CNS Spectr. 2010;15(9):579-586
Dr. Misra is assistant professor of pediatrics, Ms. Le Clair is a student, Ms. Mendes is study coordinator, Dr. Miller is associate professor of medicine, Dr. Lawson is instructor in medicine, Ms. Meenaghan is research associate, Dr. Weigel is instructor in psychiatry, Dr. Herzog is endowed professor of psychiatry in the field of eating disorders, and Dr. Klibanski is Laurie Carrol Guthart Professor of Medicine at Harvard Medical School in Boston. Dr. Ebrahimi is research associate at the Cambridge Eating Disorders Center in Massachusetts.
Faculty Disclosures: Dr. Misra has been a consultant to Ipsen. Ms. Le Clair, Ms. Mendes, Dr. Miller, Dr. Lawson, Ms. Meenaghan, Dr. Weigel, Dr. Ebrahimi, Dr. Herzog, and Dr. Klibanski report no affiliation with or financial interest in any organization that may pose a conflict of interest.
Funding/Support: Funding for this study was provided in part by National Institutes of Health grants 1 UL1 RR025758-01, DK 062249 and K23 RR018851. The study was performed in the Neuroendocrine Unit of Massachusetts General Hospital.
Submitted for publication: December 11, 2009; Accepted for publication: April 8, 2010.
Please direct all correspondence to: Madhusmita Misra, MD, MPH, Bulfinch 457, Neuroendocrine Unit, Massachusetts General Hospital, 55 Fruit St, Boston, MA 02114; Tel: 617-726-3870, Fax: 617-726-5072; E-mail: email@example.com.
Objectives: Alterations in serotonin impact bone metabolism in animal models, and selective serotonin reuptake inhibitors (SSRIs) have been associated with increased fracture risk in older adults. SSRIs are commonly used in anorexia nervosa (AN), a condition that predisposes to low bone mineral density (BMD). Our objective was to determine whether SSRI use is associated with low BMD in AN.
Methods: We examined Z-scores for spine, hip, and whole body (WB) BMD, spine bone mineral apparent density, and WB bone mineral content/height (BMC/Ht) in females 12–21 years of age with AN who had never been on SSRIs, had been on SSRIs for <6 months (<6M), or had been on SSRIs for >6 months (>6M).
Results: Subjects on SSRIs for >6M had lower spine, femoral-neck, and WBBMD Z-scores than those on SSRIs for <6M. Hip BMD and WBBMC/Ht Z-scores were lowest in subjects on SSRIs for >6M. Duration of SSRI use, duration since AN diagnosis and duration of amenorrhea inversely predicted BMD, whereas BMI was a positive predictor. In a regression model, duration of SSRI use remained an independent negative predictor of BMD.
Discussion: Duration of SSRI use >6M is associated with low BMD in AN.
Conclusion: It may be necessary to monitor BMD more rigorously when duration of SSRI use exceeds 6M.
• Use of selective serotonin reuptake inhibitors (SSRIs) for >6 months is associated with low bone density at multiple sites in anorexia nervosa.
• A longer duration of SSRI use, a longer duration of anorexia nervosa, and low weight may be deleterious to bone health.
• Prolonged SSRI use affects bone health independent of duration of anorexia nervosa and low weight.
• Clinicians should consider monitoring bone density closely in women who have been on SSRIs for >6 months.
Anorexia nervosa (AN), an eating disorder characterized by severe undernutrition, is known to have a significant negative impact on bone health,1-5 which has been attributed to the associated hypogonadism,6 low insulin-like growth factor-1 levels,1,7-9 relative hypercortisolemia,10,11 alterations in other hormones such as adiponectin, leptin, and peptide YY,12-14 and changes in body mass index1,3,15,16 and lean body mass.17-19 However, these hormonal and body composition alterations do not account for all the variability in bone mineral density (BMD) in AN, and other determinants of low BMD remain to be determined.
While it is well known that hormones such as leptin and PYY work through neural systems to impact bone metabolism,20 there is an emerging interest in the neurotransmitter serotonin (5-hydroxytryptamine [5-HT]), because of recent evidence of the presence of a functional serotonergic pathway in bone cells.21-23 Serotonin is a monoamine neurotransmitter that exerts effects on the central nervous system (CNS), gastrointestinal (GI) tract, and cardiovascular (CV) system.24 Inhibition of the serotonin transporter (5-HTT) with selective serotonin reuptake inhibitors (SSRIs) in mice has been associated with reduced bone accrual during growth, and mice expressing a null mutation in the gene encoding for the 5-HTT have evidence of decreased bone mineral accrual.25,26 Furthermore, a chronic deficiency of tryptophan, an amino acid precursor to serotonin, impairs bone metabolism and accrual in rats.27
Manipulations of serotonin levels in the CNS affect behavior and cognition,19 and serotonergic pathways have been effectively targeted for pharmacologic therapies. The SSRIs represent a class of widely prescribed antidepressants that are commonly used in populations at risk of compromised bone health, including patients suffering from eating disorders1-4,7 and major depressive disorders.28 Therefore, assessment of the impact of SSRI use on bone metabolism in these populations is important. Recent studies report increased fracture risk and lower bone density in older men and women on SSRIs for depressive disorders.29-31 However, data are lacking in adolescents and young women taking these medications.
Importantly, the adolescent years are a common time for the onset of eating disorders, and SSRI use is very prevalent in adolescents suffering from AN.5 The pubertal years are also very critical for bone accrual, and represent a very narrow window in time in which to optimize gains in bone mass towards attainment of peak bone mass, an important determinant of bone health and fracture risk in later life. Therefore, it is especially important to consider the effect of SSRI use on bone health in the adolescent population.
No studies to date have examined the effect of SSRIs, which potentially cause pharmacologic alterations in serotonin levels, on bone in adolescents and young women. In this study, we examined the potential impact of SSRI use on BMD in girls and young women with AN 12–21 years of age, and we show that duration of SSRI use is a significant and independent predictor of low BMD in AN.
Subjects and Methods
Subjects included 155 girls and young women meeting Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition criteria for diagnosis of AN32 between 12 and 21 years of age. Diagnosis of AN was confirmed by a clinical interview with the study psychiatrist. These women were recruited through referrals from primary care providers, nutritionists, psychiatrists, and therapists, and also from day-treatment eating disorder programs in and around Boston. Informed consent was obtained from all subjects. For subjects <18 years of age, consent was obtained from parents and assent from the subjects. Our Institutional Review Board approved the study.
We obtained detailed information regarding the type of SSRI used, duration of use, and medication dosage. Subjects were divided into three groups based on the duration of SSRI use, resulting in 95 subjects who had never taken an SSRI at the time of data collection (SSRI-0), 38 subjects who had been on an SSRI for <6 months at the time of data collection (SSRI<6M), and 22 subjects who had been taking an SSRI for >6 months at the time of data collection (SSRI>6M). Medications recorded in subjects’ histories and classified as SSRIs were fluoxetine, fluoxetine with olanzapine, paroxetine, duloxetine hydrochloride, sertraline hydrochloride, citalopram, escitalopram, venlafaxine, and fluvoxamine maleate.
All subjects were evaluated during a single outpatient visit to our Clinical Research Center. A complete history and physical examination was performed, and blood drawn for follicle stimulating hormone and thyroid stimulating hormone levels to ensure that amenorrhea was not a consequence of thyroid dysfunction or premature ovarian failure. A single Harpenden stadiometer was used to measure heights of subjects, using an average of three measurements. Weight was measured on a single electronic scale, with the subject in a hospital gown. Body mass index (BMI) was calculated using the formula [weight (in kilograms)]/[height (in meters)2]. Bone age (obtained for subjects ≤18 years of age) was assessed using the methods described by Greulich and Pyle33 by a single investigator (a pediatric endocrinologist) to reduce interobserver variation. Nutritionists at the Clinical Research Center obtained exercise history from subjects using the Modified Activity Questionnaire, and information regarding calcium and vitamin D intake using a Calcium Food Frequency Questionnaire.
Measurements of lumbar spine (L1–4), hip, and whole body (WB) BMD were obtained using dual-energy x-ray absorptiometry (DXA) (4500A fan-beam densitometer, software version 11.2; Hologic, Waltham, MA). We also calculated height-adjusted measures of bone density including bone mineral apparent density (BMAD) of the lumbar spine (L2–4)34 and WB bone mineral content/height (WB BMC/Ht). Body composition measures, including lean and fat mass values, were measured using the same equipment. Z-scores for BMD were calculated using reference databases available to Hologic,35 while Z-scores for BMAD of the spine and WB BMC/Ht were obtained using an online applet (http://www-stat-class.stanford.edu/pediatric-bones/). Coefficients of variation for spine BMD and WB BMD were 1.1% and 0.8%, respectively. Coefficients of variation for fat and lean mass were 2.1% and 1.0%.
Data were analyzed using the statistical software JMP 5.0.1 (SAS institute, Inc, Cary, NC), and are reported as means ±SD. P values of <.05 were considered significant and trends (P values between .05 and .10) are also reported. Analysis of variance was used to determine differences between groups (based on duration of SSRI use) and a Tukey-Kramer test was employed to correct for multiple comparisons. We performed simple correlation analysis to determine associations between bone density measures and duration of SSRI use, duration since diagnosis of AN, BMI, and duration since last menses. Because subjects with longer duration since diagnosis of AN may be more likely to be on SSRIs, we considered this variable to be a potential confounder of any detected associations between bone density measures and duration of SSRI use. BMI was used as an indicator of nutritional status, and duration of amenorrhea as an indicator of hypgonadism, both of which are important determinants of low BMD in AN.
We next performed mixed model stepwise regression analysis to determine independent predictors of bone density measures (P=.10 to enter and leave the model), and variables included in the regression model were age, BMI, duration since diagnosis of AN, duration of amernorrhea, and duration of SSRI use. This analysis was selected in order to account for potentially confounding variables and to minimize the masking of associations of various independent variables with dependent variables from confounders.
No difference was observed in bone age, percent ideal body weight, Tanner stage, or duration of amenorrhea between SSRI-0, SSRI<6M, and SSRI>6M (Table 1). Subjects in the SSRI-0 group were younger than those in the SSRI>6M group, while there was no difference in age between SSRI<6M and the other two groups. Compared with SSRI-0, subjects in the SSRI<6M group had a higher BMI. However, the BMI of subjects in the SSRI>6M group did not differ significantly from either group. Finally, duration since diagnosis of AN was significantly longer in SSRI>6M than for SSRI-0 and SSRI<6M. Exercise activity, calcium, and vitamin D intake did not differ between the groups (data not shown).
Bone Density Measures
Subjects who had been on an SSRI>6M had significantly lower spine, femoral neck, and WB BMD Z-scores than subjects who had been on an SSRI<6M (Table 2). Spine BMAD Z-scores did not differ between the groups. Hip BMD and WB BMC/Ht Z- scores were lower in subjects on an SSRI>6M compared with subjects in the other two groups.
Predictors of Bone Density Measures
Table 3 shows data from simple correlation analysis examining associations between bone density measures and co-variates of interest for the group as a whole. Overall, BMI was a positive predictor of BMD Z-scores, while duration since diagnosis of AN, duration of amenorrhea and duration of SSRI use were negative predictors of BMD Z-scores.
We next performed regression modeling to determine independent predictors of bone density Z-scores for the group as a whole. We first entered BMI, age, duration of amenorrhea, and duration of SSRI use in the regression model (Table 4). An alternate regression model was next created in which we replaced duration of amenorrhea with duration since diagnosis of AN (Table 5). In both models, duration of SSRI use was a significant negative predictor of BMD Z-scores at most sites, while BMI was the most significant positive predictor. Finally, we ran an alternate regression model in which we substituted BMI with lean mass. Independent predictors of BMD Z-scores did not change significantly in this model.
We have demonstrated that girls and young women with AN between the ages of 12 and 21 years who have taken an SSRI for >6 months have lower BMD Z-scores than those who have either never used an SSRI or used this <6 months. We have also shown that duration of SSRI use is a negative predictor of BMD Z-scores, independent of other predictors such as age, BMI, duration since diagnosis of AN, or duration of amenorrhea.
Although a functional serotoninergic pathway has been demonstrated to exist in a variety of bone cells including osteoblasts, osteocytes, and osteoclasts,26 a definitive explanation for the observed effects of SSRIs on bone remains elusive at this time, and the origin and regulation of serotonin in bone remain largely unknown. Data indicate that inhibition of the serotonin transporter 5-HTT (either from a mutation in the transporter or from use of an SSRI) is associated with decreases in bone mineral accrual via a reduction in bone formation,26 and increased levels of circulating gut serotonin caused by inhibition of the low-density lipoprotein-receptor related protein 5 (LRP5) are associated with reduced bone mass.36 However, serotonin is known to be produced and function in the CNS, the GI tract, and the CV system, and the extent to which serotonin in each of these systems affects bone remains unclear.
Normalizing blood serotonin levels in lrp5 knockout mice with a reduced tryptophan diet can normalize bone function and restore bone mass in these knockout mice.36 The reduced tryptophan diet decreases circulating serotonin levels 8–10-fold, but does not affect serotonin content in the CNS. These findings suggest that brain serotonin may not be a major factor in the observed effects of serotonin on bone. However, SSRIs are assumed to mainly affect serotonin reuptake in the brain. It is unclear at this time whether SSRIs have a more widespread effect on the activity of 5-HTT in the body, especially at peripheral sites.
An important finding in our study is that duration of SSRI use, and not SSRI use alone, predicts reduced bone density. This is in contrast to studies in older men and women, in whom SSRI use per se has been associated with lower bone density and increased fracture risk.29-31 Interestingly, Gustafsson and colleagues37 showed in vitro that administration of an SSRI (fluoxetine) at a low concentration (0.01 micromoles) had a stimulatory effect on MC3T3 –E1 (osteoblastic lineage) cell proliferation. However, when administered at high concentrations (1–10 micromoles) a marked inhibitory effect was seen. This bimodal effect of SSRI therapy may explain the difference in BMD Z-scores observed in subjects using SSRIs for varying durations. As an example, spine, femoral neck and WB BMD Z-scores were slightly higher in subjects who had been on SSRIs for <6 months as compared to subjects who had never been on an SSRI, although this was not statistically significant (Table 2). In contrast, BMD Z-scores were significantly lower in girls and women who had been on higher cumulative doses of SSRIs from having used these for >6 months.
One large study of adults >17 years of age found no association between use of antidepressants and low bone density. However, a dose or duration effect was not examined.38 In contrast, some studies in adults have demonstrated a dose effect39,40 and one study suggested a duration effect41 of SSRIs on bone density. Importantly, in our study, SSRI use was associated with a greater impact on bone density at non-vertebral sites than vertebral sites, and increased risk of low bone density at the hip has been reported in some,29,39,42 but not all studies in adults.43
Importantly, it is difficult to rule out the phenomenon of confounding by indication in cross-sectional studies, and given that depression is an independent risk factor for low bone density, it is unclear whether low bone density is a consequence of the underlying disorder necessitating SSRI use (ie, depression), or a consequence of the treatment for depression (ie, use of SSRIs). However, at least one study in adult men indicates that SSRI use is associated with low bone density independent of associated depression.31 Of importance, similar studies to ours are necessary in patients with major depressive disorders to determine the impact of SSRIs on bone density in patients receiving SSRIs for depression. Prospective studies are also necessary to determine whether it is important to monitor bone density over time in patients with major depressive disorders receiving SSRIs. Another consideration is the important therapeutic role of SSRIs in the treatment of major depressive disorders, and that the benefits of these medications likely outweigh the potential risk of low bone density at this time.
Limitations of our study include its cross-sectional nature and our inability to measure serotonin levels centrally or peripherally. In addition, our study did not assess fracture risk associated with SSRI use. However, our study does indicate the need for future prospective studies examining the impact of SSRIs on bone in adolescents and young adults. This is particularly important because patients commonly prescribed SSRIs often suffer from conditions, such as depression or eating disorders, that potentially predispose them to low bone density. It is also essential to study the impact of SSRIs on bone mass accrual in adolescents given the prevalence of use of these medications in this age group, and because the adolescent years are critical for attainment of peak bone mass.
We thus demonstrate lower BMD Z-scores in girls and young women with AN 12–21 years of age who have been on an SSRI for >6 months, when compared with subjects who have taken an SSRI for <6 months or have never been on an SSRI. Duration of SSRI use is a significant independent predictor of reduced BMD Z-scores even after controlling for duration since diagnosis of AN, duration of amenorrhea, BMI, and age. These findings support the current evidence for the presence of a functional serotonergic pathway in bone, and the potential of SSRI medications to affect bone density when given for a prolonged period. More studies are necessary to determine whether it will be important to monitor bone density in patients receiving SSRIs for long periods of time. CNS
1. Grinspoon S, Miller K, Coyle C, et al. Severity of osteopenia in estrogen-deficient women with anorexia nervosa and hypothalamic amenorrhea. J Clin Endocrinol Metab. 1999;84(6):2049-2055.
2. Hay PJ, Delahunt JW, Hall A, Mitchell AW, Harper G, Salmond C. Predictors of osteopenia in premenopausal women with anorexia nervosa. Calcif Tissue Int. 1992;50(6):498-501.
3. Lennkh C, de Zwaan M, Bailer U, et al. Osteopenia in anorexia nervosa: specific mechanisms of bone loss. J Psychiatr Res. 1999;33(4):349-356.
4. Rigotti NA, Neer RM, Skates SJ, Herzog DB, Nussbaum SR. The clinical course of osteoporosis in anorexia nervosa. A longitudinal study of cortical bone mass. JAMA. 1991;265(9):1133-1138.
5. Misra M, Aggarwal A, Miller KK, et al. Effects of anorexia nervosa on clinical, hematologic, biochemical, and bone density parameters in community-dwelling adolescent girls. Pediatrics. 2004;114(6):1574-1583.
6. Misra M, Klibanski A. Anorexia nervosa and osteoporosis. Rev Endocr Metab Disord. 2006;7(1-2):91-99.
7. Soyka LA, Grinspoon S, Levitsky LL, Herzog DB, Klibanski A. The effects of anorexia nervosa on bone metabolism in female adolescents. J Clin Endocrinol Metab. 1999;84(12):4489-4496.
8. Misra M, McGrane J, Miller KK, et al. Effects of rhIGF-1 administration on surrogate markers of bone turnover in adolescents with anorexia nervosa. Bone. 2009;45(3):493-498.
9. Misra M, Miller KK, Bjornson J, et al. Alterations in growth hormone secretory dynamics in adolescent girls with anorexia nervosa and effects on bone metabolism. J Clin Endocrinol Metab. 2003;88(12):5615-5623.
10. Lawson EA, Misra M, Meenaghan E, et al. Adrenal glucocorticoid and androgen precursor dissociation in anorexia nervosa. J Clin Endocrinol Metab. 2009;94(4):1367-1371.
11. Misra M, Miller KK, Almazan C, et al. Alterations in cortisol secretory dynamics in adolescent girls with anorexia nervosa and effects on bone metabolism. J Clin Endocrinol Metab. 2004;89(10):4972-4980.
12. Misra M, Miller KK, Cord J, et al. Relationships between serum adipokines, insulin levels, and bone density in girls with anorexia nervosa. J Clin Endocrinol Metab. 2007;92(6):2046-2052.
13. Misra M, Miller KK, Tsai P, et al. Elevated peptide YY levels in adolescent girls with anorexia nervosa. J Clin Endocrinol Metab. 2006;91(3):1027-1033.
14. Utz AL, Lawson EA, Misra M, et al. Peptide YY (PYY) levels and bone mineral density (BMD) in women with anorexia nervosa. Bone. 2008;43(1):135-139.
15. Bachrach LK, Guido D, Katzman D, Litt IF, Marcus R. Decreased bone density in adolescent girls with anorexia nervosa. Pediatrics. 1990;86(3):440-447.
16. Castro J, Lazaro L, Pons F, Halperin I, Toro J. Predictors of bone mineral density reduction in adolescents with anorexia nervosa. J Am Acad Child Adolesc Psychiatry. 2000;39(11):1365-1370.
17. Grinspoon S, Baum H, Lee K, Anderson E, Herzog D, Klibanski A. Effects of short-term recombinant human insulin-like growth factor I administration on bone turnover in osteopenic women with anorexia nervosa. J Clin Endocrinol Metab. 1996;81(11):3864-3870.
18. Soyka LA, Misra M, Frenchman A, et al. Abnormal bone mineral accrual in adolescent girls with anorexia nervosa. J Clin Endocrinol Metab. 2002;87(9):4177-4185.
19. Raymond JR, Mukhin YV, Gelasco A, et al. Multiplicity of mechanisms of serotonin receptor signal transduction. Pharmacol Ther. 2001;92(2-3):179-212.
20. Baldock PA, Sainsbury A, Allison S, et al. Hypothalamic control of bone formation: distinct actions of leptin and y2 receptor pathways. J Bone Miner Res. 2005;20(10):1851-1857.
21. Battaglino R, Fu J, Spate U, et al. Serotonin regulates osteoclast differentiation through its transporter. J Bone Miner Res. 2004;19(9):1420-1431.
22. Bliziotes MM, Eshleman AJ, Zhang XW, Wiren KM. Neurotransmitter action in osteoblasts: expression of a functional system for serotonin receptor activation and reuptake. Bone. 2001;29(5):477-486.
23. Westbroek I, van der Plas A, de Rooij KE, Klein-Nulend J, Nijweide PJ. Expression of serotonin receptors in bone. J Biol Chem. 2001;276(31):28961-28968.
24. Kroeze WK, Kristiansen K, Roth BL. Molecular biology of serotonin receptors structure and function at the molecular level. Curr Top Med Chem. 2002;2(6):507-528.
25. Warden SJ, Bliziotes MM, Wiren KM, Eshleman AJ, Turner CH. Neural regulation of bone and the skeletal effects of serotonin (5-hydroxytryptamine). Mol Cell Endocrinol. 2005;242(1-2):1-9.
26. Warden SJ, Robling AG, Sanders MS, Bliziotes MM, Turner CH. Inhibition of the serotonin (5-hydroxytryptamine) transporter reduces bone accrual during growth. Endocrinology. 2005;146(2):685-693.
27. Sibilia V, Pagani F, Lattuada N, Greco A, Guidobono F. Linking chronic tryptophan deficiency with impaired bone metabolism and reduced bone accrual in growing rats. J Cell Biochem. 2009;107(5):890-898.
28. Eskandari F, Martinez PE, Torvik S, et al. Low bone mass in premenopausal women with depression. Arch Intern Med. 2007;167(21):2329-2336.
29. Haney EM, Chan BK, Diem SJ, et al. Association of low bone mineral density with selective serotonin reuptake inhibitor use by older men. Arch Intern Med. 2007;167(12):1246-1251.
30. Richards JB, Papaioannou A, Adachi JD, et al. Effect of selective serotonin reuptake inhibitors on the risk of fracture. Arch Intern Med. 2007;167(2):188-94.
31. Williams LJ, Henry MJ, Berk M, et al. Selective serotonin reuptake inhibitor use and bone mineral density in women with a history of depression. Int Clin Psychopharmacol. 2008;23(2):84-87.
32. Diagnostic and Statistical Manual of Mental Disorders. 4th ed Washington, DC: American Psychiatric Association; 1994.
33. Greulich W PS. Radiographic Atlas of Skeletal Development of the Hand and Wrist. 2nd ed. Stanford, CA: Stanford University Press; 1959.
34. Carter DR, Bouxsein ML, Marcus R. New approaches for interpreting projected bone densitometry data. J Bone Miner Res. 1992;7(2):137-145.
35. Kelley T SB, Binkley T, et al. Pediatric BMD Reference Database for US White Children. Sorrento, Italy; 2005.
36. Yadav VK, Ryu JH, Suda N, et al. Lrp5 controls bone formation by inhibiting serotonin synthesis in the duodenum. Cell. 2008;135(5):825-837.
37. Gustafsson BI, Westbroek I, Waarsing JH, et al. Long-term serotonin administration leads to higher bone mineral density, affects bone architecture, and leads to higher femoral bone stiffness in rats. J Cell Biochem. 2006;97(6):1283-1291.
38. Kinjo M, Setoguchi S, Schneeweiss S, Solomon DH. Bone mineral density in subjects using central nervous system-active medications. Am J Med. 2005;118(12):1414.
39. Bolton JM, Metge C, Lix L, Prior H, Sareen J, Leslie WD. Fracture risk from psychotropic medications: a population-based analysis. J Clin Psychopharmacol. 2008;28(4):384-391.
40. Vestergaard P, Rejnmark L, Mosekilde L. Anxiolytics, sedatives, antidepressants, neuroleptics and the risk of fracture. Osteoporos Int. 2006;17(6):807-816.
41. Ziere G, Dieleman JP, van der Cammen TJ, Hofman A, Pols HA, Stricker BH. Selective serotonin reuptake inhibiting antidepressants are associated with an increased risk of nonvertebral fractures. J Clin Psychopharmacol. 2008;28(4):411-417.
42. Lewis CE, Ewing SK, Taylor BC, et al. Predictors of non-spine fracture in elderly men: the MrOS study. J Bone Miner Res. 2007;22(2):211-219.
43. Spangler L, Scholes D, Brunner RL, et al. Depressive symptoms, bone loss, and fractures in postmenopausal women. J Gen Intern Med. 2008;23(5):567-574.