Abstract
Objective
This study aimed to investigate the effect of smoking on the prevalence, incidence, and progression of hip osteoarthritis (OA). We used data from the Osteoarthritis Initiative (OAI) and the Cohort Hip and Cohort Knee (CHECK) studies.
Methods
We analyzed 9,386 hips from 4,716 participants (OAI cohort) and 1,954 hips from 977 participants (CHECK cohort). The primary exposure was smoking status at baseline, categorized as current, former, or never smoker. Outcomes of radiographic hip OA (RHOA) and symptomatic hip OA were evaluated both cross‐sectionally at baseline and longitudinally over a 4‐ to 5‐year follow‐up, with adjustments for major covariates.
Results
No significant differences were observed between current or former smokers and never smokers for any of the outcomes examined, either at baseline or at the 4‐ to 5‐year follow‐up. In the cross‐sectional analysis, the odds ratios with 95% confidence intervals for the prevalence of RHOA for current and former smokers were 1.29 (0.68–2.46) and 0.99 (0.70–1.40) in the OAI cohort and 1.38 (0.78–2.44) and 0.85 (0.54–1.32) in the CHECK cohort, respectively. In the longitudinal analysis, odds ratio with 95% confidence intervals for the incidence of RHOA were 1.03 (0.23–4.50) and 0.92 (0.46–1.85) in the OAI cohort and 0.61 (0.34–1.11) and 1.00 (0.69–1.44) in the CHECK cohort, respectively.
Conclusion
Our study found no clear association between smoking and the prevalence, incidence, or progression of RHOA or symptomatic hip OA, either at baseline or over a 4‐ to 5‐year period.
INTRODUCTION
Osteoarthritis (OA), a prevalent degenerative joint disease, is characterized by significant changes, including loss of cartilage, bone modifications, and formation of osteophytes, leading to joint lining inflammation. 1 Hip OA is a particularly severe manifestation of this disease, often leading to intense pain and functional loss in the hip joint. 2 Aging, obesity, and joint injury have been identified as significant contributing factors to the incidence of hip OA. 3 However, the influence of smoking—a risk factor that can be modified—on the incidence and progression of hip OA remains to be definitively understood.
Several studies, such as the systematic review by Hui et al, have investigated the potential role of smoking in hip OA, exploring whether it might have a protective effect against the incidence of hip OA. 4 Although their systematic review, including 13 studies, suggested no link between smoking and the incidence of hip OA, their conclusions were drawn from analyses that collectively considered hip OA alongside other types of OA, primarily knee OA. This approach is potentially problematic as OA can manifest differently in various joints because of distinct mechanisms. Therefore, studies that merged hip OA with other types, such as knee OA, 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 might underestimate the influence of smoking on the risk of hip OA. Indeed, hip OA has unique biomechanical and biochemical characteristics that differ from other types of OA, warranting separate investigation. For example, it has been noted that weight change influences knee and hip OA differently, with the hip joint showing more resistance to weight‐related effects. 13 , 14 This resistance has been attributed to its unique ball‐and‐socket structure, 15 , 16 lower susceptibility to metabolic damage, 13 and thinner cartilage, 17 as compared with the knee joint. These distinctions highlight the need for a separate investigation into the risk factors affecting hip OA. Furthermore, the systematic review by Hui et al 4 observed significant variability in the definition of smoking exposure, outcomes, and study designs across the 13 studies included in their review. This finding, along with a later conclusion by Felson et al suggesting a potential modest protective role of smoking against hip OA incidence based on existing data, 18 amplifies the uncertainty surrounding the specific impact of smoking on the incidence of hip OA, particularly regarding structural changes and pain.
Parallelly, studies such as the one conducted by Pearce et al explored whether smoking may slow the progression of hip OA 19 in people with hip OA. Their systematic review of studies with total hip replacement (THR) as outcome (individually or combined with total knee replacement) found no relationship between smoking and the progression to THR in participants with hip OA. Roux et al recently studied the relationship between smoking and structural progression in knee and hip OA in participants with symptomatic lower limb OA over 3 years in an observational study. 20 Although they found no influence of smoking on structural aspects of hip OA, they noted fewer osteophytes in the knees of current and former smokers. This complex relationship necessitates further exploration into the specific impacts of smoking on the progression of hip OA, including its effects on various structural changes.
Although many previous studies have primarily focused on structural changes in hip OA, research examining pain outcomes is limited. Existing studies have mostly concentrated on the development of clinically or self‐reported hip OA, largely neglecting the exploration of specific pain outcomes. The methodologies often involve cross‐sectional 21 , 22 , 23 or case‐control designs, 5 , 6 , 24 , 25 and only four studies have employed a longitudinal approach. 8 , 20 , 26 , 27 Although most of these studies involving cross‐sectional and case‐control designs show no connection between smoking and hip OA development, some exceptions exist; Mühlen et al 21 found a detrimental link, whereas Rego et al 25 reported a protective effect. Among four longitudinal studies, 8 , 20 , 26 , 27 two studies (Hootman et al 8 and Cheng et al 26 ) primarily relied on self‐reported development of hip OA. They found no association between smoking and the development of hip OA. The third longitudinal study by Roux et al 20 investigated the association of smoking with evolution of self‐reported Harris scores, which included elements of pain and function in hip OA, over 3 years. They found no association between smoking and evolution of Harris scores. The last study of the four longitudinal studies by Juhakoski et al 27 followed 840 participants for 22 years and used clinical examinations to assess hip OA. This study concluded that smoking was not a predictive risk factor for developing clinical hip OA. Despite these contributions, existing studies are limited by their reliance on self‐reported outcomes and lack of focus on pain‐related outcomes. This highlights the need for more comprehensive, in particular longitudinal, studies to address these gaps.
Given these gaps, mixed findings, and limitations in previous studies, including the potential underestimation of the influence of smoking on hip OA because of the inclusion of other OA types in the analysis and the variability in defining smoking exposure and outcomes, our study aims to provide more conclusive insights. We leverage data from two comprehensive and rigorously characterized cohorts: the Osteoarthritis Initiative (OAI) 28 and the Cohort Hip and Cohort Knee (CHECK) study. 29 By thoroughly examining both radiographic and pain data in cross‐sectional and longitudinal designs, we aim to provide a more comprehensive understanding of how smoking potentially influences the prevalence, incidence, and progression of structural changes in the hip joint and pain outcomes in hip OA. We hypothesize that smoking may influence the prevalence, incidence, and progression of radiographically assessed structural changes in hip OA and the development and resolution of pain outcomes in hip OA over time. The potential identification of smoking as a modifiable risk factor for hip OA could have significant implications for preventing and managing this condition, ultimately leading to improved clinical outcomes and public health strategies.
MATERIALS AND METHODS
Data sources and ethics
This study analyzed data from the OAI and CHECK cohorts obtained from publicly available databases. Both cohorts were part of multicenter, longitudinal, prospective observational studies focusing on individuals with or at risk of OA. The OAI enrolled 4,796 participants aged 45 to 79 from four clinical centers in the United States between February 2004 and May 2006. The CHECK cohort consisted of 1,002 participants aged 45 to 65, who were recruited from 10 clinical centers in the Netherlands between October 2002 and September 2005. The original OAI and CHECK studies received ethical approval from their institutional boards, and all participants in the original OAI and CHECK studies provided informed written consent.
Baseline and follow‐up data
We used radiographic and pain data from the OAI study's baseline and 4‐year follow‐up and from the CHECK study's baseline and 5‐year follow‐up. Beyond the 4‐year follow‐up, the OAI study only provided radiographic data for a subset of participants. Likewise, the CHECK study did not include radiographic data for the 4‐year follow‐up, only for the 5‐year follow‐up.
Selection of participants and hips
This study applied selection criteria at participant and hip levels in the OAI and CHECK cohorts. Firstly, at the participant level, we excluded participants who had missing smoking status data at baseline (Figure 1). Secondly, at the hip level, this study excluded hips that had been replaced before baseline (Figure 1). The remaining hips established the OAI smoking study cohort with 9,386 hips from 4,716 participants and the CHECK smoking study cohort with 1,954 hips from 977 participants (Figure 1). We then created four cohorts from the OAI smoking study cohort and CHECK smoking study cohort to analyze the radiographic and pain outcomes (defined below): OAI radiographic cohort, CHECK radiographic cohort, OAI pain cohort, and CHECK pain cohort. We created the OAI radiographic cohort by excluding hips with missing overall radiographic grade (a grade using a combination of modified Croft grade and individual radiographic features [IRFs], a detailed definition is below) at baseline or at the 4‐year follow‐up, and the CHECK radiographic cohort by excluding the hips with missing Kellgren Lawrence (KL) grade at baseline or the 5‐year follow‐up (Figure 1). We created the OAI pain cohort by excluding hips that either had missing pain data or showed consistent hip pain at both the baseline and 4‐year follow‐up. Similarly, for the CHECK pain cohort, hips were excluded if they had missing pain data or showed consistent hip pain at both the baseline and 5‐year follow‐up (Figure 1). Hips with consistent hip pain at both baseline and follow‐up were excluded to focus on cases showing meaningful change in pain over 4 to 5 years, thereby avoiding sample dilution.
Figure 1.
Selection of hips for the investigations of the associations between smoking and hip OA. CHECK, Cohort Hip and Cohort Knee; KL, Kellgren Lawrence, OAI, Osteoarthritis Initiative.
Radiographic measurements
Both studies used hip radiographs to assess IRFs and overall radiographic hip OA (RHOA). The grading and definition of IRFs were consistent across both studies. However, the grading and definition of overall RHOA differed between the OAI and the CHECK studies.
In the OAI study, osteophytes (located at four locations: superior acetabular, inferior acetabular, superior femoral, and inferior femoral) and joint space narrowing (JSN, either lateral or medial) were graded between 0 and 3. An IRF grade of greater than or equal to 2 was defined as ‘definite,’ whereas grade = 1 indicated ‘possible’ and grade = 0 meant ‘no’ findings. For overall RHOA in the OAI study, an overall radiographic grade was between 0 and 2. A hip was classified as having ‘definite RHOA’ if it met one of the following four conditions: (1) The modified Croft grade 30 was 2 or more, which implies the presence of 2 or more definite osteophytes, definite JSN, sclerosis, cysts, or deformity; (2) there was definite JSN and the presence of femoral osteophytes with a grade of 1 or more or acetabular osteophytes with a grade or 2 or more; (3) there were femoral osteophytes with a grade of 2 or more, regardless of other features; and (4) there was JSN lateral with a grade of 2 or more or JSN medial with a grade of 3 or more, regardless of other features. A definite RHOA was graded as overall grade 2 or more in OAI. A hip was considered to have ‘possible RHOA’ if it exhibited other individual IRFs or combinations of possible IRFs. A possible RHOA was graded as overall grade 1 in OAI. All other hips were classified as having ‘no RHOA’ and graded as overall grade 0.
In the CHECK study, osteophytes and JSNs were also graded between 0 and 3, as in the OAI study. However, the overall RHOA was assessed using the KL grade. 31 RHOA is graded between 0 and 4 in KL grade. Therefore, a KL grade of 2 or more was considered “definite RHOA,’ a KL grade = 1 was considered “possible RHOA,’ and KL grade = 0 was considered ‘no RHOA.’
Pain measurements
For the OAI cohort, we considered the presence of hip pain if a participant answered “yes” to experiencing pain, aching, or stiffness in the hip any time in the 12 months before the visit (ie, baseline or 4‐year follow‐up). The CHECK cohort determined the presence of hip pain during a clinical assessment at the time of the visit (ie, baseline or 5‐year follow‐up). We considered symptomatic hip OA as the presence of hip pain with the presence of definite RHOA.
Covariates
In this study, we adjusted the results for the following covariates: sex and baseline values of age, body mass index (BMI), and nonsteroidal anti‐inflammatory drug (NSAID) use. BMI was calculated based on weight and height measurements taken by clinic staff during baseline visits at the OAI and CHECK clinical centers. Methods for determining baseline NSAID use varied between the OAI and CHECK cohorts. In the OAI cohort, NSAID use was ascertained using the medication inventory method, 32 in which participants brought their current medications to the baseline visit for documentation by researchers. NSAID use at baseline was defined as any prescribed oral or capsule form of NSAIDs (including cyclooxygenase‐2 [COX‐2] inhibitors) used within the last 30 days. For the CHECK cohort, NSAID use was determined through self‐reporting during clinic interviews at baseline. Participants were considered NSAID users if they answered ‘yes’ to using any of the following for their hip and/or knee complaints at the time of assessment: ibuprofen, diclofenac, naproxen, celecoxib, or rofecoxib.
Exposure
In this study, the exposure variable was the smoking status of participants in both cohorts at baseline, classified as ‘current smoker,’ ‘former smoker,’ or ‘never smoker’ based on self‐report.
Outcomes
We investigated two types of outcomes, radiographic and pain outcomes, cross‐sectionally at baseline and longitudinally over 4 to 5 years and separately for the OAI and CHECK cohorts. We created four cohorts to investigate the outcomes of this study. These cohorts were OAI radiographic cohort, CHECK radiographic cohort, OAI pain cohort, and CHECK pain cohort.
Cross‐sectionally, we considered the following three outcomes: (1) the prevalence of definite RHOA, (2) the prevalence of hip pain at baseline, and (3) the prevalence of symptomatic hip OA at baseline. The prevalence of definite RHOA was investigated in the OAI radiographic cohort and the CHECK radiographic cohort. The prevalence of hip pain and the prevalence of symptomatic hip OA were investigated in the OAI pain cohort and the CHECK pain cohort.
Longitudinally, our radiographic outcomes were incidence of RHOA, progression of RHOA, and three outcomes for worsening of IRFs of the hip joint. An incidence of RHOA was defined as a definite RHOA at the 4‐ to 5‐year follow‐up while not at baseline. A progression of RHOA was defined as an increase of 1 or more overall grades (overall grades as defined in this article using modified Croft grade and IRFs for OAI and KL grade for CHECK) at the 4‐ to 5‐year follow‐up from baseline. Our three outcomes for worsening IRFs of the hip were worsening of JSN in the lateral side of the hip joint, worsening of JSN in the medial side of the hip joint, and worsening in the sum of osteophytes. The two outcomes, worsening of JSN in lateral and medial sides of the hip joint, were defined as an increase of 1 or more in IRF grade (lateral and medial, respectively) at 4‐ to 5‐year follow‐up from baseline. The third outcome of the three worsening of IRFs of the hip, the worsening in the sum of osteophytes, was defined as an increase of two or more in the sum of the IRFs grades for osteophytes in all four locations (superior acetabular, inferior acetabular, superior femoral, and inferior femoral) at 4‐ to 5‐year follow‐up from baseline. 33 All of the radiographic outcomes were investigated in the OAI radiographic cohort and the CHECK radiographic cohort.
We had four pain outcomes in the longitudinal analysis. These four pain outcomes were the development of hip pain, resolution of hip pain, development of symptomatic hip OA, and resolution of symptomatic hip pain over 4 to 5 years. Development of hip pain was defined as the presence of hip pain at the 4‐ to 5‐year follow‐up while absent at baseline. A resolution of hip pain was defined as the absence of hip pain at the 4‐ to 5‐year follow‐up while present at baseline. The development of symptomatic hip OA was defined as the presence of symptomatic hip OA at the 4‐ to 5‐year follow‐up while absent at baseline. A resolution of symptomatic hip OA was defined as the absence of symptomatic hip OA at the 4‐ to 5‐year follow‐up while present at baseline. These four pain outcomes were investigated in the OAI pain cohort and the CHECK pain cohort.
Statistical analyses
We used logistic regression with clustering of both hips within participants (ie, generalized estimating equations with a logistic link function 34 ) to investigate the association between smoking status at baseline and outcomes for cross‐sectionally at baseline and longitudinally between baseline and 4‐ to 5‐year follow‐up. Univariate (unadjusted) and multivariable (adjusted) analyses were performed. All multivariable analyses were adjusted for age at baseline, sex, and BMI at baseline. The outcomes related to pain (for both cross‐sectional and longitudinal analyses) were additionally adjusted for NSAID use at baseline.
We have performed two types of sensitivity analyses. In the first type, we performed a sensitivity analysis by rerunning the primary analysis without adjusting for BMI (while adjusting for other variables) and compared the results with our primary analysis. This sensitivity analysis approach addressed the complex relationship between smoking, BMI, and OA. Smoking can potentially influence OA risk through two distinct pathways: a direct pathway, in which smoking exerts an effect on OA independent of BMI and an indirect pathway. In the indirect pathway, smoking often leads to reduced BMI, 35 possibly because of metabolic changes or appetite suppression, which in turn can influence OA risk. 4 If we were to adjust for BMI, it might serve as a mediator in this indirect pathway, masking or diminishing the direct impact of smoking on OA. As a result, adjusting for BMI could introduce collider bias or overadjustment bias, thereby leading to inaccurate or misleading results in assessing the direct impact of smoking on OA. 18 In the second type, we performed a sensitivity analysis by rerunning the primary analysis by combining the former smoker and current smoking groups and compared the results with our primary analysis. We performed this sensitivity analysis because the relatively small proportion of current smokers, particularly in the OAI smoking study cohort (3.10%), might have limited the ability to detect any present associations in our primary analysis.
All statistical analyses were performed using STATA/BE 18.0 for Windows (64‐bit x86‐64) software. The statistical significance threshold was set as a two‐tailed P value of less than 0.05 for all statistical analyses.
RESULTS
Baseline characteristics of the OAI and the CHECK smoking study cohorts
The study included 9,386 hips from 4,716 participants in the OAI cohort and 1,954 hips from 977 participants in the CHECK cohort (Table 1). Participants in the OAI cohort were older, with a mean average age of 61.14 ± 9.16 years, compared with a younger mean average age of 55.90 ± 5.21 years in the CHECK cohort (Table 1). Both cohorts had a mean average BMI in the overweight category (OAI: 28.61 ± 4.83, CHECK: 26.18 ± 4.00 kg/m2) (Table 1). Most participants in both cohorts were White, accounting for 79.32% in the OAI and 97.54% in the CHECK cohort (Table 1). Although both cohorts were predominantly female (OAI: 58.53%, CHECK: 79.12%), the OAI cohort had a notably higher proportion of male participants among current (94.52%) and former smokers (90.02%), resulting in a less balanced sex distribution compared with the CHECK cohort (Table 1). In terms of RHOA and symptomatic indicators, most participants in both cohorts showed no radiographic severity of hip OA, classified as overall RHOA grade 0 (OAI: 75.15%, CHECK: 61.26%), reported no hip pain (OAI: 58.68%, CHECK: 60.24%), and showed no symptomatic hip OA (OAI: 97.24%, CHECK: 94.68%) at baseline (Table 1).
Table 1.
Baseline characteristics of hips in the OAI smoking study cohort and CHECK smoking study cohort, stratified by smoking status of participants*
| Baseline characteristics | OAI smoking study cohort | CHECK smoking study cohort | ||||||
|---|---|---|---|---|---|---|---|---|
| Current smoker | Former smoker | Never smoker | Total | Current smoker | Former smoker | Never smoker | Total | |
| Hips | n = 292 (3.11) | n = 1,583 (16.87) | n = 7,511 (80.02) | N = 9,386 (100.00) | n = 282 (14.43) | n = 1,052 (53.84) | n= 620 (31.73) | N = 1,954 (100.00) |
| Participants | n = 146 (3.10) | n = 796 (16.88) | n = 3,774 (80.03) | N = 4,716 (100.00) | n = 141 (14.43) | n = 526 (53.84) | n = 310 (31.73) | N = 977 (100.00) |
| Age, years | 56.82 ± 8.25 | 62.54 ± 9.43 | 61.01 ± 9.07 | 61.14 ± 9.16 | 54.72 ± 5.42 | 55.97 ± 5.20 | 56.30 ± 5.07 | 55.90 ± 5.21 |
| BMI, kg/m2 | 28.68 ± 3.92 | 29.20 ± 4.55 | 28.48 ± 4.91 | 28.61 ± 4.83 | 25.95 ± 4.30 | 26.15 ± 3.87 | 26.32 ± 4.07 | 26.18 ± 4.00 |
| Race | ||||||||
| White | 242 (82.88) | 1,386 (87.56) | 5,817 (77.45) | 7,445 (79.32) | 278 (98.58) | 1,036 (98.48) | 592 (95.48) | 1,906 (97.54) |
| Other | 50 (17.12) | 197 (12.44) | 1,694 (22.55) | 1,941 (20.68) | 4 (1.42) | 16 (1.52) | 28 (4.52) | 48 (2.46) |
| Sex | ||||||||
| Male | 276 (94.52) | 1,425 (90.02) | 2,191 (29.17) | 3,892 (41.47) | 80 (28.37) | 236 (22.43) | 92 (14.84) | 408 (20.88) |
| Female | 16 (5.48) | 158 (9.98) | 5,320 (70.83) | 5,494 (58.53) | 202 (71.63) | 816 (77.57) | 528 (85.16) | 1,546 (79.12) |
| NSAID use at baseline | ||||||||
| No | 248 (84.93) | 1,313 (82.94) | 6,161 (82.03) | 7,722 (82.27) | 224 (79.43) | 874 (83.08) | 500 (80.65) | 1,598 (81.78) |
| Yes | 44 (15.07) | 262 (16.55) | 1,333 (17.75) | 1,639 (17.46) | 58 (20.57) | 178 (16.92) | 120 (19.35) | 356 (18.22) |
| Overall grade of radiographic hip OA assessed by radiography a | ||||||||
| 0 | 216 (73.97) | 1,140 (72.02) | 5,698 (75.86) | 7,054 (75.15) | 169 (59.93) | 635 (60.36) | 393 (63.39) | 1,197 (61.26) |
| 1 | 31 (10.62) | 191 (12.07) | 820 (10.92) | 1,042 (11.10) | 64 (22.70) | 252 (23.95) | 131 (21.13) | 447 (22.88) |
| 2 | 21 (7.19) | 123 (7.77) | 389 (5.18) | 533 (5.68) | 35 (12.41) | 99 (9.41) | 65 (10.48) | 199 (10.18) |
| 3 | 0 (0.00) | 0 (0.00) | 0 (0.00) | 0 (0.00) | 2 (0.71) | 5 (0.48) | 6 (0.97) | 13 (0.67) |
| 4 | 0 (0.00) | 0 (0.00) | 0 (0.00) | 0 (0.00) | 0 (0.00) | 0 (0.00) | 0 (0.00) | 0 (0.00) |
| IRF grade of joint space narrowing in the lateral side | ||||||||
| 0 | 248 (84.93) | 1,316 (83.13) | 6,404 (85.26) | 7,968 (84.89) | 215 (76.24) | 789 (75.00) | 472 (76.13) | 1,476 (75.54) |
| 1 | 9 (3.08) | 90 (5.69) | 337 (4.49) | 436 (4.65) | 47 (16.67) | 189 (17.97) | 98 (15.81) | 334 (17.09) |
| 2 | 6 (2.05) | 41 (2.59) | 141 (1.88) | 188 (2.00) | 7 (2.48) | 18 (1.71) | 20 (3.23) | 45 (2.30) |
| 3 | 5 (1.71) | 7 (0.44) | 25 (0.33) | 37 (0.39) | 0 (0.00) | 1 (0.10) | 3 (0.48) | 4 (0.20) |
| IRF grade of joint space narrowing in the medial side | ||||||||
| 0 | 251 (85.96) | 1,296 (81.87) | 6,022 (80.18) | 7,569 (80.64) | 191 (67.73) | 708 (67.30) | 426 (68.71) | 1,325 (67.81) |
| 1 | 8 (2.74) | 93 (5.87) | 508 (6.76) | 609 (6.49) | 68 (24.11) | 264 (25.10) | 150 (24.19) | 482 (24.67) |
| 2 | 9 (3.08) | 62 (3.92) | 355 (4.73) | 426 (4.54) | 10 (3.55) | 24 (2.28) | 18 (2.90) | 52 (2.65) |
| 3 | 0 (0.00) | 3 (0.19) | 22 (0.29) | 25 (0.27) | 0 (0.00) | 1 (0.10) | 0 (0.0) | 1 (0.05) |
| Sum of osteophyte grades | 0.61 ± 1.35 | 0.60 ± 1.17 | 0.43 ± 1.09 | 0.46 ± 1.11 | 1.01 ±1.46 | 0.89 ± 1.38 | 0.87 ± 1.48 | 0.90 ± 1.42 |
| Hip pain | ||||||||
| No | 180 (61.64) | 986 (62.29) | 4,342 (57.81) | 5,508 (58.68) | 174 (61.70) | 613 (58.27) | 390 (62.90) | 1,177 (60.24) |
| Yes | 112 (38.36) | 597 (37.71) | 3,144 (41.86) | 3,853 (41.05) | 108 (38.30) | 439 (41.73) | 230 (37.10) | 777 (39.76) |
| Symptomatic hip OA | ||||||||
| No | 279 (95.55) | 1,530 (96.65) | 7,318 (97.43) | 9,127 (97.24) | 267 (94.68) | 997 (94.77) | 586 (94.52) | 1,850 (94.68) |
| Yes | 13 (4.45) | 53 (3.35) | 192 (2.56) | 258 (2.75) | 15 (5.32) | 55 (5.23) | 34 (5.48) | 104 (5.32) |
Source: Data are presented as mean ± SD or count (percentage). The percentage calculations are based on the total number of observations, including the missing values.
BMI, body mass index; CHECK, Cohort for Knee and Cohort for Hip; IRF, Individual Radiographic Features; NSAID, nonsteroidal anti‐inflammatory drug; OA, osteoarthritis; OAI, Osteoarthritis Initiative.
Assessed by KL grade in the OAI study and modified Croft grade in the CHECK study.
Associations of smoking with the prevalence of RHOA and hip pain outcomes at baseline (results from cross‐sectional analyses)
In the univariate analysis, current smokers, irrespective of the cohort, presented no statistically significant odds ratios (ORs), indicating no association with the prevalence of RHOA compared with those who never smoked (Table 2). In contrast, former smokers in the OAI cohort demonstrated a significant association with the prevalence of RHOA with an OR of 1.58 (95% confidence interval [95% CI] 1.17–2.13). The same association, however, was not observed in the CHECK cohort, in which the OR for former smokers was 0.94 (95% CI 0.61–1.47) (Table 2).
Table 2.
Associations of smoking with the prevalence of radiographic hip OA, hip pain, and symptomatic hip OA at baseline: results from univariate and multivariable cross‐sectional analyses*
| Outcome | Cohort | |||||
|---|---|---|---|---|---|---|
| OAI radiographic cohort | CHECK radiographic cohort | |||||
| Current smoker | Former smoker | Never smoker | Current smoker | Former smoker | Never smoker | |
| Hips (Participants) | N = 214 (n = 108) | N = 1,216 (n = 617) | N = 5,226 (n = 2,640) | N = 236 (n = 120) | N = 866 (n = 445) | N = 514 (n = 265) |
| Prevalence of radiographic hip OA | ||||||
| Events, n (%) | 16(7.48) | 83(6.83) | 235(4.50) | 29(12.29) | 70(8.08) | 43(8.37) |
| Univariate analysis | ||||||
| Odds ratio | 1.68 | 1.58 | 1.00 | 1.50 | 0.94 | 1.00 |
| 95% CI | 0.91–3.11 | 1.17–2.13 | reference | 0.86–2.63 | 0.61–1.47 | reference |
| Multivariate analysis a | ||||||
| Odds ratio | 1.29 | 0.99 | 1.00 | 1.38 | 0.85 | 1.00 |
| (95% CI) | 0.68–2.46 | 0.70–1.40 | reference | 0.78–2.44 | 0.54–1.32 | reference |
| OAI pain cohort | CHECK pain cohort | |||||
|---|---|---|---|---|---|---|
| Current smoker | Former smoker | Never smoker | Current smoker | Former smoker | Never smoker | |
| Hips (Participants) | N = 196 (n = 112) | N = 1,053 (n = 600) | N = 4,541 (n = 2,632) | N = 196 (n = 114) | N = 760 (n = 438) | N = 444 (n = 262) |
| Prevalence of hip pain | ||||||
| Events, n (%) | 39 (19.90) | 213 (20.23) | 1,055 (23.23) | 46 (23.47) | 201 (26.45) | 99 (22.30) |
| Univariate analysis | ||||||
| Odds ratio | 0.85 | 0.83 | 1.00 | 1.07 | 1.26 | 1.00 |
| 95% CI | 0.57–1.27 | 0.69–1.01 | reference | 0.68–1.68 | 0.93–1.72 | reference |
| Multivariate analysis b | ||||||
| Odds ratio | 1.22 | 1.11 | 1.00 | 1.10 | 1.29 | 1.00 |
| 95% CI | 0.80–1.86 | 0.89–1.38 | reference | 0.70–1.74 | 0.95–1.77 | reference |
| Prevalence of symptomatic hip OA | ||||||
| Events, n (%) | 3 (1.53) | 13 (1.23) | 56 (1.23) | 4 (2.04) | 14 (1.84) | 9 (2.03) |
| Univariate analysis | ||||||
| Odds ratio | 1.19 | 1.04 | 1.00 | 1.00 | 0.89 | 1.00 |
| 95% CI | 0.31–4.63 | 0.53–2.05 | reference | 0.30–3.39 | 0.38–2.12 | reference |
| Multivariate analysis b | ||||||
| Odds ratio | 1.20 | 0.71 | 1.00 | 0.90 | 0.83 | 1.00 |
| 95% CI | 0.30–4.81 | 0.33–1.52 | reference | 0.26–3.11 | 0.34–1.99 | reference |
Bold numbers indicate statististically significance (ie, P value less than 0.05).
CHECK, Cohort of Hip and Cohort of Knee; CI, confidence interval; OA, osteoarthritis; OAI, Osteoarthritis Initiative.
Adjusted for age at baseline, sex, and body mass index at baseline.
Adjusted for age at baseline, sex, body mass index at baseline, and nonsteroidal anti‐inflammatory drug use at baseline.
In the multivariate analysis, neither current nor former smokers showed a statistically significant association with the prevalence of RHOA, regardless of whether they belonged to the OAI or CHECK cohort (Table 2).
Regarding the prevalence of hip pain and prevalence of symptomatic hip OA, none of the analyses (univariate or multivariate) uncovered any statistically significant association for current or former smokers in either cohort (Table 2).
Associations of smoking with radiographic changes in hip OA over 4 to 5 years (results from longitudinal analyses)
For radiographic changes, the incidence of RHOA revealed no statistically significant ORs for current and former smokers in the univariate analysis (Table 3). Similarly, the multivariate analysis found no statistically significant association, with ORs for current and former smokers in the OAI cohort being 1.03 (95% CI 0.23–4.50) and 0.92 (95% CI 0.46–1.85) and in the CHECK cohort 0.61 (95% CI 0.34–1.11) and 1.00 (95% CI 0.69–1.44) respectively.
Table 3.
Associations of smoking with radiographic changes in hip OA over 4‐5 years: results from univariate and multivariable longitudinal analyses*
| Outcome | Cohort | |||||
|---|---|---|---|---|---|---|
| OAI radiographic cohort | CHECK radiographic cohort | |||||
| Current smoker | Never smoker | Current smoker | Former smoker | Never smoker | ||
| Hips (Participants) | N = 214 (n = 108) | Former smoker N = 1,216 (n = 617) | N = 5,226 (n = 2,640) | N = 236 (n = 120) | N = 866 (n = 445) | N = 514 (n = 265) |
| Incidence of radiographic hip OA | ||||||
| Events, n (%) | 2 (0.93) | 14 (1.15) | 77 (1.47) | 20 (8.47) | 112 (12.93) | 65 (12.65) |
| Univariate analysis | ||||||
| Odds ratio | 0.64 | 0.78 | 1.00 | 0.64 | 1.03 | 1.00 |
| 95% CI | 0.15–2.69 | 0.43–1.40 | reference | 0.36–1.14 | 0.72–1.48 | reference |
| Multivariate analysis a | ||||||
| Odds ratio | 1.03 | 0.92 | 1.00 | 0.61 | 1.00 | 1.00 |
| 95% CI | 0.23–4.50 | 0.46–1.85 | reference | 0.34–1.11 | 0.69–1.44 | reference |
| Progression of radiographic hip OA | ||||||
| Events, n (%) | 5 (2.34) | 47 (3.87) | 185 (3.54) | 66 (27.97) | 266 (30.72) | 158 (30.74) |
| Univariate analysis | ||||||
| Odds ratio | 0.66 | 1.09 | 1.00 | 0.88 | 1.00 | 1.00 |
| 95% CI | 0.26–1.68 | 0.78–1.54 | reference | 0.60–1.28 | 0.77–1.30 | reference |
| Multivariate analysis a | ||||||
| Odds ratio | 0.73 | 1.04 | 1.00 | 0.87 | 1.00 | 1.00 |
| 95% CI | 0.28–1.90 | 0.70–1.55 | reference | 0.59–1.27 | 0.77–1.30 | reference |
| Worsening of individual radiographic features of the hip | ||||||
| Worsening of joint space narrowing lateral | ||||||
| Events, n (%) | 7 (3.27) | 37 (3.04) | 95 (1.82) | 19 (8.05) | 79 (9.12) | 47 (9.14) |
| Univariate analysis | ||||||
| Odds ratio | 1.84 | 1.70 | 1.00 | 0.87 | 1.01 | 1.00 |
| 95% CI | 0.81–4.18 | 1.13–2.55 | reference | 0.48–1.58 | 0.67–1.51 | reference |
| Multivariate analysis a | ||||||
| Odds ratio | 1.58 | 1.31 | 1.00 | 0.87 | 1.01 | 1.00 |
| 95% CI | 0.67–3.71 | 0.81–2.11 | reference | 0.48–1.60 | 0.67–1.51 | reference |
| Worsening of joint space narrowing medial | ||||||
| Events, n (%) | 5 (2.34) | 47 (3.87) | 184 (3.52) | 43 (18.22) | 148 (17.09) | 108 (21.01) |
| Univariate analysis | ||||||
| Odds ratio | 0.65 | 1.10 | 1.00 | 0.83 | 0.77 | 1.00 |
| 95% CI | 0.24–1.74 | 0.77–1.58 | reference | 0.53–1.30 | 0.56–1.06 | reference |
| Multivariate analysis a | ||||||
| Odds ratio | 1.03 | 1.38 | 1.00 | 0.85 | 0.77 | 1.00 |
| 95% CI | 0.37–2.86 | 0.89–2.15 | reference | 0.54–1.35 | 0.56–1.06 | reference |
| Worsening in the sum of the osteophytes | ||||||
| Events, n (%) | 2 (0.93) | 9 (0.74) | 62 (1.19) | 38 (16.10) | 164 (18.94) | 95 (18.48) |
| Univariate analysis | ||||||
| Odds ratio | 0.84 | 0.64 | 1.00 | 0.88 | 1.09 | 1.00 |
| 95% CI | 0.19–3.62 | 0.30–1.34 | reference | 0.56–1.39 | 0.80–1.49 | reference |
| Multivariate analysis a | ||||||
| Odds ratio | 1.16 | 0.70 | 1.00 | 0.89 | 1.08 | 1.00 |
| 95% CI | 0.25–5.32 | 0.30–1.64 | reference | 0.55–1.42 | 0.79–1.49 | reference |
Bold numbers indicate statististically significance (ie, P value less than 0.05).
CHECK, Cohort of Hip and Cohort of Knee; CI, Confidence interval; OA, Osteoarthritis; OAI, Osteoarthritis Initiative.
Adjusted for age at baseline, sex, and body mass index at baseline.
Similar patterns were observed in the progression of RHOA and worsening of IRFs of the hip, with no statistically significant ORs found for current and former smokers across both cohorts in the univariate and multivariate analyses, with the exception of JSN in lateral compartment in OAI cohort for former smokers in the univariate analysis. Former smokers in the OAI cohort demonstrated a statistically significant association with the worsening of JSN in the lateral compartment, with an OR of 1.70 (95% CI 1.13–2.55) in the univariate analysis. The same association, however, was not observed in the CHECK cohort, in which the OR for former smokers was 1.01 (95% CI 0.67–1.51) (Table 3).
Associations of smoking with pain outcomes in hip OA over 4 to 5 years (results from longitudinal analyses)
For pain, the development of hip pain and resolution of hip pain over 4 to 5 years did not yield statistically significant associations in the univariate and multivariate analyses for current or former smokers in both the OAI and CHECK cohorts (Table 4). Similarly, the development and resolution of symptomatic hip OA did not exhibit statistically significant ORs in the univariate and multivariate analyses for current and former smokers across both cohorts (Table 4). Two and three current smokers developed symptomatic hip OA in the OAI and CHECK cohorts over 4 to 5 years, respectively. Similarly, three and four current smokers had a resolution of symptomatic hip OA in the OAI and CHECK cohorts, respectively (Table 4).
Table 4.
Associations of smoking with pain outcomes in hip OA over 4‐5 years: results from univariate and multivariable longitudinal analyses*
| Outcome | Cohort | |||||
|---|---|---|---|---|---|---|
| OAI pain cohort | CHECK pain cohort | |||||
| Current smoker | Former smoker | Never smoker | Current smoker | Former smoker | Never smoker | |
| Hips (Participants) | N = 196 (n = 112) | N = 1,053 (n = 600) | N = 4,541 (n = 2,632) | N = 196 (n = 114) | N = 760 (n = 438) | N = 444 (n = 262) |
| Development of hip pain | ||||||
| Events, n (%) | 30 (15.31) | 196 (18.61) | 963 (21.21) | 38 (19.39) | 137 (18.03) | 76 (17.12) |
| Univariate analysis | ||||||
| Odds ratio | 0.69 | 0.86 | 1.00 | 1.13 | 1.05 | 1.00 |
| 95% CI | 0.45–1.07 | 0.71–1.04 | reference | 0.69–1.84 | 0.74–1.48 | reference |
| Multivariate analysis a | ||||||
| Odds ratio | 0.74 | 0.94 | 1.00 | 1.19 | 1.10 | 1.00 |
| 95% CI | 0.47–1.15 | 0.76–1.17 | reference | 0.72–1.95 | 0.78–1.57 | reference |
| Resolution of hip pain | ||||||
| Events, n (%) | 39 (19.90) | 213 (20.23) | 1,055 (23.23) | 46 (23.47) | 201 (26.45) | 99 (22.30) |
| Univariate analysis | ||||||
| Odds ratio | 0.85 | 0.83 | 1.00 | 1.07 | 1.26 | 1.00 |
| 95% CI | 0.57–1.27 | 0.69–1.01 | reference | 0.68–1.68 | 0.93–1.72 | reference |
| Multivariate analysis a | ||||||
| Odds ratio | 1.22 | 1.11 | 1.00 | 1.10 | 1.29 | 1.00 |
| 95% CI | 0.80–1.86 | 0.89–1.38 | reference | 0.70–1.74 | 0.95–1.77 | reference |
| Development of symptomatic hip OA | ||||||
| Events, n (%) | 2 (1.02) | 22 (2.09) | 62 (1.37) | 3 (1.53) | 33 (4.34) | 13 (2.93) |
| Univariate analysis | ||||||
| Odds ratio | 0.77 | 1.56 | 1.00 | 0.52 | 1.51 | 1.00 |
| 95% CI | 0.18–3.29 | 0.94–2.61 | reference | 0.14–1.84 | 0.78–2.90 | reference |
| Multivariate analysis a | ||||||
| Odds ratio | 0.86 | 1.50 | 1.00 | 0.53 | 1.51 | 1.00 |
| 95% CI | 0.19–3.84 | 0.81–2.78 | reference | 0.15–1.92 | 0.78–2.93 | reference |
| Resolution of symptomatic hip OA | ||||||
| Events, n (%) | 3 (1.53) | 13 (1.23) | 56 (1.23) | 4 (2.04) | 14 (1.84) | 9 (2.03) |
| Univariate analysis | ||||||
| Odds ratio | 1.19 | 1.04 | 1.00 | 1.00 | 0.89 | 1.00 |
| 95% CI | 0.31–4.63 | 0.53–2.05 | reference | 0.30–3.39 | 0.38–2.12 | reference |
| Multivariate analysis a | ||||||
| Odds ratio | 1.20 | 0.71 | 1.00 | 0.90 | 0.83 | 1.00 |
| 95% CI | 0.30–4.81 | 0.33–1.52 | reference | 0.26–3.11 | 0.34–1.99 | reference |
CHECK, Cohort of Hip and Cohort of Knee; CI, confidence interval; OA, osteoarthritis; OAI, Osteoarthritis Initiative.
Adjusted for age at baseline, sex, body mass index at baseline, and nonsteroidal anti‐inflammatory drug use at baseline.
Sensitivity analysis
The results obtained from the first sensitivity analysis suggest that, excluding adjusting for BMI, there was no evidence of a statistically significant association between smoking and any of the outcomes investigated in multivariable analyses in both OAI and CHECK cohorts (Supplementary Tables 1 and 2). Similarly, the results obtained from the second sensitivity analysis by combining the current and former smokers did not show any evidence of a statistically significant association between smoking and any outcomes investigated in multivariable analyses in both OAI and CHECK cohorts (Supplementary Tables 3 and 4).
DISCUSSION
Our results indicate no clear association between smoking and the prevalence of radiographic hip OA at baseline, nor with the incidence and progression of RHOA and worsening of IRF of the hip joint over 4 to 5 years. Additionally, no clear association was found between smoking and the prevalence of hip pain or symptomatic hip OA at baseline, development of hip pain, resolution of hip pain, development of symptomatic hip OA, and resolution of symptomatic hip OA over 4 to 5 years.
As mentioned in the Introduction, previous studies investigating the impact of smoking on hip OA have produced mixed findings. These studies are also marked by gaps and methodologic limitations, which makes it challenging to consolidate the results and definitively determine the effect of smoking on hip OA. Although systematic reviews by Hui et al 4 and Pearce et al 19 found no statistically significant associations between smoking and the incidence or progression of hip OA, they highlighted the methodologic issues in the underlying studies.
Our study addressed the methodologic issues that were present in these studies. First, unlike many previous studies, we have only focused solely on the outcomes for hip OA without combining other types of OA in the cohorts, as discussed in our Introduction section of this paper. Second, previous studies had limitations due to combining former and current smokers, which may not capture the impact of smoking exposure because of the long lifespan of chondrocytes, the cells found in cartilage tissue. Chondrocytes are implicated in conditions like OA, and their long lifespan means that they could be affected by past smoking behaviors, not only current smoking. Our study has separated current and former smokers, providing more subtle variations in exposure to smoking, which might have implications for understanding its effects on OA. Additionally, the presence of current and former smokers in our cohorts provides a true representation of the general population.
Third, the multicohort setting enhances the generalizability of our findings, addressing the limitation of many single‐cohort studies. Fourth, the large sample size in our study provided robust statistical power, which is particularly critical for studying a multifactorial condition like hip OA. Fifth, our analysis includes a comprehensive range of outcomes, of structural changes and pain related, providing a more holistic view of the disease. Previous research used radiographic outcomes, such as KL grade and THR, to assess the structural changes in hip OA, which both have some issues associated with evaluating disease development and progression. KL grade puts a higher emphasis on the formation of osteophytes, 36 and THR is not an indicator of hip OA progression and is influenced by an array of extraneous factors (education, income, etc). 19 Our study included various outcomes to indicate the development and progression of hip OA, such as JSN, osteophytes, and modified Croft grade, to avoid problems with the definition of the development and progression of hip OA in previous studies. Furthermore, we have also addressed the gap on the effect of smoking in pain outcomes, which was not previously addressed specifically.
As a sixth, we maintained a relatively long follow‐up period for the longitudinal analysis, ranging between 4 to 5 years. Although this duration may be considered short for observing the long‐term effects of smoking on hip OA, previous research 37 has established that a minimum of 3 years is necessary to detect meaningful changes in functional impairment and pain related to hip OA. Moreover, other studies have demonstrated significant changes in structural outcomes of hip OA, as assessed by radiography, over a similar 5‐year period. 38 Therefore, although a more extended follow‐up could offer additional insights, we argue that our chosen study period is sufficient to meet the objectives of the current study. Lastly, our analysis considered cross‐sectional and longitudinal data, allowing for an in‐depth exploration of the relationship between smoking and hip OA.
Although our study has several strengths, including a multicohort setting, a large sample size, a relatively long follow‐up time, the use of well‐characterized cohorts, the consideration of comprehensive outcomes, and the cross‐sectional and longitudinal study design, several limitations should be acknowledged. First, the findings should be viewed as associative rather than causative because the OAI and the CHECK studies were observational. Second, self‐reported smoking status might have influenced our results because of potential misclassification or reporting bias. Third, the OAI and the CHECK cohorts were predominantly White and female adults who were 45 years or older, limiting the transferability of the findings to other populations. Fourth, the low incidence of development and resolution of symptomatic hip OA among current smokers in our cohorts should be noted, as they might have reduced the statistical power to detect significant associations for symptomatic hip OA for current smokers. As a fifth limitation, our study may be vulnerable to collider bias when examining the progression of RHOA. This vulnerability is particularly notable when associating a prevalent exposure, such as smoking, with the progression of the disease in patients who already had RHOA at baseline. Therefore, results about the progression of RHOA should be interpreted with caution. Lastly, our results did not account for the duration and intensity of smoking, which could shed further light on the relationship between smoking and hip OA.
It is worth noting that the potential relationship between smoking and hip OA is complex and involves various factors. Smoking is known to be related to systemic and local inflammation 39 , 40 , reduced BMI 35 , or decreased physical activity 41 , all of which may contribute to the development or the progression of hip OA. Although our study did not find a direct link between smoking and hip OA, it is important to recognize that these indirect effects of smoking could play a role. Further research is needed to understand how these factors interact and influence the relationship between smoking and hip OA.
In summary, this comprehensive analysis of two large cohorts found no clear association between smoking and the prevalence of RHOA or pain outcomes in hip OA at baseline. Similarly, no clear association was found between smoking and the incidence and progression of RHOA, worsening of IRF of the hip, or the development and resolution of pain outcomes over 4 to 5 years. However, further research is needed to clarify whether the indirect effects of smoking might influence hip OA. Meanwhile, given the established health risks of smoking, smoking cessation should continue to be promoted as an important public health initiative.
AUTHOR CONTRIBUTIONS
Mr. Salis drafted the article, revised it critically for important intellectual content, and approved the final version to be published.
Supporting information
Disclosure form
Table S1: Associations of smoking with radiographic measurements and pain in hip OA at baseline, as shown in multivariable analyses (results from cross‐sectional analysis)
Table S2: Associations of smoking with radiographic changes and pain in hip OA over 4‐5 years, as shown in multivariable analyses (results from longitudinal analysis)
Table S3: Associations of smoking with radiographic measurements, pain in hip OA, and symptomatic hip OA at baseline, as shown in multivariable analyses (results from cross‐sectional analysis)
Table S4: Associations of smoking with radiographic and symptomatic changes in hip OA over 4‐5 years, as shown in multivariable analyses (results from longitudinal analysis)
ACKNOWLEDGMENT
Open access publishing facilitated by University of New South Wales, as part of the Wiley ‐ University of New South Wales agreement via the Council of Australian University Librarians.
This study acknowledges the provision of data sets and/or research tools from two cohort studies: the Osteoarthritis Initiative (OAI) study and the Cohort Hip and Cohort Knee (CHECK) study. The OAI is a collaborative informatics system created by the National Institute of Mental Health and the National Institute of Arthritis, Musculoskeletal, and Skin Diseases (NIAMS) to provide a worldwide resource to quicken the pace of biomarker identification, scientific investigation, and osteoarthritis drug development. The OAI is a public–private partnership comprising five contracts (N01‐AR‐2‐2258, N01‐AR‐2‐2259, N01‐AR‐2‐2260, N01‐AR‐2‐2261, and N01‐AR‐2‐2262) funded by the NIH, a branch of the Department of Health and Human Services, and conducted by the OAI Study Investigators. Private funding partners include Merck Research Laboratories, Novartis Pharmaceuticals Corporation, GlaxoSmithKline, and Pfizer, Inc. Private sector funding for the OAI is managed by the Foundation for the NIH. This article was prepared using OAI public‐use data sets and does not necessarily reflect the opinions or views of the OAI investigators, the NIH, or the private funding partners. The OAI data repository is housed within the National Institute of Mental Health Data Archive. The Dutch Arthritis Foundation funds the CHECK study. Involved are Erasmus Medical Center Rotterdam, Kennemer Gasthuis Haarlem, Leiden University Medical Center, Maastricht University Medical Center, Martini Hospital Groningen/Allied Health Care Center for Rheumatology and Rehabilitation Groningen, Medical Spectrum Twente Enschede/Ziekenhuisgroep Twente Almelo, Reade Center for Rehabilitation and Rheumatology, Sint Maartens‐kliniek Nijmegen, University Medical Center Utrecht, and Wilhelmina Hospital Assen.
Zubeyir Salis, BEng: Geneva University Hospital and University of Geneva, Geneva, Switzerland, University of New South Wales, Kensington, New South Wales, Australia, and The University of Western Australia, Perth, Western Australia, Australia.
Additional supplementary information cited in this article can be found online in the Supporting Information section (http://onlinelibrary.wiley.com/doi/10.1002/acr2.11644).
Author disclosures are available at https://onlinelibrary.wiley.com/doi/10.1002/acr2.11644.
REFERENCES
- 1. Hunter DJ, Bierma‐Zeinstra S. Osteoarthritis. Lancet 2019;393:1745–1759. [DOI] [PubMed] [Google Scholar]
- 2. Lespasio MJ, Sultan AA, Piuzzi NS, et al. Hip osteoarthritis: a primer. Perm J 2018;22:17–084. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Bennell K. Physiotherapy management of hip osteoarthritis. J Physiother 2013;59:145–157. [DOI] [PubMed] [Google Scholar]
- 4. Hui M, Doherty M, Zhang W. Does smoking protect against osteoarthritis? Meta‐analysis of observational studies. Ann Rheum Dis 2011;70:1231–1237. [DOI] [PubMed] [Google Scholar]
- 5. Philbin EF, Ries MD, Groff GD, et al. Osteoarthritis as a determinant of an adverse coronary heart disease risk profile. J Cardiovasc Risk 1996;3:529–533. [PubMed] [Google Scholar]
- 6. Oliveria SA, Felson DT, Cirillo PA, et al. Body weight, body mass index, and incident symptomatic osteoarthritis of the hand, hip, and knee. Epidemiology 1999;10:161–166. [PubMed] [Google Scholar]
- 7. Lau EC, Cooper C, Lam D, et al. Factors associated with osteoarthritis of the hip and knee in Hong Kong Chinese: obesity, joint injury, and occupational activities. Am J Epidemiol 2000;152:855–862. [DOI] [PubMed] [Google Scholar]
- 8. Hootman JM, Macera CA, Helmick CG, et al. Influence of physical activity‐related joint stress on the risk of self‐reported hip/knee osteoarthritis: a new method to quantify physical activity. Prev Med 2003;36:636–644. [DOI] [PubMed] [Google Scholar]
- 9. Järvholm B, Lewold S, Malchau H, et al. Age, bodyweight, smoking habits and the risk of severe osteoarthritis in the hip and knee in men. Eur J Epidemiol 2005;20:537–542. [DOI] [PubMed] [Google Scholar]
- 10. Kraus VB, Stabler TV, Luta G, et al. Interpretation of serum C‐reactive protein (CRP) levels for cardiovascular disease risk is complicated by race, pulmonary disease, body mass index, gender, and osteoarthritis. Osteoarthritis Cartilage 2007;15:966–971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Grotle M, Hagen KB, Natvig B, et al. Prevalence and burden of osteoarthritis: results from a population survey in Norway. J Rheumatol 2008;35:677–684. [PubMed] [Google Scholar]
- 12. Engström G, de Verdier MG, Rollof J, et al. C‐reactive protein, metabolic syndrome and incidence of severe hip and knee osteoarthritis. A population‐based cohort study. Osteoarthritis Cartilage 2009;17:168–173. [DOI] [PubMed] [Google Scholar]
- 13. Singer SP, Dammerer D, Krismer M, et al. Maximum lifetime body mass index is the appropriate predictor of knee and hip osteoarthritis. Arch Orthop Trauma Surg 2018;138:99–103. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Salis Z, Sainsbury A. Association Between Change in Body Mass Index and Knee and Hip Replacements: A Survival Analysis of Seven to Ten Years Using Multicohort Data. Arthritis Care Res (Hoboken) 2023;75:1340–1350. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Reijman M, Pols HAP, Bergink AP, 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;66:158–162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Pamukoff DN, Dudley RI, Vakula MN, et al. An evaluation of the heel strike transient in obese young adults during walking gait. Gait Posture 2016;49:181–183. [DOI] [PubMed] [Google Scholar]
- 17. Shepherd DE, Seedhom BB. Thickness of human articular cartilage in joints of the lower limb. Ann Rheum Dis 1999;58:27–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Felson DT, Zhang Y. Smoking and osteoarthritis: a review of the evidence and its implications. Osteoarthritis Cartilage 2015;23:331–333. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Pearce F, Hui M, Ding C, et al. Does smoking reduce the progression of osteoarthritis? Meta‐analysis of observational studies. Arthritis Care Res (Hoboken) 2013;65:1026–1033. [DOI] [PubMed] [Google Scholar]
- 20. Roux CH, Coste J, Roger C, et al. Impact of smoking on femorotibial and hip osteoarthritis progression: 3‐year follow‐up data from the KHOALA cohort. Joint Bone Spine 2021;88:105077. [DOI] [PubMed] [Google Scholar]
- 21. Von Mühlen D, Morton D, Von Mühlen C, et al. Postmenopausal estrogen and increased risk of clinical osteoarthritis at the hip, hand, and knee in older women. J Womens Health Gend Based Med 2002;11:511–518. [DOI] [PubMed] [Google Scholar]
- 22. Grotle M, Hagen KB, Natvig B, et al. Obesity and osteoarthritis in knee, hip and/or hand: an epidemiological study in the general population with 10 years follow‐up. BMC Musculoskeletal Disorders 2008;9:132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Wright NC, Riggs GK, Lisse JR, et al. Self‐reported osteoarthritis, ethnicity, body mass index, and other associated risk factors in postmenopausal women‐results from the Women's Health Initiative. J Am Geriatr Soc 2008;56:1736–1743. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Sandmark H, Hogstedt, Lewold S, et al. Osteoarthrosis of the knee in men and women in association with overweight, smoking, and hormone therapy. Ann Rheum Dis 1999;58:151–155. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Rego I, Fernández‐Moreno M, Fernández‐López C, et al. Role of European mitochondrial DNA haplogroups in the prevalence of hip osteoarthritis in Galicia, Northern Spain. Ann Rheum Dis 2010;69:210–213. [DOI] [PubMed] [Google Scholar]
- 26. Cheng Y, Macera CA, Davis DR, et al. Physical activity and self‐reported, physician‐diagnosed osteoarthritis: is physical activity a risk factor? J Clin Epidemiol 2000;53:315–322. [DOI] [PubMed] [Google Scholar]
- 27. Juhakoski R, Heliövaara M, Impivaara O, et al. Risk factors for the development of hip osteoarthritis: a population‐based prospective study. Rheumatology (Oxford) 2009;48:83–87. [DOI] [PubMed] [Google Scholar]
- 28. Nevitt MC, Felson DT, Lester G. The Osteoarthritis Initiative protocol for the cohort study. Accessed December 14, 2023. https://nda.nih.gov/static/docs/StudyDesignProtocolAndAppendices.pdf
- 29. Wesseling J, Boers M, Viergever MA, et al., Cohort Profile: Cohort Hip and Cohort Knee (CHECK) study. Int J Epidemiol 2014;45:36–44. [DOI] [PubMed] [Google Scholar]
- 30. Croft P, Cooper C, Coggon D. Case definition of hip osteoarthritis in epidemiologic studies. J Rheumatol 1994;21:591–592. [PubMed] [Google Scholar]
- 31. Kellgren JH, Lawrence JS. Radiological assessment of osteo‐arthrosis. Ann Rheum Dis 1957;16:494–502. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Pahor M, Chrischilles EA, Guralnik JM, et al. Drug data coding and analysis in epidemiologic studies. Eur J Epidemiol 1994;10:405–411. [DOI] [PubMed] [Google Scholar]
- 33. Lane NE, Nevitt MC, Hochberg MC, et al. Progression of radiographic hip osteoarthritis over eight years in a community sample of elderly white women. Arthritis Rheum 2004;50:1477–1486. [DOI] [PubMed] [Google Scholar]
- 34. Ballinger GA. Using generalized estimating equations for longitudinal data analysis. Organ Res Methods 2004;7:127–150. [Google Scholar]
- 35. Piirtola M, Jelenkovic A, Latvala A, et al. Association of current and former smoking with body mass index: a study of smoking discordant twin pairs from 21 twin cohorts. PLoS One 2018;13:e0200140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Steenkamp W, Rachuene PA, Dey R, et al. The correlation between clinical and radiological severity of osteoarthritis of the knee. SICOT‐J 2022;8:14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Van Dijk GM, Dekker J, Veenhof C, et al. Course of functional status and pain in osteoarthritis of the hip or knee: a systematic review of the literature. Arthritis Rheum 2006;55:779–785. [DOI] [PubMed] [Google Scholar]
- 38. Van Berkel AC, Schiphof D, Waarsing JH, et al. 10‐year natural course of early hip osteoarthritis in middle‐aged persons with hip pain: a CHECK study. Ann Rheum Dis 2021;80:487–493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Bazzano LA, He J, Muntner P, et al. Relationship between cigarette smoking and novel risk factors for cardiovascular disease in the United States. Ann Intern Med 2003;138:891–897. [DOI] [PubMed] [Google Scholar]
- 40. Bermudez EA, Rifai N, Buring JE, et al. Relation between markers of systemic vascular inflammation and smoking in women. Am J Cardiol 2002;89:1117–1119. [DOI] [PubMed] [Google Scholar]
- 41. Conway TL, Cronan TA. Smoking, exercise, and physical fitness. Prev Med 1992;21:723–734. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Disclosure form
Table S1: Associations of smoking with radiographic measurements and pain in hip OA at baseline, as shown in multivariable analyses (results from cross‐sectional analysis)
Table S2: Associations of smoking with radiographic changes and pain in hip OA over 4‐5 years, as shown in multivariable analyses (results from longitudinal analysis)
Table S3: Associations of smoking with radiographic measurements, pain in hip OA, and symptomatic hip OA at baseline, as shown in multivariable analyses (results from cross‐sectional analysis)
Table S4: Associations of smoking with radiographic and symptomatic changes in hip OA over 4‐5 years, as shown in multivariable analyses (results from longitudinal analysis)