Osteoporosis and fracture prevention continue to receive a significant amount of attention among medical practitioners and researchers. The focus in clinical medicine has been on diagnosis and pharmacological treatment of osteoporosis to prevent potentially devastating fractures.1–3 However, given that falls, not bone fragility, have been shown to be the strongest determinant of fracture,1–6 it is critical to add fall risk assessment into clinical decision-making guidelines for fracture prevention. Aided by tools that can assess and prevent falls, physical therapists are in a unique position to impact fracture prevention. This article proposes that all ambulatory, postmenopausal women who are older than 50 years be screened and treated. To help with decision making, we propose an algorithm that combines 2 validated tools—the Fall Risk Assessment Tool (FRAX) and the Functional Gait Assessment (FGA)—to assess the risks of fractures and falls, respectively. The algorithm also includes decision points at which patients should be referred to medical providers who can order bone mineral density (BMD) testing, consider medications, or address reversible fall risks, such as polypharmacy or low vision.
FRACTURES AND FALLS
According to estimates in 2005, more than 2 million people sustained fragility (ie, low trauma) fractures in the United States, resulting in more than $19 billion of associated medical costs. By 2025, these numbers are predicted to increase by almost 50%2 with 1 of every 2 Americans older than 50 years at high risk for developing osteoporosis.3 Fragility fracture prevention has been historically associated with the diagnosis and treatment of osteoporosis, although more recently there has been a call to additionally evaluate risk factors for falls.4–6 In a study assessing bone and fall-related fracture risk over a 1-year period in individuals who experienced a fracture, van Helden and colleagues7 found that 75% had fall-related factors (history of falls, visual deficits, low activity level, etc), but only 53% had bone-related factors (low BMD, previous history of fracture, low body weight, etc). Authorities have estimated that only 10% to 44% of fractures can be attributed to osteoporosis, and the true population burden of fractures comes from individuals with osteopenia or normal BMD.8,9
Falling has been shown to be the main risk factor for fractures and other injuries in the older adults.10 Approximately 30% of community-dwelling individuals and 50% of residential-dwelling adults older than 65 years fall every year, with half experiencing more than 1 fall. Five percent of these falls result in fracture and that percentage doubles for those older than 75 years.11 Falls and fall-related injuries or even a fear of falling has been shown to result in social withdrawal, loss of independence and confidence, admission to long-term care facilities, and significant depression and anxiety.12 Fracture rates are twice as high for women as for men.11 This is especially important given the much higher rates at which women are more likely to live alone than men, creating particularly dangerous conditions for falls and fractures.13
In addition to treating patients with osteoporosis, physical therapists typically screen for fall risk in certain populations14 and engage in evidence-based interventions that help reduce falls.1 By extending this practice to a wider group of patients who may be at risk for fracture, physical therapists have a tremendous opportunity to identify individuals at risk and customize interventions aimed at preventing both falls and fractures.
DETERMINING FRACTURE RISK
Bone Mineral Density Testing
Osteoporosis is defined as “a systemic skeletal disease characterized by low bone mass and microarchitectural deterioration of bone tissue, leading to enhanced bone fragility and a consequent increase in fracture risk.”15 In 1994, a World Health Organization (WHO) working group made a distinction between the definition of osteoporosis and the diagnosis. The WHO proposed that for the purpose of epidemiological studies, osteoporosis should be diagnosed when the BMD is 2.5 standard deviations less than the mean for a healthy young female.16 The number of standard deviations more or less than that mean is known as a T-score. A T-score of −2.5 or less is the criterion for a diagnosis of osteoporosis and −1.0 to −2.5 indicates a diagnosis of osteopenia.
A T-score basically provides a “relative risk” of a fracture occurring in individuals who have the risk compared with those who do not.17 The sensitivity and fracture detection rate of BMD testing are low because fracture risk is multifactorial. Bone mineral density alone captures a small percentage of risk since it does not take into account any other risk factors for fracture, such as age, fall history, impaired mobility, and medication. For example, age alone can dramatically increase risk. The annual incidence of hip fractures increases approximately 30-fold between the ages of 50 and 90 years.18 Despite these limitations, the BMD/T-score remains the most commonly used determinant of clinical treatment.19
The WHO Fracture Risk Assessment Tool
Because the traditional approach of prescribing medications on the basis of T-score alone does not fully evaluate fracture risk, the WHO Fracture Risk Assessment (FRAX), a free online tool, has come into use. It includes multiple clinical risk factors (Figure 1) either alone or in combination with a BMD test to calculate the probability (absolute risk) of fracture over the next 10 years.18 Absolute fracture risk rather than relative fracture risk (T-score) better identifies patients for whom treatment is most cost-effective.20 The National Osteoporosis Foundation recommends that patients should receive medication if their 10-year probability of sustaining a hip or other major osteoporotic fracture equals or exceeds 3% and 20%, respectively.21
The FRAX calculator offers several important advantages: it allows for identification of individuals with comorbidities that increase their fracture risk; it can assess risk in men; it has been validated for use in people of various races; and it stratifies individuals with a low BMD into those with a high risk of fracture from those with a lower risk of fracture, some of whom may not be appropriate for pharmacological intervention.6,22
An example of the clinical utility of the FRAX calculator is shown in Figure 1. Identical information was entered into the FRAX calculator with the exception of age. One patient (Figure 1a) is 55 years old and another patient (Figure 1b) is 75 years old. Both patients are postmenopausal women with the added risk factor of “parental hip fracture” and both have a T-score of −2.0 (osteopenia). The 55-year-old woman has a 1% probability of a hip fracture and a 13% probability of a major osteoporotic fracture (spine, forearm, hip, or shoulder fracture) over 10 years, while the 75-year-old woman has a 13% probability of a hip fracture and a 21% probability of a major osteoporotic fracture over the same time period. Without using the FRAX, both of these women would have been diagnosed with osteopenia on the basis of their T-scores. Using the previous 2003 National Osteoporosis Foundation guideline23 (before FRAX), the younger woman would have been offered treatment based on her BMD even though her fracture risk was low. For her, the long-term costs and risks of medication could outweigh the benefits.5,6
As with all tools, there are limitations to the FRAX. In addition to the lack of fall risk assessment (discussed later), the FRAX is based on the total hip BMD as opposed to vertebral density, thus missing some individuals with low spinal density. It is intended only for use in postmenopausal women and men aged 50 years and older and most of the risk factor questions are binary (yes/no) and therefore do not allow for gradation of risk.5,6 Further limitations are that other medications that can potentially decrease (eg, antivirals) or increase (eg, hormone replacement therapy, osteoporosis medications) bone density are not included; bone turnover rate is not taken into account; FRAX cannot be used to monitor the patient's response to medical treatment,5,6 and vitamin D intake is omitted.4
Multiple authors have identified the exclusion of fall risk in the FRAX tool as a serious limitation,4–6,17,24 since falling is the strongest single predictor of fractures.1,7,10 Yet fall risk is measurable14 and it can be reduced through customized interventions such as supervised, challenging balance, and exercise training.25,26 It may be useful to combine fracture risk and fall risk to best differentiate individuals most at risk for fracture. Patients who are at both high fall and high fracture risk may benefit the most from medical (eg, pharmacological) treatment to improve bone health and physical therapy to reduce fall risk. Physical therapists are logical practitioners to determine both of these risks, make appropriate medical referral, and provide rehabilitative care.
FALL AND FRACTURE PREVENTION
Conventional approaches to the treatment, or in some instances prevention, of osteoporosis have included exercise and medication. Research on the effect of exercise on bone density or strength has demonstrated varying results in postmenopausal women on the basis of type, frequency, and intensity of exercise. To summarize, systematic reviews and meta-analyses have shown a 1% to 2% increase in BMD or 2% to 3% increase in bone strength.27–29 The best results were found with exercise programs that included high impact weight bearing, resistance exercise, and very good compliance.30 A more comprehensive review of the role of exercise on bone health was recently published in Physical Therapy.26
Osteoporosis medications such as bisphosphonates, selective estrogen receptor modulators, hormone replacement therapy, calcitonin, and recombinant human parathyroid hormone have demonstrated increases in BMD and reductions in fracture risk.31–33 These medications have variable effectiveness and side effect profiles.27–29,32 In addition, they are relatively expensive and have issues with compliance and the evidence remains unclear on the optimal duration of safe use.34 Recent literature suggests that bisphosphonate use increases the risk of subtrochanteric or femoral shaft fracture, and while low, this risk increases with length of exposure.27,31,35 Benefits should clearly outweigh the risks for individual patients, especially for use over 3 to 5 years.34,36 The reader is referred to the report of the American College of Physicians for a comprehensive table regarding effectiveness and side effects of osteoporosis medications.37
Fracture Prevention Screening Algorithm
Prevention of falls and fractures is a critical health care initiative, particularly as the percentage of aged population increases.38,39 While the role of physical therapists in fall screening is well-established, they may not take fracture risk into account, other than considering a (often secondary) diagnosis of osteoporosis based on a T-score. There are limitations to this broad label and diagnostic approach, as was previously explained. The FRAX includes the multifactorial risks of fracture beyond bone density, yet does not consider fall risk. Given this relatively new and simple way to estimate a woman's fracture risk, we set out to devise a clinical decision-making algorithm that would inform and enhance the physical therapist's ability to screen, examine, treat, and refer postmenopausal women. In fact, we propose that all ambulatory, postmenopausal women older than 50 years, who attend physical therapy, should be screened for fracture and fall risk. Such screening is consistent with the role of the physical therapist in prevention and health and wellness promotion.40 The Fracture Prevention Screening Algorithm can guide therapists in determining whether each woman would benefit from exercise and education to promote bone health, versus those who would also benefit from a customized physical therapy program to reduce fall risk (Figure 2).
Using FRAX, FGA, and a 2-question fall screening, the algorithm stratifies patients into 1 of 4 categories on the basis of fracture and fall risk (Table 1). Both the FRAX and the FGA have “cut points” that classify fracture and fall risk, respectively. Individually, higher versus lower risk for fracture and fall suggests different treatment and referral recommendations. These recommendations are further refined by placing risk into a 2 × 2 table (Table 1). While both of the tools in the Fracture Prevention Screening Algorithm dichotomize risk into high and low, it must be stated that each patient's individual risk will lie somewhere on a continuum.
Use of the Algorithm
If a woman is older than 50 years and postmenopausal, hip fracture risk should be calculated using the FRAX. Hip fracture is used because it is the most devastating fragility fracture since its sequelae can result in high morbidity and even mortality.41 The FRAX tool can be found at www.shef.ac.uk/FRAX or through a smart phone application. Table 2 shows the data required; probabilities can be determined with or without a BMD. A patient with a 10-year probability of hip fracture that is 3% or greater is considered to have a high fracture risk. She should be referred to her physician for further evaluation (eg, BMD testing or medications).
In addition to the referral, this patient should be evaluated for fall risk. Commonly used measures of fall risk have been validated primarily on community-dwelling elders and individuals with neurological dysfunction.42,43 Ceiling effects have been found for the Berg Balance Scale in subjects with acute stroke and with spinal cord injury.44,45 Similarly, norms for the Timed Up and Go in subjects ranging from 60 to 99 years are all less than the cutoff score for fall risk.46 A population of postmenopausal patients has the potential to be very high functioning, although a range of abilities could be represented, depending on an individual's comorbidities. The FGA is composed of highly dynamic elements and may be the most useful to determine fall risk in this population.47 The FGA is a 10-item test that assesses postural stability during various walking tasks (Table 3). Each task is scored from 3 (normal) to 0 (severe impairment), with a total possible score of 30 points.
Normative data for the FGA exist for adults aged 40 to 89 years, and interrater reliability in this same group was excellent (intraclass correlation coefficient = 0.93).48 The FGA has good concurrent validity with other common tools used for fall risk: Activities of Balance Confidence Scale (r = 0.53), Berg Balance Scale (r = 0.84), and the Timed Up and Go (r = −0.84).49 Finally, FGA scores of 22/30 or less effectively predicted falls in community-dwelling adults aged 60 to 90 years.49 With experience, the FGA takes approximately 5 minutes to perform.
High Fracture/High Fall Risk
A patient at high-fracture risk who scores 22 of 30 or lower on the FGA would be categorized as high fracture/high fall risk. In addition to the medical referral mentioned previously, this woman should be further evaluated by her physical therapist to determine the causes for her fall risk (eg, impaired balance, lower limb weakness, visual impairment, and sensory loss). This detailed examination will provide the basis for a customized exercise program that challenges the patient's deficits in a progressive and supervised manner.26 Certain deficits may merit a referral to an ophthalmologist or a primary care provider to address reversible risk factors. To optimize her bone health, this patient would also benefit from education about nutrition, home safety, exercise, and health modifications26 (Table 4).
Multiple randomized controlled trials have evaluated the effect of exercise on fall risk and rate in women with osteoporosis. Balance and strengthening exercises have reduced fall risk factors and the number of falls in this population.51–53 Exercise elements that seem to be most critical to reduce falls include customization to the limitations of the patient; high dosage and supervised progression in difficulty over time; highly challenging balance activities; and 2 or more of strength, balance, flexibility, and/or endurance exercises.25,54
High Fracture/Low Fall Risk
For a patient who is high fracture/low fall risk (FGA score of 23-30), targeted intervention to prevent falls may not be justified. However, because of her fracture risk, such a patient should receive education about bone health and a medical referral for possible BMD testing and pharmacological treatment.
Low Fracture/Low Fall Risk
Women whose FRAX score indicates a less than 3% probability of hip fracture over the next 10 years are asked 2 self-report questions about balance and falls: (1) Do you have any problems with your balance? and (2) have you fallen in the past 12 months? Separately, these 2 questions have been found to accurately screen for fall risk.50,55 If the patient answers no to both questions, she is categorized as low fracture/low fall risk. Because of her postmenopausal status, however, she would still benefit from education about bone health. If the patient answers yes to one or both of the self-report questions, it is recommended to determine whether she is at risk for falls by administering the FGA.
Low Fracture/High Fall Risk
A patient with high fall risk (FGA score of 22/30 or lower), irrespective of fracture risk, should be further evaluated to determine the cause of fall risk and, if appropriate, implement a customized fall-prevention program as mentioned previously.
Examples of specific patient cases from each category are shown in Table 5.
This evidence-based screening algorithm incorporates the FRAX to determine a patient's absolute fracture risk and the Functional Gait Assessment to screen for fall risk. Using these tools together should help identify patients who are at greatest risk for fracture. This will inform the development of targeted preventive interventions. Such interventions may consist of education about bone health and exercise that maintains bone density in the postmenopausal woman who is at low risk for falls or customized balance and strengthening exercise programs for the postmenopausal woman who is at high risk for falls. In addition, it helps identify when a medical referral is warranted.
The Fracture Prevention Screening Algorithm can enhance awareness of fracture and fall risk in physical therapists. Since we suggest that the algorithm be used with ambulatory, postmenopausal women older than 50 years, therapist recommendations and interventions have the potential to prevent fractures and falls in even the nontypical patient, versus just those with confirmed osteoporosis or a history of falls. This algorithm must be further tested to validate if, when used together, the FRAX and the FGA lead to treatment that prevents falls and/or fractures better than either one alone. We recommend such studies as future avenues for research.
1. Karinkanta S, Piirtola M, Sievanen H, Uusi-Rasi K, Kannus P. Physical therapy approaches to reduce fall and fracture risk among older adults. Nat Rev Endocrinol. 2010;6(7):396–407.
2. Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A, Tosteson A. Incidence and economic burden of osteoporosis-related fractures in the United States, 2005-2025. J Bone Mineral Res. 2007;22(3):465–475.
3. Department of Health and Human Services. Bone Health and Osteoporosis: A Report of the Surgeon General. Rockville, MD; 2004.
4. Watts NB, Ettinger B, LeBoff MS. Perspective: FRAX facts. J Bone Miner Res. 2009;24(6):975–979.
5. Hamdy RC. FRAX: a fracture risk-assessment tool. Aging Health. 2009;5:489–494.
6. Adami S. The FRAX: critical appraisal. Int J Clin Rheumatol. 2009;4(6):645–650.
7. van Helden S, van Geel A, Geusens P, Kessels A, Nieuwenhuijzen Kruseman A, Brink P. Bone and fall-related fracture risks in women and men with a recent clinical fracture. J Bone Joint Surg Am. 2008;90(2):241–248.
8. Pasco J, Seeman E, Henry M. The population burden of fractures originates in women with osteopenia, not osteoporosis. Osteoporos Int. 2006;17(9):1404–1409.
9. Stone KL, Seeley DG, Lui L, et al. BMD at multiple sites and risk of fracture of multiple types: long-term results from the study of osteoporotic fractures. J Bone Miner Res. 2003;18(11):1947–1954.
10. Kannus P, Sievanen H, Jarvinen M, Parkkari J. Prevention of falls and consequent injuries in elderly people. Lancet. 2005;366(9500):1885–1893.
11. Centers for Disease Control and Prevention. Falls Among Older Adults: An Overview. Atlanta, GA: Centers for Disease Control and Prevention; 2009.
12. Kannus P, Uusi-Rasi K, Palvanen M, Parkkari J. Non-pharmacological means to prevent fractures among older adults. Ann Med. 2005;37(4):303–310.
13. Department of Health and Human Services. A Profile of Older Americans: Adminstration on Aging. Washington, DC: Department of Health and Human Services; 2011.
14. Shumway-Cook A, Woollacott M. Motor Control. Translating Research Into Clinical Practice. 4th ed. Philadelphia: Wolters Kluwer/Lippincott Williams & Wilkins; 2012.
15. Scientific Advisory Board, Osteoporosis Society of Canada. Clinical practice guidelines for the diagnosis and management of osteoporosis. Can Med Assoc J. 1996;155:1113–1133.
16. Kanis JA. Assessment of fracture risk and its application to screening for postmenopausal osteoporosis: synopsis of a WHO report. WHO Study Group. Osteoporos Int. 1994;4(6):368–381.
17. McCloskey E. FRAX: Identifying People at High Risk of Fracture. International Osteoporosis Foundation; Nyon, Switzerland: 2009.
18. Kanis J, Johnell O, Oden A. FRAX and the assessment of fracture probability in men and women from the UK. Osteoporos Int. 2008;19(4):385–397.
19. Watts N. The Fracture Risk Assessment Tool (FRAX): applications in clinical practice. J Women's Health. 2011;20(4):525–531.
20. Lewiecki M, Watts NB. New guidelines for the prevention and treatment of osteoporosis. South Med J. 2009;102(2):175–179.
21. National Osteoporosis Foundation. Physician's guide to prevention and treatment of osteoporosis. http://www.nof.org/physguide
. 2008. Accessed October 10, 2011.
22. Siris E, Delmas P. Assessment of 10-year absolute fracture risk: a new paradigm with worldwide application. Osteoporos Int. 2008;19(4):383–384.
23. Physician's Guide to Prevention and Treatment of Osteoporosis. Washington, DC; 2003.
24. Jarvinen TL, Sievanen H, Khan KM, Heinonen A, Kannus P. Shifting the focus in fracture prevention from osteoporosis to falls. BMJ. 2008;336(7636):124–126.
25. Gillespie LD, Roberston MC, Gillespie WJ, et al. Interventions for preventing falls in older people living in the community. Cochrane Database Syst Rev. 2009;2:CS007146.
26. Perry S, Downey P. Fracture risk prevention: a multidimensional approach. Phys Ther. 2012;92(1):164–178.
27. Lenart B, Lorich D, Lane JM. Atypical fractures of the femoral diaphysis in postmenopausal women. N Engl J Med. 2008;358(12):1304–1306.
28. Kennel K, Drake M. Adverse effects of bisphosphonates: implications for osteoporosis management. Mayo Clin Proc. 2009;84(7):632–638.
29. Solomon D, Rekedal L, Cadarette S. Osteoporosis treatments and adverse events. Curr Opin Rheumatol. 2009;21(4):363–368.
30. Nikander R, Sievanen H, Heinonen A, Daly RM, Uusi-Rasi K, Kannus P. Targeted exercise against osteoporosis: a systematic review and meta-analysis for optimising bone strength throughout life. BMC Med. 2010;8:47.
31. Gehrig L, Lane J, O'Connor M. Osteoporosis: management and treatment strategies for orthopaedic surgeons. J Bone Joint Surg. 2008;90-A(6):1362–1374.
32. Iwamoto J, Tsuyoshi T, Yoshihiro S. Efficacy and safety of alendronate and risedronate for postmenopausal osteoporosis. Curr Med Res Opin. 2006;22(5):919–928.
33. Stroup J, Kane M, Abu-Baker A. Teriparatide in the treatment of osteoporosis. Am J Health-Syst Pharm. 2008;65:532–539.
34. Ott SM. What is the optimal duration of bisphosphonate therapy? Clev Clin J Med. 2011;78:619–630.
35. Shane E, Burr D, Ebeling PR, et al. Atypical subtrochanteric and diaphyseal femoral fractures: report of a task force of the American Society for Bone and Mineral Research. J Bone Miner Res. 2010;25(11):2267–2294.
36. Black D. Continuing bisphosphonate treatment for osteoporosis—for whom and for how long? N Engl J Med. 2012;366(22):2051–2053.
37. Qaseem A, Snow V, Shekelle P, Hopkins R, Forciea M, Owens D. Pharmacologic treatment of low bone density or osteoporosis to prevent fractures: a clinical practice guideline from the American College of Physicians. Ann Intern Med. 2008;149(6):404–415.
38. Michael YL, Whitlock EP, Lin JS, Fu R, O'Connor EA, Gold R. Primary care-relevant interventions to prevent falling in older adults: a systematic evidence review for the U.S. Preventive Services Task Force. Ann Intern Med. 2010;153(12):815–825.
39. Leipzig R, Whitlock E, Wolff T, et al. Reconsidering the approach to prevention recommendations for older adults. Ann Intern Med. 2010;153(12):809–814.
41. Melton LJ. Adverse outcomes of osteoporotic fractures in the general population. J Bone Miner Res. 2003;18(6):1139–1142.
42. Bogle-Thorbahn L, Newton R. Use of the Berg Balance Test to predict falls in elderly persons. Phys Ther. 1996;76(6):576–583.
43. Shumway-Cook A, Baldwin M, Polissar N, Gruber W. Predicting the probability for falls in community-dwelling older adults. Phys Ther. 1997;77(8):812–819.
44. Salbach NM, Mayo NE, Higgins J, Ahmed S, Finch LE, Richards CL. Responsiveness and predictability of gait speed and other disability measures in acute stroke. Arch Phys Med Rehabil. 2001;82(9):1204–1212.
45. Wirz M, Muller R, Bastiaenen C. Falls in persons with spinal cord injury: validity and reliability of the Berg Balance Scale. Neurorehabil Neural Repair. 2010;24(1):70–77.
46. Bohannon RW. Reference values for the Timed Up and Go test: a descriptive meta-analysis. J Geriatr Phys Ther. 2006;29(2):64–68.
47. Wrisley D, Marchetti G, Kuharsky D, Whitney S. Reliability, internal consistency, and validity of data obtained with the Functional Gait Assessment. Phys Ther. 2004;84(10):906–918.
48. Walker ML, Austin AG, Banke GM, et al. Reference group data for the Functional Gait Assessment. Phys Ther. 2007;87(11):1468–1477.
49. Wrisley D, Kumar N. Functional Gait Assessment: concurrent, discriminative, and predictive validity in community-dwelling older adults. Phys Ther. 2010;90(5):761–773.
50. Summary of the Updated American Geriatrics Society/British Geriatrics Society Clinical Practice Guideline for the Prevention of Falls in Older Persons. J Am Geriatric Soc. 2011;59:148–157.
51. Carter ND, Kham KM, McKay HA, et al. Community-based exercise program reduces risk factors for falls in 65- to 75-year-old women with osteoporosis: randomized controlled trial. CMAJ. 2002;167(9):997–1004.
52. Madureira MM, Takayama L, Gallinaro AL, Caparbo VF, Costa RA, Pereira RMR. Balance training program is highly effective in improving functional status and reducing the risk of falls in elderly women with osteoporosis: a randomized controlled trial. Osteoporos Int. 2007;18(4):419–425.
53. Teixeira LEPP, Silva KNG, Imoto AM, et al. Progressive load training for the quadriceps muscle associated with proprioception exercises for the prevention of falls in postmenopausal women with osteoporosis: a randomized controlled trial. Osteoporos Int. 2010;21(4):589–596.
54. Sherrington C, Whitney JC, Lord SR, Herbert RD, Cumming RG, Close JCT. Effective exercise for the prevention of falls: a systematic review and meta-analysis. J Am Geriatr Soc. 2008;56(12):2234–2243.
55. Thurmann DJ, Stevens JA, Rao JK. Practice parameter: assessing patients in a neurology practice for risk of falls (an evidence-based review). Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology. 2008;70(6):473–479.