Abstract
Objectives
The purpose of this study was to determine whether clinical, health-related quality of life (HRQL), and gait characteristics in adults with knee osteoarthritis (OA) differed by obesity category.
Methods
This cross-sectional analysis of 823 older adults (mean age, 64.6 years, SD, 7.8) with knee OA and overweight/obesity compared clinical, HRQL, and gait outcomes between obesity classifications (overweight/class I, BMI,27.0–34.9 kg/m2; class II, BMI,35.0–39.9 kg/m2; class III, BMI ≥ 40.0 kg/m2).
Results
Patients with class III obesity had worse Western Ontario McMasters Universities Arthritis Index knee pain (0 to 20) than the overweight/class I (mean 8.6 vs 7.0, difference, 1.5, 95% CI, 1.0 to 2.1, p < .0001) and class II (mean 8.6 vs 7.4, difference, 1.1, 95%CI, 0.6 to 1.7, p=.0002) obese groups. The SF-36 physical HRQL measure was lower in the class III obese group compared to the overweight/class I (mean 31.0 vs 37.3, difference, −6.2,95% CI,−7.8 to −4.7, p < .0001) and class II (mean 31.0 vs 35.0, difference −3.9, 95%CI −5.6 to −2.2, p<.0001) obese groups. The class III obese group had a base of support (cm) during gait that was wider than the overweight/class I (mean 14.0 vs 11.6, difference 3.3, 95%CI 2.6 to 4.0, p<.0001) and class II (mean 14.0 vs 11.6, difference 2.4, 95%CI 1.6 to 3.2, p<.0001) obese groups.
Conclusions
Among adults with knee OA, those with class III obesity had significantly higher pain levels, and worse physical HRQL and gait characteristics compared to adults with overweight/class I or class II obesity.
Keywords: Osteoarthritis, Knee Pain, Obesity, Gait, Physical Function
Graphical Abstract
Introduction
Knee osteoarthritis (OA) is the most common and persistent cause of mobility dependency and disability, with an estimated prevalence exceeding 250 million worldwide. (1) OA develops from a complex interaction of biomechanical and inflammatory disease pathways. These pathways are both common to obesity, resulting in increased mechanical joint stress and increased production of proinflammatory cytokines and adipokines that can promote both joint structural changes and pain. (2) Obesity is a major risk factor for knee OA; the prevalence of obesity has risen from 30% in 2000 to 42% in 2020, with class III obesity (BMI ≥ 40.0 kg/m2) nearly doubling from 4.7% to 9.2%. (3) Among patients with OA, a higher BMI, from overweight (25 kg/m2 to 29.9 kg/m2) to class III obesity, was associated with significantly greater analgesic medication use, rates of all-cause hospitalization, and total healthcare expenditures. (4) Adults with class III obesity also make adaptations to their gait compared to adults with less obesity, most notably people with severe obesity walk with a shorter step length resulting in a slower gait velocity that lowers knee joint loads relative to their BMI. (5,6) However, the effect of OA on individuals’ quality of life appears to be affected similarly across obesity classes. (7) Unfortunately, previous studies of obesity and knee OA have generally lacked a sufficiently large sample of patients to examine class III obesity specifically. (7) This limited variability in BMI range could obscure the clinical and mental health impact obesity has among the most severely affected. (8)
With a large sample of adults with knee OA, including a large subsample with class III obesity, this study determined if patients with knee OA and class III obesity had clinically important differences in pain, function, mobility, HRQL, and gait compared to people with knee OA and overweight/class I and class II obesity. These data could inform the costs and benefits of various surgical, pharmacological, and behavioral weight-loss solutions and whether patients with various levels of obesity respond similarly.
Methods
Study Design
The Weight-loss and Exercise for Communities with Arthritis in North Carolina (WE-CAN) trial is a Phase III, assessor-blinded, three center (Forsyth County, NC; Haywood County, NC, Johnston County, NC) randomized clinical trial with two parallel groups followed for 18 months. The present analysis utilized the baseline data from all study participants. The study protocol was reviewed and approved by the Human Subjects Committees of Wake Forest Health Sciences and The University of North Carolina at Chapel Hill and is in compliance with the terms and conditions set forth in the Helsinki Declaration. The trial design has been published previously. (9)
Participants
The pragmatic design of the parent trial resulted in few exclusion criteria for participant eligibility. Participants included 823 ambulatory, community-dwelling, men and women with a BMI ≥ 27 kg/m2 and who met the American College of Rheumatology (ACR) clinical criteria for knee OA. These included age ≥ 50 years, knee pain on most days of the week, plus at least two of the following: stiffness < 30 min/day; crepitus; bony tenderness; bony enlargement; no palpable warmth. (10) Key exclusions were symptomatic coronary artery disease, type 1 diabetes, and body mass index (BMI) < 27 kg/m2. Enrollment occurred between May 2016 and August 2019. Participants lived in three counties in North Carolina with diverse residential (urban, rural) and socioeconomic composition.
Measurements
Self-reported knee pain was measured using the Likert version of the Western Ontario McMasters Universities Osteoarthritis Index (WOMAC) which assesses knee pain over the last 48 hours. (11–13) The total score ranges from 0 to 20 (higher scores indicate greater pain). Prior work has suggested categorizing WOMAC Pain (0 to 20 scale) as 2 to 8 mild; ≥ 8 to 14 moderate; ≥14 to 20 severe. (14) WOMAC function assessed the degree of difficulty with activities of daily living in the last 48 hours with the total score ranging from 0 to 68; higher scores indicated poorer function with a score of 21 or higher indicating physical work limitations. (15) Six-minute walk distance assessed the distance a participant could walk along a standardized walkway in 6 minutes. (22) The expanded Short Physical Performance Battery (SPPB) assessed lower extremity function using balance, gait speed, and chair stand tests. (17) Test-retest reliability coefficient on a sample of adults age 65 – 75 years was 0.87. (18)
We used a GAITRite 6.1 m instrumented mat (Franklin, NJ, USA) to measure spatio-temporal gait variables including walking velocity, stride length, swing time, stance time, base of support, and foot abduction (Figure 1). Three trials within 3.5% of a freely chosen walking speed were averaged to yield a representative trial for each participant. Of the 823 participants, 648 completed a gait analysis due to the limited availability of the testing equipment at one of the testing sites. The GAITRite instrument has a one-week test-retest reliability of ≥ 0.92 for all spatio-temporal gait measurements except base of support which has an intraclass correlation coefficient of 0.80. (8)
Figure 1.
Gait characteristics.
The SF-36 measures health-related quality of life (HRQL) using two broad summary scores: physical and mental health, scaled from 0 (worst) to 100 (best). (19) The CES-D-10 (range, 0–30) assessed depressive symptoms during the last week. Ten items are scored from 0 (rarely or none of the time) to 3 (most or almost all the time). (20) The range for internal consistency is 0.80 – 0.88 (Cronbach’s alpha). (21)
Statistical Analysis
Baseline data were used to classify participants into BMI categories: 27 – 34.9 kg/m2, overweight/class I obese; 35 −39.9 kg/m2, class II obese; and ≥ 40.0 kg/m2, class III obese. (4) Means and 95% confidence intervals described continuous characteristics, ANCOVAs compared BMI categories adjusted for age, race, sex, crepitus, education, and depression. Discrete characteristics were presented as counts and percentages, with group comparisons performed using Chi-squared (χ2) tests and T-tests using a Type I error rate of 0.05. Categorical outcomes were described using unadjusted odds ratios from logistic regression models, and pairwise comparisons for 3-group BMI outcomes were deemed significant at a Bonferroni-adjusted 0.0167 level. The half-SD formula was used as a proxy to calculate minimally clinically important differences (MCID) between obesity categories. (22)
Results
Participant characteristics by obesity class.
Mean BMI was overweight/class I obese (n = 389), 31.5 kg/m2; class II obese (n = 230), 37.3 kg/m2; and class III obese (n = 204) 46.4 kg/m2 (p < 0.001) (Table 1). The class III obese group was significantly younger than the overweight/class I (difference −4.9 yrs., 95% CI, −6.2 to −3.6, p < .001) and class II (difference −3.8 yrs., 95% CI, −5.2 to −2.4, p < 0.001) obese groups. The class III obese group compared to the overweight/class I obese group had lower odds of being married or living as married (OR 0.63, 95% CI, 0.45 to 0.89, p = .008), had lower odds of having a bachelor’s degree or more (OR 0.56, 95% CI, 0.40 to 0.80, p = .001), and had lower odds of having an annual income greater than $75,000 (OR 0.44, 95% CI, 0.29 to 0.67, p <.001). Class III obese participants also had higher odds of reporting depressive symptoms (OR 1.94, 95% CI, 1.31 to 2.85, p<.001) and having type II diabetes (OR 1.77, 95% CI, 1.18 to 2.66, p=.006).
Table 1.
Participant Characteristics by BMI Category
| Overall | Overweight/Class I Obese | Class II Obese | Class III Obese | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Variable | n | aMean ± SD or n (%) | n | bMean (95% CI) or n (%) | n | Mean (95% CI) or n (%) | n | Mean (95% CI) or n (%) | cp-value |
| Anthropometric Measurements | |||||||||
| Age (years) | 823 | 64.6 (7.8) | 389 | 66.1 (65.4, 66.9) | 230 | 65.0 (64.0, 66.0) | 204 | 61.2 (60.2, 62.2) | <.001 |
| Sex (Men) | 823 | 186 (22.6) | 389 | 94 (24.2) | 230 | 52 (22.6) | 204 | 40 (19.6) | 0.45 |
| Weight (kg) | 823 | 100.9 ± 21.3 | 389 | 86.7 (85.3, 88.1) | 230 | 101.8 (100.0, 103.6) | 204 | 126.9 (125.0, 128.8) | <.001 |
| BMI (kg/m2) | 823 | 36.8 (6.9) | 389 | 31.5 (31.1, 31.8) | 230 | 37.3 (36.8, 37.7) | 204 | 46.4 (45.9, 46.9) | <.001 |
| Demographic Characteristics | |||||||||
| Married or Marriage-like | 821 | 492 (59.9) | 387 | 250 (64.6) | 230 | 133 (57.8) | 204 | 109 (53.4) | 0.02 |
| Income (75k+) | 793 | 232 (29.3) | 377 | 132 (35.0) | 219 | 62 (28.3) | 197 | 38 (19.3) | <.001 |
| Education (Bachelors or more) | 820 | 545 (66.5) | 387 | 279 (72.1) | 229 | 145 (63.3) | 204 | 121 (59.3) | 0.004 |
| Medically Related Characteristics | |||||||||
| CES-D-10 (0–30.) | 818 | 5.5 (5.0) | 385 | 4.7 (4.2, 5.2) | 230 | 5.6 (5.0, 6.3) | 203 | 6.7 (6.0, 7.4) | <.001 |
| Crepitus (Y/N) | 823 | 713 (86.6) | 389 | 335 (86.1) | 230 | 199 (86.5) | 204 | 179 (87.7) | 0.86 |
| Total Medications/Supplements | 823 | 8.2 ± 5.0 | 389 | 8.3 (7.8, 8.8) | 230 | 8.1 (7.5, 8.8) | 204 | 8.2 (7.5, 8.9) | 0.93 |
| Diabetes (Y/N) | 823 | 172 (20.9) | 389 | 67 (17.2) | 230 | 50 (21.7) | 204 | 55 (27.0) | 0.02 |
| Number of Comorbidities | 822 | 2.0 ± 1.6 | 389 | 1.9 (1.7, 2.0) | 229 | 2.0 (1.8, 2.2) | 204 | 2.4 (2.2, 2.6) | 0.001 |
Abbreviation: CES-D-10, Center for Epidemiological Studies – Depression Scale – 10
Means and standard deviations calculated for continuous variables; counts and percentages calculated for categorical variables.
Means and 95% confidence intervals calculated for continuous variables; counts and percentages calculated for categorical variables.
p-values were calculated with ANOVAs and indicate an overall significant difference between groups if p ≤ 0.05.
Obesity effects on clinical, HRQL, and mobility outcomes.
Post hoc pairwise comparisons revealed that the class III obese group had greater WOMAC knee pain than the overweight/class I (difference, 1.5, 95% CI, 1.0 to 2.1, p < .0001) and class II (difference 1.1, 95% CI, 0.6 to 1.7, p = .0002) obese groups. WOMAC function was also significantly worse in the class III obese group compared to the overweight/class I (difference 7.3, 95% CI, 5.4 to 9.2, p < .0001) and class II (difference 4.5, 95% CI, 2.3 to 6.6, p < .0001) obese groups (Figure 2). The SF-36 physical HRQL measure was lower in the class III obese group compared to the overweight/class I (difference −6.2, 95% CI, −7.8 to −4.7, p < .0001) and class II (difference −3.9, 95% CI, −5.6 to −2.2, p < .0001) obese groups, while the class II obese group had significantly worse scores on the HRQL physical dimension than the overweight/class I obese group (difference −2.3, 95% CI, −3.8 to −0.8, p = .002) (Table 2).
Figure 2.
(a) WOMAC pain scores for each BMI category (range, 0 [no pain] to 20 [severe pain]); (b) WOMAC function scores for each BMI category (range, 0 [best function to 68 [worse function]. Box plots are shown in which the middle line represents median values; x, mean values; and boxes, interquartile ranges (IQRs). Whiskers extend to the most extreme observed values within 1.5 times the IQR of the nearer quartile, and dots represent observed values outside the range.
Table 2.
Clinical Outcomes by BMI Category. Pairwise Comparisons Were Deemed Significant at a Bonferroni-Adjusted 0.0167 Level.
| Overall | Overweight/Class I Obese | Class II Obese | Class III Obese | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Variable | n | Mean ± SD | n | Mean (95% CI) | n | Mean (95% CI) | n | Mean (95% CI) | Mean Differences (95% CI) | p-value |
| WOMAC Pain (0–20) | 823 | 7.5 ± 3.2 | 389 | 7.0 (6.7, 7.4) | 230 | 7.4 (7.0, 7.8) | 204 | 8.6 (8.1, 9.0) | <.001 a | |
| b Class III vs. Overweight/Class I | 1.5 (1.0, 2.1) | <.0001 | ||||||||
| Class III vs. Class II | 1.1 (0.6, 1.7) | 0.0002 | ||||||||
| Class II vs Overweight/Class I | 0.4 (−0.1, 0.9) | 0.13 | ||||||||
| WOMAC Function (0–68) | 823 | 25.9 ± 11.7 | 389 | 23.4 (22.2, 24.5) | 230 | 26.2 (24.7, 27.6) | 204 | 30.6 (29.1, 32.2) | <.001a | |
| Class III vs. Overweight/Class I | 7.3 (5.4, 9.2) | <.0001 | ||||||||
| Class III vs. Class II | 4.5 (2.3, 6.6) | <.0001 | ||||||||
| Class II vs Overweight/Class I | 2.8 (1.0, 4.7) | 0.003 | ||||||||
| 6 Minute Walk Distance (m) | 819 | 371.6 ± 93.1 | 387 | 408 (400, 417) | 229 | 365 (354, 376) | 203 | 309 (298, 321) | <.001a | |
| Class III vs. Overweight/Class I | −99 (−113, −84) | <.0001 | ||||||||
| Class III vs. Class II | −56 (−72, −40) | <.0001 | ||||||||
| Class II vs Overweight/Class I | −43 (−57, −29) | <.0001 | ||||||||
| SPPB Total Score (0–12) | 822 | 9.6 ± 2.1 | 388 | 10.0 (9.8, 10.2) | 230 | 9.5 (9.2, 9.7) | 204 | 8.7 (8.5, 9.0) | <.001a | |
| Class III vs. Overweight/Class I | −1.3 (−1.6, −1.0) | <.0001 | ||||||||
| Class III vs. Class II | −0.8 (−1.1, −0.4) | ..0001 | ||||||||
| Class II vs Overweight/Class I | −0.5 (−0.9, −0.2) | 0.001 | ||||||||
| SF-36 Physical (0–100) | 823 | 35.1 ± 9.3 | 389 | 37.3 (36.4, 38.2) | 230 | 35.0 (33.8, 36.1) | 204 | 31.0 (29.8, 32.3) | <.001a | |
| Class III vs. Overweight/Class I | −6.2 (−7.8, −4.7) | <.0001 | ||||||||
| Class III vs. Class II | −3.9 (−5.6, −2.2) | <.0001 | ||||||||
| Class II vs Overweight/Class I | −2.3 (−3.8, −0.8) | 0.002 | ||||||||
| SF-36 Mental (0–100) | 823 | 54.9 ± 10.0 | 389 | 55.4 (54.4, 56.4) | 230 | 54.8 (53.6, 56.1) | 204 | 54.2 (52.8, 55.5) | 0.36a | |
| Class III vs. Overweight/Class I | −1.2 (−2.9, 0.5) | 0.16 | ||||||||
| Class III vs. Class II | −0.7 (−2.6, 1.2) | 0.47 | ||||||||
| Class II vs Overweight/Class I | −0.5 (−2.8, 1.1) | 0.52 | ||||||||
p-values indicate overall significant difference between groups
Pairwise comparisons significantly different at p ≤ .0167 to account for multiple comparisons. Values adjusted for age, race, sex, crepitus, education, and depression.
Six-minute walk distance (m) was significantly shorter in the class III obese group compared to the overweight/class I (difference −99, 95% CI, −113 to −84, p < .0001) and class II (difference −56, 95% CI, −72 to −40, p < .0001) obese groups (Table 2). The class III obese group had a shorter stride length (cm) (difference −15, 95% CI, −19 to −12, p < .0001), a wider base of support (cm) (difference 3.4, 95% CI, 2.7 to 4.1, p < .0001), and more foot abduction (degrees) (difference 3.1, 95% CI, 2.0 to 4.1, p < .0001) than the overweight/class I obese group; similar significant differences were present between the class III and class II obese groups (Table 3).
Table 3.
Gait Characteristics by BMI Category. Pairwise Comparisons Were Deemed Significant at a Bonferroni-Adjusted 0.0167 Level.
| Overall | Overweight/Class I Obese | Class II Obese | Class III Obese | |||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Variable | n | Mean ± SD | n | Mean (95% CI) | n | Mean (95% CI) | n | Mean (95% CI) | Mean Differences (95% CI) | p-value |
| Walking Velocity (m/s) | 648 | 1.01 ± 0.21 | 317 | 1.08 (1.06, 1.10) | 173 | 0.99 (0.96, 1.02) | 158 | 0.89 (0.86, 0.92) | <.001a | |
| bClass III vs. Overweight/Class I | −0.18 (−0.22, −0.14) | <.0001 | ||||||||
| Class III vs. Class II | −0.10 (−0.14, −0.05) | <.0001 | ||||||||
| Class II vs Overweight/Class I | −0.08 (−0.12, −0.04) | <.0001 | ||||||||
| Stride Length (cm) | 648 | 116.4 ± 18.0 | 317 | 122.1 (120.2, 123.9) | 173 | 114.9 (112.4, 117.4) | 158 | 106.7 (104.1, 109.4) | <.001 | |
| Class III vs. Overweight/Class I | −15.4 (−18.6, −12.1) | <.0001 | ||||||||
| Class III vs. Class II | −8.2 (−11.8, −4.5) | <.0001 | ||||||||
| Class II vs Overweight/Class I | −7.2 (−10.3, −4.0) | <.0001 | ||||||||
| Swing Time (s) | 648 | 0.38 ± 0.04 | 317 | 0.39 (0.38, 0.39) | 173 | 0.38 (0.37, 0.38) | 158 | 0.37 (0.37, 0.38) | <.001 | |
| Class III vs. Overweight/Class I | −0.01 (−0.02, −0.01) | 0.0009 | ||||||||
| Class III vs. Class II | −0.00 (−0.01, 0.01) | 0.6841 | ||||||||
| Class II vs Overweight/Class I | −0.01 (−0.02, −0.00) | 0.0031 | ||||||||
| Stance Time (s) | 648 | 0.79 ± 0.13 | 317 | 0.75 (0.74, 0.77) | 173 | 0.79 (0.77, 0.81) | 158 | 0.84 (0.82, 0.86) | <.001 | |
| Class III vs. Overweight/Class I | 0.09 (0.06, 0.11) | <.0001 | ||||||||
| Class III vs. Class II | 0.05 (0.02, 0.08) | 0.0005 | ||||||||
| Class II vs Overweight/Class I | 0.04 (0.02, 0.06) | 0.0013 | ||||||||
| Base of Support (cm) | 648 | 11.8 ± 3.9 | 317 | 10.7 (10.3, 11.1) | 173 | 11.6 (11.1, 12.2) | 158 | 14.0 (13.5, 14.6) | <.001 | |
| Class III vs. Overweight/Class I | 3.4 (2.7, 4.1) | <.0001 | ||||||||
| Class III vs. Class II | 2.4 (1.6, 3.2) | <.0001 | ||||||||
| Class II vs Overweight/Class I | 0.9 (0.3, 1.6) | 0.0071 | ||||||||
| Foot Abduction (degrees) | 648 | 8.1 ± 5.4 | 317 | 7.1 (6.5, 7.7) | 173 | 8.0 (7.2, 8.8) | 158 | 10.2 (9.3, 11.0) | <.001 | |
| Class III vs. Overweight/Class I | 3.1 (2.0, 4.1) | <.0001 | ||||||||
| Class III vs. Class II | 2.1 (1.0, 3.3) | 0.0003 | ||||||||
| Class II vs Overweight/Class I | 0.9 (−0.1, 1.9) | 0.0672 | ||||||||
p-values indicate overall significant difference between groups
Pairwise comparisons significantly different at p ≤ .0167.
The pairwise comparisons between the class III and overweight/class I groups reached the MCID threshold for most outcomes including pain, function, 6-minute walk distance, SPPB, SF-36 physical, walking velocity, and stride length (Tables 2, 3). The class III versus class II pairwise differences in 6-minute walk distance and walking velocity also reached the MCID threshold. None of the pairwise comparisons between the two lower obesity classes reached the MCID threshold.
Discussion
In this cross-sectional analysis of the WE-CAN trial, in participants with knee OA and overweight and obesity, those with the most severe obesity had the worst pain, poorest function, shortest 6-minute walk distance, and gait deviations that were reflective of a slower walking velocity. The group with the most severe obesity also exhibited gait mechanics that indicated an apparent attempt to maintain stability by widening the base of support via an increased stride width and greater forefoot abduction. A slow walking velocity and short stride length to reduce knee loads were likely adopted in response to the greater knee pain and higher body weight in this group. (23)
Higher levels of obesity are associated with more intense chronic musculoskeletal pain. (24) This association is robust, even after controlling for age and sex, with greater joint loads during functional activities like gait and more systemic inflammation the most common contributors. (25) Obesity and knee OA were first linked in 1945, a relationship that has been confirmed repeatedly. (26–29) In knee OA patients, those with obesity report more severe knee pain than patients with overweight; (27,30) however, the clinical consequences for people in the highest obesity class are less clear. Raud et al. (30) reported a dose-response in pain as BMI increased from overweight to obesity class I to obesity class II/III in 391 patients with knee OA. However, only 15% (n = 57) of the cohort were in the combined class II/III group; there was no clinically or statistically significant difference in pain between the class II/III and the class I groups; and the distribution of BMIs within the heaviest group (e.g., those under and over 40 kg/m2) was not reported. In the present study, the class III obese group (BMI ≥ 40 kg/m2) had significantly more pain than either the class II or overweight/class I groups; however, there was no statistical difference in pain between the overweight/class I and class II obese groups.
The association between obesity and poor physical function is well established across middle age and older adults, with more severe obesity generally related to poorer mobility, including slower walking speeds, and longer chair rise times. (31,32) The class III obese group in the present study had significantly worse function and HRQL, as well as a shorter 6-minute walk distance than the class I and class II obesity groups. Taken together, these data suggest that interventions designed to reduce obesity from a severe Class III level to a moderate Class II level could result in clinically meaningful reductions in pain and improvements in function and HRQL in adults with knee OA. (15)
Gait characteristics of the three groups supported the clinical outcomes. The class III obesity group had the slowest walking velocity (mean = .89 m/s). Slow walking velocities indicate reduced physical functioning, with velocities of .8 m/s and below also predictive of a life expectancy below the median age. (33) In addition, the class III obese group exhibited shorter strides and the associated increased stance and reduced swing phase times, a wider stride width (base of support), and more forefoot abduction during walking compared to the other obesity classes. These gait adjustments indicate an attempt by the class III group to reduce joint loads, maintain stability, and reduce the metabolic cost of walking due to the greater obesity level.
The statistically and clinically important differences that separate the class III obesity group from the groups with less obesity suggest that older adults with knee OA and class III obesity should be included in sufficient numbers and analyzed as a separate group in future weight-loss clinical trials. By partitioning out Class III obese individuals, more accurate outcomes may be observed across all obesity classes.
Like other diseases, the prevalence of arthritis and its associated limitations is highest among adults with mobility limitations, low socioeconomic class, and poor overall health, including higher BMI levels. (34,35) These characteristics describe our study participants with class III obesity. Data on class III obese patient subgroups could be critical in accurately determining the costs and benefits of various surgical, pharmacological, and behavioral weight-loss solutions for patients with knee OA. For example, do knee OA patients with class III obesity respond better (or worse) to weight loss than knee OA patients with less obesity? Also, is the relationship between weight loss and knee pain reduction linear, and is it similar across all obesity levels? This more disabled subgroup should be represented in knee OA weight-loss clinical trials in sufficient numbers to facilitate adequately powered subgroup analyses.
Limitations
This study has several limitations. First, the cross-sectional design prevents any observations regarding causation. Second, this cohort was recruited for a pragmatic randomized clinical trial; hence, there were limitations in the data collected, including the absence of radiographs. Consequently, it is likely some participants did not meet the criteria for structural OA. Third, absence of a normal weight group reduces the spread of the data, making correlational analyses less effective. Fourth, we did not account for all factors that may explain differences in pain by obesity class, such as fat mass. BMI is strongly correlated with body fat measures, but does not distinguish between lean and fat mass and provides no indication of body fat distribution. (36)
Conclusions
Among participants with overweight, obesity, and clinical knee OA, the most severe obesity group had the worst pain and physical HRQL, the poorest function and mobility, and gait deviations that reflected the slowest walking velocity. The group with the most severe obesity also exhibited several compensatory gait adaptations that indicated an apparent attempt to maintain stability. These health differences warrant inclusion of people with knee OA and class III obesity in future behavioral weight-loss trials.
Acknowledgements
Research reported in this publication was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases of the National Institutes of Health under Award Number 1 U01 AR068658-01. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
We thank General Nutrition Centers for their generosity in supplying the Lean Shakes to us at no cost. They have not funded the study in any other way and have not contributed to and have no influence over study design, publications, or dissemination of results.
The WE-CAN investigators wish to acknowledge the tremendous efforts of our staff in making this project come to fruition: Natalia Favoreto, Nathan Fiore, James Gerosa, Daniel Hamm, Ryan Hill, Elyse Howdershell, Erika Janssen, Monica Love, Chris Mygrant, Alex Nielson, Carol Patterson, Elena Wright, Caroline Wyrosdick, and project managers Betsy Hackney and Jovita Newman.
Footnotes
Trial Registration:clinicaltrials.gov Identifier: NCT02577549 October 12, 2015
Ethical Approval and Consent to Participate
The Wake Forest Health Sciences and The University of North Carolina at Chapel Hill Institutional Review Boards (IRB) reviewed all research involving humans to ensure that participants were informed of all known risks posed by the research study and that these studies were conducted in accordance with the ethical standards put forward by the Belmont Report, and federal, state, and local regulations, and policies governing human research. All subjects gave informed consent to participate in the study.
Competing Interests
MEG is employed by Xcenda, LLC. DJH provides consulting advice on scientific advisory boards for Pfizer, Lilly, TLCBio, Novartis, Tissuegene, and Biobone. DJH is the co-director of the Sydney Musculoskeletal Health Flagship. In addition, DJH is the editor of the osteoarthritis section for UpToDate and co-Editor in Chief of Osteoarthritis and Cartilage.
Availability of Data and Material
The investigators support data sharing and will comply with all NIH guidelines as outlined in the NIH Data Sharing Policy and Implementation Guidance document. Any data-sharing agreement will require that the data be used only for research purposes; no attempts will be made to identify individual participants; the data will be kept secure; the user will not distribute the data to other researchers; the user will return the files or destroy them once the project is completed; and the user will acknowledge the data source.
References
- 1.Vos T, Flaxman AD, Naghavi M, Lozano R, Michaud C, Ezzati M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet 2012;380:2163–2196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Griffin TM, Guilak F. The role of mechanical loading in the onset and progression of osteoarthritis. ExercSport SciRev 2005;33:195–200. [DOI] [PubMed] [Google Scholar]
- 3.Stierman B, Afful J, Carroll M, Te-Ching C, Orlando D, Fink S, et al. NHSR 158. National Health and Nutrition Examination Survey 2017–March 2020 Pre-pandemic Data Files National Center for Health Statistics; (U.S.); 2021. Available at: https://stacks.cdc.gov/view/cdc/106273. Accessed April 19, 2023. [Google Scholar]
- 4.Johnston SS, Ammann E, Scamuffa R, Samuels J, Stokes A, Fegelman E, et al. Association of body mass index and osteoarthritis with healthcare expenditures and utilization. Obes Sci Pract 2020;6:139–151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Devita P, Hortobagyi T. Obesity is not associated with increased knee joint torque and power during level walking. JBiomech 2003;36:1355–1362. [DOI] [PubMed] [Google Scholar]
- 6.Okoro CA, Hootman JM, Strine TW, Balluz LS, Mokdad AH. Disability, arthritis, and body weight among adults 45 years and older. Obes Res 2004;12:854–861. [DOI] [PubMed] [Google Scholar]
- 7.Sendi P, Brunotte R, Potoczna N, Branson R, Horber FF. Health-related quality of life in patients with class II and class III obesity. Obes Surg 2005;15:1070–1076. [DOI] [PubMed] [Google Scholar]
- 8.van Uden CJT, Besser MP. Test-retest reliability of temporal and spatial gait characteristics measured with an instrumented walkway system (GAITRite). BMC Musculoskelet Disord 2004;5:13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Messier SP, Callahan LF, Beavers DP, Queen K, Mihalko SL, Miller GD, et al. Weight-loss and exercise for communities with arthritis in North Carolina (we-can): design and rationale of a pragmatic, assessor-blinded, randomized controlled trial. BMC musculoskeletal disorders 2017;18:91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Altman R, Asch E, Bloch D, Bole G, Borenstein D, Brandt K, et al. Development of criteria for the classification and reporting of osteoarthritis. Classification of osteoarthritis of the knee. Arthritis Rheum 1986;29:1039–1049. [DOI] [PubMed] [Google Scholar]
- 11.Bellamy N Outcome measurement in osteoarthritis clinical trials. JRheumatolSuppl 1995;43:49–51. [PubMed] [Google Scholar]
- 12.Bellamy N, Buchanan WW, Goldsmith CH, Campbell J, Stitt LW. Validation study of WOMAC: a health status instrument for measuring clinically important patient relevant outcomes to antirheumatic drug therapy in patients with osteoarthritis of the hip or knee. JRheumatol 1988;15:1833–1840. [PubMed] [Google Scholar]
- 13.Griffiths G, Bellamy N, Kean WF, Campbell J, Gerecz-Simon E. A study of the time frame dependency of responses to the WOMAC osteoarthritis index. Inflammopharmacology 1993;2:85–87. [Google Scholar]
- 14.Kapstad H, Hanestad BR, Langeland N, Rustøen T, Stavem K. Cutpoints for mild, moderate and severe pain in patients with osteoarthritis of the hip or knee ready for joint replacement surgery. BMC Musculoskelet Disord 2008;9:55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Siviero P, Limongi F, Gesmundo A, Zambon S, Cooper C, Dennison EM, et al. Minimal clinically important decline in physical function over one year: EPOSA study. BMC Musculoskeletal Disorders 2019;20:227. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Bohannon RW, Crouch R. Minimal clinically important difference for change in 6-minute walk test distance of adults with pathology: a systematic review. J Eval Clin Pract 2017;23:377–381. [DOI] [PubMed] [Google Scholar]
- 17.Guralnik JM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer DG, 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:M85–M94. [DOI] [PubMed] [Google Scholar]
- 18.Gómez JF, Curcio C-L, Alvarado B, Zunzunegui MV, Guralnik J. Validity and reliability of the Short Physical Performance Battery (SPPB): a pilot study on mobility in the Colombian Andes. Colomb Med (Cali) 2013;44:165–171. [PMC free article] [PubMed] [Google Scholar]
- 19.Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. MedCare 1992;30:473–483. [PubMed] [Google Scholar]
- 20.Andresen EM, Malmgren JA, Carter WB, Patrick DL. Screening for depression in well older adults: evaluation of a short form of the CES-D (Center for Epidemiologic Studies Depression Scale). Am J Prev Med 1994;10:77–84. [PubMed] [Google Scholar]
- 21.Zhang W, O’Brien N, Forrest JI, Salters KA, Patterson TL, Montaner JSG, et al. Validating a shortened depression scale (10 item CES-D) among HIV-positive people in British Columbia, Canada. PLoS One 2012;7:e40793. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Norman GR, Sloan JA, Wyrwich KW. Interpretation of changes in health-related quality of life: the remarkable universality of half a standard deviation. Med Care 2003;41:582–592. [DOI] [PubMed] [Google Scholar]
- 23.Browning RC, Kram R. Effects of obesity on the biomechanics of walking at different speeds. MedSciSports Exerc 2007;39:1632–1641. [DOI] [PubMed] [Google Scholar]
- 24.Costa ABP, Machado LAC, Telles RW, Barreto SM. Dose-response associations of clinical markers of obesity and duration of exposure to excess weight with chronic musculoskeletal pain: cross-sectional analysis at baseline of ELSA-Brasil Musculoskeletal cohort. Rheumatol Int 2020;40:881–891. [DOI] [PubMed] [Google Scholar]
- 25.Stone AA, Broderick JE. Obesity and Pain Are Associated in the United States. Obesity 2012;20:1491–1495. [DOI] [PubMed] [Google Scholar]
- 26.Lewis-Faning E, Fletcher E. Chronic rheumatic diseases: statistical study of 1000 cases of chronic rheumatism. Postgraduate Medical Journal 1945;21:137. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Felson DT, Anderson JJ, Naimark A, Walker AM, Meenan RF. Obesity and knee osteoarthritis. The Framingham Study. AnnInternMed 1988;109:18–24. [DOI] [PubMed] [Google Scholar]
- 28.Felson DT, Zhang Y, Anthony JM, Naimark A, Anderson JJ. Weight loss reduces the risk for symptomatic knee osteoarthritis in women. The Framingham Study. AnnInternMed 1992;116:535–539. [DOI] [PubMed] [Google Scholar]
- 29.Blagojevic M, Jinks C, Jeffery A, Jordan KP. Risk factors for onset of osteoarthritis of the knee in older adults: a systematic review and meta-analysis. Osteoarthritis and cartilage / OARS, Osteoarthritis Research Society 2010;18:24–33. [DOI] [PubMed] [Google Scholar]
- 30.Raud B, Gay C, Guiguet-Auclair C, Bonnin A, Gerbaud L, Pereira B, et al. Level of obesity is directly associated with the clinical and functional consequences of knee osteoarthritis. Sci Rep 2020;10:3601. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Rogers NT, Power C, Pinto Pereira SM. Birthweight, lifetime obesity and physical functioning in mid-adulthood: a nationwide birth cohort study. Int J Epidemiol 2020;49:657–665. [DOI] [PubMed] [Google Scholar]
- 32.Hergenroeder AL, Brach JS, Otto AD, Sparto PJ, Jakicic JM. The Influence of Body Mass Index on Self-report and Performance-based Measures of Physical Function in Adult Women. Cardiopulm Phys Ther J 2011;22:11–20. [PMC free article] [PubMed] [Google Scholar]
- 33.Studenski S, Perera S, Patel K, Rosano C, Faulkner K, Inzitari M, et al. Gait speed and survival in older adults. JAMA 2011;305:50–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Theis KA. Prevalence of Arthritis and Arthritis-Attributable Activity Limitation — United States, 2016–2018. MMWR Morb Mortal Wkly Rep 2021;70. Available at: https://www.cdc.gov/mmwr/volumes/70/wr/mm7040a2.htm. Accessed October 7, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Johnson-Lawrence V, Kaplan G, Galea S. Socioeconomic mobility in adulthood and cardiovascular disease mortality. Ann Epidemiol 2013;23:167–171. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Adab P, Pallan M, Whincup PH. Is BMI the best measure of obesity? BMJ 2018;360:k1274. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The investigators support data sharing and will comply with all NIH guidelines as outlined in the NIH Data Sharing Policy and Implementation Guidance document. Any data-sharing agreement will require that the data be used only for research purposes; no attempts will be made to identify individual participants; the data will be kept secure; the user will not distribute the data to other researchers; the user will return the files or destroy them once the project is completed; and the user will acknowledge the data source.