Squatting, sitting on the floor, or cycling: Are life-long daily activities risk factors for clinical knee osteoarthritis? Stage III results of a community-based study




To evaluate the association between occupations, sports, life-long daily activities, and knee osteoarthritis.


In this case–control study, we randomly recruited 480 subjects with knee osteoarthritis, who had participated in the first stage of a community-based study in Tehran, and compared them with 490 controls. A questionnaire was used to record all occupations, sports, and the details of 10 daily activities. The mean ± SD hours/day spent on these activities were calculated and compared using the lightest activity as reference. Generalized estimation equation was used with each knee as the unit of analysis.


The mean ± SD age and percentage of female distribution was 57 ± 12 years, 69.8% women for cases, and 46.8 ± 15 years, 64.1% women for controls. After adjusting for age, sex, and body mass index, 2 activities were risk factors for knee osteoarthritis: prolonged squatting (odds ratio [OR] 1.51, 95% confidence interval [95% CI] 1.12–2.04) and cycling (OR 2.06, 95% CI 1.23–3.45). Knee-bending had borderline significance (OR 1.98, 95% CI 0.98–3.99). Carrying loads (OR 1.24, 95% CI 0.87–1.76) or climbing stairs (OR 0.99, 95% CI 0.69–1.42) showed no extra risk for knee osteoarthritis. Prolonged standing, sitting on the floor, and walking up/downhill were not risks for knee osteoarthritis. Housewives were at greater risk (borderline-significant) of developing knee osteoarthritis (OR 1.68, 95% CI 0.93–3.03) than women whose main occupation was outside the home. Other types of jobs and sports did not show an extra risk of knee osteoarthritis.


Our findings support the role of lifestyle in the pathogenesis of knee osteoarthritis for squatting and cycling. Education on preventable risk factors should be considered in order to ensure people use knee joints appropriately and avoid overuse.


In Iran, as in various other countries, coordinated efforts by the World Health Organization and the International League Against Rheumatism have led to a Community-Oriented Program for Control of Rheumatic Diseases (COPCORD) (1). In Stage I of the Tehran COPCORD study, a total of 10,291 people from an urban community were screened for the presence of any rheumatologic symptoms. The study's findings showed that ∼25.5% had experienced knee symptoms during the previous 7 days, with 15.3% of these symptoms being identified as knee osteoarthritis (1).

Knee osteoarthritis is one of several frequent and functionally impairing disorders of the musculoskeletal system. As stated by Bijlsma et al, “Osteoarthritis is commonly associated with disability and can be improved with a wide variety of rehabilitative interventions: joint-specific exercises, physical fitness, and physical modalities. As no cure is advisable, preventable risk factors of knee osteoarthritis are of great interest” (2).

Various occupational risk factors related to the development and progress of knee-joint diseases, including kneeling and squatting, climbing stairs, and lifting and carrying heavy loads have been discussed in the literature (3). Since the measurement of daily activities has not always been complete, exposure has been surmised indirectly from the subject's job and leisure-time activities description.

The present study, which is the third stage of the COPCORD study, was therefore designed as a case–control study to evaluate the risk factors of knee osteoarthritis in participants recruited from the Stage 1 Tehran COPCORD study. To generate further data on daily activities that could complement data on other factors, we collected detailed lifetime histories of exposure to 10 specified daily activities performed in the workplace or during leisure-time. This study presents the role of occupation, sports, and daily activities with relationship to knee osteoarthritis.


Study population.

This study is an extension of the Stage III Tehran COPCORD, which followed Stage I to evaluate the risk factors of knee osteoarthritis. It was conducted by Tehran University's Rheumatology Research Center, and was approved by the Medical Ethics Committee of the Iranian Ministry of Health and Medical Education.

In Stage I of the Tehran COPCORD study, which took place in 2004 and 2005, a total of 10,291 subjects from 50 clusters in Tehran were screened for rheumatologic symptoms (1). In the current study, we randomly selected 480 of the 1,532 participants with knee OA who had been identified at that stage. For each of these cases, we sought a matching control who had been a participant without knee osteoarthritis from the same cluster in Stage I, and who would therefore be representative of the same socioeconomic class (n = 490). Controls were also matched as closely as possible to cases by age decade and sex.

In a telephone call, trained interviewers explained the study and arranged for a home visit. Patients and controls who agreed were visited at home by a team of 1 trained nurse and 1 physician. Using a structured questionnaire and a physical examination sheet, the team interviewed and examined the cases and controls.


Knee osteoarthritis. Participants were eligible for the case group if they fulfilled the American College of Rheumatology (ACR; formerly known as the American Rheumatism Association) criteria for knee OA (4), i.e., knee pain and at least 3 of the 6 following ACR criteria: (age ≥50 years, morning stiffness lasting <30 minutes, crepitus, bony tenderness, bony enlargement, or no palpable warmth).Subjects with any history of inflammatory rheumatic disorders were excluded from the case group.

Risk factors for knee osteoarthritis.

The core questionnaire and examination sheet developed by the Asia Pacific League of Associations for Rheumatology (APLAR) and COPCORD on identifying risk factors for knee OA was developed by COPCORD study groups in 2 Asian countries (Iran and Bangladesh). Although investigators in each country worked separately, they were in frequent communication and exchanged views and opinions on the successive steps in developing the questionnaire. These steps included item-generation, the development of a common preliminary core questionnaire in English, translation into the local language (followed by back translation), the development of a common English version, content analysis and item reduction, adaptation of the local-language version, validity and reliability testing, and finalization of the questionnaire.

Age, sex, and body mass index (BMI).

Age and sex were recorded at the time of the interview. Current weight and height were measured at the start of the complete physical examination by the trained physician. BMI was calculated as weight in kilograms divided by height in square meters.


Information about jobs held since leaving school was recorded. Jobs were classified into 3 categories: sedentary work (typically done by office workers, students, managers, accountants, and other professionals who work mainly in an office), laborious jobs (typically done by unskilled workers, technical workers, porters, farmers, constructive workers, mechanics, factory workers, cooks, and workers in supermarkets and fruit or grocery shops), and housewives, who were categorized in a separate job category that was reported only by women in the study population. The time that each participant had spent in their occupation was recorded, and the type of job that each participant had performed for the majority of his or her working life was recorded.

Sports activities.

All sports activities practiced for at least 6 months were recorded. The association between the kind of sport and knee osteoarthritis was evaluated in 4 categories: running/jogging, fitness/body-building, football/volleyball, and other kinds of sports. If participants practiced more than 1 type of sport, they were categorized into more than 1 group.

Life-long daily activities.

Trained interviewers used the APLAR-COPCORD core questionnaire to question participants on their life-long daily activities, as well as jobs and leisure-time activities held since leaving school. For each job/leisure time activity reported, the interviewers asked about the weekly frequency of 10 specific occupational activities: standing, walking on flat ground, walking up/downhill, sitting on the floor, sitting on a chair, squatting, knee bending, cycling, climbing stairs, and carrying loads.

Statistical analysis.

Univariate logistic regression analysis was used initially to examine associations between knee osteoarthritis and daily activities as risk factor. Associations were expressed as odds ratios (OR) with 95% confidence intervals (95% CI). We used logistic regression for repeated measurement (generalized estimation equation [GEE]) to account for the contribution of each knee as a unit of analysis.

By assumption of the wear-and-tear theory, we summed the amount of each activity for its lifetime, calculated the mean time spent on each activity per day, and placed each into 2–4 categories, in which the lightest activity was the reference group of comparison. For the study population, the distribution of time spent on each daily activity was plotted, and the cutoff point was chosen on the basis of 25%, 50%, and 75%. Since certain daily activities were reported by only a small percentage of the study population, we selected a 90% cutoff point for comparison. The Stata statistical software, version 10 (Stata, College Station, TX) was used for all analysis.


A total of 607 cases were contacted over the course of the study, and 480 (80%) were eventually interviewed and examined. Among the 1,049 controls we attempted to recruit, the response rate was 490 (47%) with the nonparticipation occurring either because the case or control subject could not be contacted (changed residence or telephone number), or because they declined to participate (3% for cases, 2% for controls). The age range of the 480 cases was 22–88 years (mean ± SD 57 ± 12 years), and the age range of the controls was 17–87 years (mean ± SD 46.8 ± 15 years). The mean ± SD age and sex distribution of the nonrespondents were 57 ± 13.6 and 59.1% for the case group, and 46 ± 14 and 43% for the control group. Table 1 summarizes the baseline characteristics of the study population.

Table 1. Baseline characteristics of the case and control groups*
CharacteristicsCases (n = 480)Controls (n = 490)P
  • *

    Values are the number (percentage) unless otherwise indicated.

  • By t-test or chi-square test.

  • The job held longest by a participant.

  • §

    Each participant may practice more than 1 type of sport.

Female, %69.864.10.058
Age, mean ± SD years57 ± 1246.8 ± 150.000
Body mass index, mean ± SD kg/m229.8 ± 5.227.4 ± 5.30.000
 Sedentary49 (10.6)51 (12.8)0.28
 Housekeeping (housewives)307 (66.3)214 (53.5)0.000
 Laborious95 (20.5)113 (28.3)0.003
 Never participated323 (67.3)291 (59.4) 
 Practiced sport157 (32.7)199 (40.6)0.01
Knee osteoarthritis   
 Right knee88 (18.3)  
 Left knee90 (18.7)  
 Both knees302 (63)  

Participants who had worked most of their life in laborious jobs were not at greater risk of knee osteoarthritis. Further evaluation of the association between laborious jobs and knee osteoarthritis for men and women revealed no difference between the 2 groups (OR 0.97, 95% CI 0.6–1.56 for men and OR 1.03, 95% CI 0.41–2.58 for women). Further adjustment for knee injuries did not change these results.

Housewives had a borderline significant risk for knee osteoarthritis (OR 1.68, 95% CI 0.93–3.03) (Table 2). Further adjustment for the number of years spent at a job did not change the results. In a stratified analysis, we also evaluated whether this association was different in people age <50 years, which was not the case. People who had sedentary jobs were also not protected against knee osteoarthritis (OR 1.0, 95% CI 0.63–1.65).

Table 2. Association of knee osteoarthritis with jobs held since leaving school*
 OR (95% CI)Adjusted OR (95% CI)
  • *

    All analyses were evaluated using generalized estimating equation technique. OR = odds ratio; 95% CI = 95% confidence interval.

  • Adjusted for age, sex, and body mass index.

  • See Subjects and Methods for category descriptions.

  • §

    Measured only in women; adjusted for age and body mass index (kg/m2).

  • P < 0.05.

  • #

    P < 0.087.

 Sedentary0.81 (0.53–1.23)1.0 (0.63–1.65)
 Housekeeping  (housewife)§2.6 (1.51–4.5)1.68 (0.93–3.03)#
 Laborious0.72 (0.54–0.96)0.97 (0.63–1.48)
 Running/jogging0.95 (0.64–1.39)1.05 (0.70–1.58)
 Fitness/body- building0.75 (0.54–1.06)0.97 (0.67–1.39)
 Football/volleyball0.39 (0.23–0.68)1.09 (0.58–2.04)
 Other sports0.69 (0.47–1.02)1.09 (0.70–1.7)

A total of 157 (32.7%) participants in the case group practiced, or had practiced sports during their lifetime, compared with 199 (40.6%) members of the control group. After adjustment for age, sex, and BMI, no extra risk for knee osteoarthritis appeared to have resulted from any type of sports activity during lifetime, whether running/jogging, fitness/body-building, football/volleyball, or other sports (Table 2).

Table 3 summarizes the relationship of knee osteoarthritis to the 10 specified daily activities people had performed either in their work or during their leisure time since leaving school. We found none of the daily activities, such as standing, walking on flat ground, walking up/down hills, or sitting on the floor, to be a risk for knee osteoarthritis (Table 2).

Table 3. Association of knee osteoarthritis with life-long daily activities performed for at least 1 year*
ActivityOR (95% CI)Adjusted OR (95% CI)
  • *

    All analyses were evaluated using the generalized estimating equation technique. Daily activities were recorded for all jobs and leisure time activities held since leaving school. OR = odds ratio; 95% CI = 95% confidence interval.

  • Adjusted for age, sex, and body mass index.

  • P < 0.05.

  • §

    P 0.05–0.066.

 <1 hour/day (reference)  
 1–2 hours/day0.90 (0.63–1.28)0.97 (0.65–1.44)
 2–3 hours/day0.94 (0.65–1.35)0.75 (0.50–1.14)
 >3 hours/day1.24 (0.90–1.7)0.85 (0.58–1.24)
Walking on flat ground  
 <1 hour/day (reference)  
 1–2 hours/day0.74 (0.54–1.02)0.60 (0.42–0.85)
 2–3 hours/day1.15 (0.82–1.61)0.72 (0.49–1.07)
 >3 hours/day1.58 (1.12–2.22)0.92 (0.62–1.37)
Walking up/downhill  
 <30 minutes/day (reference)  
 >30 minutes/day0.94 (0.61–1.44)1.06 (0.66–1.69)
Sitting on the floor  
 <1 hour/day (reference)  
 1–2 hours/day0.94 (0.7–1.25)0.94 (0.69–1.29)
 2–3 hours/day0.89 (0.64–1.26)0.83 (0.57–1.21)
 >3 hours/day0.92 (0.65–1.29)0.82 (0.56–1.19)
Sitting on a chair  
 <1 hour/day (reference)  
 1–2 hours/day0.51 (0.35–0.74)§0.54 (0.36–0.80)
 2–3 hours/day0.57 (0.41–0.80)§0.87 (0.61–1.25)
 >3 hours/day0.58 (0.43–0.77)§0.94 (0.68–1.29)
 <30 minutes/day (reference)  
 >30 minutes/day1.94 (1.46–2.58)1.51 (1.12–2.04)
 <30 minutes/day (reference)  
 >30 minutes/day1.73 (0.89–3.37)1.98 (0.98–3.99)§
 <30 minutes/day (reference)  
 >30 minutes/day1.41 (0.88–2.27)2.06 (1.23–3.45)
 <2 kg/day (reference)  
 2–4 kg/day1.20 (0.92–1.56)1.12 (0.84–1.50)
 >4 kg/day1.22 (0.90–1.66)1.24 (0.87–1.76)
 <3 stories/day (reference)  
 3–5 stories/day0.62 (0.43–0.90)0.91 (0.61–1.37)
 5–10 stories/day0.77 (0.58–1.02)§1.06 (0.78–1.45)
 >10 stories/day0.72 (0.52–1.0)§0.99 (0.69–1.42)

We found that participants who sat in a chair between 1 and 2 hours a day were protected against knee osteoarthritis (OR 0.50, 95% CI 0.36–0.80), unlike people who spent less time sitting. However, for people who sat on a chair/sofa longer, we found the association was no longer significant (Table 2).

The risk of knee osteoarthritis was high in people who reported prolonged squatting for >30 minutes/day (OR 1.51, 95% CI 1.12–2.04). Cycling was also a significant risk factor (OR 2.06, 95% CI (1.23–3.45). Further investigation in men and women separately showed that cycling was a risk for knee osteoarthritis in men (OR 2.6, 95% CI 1.43–4.7), but not for women (OR 0.97, 95% CI 0.33–2.88). Work involving knee-bending was of borderline significance for risk of knee osteoarthritis.

Lifting loads heavier than 4 kg/day was not a risk factor for knee osteoarthritis (OR 1.24, 95% CI 0.87–1.76), and standing, walking on flat ground, walking up/downhill, or climbing stairs were also not considered risk factors.


These findings support our previous findings that squatting is a risk factor for knee osteoarthritis, irrespective of the participant's age, sex, and BMI. They also showed cycling to be a risk factor. Two other possible risk factors were knee-bending or working as a housewife. After additional adjustment for BMI, there was no longer an association between lifting heavy loads and knee osteoarthritis. The traditional Iranian way of sitting on the floor was not a risk factor for knee osteoarthritis, neither were other daily activities, including standing, walking on flat ground, walking up/downhill, climbing stairs.

This is the first community-based study conducted in Iran that evaluates risk factors for knee osteoarthritis in an urban area, and it has 3 particular strengths. First, the data were collected from 50 random clusters in different socioeconomic classes in Tehran. Second, to obtain detailed exposure information on risk factors, we used a life-course approach previously used by Dawson et al (5). Third, to help account for the contribution of site-specific determinants such as knee trauma, history of fracture, and muscle strength for the knee on the same side, we used logistic regression for repeated measurement (GEE). This technique calculates the relationship of each knee as the unit of analysis, but accounts for the correlation between both knees. By considering all these strengths mentioned above, we believe that we present a unique dataset on the risk factors for knee osteoarthritis that we hope will contribute significantly to the knowledge of the etiology of knee osteoarthritis.

Similar to the results of other studies, we demonstrated that squatting is associated with knee osteoarthritis (6, 7), although it was not confirmed by Tangtrakulwanich et al, who investigated squatting in Buddhist monks (8). Zhang et al reported a borderline significant association between radiologic osteoarthritis and squatting at age 25. They also reported a substantial difference in the prevalence of tibiofemoral osteoarthritis between Chinese and white subjects in a Framingham study, which, they concluded, was due to characteristically Asian postures such as squatting (9).

In a recent study, Thambyah et al showed that squatting created peak external moments that were more than 2.5-times greater than those during walking. The author suggested that squatting involves a reversal in the tibiofemoral shear-reaction force that occurs not only under full compressive load, but also within a fraction of a second. They believe that repeated squatting results in so many such reversals in shear reactions that they may ultimately have significant implications for the long-term mechanical function and structural integrity of the joint cartilage (10).

Although Coggon et al reported that jobs involving knee-bending were a risk factor for knee osteoarthritis (6), our own study shows a borderline significant association between knee-bending and knee osteoarthritis. However, this borderline significance may have been due to the fact that knee-bending was reported by only a few of the participants, which therefore resulted in less statistical power. Our study also showed an extra risk of knee osteoarthritis in people who cycled >30 minutes a day. Although Yoshimura et al found an association between cycling and knee osteoarthritis, this was no longer evident after adjustment for other factors (11). Although differentiation for men and women showed that cycling was a risk for knee osteoarthritis especially in men, but not in women, this may have been due to the small sample size of the women who used a bicycle.

After adjustment for BMI, the association we found between heavy lifting and knee osteoarthritis was no longer evident. This meant that our result contrasted with that of Coggon et al (6), who found lifting load as a risk for knee osteoarthritis. Unlike Coggon et al, we did not find that climbing stairs is a risk factor for knee osteoarthritis. This may be due to the fact that until recently, few buildings in Tehran were higher than 2 stories. Although these have now been superseded by multistory buildings, few of the people in our study, especially those who are older and experiencing knee osteoarthritis, have lived in a high building.

Anderson et al found an association between knee osteoarthritis and a history of heavy physical labor, such as farming (12). Although we expected laborious jobs to harm the knee and to be a risk factor for knee osteoarthritis, this association was not confirmed in our study. Speculating that participants who worked in laborious jobs might be younger people who had been doing such work for a shorter period, we investigated whether the association would be altered by adjusting for the number of years spent in such work. There was no change in our findings. Restricted analysis in participants >50 years old also showed that people with laborious jobs were not at greater risk than those in other jobs.

Control for age, sex, and BMI showed that women working as housewives showed only a borderline significant risk of knee osteoarthritis than those occupied outside the home. However, we speculate that this association might have become diluted since ongoing technological developments have reduced the burden of housekeeping. For example, Iranian women used to maintain a squatting position to hand wash the laundry. This is no longer the case. Similarly, cleaning the floor is no longer a task performed while on the knees.

The raw data in Table 1 shows that the control group participated in more sports and had more laborious jobs, and this difference was statistically significant. Although this difference might lead to the assumption that these factors are protective against knee osteoarthritis, adjustment for age and sex showed that these associations were no longer statistically significant, and that this effect was underlain largely by the 2 important risk factors of age and sex. Our study did not confirm the assumption made by Yoshimura et al that sedentary work is protective against knee osteoarthritis (11).

Data were collected by interviews according to a life-course approach. The fact that occupation and recreational activity can change throughout a person's life made it difficult to find associations with knee osteoarthritis, which develops only gradually. As Dawson et al also reported, a life-course approach helps reduce these problems by counting the years of exposure to each risk factor (5). However, despite this approach, we were still unable to confirm these associations for some of the risks we had expected, such as laborious work and climbing stairs. Details on developing the questionnaire, the examination sheet, and the validation have been submitted for publication by the APLAR-COPCORD group, with the participation of the Tehran COPCORD group, and a summary of the project has been published elsewhere (13).

Some bias may be inherent to this study. One possible source of bias is using a questionnaire to collect information about daily activities. To reduce the recall bias caused by asking about each daily activity participants had performed at work or during leisure time since leaving school, we recorded their life-time occupation and leisure-time activities on a life grid, i.e., the starting and finishing dates of jobs and periods of unemployment, of being a housewife, or of recreational periods. This grid was a source of reference for the trained interviewer to remind participants and to fill in the tables that followed recording the time participants spent on the 10 daily activities.

It is also possible that risk estimates might have been biased if the early symptoms of osteoarthritis had caused some cases to change their activities. If this had happened, the bias would underestimate the risk. However, because data were collected using a life-long approach, all analyses were adjusted by counting the number of years of exposure to particular variable analysis.

A third potential source of bias is that, on average, controls were 10 years younger, and therefore had fewer years of exposure to daily activity. Some controls may also have been too young to develop signs of clinical osteoarthritis. This limitation would have affected the associations found. However, to eliminate this effect, we matched controls to cases within the same age decades as closely as possible. We also adjusted all analyses for age, sex, and BMI. Having therefore minimized the effect of confounders as much as possible, we believe that the associations presented are very close to reality.

In conclusion, our findings support the theory that osteoarthritis has an overuse component in its pathogenesis. Regardless of age, sex, and excess load due to being overweight, repetitive movements in some daily activities increase the risk of knee osteoarthritis.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Dahaghin had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Tehrani-Banihashemi, Jamshidi, Davatchi.

Acquisition of data. Dahaghin, Faezi.

Analysis and interpretation of data. Dahaghin, Tehrani-Banihashemi, Davatchi.


The authors are grateful to the participants and staff of this study.