG from pure androgens: the former elevates TBG while the latter lowers it. If the primary effects of AAS on measured thyroid parameters result from changes in TBG, then testosterone would be expected to have only a minor effect, the increased androgen and estrogen levels tending to cancel each other’s effects. Indeed this was the case in (5). Testosterone had no significant effect on any parameter measured (T4; TSH; T3 resin uptake; or free thyroxine index, a calculated measure of free T4).

Oxandrolone on the other hand does not aromatize. The oxandrolone group showed a significant decrease in T4 and T3 resin uptake, with no change in TSH. Referring to Table 1 above, we see that the combination of low T4 and low T3 resin uptake is characteristic of hypothyroidism. This is the conclusion the authors arrived at as well, that the oxandrolone group experienced mild hypothyroidism. Again, as in the study by Deyssig & Weissel, even though T4 and T3 resin uptake were low relative to placebo the values fell within the normal range, making the diagnosis of “hypothyroidism” a relative one rather than a clinical one. Clinically all subjects would be classified as euthyroid. And as in the other studies the authors here concluded “these changes were most likely due to the effects of sex steroids on thyroid binding globulin (TBG).” The authors also observed that the Free T4 Index was higher in the oxandrolone group than either the placebo or testosterone groups.

A recent study looked at the effects of short-term methyltestosterone administration to normal subjects (6). The researchers found that total T4, total T3, and TBG were lowered, as we have come to expect. However, TSH and free T4 were elevated as well compared to the subjects’ baseline values. Again all hormone values remained within the normal range. The authors speculate that the elevated TSH and free T4 could be due to increased sensitivity of the thyroid to TRH or decreased sensitivity to hormonal feedback, suggesting some form of mild impaired thyroid function. As with the study by Deyssig, if this functional impairment were real, it would be subclinical and of dubious relevance. Note also that the elevated TSH measured by Daly et al differs from the depressed TSH observed by Alen et al. Daly et al speculate that this may be due to the fact that they sampled blood after six days vs 4 weeks and longer in the study by Alen et al.

A 1984 study by Small et al examined the effects of 10 mg daily of stanozolol, another nonaromatizing steroid, for 14 days in nine healthy subjects (7). This dosage was enough to lower testosterone by 50% and LH by 30%. TBG, T4 and T3 were lowered significantly, with no change in free T4 or TSH. This is the “standard model” of action of androgens on thyroid parameters stressed in endocrinology texts: no change in thyroid function, merely a lowering of TBG with the expected lowering of total thyroid hormone levels, but no effect on the physiologically relevant free hormone levels. To quote from the study,

“The changes found in thyroid hormones are in accord with the well known effects of anabolic steroids on thyroid function tests. Both T3 and T4 fell as a result of the reductions in TBG levels. The lack of change in TSH or Free T4 indicates that important physiological changes of thyroid function do not occur during treatment with stanozolol.”

We can summarize the results of the studies for comparative purposes by tabulating the data in Fig 2.

T4 T3 TSH T3 resin uptake Free T4 TBG
Deyssig low low 0 - 0 low
Alen low low low high low low
Lovejoy (test) 0 - 0 0 0 -
Lovejoy (ox) low - 0 low high -
Malarkey low - high - 0 low
Daly low low high - high low
Small low low 0 - 0 low

Table 2. Summary of measured thyroid parameters ( 0, no change; - unmeasured; )

Fluoxymesterone at 10 mg per day caused the by now familiar drop in TBG and total thyroid hormone levels with no effect on free parameters (8). Quoting from this study,

“Fluoxymesterone administration was accompanied by a reduction in thyroid binding globulin (with associated decreases in T3 and increases in T3 resin uptake). The free T4 index was unaltered, which implies that thyroid function was unchanged.”

Thus far we have looked at studies involving humans. One study that is often cited in the bodybuilding literature as evidence that trenbolone in particular suppresses thyroid function was done in sheep (9). However, in this study only total T4 was measured, not free T4, so we cannot conclude from this research that bioavailable T3 was affected in any way.

Can we make any sense out of the seeming hodgepodge of conflicting data? The only parameters that are consistent from study to study, where they were measured, are depressed total T3 and T4, and TBG. As we have discussed, androgens typically lower TBG, along with total T4 and T3 since the latter are a function of TBG levels. This however does not necessarily reflect thyroid dysfunction since the physiologically significant free fractions of these hormones typically remain in the normal range. If TBG levels change rapidly, however, a period of disequilibrium will exist during which thyroid function will be perturbed. This could explain the low free T4 and TSH observed by Alen as follows: The abrupt drop in TBG leads to a drop in bound T4, but free T4 remains elevated. This causes a shift in hormone from the blood to tissues because of a steeper free T4 concentration gradient. This increases the degradation rate of hormone in peripheral tissues. The increased tissue concentration of T4 signals the pituitary to lower TSH production, which will be reflected by temporarily lowered free T4 until the appropriate thyroid hormone/TBG ratio, and plasma/tissue ratio is reestablished. Alen et al discuss this possibility, and the process is illustrated graphically here:

In conclusion then AAS seem to have little if any effect on thyroid function per se. The reports by Deyssig & Weissel, and Daly et al suggest the possibility of a direct action of AAS on the thyroid or pituitary, but their results are inconsistent: The former researchers detected elevated stimulated TSH while the latter saw an increase in basal TSH. Free T4 was unchanged in former group, while it was elevated in the latter. The only consistently reported effect is a depression in total T4, total T3 and TBG. If there is a direct effect of AAS on the thyroid, pituitary, or hypothalamus the studies conducted so far shed little light on the mechanism due to their inconsistent results. And as stressed by Deyssig & Weissel any direct effect of anabolic steroids on the thyroid would likely be of no clinical significance due to its small magnitude.

From a practical standpoint for those concerned that anabolic steroids might suppress the thyroid it is a simple matter to incorporate low dose (25 to 50 mcg/day) T3 into a cycle to enhance fat loss while at the same time only minimally if it all compromising gains in muscle mass (10). In (10) one group of subjects was given T3 alone while the other was given a combination of T3 and testosterone enanthate, 200 mg/week. After 28 days of bed rest, the men in the T3 group lost an average of 3.9 kg of body weight (i.e. from 82.0 ± 7.1 to 78.1 ± 7.1 kg). Body weight in the T3 plus testosterone-treated subjects declined by only 1.0 kg (78.9 ± 4.9 to 77.9 ± 4.9 kg). Lean body mass declined by 1.5 kg in the T3 group, whereas the T3 plus testosterone-treated subjects experienced nearly a 2-kg increase in lean mass (i.e. 1.7 ± 0.9 kg). Of course we don’t know how much mass the test plus T3 group would have gained had they foregone the T3. Nevertheless these are still impressive gains considering the subjects were forced to lie in bed for 28 days with no exercise, and considering that no special dietary measures were imposed to preserve or increase muscle mass.