Abstract
Objective:
Our aim was to define the association of weight change (weight loss or weight gain) with the incidence and progression of hand osteoarthritis (OA), assessed either by radiography or by pain, using data from the Osteoarthritis Initiative.
Methods:
Among the 4,796 participants, we selected 4,598 subjects, excluding those with cancer or rheumatoid arthritis, or a Body Mass Index under 18.5 kg/m2. We investigated the association of weight change with incidence and progression of radiographic hand OA and the development and resolution of hand pain. Utilizing multivariable logistic regression, we investigated the association of weight change from baseline to the 4-year follow-up with the incidence and progression of radiographic hand OA at the 4-year follow-up. Additionally, multivariable repeated-measure mixed-effects logistic regression analyzed the association of weight change with the development and resolution of hand pain across 2-year, 4-year, 6-year, and 8-year follow-ups.
Results:
No statistically significant associations were observed between weight change and the investigated outcomes. Specifically, for each 5% weight loss, the odds ratios for the incidence and progression of radiographic hand OA were 0.90 (95%CI: 0.67-1.23) and 0.92 (95%CI: 0.84-1.00), respectively. Similarly, for each 5% weight loss, the odds ratios for the development and resolution of hand pain at the 8-year follow-up were 1.00 (95%CI: 0.92-1.09) and 1.07 (95%CI: 0.91-1.25), respectively.
Conclusion:
Our study found no evidence of an association between weight change and the odds of incidence or progression of radiographic hand OA over 4 years, nor the development or resolution of hand pain over 8 years.
Keywords: Osteoarthritis, Hand 1, Weight Loss 2, Weight Gain 3
Introduction
Hand osteoarthritis (OA) is a debilitating joint disease [1, 2]. It is the second-most prevalent form of OA, with only knee OA being more prevalent [3]. By the time they reach 85 years of age, 40% of adults are estimated to be affected by hand OA [4]. Hand OA adversely affects people’s day-to-day activities, including work activities, due to pain, decreased grip force, and impaired function [5, 6]. Unfortunately, there is no cure for hand OA. The risk factors for hand OA are age, female sex, race, occupation (involving repetitive high force on hand joints), and obesity [7, 8].
The association of obesity with hand OA – both the incidence and progression thereof - has been suggested by several studies [9–12]. As there is no cure for hand OA, weight loss or the prevention of weight gain could potentially be important strategies for managing hand OA. However, the association of weight change (weight loss and weight gain) with the development and progression of hand OA has yet to be studied. An informative study requires examining the long-term impact of weight change and hand outcomes because hand OA outcomes manifest slowly, and the impact of weight change may not be immediate. Therefore, we performed analyses to define the long-term relationship of weight loss and weight gain with the incidence and progression of hand OA, either assessed by radiography or by symptoms, using data from the Osteoarthritis Initiative (OAI).
Methods
Data
We obtained data from the OAI [13], a prospective cohort study of people with or at risk of knee OA. Ethical approval was obtained by the institutions undertaking the original OAI study. In addition, written informed consent was obtained from all participants in the OAI.
Baseline and follow-up data
We used hand radiographic data of participants at the baseline and 4-year follow-up. Hand radiographic data was available primarily for the dominant hand of a participant. For the pain, we used data at the baseline, 2-year, 4-year, 6-year, and 8-year follow-up. Pain data was available for both hands (left and right) of a participants.
Exposure
For hand radiographic outcomes, our exposure of interest was weight change between baseline and 4-year follow-up. For pain outcomes, we used weight change between baseline and respective follow-up of 2-year, 4-year, 6-year, and 8-year. Weight change was expressed as a percent of baseline weight. (‘percentage weight change’). Weight was measured at clinical centers (not self-reported).
Outcomes
We investigated two types of outcomes: radiographic and pain outcomes. Specifically, we had the following four radiographic outcomes: 1) incidence of hand radiographic OA (ROA); 2) progression of hand ROA; 3) worsening of joint space narrowing (JSN) of hand joints assessed by radiography; 4) worsening of osteophytes of hand joints assessed by radiography. For the pain outcomes, we investigated development of hand pain and resolution of hand pain. All of our outcomes were binary (i.e., Yes or No). To investigate these radiographic and pain outcomes, we created 4 cohorts: a ‘hand ROA incidence cohort’; a ‘hand ROA progression cohort’; a ‘hand pain development cohort’; and a ‘hand pain resolution cohort’.
We assessed the incidence and progression of hand ROA based on Kellgren-Lawrence (KL) grades [14]: 0 = no hand ROA (no JSN and no osteophytes); 1 = doubtful hand ROA (presence of questionable JSN or questionable osteophytes); 2 = mild hand ROA (presence of mild JSN or small osteophyte(s); 3 = moderate hand ROA (presence of moderate JSN or moderate osteophyte(s)); and 4 = severe hand ROA (presence of major JSN or major osteophyte(s)) [15]. Our reader demonstrated good intra-reader agreement, evidenced by a weighted kappa exceeding 0.84. This agreement was tested using 100 randomly selected pairs of hand radiographs read twice, with a 2 to 3-month interval between readings. The reader examined paired images in a blinded fashion, without access to time or clinical information. A hand joint was considered to have hand ROA if the joint had a KL grade of ≥ 2. ‘Overall Hand ROA’ was considered if a hand had OA in ≥ 2 hand joints on 2 different digits (rays, consisting of the thumb and the fingers). This is because OA in two hand joints on the same digit (ray) may indicate trauma-related OA, but we wanted to include only joint OA in our analyses due to the natural course of the disease. For hand ROA, we investigated 14 hand joints: 5 metacarpophalangeal (MCP) joints, the thumb interphalangeal joint, 4 proximal interphalangeal (PIP) joints, and 4 distal interphalangeal (DIP) joints. We did not investigate the joints at the base of the thumb (i.e., the carpometacarpal [CMC] and scaphotrapeziotrapezoid [STT] joints, which are close to the wrist). This is because OA of the CMC and STT joints is most common in the non-dominant hand [16–18], and we did not have data for the most non-dominant hand, therefore including these joints would have diluted our sample and made it prone to error in findings. A second reader provided semi-quantitative scoring for JSN and osteophytes (graded 0-3). The reader had a good intra-reader agreement for JSN (weight kappa = 0.75 to 0.96 [median weight kappa = 0.86]) and osteophytes (weight kappa = 0.65 to 1.00 [median weight kappa = 0.92]).
Our first outcome is the incidence of overall hand ROA, which we defined as a participant having it at the 4-year follow-up but not at baseline. We investigated this outcome in the ‘hand ROA incidence cohort,’ consisting of hands without overall hand ROA at baseline. Our second outcome of progression of overall hand ROA was investigated in the ‘hand ROA progression cohort’, which consisted of hands that had overall hand ROA at baseline. Progression of overall hand ROA was defined as an increase of ≥ 1 in the sum of the KL grades in hand joints for an entire hand between baseline and the 4-year follow-up. We aimed to exclude hands where all joints had a KL grade of 4 (highest possible grade) from the progression cohort, as these hands would not be subject to further progression based on the KL grading system. However, no hands in our dataset met this exclusion criterion.
The outcomes of worsening of JSN and worsening of osteophytes of hand joints were investigated in the ‘hand ROA incidence cohort’ and the ‘hand ROA progression cohort’. The worsening of JSN and worsening of osteophytes of hand joints were defined as an increase ≥ 1 in the sum of the grades of JSN and osteophytes, respectively, on hand joints between baseline and the 4-year follow-up. We aimed to exclude hands where all joints had a JSN grade of 3 or osteophyte grade of 3 (highest possible grades) from the cohorts, as these hands would not be subject to further progression based on the JSN and osteophytes grading system. However, no hands in our dataset met these specific exclusion criteria.
For hand pain, the reader is reminded that our 2 outcomes were: development of hand pain; and resolution of hand pain, and we used the pain data at the follow-up of 2-year, 4-year, 6-year, and 8-year. Hand pain was defined in the OAI dataset as the presence of “hand/finger pain, aching or stiffness: more than half the days, past 30 days”. Development of hand pain was defined as the presence of hand pain at the respective follow-up when it had not been present in the same hand at baseline. The outcome of development of hand pain was investigated in the ‘hand pain development cohort’, which consisted of hands without hand pain at baseline. Resolution of hand pain was defined as the absence of hand pain at the respective follow-up when it had been present in the same hand at baseline. The outcome of resolution of hand pain was investigated in the ‘hand pain resolution cohort’, which consisted of hands with hand pain at baseline.
Selection of participants (hands)
Before creating the 4 cohorts mentioned above, we excluded participants with the following characteristics: cancer or rheumatoid arthritis at baseline; rheumatoid arthritis at baseline, and Body Mass Index (BMI) in the underweight category (BMI < 18.5 kg/m2) at baseline (Figure 1). We excluded participants with cancer or rheumatoid arthritis at baseline as these diseases may impact the participant’s weight. In addition, we excluded participants with BMI < 18.5 kg/m2 because it may indicate an underlying pathologic disease. From the remaining 4,598 participants, we created a hand ROA cohort and a hand pain cohort. Hand ROA cohort had consisted of participants (n=3,339) with available hand radiographs (primarily dominant hand) and with complete weight data at baseline and 4-year follow-up (Figure 1). Hand pain cohort had consisted of participants (n=2,580) with complete hand pain data (for both hands, left and right) and weight data at baseline and follow-up of 2-year, 4-year, 6-year, and 8-year (Figure 1). From the hand ROA cohort, we created the ‘hand ROA incidence cohort’ by excluding participants with hand ROA at baseline, and the ‘hand ROA progression cohort’ by excluding participants that did not have hand ROA at baseline (Figure 1). From the hand pain cohort, we created the ‘hand pain development cohort’ by excluding hands with pain at baseline, and the ‘hand pain resolution cohort’ by excluding hands with no hand pain at baseline (Figure 1). Note that these 4 cohorts were not mutually exclusive.
Figure 1.
Selection of participants for our investigations. Only data from dominant hand was available for each participant in the Hand ROA cohort. BMI: body mass index; OA: osteoarthritis; OAI: Osteoarthritis Initiative; ROA: radiographic osteoarthritis (i.e., osteoarthritis assessed by radiography).
Statistical analyses
We used logistic regression to estimate the odds ratios (ORs) and 95% confidence intervals (CIs) for the association between our exposure of percentage weight change and our 4 radiographic outcomes (i.e., incidence of hand ROA; progression of hand ROA; worsening of JSN; worsening of osteophytes) over 4 years. Adjusted for sex, race, and the baseline values of the following variables: age, obesity status (BMI) ≥ 30 kg/m2), comorbidity score, presence of hand pain, Physical Activity Score for Elderly (PASE), smoking status, and the sum of KL grades of hand joints investigated in this study. These variables were selected because of their association with exposure and outcomes. Additionally, we adjusted our results for the variability in weight change over time for the investigation of radiographic outcomes. For this purpose, we used standard deviation (SD) of annual weight change over a 4-year follow-up period change for each individual. A higher standard deviation would indicate more variability in weight over time. The assumption of linearity in the association between weight change and our radiographic outcomes of interest was tested using the Box-Tidwell method [19]. There was no violation of this assumption of linearity.
We used mixed-effects logistic regression to estimate the ORs and 95% CIs for the association between percentage weight change from baseline and the development and resolution of hand pain across 2-year, 4-year, 6-year, and 8-year follow-ups. We explored the influence of quadratic and cubic transformations of time variable (i.e., follow-up year) in the models to account for potential non-linear changes over time. As these transformations were statistically significant, we retained the quadratic and cubic forms of time variable in the models, as well as the linear form. To address the inherent correlation between the left- and right-hand measurements within each participant, we specified a random effect for participants. This effectively clustered the data for left and right hands within the same individual, acknowledging their interdependence. Moreover, we incorporated a random effect of hand side to handle the dependence in repeated observations on the same hand over the follow-up duration, and a random slope for the effect of time variable to allow for differences in trajectory over time. We entered the weight change, time (linear form), and the interaction between weight change and linear time variable as predictor variables in the models. The interactions between weight change and non-linear forms of time (i.e., quadratic and cubic) were left out from the model as they were not statistically significant. Our estimates for pain outcomes were adjusted for sex, race, and the baseline values of the following variables: age, obesity status ( (BMI ≥ 30 kg/m2), comorbidity score, PASE, smoking status, and the sum of KL grades of hand joints.
Our study is designed to explore the long-term impact of changes in body weight on changes in hand OA outcomes. Given this emphasis on longitudinal effects, we have selected a methodology that compares all subsequent time points to a common baseline for both the predictors and the outcomes of the development and resolution of hand pain. This approach serves to align our analyses closely with our core research objective, which is to focus on significant changes over extended periods rather than on short-term fluctuations. Furthermore, using a common baseline aids in the straightforward interpretation of clinical relevance, enabling us to make clear statements about how specific changes in weight from baseline are associated with particular outcomes. Additionally, by defining our primary outcomes (i.e., the development and resolution of hand pain) in relation to baseline measurements, this methodology ensures internal consistency in our analyses.
In our analyses, percentage weight change was treated as a continuous variable. We calculated and reported odds ratios for our outcomes based on point estimates of 5% weight loss (because previous studies suggest that this degree of weight loss is clinically relevant [20–22]) and 5% weight gain from baseline to the respective follow-up. While we report the odds ratios based on each 5% weight loss and 5% weight gain, we reported the descriptive statistics for each cohort (i.e., baseline characteristics (Table 1), and number of cases (Table 2), stratified by 3 groups: weight loss (5% or more from baseline to the respective follow-up); stable weight (less than 5% weight change from baseline to the respective follow-up); and weight gain (5% or more from baseline to the respective follow-up).
Table 1.
Baseline characteristics of participants in the hand ROA incidence cohort, hand ROA progression cohort, hand pain development cohort, and hand pain resolution cohort.
| Characteristics | Hand ROA incidence cohort | Hand ROA progression cohort | Hand pain development cohort | Hand pain resolution cohort |
|---|---|---|---|---|
| Participants | n = 1,976 | n = 1,363 | n = 2,178 | n = 654 |
| Hands | n = 1,976 | n = 1,363 | n = 4,104 | n = 1,056 |
| Sex | ||||
| Male | 915 (46.31) | 515 (37.78) | 1,014 (46.56) | 198 (30.28) |
| Female | 1,061 (53.69) | 848 (62.22) | 1,164 (53.44) | 456 (69.72) |
| Race | ||||
| Non-White | 427 (21.61) | 167 (12.25) | 356 (16.35) | 126 (19.27) |
| White | 1,549 (78.39) | 1,196 (87.75) | 1,822 (83.65) | 528 (80.73) |
| Age, mean ± SD years | 57.08 ± 8.06 | 66.33 ± 7.66 | 59.98 ± 8.89 | 61.14 ± 8.62 |
| Obesity status (BMI ≥ 30 kg/m2) | ||||
| No | 1,249 (63.21) | 879 (64.49) | 1,437 (65.98) | 413 (63.15) |
| Yes | 727 (36.79) | 484 (35.51) | 741 (34.02) | 241 (36.85) |
| Obesity status (abdominal circumference ≥ 102 cm for men and ≥ 88 cm for women) | ||||
| No | 620 (32.55) | 303 (22.41) | 671 (31.71) | 164 (25.23) |
| Yes | 1,285 (67.45) | 1,049 (77.59) | 1,445 (68.29) | 486 (74.77) |
| Comorbidity score | ||||
| 0 | 1,596 (81.93) | 1,052 (77.75) | 1,770 (81.64) | 504 (77.90) |
| 1 | 270 (13.86) | 214 (15.82) | 304 (14.02) | 108 (16.69) |
| 2 or more | 82 (4.21) | 87 (6.43) | 94 (4.34) | 35 (5.41) |
| Presence of hand pain | ||||
| No | 1,658 (83.91) | 965 (70.80) | 2,178 (4,104 hands) (100.00) | 0 (0 hands) (0.00) |
| Yes | 318 (16.09) | 398 (29.20) | 0 (0.00) | 654 (1,056 hands) (100.00) |
| PASE, mean ± SD score | 175.13 ± 85.26 | 148.54 ± 74.22 | 169.41 ± 81.92 | 166.63 ± 81.85 |
| Smoking status | ||||
| Never smoked | 1,110 (56.92) | 694 (51.45) | 1,227 (56.70) | 338 (51.92) |
| Past smoker | 709 (36.36) | 601 (44.55) | 823 (38.03) | 277 (42.55) |
| Current smoker | 131 (6.72) | 54 (4.00) | 114 (5.27) | 36 (5.53) |
| Sum of KL of the hand joints, mean ± SD | 1.90 ± 1.96 | 12.90 ± 6.39 | 5.15 ± 6.08 | 8.75 ± 8.35 |
Except where indicated otherwise, values are the number (%). The percentage calculations are based on complete case (i.e., excluding the missing values). BMI: Body Mass Index; KL: Kellgren-Lawrence Grade; PASE: Physical Activity Score for Elderly; SD: Standard Deviation, ROA: Radiographic Osteoarthritis, i.e., osteoarthritis assessed by radiography.
Table 2.
The number of cases of incidence of hand ROA, progression of hand ROA, hand pain development, and hand pain resolution, stratified by weight change from baseline.
| Outcome | Weight loss (5% or more from baseline) | Weight stable (Less than 5% weight change from baseline) | Weight gain (5% or more from baseline) | Total |
|---|---|---|---|---|
| Hand ROA incidence cohort | ||||
| Hands (%) | n = 318 (16.09) | n = 1,267 (64.12) | n = 391 (19.79) | n = 1,976 (100.0) |
| Incidence of hand ROA | ||||
| Year 4 | ||||
| No | 314 (98.74) | 1,243 (98.11) | 383 (97.95) | 1,940 (98.18) |
| Yes | 4 (1.26) | 24 (1.89) | 8 (2.05) | 36 (1.82) |
| Worsening of JSN | ||||
| Year 4 | ||||
| No | 275 (86.48) | 1,103 (87.06) | 337 (86.19) | 1,715 (86.79) |
| Yes | 43 (13.52) | 164 (12.94) | 54 (13.81) | 261 (13.21) |
| Worsening of osteophytes | ||||
| Year 4 | ||||
| No | 245 (77.04) | 981 (77.43) | 294 (75.19) | 1,520 (76.92) |
| Yes | 73 (22.96) | 286 (22.57) | 97 (24.81) | 456 (23.08) |
| Hand ROA progression cohort | ||||
| Hands | n = 252 (18.49) | n = 907 (66.54) | n = 204 (14.97) | n = 1,363 (100.0) |
| Progression of hand ROA | ||||
| Year 4 | ||||
| No | 154 (61.11) | 533 (58.77) | 112 (54.90) | 799 (58.62) |
| Yes | 98 (38.89) | 374 (41.23) | 92 (45.10) | 564 (41.38) |
| Worsening of JSN | ||||
| Year 4 | ||||
| No | 134 (53.17) | 442 (48.73) | 109 (53.43) | 685 (50.26) |
| Yes | 118 (46.83) | 465 (51.27) | 95 (46.57) | 678 (49.74) |
| Worsening of osteophytes | ||||
| Year 4 | ||||
| No | 138 (54.76) | 519 (57.22) | 106 (51.96) | 763 (55.98) |
| Yes | 114 (45.24) | 388 (42.78) | 98 (48.04) | 600 (44.02) |
| Hand pain development cohort | ||||
| Development of hand pain | ||||
| Year 2 | ||||
| Hands | n = 495 (12.06) | n = 3,075 (74.93) | n = 534 (13.01) | n = 4.104 (100.0) |
| No | 404 (81.62) | 2,638 (85.79) | 443 (82.96) | 3,485 (84.92) |
| Yes | 91 (18.38) | 437 (14.21) | 91 (17.04) | 619 (15.08) |
| Year 4 | ||||
| Hands | n = 648 (15.79) | n = 2,687 (65.47) | n = 769 (18.74) | n = 4.104 (100.0) |
| No | 522 (80.56) | 2,300 (85.60) | 621 (80.75) | 3,443 (83.89) |
| Yes | 126 (19.44) | 387 (14.40) | 148 (19.25) | 661 (16.11) |
| Year 6 | ||||
| Hands | n = 867 (21.13) | n = 2,478 (60.38) | n = 759 (18.49) | n = 4.104 (100.0) |
| No | 706 (81.43) | 2,072 (83.62) | 618 (81.42) | 3,396 (82.75) |
| Yes | 161 (18.57) | 406 (16.38) | 141 (18.58) | 708 (17.25) |
| Year 8 | ||||
| Hands | n = 1,039 (25.32) | n = 2,193 (53.44) | n = 872 (21.25) | n = 4.104 (100.0) |
| No | 827 (79.60) | 1,820 (82.99) | 708 (81.19) | 3,355 (81.75) |
| Yes | 212 (20.40) | 373 (17.01) | 164 (18.81) | 749 (18.25) |
| Hand pain resolution cohort | ||||
| Resolution of hand pain | ||||
| Year 2 | ||||
| Hands | n = 153 (14.49) | n = 735 (69.60) | n = 168 (15.91) | n = 1.056 (100.0) |
| No | 90 (58.82) | 373 (50.75) | 95 (56.55) | 558 (52.84) |
| Yes | 63 (41.18) | 362 (49.25) | 73 (43.45) | 498 (47.16) |
| Year 4 | ||||
| Hands | n = 198 (18.75) | n = 667 (63.16) | n = 191 (18.09) | n = 1,056 (100.0) |
| No | 107 (54.04) | 334 (50.07) | 98 (51.31) | 539 (51.04) |
| Yes | 91 (45.96) | 333 (49.93) | 93 (48.69) | 517 (48.96) |
| Year 6 | ||||
| Hands | n = 265 (25.09) | n = 594 (56.25) | n = 197 (18.66) | n = 1,056 (100.0) |
| No | 120 (45.28) | 316 (53.20) | 102 (51.78) | 538 (50.95) |
| Yes | 145 (54.72) | 278 (46.80) | 95 (48.22) | 518 (49.05) |
| Year 8 | ||||
| Hands | n = 283 (26.80) | n = 549 (51.99) | n = 224 (21.21) | n = 1,056 (100.0) |
| No | 134 (47.35) | 275 (50.09) | 125 (55.80) | 534 (50.57) |
| Yes | 149 (52.65) | 274 (49.91) | 99 (44.20) | 522 (49.43) |
Data are presented as count (percentage). JSN: joint space narrowing; OA: osteoarthritis, ROA: Radiographic Osteoarthritis, i.e., osteoarthritis assessed by radiography.
We used STATA/BE 17.0 for Windows (64-bit x86-64) for our analyses. We set our threshold for statistical significance as a two-tailed P value of less than 0.05 in all analyses.
Sensitivity Analyses
We have performed two type of sensitivity analysis. In the first type of sensitivity analysis, we used change in abdominal circumference between baseline and respective follow-up as our exposure, instead of weight change. In our study cohort (i.e., OAI), while obesity defined by BMI (≥ 30 kg/m2) was not associated with incidence and progression of hand ROA, obesity defined by abdominal circumference (≥ 88 cm for female participants and ≥ 102 cm for male participants) was associated with progression, but not incidence, of hand ROA. Therefore, we performed this sensitivity analysis to investigate any association between change in abdominal adiposity and our outcomes of interest. In the second type of sensitivity analysis, we chose to adjust the estimates using BMI at baseline as a continuous variable, as opposed to utilizing obesity defined by BMI (≥ 30 kg/m2) as in our primary analysis. While our primary analysis focused on adjusting for obesity defined by BMI (≥ 30 kg/m2), which is often considered more clinically relevant, it is important to acknowledge that categorizing continuous variables can potentially impact results and lead to different conclusions [23]. By comparing the outcomes of this sensitivity analysis with those of our primary analysis, we aimed to verify if using obesity defined by BMI (≥ 30 kg/m2) in our primary analysis led to different results.
Results
Characteristics of participants in our four cohorts
Table 1 shows baseline characteristics of the participants in each of our four cohorts, stratified by weight change categories between baseline and the respective follow-up. In each of our four cohorts, most participants were female and White. The mean age in four cohorts ranged between 57.08 ± 8.06 and 66.33± 7.66 years old. While majority of participants were not in obesity category by BMI (BMI ≥ 30 kg/m2), they were in obesity category by abdominal circumference (abdominal circumference ≥ 102 cm for men and ≥ 88 cm for women).
Weight change and the incidence of hand ROA (hand ROA incidence cohort)
In the hand ROA incidence cohort (i.e., the cohort free from overall hand ROA at baseline), the mean percentage weight change over four years was −9.7 ± 5.0 for those who experienced a weight loss of 5% or more, and 8.9 ± 4.1 for those who gained 5% or more in weight (Table S1). At the 4-year follow-up, overall hand ROA was observed in 36 (1.82%) out of the 1,976 hands (Table 2). With an incidence rate of 1.82%, it was relevant to evaluate the statistical power of our study. Using a standard significance level of 0.05 and an expected hand ROA incidence rate of 5.6% from a previous study by Eaton et a [24], our power calculation indicates that our study was adequately powered to detect statistically significant differences if present.
There was no evidence of an association between weight loss and weight gain with the odds of incidence of hand ROA by the 4-year follow-up (Table 3). Similarly, there was no evidence of an association between weight loss and weight gain with the odds of worsening of JSN or worsening of osteophytes over 4 years (Table 3). There was no visible difference in the trajectory of annual mean percentage weight change from baseline over 4 years between those who had incidence of hand ROA and who did not have incidence of hand ROA at 4-year follow-up (Figure 2a).
Table 3.
Association of weight loss and gain with the odds of outcomes investigated.
| Outcome | 5% weight loss | 5% weight gain |
|---|---|---|
| Hand ROA incidence cohort | ||
| Incidence of hand ROA (Year 4) | ||
| Odds Ratio (95% CI, P-value) | 0.90 (0.67-1.23, P = 0.52) | 1.11 (0.82-1.50, P = 0.52) |
| Worsening of JSN (Year 4) | ||
| Odds Ratio (95% CI, P-value) | 0.94 (0.84-1.06, P = 0.30) | 1.06 (0.95-1.19, P = 0.30) |
| Worsening of osteophytes (Year 4) | ||
| Odds Ratio (95% CI, P-value) | 0.96 (0.88-1.05, P = 0.35) | 1.04 (0.96-1.14, P = 0.35) |
| Hand ROA progression cohort | ||
| Progression of hand ROA (Year 4) | ||
| Odds Ratio (95% CI, P-value) | 0.92 (0.84-1.00, P = 0.06) | 1.09 (1.00-1.19, P = 0.06) |
| Worsening of JSN (Year 4) | ||
| Odds Ratio (95% CI, P-value) | 0.97 (0.88-1.06, P = 0.52) | 1.03 (0.94-1.13, P = 0.52) |
| Worsening of osteophytes (Year 4) | ||
| Odds Ratio (95% CI, P-value) | 0.93 (0.85-1.02, P = 0.12) | 1.08 (0.98-1.18, P = 0.12) |
| Hand pain development cohort | ||
| Development of hand pain | ||
| Year 2 | ||
| Odds Ratio (95% CI, P-value) | 1.00 (0.98-1.02, P = 0.98) | 1.00 (0.98-1.02, P = 0.98) |
| Year 4 | ||
| Odds Ratio (95% CI, P-value) | 1.00 (0.96-1.04, P = 0.98) | 1.00 (0.96-1.04, P = 0.98) |
| Year 6 | ||
| Odds Ratio (95% CI, P-value) | 1.00 (0.94-1.07, P = 0.98) | 1.00 (0.94-1.07, P = 0.98) |
| Year 8 | ||
| Odds Ratio (95% CI, P-value) | 1.00 (0.92-1.09, P = 0.98) | 1.00 (0.92-1.09, P = 0.98) |
| Hand pain resolution cohort | ||
| Resolution of hand pain | ||
| Year 2 | ||
| Odds Ratio (95% CI, P-value) | 1.02 (0.98-1.06, P = 0.44) | 0.99 (0.95-1.02, P = 0.44) |
| Year 4 | ||
| Odds Ratio (95% CI, P-value) | 1.03 (0.95-1.12, P = 0.44) | 0.97 (0.89-1.05, P = 0.44) |
| Year 6 | ||
| Odds Ratio (95% CI, P-value) | 1.05 (0.93-1.18, P = 0.44) | 0.95 (0.84-1.08, P = 0.44) |
| Year 8 | ||
| Odds Ratio (95% CI, P-value) | 1.07 (0.91-1.25, P = 0.44) | 0.94 (0.80-1.10, P = 0.44) |
Results are reported as point estimates of 5% weight loss and gain from baseline to respective follow-up. Adjusted for sex, race, standard deviation (SD) of weight change over respective follow-up period (radiographic outcomes only), time (i.e., follow-up period, only in pain outcomes), and the baseline values of the following variables: age, obesity status (Body Mass Index (BMI) ≥ 30 kg/m2), comorbidity score, presence of hand pain (radiographic outcomes only), Physical Activity Score for Elderly (PASE), smoking status, and the sum of Kellgren-Lawrence (KL) grades of hand joints. CI: Confidence Interval; JSN: Joint Space Narrowing; OA: Osteoarthritis; ROA: Radiographic Osteoarthritis, i.e., osteoarthritis assessed by radiography.
Figure 2.
Weight change trajectories by outcomes investigated. The mean percentage weight change (SD) values are calculated from the raw values. ROA: radiographic osteoarthritis (i.e., osteoarthritis assessed by radiography). a) by incidence of hand ROA at 4-year follow-up; b) by progression of hand ROA at 4-year follow-up; c) by development of hand pain at 2-year, 4-year, 6-year, and 8-year follow-up; and d) by resolution of hand pain at 2-year, 4-year, 6-year, and 8-year follow-up. SD: Standard Deviation.
Weight change and the progression of hand ROA (hand ROA progression cohort)
In the hand ROA progression cohort (i.e., the cohort with overall hand ROA at baseline), the mean percentage weight change over four years was −9.0 ± 3.9 for those who experienced a weight loss of 5% or more, and 9.6 ± 5.3 for those who gained 5% or more in weight (Table S1). At the 4-year follow-up, progression of hand ROA was observed in 564 (41.38%) out of the 1,363 hands (Table 2).
There was no evidence of association between weight loss and weight gain with the odds of progression of overall hand ROA by the 4-year follow-up (Table 3). Similarly, no evidence of an association was observed for worsening of JSN or worsening of osteophytes over 4 years (Table 3). There was no visible difference in the trajectory of annual mean percentage weight change from baseline over 4 years between those who had progression of hand ROA and who did not have progression of hand ROA at 4-year follow-up (Figure 2b).
Weight change and the development of hand pain (hand pain development cohort)
In the hand pain development cohort (i.e., the cohort free from hand pain at baseline), the mean percentage weight change over eight years was between −9.0 ± 4.5 and −10.3 ± 4.9 for those who experienced a weight loss of 5% or more, and was between 8.6 ± 5.0 and 10.0 ± 6.9 for those who gained 5% or more in weight (Table S1). During the 8 years follow-up, development of hand pain was observed in between 619 (15.08%) and 749 (18.25%) out of the 4,104 hands (Table 2).
There was no evidence of association between weight loss and weight gain with the odds of development of hand pain at any time points (Table 3). There was no visible difference in the trajectory of biennial (every two years) mean percentage weight change from baseline over 8 years between those who had development of hand pain and who did not have development of hand pain at the 2-year, 4-year, 6-year, and 8-year follow-ups (Figure 2c).
Weight change and the resolution of hand pain (hand pain resolution cohort)
In the hand pain resolution cohort (i.e., the cohort with hand pain at baseline), the mean percentage weight change over eight years was between −9.5 ± 4.4 and −10.2 ± 4.6 for those who experienced a weight loss of 5% or more, and was between 7.6 ± 2.9 and 9.9 ± 4.0 for those who gained 5% or more in weight (Table S1). During the 8 years follow-up, development of hand pain was observed in between 498 (47.16%) and 522 (49.43%) out of the 1,056 hands (Table 2).
There was no evidence of association between weight loss and weight gain with the odds of resolution of hand pain at any time points (Table 3). There was no visible difference in the trajectory of biennial (every two years) mean percentage weight change from baseline over 8 years between those who had resolution of hand pain and who did not have development of hand pain at the 2-year, 4-year, 6-year, and 8-year follow-ups (Figure 2d).
Sensitivity analyses
Our first type of sensitivity analysis showed no evidence of an association between change in abdominal circumference and any of the outcomes (Table S2). We reported the mean absolute change in abdominal circumference in four cohorts in Table S3, and the number of cases of outcomes stratified by abdominal circumference change from baseline in Table S4. In the second type of sensitivity analysis, in which we adjusted the estimates using BMI as a continuous variable, the results closely mirrored those obtained in our primary analysis, where we adjusted the estimates using obesity defined by BMI (≥ 30 kg/m2) (Table S5).
Discussion
Our study revealed no association between weight change (weight loss or gain) and either the incidence or progression of radiographic hand OA over 4 years or the development or resolution of hand pain over 8 years.
This research provides the first evidence in support of current clinical guidelines [25–28] that refrain from recommending weight management as a treatment for hand OA due to previous research gaps. However, recommending weight loss for people with hand OA is sensible if hand OA coexists with knee OA, given the known benefits of weight loss for knee OA [29, 30]. Although the prevalence of the co-existing knee and hand OA is lower than that of isolated hand OA, a substantial percentage of people have both conditions. For example, data from the Canadian Longitudinal Study on Aging (CLSA) [31] showed that of 2,212 people with hand OA, 1,514 (68.4%) had hand OA only and 698 (31.6%) had a coexistence of knee and hand OA. Given this data, healthcare professionals should be cognizant of the potential coexistence of hand and knee OA and tailor their weight management recommendations accordingly for comprehensive patient care.
The absence of a relationship between weight change and hand OA could be attributed to hand joints not bearing any force from excess weight [32]. Therefore, the effect of weight change on OA in joints could be considered to occur through biomechanical mechanisms. This theory is supported by the observational studies from our team [33–37] and others [20, 30, 38–44] that have shown that while weight change is linked to structural defects of the joint [20, 34–36, 38–41] and symptoms [20, 30, 42–44] of knee OA, a weight-bearing joint, it is not associated with structural defects of the joint [20, 34, 35, 37, 41] and symptoms [20, 35, 37] of hip OA, which is less sensitive to mechanical loading from excess weight [45, 46]. Additionally, in one of our studies mentioned above [36], we showed that weight change is linked to structural defects on the medial side of the knee, but not on the lateral side, which bears less force from weight than the medial side of the knee [47]. Furthermore, our sensitivity analysis found that change in abdominal circumference, which is an indicator for metabolic factors [48], was not linked to the outcomes of hand OA. Therefore, it appears that neither changes in biomechanical factors nor metabolic factors were associated with hand OA over 4 to 8 years. These findings support the conclusion that weight change affects joints through biomechanical factors.
While our study did not find any relationship between weight change and hand OA, and we did not specifically investigate inflammation, the connection between weight change and OA through inflammation remains an area of interest [49, 50]. Significant weight gain has been linked to systemic inflammation [51], and it can exacerbate OA in weight-bearing joints like the knee and hip [52]. However, the effect on hand OA may be different and more complex, possibly involving unique inflammatory pathways or other factors may be linked to inflammation such as mechanical stress [53] and genetics [54]. Though our study did not directly explore these connections, understanding how weight change might influence inflammatory pathways in hand OA could still be important for future research and treatments. More targeted studies are needed to explore this complex relationship.
This study has several limitations. Firstly, our findings are associative due to the nature of observational studies. Secondly, there were possibly latent confounders that were not captured in our analyses or in the OAI cohort from which we sourced the data for this study. Thirdly, our study included people with or at risk of knee OA; therefore, the generalizability of our findings is limited to this population. A fourth limitation of the current study is the lack of data on occupation within the OAI dataset. This limitation meant that we could not adjust for occupation in our analysis, even though certain occupations may contribute to the risk of hand OA. Therefore, any conclusions drawn from our findings must take this limitation into account. To lessen the effect of this limitation, we performed our main analyses with an extra adjustment for whether people were working for pay or not at the start of the study (a binary yes-or-no variable). The results from these adjusted analyses were mainly the same or very close to the results from our main analyses. As a fifth limitation, the 4-year follow-up period for radiographic changes in our study might not have been sufficient for some participants to develop or progress any structural changes in hand joints due to slow progression of OA. Despite this limitation, our study remains relevant. The 5.6% incidence of radiographic hand OA over 48 months in the OAI dataset supports our approach. Given that most clinical trials in this field rarely exceed 2 or 3 years, our study aligns with standard research practices. While a longer observation period could provide stronger evidence, we believe it is unlikely to alter our overall findings, as we have identified enough progression cases within our sample. As a final limitation, obesity was not associated with hand OA in our study cohort (i.e., OAI) [7, 55]. Therefore, future research to investigate the association of weight change with hand OA should be conducted in a cohort in which obesity is associated with hand OA. Furthermore, in our study cohort, only 85 of 4,796 people (1.8%) had a BMI ≥ 40 kg/m2. Therefore, our data may not be generalized to those with BMI ≥ 40 kg/m2. Despite this limitation, this study may yield valuable insights for the large population with a BMI between 18.5 to 40 kg/m2.
In conclusion, our study found no association between weight change (whether loss or gain) and either structural changes in hand OA as detected by radiography or hand pain over a span of 4 to 8 years.
Supplementary Material
Significance and Innovations.
This study is the first, to our knowledge, to investigate the association between weight change (weight loss or weight gain) and hand OA using a large, population-based cohort.
Weight loss or weight gain over 4 years shows no significant association with the incidence or progression of radiographic hand osteoarthritis.
Weight loss or weight gain over 8 years show no significant association with the development or resolution of hand pain.
Acknowledgments
This article was prepared using an Osteoarthritis Initiative (OAI) public-use data set, and its contents do not necessarily reflect the opinions or views of the OAI Study Investigators, the NIH, or the private funding partners of the OAI. The OAI data repository is housed within the National Institute of Mental Health Data Archive. The OAI is a public–private partnership between the NIH (contracts N01-AR-2-2258, N01-AR-2-2259, N01-AR-2-2260, N01-AR-2-2261, and N01-AR-2-2262) and private funding partners (Merck Research Laboratories, Novartis Pharmaceuticals, GlaxoSmithKline, and Pfizer, Inc.) and is conducted by the OAI Study Investigators. Private sector funding for the OAI is managed by the Foundation for the NIH. The authors of this article are not part of the OAI investigative team.
We would like to express our gratitude to the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) of the National Institutes of Health (NIH), which provided support under Award Number R01 AR066378 (primary investigator: Dr. McAlindon). This grant enabled the reading of X-rays for hand osteoarthritis in this study.
Conflicts of interest
Zubeyir Salis and Amanda Sainsbury own 50% each of the shares in Zuman International Pty Ltd, which receives royalties and other payments for educational resources and services in adult weight management and research methodology. Amanda Sainsbury additionally reports receiving presentation fees and travel reimbursements from Eli Lilly and Co, the Pharmacy Guild of Australia, Novo Nordisk, the Dietitians Association of Australia, Shoalhaven Family Medical Centres, the Pharmaceutical Society of Australia, and Metagenics, and serving on the Nestlé Health Science Optifast VLCD advisory board from 2016 to 2018. JBD declares that he is a consultant for Pfizer Inc and Eli Lilly and Company. TEM declares that he is a consultant for Remedium-Bio, Anika, Chemocentryx, Grunenthal, Kolon Tissue Gene, Novartis, BioSplice, Organogenesis, and Pfizer Inc.
Funding sources
None. No funding source had a role in the design, analyses, interpretation of data, or decision to submit the results in this study.
Data availability
The datasets were derived from the Osteoarthritis Initiative in the public domain and can be accessed through the National Institute of Mental Health Data Archive.
References
- 1.Magni N, et al. , Neuropathic Pain in Hand Osteoarthritis: A Cross-Sectional Study. J Clin Med, 2021. 10(19). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Michon M, Maheu E, and Berenbaum F, Assessing health-related quality of life in hand osteoarthritis: a literature review. Ann Rheum Dis, 2011. 70(6): p. 921–8. [DOI] [PubMed] [Google Scholar]
- 3.Hunter DJ and Bierma-Zeinstra S, Osteoarthritis. Lancet, 2019. 393(10182): p. 1745–1759. [DOI] [PubMed] [Google Scholar]
- 4.Qin J, et al. , Lifetime Risk of Symptomatic Hand Osteoarthritis: The Johnston County Osteoarthritis Project. Arthritis Rheumatol, 2017. 69(6): p. 1204–1212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kloppenburg M and Kwok W-Y, Hand osteoarthritis—a heterogeneous disorder. Nature Reviews Rheumatology, 2012. 8(1): p. 22–31. [DOI] [PubMed] [Google Scholar]
- 6.Terpstra SES, et al. , Hand osteoarthritis is associated with limitations in paid and unpaid work participation and related societal costs: the HOSTAS cohort. RMD Open, 2022. 8(2): p. e002367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Eaton CB, et al. , Prevalence, Incidence, and Progression of Radiographic and Symptomatic Hand Osteoarthritis: The Osteoarthritis Initiative. Arthritis Rheumatol, 2022. [DOI] [PubMed] [Google Scholar]
- 8.Plotz B, et al. , Current Epidemiology and Risk Factors for the Development of Hand Osteoarthritis. Curr Rheumatol Rep, 2021. 23(8): p. 61. [DOI] [PubMed] [Google Scholar]
- 9.Yusuf E, et al. , Association between weight or body mass index and hand osteoarthritis: a systematic review. Annals of the Rheumatic Diseases, 2010. 69(4): p. 761–765. [DOI] [PubMed] [Google Scholar]
- 10.Rydberg M, et al. , High body mass index is associated with increased risk for osteoarthritis of the first carpometacarpal joint during more than 30 years of follow-up. RMD Open, 2020. 6(3). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Reyes C, et al. , Association Between Overweight and Obesity and Risk of Clinically Diagnosed Knee, Hip, and Hand Osteoarthritis: A Population-Based Cohort Study. Arthritis and Rheumatology, 2016. 68(8): p. 1869–1875. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Jiang L, et al. , Body mass index and hand osteoarthritis susceptibility: an updated meta-analysis. Int J Rheum Dis, 2016. 19(12): p. 1244–1254. [DOI] [PubMed] [Google Scholar]
- 13.Osteoarthritis Initiative (OAI) Study Protocol. https://nda.nih.gov/static/docs/StudyDesignProtocolAndAppendices.pdf.
- 14.Kellgren JH and Lawrence JS, Radiological assessment of osteo-arthrosis. Ann Rheum Dis, 1957. 16(4): p. 494–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Haugen IK, et al. , Prevalence, incidence and progression of hand osteoarthritis in the general population: the Framingham Osteoarthritis Study. Ann Rheum Dis, 2011. 70(9): p. 1581–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Manninen P, et al. , Weight changes and the risk of knee osteoarthritis requiring arthroplasty. Annals of the Rheumatic Diseases, 2004. 63(11): p. 1434. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Acheson RM, Chan YK, and Clemett AR, New Haven survey of joint diseases. XII. Distribution and symptoms of osteoarthrosis in the hands with reference to handedness. Ann Rheum Dis, 1970. 29(3): p. 275–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Jónsson H, et al. , High hand joint mobility is associated with radiological CMC1 osteoarthritis: the AGES-Reykjavik study. Osteoarthritis Cartilage, 2009. 17(5): p. 592–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zhang S, Su P, and Liu S, Fusion of Cognitive Information: Evaluation and Evolution Method of Product Image Form. Comput Intell Neurosci, 2021. 2021: p. 5524093. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Joseph GB, et al. , Effects of weight change on knee and hip radiographic measurements and pain over 4 years: Data from the Osteoarthritis Initiative. Arthritis Care & Research, 2022. n/a(n/a). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ensrud KE, et al. , Intentional and Unintentional Weight Loss Increase Bone Loss and Hip Fracture Risk in Older Women. Journal of the American Geriatrics Society, 2003. 51(12): p. 1740–1747. [DOI] [PubMed] [Google Scholar]
- 22.Williamson DA, Bray GA, and Ryan DH, Is 5% weight loss a satisfactory criterion to define clinically significant weight loss? Obesity (Silver Spring), 2015. 23(12): p. 2319–20. [DOI] [PubMed] [Google Scholar]
- 23.Salis Z, Gallego B, and Sainsbury A, Researchers in rheumatology should avoid categorization of continuous predictor variables. BMC Medical Research Methodology, 2023. 23(1): p. 104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Eaton CB, et al. , Prevalence, Incidence, and Progression of Radiographic and Symptomatic Hand Osteoarthritis: The Osteoarthritis Initiative. Arthritis Rheumatol, 2022. 74(6): p. 992–1000. [DOI] [PubMed] [Google Scholar]
- 25.Kolasinski S, et al. , American College of Rheumatology/Arthritis Foundation guideline for the Management of Osteoarthritis of the hand, hip, and knee. Arthritis Rheumatol. 2020; 72 (2): 220–33. Updated evidence-based recommendations for the management of hand, hip and knee osteoarthritis based on systematic review of published randomized clinical trials, 2019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kloppenburg M, et al. , 2018 update of the EULAR recommendations for the management of hand osteoarthritis. Annals of the rheumatic diseases, 2019. 78(1): p. 16–24. [DOI] [PubMed] [Google Scholar]
- 27.Ariani A, et al. , The Italian Society for Rheumatology clinical practice guidelines for the diagnosis and management of knee, hip and hand osteoarthritis. Reumatismo, 2019. 71(S1): p. 5–21. [DOI] [PubMed] [Google Scholar]
- 28.Rillo O, et al. , PANLAR consensus recommendations for the management in osteoarthritis of hand, hip, and knee. JCR: Journal of Clinical Rheumatology, 2016. 22(7): p. 345–354. [DOI] [PubMed] [Google Scholar]
- 29.Christensen R, et al. , Effect of weight reduction in obese patients diagnosed with knee osteoarthritis: a systematic review and meta-analysis. Ann Rheum Dis, 2007. 66(4): p. 433–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Messier SP, et al. , Effects of intensive diet and exercise on knee joint loads, inflammation, and clinical outcomes among overweight and obese adults with knee osteoarthritis: the IDEA randomized clinical trial. Jama, 2013. 310(12): p. 1263–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Badley EM, et al. , The relationship between body mass index and osteoarthritis for single and multi-site osteoarthritis of the hand, hip, or knee: findings from the CLSA. Arthritis Care & Research, 2021. [DOI] [PubMed] [Google Scholar]
- 32.Yusuf E, Metabolic factors in osteoarthritis: Obese people do not walk on their hands. Arthritis Research and Therapy, 2012. 14(4). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Salis Z, et al. , Weight loss is associated with reduced risk of knee and hip replacement: a survival analysis using Osteoarthritis Initiative data. International Journal of Obesity, 2022. [DOI] [PubMed] [Google Scholar]
- 34.Salis Z and Sainsbury A, Association between change in body mass index and knee and hip replacements: a survival analysis of 7 to 10 years using multi-cohort data. Arthritis Care and Research, 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Salis Z and Sainsbury A, Association of change in BMI with incidence and progression of the structural defects of hip osteoarthritis: data from the OAI and CHECK studies. Arthritis Care & Research. n/a(n/a). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Salis Z, et al. , Decrease in body mass index is associated with reduced incidence and progression of the structural defects of knee osteoarthritis: a prospective multi-cohort study. Arthritis & Rheumatology. n/a(n/a). [DOI] [PubMed] [Google Scholar]
- 37.Salis Z, et al. , Investigation of the association of weight loss with radiographic hip osteoarthritis in older community-dwelling female adults. Journal of the American Geriatrics Society, 2023. 71(8): p. 2451–2461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Felson DT, et al. , Risk factors for incident radiographic knee osteoarthritis in the elderly: the Framingham Study. Arthritis Rheum, 1997. 40(4): p. 728–33. [DOI] [PubMed] [Google Scholar]
- 39.de Vos BC, et al. , Long-term effects of a lifestyle intervention and oral glucosamine sulphate in primary care on incident knee OA in overweight women. Rheumatology (United Kingdom), 2017. 56(8): p. 1326–1334. [DOI] [PubMed] [Google Scholar]
- 40.Jin X, et al. , Does weight loss reduce the incidence of total knee and hip replacement for osteoarthritis?-A prospective cohort study among middle-aged and older adults with overweight or obesity. Int J Obes (Lond), 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Heuts EAF, et al. , The influence of bariatric surgery on hip and knee joint pain: a systematic review. Surg Obes Relat Dis, 2021. 17(9): p. 1637–1653. [DOI] [PubMed] [Google Scholar]
- 42.Messier SP, et al. , Intentional Weight Loss in Overweight and Obese Patients With Knee Osteoarthritis: Is More Better? Arthritis Care & Research, 2018. 70(11): p. 1569–1575. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Atukorala I, et al. , Is There a Dose-Response Relationship Between Weight Loss and Symptom Improvement in Persons With Knee Osteoarthritis? Arthritis Care Res (Hoboken), 2016. 68(8): p. 1106–14. [DOI] [PubMed] [Google Scholar]
- 44.Messier SP, et al. , Exercise and Dietary Weight Loss in Overweight and Obese Older Adults with Knee Osteoarthritis: The Arthritis, Diet, and Activity Promotion Trial. Arthritis and Rheumatism, 2004. 50(5): p. 1501–1510. [DOI] [PubMed] [Google Scholar]
- 45.Singer SP, et al. , Maximum lifetime body mass index is the appropriate predictor of knee and hip osteoarthritis. Archives of Orthopaedic and Trauma Surgery, 2018. 138(1): p. 99–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Reijman M, et al. , Body mass index associated with onset and progression of osteoarthritis of the knee but not of the hip: the Rotterdam Study. Ann Rheum Dis, 2007. 66(2): p. 158–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Mündermann A, et al. , In vivo knee loading characteristics during activities of daily living as measured by an instrumented total knee replacement. Journal of Orthopaedic Research, 2008. 26(9): p. 1167–1172. [DOI] [PubMed] [Google Scholar]
- 48.Shen W, et al. , Waist circumference correlates with metabolic syndrome indicators better than percentage fat. Obesity (Silver Spring), 2006. 14(4): p. 727–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Zhai G, et al. , Reduction of leucocyte telomere length in radiographic hand osteoarthritis: a population-based study. Ann Rheum Dis, 2006. 65(11): p. 1444–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Sokolove J and Lepus CM, Role of inflammation in the pathogenesis of osteoarthritis: latest findings and interpretations. Ther Adv Musculoskelet Dis, 2013. 5(2): p. 77–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Das UN, Is obesity an inflammatory condition? Nutrition, 2001. 17(11-12): p. 953–66. [DOI] [PubMed] [Google Scholar]
- 52.Felson DT and Chaisson CE, Understanding the relationship between body weight and osteoarthritis. Baillieres Clin Rheumatol, 1997. 11(4): p. 671–81. [DOI] [PubMed] [Google Scholar]
- 53.Visser AW, et al. , The relative contribution of mechanical stress and systemic processes in different types of osteoarthritis: the NEO study. Annals of the Rheumatic Diseases, 2015. 74(10): p. 1842–1847. [DOI] [PubMed] [Google Scholar]
- 54.Felson DT, et al. , Evidence for a Mendelian gene in a segregation analysis of generalized radiographic osteoarthritis: the Framingham Study. Arthritis Rheum, 1998. 41(6): p. 1064–71. [DOI] [PubMed] [Google Scholar]
- 55.Haugen IK, et al. , Neither obesity, smoking and alcohol are associated with hand osteoarthritis (OA)-longitudinal data from osteoarthritis initiative (OAI). Annals of the Rheumatic Diseases, 2016. 75 (Supplement 2): p. 827. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets were derived from the Osteoarthritis Initiative in the public domain and can be accessed through the National Institute of Mental Health Data Archive.