McGough, Ellen L. PT, PhD; Logsdon, Rebecca G. PhD; Kelly, Valerie E. PT, PhD; Teri, Linda PhD
The prevalence of functional mobility limitations and falls is higher in people with dementia compared with cognitively healthy older adults,1 and both are associated with gait and motor impairments.2 The prevalence of gait impairments increases with advancing stages of dementia.3 Slowed gait speed, alone, is associated with increasing dementia severity and associated disability.4 Gait and balance impairments and fall-related complications frequently result in hospitalization and transitions from assisted living to skilled nursing care,2,5,6 where increased assistance is needed for basic activities of daily living (ADL) related to limitations in functional mobility.7 Thus, effective strategies for improving mobility and decreasing fall risk would improve care and decrease costs associated with caring for individuals with advanced dementia. To facilitate the implementation of such fall and injury prevention programs, brief physical performance assessments are needed to identify contributing gait and balance impairments in older adults with dementia.
Physical performance assessments are less reliable in older adults with dementia than in cognitively healthy older adults, due to the impact of cognitive impairment on their ability to understand and follow complex directions. A temporal relationship between cognitive decline and subsequent development of mobility impairments has been reported.8 Therefore, owing to progressive cognitive decline, individuals with dementia are particularly vulnerable to mobility disability and fall-related injuries. Thus, brief physical performance assessments are important to assist in identifying functional mobility limitations and fall risk in people with dementia, and for identifying and intervening with individuals who are at risk for adverse events that may lead to nursing home placement.9
The Berg Balance Scale10 and the Short Physical Performance Battery (SPPB)11 have been used widely in older adults with chronic health conditions, but minimally in older adult with dementia, especially in moderate to severe dementia. At present, little is known about the utility of physical performance tests, such as the Berg and SPPB, for identifying functional mobility and fall risk in older adults with moderate to severe dementia. Using a modified version of the Berg Balance Scale, a test of balance and fall risk, to assess older adults living in dementia-specific assisted living, Kenny et al.2 reported that it was predictive of transfers from assisted living to skilled nursing care. This finding suggests potential utility of this test to identify balance problems in older adults with dementia.
The SPPB, a test of lower extremity function, has been shown to be predictive of short-term mortality and nursing home admissions in community-based older adults, but has not been used extensively in studies of older adults with dementia.10,12,13 The SPPB is a quick and easy to administer assessment that provides information on 3 domains of lower extremity physical function—gait speed, balance, and lower extremity strength and power, providing important information about functional mobility related to basic ADL.
Spatiotemporal gait parameters measured with an electronic walkway have demonstrated good to excellent reliability in older adults with frontal lobe dementia [intraclass correlation coefficients (ICC) = 0.65],14 Alzheimer disease (ICC = 0.86−0.97),15 and Huntington disease (ICC = 0.86−0.95).16 Although gait disorders are prevalent and worsen over time in older adults with dementia,17 their relationship with physical performance assessment scores, functional mobility, and falls is not well understood.
The purpose of this study was 2-fold: (1) to examine concurrent validity of 2 physical performance assessments, a modified Berg Balance Scale (mod-Berg) and the SPPB, with spatiotemporal gait parameters measured with the GAITRite Walkway System in older adults with advanced dementia, and (2) to prospectively examine their relationship to functional mobility limitations and falls over a 4-month period. A better understanding of these relationships will allow health care providers to develop strategies for detecting and addressing gait, balance, and mobility problems in older adults with dementia.
Study Design and Participants
This study was a cross-sectional and prospective analysis of functional mobility limitations and falls in older adults with moderate to severe dementia living in assisted living residences. A cross-sectional analysis of concurrent validity between baseline physical performance assessments and quantitative gait measures was conducted. This was followed by a prospective analysis to determine whether baseline physical performance and gait measures were predictive of 4-month functional mobility limitations and falls.
Older adults living in 4 dementia-specific assisted living residential units in Western Washington State participated in this study. Data were collected between September 2010 and March 2011. All study participants attended a structured physical activity program within their residence and were approved for participation in this investigation by the nursing supervisor. The Washington State Internal Review Board approved all study procedures. Informed consent was received from both the participant and their legally authorized representative.
All participants had a diagnosis of dementia, were able to walk indoor distances (at least 10 m) with or without an assistive device, and were able to follow simple instructions. Residents who were unable to walk indoor distances or who had motor impairment due to central nervous system pathology (eg, cerebral vascular accident, Parkinson disease, multiple sclerosis) were excluded from this investigation.
All measures were taken within the participant's assisted living residence in a familiar, quiet location. The physical performance assessments and quantitative gait measures were administered by a physical therapist. The nursing supervisor for the unit completed a functional mobility rating questionnaire and monthly fall report. A trained research assistant administered the Mini-Mental State Examination (MMSE). The physical therapist was blinded to the results of the functional mobility rating and MMSE to reduce researcher bias.
Global Cognitive Function
The MMSE18 is a 30-point screening measure widely used to assess global cognition in older adults. Higher scores indicate better cognitive function. The MMSE was administered to obtain an estimate of overall cognitive level of each participant. Although the diagnosis of dementia depends on a combination of factors, the general MMSE ranges for disease severity are normal 30 to 27, mild dementia 26 to 20, moderate 19 to 10, severe ≤ 10)18
Modified Berg Balance Scale
A modified version of the Berg Balance Scale,11 a 14-item assessment designed to assess balance and monitor motor changes in older adults was used for this study. The Berg has been used extensively in older adults with and without neurological disease, and good reliability has been reported in community-dwelling older adults.19 Very high test-retest reliability (ICC = 0.97) was reported in 45 older adults who were dependent in ADL and living in a residential care setting, but potential subjects with MMSE scores <10 were excluded from the analysis.20 The Berg was modified for this study, as per Kenny et al.,2 by removing 3 items (chair-to-chair transfer, the forward reach, and alternate stepping) to allow for brevity, consistency, safety, and the ability to convey instructions to individuals with moderate to severe dementia. The mod-Berg was an 11-item, 44-point assessment with a higher score indicating better balance. For the current study, we selected this modified version of the Berg because it had previously been applied in older adults with advanced dementia (mean MMSE = 14) living in dementia-specific assisted living, similar to our anticipated study population.
Short Physical Performance Battery
The SPPB is a brief assessment of lower extremity function that includes standing balance (narrow base, semitandem and tandem stance, maximum 10 seconds each), usual walking speed (4 m) and 5-Times-Sit-to-Stand Test.11 The total score ranges from 0 to 12 with a higher score indicating better function. Test components were first demonstrated, and verbal reminders were provided as needed. No physical assistance (beyond contact guarding) was provided, and no guidance, for pacing or timing, was provided during physical performance assessment. The SPPB has been used widely across healthy older adults and chronic disease, demonstrating strong associations between with morbidity and mortality, but has been used minimally in older adults with dementia.13 Because the instructions were simple and the test was brief, we chose the SPPB to test lower extremity function in older adults with moderate to severe dementia. Individual components of usual gait speed and the 5-Times-Sit-to-Stand Test have demonstrated validity in this population.21,22
Quantitative Gait Analysis
Quantitative gait measures were collected with the GAITRite Walkway System (MAP/CIR Inc, Haverton, Pennsylvania), a portable electronic mat that is 3.66-m long. The walkway contains pressure sensors arranged to measure foot contacts at a frequency of 80 Hz. Gait measures obtained in this study were determined a priori based on previous studies of fall risk in older adults (speed, cadence, stride velocity, step width, stride length variability, swing time variability, and percentage of gait cycle in double support).23 Stride-to-stride variability, defined as the coefficient of variation (CoV = standard deviation/mean × 100), of swing time and stride length were calculated for usual paced walking.24 Increased variability of spatiotemporal gait parameters, particularly stride time and stride length variability, is considered and indicator of decreased gait control.23,25,26 High test-retest reliability of spatiotemporal gait parameters has been reported in people with Alzheimer disease (ICC > 0.86.15 The reliability of stride time variability during single task, usual paced walking has been reported as excellent in cognitively healthy (ICC = 0.96) and good in older adults with dementia (ICC = 0.65).14
The GAITRite mat was placed in a quiet corridor of the unit, and distractions were minimized. To allow for a zone of acceleration, the participant was positioned at a distance of 1.5 m from the recording area of the GAITRite mat to allow for acceleration and instructed to walk until they reach a target at least 2 m past the recording area of the mat. The researcher first demonstrated walking on the GAITRite mat and participants completed 1 practice walk to ensure that they understood the task. The participant was instructed to “walk at your comfortable pace, as if you're walking down the street”. Two consecutive lengths of the mat, walking at usual pace, were recorded for each participant. Walking instructions were repeated after each length of the mat to accommodate for lapses in memory and attention. A physical therapist provided guarding for balance during walking (no physical assistance), and care was taken not to influence the participant's speed. To assess test-retest reliability, a subgroup of 8 participants repeated the usual paced walking on the GAITRite mat approximately 4 hours after initial testing.
Functional Mobility Limitations
The Functional Mobility Scale was a 10-item, 40-point assessment developed for this study to measure daily function of participants within their assisted living environment. The functional mobility rating was completed at baseline and after 4 months by the nursing supervisor on the assisted living unit. The assisted living units were small enough that the nursing supervisors were involved in direct daily care of the residents. The items on the functional mobility rating ranged from ability to climb stairs and outdoor walking to bathroom transfers and bed mobility. Each functional mobility item was rated on a 4-point scale: (1) unable, (2) very difficult, (3) somewhat difficult, and (4) not difficult, so that higher scores indicated better functional mobility (Figure 1).
A fall was defined as an event that resulted in a person coming to rest inadvertently on the ground or floor or other lower surface.22 Falls were recorded in the resident's chart when they occurred (per usual facility record keeping), and each unit nursing supervisor completed a monthly report of each participant's health status changes and falls for each of the 4 months after the baseline assessment.
Data were analyzed using SPSS 16.0 for MAC (SPSS, Inc, Chicago, Illinois). Descriptive statistics were examined to check for missing data. Preliminary analyses were performed to ensure that assumptions for statistical tests were not violated. Linearity and homoscedasticity of the mod-Berg and SPPB across all participants were evaluated using scatter plots. Normality of the mod-Berg and SPPB was evaluated in groups based on high versus low functional mobility, and falls versus no falls were evaluated with histograms, skewness, and kurtosis. Box plots were used to further examine for outliers.
Intraclass correlations coefficients (ICC), based on 2-way random effects analysis of variance, were used to quantify test-retest reliability of a subgroup of 8 participants who repeated the usual paced walking on the GAITRite mat. To interpret ICC values, we used bench marks suggested by Portney and Watkins27 (ICC < 0.75 were indicative of poor to moderate reliability and ICC > 0.75 good reliability. ICC > 0.90 were considered excellent. The level of statistical significance was set at 0.05.
Bivariate and partial correlations were used to examine the relationship between physical performance assessments (SPPB and mod-Berg) and quantitative gait measures (usual pace walking) while controlling for potential confounders, including age and MMSE score.
To examine whether baseline physical performance scores and quantitative gait measures differed in individuals with high versus low functional mobility ratings, at the 4-month follow-up, the nonparametric Mann-Whitney U test was applied (α = .05). A nonparameteric test was applied because the group scores of the mod-Berg and SPPB (low vs high functional mobility) were not normally distributed. The independent variable was a bivariate variable created by using a receiver operating characteristic curve to determine the cut point of 21 of the functional mobility rating score [low (= 1) vs high (= 0)]. Independent T tests were used to determine whether there were group differences in age and MMSE in participants with high versus low functional mobility. Dependent variables consisted of (1) baseline physical performance scores on the mod-Berg and SPPB and (2) spatiotemporal gait measures and stride to stride variability (coefficient of variability of stride length and swing time) during usual paced walking.
To examine whether baseline physical performance scores and quantitative gait measures differed in fallers versus nonfallers, at the 4-month follow-up, the nonparametric Mann-Whitney U test was applied (α = .05). A nonparameteric test was applied because the group scores of the mod-Berg and SPPB (fallers vs nonfallers) were not normally distributed. The independent variable was a bivariate variable for falls during the 4 months follow-up [≥ 1 fall (= 1) vs no falls (= 0)]. Independent T tests were used to determine whether there were differences in age and MMSE in fallers versus nonfallers. Dependent variables were (1) baseline physical performance scores on the mod-Berg and SPPB and (2) spatiotemporal gait measures and stride-to-stride variability (CoV of stride length and swing time) during usual paced walking.
Thirty-one older adults (ages 67–96 years, 93% female) living in 4 dementia-specific assisted living units participated in this study. As shown in Table 1, the mean age of participants was 83.6 years (SD = 7.01, and cognitive functioning was moderately to severely impaired with a mean MMSE of 12.4 (SD = 6.96; range 2–26), in all except 1 participant who was mildly demented. Three subjects were lost to follow-up (2 to death and 1 to illness). Thirty-two percent (n = 10) of participants had at least 1 fall over the 4-month follow-up period. Mean baseline SPPB was 3.6 (SD 2.8) in fallers compared with 5.6 (SD 2.9) in nonfallers. Mean mod-Berg was 19 (SD 12.6) in fallers compared with 30.2 (SD 6.8) in nonfallers. Mean baseline SPPB was 2.7 (SD 2.0) in the low function group compared with 6.7 (SD 2.5) in the high function group. Mean baseline mod-Berg was 18.2 (SD 2.5) in the low function group compared with 32.9 (SD 4.3) in the high function group.
Physical Performance Assessments and Spatiotemporal Gait Measures
Older age and lower MMSE scores were associated with greater impairment on both balance and physical performance tests, with significant bivariate correlations between mod-Berg, age (r = −0.45, P = .017), and MMSE (r = 0.44, P = .018); and between SPPB, age (r = −0.43, P = .015), and MMSE (r = 0.51, P = .003). However, neither age nor MMSE was significantly correlated with any of the quantitative gait measures. Test-retest on the GAITRite Walkway System resulted in ICC of >0.90 for spatiotemporal gait measures (Table 2).
Spatiotemporal gait measures of gait speed and cadence were positively associated with performance on the mod-Berg and SPPB (Figure 2; Table 3). In contrast, percentage of gait cycle in double stance, stride time variability, and stride length variability were negatively associated with the performance on the mod-Berg and SPPB (Figure 2; Table 3). After adjusting for the age and MMSE, partial correlations between spatiotemporal gait measures and the mod-Berg as well as the SPPB remained statistically significant and all were in the expected direction (Table 3).
Low Versus High Functional Mobility Rating
Participants with low functional mobility ratings had significantly lower scores on the mod-Berg (P ≤ .001) and SPPB (P ≤ .001). In addition, participants with low functional mobility ratings demonstrated slower gait speed, lower cadence, higher swing time variability, and a greater percentage of gait cycle in double support (P ≤ .01) (Table 4). MMSE was significantly lower (P = .005), but the age did not differ significantly (P = .14) in the low versus high functional mobility group.
Fallers Versus Nonfallers
Participants with at least 1 fall, compared with those who did not fall over the 4-month follow-up, had a lower mean score on the mod-Berg (P = .02), lower cadence (P = .048), and greater stride length variability (P = .035). Differences in percentage of gait cycle in double support and gait speed were not statistically different in fallers compared to nonfallers (Table 5). There were no statistically significant differences in age (P = .30) or MMSE (P = .19) between fallers and nonfallers.
The results of this study demonstrate a strong association between physical performance assessments (mod-Berg and SPPB) and spatiotemporal gait parameters after adjusting for age and MMSE. In addition, excellent ICC values of spatiotemporal gait parameters were demonstrated in a small subgroup of subjects in this study. These findings, as well as good test-retest reliability of spatiotemporal gait parameters reported in previous studies of older adults with dementia, provide support for concurrent validity between the SPPB and mod-Berg and spatiotemporal gait parameters. Therefore, the SPPB and mod-Berg scores were strongly correlated with reliable spatiotemporal gait parameters that have been associated with instability and increased fall risk in previous studies.23,28,29
The results of this study also demonstrate that lower baseline scores on the mod-Berg and the SPPB were associated with low functional mobility at the 4-month follow-up. This finding is consistent with a study in community-dwelling older adults that demonstrated that both the Berg Balance Scale and the SPPB were predictive of ADL difficulty.30 Although the SPPB has been shown to be predictive of short-term mortality and nursing home admissions in community-based older adults,31 its relationship to adverse events in people with dementia has not previously been reported.
Our findings of lower cadence (fewer steps per min) in fallers compared with nonfallers is consistent with a previous study involving a 1-year prospective examination of falls in assisted living residents with dementia, in which lower cadence during usual paced walking was reported in fallers compared with nonfallers.28 We also found that stride length variability was greater in fallers compared with nonfallers. This finding is consistent with studies showing associations between greater stride length variability and instability in community-dwelling older adults.23 Although our study examined only usual paced walking without distractions (single task walking), previous studies report that gait instability worsens under conditions that are distracting, especially when cognitive processes needed for attention and executive functions are challenged (dual task walking).32–34
We demonstrated that physical performance assessments and quantitative gait analysis using an electronic walkway system were feasible in older adults with moderate to severe dementia within an assisted living residential unit. The assessments were conducted in a familiar and minimally distracting environment, with repeated verbal cueing to facilitate accurate assessment. Gait speed, Timed Up and Go, and the 6-Minute Walk Tests were examined by Ries et al.35 across individuals from mild to moderate dementia.35 Very high ICC values were reported for all 3 tests. The authors discuss the role of providing extra instructions and demonstrations to ensure that verbal instructions do not interfere with the participant's ability to complete the task. Individual spatiotemporal gait parameters required minimal instruction and were reliable in our study.
Previous studies have reported feasibility and reliability of balance measures in older adults with mild to moderate dementia; however, participants in these studies had a higher level of cognitive function (mean MMSE = 17−21) than in our study (mean MMSE = 12). Since physical function is known to decline with increasing severity of dementia8 further studies of reliability and validity are needed to evaluate modified tests, such as the mod-Berg, across stages of dementia. Although the GAITRite Walkway System demonstrated test-retest reliability in our study, use of an electronic walkway is not feasible in many home or clinical environments. Concurrent validity demonstrated between quantitative gait parameters, associated with instability and falls, and physical performance tools in this study support the application of the mod-Berg and SPPB in more advanced stages of dementia. Further research in this area has implication for improving the care of older adults with dementia by developing brief tools to identify and evaluate individuals who can benefit from physical therapy interventions to improve functional mobility and prevent falls.
Further studies with larger samples and longer follow-up would be beneficial. The Berg Balance Scale, modified by omitting 3 items, may have altered psychometric properties. However, the modified scale remained significantly correlated with spatiotemporal gait parameters and was associated with functional mobility limitations as well as falls. Modification of the Berg also allowed for a shorter assessment with less equipment and fewer instructions. Reminders were provided as needed to stay on task, but the standard instructions for the SPPB allow for repeating instructions and counting repetitions. While cues were provided to accommodate for attention and memory deficits, we were careful not to influence pacing of the activity. We do not have detailed health history or demographic data, such and body mass index, on participants. Finally, the results of this study apply specifically to older adults living in dementia-specific assisted living residences.
The results of this study provide support for the application of brief physical performance assessments to identify functional mobility limitations and fall risk in dementia-specific assisted living residents. People who fell over a 4-month period had lower scores on a modified Berg Balance Scale. People with low functional mobility, as rated by nursing supervisors, had significantly lower scores on both the mod-Berg and the SPPB. Spatiotemporal gait impairments were associated with physical performance tests as well as functional mobility limitations and falls. The mod-Berg and SPPB may provide valuable information for physical therapists in making recommendations for institutional strategies and individual care plans targeting functional mobility enhancement and fall prevention within dementia-specific assisted living centers.
We thank Kenneth C. Pike, PhD, for statistical support, our research coordinator, Amy Moore, our skilled staff, Becky Archer and Ray Houle, our partners at Washington State Aging and Disability Services Administration, Candace Goehring and Lynne Korte, our participating sites, The Hampton, lodge at Roo Lan, Marine Courte, Orchard Pointe, and most importantly the individuals with dementia who contributed their time and energy to this project.
1. Allan LM, Ballard CG, Rowan EN, Kenny RA. Incidence and prediction of falls in dementia: a prospective study in older people. PLoS One. 2009;4(5):e5521.
2. Kenny AM, Bellantonio S, Fortinsky RH, et al. Factors associated with skilled nursing facility transfers in dementia-specific assisted living. Alzheimer Dis Assoc Disord. 2008;22(3):255–260.
3. O'Keeffe ST, Kazeem H, Philpott RM, Playfer JR, Gosney M, Lye M. Gait disturbance in Alzheimer's disease: a clinical study. Age Ageing. 1996;25(4):313–316.
4. Beauchet O, Allali G, Berrut G, Hommet C, Dubost V, Assal F. Gait analysis in demented subjects: Interests and perspectives. Neuropsychiatr Dis Treat. 2008;4(1):155–160.
5. Morris JC, Rubin EH, Morris EJ, Mandel SA. Senile dementia of the Alzheimer's type: an important risk factor for serious falls. J Gerontol 1987;42(4):412–417.
6. Tinetti ME, Kumar C. The patient who falls: “it's always a trade-off.” JAMA. 2010;303(3):258–266.
7. Tornatore JB, Hedrick SC, Sullivan JH, Gray SL, Sales A, Curtis M. Community residential care: comparison of cognitively impaired and noncognitively impaired residents. Am J Alzheimers Dis Other Demen. 2003;18(4):240–246.
8. Buchman AS, Boyle PA, Leurgans SE, Barnes LL, Bennett DA. Cognitive function is associated with the development of mobility impairments in community-dwelling elders. Am J Geriatr Psychiatry. 2011;19(6):571–580.
9. van Iersel MB, Benraad CEM, Olde Rikkert MGM. Validity and reliability of quantitative gait analysis in geriatric patients with and without dementia. J Am Geriatr Soc. 2007;55(4):632–633.
10. Berg KO, Wood-Dauphinee SL, Williams JI, Maki B. Measuring balance in the elderly: validation of an instrum ent. Can J Public Health. 1992;83(suppl 2):S7–S11.
11. Guralnik JM, Simonsick EM, Ferrucci L, et al. A Short Physical Performance Battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol. 1994;49(2):M85–94.
12. Guralnik JM, Ferrucci L, Simonsick EM, Salive ME, Wallace RB. Lower-extremity function in persons over the age of 70 years as a predictor of subsequent disability. N Engl J Med. 1995;332(9):556–561.
13. Guralnik JM, Ferrucci L, Pieper CF, et al. Lower extremity function and subsequent disability: consistency across studies, predictive models, and value of gait speed alone compared with the Short Physical Performance Battery. J Gerontol A Biol Sci Med Sci. 2000;55(4):M221–231.
14. Beauchet O, Freiberger E, Annweiler C, Kressig RW, Herrmann FR, Allali G. Test-retest reliability of stride time variability while dual tasking in healthy and demented adults with frontotemporal degeneration. J Neuroeng Rehabil. 2011;8(1):37.
15. Wittwer JE, Webster KE, Andrews PT, Menz HB. Test-retest reliability of spatial and temporal gait parameters of people with Alzheimer's disease. Gait Posture. 2008;28(3):392–396.
16. Rao AK, Quinn L, Marder KS. Reliability of spatiotemporal gait outcome measures in Huntington's disease. Movement Disord. 2005;20(8):1033–1037.
17. Wittwer JE, Webster KE, Menz HB. A longitudinal study of measures of walking in people with Alzheimer's Disease. Gait Posture. 2010;32(1):113–117.
18. Folstein MF, Folstein SE, McHugh PR. “Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189–198.
19. Steffen TM, Hacker TA, Mollinger L. Age- and gender-related test performance in community-dwelling elderly people: Six-Minute Walk Test, Berg Balance Scale, Timed Up & Go Test, and gait speeds. Phys Ther. 2002;82(2):128–137.
20. Conradsson M, Lundin-Olsson L, Lindelof N, et al. Berg Balance Scale: intrarater test-retest reliability among older people dependent in activities of daily living and living in residential care facilities. Phys Ther. 2007;87(9):1155–1163.
21. Whitney SL, Wrisley DM, Marchetti GF, Gee MA, Redfern MS, Furman JM. Clinical measurement of sit-to-stand performance in people with balance disorders: validity of data for the Five-Times-Sit-to-Stand Test. Phys Ther. 2005;85(10):1034–1045.
22. Shumway-Cook A, Baldwin M, Polissar NL, Gruber W. Predicting the probability for falls in community-dwelling older adults. Phys Ther. 1997;77(8):812–819.
23. Verghese J, Holtzer R, Lipton RB, Wang C. Quantitative gait markers and incident fall risk in older adults. J Gerontol A Biol Sci Med Sci. 2009;64(8):896–901.
24. Hausdorff JM. Gait variability: methods, modeling and meaning. J Neuroeng Rehabil. 2005;2:19.
25. Beauchet O, Allali G, Annweiler C, et al. Gait variability among healthy adults: low and high stride-to-stride variability are both a reflection of gait stability. Gerontology. 2009;55(6):702–706.
26. Hausdorff JM, Rios DA, Edelberg HK. Gait variability and fall risk in community-living older adults: a 1-year prospective study. Arch Phys Med Rehabil. 2001;82(8):1050–1056.
27. Portney LG, Watkins MP. Foundations of Clinical Research: Applications to Practice. 3rd ed. Upper Saddle River, NJ: Prentice Hall; 2009.
28. Camicioli R, Licis L. Motor impairment predicts falls in specialized Alzheimer care units. Alzheimer Dis Assoc Disord. 2004;18(4):214–218.
29. Verghese J, Robbins M, Holtzer R, et al. Gait dysfunction in mild cognitive impairment syndromes. J Am Geriatr Soc. 2008;56(7):1244–1251.
30. Wennie Huang WN, Perera S, Van Swearingen J, Studenski S. Performance measures predict onset of activity of daily living difficulty in community-dwelling older adults. J Am Geriatr Soc. 2010;58(5):844–852.
31. Guralnik JM. Assessment of physical performance and disability in older persons. Muscle Nerve Suppl. 1997;5:S14–S16.
32. Sheridan PL, Solomont J, Kowall N, Hausdorff JM. Influence of executive function on locomotor function: divided attention increases gait variability in Alzheimer's disease. J Am Geriatr Soc. 2003;51(11):1633–1637.
33. Webster KE, Merory JR, Wittwer JE. Gait variability in community dwelling adults with Alzheimer disease. Alzheimer Dis Assoc Disord. 2006;20(1):37–40.
34. Allali G, Assal F, Kressig RW, Dubost V, Herrmann FR, Beauchet O. Impact of impaired executive function on gait stability. Dement Geriatr Cog Disord. 2008;26(4):364–369.
35. Ries JD, Echternach JL, Nof L, Gagnon Blodgett M. Test-retest reliability and minimal detectable change scores for the timed “up & go” test, the six-minute walk test, and gait speed in people with Alzheimer disease. Phys Ther. 2009;89(6):569–579.
dementia; falls; functional assessment; mobility limitation; spatiotemporal gait measures