Newer antidepressants may be associated with an increased rate of bone loss
in older men and women, according to two studies in the June 25 Archives
of Internal Medicine.
Researchers cautioned, however, against changing or stopping treatment
because of the results, which they called "preliminary."
"We cannot say for certain that SSRIs increase the rate of bone
loss," said one of the studies' authors, Susan Diem, M.D., an assistant
professor of medicine at the University of Minnesota. "It could be
depression itself that accelerates the bone loss," she told
Diem and her colleagues speculated that a possible mechanism for the
association between SSRIs and a drop in bone density could be attributed to
the role of serotonin transporters in osteoblast production.
In one of the two studies, Diem and her colleagues analyzed data on a group
of women recruited from 1986 to 1989 from population-based listings in
Maryland, Minnesota, Oregon, and Pennsylvania to take part in the Study of
The findings came from a cohort of 2,722 women over age 65 from the larger
sample who returned for two visits—one in 1997 or 1998 and the other
from 2002 to 2004, during which they completed a medication inventory and the
Geriatric Depression Scale.
In addition, researchers completed bone mineral density scans of the hip
and two femoral subregions at both visits. The average age of women in the
study was about 78.
Because depressive symptoms have been associated with lower bone mineral
density in previous studies, Diem said, researchers omitted women who had a
high score on the depression scale at one of the two visits from 2002 to
Diem found that the 198 women in her sample who reported using SSRIs from
2002 to 2004 had 1.6 times the rate of bone loss than those who had not used
the medications, even after controlling for age, race, smoking,
calcium-supplement use, and health status, among other factors.
In women who did not report SSRI use, bone density was lost at the rate of
about 0.5 percent per year. In those who reported SSRI use during one of the
two periods during which they had bone scans, they lost bone at a rate of
about 0.8 percent per year, a statistically significant difference.
Diem also called it a "clinically significant" change because
the higher rate of bone loss may translate to an increased risk in
A total of 118 women in her sample reported taking tricyclic
antidepressants, and they had the same rates of bone loss as non-SSRI
She noted that in laboratory studies serotonin has been shown to encourage
proliferation of osteoblasts, cells that are responsible for bone formation.
Since SSRIs have a net effect of blocking serotonin, they theoretically may
hamper the action of osteoblasts.
The study had its limitations, Diem acknowledged.
For instance, Diem and her colleagues had no information about the dosage
or length of time participants took SSRIs before they were evaluated. In
addition, the number of participants who reported taking SSRIs during both of
the two visits was relatively small (198), which could affect the strength of
Diem said future research needs to replicate the findings in larger
samples, gather information on a prospective basis, and collect more detailed
information on SSRI use.
For the present, she said, "I would not recommend that patients
discontinue antidepressant treatment based on these results, which are
The second study examined the association between SSRI use in men and
whether that use was associated with bone loss.
Elizabeth Haney, M.D., and colleagues analyzed findings from the
Osteoporotic Fractures in Men Study, in which almost 6,000 men aged 65 and
older were recruited between 2000 and 2002 in the Birmingham, Ala.;
Minneapolis; Pittsburgh; Palo Alto, Calif.; Portland, Ore.; and San Diego
Haney is assistant professor of medicine at the Oregon Health and Science
University in Portland.
Interviewers met with study participants at their baseline visit to
determine what prescription medications they had used in the prior month and
how often. They also gathered information on participants' age, race,
ethnicity, level of physical activity, medical conditions, and alcohol and
tobacco use. Men in the study had an average age of about 74.
In addition, clinicians used imaging techniques to measure bone-mineral
density along two regions of the hip and lumbar spine.
A total of 160 men reported using SSRIs and 99 used tricyclic
Haney found that total bone-mineral density in the hip was 3.9 percent
lower among those who reported using SSRIs than those who didn't. Total spine
bone-mineral density was 5.9 percent lower among SSRI users. In Haney's study
as well as in Diem's, subjects who reported using tricyclic antidepressants
had no elevated rates of bone loss.
The authors acknowledged that symptoms of depression or lifestyle changes
brought on by these symptoms could play a role in the reduction in bone
In an editorial accompanying the two studies, Kenneth Saag, M.D., M. Sc.,
an associate professor of medicine at the University of Alabama at Birmingham,
noted that "depressed persons may eschew exercise, healthy eating
(leading to lower calcium intake and weight loss), and sun
Saag also said that even while taking into account the studies'
limitations, "the magnitude of the associations partially support a
causal relationship" between SSRI use and bone loss and suggest that
SSRI use be added to the list of risk factors "that prompt clinicians to
more carefully consider bone health."
He also suggested that people receiving SSRIs and who have other risk
factors for osteoporosis have bone mineral density measurements done as a
matter of course and noted that "good clinical acumen and thoughtful
adverse-event monitoring can help avoid having healthier minds at the expense
of sicker bones."
An abstract of "Use of Antidepressants and Rates of Hip Bone
Loss in Older Women: The Study of Osteoporotic Fractures is posted at<archinte.ama-assn.org/cgi/content/abstract/167/12/1240>;
an abstract of "Association of Low Bone Mineral Density With Selective
Serotonin Reuptake Inhibitor Use by Older Men" is posted at<archinte.ama-assn.org/cgi/content/abstract/167/12/1246>.▪