Skip Navigation


Age and Ageing Advance Access originally published online on December 15, 2006
Age and Ageing 2007 36(1):78-83; doi:10.1093/ageing/afl147
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
36/1/78    most recent
afl147v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Morris, R.
Right arrow Articles by Masud, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Morris, R.
Right arrow Articles by Masud, T.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Copyright © The Author 2007. Published by Oxford University Press on behalf of the British Geriatrics Society.

A comparison of different balance tests in the prediction of falls in older women with vertebral fractures: a cohort study

Rob Morris1, Rowan H. Harwood1, Ros Baker1, Opinder Sahota2, Sarah Armstrong3 and Tahir Masud1,4

1 Nottingham City Hospital, UK
2 Queen's Medical Centre, Nottingham, UK
3 Trent Research and Development Support Unit, University of Nottingham, UK
4 University of Derby, UK

Address correspondence to: R. Morris. Tel: +44 (0)1159 691169. Fax: +44 (0)1159 608409. Email: rmorris2{at}ncht.trent.nhs.uk; drbob{at}innotts.co.uk


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 Key points
 References
 
Background: people with vertebral fractures are at high risk of developing hip fractures. Falls risk is important in the pathogenesis of hip fractures.

Aim: to investigate if balance tests, in conjunction with a falls history, can predict falls in older women with vertebral fractures.

Methods: a cohort study of community-dwelling women aged over 60 years, with vertebral fractures. Balance tests investigated were: 5 m-timed-up-and-go-test (5 m-TUG), timed 10 m walk, TURN180 test (number of steps to turn 180°), tandem walk, ability to stand from chair with arms folded. Leg extensor power was also measured.

Outcome measure: fallers (at least one fall in a 12 month follow-up period) versus non-fallers.

Results: one hundred and four women aged 63-91 years [mean = 78 ± 7], were recruited. Eighty-six (83%) completed the study. Four variables were significantly associated with fallers: previous recurrent faller (2+ falls) [OR = 6.52; 95% CI = 1.69–25.22], 5 m-TUG test [OR = 1.03; 1.00–1.06], timed 10 m walk [OR = 1.07; 1.01–1.13] and the TURN180 test [OR = 1.22; 1.00–1.49] [P <0.05]. Multi-variable analysis showed that only two variables, previous recurrent faller [OR = 5.60; 1.40–22.45] and the 5 m-TUG test [OR = 1.04; 1.00–1.08], were independently significantly associated with fallers. The optimal cut-off time for performing the 5 m-TUG test in predicting fallers was 30 s (area under ROC = 60%). Combining previous recurrent faller with the 5 m-TUG improved prediction of fallers [OR = 16.79, specificity = 100%, sensitivity = 13%].

Conclusions: a previous history of recurrent falls and the inability to perform the 5 m-TUG test within 30 s predicted falls in older women with vertebral fractures. Combining these two measures can predict fallers with a high degree of specificity (although a low sensitivity), allowing the identification of a group of patients suitable for fall and fracture prevention measures.

Keywords: accidental falls, compression fractures, musculoskeletal equilibrium, elderly


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 Key points
 References
 
Vertebral fractures are common consequences of osteoporosis. In the presence of vertebral fractures, the risk of hip fracture increases by between 2.3- and 4.5-fold [1–3]. These patients are therefore an ideal group in which to consider hip fracture prevention. In current clinical practice, most clinicians dealing with established vertebral osteoporosis focus their attentions on bone mineral density and rarely consider fall prediction or prevention. Targeting fall prevention strategies in this group of patients may reduce the risk of further fractures.

Someone with a history of previous falls has a two-thirds chance of having a fall in the subsequent year [4]. In addition, a number of balance tests have been shown to predict future falls in older people. These include the following simple tests, which may be used in a busy clinical setting: the ‘Chair Stand test’ (ability to rise from a chair with the arms folded) [5], The ‘Timed Up and Go’ (TUG) test (time taken to rise from a chair, walk a set distance e.g. 3, 5 or 10 m, turn around, walk back and sit down) [6], the ‘Tandem Walk test’ (walking heel-to-toe in a straight line) [7], ‘The Timed 10 m Walk’ [8], the ‘TURN180 test’ (number of steps needed to turn through 180°) [9]. Slightly more complex tests, which can be used in a day hospital setting include: leg extensor power [10], the ‘Berg Balance scale’ [11], ‘Functional Reach’ test [12], and ‘Body Sway’ as a measure of postural stability [13].

The aim of this study was to investigate if the use of simple balance tests, in conjunction with a falls history, can predict the likelihood of falling in older women with vertebral fractures. We included the ‘Chair Stand’, ‘5 m-TUG’ (5 m-TUG), ‘Tandem Walk’, ‘Timed 10 m Walk’ and the ‘TURN 180’ tests. The ‘leg extensor power’ was also included as this can be measured using a dynamometer, which can be kept in the clinic.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 Key points
 References
 
The design was a cohort study with the outcome being fallers (at least one fall in a 12 month follow-up period) versus non-fallers. Study participants were community-dwelling women aged 60 years or over, with at least one vertebral fracture, who were referred by general practitioners to the osteoporosis clinic. Consecutively, referred patients were approached for recruitment, to minimise selection bias. A ‘vertebral fracture’ was defined as ‘a ≥25% reduction in vertebral height on lateral radiographic imaging’.

A ‘fall’ was defined as ‘an unexpected event where a person comes to rest on the ground from an upper level or the same level’ [14]. A ‘faller’ was defined as someone who fell once or more during the follow-up period. A ‘previous faller’ was someone who fell once or more in the preceding year, and a ‘previous recurrent faller’ was someone who fell twice or more in the preceding year. ‘Previous recurrent fallers’ were a subgroup of ‘previous fallers’. Power calculations (using Egret/Siz DOS package) showed that investigating 100 subjects and assuming falls to occur in a third of subjects, the study would have 80% power to detect an odds ratio (OR) of 3.5 for an effect of a risk factor, which occurs in a third of the population. The study was approved by the Nottingham City Hospital, Local Research Ethics Committee and all patients gave written, informed consent.

Baseline data included patient demographics (age, height, weight, smoking and alcohol history, mobility aid use) and the number of falls recalled during the previous year. The researcher then performed the five balance tests and measured leg extensor power, in a standardised manner.

Follow-up was by telephone interviews conducted at 1, 2, 3, 6, 9 and 12 months. The number of falls, since the last telephone call (or since recruitment for the first telephone call) were ascertained. Each participant was categorised as a ‘non-faller’ or a ‘faller’. Data analysis was performed using SPSS version 10. Logistic regression analysis was used to examine the relation between balance tests and risk of falling. Firstly, univariate logistic regression, using ‘faller/non-faller’ as the dependent variable, was employed to investigate the relationship of the test variables. The ORs were calculated for each explanatory variable. Some variables were treated as categorical and others, where appropriate, were dealt with as continuous. The following variables were included in a categorical, dichotomous manner: current smoking, alcohol consumption in excess of 14 units per week, use of any mobility aid, ability to perform a Tandem Walk test, ability to complete a Chair Stand test, previous falls and previous recurrent falls (Table 2). Then multivariate logistic regression using stepwise-automated methods was performed to investigate the independent effects of the appropriate tests in predicting fallers. Sensitivities, specificities, positive and negative predictive values of different cut-off times were calculated for any non-categorical variable shown to have an independent effect on predicting fallers.


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 Key points
 References
 
One hundred and ten women with vertebral fractures were asked to participate in the study. Of these, 104 (95%) agreed to take part. The six women who declined did so because they did not like the idea of having to perform balance tests. Baseline socio-demographic, falls history and balance test data are shown in Table 1.


View this table:
[in this window]
[in a new window]

  		Table 1. Baseline demographic, previous falls history and balance tests data

 
Eighty-six (82.7%) women completed the study at 12 months. Seven died, four were not traceable and seven refused further involvement beyond the initial interview. The non-completers were significantly older [mean 4.7 years (95% CI 1.5–7.9)], but there were no other significant differences between the completers and non-completers.

Thirty-nine (45.3%) women had fallen by first year. In the univariate analysis, four variables were significantly associated with the risk of falling in the next year: previous recurrent faller, ‘5 m-TUG Test’, ‘Timed 10 m Walk’ and ‘TURN180 test’ (Table 2). Previous recurrent fallers had 6.5 times increased risk. Each second longer it took to perform the ‘5 m-TUG’ test was associated with a 3% increase, each extra second to perform the ‘Timed 10 m walk’ was associated with a 7% increase, and each extra step it took to perform the ‘TURN 180’ test was associated with a 22% increase in risk of becoming a faller.


View this table:
[in this window]
[in a new window]

  		Table 2. Univariate logistic regression with fallers at first year as the dependent

 
Multi-variable logistic regression, using the four variables found to be significantly associated in univariate analysis, was performed using a model employing the forward stepwise procedure. Two variables, previous recurrent faller [adjusted OR 5.60 (95% CI 1.40–22.45; P = 0.02)] and the ‘5 m-TUG’ test [adjusted OR 1.04 (95% CI 1.00–1.08; P = 0.04)] remained in the final model, suggesting that these two variables are independent predictors. Adjusted OR for the ‘5 m-TUG’ test showed that every second longer it took to perform the test was associated with a 4% increased chance of being a faller. We checked the multivariate logistic regression analysis using the ‘methods = enter’ methodology and the results were similar, with no other variable having an independent significant effect in predicting fallers.

A cut-off of 10 s for the 5 m-TUG test gives 95% sensitivity, but low specificity (11%). A 40 s cut-off gives 94% specificity but the sensitivity is considerably reduced (20%) (Table 3). Someone with a time of more than 40 s is almost certain to fall, whereas someone with a time less than 10 s is very unlikely to fall.


View this table:
[in this window]
[in a new window]

  		Table 3. Sensitivities, specificities, positive predictive values, negative predictive values, odds ratios (± 95% CIs) [P] of different cut-offs for the timed up and go (TUG)test in predicting fallers

 
The receiver operator characteristic curve for the ‘5 m-TUG’ test in predicting fallers estimated the optimum cut-off as 30 s (area under the curve 60%). Using this cut-off, the OR (95% CIs) for the inability to complete the ‘5 m-TUG’ test within 30 s for predicting fallers at 1 year was 2.86 (1.01–8.11) [P = 0.049].

With both explanatory variables in a logistic regression model, the adjusted OR for previous recurrent fallers and the inability to perform the ‘5 m-TUG’ test within 30 s were 6.24 (1.58–24.60) [P <0.01] and 2.69 (0.90–8.02) [P = 0.07], respectively. These are very similar to the univariate ORs, indicating that they act essentially independently of each other. The OR for the combination of previous recurrent fallers and the inability to perform the ‘TUG’ test within 30 s was 16.79.

Only 5 of the 86 subjects who completed the study had both a history of recurrent falling and were unable able to complete the ‘5 m-TUG’test within 30 s. All five subjects had at least one fall at first year (positive predictive value and specificity of 100%). Sensitivity was only 13%, however, implying 87% of fallers are missed.

Thirty subjects either had previous recurrent falls or inability to perform the ‘5 m-TUG’ test within 30 s. Of these, 20 subjects were fallers at first year (sensitivity 51%, specificity 79%, ORs 3.90 (1.52–9.96) [P <0.01]). The either/or test thus correctly predicts half of future fallers, and falsely predicts that 21% of non-fallers will fall.


   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
Key points
 References
 
The results of this study suggest that, out of the various performance-based balance tests investigated, the 5 m-TUG test performed the best in predicting falls in older women with vertebral fractures. The ‘Up and Go test’ (not timed) was originally described as a 5-point subjective rating scale and was assessed to be a satisfactory clinical measure of balance in a group of 40 older people [15]. Subsequently, the test was modified into a timed, 3 m version as a method of assessing and quantifying locomotor performance [6]. The TUG test has been shown to have good re-test reliability in different settings [16, 17], is an independent predictor of nursing home placement [18], and is sensitive enough to show improvement in balance in patients with rheumatoid arthritis undergoing knee extensor and flexor muscle training [19]. Its usefulness in cognitively impaired people has been questioned [20] and the height and type of chair used (armrests or not) can affect performance [21].

Cross-sectional and case-controlled studies have shown that the TUG test is associated with an increased risk of falls [22–25]. This study is the first prospective study of predictors of falling in women with vertebral fractures. The 5 m-TUG and a history of previous recurrent falling were independent risk factors for falling. The chance of being a faller in the next year increases by 4% for every additional second it takes to complete the 5 m-TUG test, after adjustment for previous recurrent fallers. Failing to complete the 5 m-TUG in 30 s is associated with a 3-fold increase in risk of being a future faller. However, with this cut-off, the sensitivity is only 33%, and specificity 85%, implying that it will fail to predict falls in two-third of future fallers, but will falsely predict falls in only 15% of non-fallers.

A better predictor, than the TUG, of future fallers was previous recurrent fallers. Thus, those subjects with two or more falls in the previous year were 5.6 times more likely to have at least one fall in the follow-up year, after adjustment for the 5 m-TUG. Single previous falls were not significantly associated with future risk although some, but not all, other data suggest that a previous faller is at a higher risk [4, 26].

Combining the independent predictors improved fall prediction to a clinically useful level. The inability to complete the 5 m-TUG within 30 s combined with a previous history of recurrent falls gave a specificity of 100% (i.e. all fell and should therefore be considered for fall and fracture prevention measures). The sensitivities of this combination in predicting fallers at first year, however, was low (13%), implying that the combination fails to predict most falls. These sensitivities can be improved to 51% by using the ‘either or’ combination method. Intervening on these patients would represent reasonable targeting as only 21% of people ‘screened in’ as at high risk would not have subsequently fallen.

Univariate analysis showed that some of the other performance-based tests also had some value in predicting falls, although they were inferior to the TUG test. A previous study showed that an increase in the number of steps while turning 180° is a marker of turning difficulty, which in itself is associated with fall-risk [27], and another recent study has shown that the number of steps taken to turn has a high sensitivity for identifying multiple fallers [28]. Two studies have shown that poor tandem walk ability was a predictor of falls [4, 29]. A French-longitudinal study of 7,575 women aged 75 years and over, showed that, after adjustment for bone density, poor tandem walk ability, as well as slow gait speed and poor vision, were independent predictors of hip fracture [7]. In our study the Chair Stand test was unable to predict falls in contrast to other studies showing that the inability to perform this test can predict fall and hip fracture risk [4, 5].

The TUG test includes elements of the other performance-based tests studied, including getting up from a chair, walking and turning. A high degree of correlation and co-linearity therefore would be expected between these tests, which is why on multi-variable analysis, most of the tests that showed a significant association with fall-risk were no longer independent predictors. We chose to test the 5 m version of the TUG test in contrast to the 3 m version that has been studied by most previous investigators. The 5 m version is more in line with what clinicians ask their patients to do routinely in clinics and on the wards to make a gait assessment. Observing a gait for only 3 ms may not be enough to make a proper qualitative gait assessment. The only situation we believe the 3 m version has an advantage is when there may be a limitation of space in a small clinic room. A recent study by Bischoff et al., in 413 community-dwelling and 78 institutionalised women aged 65–85 years, defined a cut-off of 12 s for the 3 m version of the TUG, although this cut-off was suggested as a screening tool to make a further in-depth mobility assessment rather than as a predictor of falls [30].

The study was slightly underpowered to detect a significant effect of some risk factors. Numbers recruited were based on power calculations for detecting an odds ratio of 3.5 for an effect of a risk factor occurring in a third of a population. Many of the risk factors studied may have had a prevalence of less than a third, and also gave an odds ratio of 2 or more (but less than 3.5) in predicting fallers and recurrent fallers. A larger study may have shown a significant positive association between these risk factors and fallers. However, the aim of the study was to find the most important predictors of falls, which the clinician can use in a busy clinical scenario, whereas risk factors, which only have a minor role in fall prediction would be less relevant (unless used in combination as a risk score).

A further limitation is that we took no account of co-morbidity and the effects of drugs. Drugs are a well-known risk factor for falling, which clinicians are generally aware of and usually try and rationalise in subjects at high risk of falls. However, the TUG and falls history are ‘final common pathways’ for a number of mechanisms that mark someone out as at increased falls risk, and highlight the need for intervention and perhaps the need to compromise on an otherwise-indicated drug therapy.

In conclusion, we found that the combination of the falls history and the 5 m-TUG test have potential to provide a clinically useful degree of selection for falls prevention intervention in this high-risk group of patients.


   Key points
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
References
 

  • Older women with vertebral fractures are at an increased risk of hip fractures.
  • A history of previous recurrent falling and the inability to perform the 5 m-TUG test within 30 s (‘timed up and go’) can independently predict fallers in women with known vertebral fractures.
  • Combining these two parameters can improve fall prediction further.
  • Identifying people at a high risk of falls should allow targeting of fall and fracture prevention measures to a high-risk group.


   Acknowledgements
 
We thank the Clinic Staff at Nottingham City Hospital for assisting with the study.


   Conflicts of interest
 
None


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 Key points
 References
 

  1. Melton LJ III, Atkinson EJ, Cooper C, et al. (1999) Vertebral fractures predict subsequent fractures. Osteoporos Int 10 214–21.[CrossRef][ISI][Medline]
  2. Black DM, Arden NK, Palermo L, et al. Study of Osteoporotic Fractures Research Group. (1999) Prevalent vertebral deformities predict hip fractures and new vertebral deformities but not wrist fractures. J Bone Miner Res 14 821–8.[CrossRef][ISI][Medline]
  3. Ismail AA, Cockerill W, Cooper C, et al. (2001) Prevalent vertebral deformity predicts incident hip though not distal forearm fracture: results from the European Prospective Osteoporosis Study. Osteoporos Int 12 85–90.[CrossRef][ISI][Medline]
  4. Nevitt MC, Cummings SR, Kidd S, Black D. (1989) Risk factors for recurrent non-syncopal falls. A prospective study. JAMA 261 2663–8.[Abstract]
  5. Cummings SR, Nevitt MC, Browner WS, et al. (1995) Risk factors for hip fracture in white women. N Engl J Med 332 767–73.[Abstract/Free Full Text]
  6. Podsiadlo D and Richardson S. (1991) The Timed "Up and GO". A test of basic functional mobility for frail elderly persons. J Am Geriatr Soc 39 142–8.[ISI][Medline]
  7. Dargent-Molina P, Favier F, Grandjean H, et al. (1996) Fall-related factors and risk of hip fracture: the EPIDOS prospective study. Lancet 348 145–9.[CrossRef][ISI][Medline]
  8. Kressig RW, Wolf SL, Sattin RW, et al. (2001) Associations of demographic, functional, and behavioural characteristics with activity-related fear of falling among older adults transitioning to frailty. J Am Geriatr Soc 49 1456–62.[CrossRef][ISI][Medline]
  9. Simpson JM, Worsefield C, Reilly E, Nye N. (2002) A standard procedure for using TURN180 to test dynamic postural stability among elderly people. Physiotherapy 88 342–51.[CrossRef]
  10. Bassey EJ, Fiatarone MA, O’Neill EF, Kelly M, Evans WJ, Lipsitz LA. (1992) Leg extensor power and functional performance in very old men and women. Clin Sci 82 321–7.
  11. Thorbahn LD and Newton RA. (1996) Use of the Berg Balance Test to predict falls in elderly persons. Phys Ther 76 576–83.[Abstract/Free Full Text]
  12. Duncan PW, Studenski S, Chandler J, Prescott B. (1992) Functional reach: predictive validity in a sample of elderly male veterans. J Gerontol 47 M93–8.[ISI][Medline]
  13. Lord SR, Clark RD, Webster IW. (1991) Physiological factors associated with falls in an elderly population. J Am Geriatr Soc 39 1194–1200.[ISI][Medline]
  14. WHO. (1977) Manual of the International Classification of Diseases, Injuries, and Causes of Death, Ninth revision. Geneva World Health Organization vol. 1.
  15. Mathias S, Nayak USL, Isaacs B. (1986) Balance in the elderly patient: the get up and go test. Arch Phys Med Rehabil 67 387–9.[ISI][Medline]
  16. Steffan TM, Hacker TA, Mollinger L. (2002) Age- and gender-related test performance in community-dwelling elderly people: six minute walk test, Berg balance scale, timed up and go test, and gait speeds. Phys Ther 82 128–37.[Abstract/Free Full Text]
  17. Morris S, Morris ME, Iansek R. (2001) Reliability of measurements obtained with the timed "up and Go" test in people with Parkinson's disease. Phys Ther 81 810–8.[Abstract/Free Full Text]
  18. Nikolaus T, Bach M, Oster P, Schlierf G. (1996) Prospective value of self-report and performance based tests of functional status for 18-month outcomes in elderly patients. Aging 8 271–6.
  19. McMeeken J, Stillman B, Story I, Kent P, Smith J. (1999) The effects of knee extensor and flexor muscle training on the timed-up-and-go test in individuals with rheumatoid arthritis. Physiother Res Int 4 55–67.[Medline]
  20. Rockwood K, Awalt E, Carver D, MacKnight C. (2000) Feasibility and measurement properties of the functional reach and the timed up and go tests in the Canadian study of health and ageing. J Gerontol A Biol Sci Med Sci 55 M70–3.[Abstract]
  21. Siggeirsdottir K, Jonsson BY, Jonsson H Jr, Iwarsson S. (2002) The timed "up and go" is dependent on chair type. Clin Rehabil 16 609–16.[Abstract/Free Full Text]
  22. Anacker SL and Di Fabio RP. (1992) Influence of sensory inputs on standing balance in community-dwelling elders with a recent history of falling. Phys Ther 72 575–81.[Abstract/Free Full Text]
  23. Gunter KB, White KN, Hayes WC, Snow CM. (2000) Functional mobility discriminates nonfallers from one-time and frequent fallers. J Gerontol A Biol sci Med Sci 55 M672–6.[Abstract/Free Full Text]
  24. Shumway-Cook A, Brauer S, Woollacott M. (2000) Predicting the probability for falls in community-dwelling older adults using the timed up and go test. Phys Ther 80 896–903.[Abstract/Free Full Text]
  25. Chiu AY, Au-Yeung SS, Lo SK. (2003) A comparison of four functional tests in discriminating fallers from non-fallers in older people. Disabil Rehabil 25 45–50.[ISI][Medline]
  26. Lord SR, Ward JA, Williams P, Anstey KJ. (1994) Physiological factors associated with falls in older community-dwelling women. J Am Geriatr Soc 42 1110–17.[ISI][Medline]
  27. Thigpen MT, Light KE, Creel GL, Flynn SM. (2000) Turning difficulty characteristics of adults aged 65 years or older. Phys Ther 80 1174–87.[Abstract/Free Full Text]
  28. Dite W and Temple VA. (2002) Development of a clinical measure of turning for older adults. Am J Phys Med Rehabil 81 857–66.[CrossRef][ISI][Medline]
  29. Chu LW, Pei CK, Chiu A, et al. (1999) Risk factors for falls in hospitalised older medical patients. J Gerontol A Biol Sci 54 M38–43.
  30. Bischoff HA, Stahelin HB, Monsch AU, et al. (2003) Identifying a cut-off point for normal mobility: a comparison of the timed "up and go" test in community-dwelling and institutionalised elderly women. Age Ageing 32 315–20.[Abstract/Free Full Text]
Received 22 February 2006; accepted in revised form 20 October 2006.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
36/1/78    most recent
afl147v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Morris, R.
Right arrow Articles by Masud, T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Morris, R.
Right arrow Articles by Masud, T.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?