Levothyroxine dose and fractures in older adults

BMJ 2011; 342 doi: http://dx.doi.org/10.1136/bmj.d2250 (Published 28 April 2011) Cite this as: BMJ 2011;342:d2250
  1. Graham P Leese, professor of endocrinology and diabetes,
  2. Robert V Flynn, research pharmacist
  1. 1Ninewells Hospital and Medical School, Dundee DD1 9SY, UK
  1. grahamleese{at}nhs.net

Risk may be higher in certain subgroups

Hyperthyroidism is a risk factor for osteoporosis and fractures, so are patients who take long term thyroxine replacement at increased risk of fractures? Some studies have shown an increased risk of fracture for patients with low concentrations of thyroid stimulating hormone (TSH) who take long term thyroxine,1 2 but others have not.3 4 This discrepancy may be due to confounding as a result of patient selection, the background fracture risk in the population, the power of the studies, and inclusion of patients with a history of hyperthyroidism who are on long term thyroxine replacement.

A recent cohort study included 17 684 patients with primary hypothyroidism and serial TSH estimations rather than baseline measurements only.5 It found a statistically significant increased risk of fractures in patients of all ages taking long term thyroxine who had suppressed serum TSH concentrations (<0.03 mU/L) but not in patients with a low TSH (0.04-0.4 mU/L) or TSH in the reference range.

In the linked nested case-control study of 213 511 patients over the age of 70 years who were taking thyroxine for primary hypothyroidism (doi:10.1136/bmj.d2238), Turner and colleagues report that current use of thyroxine was associated with a significantly increased risk of fracture (adjusted odds ratio 1.88, 95% confidence interval 1.71 to 2.05) compared with use in the remote past.6 The risk of fracture was closely associated with the dose of thyroxine. Even patients taking “medium” doses (mean dose of 76 µg daily) were at higher risk than those taking “low” doses (46 µg daily). Many patients over 70 are on even higher doses. However, the choice of dose in clinical practice is decided on the basis of serum TSH concentration and clinical symptoms, not by age. Turner and colleagues do not provide data on TSH concentrations, but higher thyroxine doses are probably related to lower, and possibly suppressed, serum TSH concentrations.

How should these data be interpreted for routine clinical practice? Certain subgroups of patients on long term thyroxine are likely to have a small increase in the risk of fracture. Such groups might include those with a relatively high background risk of fracture, such as elderly people and postmenopausal women, who are also on thyroxine doses that suppress their serum TSH concentration. Some previous studies have shown a trend towards an increased risk of fracture in postmenopausal women,7 8 and the current study shows a clear association between the thyroxine dose and fracture in elderly people. Age is a risk factor for fracture in the general population, so the relatively small additional effects of thyroxine become more evident.

The interaction between age and TSH concentration may be crucial in this context. The distribution of serum TSH concentration in the general population is skewed to the right. One possible reason is because it reflects undiagnosed hypothyroidism. Patients with a TSH of 2-4 mU/L are more likely to develop hypothyroidism during 20 years of follow-up.9 However, most people with a TSH value in this range do not progress to hypothyroidism and do not have subclinical thyroid disease. The right sided skew may also reflect different subpopulations with different “normal” TSH distributions—for example, elderly people and different ethnic groups.10 The 97.5th centile for the distribution of serum TSH concentration in patients 80 years or older in the United States is 7.49 mU/L, not 4 mU/L as seen in younger people.10 Thus, in the US 14.5% of people over 80 without apparent thyroid disease have a TSH greater than 4.5 mU/L.10 This may be because many elderly patients lack clear symptoms, or it may reflect a different “normal” range in elderly people.

What is the practical implication of this? Current evidence suggests that elderly people need relatively low thyroxine doses, so serum TSH should be regularly monitored and a suppressed TSH should be avoided in such patients. Further work is needed to see whether current TSH reference ranges (usually 0.4-4.0 mU/L) are appropriate for use in elderly people. If these ranges are not appropriate, this could theoretically exacerbate the risk of overtreatment in elderly people, with increased risk of bone loss and fractures in this high risk group. Also, a greater number of elderly patients may be started on thyroxine replacement than is warranted. This concern could be studied by assessing changes in bone mineral density and bone markers at different thyroxine doses and different TSH concentrations in elderly patients.

The increased risk of fracture for patients taking long term thyroxine is small, with the main risk likely to be in elderly patients with a suppressed serum TSH concentration. Current guidelines of aiming for a TSH value within the reference range should be adhered to.11 Ideal thyroxine doses may vary with age and be unexpectedly low in elderly people. It is 120 years since the effect of excess thyroid hormone on bone was first described, yet research in this area still lacks funding. With the prevalence of treated hypothyroidism increasing,12 and the annual economic burden of fractures in the United Kingdom currently estimated at €5.8bn (£5.1bn; $8.4bn), such research warrants a higher priority.


Cite this as: BMJ 2011;342:d2250


  • Research, doi:10.1136/bmj.d2238
  • Competing interests: All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.