Dr. Keysor has received consultant fees, speaking fees, and/or honoraria (less than $10,000 each) from the Association of Rheumatology Health Professionals and the American Physical Therapy Association.
When it hurts, a positive attitude may help: association of positive affect with daily walking in knee osteoarthritis. Results from a multicenter longitudinal cohort study
Article first published online: 27 AUG 2012
Copyright © 2012 by the American College of Rheumatology
Arthritis Care & Research
Volume 64, Issue 9, pages 1312–1319, September 2012
How to Cite
White, D. K., Keysor, J. J., Neogi, T., Felson, D. T., LaValley, M., Gross, K. D., Niu, J., Nevitt, M., Lewis, C. E., Torner, J. and Fredman, L. (2012), When it hurts, a positive attitude may help: association of positive affect with daily walking in knee osteoarthritis. Results from a multicenter longitudinal cohort study. Arthritis Care Res, 64: 1312–1319. doi: 10.1002/acr.21694
- Issue published online: 27 AUG 2012
- Article first published online: 27 AUG 2012
- Accepted manuscript online: 13 APR 2012 12:07PM EST
- Manuscript Accepted: 26 MAR 2012
- Manuscript Received: 15 DEC 2011
- NIH. Grant Numbers: AG-18820, AG-18832, AG-18947, AG-19069, AR-007598, AR-47885, R01-AG-18037, R01-AG-028144, R01-AG-028556, R21-AT-002959
- National Institute of Arthritis and Musculoskeletal and Skin Diseases. Grant Number: K23-AR-055127
- American College of Rheumatology/Research and Education Foundation Rheumatology Investigator Award (White)
- Arthritis Foundation Arthritis Investigator Award (Neogi)
- Boston Claude D. Pepper Older Americans Independence Center. Grant Number: P30-AG-031679
- Foundation for Physical Therapy
While depressive symptoms and knee pain are independently known to impede daily walking in older adults, it is unknown whether positive affect promotes daily walking. This study investigated this association among adults with knee osteoarthritis (OA) and examined whether knee pain modified this association.
This study is a cross-sectional analysis of the Multicenter Osteoarthritis Study. We included 1,018 participants (mean ± SD age 63.1 ± 7.8 years, 60% women) who had radiographic knee OA and had worn a StepWatch monitor to record their number of steps per day. High and low positive affect and depressive symptoms were based on the Center for Epidemiologic Studies Depression Scale. Knee pain was categorized as present in respondents who reported pain on most days at both a clinic visit and a telephone screening.
Compared to respondents with low positive affect (27% of all respondents), those with high positive affect (63%) walked a similar number of steps per day, while those with depressive symptoms (10%) walked less (adjusted β −32.6 [95% confidence interval (95% CI) −458.9, 393.8] and −579.1 [95% CI −1,274.9, 116.7], respectively). There was a statistically significant interaction of positive affect by knee pain (P = 0.0045). Among the respondents with knee pain (39%), those with high positive affect walked significantly more steps per day (adjusted β 711.0 [95% CI 55.1, 1,366.9]) than those with low positive affect.
High positive affect was associated with more daily walking among adults with painful knee OA. Positive affect may be an important psychological factor to consider for promoting physical activity among people with painful knee OA.
Walking is the most common means by which older adults stay physically active (1) and can reduce the risk of cardiovascular disease and other adverse health outcomes (2). Daily walking, or the number of steps one takes per day, can be difficult for older adults with knee osteoarthritis (OA), since knee OA is the leading cause of functional limitations (3). While knee pain is considered the primary reason for limited walking among older adults with knee OA (4), psychological factors, notably depressive symptoms, are also associated with decreased physical activity in this population (5).
Among older adults, persons with depression have more limitations with physical functioning (6) and lower levels of physical activity (7) than persons without depression. However, the absence of depressive symptoms may not be sufficient to promote daily walking. Rather, high positive affect (i.e., the absence of depressive symptoms coupled with a sense of emotional well-being or a positive attitude) may be a determining factor. Positive affect is associated with faster walking speed (8–10) and with lower rates of frailty and mortality in older adults (11). Nevertheless, the relationship between positive affect and daily walking is not known; understanding this association is important. People with knee OA commonly have insufficient levels of physical activity (2), and the identification of factors that may promote daily walking and physical activity in general is needed.
The association of positive affect with daily walking in adults with knee OA may also differ depending on the presence of knee pain. Knee pain is common and is prevalent in 20–25% of Americans ages >60 years (12). The amount of daily walking varies among people with the same amount of knee pain, with some walking more than others (13). Positive affect may be a reason for this variation. In particular, persons with knee pain may need to draw on psychological resources, such as high positive affect, to continue daily walking habits. In contrast, positive affect may be less important for maintaining daily walking habits among OA patients without knee pain. For these individuals, obstacles to daily walking may be more related to lifestyle preferences, scheduling, and environmental and policy influences found to be associated with physical activity in the general population (14, 15).
The current study investigated the association of positive affect with daily walking in a large, multisite sample of people with knee OA. We hypothesized that respondents with high positive affect would walk more per day than those with low positive affect. Moreover, we hypothesized that the association between positive affect and daily walking would be stronger among respondents with knee pain compared with those without knee pain.
Significance & Innovations
Among persons with knee pain, those with high positive affect walked more steps per day compared to those with low positive affect.
The absence of depressive symptoms alone may not be enough to encourage daily walking among people with symptomatic knee osteoarthritis; positive affect should instead be considered.
PATIENTS AND METHODS
The study sample consisted of participants from the Multicenter Osteoarthritis (MOST) Study, a large multisite longitudinal cohort study of community-dwelling adults, ages 50–79 years, who have or are at high risk for knee OA (16). The MOST Study sample was recruited from 2 communities, Birmingham, Alabama and Iowa City, Iowa, through mass letter mailings, study brochures, and media and community outreach programs. Inclusion criteria were the presence of known risk factors for knee OA, including being age >50 years, female sex, having a previous knee injury or operation, and a body weight higher than the median weight for each age- and sex-specific group, based on data from the Framingham OA study (17).
Daily walking subsample.
The present cross-sectional study utilized data from a 60-month followup examination conducted between May 2009 and January 2011, since objective data on daily walking were only collected at this visit. The sample for these analyses was restricted to participants with knee OA based on radiographic findings in either the tibiofemoral (TF) or patellofemoral (PF) joints. For the TF joint, knee OA was defined as a Kellgren/Lawrence grade ≥2, and for the PF joint, an osteophyte score ≥2 or any joint space narrowing score ≥2 with any osteophyte, sclerosis, or cyst score of ≥1 on a lateral plain view film (18, 19). Furthermore, we restricted our sample to those participants who wore a StepWatch monitor (Orthocare Innovations) with a minimum of 3 days of valid walking data, since previous studies have found this to be the minimum number of days needed for a reliable estimate of physical activity (20, 21).
The MOST Study protocol was approved by the Institutional Review Boards at the University of Iowa in Iowa City, the University of California in San Francisco, the University of Alabama in Birmingham, and Boston University Medical Center.
Positive affect was based on the Center for Epidemiologic Studies Depression Scale (CES-D) (22), following previously published methodology (10, 23). In brief, the 20-item CES-D includes a positive affect factor, comprising 4 positively worded items: “I felt that I was just as good as other people,” “I felt hopeful about the future,” “I was happy,” and “I enjoyed life.” The participants were grouped into 3 mutually exclusive categories: depressive symptoms, low positive affect, or high positive affect. Depressive symptoms were defined as an overall CES-D score of ≥16, irrespective of the number of positive items participants reported, which is the traditional cut point for depressive symptoms. The respondents with overall CES-D scores <16 were categorized as low positive affect if they reported experiencing ≤2 positive affect items most of the time in the previous week. We classified respondents as high positive affect if they reported experiencing 3 or 4 of the items most of the time in the previous week. Therefore, nobody categorized with high or low positive affect could have depressive symptoms (CES-D score ≥16).
We measured daily walking with a StepWatch monitor, which is a small (70 × 50 × 20 mm, 38 gm), waterproof, self-contained device that is worn on the ankle and records the number of steps taken every minute, while providing no feedback to the user. The StepWatch monitor has high concurrent validity in comparison with several standard measures of step frequency in older adults and high convergent validity when compared with Short Form 36 scores among participants with OA (24).
Following the collection of data at the clinic visit where positive affect and other study variables were collected, each study participant was fitted with the StepWatch monitor and provided with written and verbal instructions for attaching the monitor each morning and removing it at bedtime for 7 consecutive days (plus part of the day that the participant received the device and the day it was returned). To determine whether the participants wore the monitor long enough to be counted as a full day, we adopted a published method for processing accelerometry data (25) and defined 10 hours of monitoring as the minimum amount of time needed for a full monitored day. The 10-hour requirement represents >66% of waking hours and has been utilized as a threshold in studies of physical activity in the general adult population (26) and in people with knee OA (25). The time worn was counted from the first step recorded in the morning to the last step recorded in the evening. To exclude the times that participants may have taken the StepWatch monitor off during the day, we omitted times where the monitor registered no steps for 180 consecutive minutes during the day (25, 27). Finally, we restricted our sample to those participants who had a minimum of 3 days of valid data, since previous studies have found this to be the minimum number of days needed for a reliable estimate of physical activity (20, 21).
We quantified walking as the average number of steps per day. We calculated steps per day by totaling the number of steps taken on each valid day of monitoring divided by the number of valid days.
Subjects were asked about knee pain during a telephone screening and again at the clinic visit (28). In particular, subjects were asked, “During the past 30 days, have you had pain, aching, or stiffness in your knee on most days?” The telephone screening and clinic visit were separated by a median of 33 days. The subjects who responded yes to this question at both visits were defined as having knee pain.
The following factors were treated as potential confounders based on their association with function in previous studies (29–31): age, sex, race (white or other), living situation (alone or with someone), education (at least some college [yes/no]), body mass index (BMI) computed from standardized weight and height assessments classified into World Health Organization categories (32), severity of knee pain measured from a visual analog scale with 0 and 100 as end points, knee strength as determined from the mean of 4 isokinetic knee extensor torque repetitions at 60° per second using Cybex exercise equipment categorized by strength of the stronger knee into sex-specific and weight-adjusted quartiles, self-reported comorbidities based on the modified Charlson comorbidity index (none or ≥1) (33), and widespread pain (absent or present) defined by the American College of Rheumatology as pain above and below the waist, pain on the right and left sides of the body, and back pain using subject-marked pain patterns on a figure of the human body (34).
We compared the characteristics of participants with different levels of positive affect by performing t-tests for continuous covariables and chi-square tests for categorical covariables. Next, we examined the levels of daily walking by using descriptive statistics. To examine the association of positive affect with daily walking, we calculated the mean number of steps for those with high positive affect, low positive affect, and depressive symptoms and their 95% confidence intervals using multivariable linear regression adjusting for potential confounders, which is equivalent to an analysis of covariance. Low positive affect was designated as a reference to examine the association of low and high positive affect (i.e., those without depressive symptoms) with daily walking. To examine if this association differed by knee pain, we examined the interaction of positive affect and knee pain with daily walking. We then repeated the analyses stratified by knee pain and adjusted for potential confounders and confirmed that this was normally distributed from visual inspection of steps per day in the sample. Whether each covariable was a potential confounder was assessed by determining if the addition of each to the regression model, containing only the positive affect variable, changed the association between positive affect with daily walking by ≥10% (35). The following variables were evaluated as potential confounders: race, education, strength, comorbidity, and widespread pain. All of these covariables met the criteria for a confounder and were included along with age, sex, BMI, site (Alabama or Iowa), and knee pain severity in multivariable linear models.
Of the 2,330 participants who completed the 60-month clinic visit, 1,343 had radiographic knee OA and a measure of positive affect and were eligible for the study. Of these participants, 83% (n = 1,116) received a StepWatch monitor and 93% (n = 1,018) wore it for at least 3 days and were included in the analyses. Of the 229 who did not receive a StepWatch monitor, 72% refused, 16% had medical or physical impairments preventing use, 7% had no device available to them, and 5% had other reasons. The majority of the participants included in the study were women (60%) and white (89%) (Table 1). Their mean ± SD age was 63.1 ± 7.8 years, and their mean ± SD BMI was 31.7 ± 6.3 kg/m2. Generally, participants who were not included were more likely to have worse health indicators (higher BMI, less strength, more widespread pain, more knee pain, and higher CES-D scores) compared to participants included in the analyses (Table 1). The mean ± SD number of steps taken per day was 8,394.5 ± 3,469.0 (interquartile range 5,888–10,359).
|Included (n = 1,018)||Not included (n = 325)||P|
|Age, mean ± SD years||63.1 ± 7.8||63.0 ± 8.1||0.81|
|At least some college education||43||43||0.96|
|BMI, mean ± SD kg/m2||31.7 ± 6.3||33.4 ± 7.4||0.0006|
|Strength, mean ± SD Nm/kg||0.95 ± 0.40||0.84 ± 0.38||0.006|
|From Alabama vs. Iowa||39||85||< 0.0001|
|Frequent knee pain||41||49||0.02|
|Knee pain severity (VAS range 0–100), mean ± SD||23.7 ± 22.4||32.9 ± 27.8||< 0.0001|
|CES-D, mean ± SD||6.6 ± 6.8||9.5 ± 8.6||< 0.0001|
|High positive affect||63||52|
|Low positive affect||27||27|
Association of positive affect with daily walking.
Respondents with high positive affect walked the greatest number of steps per day (mean ± SD 8,560.3 ± 3,317.5), followed by respondents with low positive affect (mean ± SD 8,401.8 ± 3,864.2) and those with depressive symptoms (mean ± SD 7,305.9 ± 3,083.9). In unadjusted analyses, respondents with high positive affect walked an average of 158.5 more steps per day compared with those with low positive affect, although this was not statistically significant. Those with depressive symptoms walked 1,095.9 fewer steps per day, on average, which was statistically significant. After adjusting for confounders, high positive affect was not associated with walking more steps per day and depressive symptoms were associated with walking ∼579 fewer steps per day compared with those with low positive affect, although this difference was not statistically significant (Table 2).
|No.||Steps per day, mean ± SD||β (95% CI)||Adjusted β (95% CI)†|
|High positive affect||640||8,560.3 ± 3,317.5||158.5 (−327.9, 644.9)||−32.6 (−458.9, 393.8)|
|Low positive affect||279||8,401.8 ± 3,864.2||Ref.||Ref.|
|Depressive symptoms||99||7,305.9 ± 3,083.9||−1,095.9 (−1,889.0, −302.8)||−579.1 (−1,274.9, 116.7)|
|Age, 10-year increase||−129.3 (−156.3, −102.3)|
|Sex, men vs. women||576.2 (189.7, 962.6)|
|Race, white vs. nonwhite||38.8 (−612.5, 690.1)|
|<25, healthy weight||Ref.|
|25–29.9, overweight||−764.6 (−1,400.1, −129.2)|
|30–34.9, obese class I||−1,826.2 (−2,479.2, −1,173.2)|
|≥35, obese class II to III||−3,010.6 (−3,708.3, −2,312.9)|
|Comorbidity, ≥1 vs. none||−588.0 (−981.2, −194.9)|
|Knee extensor strength|
|Middle tertile||−361.2 (−901.7, −179.3)|
|Weakest tertile||−1,066.6 (−1,624.7, −508.6)|
|Knee pain severity, 10-point increase||−14.7 (−23.7, −5.7)|
|Widespread pain, yes vs. no||−232.3 (−163.6, 628.2)|
|Study site, Iowa vs. Alabama||1,125.1 (709.6, 1,540.6)|
Association of positive affect with daily walking stratified by knee pain.
The interaction of positive affect by knee pain was statistically significant (P = 0.0045). Approximately 39% of the respondents reported knee pain. Among these respondents, those with high positive affect walked 1,157.7 more steps per day, which was statistically significant, and those with depressive symptoms walked 59.5 fewer steps per day, which was not statistically significant compared with respondents with low positive affect. After adjusting for confounders, including knee pain severity, high positive affect remained significantly associated with walking >700 steps per day more than respondents with low positive affect.
Among respondents without knee pain, high positive affect was not associated with daily walking in adjusted analyses. Participants with depressive symptoms walked 1,197 fewer steps per day compared with those with low positive affect, which was statistically significant after adjustment for confounders (Table 3).
|No.||Steps per day, mean ± SD||β (95% CI)||Adjusted β (95% CI)†|
|Knee pain (n = 393)|
|High positive affect||235||8,541.5 ± 3,268.0||1,157.7 (431.2, 1,884.1)||711.0 (55.1, 1,366.9)|
|Low positive affect||109||7,383.8 ± 3,090.8||Ref.||Ref.|
|Depressive symptoms||49||7,324.3 ± 3,007.3||−59.5 (−1,137.7, 1,018.7)||85.2 (−871.6, 1,042.1)|
|No knee pain (n = 588)|
|High positive affect||367||8,646.3 ± 3,403.6||−514.7 (−1,201.7, 172.3)||−523.7 (−1,129.0, 81.5)|
|Low positive affect||151||9,168.5 ± 4,211.8||Ref.||Ref.|
|Depressive symptoms||44||7,405.1 ± 3,277.1||−1,755.9 (−2,975.3, −536.6)||−1,196.7 (−2,278.6, −114.8)|
We found that positive affect was not associated with more daily walking among adults with knee OA overall. However, in the presence of knee pain, respondents with high positive affect walked 711 more steps per day compared with those with low positive affect, representing ∼8.5% more steps from a mean of 8,400 steps per day. Increasing daily walking by this amount is clinically meaningful since it could mean attaining a higher level of physical activity based on pedometer-determined physical activity indices (e.g., being sedentary or at a low active level of physical activity) (36). In contrast, respondents without knee pain walked a similar amount per day whether or not they had high or low positive affect. These findings suggest that the absence of depressive symptoms may not be sufficient to promote daily walking in people with knee pain. Rather, high positive affect is an important psychological feature associated with daily walking and may be needed to increase daily walking among people with both knee OA and knee pain.
Our findings highlight the fact that knee pain modifies the relationship between positive affect and walking among adults with knee OA. Positive psychological factors could be particularly important when health is challenged, and may be why positive affect was associated with daily walking in respondents with knee pain, as opposed to those without knee pain. These findings support the broaden and build theory, which hypothesizes that persons with positive affect are able to better marshal a broad range of physical and psychological resources in response to stress compared to persons without positive affect (37, 38). For example, it may be that people with high positive affect are better able to cope with and work through knee pain and achieve more daily walking compared to people with low positive affect. However, given the cross-sectional nature of our study, we could not determine the temporal relationships between positive affect and walking; longitudinal studies are needed to better understand this association.
Our results are consistent with previous observational studies that found protective effects for positive affect with health-related outcomes in older adults (10, 11, 37). Existing evidence suggests that negative psychological attributes, such as fear-avoidance beliefs and catastrophizing, influence functional outcomes in adults with low back pain (39). Our findings extend this evidence from previous research by suggesting that positive psychological attributes influence beneficial health behaviors among adults with knee OA. These findings have important clinical implications. In contrast to the theories that view positive affect as a trait, and therefore unlikely to be modifiable, recent intervention studies have increased positive affect in persons with cardiopulmonary disease (40) and with newly diagnosed human immunodeficiency virus (41). Moreover, recent trials have shown that interventions aimed at increasing positive affect resulted in increased physical activity in people with asthma (42) and adults after percutaneous coronary interventions (43), as well as increased adherence to medication in African Americans with hypertension (44). These studies suggest that people with OA could be taught techniques to increase positive affect and other positive emotions, which might improve walking and other health behaviors, in addition to psychological well-being.
We found that depressive symptoms were associated with less daily walking among respondents without knee pain, which is consistent with previous studies (45, 46); however, we did not find this association among people with knee pain. One possible reason for this is that walking decreased overall for people with knee pain and there was little difference in walking between those with low positive affect and depressive symptoms. For instance, in the crude analysis, people with low positive affect walked 9,168.5 steps per day without knee pain and 7,383.8 steps per day with knee pain. For those with depressive symptoms, there was less difference between those with and without knee pain. Therefore, using persons with low positive affect as the reference group may have led to larger observed differences in walking in the group with depressive symptoms without knee pain and smaller differences in the group with depressive symptoms and knee pain. Nonetheless, the crude analyses showed that persons with depressive symptoms walked less than the other respondents in each subgroup, suggesting that persons with depressive symptoms and knee OA walked less than other respondents, regardless of the amount of knee pain.
The observed step counts in this sample were comparable to those from previous studies in older adults measured with a StepWatch monitor. For instance, Cavanaugh et al reported an average of 9,981 steps per day among healthy older adults and 7,681 steps per day among older adults with self-reported functional limitations (47). Our sample took on average 8,395 steps per day, fitting between the healthy and functionally limited groups. This result was expected, given that our sample comprised respondents with and without functional limitations (48). Furthermore, it is important to note that the steps per day collected in our study cannot be easily compared with previous studies utilizing pedometers. Pedometers have been shown to underestimate the number of steps taken at slow speeds up to 33% when compared to a StepWatch monitor (49). Therefore, the step counts in our study are higher than those reported for pedometer-based studies.
The limitations of our study should be acknowledged. First, 24% of the MOST Study cohort with radiographic knee OA who attended the 60-month clinic visit chose not to wear the accelerometer or had insufficient data for these analyses. Those included in our study were more likely to have better health indicators compared to those not included in the analysis; therefore, our study findings may have limited generalizability to all people with knee OA. Second, we defined the presence of knee pain using the presence of pain at 2 time points, a clinic visit and a preceding telephone screening. We performed a sensitivity analysis defining knee pain using severity measured on a visual analog scale with a cut point of 10 of 100 and found similar results. Furthermore, we repeated the analysis including all study participants, irrespective of whether they had radiographic knee OA, and found similar results among those with knee pain defined both ways. Third, in addition to knee pain, lower body pain at the feet, ankles, or hips can also alter daily walking. Although we adjusted for widespread pain in our analyses, this may not have fully accounted for lower body pain. A sensitivity analysis adjusting for the presence of lower body pain was performed and we found similar results. Fourth, the participants may have changed their daily walking habits because of the knowledge that their habitual walking was being recorded. Previous research suggests that this testing effect is greatest when participants wear an unsealed monitor (i.e., when participants are aware of how many steps are being recorded) (50). We believe that any increases in daily walking due to a testing effect were minimized, since the StepWatch monitor did not display recorded data to study participants. Fifth, other measures of positive affect exist (51); however, these measures were not available within the MOST Study 60-month examination. Using questions from the CES-D has been found to be a valid measure of positive affect (52) and has been employed as such in previous studies (10, 37). Last, our study was cross-sectional in design, which precluded inferring causality from our findings.
Despite these limitations, our study has several important implications. Our findings support the notion that positive affect is associated with daily walking, although this association appears dependent on the presence of knee pain. Furthermore, people with high positive affect and low positive affect walk considerably more if they do not have knee pain. Therefore, in the context of knee pain, high positive affect enhances walking among persons with knee pain. Thus, future studies examining the association of psychological health with physical activity or other health outcomes in older adults should consider the effect of pain as a potential modifier.
In conclusion, we found that high positive affect was associated with daily walking among adults with knee OA and knee pain. Given that physical activity is recommended to reduce the risk of poor health outcomes in adults, it is important to identify factors, such as high positive affect, to promote an active lifestyle. Our findings suggest that the absence of depressive symptoms may not be enough to promote daily walking and that high positive affect may be an important factor among older adults with knee pain. Longitudinal studies are needed to investigate the temporal relationship between positive affect and daily walking among older adults.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. White had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. White, Felson, Nevitt, Fredman.
Acquisition of data. White, Neogi, Felson, Gross, Niu, Nevitt, Lewis, Torner.
Analysis and interpretation of data. White, Keysor, Neogi, LaValley, Gross, Niu, Lewis, Fredman.
- 22The CES-D Scale: a self-report depression scale for research in the general population. Appl Psychol Meas 1977; 1: 385–401..
- 32World Health Organization. Obesity: preventing and managing the global epidemic. Report of a WHO consultation. World Health Organ Tech Rep Ser 2000; 894: 1–253.
- 35Modern epidemiology. 2nd ed. Philadelphia: Lippincott Williams & Wilkins; 1998., .
- 41A positive affect intervention for people experiencing health-related stress: development and non-randomized pilot test. J Health Psychol 2011. E-pub ahead of print., , , , , , et al.