I asked several colleagues whether they had added bone mass density (BMD) loss to the pantheon of practice-altering medication risks we have learned to consider when prescribing antidepressants, or even to the list of disturbing and more or less common experiences (eg, sexual dysfunction, sweating, nightmares, myoclonus, apathy, discontinuation syndrome, hyponatremia, bradycardia/hypotension) that taint SSRI benefits for some patients. My informal inquiry and review of the literature suggested that a column on this topic would be timely.
Osteoporosis—the development of bone porosity—results from the deterioration of bone tissue that predisposes to low-trauma fractures, especially of the hip, spine, or wrists. Early BMD loss and osteoporosis are asymptomatic and often develop over a number of years, reflecting the convergence of multiple disruptive influences on the process of bone remodeling.1 Without directed attempts to diagnose osteoporosis in asymptomatic persons at risk for BMD loss, clinical recognition may be delayed until the occurrence of backache, height loss, or a fracture associated with minimal physical trauma (such as a fall from standing height).2 Such “fragility fractures” added an estimated 5.8 million disability-adjusted life-years to the global health burden in 2000.2
Background on risk
Osteoporosis is present in 15% of whites in their 50s and in 70% of those older than 80 years.3 Although 80% of persons with osteoporosis are women, men affected by hypogonadism or a host of other disorders are vulnerable as well. Osteoporosis occurs at the highest rates among non-Hispanic white and Asian women aged 50 or older, but no ethnic group is unaffected.4
In certain at-risk subpopulations, bone loss is accelerated. Various sources, including the informative Web site of the National Osteoporosis Foundation (www.nof.org) provide lists of modifiable and nonmodifiable risk factors (Table 1). Among the prescribed medications that compound modifiable risk (Table 2), the glucocorticoids and antiepileptic drugs are most widely known. However, 2 additional medication classes that figure prominently in psychiatric practice have also been reported to increase the risk of osteoporosis: the prolactin-increasing antipsychotics5 and the SSRIs.6-9 The SSRIs, about which recent data are available, will be the focus here.
SSRI effects on BMD
A need for examination of the effects of SSRIs on bone integrity was suggested as early as 1998 in a case-control study that linked use of SSRIs or tricyclic antidepressants (TCAs) with increased risk for hip fractures in elderly persons.9 Concern flared again last year with the publication of several important studies. Haney and colleagues6 reported the results of a cross-sectional analysis of data from a prospective observational study of 5995 men, 65 years and older, of whom 160 were SSRI users. Compared with men who were taking no antidepressant, the SSRI users (but not users of trazodone or TCAs) had significantly lower BMD in the hip and lumbar spine. The magnitude of the association appeared to be similar to that linked with glucocorticoid use in this cohort, although the meaning of this finding was unclear because detailed data about the amount and route of glucocorticoid use were too limited. After considering the roles of several potential confounding factors, these authors concluded that the association between low BMD and SSRI use was not explained by the presence of depressed mood, as measured by the score from the 12-item Short Form of the Medical Outcomes Study, Mental Component Scale.
A prospective study compared serial measurements of BMD in a cohort of 2722 women 65 years and older who were taking an SSRI or a TCA.7 The mean interval between the BMD measurement visits was 4.9 years. Depression was measured with the Geriatric Depression Scale (GDS), and the use of concurrent medications and calcium supplements was recorded. The SSRI users, but again not TCA users, showed a higher rate of bone loss in total hip measurements compared with nonusers. Because the SSRI-user cohort included more patients with a GDS score of at least 6 than the TCA-user cohort, the results were re-analyzed; patients whose GDS score was at least 6 during one of the BMD visits were excluded. This exclusion decreased the magnitude but did not eliminate the significance of the greater decrease in BMD observed in SSRI users.
