HomeLarge Type Edition
HOME ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Download to citation manager
PubMed
Right arrow PubMed Citation
The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 61:488-494 (2006)
© 2006 The Gerontological Society of America

Depressive Symptoms and Changes in Body Weight Exert Independent and Site-Specific Effects on Bone in Postmenopausal Women Exercising for 1 Year

Laura A. Milliken, Jennifer Wilhelmy, Catherine J. Martin, Nuris Finkenthal, Ellen Cussler, Lauve Metcalfe, Terri Antoniotti Guido, Scott B. Going and Timothy G. Lohman

1 Department of Exercise and Health Sciences, University of Massachusetts, Boston.
Departments of 2 Physiology and 3 Nutritional Sciences, University of Arizona, Tucson.

Address correspondence to Laura A. Milliken, PhD, Department of Exercise and Health Sciences, University of Massachusetts Boston, 100 Morrissey Blvd., Boston, MA 02125. E-mail: laurie.milliken{at}umb.edu


   Abstract
Top
 Abstract
Methods
 Results
 Discussion
 References
 
Background. Lower bone mineral density (BMD) has been documented in clinically depressed populations, and depression is the second most common chronic medical condition in general medical practice. Therefore, the purpose of this study was to determine whether depressive symptoms, vitality, and body weight changes were related to 1-year BMD changes after accounting for covariates.

Methods. Healthy postmenopausal women (n = 320; 40–65 years) were recruited, and 266 women completed the study. Participants were 3–10 years postmenopausal, sedentary, and either taking hormone replacement therapy (1–3.9 years) or not taking it (at least 1 year). Exclusion criteria were: current smoking status, history of fractures, low BMD, body mass index >32.9 or <19.0, or use of bone altering medications. Regional BMD was measured from dual-energy x-ray absorptiometry at baseline and 1 year. Self-reported depressive symptoms and vitality were measured using standard questionnaires.

Results. Both the vitality and depressive symptoms scores were related to BMD changes at the femur neck but not at the greater trochanter or spine. Weight change was a predictor of BMD changes in the trochanter and spine but not in the femoral neck. Weight change and vitality and/or depressive symptoms had differential and site-specific effects on BMD changes at the hip. Vitality and depressive symptoms related to femoral neck changes and weight change related to greater trochanter changes.

Conclusions. The negative impact of depressive symptoms on BMD in this population of postmenopausal women was independent of body weight or other behavioral factors such as calcium compliance or exercise.

THE loss of bone mineral density (BMD) with aging is a result of the complex interactions of hormonal, environmental, nutritional, and genetic factors. In recent years, psychological status has been identified as another factor possibly related to the loss of BMD (1,2). In clinically depressed populations, significantly lower BMD has been found compared to nondepressed controls (3–7). Lower BMD in depressed populations could be related to depression itself or to other behavioral disturbances that occur as a result of depression. Factors sometimes associated with depression such as lower levels of physical activity, changes in body weight, lower calcium compliance, or use of antidepressant medications have been postulated as underlying causes of bone loss (8). It is also possible that the depressed state alters the metabolic hormonal milieu resulting in bone mineral loss (9).

The difference in BMD between depressed and nondepressed populations has been documented in both medicated and nonmedicated individuals, in males and females, and in persons who were clinically depressed as well as in persons with undiagnosed but severe depressive episodes (3–7). Despite these findings, there have been no large prospective trials examining bone loss and depressive symptoms. Additionally, no studies have focused on depressive symptoms measured in an apparently healthy population nor have studies accounted for the potential behavioral factors (changes in body weight or low calcium compliance) that are also likely to contribute to bone loss. The clinical relevance is unambiguous considering that depression is the second most common chronic condition encountered in general medical practice (10). Therefore, the purpose of this study was to determine whether depressive symptoms, indicators of well being, and changes in body weight were significantly related to 1-year BMD changes after accounting for behavioral factors (e.g., calcium compliance, exercise, and use of hormone replacement therapy [HRT]) and other related factors (baseline BMD, age). In addition, in a subset of women who were exercising, we sought to determine whether depressive symptoms were associated with bone changes after accounting for body weight change, exercise compliance, and the cumulative amount of weight lifted over 1 year of training, as well as other important covariates.


   METHODS
Top
 Abstract
 Methods
Results
 Discussion
 References
 
Postmenopausal women aged 40–65 years were recruited using radio, television, and newspaper advertisements in the community. Four hundred and twenty-eight women inquired, 320 enrolled at baseline, and 266 women completed the 1-year study. Of the 108 who did not enroll, 34 had scheduling problems, 43 did not meet the inclusion criteria, 11 did not agree to the requirements of the group to which they were randomly assigned, 9 had personal issues, and 11 did not give a reason. Women were 3–10 years past menopause (natural or surgical), participated in less than 120 minutes of exercise per week, and were willing to be randomized to an exercise or control group. Exclusion criteria included: current smoking status, a history of fractures, low BMD (z score of –3.0 or less), body mass index (BMI, kg/m2) >32.9 or <19.0, or use of any bone altering medication (except HRT). At the study's start, participants were either taking HRT (for 1–3.9 years) or not taking HRT (for at least 1 year) and were randomized within group to either a 1-year supervised exercise training program or to the control group (Figure 1). All women received 800 mg of calcium citrate daily (Citracal; Mission Pharmacal, San Antonio, TX), and compliance was measured through pill counts. Exercise compliance and weight lifted throughout 1 year were monitored using workout logs. All protocols were approved by the University of Arizona's Institutional Review Board, and informed consent was obtained from each participant. The main effects of exercise with and without HRT on BMD have been published (11). The present study is a secondary analysis of the original database. Two women were excluded from analyses due to noncompliance to the study protocol and their sizable influence on the regression results. Both women lifted 30% more than did the next highest lifters, one by attending 50% more sessions and the other by performing up to 75% more repetitions on most exercises. The final sample size was 264.


Figure 01
View larger version (19K):
[in this window]
[in a new window]
  		Figure 1. Participant recruitment and randomization. HRT = hormone replacement therapy

 
Lumbar spine, femoral neck, and greater trochanter BMDs (g/cm2) were measured in duplicate (within 7 days) on medium speed at baseline and at 1 year using a Lunar DPX-L (version 1.3y; Lunar Radiation Corporation, Madison, WI) dual-energy x-ray absorptiometer (DXA). The average of the two scans was used in all analyses. Scan analysis was performed by one certified technician using the extended research analysis feature. DXA calibration was performed daily using a calibration block supplied by the manufacturer. The coefficient of variation for this block was 0.6%. BMD precision, expressed as a percentage of mean BMD, was less than 2.4% for each BMD site.

Participants completed several questionnaires at baseline including the Medical Outcomes Study 36-item Short-Form Health Survey (SF-36) (12) and the 21-item Beck Depression Inventory (BDI) (13). Demographic information was also collected through a questionnaire. Body weight (kg) was measured to the nearest 0.1 kg at baseline and at 1 year using a digital scale (model 770; SECA, Hamburg, Germany), and height was measured to the nearest 0.1 cm with a Schorr measuring board (Schorr Products, Olney, MD).

Statistical Analysis
Multiple regression was used to determine whether 1-year changes in body weight and baseline BDI score, vitality (from SF-36), or general well-being (from SF-36) could significantly account for variability in 1-year changes in femoral neck, greater trochanter, and spine BMD. After adjusting for covariates (baseline BMD, baseline body weight, age, exercise group assignment, HRT use, and calcium compliance), changes in body weight plus either BDI score, vitality, or general well-being variables were added to predict the 1-year changes in regional BMD. For the subset of women who exercised, the impact of the cumulative amount of weight lifted during the 1-year exercise program, exercise compliance, changes in body weight, along with BDI score or vitality were tested, adjusted for the above covariates except exercise group assignment. All analyses were carried out using the Statistical Package for Social Sciences (v. 11.5; SPSS, Chicago, IL).


   RESULTS
Top
 Abstract
 Methods
 Results
Discussion
 References
 
Participant physical characteristics (n = 264) for BMI, regional BMD, and baseline values for vitality and depressive symptoms are given in Table 1. For those 54 women who dropped out, more dropped out from the exercise group than from the control group (20% vs 13%), and more dropped out from the no HRT group than from the HRT group (19% vs 14%). Independent t tests were used to compare characteristics of women who dropped out to those of women who completed the study. The dropouts were 2 years younger but were not different for number of years past menopause; estrogen levels; height; weight; percent fat, lean soft tissue, total body BMD and regional BMDs. Bone density values for our sample of postmenopausal women are similar to others measured using similar technology (14–17). Figures 2 and 3 show frequency distributions for BDI scores and changes in body weight. Table 2 shows the standardized regression coefficients for the three models predicting 1-year changes in each regional BMD site with the change in body weight plus either vitality or depressive symptoms as predictors. Both the vitality and BDI scores were significantly related to BMD changes at the femur neck but not at the greater trochanter or spine. Vitality was a positive predictor of femoral BMD changes over 1 year (p =.034). BDI was a negative predictor of 1-year BMD changes at the femoral neck (p =.026). General well-being (from the SF-36) was not a significant predictor of BMD changes. The results for depressive symptoms were substantiated by replacing BDI score in the regression model with either the probability of depression (calculated using the Women's Health Initiative's formula) or a single question about depressive symptoms ("Have you felt depressed or sad much of the time in the past year?" n = 243). These alternate variables also significantly predicted femoral neck BMD changes (p =.09 and.01, respectively).


View this table:
[in this window]
[in a new window]
  		Table 1. Participant Physical Characteristics (N = 264).

 

Figure 02
View larger version (10K):
[in this window]
[in a new window]
  		Figure 2. Frequency distribution for the Beck Depression Inventory (n = 264)

 

Figure 03
View larger version (10K):
[in this window]
[in a new window]
  		Figure 3. Frequency distribution for the 1-year change in body weight (n = 264)

 

View this table:
[in this window]
[in a new window]
  		Table 2. Standardized Regression Coefficients for the Prediction of Bone Mineral Density (BMD) Changes Over 1 Year for Covariates Plus the Change in Body Weight and Either Vitality or Depressive Symptoms (N = 264).

 
After accounting for the effects of baseline bone density, baseline body weight, age, calcium compliance, HRT use, and exercise group assignment, weight change was a predictor of 1-year BMD changes in the trochanter (p <.01) and spine (p <.10) but not the femoral neck. Weight change after 1 year and vitality or BDI score had differential and site-specific effects on bone density changes at the hip, with vitality and BDI score related to femoral neck changes and weight change related to BMD changes in the greater trochanter. When comparing the standardized regression coefficients (ß), we found that the magnitude of the effect for BDI score (ß = –0.139) exceeded the effects of exercise in HRT users and nonusers (ß = 0.070 and 0.101, respectively). At the greater trochanter, the change in body weight (ß = 0.211) was a more powerful predictor of BMD change than were HRT use (ß = 0.085) and exercise group assignment in HRT users and nonusers (ß = 0.131 and 0.190, respectively). Weight increases were associated with a greater change in trochanter BMD, whereas weight decreases were associated with a smaller change in trochanter BMD. Figures 3 and 4 illustrate the result on bone in response to arbitrarily selected high and low values of BDI, vitality, and body weight changes.


Figure 04
View larger version (13K):
[in this window]
[in a new window]
  		Figure 4. Changes in bone mineral density (BMD) for high and low depression and vitality scores and exercise group status. BDI = Beck Depression Inventory; SF-36 = Medical Outcomes Study 36-item Short-Form Health Survey

 
In the subset of women who exercised (n = 140), the effects of vitality and BDI score were examined after accounting for age, baseline BMD, HRT use, change in body weight, baseline body weight, exercise compliance, calcium compliance, and the cumulative amount of weight lifted over 1 year. Increased vitality was associated with greater BMD changes at the femoral neck (p =.065). At all other sites, neither vitality nor BDI score (or scores on alternate depression indices) were related to BMD changes. Changes in body weight also were not related to BMD changes at any BMD site. The cumulative amount of weight lifted during the 1-year program predicted BMD changes (p <.01) at the greater trochanter.


   DISCUSSION
Top
 Abstract
 Methods
 Results
 Discussion
References
 
Factors that affect the loss of BMD in postmenopausal women are numerous and include age, exercise, HRT use, calcium intake, loss of body weight, and psychological factors. The link between depression and bone loss has been documented in clinically depressed populations, mostly based on cross-sectional studies examining either women acutely ill with a major depressive episode (3,7,18) or women identified as depressed using standardized diagnostic checklists or interviews (3–5,19,20). In each case, lower regional BMD or elevated bone remodeling, a precursor to bone loss, was found in the depressed groups versus controls. Robbins and colleagues (21) found a significant negative association between depression and hip BMD (measured once 2 years after the assessment of depression) in a large (n = 1566) random sample of males and females aged 65–100 years, 16% of whom were clinically depressed. In a sample of 102 Portuguese white women selected for elevated depression, those with osteoporosis were significantly more depressed (BDI score = 16.6) than were women without osteoporosis (BDI score = 13; 22). In the only prospective study, Schweiger and colleagues (6) found greater 2-year spine BMD loss in 18 depressed patients (receiving medication and outpatient treatment) compared to 21 controls. In contrast to most studies, Amsterdam and Hooper (23) and Reginster and colleagues (24) did not find a BMD depression relationship. However, in both studies, most (24) or all (23) of the analyses may have been limited due to low statistical power. Unique to our longitudinal study was an association between 1-year BMD changes and initial levels of self-reported depressive symptoms and vitality, after accounting for several important covariates.

The present results, a secondary analysis of data from a previously published clinical trial (11), provide evidence that this link between depressive symptoms and bone loss exists in a population that exhibited much lower levels of depressive symptoms than participants in previous studies (Figure 2). For example, the mean score on the BDI was 4.5 with a range of 0–27. The percentage of individuals who scored >9 on the BDI in the present study was 12.5%, whereas 64.7% scored >9 in the study by Coehlo and colleagues (22). We found that depressive symptoms were significantly related to BMD changes at the femoral neck and accounted for an additional 1.8% (1.3% adjusted) of variation at that site. This is consistent with Michelson and colleagues (4) and Reginster and colleagues (24) who reported that the largest differences between the BMD of the depressed and nondepressed participants occurred at the femur. Also consistent with the present study, Robbins and colleagues (21) reported that depression accounted for an additional 2% of the variability in the total hip BMD in a regression model accounting for age, race, gender, alcohol use, smoking status, estrogen use, and BMI. Similar to both Amsterdam and Hooper (23) and Reginster and colleagues (24), we did not find significant effects of depressive symptoms on the BMD loss at the lumbar spine.

Depression is often associated with behavioral factors such as low activity levels or changes in body weight that could influence bone independently of other depressive symptomatology. Although these behavioral factors contribute to bone loss in individuals who are depressed, depressive symptoms increase the 1-year loss of BMD even after accounting for the influence of behavioral factors such as exercise group assignment, calcium supplement compliance, and changes in body weight. In addition, the influence of depressive symptoms on femoral neck BMD was larger than that of the exercise effect (Figure 4). Hence, depressive symptoms may exert a negative effect on bone through mechanisms that are, at least in part, unrelated to behavior. However, in the present study, though women in the control group were asked to maintain their baseline exercise behaviors (<120 minutes per week), decreases in exercise could have occurred and affected BMD. Although 1-year effects of depressive symptoms are generally small, the negative impact may become substantial if this loss persists long-term or if it is combined with other negative factors affecting BMD.

Subanalyses performed on the data of the women who exercised (n = 140) showed the impact of depressive symptoms or vitality on BMD changes after accounting for exercise compliance and amount of weight lifted, rather than simply the exercise group assignment. BMD changes at the femoral neck were associated with self-reported vitality (p =.065), but not depressive symptoms (p =.199). Because of the small number of participants relative to the number of independent variables in this analysis, replication in a larger sample of exercising women is needed to determine whether depression and vitality influence BMD changes after accounting for the amount of exercise performed. Noteworthy was the persistence of vitality as a predictor of BMD changes even after accounting for the effects of HRT use, exercise program compliance, calcium supplement compliance, and the amount of weight lifted over the 1-year program.

Although the present study was not designed to reduce body weight and the average body weight change was negligible (0.22 kg), there was a considerable range of changes in weight (–13.9 to +12.2 kg; Figure 3), and these changes in weight had moderate effects on BMD changes. The effect of weight change was a more important factor affecting femoral neck bone changes than both exercise and HRT use (Figure 5). When only the exercisers were examined, the amount of weight lifted was the most important factor affecting trochanter BMD changes, and changes in body weight were not significant. This finding should be replicated in a larger sample.


Figure 05
View larger version (11K):
[in this window]
[in a new window]
  		Figure 5. Changes in femur and trochanter bone mineral density (BMD) for the mean and 2 standard deviations (SDs) above and below the mean for the change in weight

 
The phenomenon of weight loss-associated bone loss has been reported consistently (25–31). Because of this phenomenon, bone density clinical trials generally attempt to minimize weight loss throughout the intervention. Despite any investigators' best efforts, moderate weight changes may occur and will confound the overall results. As covariates of bone density changes (such as depressive symptoms or weight loss) are identified, investigators should account for their confounding effects when interpreting the final results of long-term exercise trials. Also, because weight loss and depression have differential site-specific effects, it is important for investigators to examine the femoral neck and greater trochanter areas rather than only "total hip" BMD.

Summary
The negative impact of depressive symptoms on BMD in this population of postmenopausal women was independent of changes in body weight and varying calcium compliance, and was not ameliorated by assignment to a strength-training program. Furthermore, the impact of depressive symptoms (at the femoral neck) and weight change (at the greater trochanter) was comparable to or exceeded the impact of hormone use and exercise training. These results suggest that the presence of depressive symptoms may be clinically relevant with regard to bone health. Future bone density clinical trials should control for the change in body weight and for depressive symptoms when assessing the osteogenic potential of any intervention.


   Acknowledgments
Top
 Abstract
 Methods
 Results
 Discussion
 References
 
This work was supported by the National Institute for Arthritis, and Musculoskeletal and Skin Diseases, National Institutes of Health (AR39559), and by Mission Pharmacal.


   Footnotes
Top
 Abstract
 Methods
 Results
 Discussion
 References
 
Decision Editor: Luigi Ferrucci, MD, PhD

Received April 27, 2005

Accepted September 21, 2005


   References
Top
 Abstract
 Methods
 Results
 Discussion
 References
 

  1. Cizza G, Ravn P, Chrousos GP, Gold P. Depression: a major, unrecognized risk factor for osteoporosis? Trends Endocrinol Metab. 2001;12:198-203.[Medline]
  2. Lyles KW. Osteoporosis and depression: shedding more light upon a complex relationship. J Am Geriatr Soc. 2001;49:827-828.[Medline]
  3. Halbreich U, Rojansky N, Palter S, et al. Decreased bone mineral density in medicated psychiatric patients. Psychosom Med. 1995;57:485-491.[Abstract/Free Full Text]
  4. Michelson D, Stratakis C, Hill L, et al. Bone mineral density in women with depression. N Engl J Med. 1996;335:1176-1181.[Abstract/Free Full Text]
  5. Schweiger U, Deuschle M, Korner A, et al. Low lumbar bone mineral density inpatients with major depression. Am J Psychiatry. 1994;151:1691-1693.[Abstract/Free Full Text]
  6. Schweiger U, Weber B, Deuschle M, Heuser I. Lumbar bone mineral density in patients with major depression: evidence of increased bone loss at follow-up. Am J Psychiatry. 2000;157:118-120.[Abstract/Free Full Text]
  7. Yazici KM, Akinci A, Sütçü A, Özçakar L. Bone mineral density in premenopausal women with major depressive disorder. Psychiatry Res. 2003;117:271-275.[Medline]
  8. Sobin C, Sackeim HA. Psychomotor symptoms of depression. Am J Psychiatry. 1997;154:4-17.[Abstract]
  9. Halbreich U, Palter S. Accelerated osteoporosis in psychiatric patients: possible pathophysiological processes. Schizophr Bull. 1996;22:447-454.[Abstract/Free Full Text]
  10. Whooley MA, Simon GE. Primary care: managing depression in medical outpatients. N Engl J Med. 2000;343:1942-1950.[Free Full Text]
  11. Going S, Lohman T, Houtkooper L, et al. Effects of exercise on bone mineral density in calcium-replete postmenopausal women with and without hormone replacement therapy. Osteporos Int. 2003;14:637-643.
  12. Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992;30:473-483.[Medline]
  13. Lasa L, Ayuso-Mateos JL, Vasquez-Barquero JL, Diez-Manrique FL, Dowrick CF. The use of the Beck Depression Inventory to screen for depression in the general population: a preliminary analysis. J Affect Disord. 2000;57:261-265.[Medline]
  14. Ryan A, Nicklas B, Dennis K. Aerobic exercise maintains regional bone mineral density during weight loss in postmenopausal women. J Appl Physiol. 1998;84:1305-1310.[Abstract/Free Full Text]
  15. Kroger H, Tuppurainen M, Honkanen R, Alhava E, Saarikoski S. Bone mineral density and risk factors for osteoporosis–a population-based study of 1600 perimenopausal women. Calcif Tissue Int. 1994;55:1-7.[Medline]
  16. Puntila E, Kroger H, Lakka T, Honkanen R, Tuppurainen M. Physical activity in adolescence and bone density in peri- and postmenopausal women: a population-based study. Bone. 1997;21:363-367.[Medline]
  17. Bassey EJ, Rothwell MC, Littlewood JJ, Pye DW. Pre- and postmenopausal women have different bone mineral density responses to the same high-impact exercise. J Bone Miner Res. 1998;13:1805-1813.[Medline]
  18. Herran A, Amado JA, Garcia-Unzueta MT, Vazquez-Barquero JL, Perera L, Gonzalez-Macias J. Increased bone remodeling in first-episode major depressive disorder. Psychosom Med. 2000;62:779-782.[Abstract/Free Full Text]
  19. Kavuncu V, Kuloglu M, Kaya A, Sahin S, Atmaca M, Firidin B. Bone metabolism and bone mineral density in premenopausal women with mild depression. Yonsei Med J. 2000;43:101-108.
  20. Whooley MA, Kip KE, Cauley JA. Study of Osteoporotic Fractures Research Group. Depression, falls, and risk of fracture in older women. Arch Intern Med. 1999;159:484-490.[Abstract/Free Full Text]
  21. Robbins J, Hirsch C, Whitmer R, The Cardiovascular Health Study. The association of bone mineral density and depression in an older population. J Am Geriatr Soc. 2001;49:732-736.[Medline]
  22. Coelho R, Silva C, Maia A, Prata J, Barros H. Bone mineral density and depression: a community study in women. J Psychosom Res. 1999;46:29-35.[Medline]
  23. Amsterdam JD, Hooper MB. Bone density measurement in major depression. Prog Neuropsychopharmacol Biol Psychiatry. 1998;22:267-277.[Medline]
  24. Reginster JY, Deroisy R, Paul I, Hansenne M, Ansseau M. Depressive vulnerability is not an independent risk factor for osteoporosis in postmenopausal women. Maturitas. 1999;33:133-137.[Medline]
  25. Chao D, Espeland MA, Farmer D, et al. Effect of voluntary weight loss on bone mineral density in older overweight women. J Am Geriatr Soc. 2000;48:753-759.[Medline]
  26. Cundy T, Evans MC, Kay RG, Dowman M, Wattie D, Reid IR. Effects of vertical-banded gastroplasty on bone and mineral metabolism in obese patients. Br J Surg. 1996;83:1468-1472.[Medline]
  27. Holbrook TL, Barrett-Connor E. The association of lifetime weight and weight control patterns with bone mineral density in an adult community. Bone Miner. 1993;20:141-149.[Medline]
  28. Jensen LB, Quaade F, Sørensen OH. Bone loss accompanying voluntary weight loss in obese humans. J Bone Miner Res. 1994;9:459-463.[Medline]
  29. Langlois JA, Mussolino ME, Visser M, Looker AC, Harris T, Madans J. Weight loss from maximum body weight among middle-aged and older white women and the risk of hip fracture: the NHANES I epidemiologic follow-up study. Osteoporos Int. 2001;12:763-768.[Medline]
  30. Pritchard JE, Nowson CA, Wark JD. Bone loss accompanying diet-induced or exercise-induced weight loss: a randomized controlled study. Int J Obes Relat Metab Dis. 1996;20:513-520.[Medline]
  31. Salamone LM, Cauley JA, Black DM, et al. Effect of a lifestyle intervention on bone mineral density in premenopausal women: a randomized trial. Am J Clin Nutr. 1999;70:97-103.[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Download to citation manager
PubMed
Right arrow PubMed Citation

HOME ARCHIVE SEARCH TABLE OF CONTENTS