Falls in Older People: Effects of Age and Blurring Vision on the Dynamics of Stepping

Abstract

purpose. The risk of falling increases dramatically with age, and visual impairment is known to be an important risk factor. Therefore, it is highly pertinent to assess the effects of age and vision on the performance of everyday tasks linked to falling, such as stepping from one level to another.

methods. Nine young (age, 26 ± 4 years) and ten elderly (age, 72 ± 5 years) subjects performed a stepping-up task of three different heights. Their stepping strategies with blurred and optimally corrected vision were compared. Center of mass (CM), center of pressure (CP) dynamics (in the mediolateral and anteroposterior directions), and foot clearance parameters were determined, and statistical regression modeling was applied.

results. Elderly subjects spent 20% more time (P = 0.03) than young subjects during double support and they had reduced anteroposterior CM-CP divergence (P < 0.001) during double support and slower anteroposterior (P < 0.001) and mediolateral (P = 0.002) CM velocities during initiation of movement and single limb support. Blur caused similar adaptations, such as increased toe clearance, across both age groups, though mediolateral (ML) CM-CP divergence in elderly subjects was significantly more reduced than in young subjects (P < 0.001).

conclusions. Findings indicate, in general, that older subjects used a more cautious and controlled stepping strategy. However, the lack of significant age differences in toe clearance suggests this strategy was mainly aimed at reducing ML instability rather than increasing margins of safety regarding toe clearance.

The risk of falling increases dramatically with age, with one third of community-dwelling people 65 years and older experiencing one fall each year. ^{1 2} A high proportion of falls occur on steps and stairs. ^{3 4} Because of that, research pertaining to the effects of age on balance and mobility has focused on understanding the mechanics of stepping, either up or down to a new level, and on avoiding obstacles during gait. ^{5 6 7 8 9 10 11 12} Collectively, these studies reveal that compared with young people, elderly people have a more restrained stepping pattern and consequently display slower execution times, ^{6 7 11} reduced anteroposterior (AP) ground contact force during push off and foot strike, ¹⁰ decreased AP center of mass (CM) velocity, ¹¹ and reduced AP separation between the horizontal position of CM and center of pressure (CM-CP) when traversing an obstacle. ⁸ Explanations proposed for the stepping strategy adaptations observed in older people include fear of falling, ¹¹ decrease in muscle strength ⁸ or in flexibility as a result of aging, ¹² and desire to reduce mechanical loading of the stance limb during single support. ⁸  

Most of these studies have also looked at lead limb step/obstacle clearance (the minimum distance between the foot and the obstacle or step) because of its critical importance in the safe avoidance of tripping. ^{5 6 7} Indeed, one such study has concentrated solely on this parameter. ⁹ However, many of these foot clearance findings appear contradictory, with studies suggesting foot clearance increases with age, ⁶ decreases with age, ^{5 9} or does not change. ⁷  

Visual impairment has been strongly associated with an elevated risk for falling, ^{13 14 15} and the incidence of visual impairment increases with age. ¹⁶ We have recently completed a series of studies investigating how vision affects stepping dynamics and foot clearance parameters in healthy older subjects when stepping to a new level. ^{17 18} In the initial study, we found that a twofold safety strategy was used by subjects under blurred conditions, demonstrated by a reduction in the mediolateral (ML) separation between the CM and the CP while in double support and an increase in toe clearance to reduce the possibility of tripping. ¹⁷  

The aim of the present study was to expand on our previous work examining the effects of blurred vision on stepping by investigating the role of age. We hypothesized that the elderly subjects in the present study would adopt a stepping strategy similar to that previously demonstrated under blurred conditions by reducing ML CM-CP divergence and increasing minimum toe clearance and that this would be performed by the elderly subjects to a greater extent than it was performed by the young subjects under blurred conditions. Therefore, the stepping strategies of young and elderly subjects were compared by examining CM dynamics, temporal interactions between the CM and CP, and minimum toe clearance parameters during a stepping-up task, with vision optimally corrected or blurred. 

Methods

Subjects

Nine young (6 women, 3 men; age, 25.7 ± 3.5 years; height, 170.3 ± 5.6 cm; weight, 71.3 ± 9.3 kg) and ten elderly (3 women, 7 men; age, 71.6 ± 4.9 years; height, 161.6 ± 8.7 cm; weight, 69.3 ± 8.8 kg) volunteers participated in the study. Participants were screened through an in-house self-report health status questionnaire to ensure that they had no mobility problems, such as joint replacements or previous injuries affecting their gait, and that they had no more than one comorbidity. Subjects with corrected binocular visual acuity (VA) worse than 0.0 logMAR (Snellen 20/20 equivalent) were also excluded. All subjects were freely mobile and physically active, as determined through the activity scale of the Allied Dunbar fitness survey 1992. ¹⁹ Written informed consent was obtained from all subjects, and the study gained ethical approval from the university’s Research Ethics Committee. The study was conducted in accordance with the tenets of the Declaration of Helsinki. 

Visual Assessment

Trial frames and trial case lenses were used to obtain the optimal subjective refractive correction of each subject at 4 m. Binocular VA was measured with the Early Treatment Diabetic Retinopathy Study (ETDRS) logMAR chart using a by-letter scoring system and a 4-m working distance. ²⁰ Binocular contrast sensitivity (CS) was measured using the Pelli-Robson chart at 1 m using a by-letter scoring system and a chart luminance of 200 cd/m². ²¹ Binocular VA and CS were remeasured with each participant’s vision blurred by the placement of light-scattering lenses (Vistech Consultants Inc., Dayton, OH) over the trial frame. ^{22 23} Throughout the experimental session, trial frames replaced subjects’ own spectacles to standardize the optimal correction, especially because multifocal lenses have been previously reported to increase the likelihood of falls. ²⁴  

Protocol

A five-camera, three-dimensional motion capture system (Vicon 250; Oxford Metrics Ltd., Oxford, UK) recorded (50 Hz) the subjects stepping up to a new level. 

Two force plates (AMTI OR67-1000; Advanced Mechanical Technologies, Inc., Boston, MA), FP 1 and FP 2, recorded ground reaction force (GRF) data at 100 Hz. Step heights (73 mm, 146 mm, 218 mm) corresponded to a roadside curb, an indoor stair, and a bus step, respectively. The steps were positioned directly on FP 2 (Fig. 1) , and each subject stood with both feet positioned on FP 1, behind a line that equaled half the subject’s foot length away from the step edge. A voice command to step up was given, at which the subjects stepped up and resumed a stationary position on the top of the step. For each step height, trials were repeated with vision optimally corrected either with or without the addition of light-scattering lenses (blurred condition). All trials were repeated 3 times, and the order of visual condition and step height was randomized. The leading leg was self-selected by the participants, after which they were asked to continue with the same lead limb; trials in which this did not occur were discarded and repeated. Subjects used the shoes they most frequently wore outside. An observer was present as a precautionary measure to ensure that if any subjects tripped, they did not fall and injure themselves. 

For every frame of the stepping action, the following data were exported in ASCII format (50 Hz) for further analysis: global body CM coordinates (x, y), GRF from each force platform (Fx, Fy, Fz), CP coordinate (x, y) from force platform 1, lead limb heel marker coordinates (x, y, z), lead limb’s shoe tip marker coordinates (x, y, z), and lead limb ankle flexion/extension joint angle (θy). See Figure 1for the coordinate reference system and refer to previous work for detailed definitions. ^{17 18}  

Statistical Analysis

The stepping action was divided into discrete phases: anticipatory, initial swing, terminal swing, and weight transfer. ¹⁷ Data were analyzed using a random effects population averaged model (Stata version 7.0; Stat Corp., College Station, TX). This multivariate model was obtained using the generalized least squares (GLS) random-effects estimator, which produces a matrix-weighted average of between-subjects and within-subjects results. An exchangeable correlation structure was judged to be appropriate, given the experimental design, and because of the exploratory nature of the study, no type 1 error adjustment of the alpha level was deemed necessary. Thus, level of significance was set at P < 0.05. Factors of interest were incorporated sequentially, and their statistical significance was tested using a likelihood ratio test. Factors with a P < 0.1 were provisionally retained, whereas those greater than 0.1 were dropped. The final model adopted was the most parsimonious one that was felt to adequately explain the data. A t test revealed that young subjects were significantly taller than elderly subjects (P = 0.01); therefore, subject height was deemed to be a confounding variable and was added as a confounder to the model. There was no difference in mass between the two groups (P > 0.1). P values quoted in the text of the paper are those associated with the specific terms in the final regression model, which were:

1.  
   Vision: a fixed factor with 2 levels—normal (optimal correction) and blurred (diffuse blur);
2.  
   Repetition: a fixed factor with 3 levels—trials one, two, and three;
3.  
   Step height: a fixed factor with 3 levels—low, medium, and high step;
4.  
   Phase: a fixed factor with four levels—anticipatory, initial swing, terminal swing, and weight transfer;
5.  
   Age: a fixed factor with two levels—young and elderly;
6.  
   Subject height: a confounding factor.

Results

Group mean VA for young subjects was −0.14 ± 0.09 logMAR (Snellen equivalent 20/14), and mean contrast sensitivity (CS) was 1.88 ± 0.06 logCS, with vision optimally corrected. Group mean VA for elderly subjects was −0.08 ± 0.04 logMAR (Snellen equivalent 20/16), and mean CS was 1.71 ± 0.09 logCS. Under the blurred condition, VA and CS were reduced for both age groups by 0.14 to 0.18 logMAR VA and by approximately 0.70 logCS. Mean (±1 SD) values for foot clearance parameters and CM dynamics per age group, vision condition, and phase are shown in Table 1 . Repetition had no effect on any of the variables of interest (P > 0.1). 

Effects of Age on Stepping

Temporal Parameters.

Age had no effect on duration across the phases (P = 0.23), but there was a significant age-phase interaction (P = 0.007). This indicated that elderly subjects spent more time in the anticipatory (7%) and weight transfer (13%) phases than did young subjects, but duration times were similar during the swing phases. 

Foot Clearance Parameters.

Although elderly subjects displayed less vertical (−3.8 mm) and greater horizontal toe clearance (+8.3 mm) than young subjects, neither value was statistically significant (P = 0.83 and P = 0.36, respectively). 

CM Dynamics.

Elderly subjects had significantly less AP CM-CP divergence (−8.7 mm), regardless of visual condition or step height (P < 0.001). An age-phase interaction indicated that elderly subjects reduced AP CM-CP divergence during both the anticipatory (−11.4 mm) and the initial (−6.1 mm; P = 0.01) swing phases. ML CM-CP divergence showed no significant main effect for age (P = 0.54). 

Age had a significant effect on the CM AP peak velocity in all four phases (P < 0.01), indicating that across step heights and visual conditions, velocity in elderly subjects was on average 40.2 mm/s slower than it was in young subjects. There was a significant age-phase interaction for AP (P < 0.001) and ML (P = 0.001) CM velocity, indicating that the reduction in AP velocity in elderly subjects was greater in each of the phases, with differences increasing from phase 1 to phase 4. In addition, elderly subjects had slower (by 148 mm/s) ML velocity in the anticipatory phase yet faster ML velocity (by 4.6 mm/s) during the weight transfer phase. The finding that elderly subjects had significantly reduced CM AP peak velocity in each phase, even though phase duration was only increased in the initial and weight transfer phases, may seem contradictory. However, AP CM velocity was determined as the peak, not the average, value occurring in each phase; average values might have been similar. 

Effects of Blur on Stepping

Duration of all four phases increased with blur in both age groups (by 0.04 s; P < 0.001). Minimum toe clearance increased with blur in both groups in both the horizontal (19.7 mm; P < 0.001) and the vertical (18.6 mm; P < 0.001) directions. ML CM-CP divergence was reduced significantly by 5.9 mm because of blur (P < 0.001), primarily because of a significant reduction in ML CP displacement (−8.6 mm; P < 0.001). Blur had no effect on CM AP velocity but led to a reduction in CM ML velocity of 11.3 mm/s (P < 0.001) in both groups. However, a significant age-blur interaction indicated that during blur in elderly subjects, ML CM-CP divergence decreased by an average of 8.2 mm compared with a 3.6-mm reduction during blur in the young group (P < 0.001). 

Discussion

Findings indicate that elderly subjects, irrespective of step height and visual condition, spent significantly longer time in the anticipatory and weight transfer phases than did young subjects, but there was no difference between the age groups in duration of initial and terminal swing phases (P = 0.007). The swing phases constituted the period of single support, whereas the anticipatory and weight transfer phases were periods of double support. It has recently been shown that during the initial period of double support, the central nervous system (CNS) organizes postural adjustments to control, in a ballistic manner, the future motion of the body during the subsequent single support phase. ²⁵ Therefore, the increased anticipatory phase duration indicates that elderly subjects needed more time to prepare and accurately judge how much “throw” to give to the CM to move it the required distance laterally toward the supporting limb during single support. Furthermore, the reductions in anticipatory phase ML (P = 0.001) and AP (P < 0.001) peak CM velocity and in AP CM-CP divergence (P = 0.01) indicate that the throw given to the CM was achieved in a more cautious manner, presumably to ensure single support would be completed safely. 

The findings relating to the effects of blur, which corroborate the findings of our previous study, ¹⁷ indicate that ML CM-CP divergence and ML CP displacement decreased when vision was blurred, whereas phase duration increased. An age-blur interaction in ML CM-CP divergence (P < 0.001) identified that, although blur caused a reduction in ML CM-CP divergence during the anticipatory phase in both age groups, divergence was reduced more significantly in elderly subjects than in young subjects (elderly subjects, 8.2 mm; young subjects, 3.6 mm). This finding not only further supports the notion that elderly subjects required more time to organize the postural adjustments necessary to initiate movement (particularly when sensory information was disrupted) but that elderly subjects perceived ML instability as a greater threat than young subjects did. A sideways fall has been shown to be a potent risk factor for hip fractures in older persons. ²⁶ Given that sideways imbalance is harder to correct than forward imbalance, where taking a step forward can be undertaken, older subjects appeared to adopt a strategy to minimize their risk for ML instability. 

We previously determined that when the vision of healthy older persons is blurred, they increase toe clearance when stepping to a new level. ¹⁷ This was seen as a safety strategy to increase margins of error, thus acting as a safeguard against potential trips caused by step-edge contact. In the present study, it was expected that young subjects would also implement this safety strategy but that the elderly subjects, because they were more cautious, would do so to a greater extent. Findings indicated a significant effect for blur across both age groups, but age had no significant main effect on either horizontal or vertical toe clearance, and there was no age-blur interaction effect. 

The lack of an increase in toe clearance in the older group may be linked to an adaptive response toward minimizing single support duration, which is a period of increased instability during stepping, particularly when stepping up to the highest step height. ²⁷ It follows that executing a larger toe clearance, to produce the expected safety margin for reducing tripping error, would require greater time spent in single support and ultimately would prolong the period of greatest instability. Alternatively, the lack of an age-related increase in toe clearance could be related to the significant reduction in ML CM-CP divergence determined for the older group. This reduction in sideways momentum of the body would mean more energy would be required to lift the swing limb to ensure adequate step-edge clearance because elevation of the swing limb occurs with a sideways tilting motion of the body. ²⁸ These findings suggest that to minimize fall risk and to ensure safety during stepping up to a new level, older persons have to organize a tradeoff between stability-orientated safety strategies and tripping-avoidance strategies. The significant reduction in ML CM-CP divergence and the lack of an age-related difference in toe clearance infers that older subjects perceived ML instability as the greater threat. Having to organize such a tradeoff may explain the increased duration in the anticipatory phase for older subjects. It may also shed some light on the high rates of falls that result from tripping incidents, especially in those older people who, after implementing reduced safety-driven ML dynamics, cannot meet the additional energy demand to lift the swing limb sufficiently to clear the step edge. 

Because of the difference in gender composition between the two age groups, multivariate regression modeling was repeated with the inclusion of gender as an outcome factor. Only AP CM velocity was significantly different between men and women (P = 0.001), with a faster average CM velocity in women (by approximately 29 mm/s). However, because of the small number of subjects, it was felt that this finding could not be conclusive. Similarly, subject height, which was included as a confounder variable because of the significant difference in height between groups, was found to have little effect on findings. We decided not to include body mass index as a confounder because of our belief that subject height should be treated as an independent confounder. If subject height were substituted for BMI values, we could not have explored the effect of subject height because BMI calculations effectively normalize height to weight. 

Given that this was an exploratory study investigating several outcome measures, the power of the study was not addressed because it was unclear which outcome measure should be chosen as the primary outcome measure for the purpose of a power calculation. Another factor that made prospective power calculation difficult was the lack of readily available techniques for power/sample size estimation for multivariate repeated measures models. Although we believe the power of the study did not influence how results were interpreted, we recognize that clinically important effect sizes might have been missed if indeed the study was underpowered. 

In conclusion, the present study confirms to a certain extent our main hypothesis, which is that elderly subjects used a more cautious, controlled stepping strategy with longer time taken in the anticipatory phase to ensure that the throw given to the CM for the subsequent single support swing phases was well within limits of safety. However, the lack of significant age differences in toe clearance suggests that adopting this strategy was mainly aimed at reducing ML instability rather than increasing margins of safety regarding toe clearance. 

Because of the rigorous selection criteria, elderly subjects were representative of healthy, active older people who are free of common age-related ailments; therefore, the findings of the study may be conservative estimations of the stepping adaptations within an older population. 

 Supported by Grant 3991/882 from The Health Foundation United Kingdom.

 Submitted for publication January 17, 2005; revised May 16 and June 21, 2005; accepted August 22, 2005.

 Disclosure: K. Heasley, None; J.G. Buckley, None; A. Scally, None; P. Twigg, None; D.B. Elliott, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Step dimensions. Subjects started with their feet placed behind the dashed line, equivalent to a distance half their own foot length away from the edge of the step.

Table 1.

 

Mean (± 1 SD) Values for Minimum Toe Clearance, Peak CM-CP Divergence, Peak CP Displacement, Peak CM Velocity, and Phase Duration

	Age Group			Vision			Phase
		Young	Elderly	Normal	Blur	1	2	3	4
X toe clearance (mm)	64	67	54	74‡
	(29)	(27)	(29)	(33)
Z toe clearance (mm)	47	43	35	55‡
	(25)	(22)	(18)	(25)
CM velocity AP (mm/s), *	329.6	277.9‡	309.2	308.2	58.4	222.8	379.0	568.1
	(204.5)	(184.3)	(193.9)	(202.7)	(25.4)	(47.1)	(86.4)	(68.1)
CM velocity ML (mm/s), *	155.1	145.7	155.9	146.6‡	115.7			187.0‡
	(52.5)	(51.3)	(52.6)	(51.4)	(35.2)			(40.8)
Divergence AP (mm), *	45.3	36.3‡	41.4	42.0	26.3	56.8		‡
	(19.5)	(20.1)	(21.1)	(19.3)	(12.7)	(13.8)
Divergence ML (mm), †	40.7	35.9	41.4	36.1‡	37.8	39.7		‡
	(11.7)	(14.0)	(12.7)	(12.4)	(14.3)	(11.1)
CP displacement ML (mm)	21.7	21.5	25.7	17.3‡
	(12.5)	(14.1)	(13.0)	(11.9)
Duration (s), *	0.41	0.43	0.40	0.44‡	0.52	0.40	0.36	0.38‡
	(0.11)	(0.14)	(0.10)	(0.15)	(0.15)	(0.09)	(0.09)	(0.09)

 Data are averaged across repetitions, step heights, and associated outcome measures to illustrate the effects of age group, vision, and phase. Main effect,

‡   P < 0.001. Age/blur interaction,

†   P < 0.05. Age/phase interaction,

*   P < 0.05.

The authors thank Caroline Wilson for her help during data collection.