Is There a Better Way to Estimate Levothyroxine Doses for Hypothyroid Patients?

Primary hypothyroidism is seen routinely in clinical practice, and levothyroxine (LT4) is one of the most commonly prescribed medications in the United States. LT4 doses required to normalize serum thyrotropin (TSH) levels depend on a number of factors, including weight, age, sex, and amount of residual functioning thyroid tissue. ¹ Various formulas attempt to predict optimal initial LT4 doses in athyrotic patients, but these are usually simplified to weight-based algorithms for ease of use. Thus, current recommendations state that athyrotic patients require LT4 doses of 1.6–1.8 mcg/kg body weight/day. ² Starting doses are further tailored based on patient characteristics and then adjusted based on follow-up TSH levels until a normal TSH is achieved.

This approach works well for many patients but is based on studies in younger patients with no functional thyroid tissue, and it can cause overtreatment in other populations, leading to adverse health outcomes. In particular, women and older patients have higher rates of iatrogenic thyrotoxicosis compared with men and younger patients at the same average body weight-based LT4 doses. ^3,4 Weight-based LT4 dosing schemes correctly predict only 51% of euthyroid LT4 doses. ⁵ Attempts to refine dosing algorithms using multiple clinical variables and complex formulas result in lower rates of thyrotoxicosis in clinical simulations, but rates remain above 25%. ^5,6

Studies have attempted to improve these rates by utilizing two alternatives to body weight that may be more closely associated with optimal LT4 dosing: ideal body mass (IBM) or lean body mass (LBM). IBM is based on actuarial data as the weight associated with the greatest life expectancy; equations predict IBM as a linear function of height. ⁷ LBM is the difference between total body mass and body fat mass and includes muscle, bone, skin, and internal organs. LBM can be measured by dual-energy X-ray absorptiometry (DXA) or other radiographical techniques or calculated by formulas validated against DXA measurements. ⁸ Most IBM and LBM formulas incorporate height, weight, sex (or have separate formulas for women and men), and sometimes age. They illustrate differences in body composition among people with similar weights that may affect drug dosing; for example, 1.6 mcg/kg actual body weight translates to 2.3 mcg/kg LBM for an average man and 2.8 mcg/kg LBM for an average woman.

The study by Adams and Mammen in the current issue of Thyroid ⁹ investigated whether higher rates of iatrogenic thyrotoxicosis in older LT4-treated hypothyroid women (compared to men) could be mitigated by using IBM or LBM rather than actual body weight. They conducted a retrospective cohort study of over 20,000 LT4-treated patients (average age 65 years, 77% women) followed for a mean of 4.7 years in the Johns Hopkins Health System electronic health record. They utilized validated formulas for IBM and LBM, corrected for multiple clinical variables, and conducted sensitivity analyses in patients with at least one TSH level above 10 mU/L (to exclude possible mild hypothyroidism) or patients with at least one TSH level below 0.1 mU/L (to focus on patients most likely to experience adverse effects).

Mean body weight was 79 kg in women and 93 kg in men, while LBM was 71% in women and 58% in men. Daily LT4 doses were 1.2 mcg/kg in women and 1.1 mcg/kg in men (9% higher in women) based on actual body weight, 1.7 mcg/kg in women and 1.4 mcg/kg in men based on IBM (21% higher in women), and 2.2 mcg/kg in women and 1.6 mcg/kg in men based on LBM (38% higher in women).

Overall, 34% of the patients had at least one low TSH level during the study period, with 14% below 0.1 mU/L; 25% had at least one TSH level above 10 mU/L. Thirty-seven percent of the women had one or more low TSH levels, compared to 24% of the men (adjusted hazard ratio [HR] = 1.50). Adjusting for IBM rather than actual weight attenuated this HR to 1.30, while adjusting for LBM eliminated the excess risk in women entirely (HR = 1.06). Results did not differ in the subgroup with TSH <0.1 mU/L but were attenuated in the subgroup with TSH >10 mU/L. Thyrotoxicosis was more common with higher LT4 doses, but the gap between women and men persisted at all doses and was fully mediated by body composition.

A particular strength of this study was the use of real-world data in a large ambulatory population. Data extraction was carefully checked, and numerous comorbidities that likely affect LT4 dosing were accounted for. Sub-analyses addressed relevant questions over a range of TSH levels and LT4 doses. The study only included older patients, but this decision is understandable, since older patients may be more likely to suffer ill effects of LT4 overtreatment. Other minor limitations are acknowledged in the article, but overall, this was a carefully conducted and comprehensive study.

How to translate these findings into clinical practice? The authors did not provide a dose calculator or formula for estimating LT4 doses based on LBM, which would be ideal. In the meantime, they recommended using an online tool to calculate LBM, of which several are available. Alternatively, the formula used by the authors to calculate LBM is complex but well validated against DXA-derived LBM measurements ⁸ and could be incorporated into an automated calculation in the electronic health record. Once LBM is derived, the authors recommend starting an LT4 dose (for an athyrotic patient) at 1.6 mcg/kg LBM/day. They do not explicitly explain this recommendation, but examination of Figure 3 in the article provides some insights. At 1.6 mcg/kg LBM/day, 4.4% of TSH levels would have been low in their cohort, compared to 34% with weight-based dosing. Figure 3 also shows a stable rate of high TSH levels between 0.3 and 2.0 mcg/kg LBM/day, providing reassurance that this approach would not increase undertreatment rates.

This publication provides valuable information that will allow care providers to more accurately estimate starting LT4 doses in hypothyroid patients, likely reducing adverse effects of overtreatment. The authors are to be commended for their contribution to this important clinical question.