Purpose: The impact of long-term L-thyroxine replacement therapy on skeletal integrity is a growing concern because of the large number of women receiving thyroid hormone therapy. The purpose of this study was to examine the hypothesis that long-term L-thyroxine therapy in which the free thyroxine index (FT4I) is maintained within a physiologic range has minimal impact on vertebral or femoral bone mineral density in both premenopausal and postmenopausal women.
Patients and methods: We measured hip integral and spinal trabecular and integral bone densities in 28 premenopausal and 28 postmenopausal women who had been receiving L-thyroxine therapy for a median of 12 and 15 years, respectively, and in whom therapy was titrated to keep the FT4I within the normal range. The relationship between bone density parameters and thyroid hormone status was examined using univariate and multivariate statistical methods.
Results: Seventy-nine percent of the premenopausal women and 86% of the postmenopausal women had FT4I values in the normal range at the time of bone density determination. Moreover, throughout the study's duration, the majority of annually measured values were in the normal range for more than 80% of subjects. Premenopausal women had slightly lower bone density than would be expected for age: -6.7% (z = -0.39 +/- 0.74 [mean +/- SD], p less than 0.01), -3.1% (z = -0.22 +/- 0.78, p = 0.15), and -5.1% (z = -0.36 +/- 0.74, p less than 0.02) for spinal trabecular, spinal integral, and hip integral bone density, respectively. Postmenopausal women likewise had slightly lower bone density values that were significant only at the hip: -0.2% (z = -0.01 +/- 1.01, p = 0.95), -1.0% (z = -0.05 +/- 1.11, p = 0.80), and -6.2% (z = -0.39 +/- 0.80, p less than 0.02) for spinal trabecular, spinal integral, and hip integral bone density, respectively. When patients with previously treated Graves' disease (n = 4 in each group) were eliminated, the differences in bone density at the hip were no longer seen. Correlation analysis revealed only weak and generally nonsignificant relationships between parameters of thyroid hormone status and bone density at any site in either subgroup. Results of multiple regression analysis among the pooled data of all subjects showed that age provided a consistently significant contribution (R2 = 0.18 to 0.66) to the variability in bone density at the spine and the hip, but parameters of thyroid hormone status did not.
Conclusion: These data provide the first supportive evidence that long-term L-thyroxine therapy that maintains the FT4I in the physiologic range is associated with a statistically significant, but clinically minimal, decrement in spinal and hip bone density in both premenopausal and postmenopausal women. The decrement at the hip was entirely due to the inclusion of patients with treated Graves' diseases. Thus, the changes in bone density in women receiving long-term L-thyroxine therapy are minimal at most and should not be a contraindication to therapy.