Gait-related Self-efficacy is Low in Older Adults with Knee Osteoarthritis: A Preliminary Study

Abstract

Objective: To investigate the differences in self-efficacy (SE) for walking tasks between older patients with knee osteoarthritis (OA) and older adults without knee OA. Methods: A cross-sectional design was employed. Older patients with radiographic knee OA and community-dwelling older adults without knee OA as controls were enrolled in the study. SE for the walking task was assessed using the modified gait efficacy scale (mGES). A Wilcoxon rank-sum test was used to compare the mGES between the groups of participants. A Tobit regression model was used to estimate the difference in mGES. The presence of radiographic knee OA was used as an independent variable. Sex (women), age, and body mass index were used as potential confounding variables in the model. Results: After exclusion, 78 participants (n=40 with knee OA, n=38 controls) were included. The mGES was lower in patients with knee OA than in controls. In the Tobit regression model adjusted for confounding factors, mGES in patients with knee OA was estimated to be 26.8 (95% confidence interval [CI]: 15.8-37.8) points lower than in controls. Conclusion: This study demonstrated that mGES was lower in older patients with knee OA than in older adults without knee OA.

Knee osteoarthritis (OA) is one of the most common joint diseases and is clinically characterized by chronic pain, functional disability, and limitations in activities of daily living. There are many studies that have investigated physical problems such as gait kinematics in patients with knee OA¹⁾. However, recent studies focusing on psychological aspects such as pain catastrophizing and self-efficacy (SE) have also been investigated^2,3). Several SE scales have been used for patients with knee OA, including fear of falling and SE for disease-specific symptoms, such as the fall efficacy scale and the arthritis self-efficacy scale, respectively^3,4). The modified gait efficacy scale (mGES), which is specifically used for self-reporting confidence in carrying out 10 common walking tasks, was developed⁵⁾. Walking is the most important for mobility and is essential for many daily activities. Walking-related symptoms (i. e., pain) and dysfunction are common problems for patients with knee OA; therefore, SE scales for walking tasks will be very informative. Consequently, reductions in SE for walking tasks can adversely affect various activities of daily living, occupations, as well as general physical activity (PA). However, it is unclear how much lower the SE for walking tasks is in patients with knee OA when compared to without knee OA. The purpose of this preliminary study was to provide basic data on the SE for walking tasks among patients with knee OA, comparing it with that in people without knee OA.

Materials and Methods

Study design and ethics

The present study employed a cross-sectional design. All participants underwent the following measurements: measurement of SE for the walking task, objective measurement of PA, and measurement of gait speed. The present study was approved by the medical ethics committee of Akita City Hospital, 2017 (approval No. 6), and the study was conducted in accordance with the Declaration of Helsinki. The objectives and procedures used in the study were orally explained to the participants, and explanatory documents were also provided. Written consent was obtained from the participants after they were informed that participation in the study was voluntary and that their privacy would be protected.

Participants

Patients aged 65 years or older with radiological knee OA (Kellgren-Lawrence [K-L] grade ≥ 2) who visited the Department of Orthopedics at Akita City Hospital between April 2017 and March 2019 were enrolled. As a control group, community-dwelling older adults were recruited by poster at a local community center as an opportunity to assess physical performance. Participants as the control group who had symptoms of knee pain or had been diagnosed with knee OA were excluded from the analysis. The common exclusion criteria for participation were as follows: diagnosis of severe and/or unstable cardiac diseases, hip or ankle osteoarthritis, mental disorders, disorders other than OA that could affect PA level (i.e., chronic obstructive pulmonary disease), not living independently, being wheelchair-bound, and the inability to operate the activity monitor.

Measurements

Self-efficacy for walking task

The SE for the walking task was assessed using the Japanese version of the mGES, which has been established reliability^5,6). The mGES, which represents the SE for performing safe walking tasks, consists of 10 items that address an individual's perception of their level of confidence in each task. The items include level-ground walking, climbing stairs, and long-distance walking. The items were scored individually on a 10-point scale, with one indicating no confidence and 10 indicating perfect confidence. The total score ranged from 10 to 100 points.

Other characteristics

A ten-meter walking test was performed to measure gait speed (GS). The participants were instructed to walk at a comfortable pace. The test was performed twice for each subject, with a 10-second break between the trials. The faster speed of the two trials was accepted as GS. The intensity of knee pain for level-ground walking was assessed after the trials using the visual analog scale (VAS) in patients with knee OA.

The objective measurement of PA was performed using an accelerometer-based activity monitor (Lifecorder GS4, Suzuken Co., Ltd., Tokyo, Japan). The validity and reliability of this activity monitor has been previously demonstrated⁷⁾. Participants wore the monitor on the belt or waistband of their clothing for at least nine consecutive days. The device was worn throughout the day, except during bathing and at bedtime, and data on the participants' daily activities for at least 7 days were obtained. Days with at least 10 hours of wearing time were considered valid⁸⁾. The participants' average daily step count was calculated as their PA level.

The participants' sex, age, height, weight, and body mass index (BMI) were assessed or retrieved from medical records. The K-L grade and femorotibial angle (FTA) were assessed by a physician using a full-length standing radiograph from the hip to the ankle joints in patients with knee OA.

Statistical analysis

Statistical analysis was performed using R ver. 3.6.1 (The R Foundation for Statistical Computing, Vienna, Austria) and RStudio ver. 1.2.5042 (RStudio, PBC., MA, US). Statistical significance was set at P < 0.05. The assumptions of normality were assessed using the Shapiro-Wilk test. We performed Welch's t-tests or Wilcoxon rank-sum tests for interval and ordinal variables, respectively, and Pearson's chi-squared tests for categorical variables to assess differences between the two groups of participants. A general linear regression model or generalized linear regression model was used to estimate the difference in mGES between the two groups. Radiographic knee OA was adopted as an independent variable in the initial model (Model 1). In the second model (Model 2), sex (women) and age as demographic factors and BMI as obesity level were adopted as confounding variables and added to Model 1. Categorical variables (the presence of radiographic knee OA and sex) were transformed into dummy variables.

Results

Eighty-eight participants were enrolled in this study. Ten patients with knee OA were excluded for the following reasons: previous history of knee surgery (n=2), unwilling to participate in the study (n=1), inability to walk independently (n=1), and a lack of available PA data (n=6). In the control group, there were no individuals excluded from the analysis. Consequently, 78 participants (n=40 with knee OA; n=38 controls) were included in the study.

Comparisons of the clinical and demographic data for the participants in the two groups are presented in Table 1. The mGES was lower in patients with knee OA than the controls. In the controls, 6 individuals (15.8%) were full score of the mGES. Patients with knee OA were older and had higher body weight and BMI than the controls. GS was faster, and the daily step count was higher in the control group compared to the patients with knee OA.

Table 1.

Comparison of the clinical and demographic characteristics and measurement values for patients with knee osteoarthritis and controls

	All	Knee OA	Control
	(n=78)	(n=40)	(n=38)	p-value
The data are presented as mean ± standard deviation, median (25, 75 percentile), or number (%). The p-values are results of Welch’s t-test, Wilcoxon rank-sum test, or chi-squared test. The K/L grade and VAS in the control group were not evaluated. Abbreviations: OA, osteoarthritis; BMI, body mass index; K/L grade, Kellgren-Lawrence grade; VAS, visual analog scale of knee pain; mGES, modified gait efficacy scale; GS, gait speed.
Sex, women (n)	58 (74)	34 (85)	24 (63)	0.051
Age (years old)	74 ± 6	76 ± 6	72 ± 5	0.008
Height (cm)	153.2 ± 7.4	152.7 ± 7.7	153.9 ± 7.0	0.475
Weight (kg)	58.5 ± 9.8	61.4 ± 11.3	55.6 ± 6.8	0.007
BMI (kg/m2)	24.9 ± 3.3	26.2 ± 3.6	23.5 ± 2.2	< 0.001
K/L grade (n)
III		14 (35)	―
IV		26 (65)	―
VAS (mm)		55.8 ± 24.1	―
mGES	74 (40, 91)	42 (30, 72)	88 (79, 100)	< 0.001
GS (m/sec)	1.28 ± 0.37	0.99 ± 0.25	1.59 ± 0.19	< 0.001
Daily step counts (steps/day)	4716 (2691, 7004)	3093 (2256, 4950)	6486 (4690, 9595)	< 0.001

The Shapiro-Wilk test determined that a normal distribution was not present for mGES (W = 0.905, p < 0.001). The distribution of mGES was slightly skewed (−0.330) and censored (range: 16-100); some participants were at the upper limit. As a result of the above reasons, a Tobit regression model was used (Table 2). In Model 1 (chi-square = 42.165, p < 0.001), mGES was lower in patients with knee OA than in the controls (Table 2). These findings were similar for Model 2 (chi-square = 52.481, p < 0.001), which was adjusted for sex, age, and BMI. Model 2 estimated that mGES was 26.8 (95% confidence interval [CI]: 15.8-37.8) points lower in patients with knee OA than in the controls (Table 2).

Table 2.

Results of the Tobit regression analysis

			95% CI
	Variables	B	lower	upper	SE	β	z-value	p-value	AIC
The dependent variable is modified gait efficacy scale Abbreviations: B, partial regression coefficient; CI, confidence interval; SE, standard error; β，standardized partial regression coefficient; AIC, Akaike information criteria; OA, osteoarthritis; BMI, body mass index.
Model 1	Knee OA	-36.612	-46.187	-27.036	4.886	-0.647	-7.494	< 0.001	664.032
Model 2	Knee OA	-26.793	-37.778	-15.809	5.605	-0.474	-4.781	< 0.001	659.715
	Women	-10.734	-21.600	0.132	5.544	-0.166	-1.936	0.053
	Age	-1.154	-2.139	-0.169	0.503	-0.226	-2.295	0.022
	BMI	-1.352	-2.834	0.131	0.757	-0.155	-1.786	0.074

Discussion

The results of this study demonstrate that mGES was lower in patients with knee OA than in the controls. Additionally, gait speed and the daily step count were lower in patients with knee OA, as shown in previous studies⁹⁾.

Patients with knee OA who have various challenges related to walking, such as pain, poor balance and lower gait function, may also have problems with SE for walking tasks. The influence of psychological aspects, such as pain catastrophizing, has been studied. A previous study showed that pain catastrophizing is associated with physical function, activities of daily living, and PA^10-12). The current study found that patients with knee OA may have psychological problems as well as pain and poor gait function during walking. Physical function was associated with pessimism, SE for walking tasks, and activity-related balance efficacy, but the association with pessimism was attenuated after controlling for both SE in older adults¹³⁾. Additionally, one study found that the SE for walking tasks has a partially mediating role in the pathway from PA to functional limitations in older adults¹⁴⁾. Thus, the SE for walking tasks was associated with physical function and PA in older adults. Future studies are therefore warranted among patients with knee OA.

Total knee arthroplasty (TKA) is one of the methods used to improve pain and physical function in patients with knee OA. Hiyama et al. reported that although there were improvements in physical function and pain in patients 6 months after TKA, there was no difference compared with the preoperative scores for life-space mobility and mGES¹⁵⁾. These results suggest that improving pain and physical function alone is not enough to improve the SE for walking tasks and that interventions to improve confidence in the walking task itself (e.g., task-oriented training) are needed. The mGES can assess the SE for walking tasks in 10 different situations, including level surface walking, rough terrain walking, stair climbing, and more⁵⁾. Consequently, we might be able to assess not only overall confidence in walking, but also more specific problems in SE for individual walking challenges. These findings are informative for personalized, task-oriented gait training conducted by physicians, physical therapists, and others.

This study was subject to several limitations that merit mentioning here. First, because this study was conducted at a single center, the participants with knee OA had the following characteristics: 74% of the subjects were female; and all patients had severe radiographic OA changes. These limitations might have imposed selection bias and might limit the validity and generalizability of our findings. Further study is needed, particularly on mild and moderate radiographic knee OA. Second, we did not perform radiological examinations among the community-dwelling participants. Lastly, the community-dwelling participants was recruited with posters, so their interest in health may have been higher than average, and their gait speed was faster than in previous studies. There was no knee pain in controls and median mGES was comparable to that reported in previous studies⁵⁾. However, the number of participants with full score was twice that of the previous study, which may be due to differences in physical function⁵⁾. Therefore, the controls in this study may have higher physical function and the SE for walking tasks than the general community-dwelling older adults.

In summary, the present preliminary data support that older patients with severe knee OA have lower SE for walking tasks compared to older adults without knee OA. The impact of lower SE for walking tasks on physical function, PA, and patient-reported outcome measures requires further study.

Conflict of Interest

All authors report no conflicts of interest.