Levothyroxine brand and weight-adjusted dose requirements in primary hypothyroidism: real-world evidence of dose differences without pharmacodynamic variation

Abstract

Levothyroxine (LT4) is commonly considered interchangeable across brands, yet small formulation differences may translate into different dose requirements in routine care. We compared three LT4 brands in adults with biochemically controlled hypothyroidism, focusing on weight-adjusted dose needed to maintain normal TSH and pharmacodynamic response. Materials and methods: observational cross-sectional study using a private outpatient database. Adults with primary hypothyroidism, TSH within the reference range, and etiology limited to Hashimoto’s thyroiditis or post-thyroidectomy were included. Exposure was LT4 brand (A, B, C). Primary outcome was LT4 dose per kg (µg/kg/day). Pharmacodynamic outcome was log-transformed TSH modeled as a function of brand, dose per kg, age, sex, and etiology. Robust standard errors and propensity score-based inverse probability of treatment weighting (IPTW) were used to address confounding by non-random brand assignment. Brand×etiology interaction was explored. Results: we included 274 adults (A n=102, B n=152, C n=20). In adjusted models, brand was associated with dose per kg; Brand A required a higher dose than Brand B (adjusted difference ≈0.14 µg/kg/day), while comparisons involving Brand C were imprecise and non-significant. Dose per kg decreased with age and was higher in post-thyroidectomy. In pharmacodynamic models, brand was not associated with log(TSH) after adjustment (global p=0.58), and results were consistent in IPTW-weighted analyses (global p=0.68). No significant brand×etiology interaction was detected. Discussion: LT4 brand was associated with small differences in weight-adjusted dose requirements, but not with adjusted TSH response. These observational findings should be interpreted cautiously and support biochemical monitoring when treatment changes occur.