Two studies have concluded that use of SSRI antidepressive drugs may be associated with significant reductions in hip and spine bone mass density (BMD).
The studies reported, in the Archives of Internal Medicine, raise concerns that the SSRI antidepressive drugs may have significant adverse effects on bone.
Recent research has found that there are serotonin transporters in bone cells, and animal studies suggest disruption of this system has adverse effects on bone density. There is therefore the possibility that use of drugs affecting serotonin transport may affect bone mineral density (BMD) in humans. The two reported studies investigate this using data from two related existing prospective cohort studies.
The studies, funded through the US National Institutes of Health, are the Study of Osteoporotic Fractures (SOF, women) and the Osteoporotic Fractures in Men Study (MrOS). The first has been running for nearly two decades, the second for more than five years.
The authors conclude that in their cohort, use of SSRI was associated with faster bone loss compared to non-use, whereas use of TCA was not. They suggest this may be due to a direct adverse effect of SSRI on bone.
They note that depression itself is associated with reduction in BMD, so is a confounding factor: when those with the highest depression scores were excluded from analysis the association was reduced but not eliminated, suggesting at least some confounding by indication. Nevertheless, the authors consider that their results indicate the need for further research.
The second study reported aimed to determine whether SSRI use was associated with effects on BMD in older men. Previous analysis of data from MrOS suggested an association, and this cross-sectional analysis aimed to clarify that.
Analysis including a wide range of potentially relevant variables in the model made no significant difference to the association. The authors conclude that SSRI use in this cohort was associated with significant reductions in hip and spine BMD compared to non-users.
They note that this was not affected by potential confounding variables, and consider that it represents a real and clinically significant effect. They comment that it is similar in magnitude to the effect of corticosteroids reported in this population, and say further studies are needed to confirm their results.
Arch Intern Med 2007;167:1240-51;1231-2