Antidepressants: risk of fractures

Healthcare professionals should be aware of epidemiological data showing a small increased risk of fractures associated with the use of TCAs and SSRIs, and should take this risk into account in their discussions with patients and in prescribing decisions

Article date: May 2010

A review of epidemiological studies, mainly in patients age 50 years or older, shows an increased risk of bone fractures in patients receiving SSRIs (selective serotonin reuptake inhibitors) and TCAs (tricyclic antidepressants). The mechanism leading to this increased risk is unclear.

Nine reviewed observational studies 1 2 3 4 5 6 7 8 9 give a range of odds ratios for fractures associated with current use of SSRIs (irrespective of dose or duration of use) between 1·4 (95% CI 0·93–2·24) and 2·4 (2·0–2·7). These risks were significant in most studies. The odds ratios for studies that included TCAs varied between 1·2 (0·7–2·2) and 2·2 (1·8–2·8), and in all but one study were lower than the estimates for SSRIs.

Six of the nine studies assessed dose response and observed a trend more consistently for SSRIs compared with TCAs. Six of the nine studies assessed duration of use and found that the risk of fracture associated with SSRIs seems to increase initially to a peak within the first 6–12 months; risk subsequently decreases, but remains elevated with prolonged use (>1·5 years). Risk with TCAs peaks shortly after initiation (1–2 months) and disappears after prolonged use (>6–12 months). The persistence of the effect after cessation of use was assessed in four of the nine studies. The increased risk observed for SSRIs and TCAs disappears relatively shortly after discontinuation (3–12 months). None of the studies investigated the effects of dose and duration simultaneously.

Risk of falls

The relationship between SSRIs and fall risk is not clear.10Several studies show that SSRI use is associated with an increased fall risk, and that this risk is higher than for other types of antidepressants, including mainly TCAs. 11 12 13 14 However, other studies show that SSRIs confer equivalent or lower fall risk to TCAs. 15 An Italian study of patients receiving home care showed no increased risk of falls with any antidepressant. 16

SSRIs and decreased bone mineral density

Some studies17 18 19 20 have found an association between antidepressant use and decreased bone mineral density; however, other studies have found no such association. 21 22

Conclusions of the review

Taking into account the strengths and limitations of the available evidence, the review concluded that product information should be updated with a statement on epidemiological findings of an increased risk of bone fractures with TCAs and SSRIs.

From the available data, no definite conclusions could be drawn regarding a dose-response relation, time relation, or the underlying mechanism.

Advice for healthcare professionals

A review of epidemiological studies, mainly in patients age 50 years or older, shows an increased risk of bone fractures in patients receiving SSRIs and TCAs. The mechanism leading to this increased risk is unclear.

Healthcare professionals should be aware of a small increased risk of fractures associated with the use of TCAs and SSRIs, and should take this risk into account in their discussions with patients and in prescribing decisions.

For further information see the monthly report from the EU Pharmacovigilance Working Party

Article citation: Drug Safety Update May 2010, vol 3 issue10: 3.

  1. Liu B, et al. Lancet 1998; 351:1303–07.

  2. Hubbard R, et al. Am J Epidemiol2003; 158: 77–84.

  3. French DD, et al. Drugs Aging 2005;22: 877–85.

  4. Vestergaard P, et al. Calcif Tissue Int 2008; 82: 92–101.

  5. Bolton JM, et al. J Clin Psychopharmacol 2008; 28: 384–90.

  6. Ensrud KE, et al. Arch Intern Med 2003; 163: 949–57.

  7. Richards JB, et al. Arch Intern Med 2007; 167: 188–94.

  8. Ziere G, et al. J Clin Psychopharmacol 2008; 28: 411–17.

  9. van den Brand MW, et al. Osteoporos Int 2009; 20: 1705–13.

  10. Haney EM, et al. Bone 2009. Published online Aug 5, 2009. doi.org/10.1016/j.bone.2009.07.083.

  11. Yadav YK, et al. Cell 2008; 135: 825–37.

  12. Kerse N, et al. PloS One 2008; 3: e2423.

  13. Sterke CS, et al. Int Psychogeriatr 2008; 20: 890–910.

  14. Hartikainen S, et al. J Gerontol Med Sci 2007; 62A: 1172–81.

  15. Thapa PB, et al. N Engl J Med 1998; 339: 875–82.

  16. Landi F, et al. J Gerontol A Biol Sci Med Sci 2005; 60: 622–26.

  17. Williams LJ, et al. Int Clin Pscyhopharmacol 2008; 23: 84–87.

  18. Mezuk B, et al. J Gerontol A Biol Sci Med Sci 2008; 63: 1410–15.

  19. Haney EM, et al. Arch Intern Med 2007; 167: 1246–51.

  20. Diem SJ, et al. Arch Intern Med 2007; 167: 1240–45.

  21. Kinjo M, et al. Am J Med 2005; 118: 1414.e7–e12.

  22. Spangler L, et al. J Gen Intern Med 2008; 23: 567–74.

Help us improve GOV.UK

Don’t include personal or financial information, eg your National Insurance number or credit card details.