J Occup Rehabil (2010) 20:489–501 DOI 10.1007/s10926-010-9242-8

A Review of Occupational Knee Disorders Christopher R. Reid • Pamela McCauley Bush Nancy H. Cummings • Dianne L. McMullin • Samiullah K. Durrani

•

Published online: 20 May 2010 Ó Springer Science+Business Media, LLC 2010

Abstract Introduction Lower extremity knee disorders, like other cumulative disorders of the body, build up over time through cumulative exposures. 2006 data from the U.S. Bureau of Labor Statistics reveal that cumulative knee disorders account for 65% of lower extremity musculoskeletal disorders and 5% of total body musculoskeletal disorders. Methods The objective of the literature review was to find papers on work-related musculoskeletal disorders (WMSDs) common to the knee region. From these, symptoms of the disorders, affected industries, and potential risk factors were assessed. Results A review of the literature divulges that knee disorders primarily consist of bursitis, meniscal lesions or tears, and osteoarthritis. Though kneeling and squatting are considered to be two of the primary risk factors correlated to these knee disorders, 12 other risk factors should also be contemplated. These 14 contributing risk factors include both occupational

C. R. Reid (&)
P. M. Bush
S. K. Durrani Department of Industrial Engineering & Management Systems, University of Central Florida, 4000 Central Florida Blvd., Orlando, FL 32816-2993, USA e-mail: [email protected] P. M. Bush e-mail: [email protected] S. K. Durrani e-mail: [email protected] N. H. Cummings Athletic Training Education Program, Florida Southern College, Lakeland, FL 33801-5698, USA e-mail: [email protected] D. L. McMullin Ergonomics Applications, 2 Tradewinds Way, Salem, SC 29675, USA e-mail: [email protected]

(extrinsic) and personal (intrinsic) variables that affect the labor industries. Example industries include mining, construction, manufacturing, and custodial services where knee bending postural activities exist as a commonality. Conclusion The understanding of the types of knee disorders, the affected occupations, and the job related risk factors will allow ergonomic practitioners and researchers to create and adjust work environments for the detection and lessening of knee work-related musculoskeletal risk. Further studies need to be conducted to (1) justify the presence of risk from certain risk factors and (2) enhance the understanding of risk factor dose–response levels and their temporal development. Keywords Lower extremity
Occupational knee disorder
Knee bursitis
Knee meniscal lesion
Knee osteoarthritis
Work-related musculoskeletal disorders

Introduction The latest version of Injury Facts from the National Safety Council [1] states that in 2006, there were 145.6 million reported workers in US industries nation wide. Of these, there were 4.1 million total OSHA-recordable cases. These cases resulted in 1.2 million injury and illness Lost Work Day Cases (LWDCs) and accumulated 80 million lost work days. To put this into perspective for Work-related Musculoskeletal Disorders (WMSDs), the U.S. Bureau of Labor Statistics (BLS) mentions that for the same year there were 357,160 LWDCs [2]. The BLS continues on, revealing that 29,390 of these cases were related to the lower extremities (LE) of the body and that 19,070 of these LE cases alone were related to the knee region. Quick math on these data

123

490

divulge that cumulative knee disorders make up approximately 5 and 65% of total body WMSD cases and LE WMSD cases for that year respectively. In a similar review of Norway’s offshore petroleum industry, investigators reviewed 3,050 cases that occurred from 1992 to 2003 [3]. Their results found that 16% of the cases involved disorders to the LE. Intriguingly, 12% of all their cases were related to knee disorders in particular. In acknowledgement of incident rates, the BLS data reported knee disorders to be at 2.1 occurrences per 10,000 full-time employees with median lost work days per case at 20 days [2]. Cost related information for the year of 2004 to 2005 shows that incidents related to the knee, regardless of whether they are categorized as traumatic or cumulative, cost $19,785 per incident on average [1]. This figure, an increase of $1290 from the previous year [4], includes the costs of both indemnity and medical. A study conducted in the United Kingdom [5] found that 1% of their surveyed population had recorded job loss days due to the limitations caused by their knee disorders. Of those surveyed, 14% had a median of 14 lost work days. A review of the literature exposes the types of occupationally related disorders that exist for the knee and the industries that are concerned. This list of knee disorders includes knee bursitis, knee meniscal disorders, and knee osteoarthritis [5, 6]. The review of the epidemiology of these disorders discloses a list of risk factors that are both intrinsic (personal) and extrinsic (occupational) in nature to the worker being assessed. The evidence provided in this paper suggests that musculoskeletal disorders (MSDs), regardless of resultant body region, match the National Research Council’s (NRC) MSD conceptual model for causations of discomforts and disabilities [7]. Additionally the dose exposure–response relationships are reviewed and when applicable, they are consolidated into a set of risk guidelines for practitioner’s use.

Methods The objective of conducting a literature review was to determine the WMSDs that were related to the knee, detail potential risk factors, identify affected industries and list likely symptoms. Although the BLS suggests that it is a substantial problem, limited research has focused on the ergonomics of the lower extremity, the review from the BLS data pinpointed a region to focus efforts. Due to this, a wide berth was given to knee disorders when it came to year of publication. Books, conference papers, journals, government documents, and websites written in English from 1950 to 2009 that spoke of

123

J Occup Rehabil (2010) 20:489–501

disorders associated to these initial set of criteria were taken into consideration. Search Methods Online article databases were used to find conference papers, government documents, and journals. These databases included PubMed, eMedicine, WebMD, Science Direct, Ergonomics Abstracts, Google Scholar, and Ingenta. Books were gathered using the state of Florida’s university search tools that allow interlibrary searches and loans. 2006 Injury and Illness data was gathered by both contacting the U.S. Department of Labor’s Bureau of Labor Statistics for personal inquiry as well as using their website (http://www.bls.gov/). Body Mass Index information was gathered from the U.S. Department of Health and Human Service’s website of Centers for Disease Control and Prevention (CDC): http://www.cdc.gov/. Keywords used in the search included ‘knee’, ‘knee injury’, ‘knee disorder’, ‘knee illness’, ‘occupational knee injury’, ‘occupational knee disorder’, ‘cumulative knee injury’, ‘work-related disorder’, ‘musculoskeletal disorder’, and ‘work-related musculoskeletal disorder’. Combinations of the aforementioned terms were also used (e.g., knee ? musculoskeletal disorder). Initial results revealed specific types of knee disorders such as knee osteoarthritis, bursitis, knee bursitis, beat knee, meniscus lesion or tears, and meniscal lesion or tear which were then used for additional searches. Data Evaluation The review needed to be relevant and be applicable to industry, therefore it was essential that the results of the data be screened for recorded association between the disorders and an industry or job title. Results related to sports injuries were not focused on in this work, although athletic activities themselves were included when it was referred to as a personal risk by researchers. Following this initial screening criteria, references were then mined for etiology, epidemiology, and disorder symptoms. The last and most stringent screening stage required not only reported statistical correlation (odds ratios or prevalence ratios) between risk factor and disorder, but also a relationship to a given dose (exposure quantity) such as duration or frequency. This last screening stage proved to be too exclusive. The research on knee bursitis for example, lacked adequate information on exposure levels. As a result, as long as the disorders were considered occupationally related and they had published risk factor associations, then the disorder was included for the sake of increasing occupational knee disorder awareness.

J Occup Rehabil (2010) 20:489–501

Results Knee Bursitis The knee has two bursae (the prepatellar and infrapatellar) along its anterior portion [8]. Myllymaki et al. [8] describe bursitis as the inflammation of a bursa sac accompanied by swelling (through fluid retention in the bursa sac) and thickening of the bursa walls. What is distinctive for knee bursitis is that its incidence is so common to certain industries that nicknames from those industries have been given to it. For example, in the coal mining industry, knee bursitis has come to be known as either ‘‘miner’s knee’’ [9] or ‘‘beat knee’’ [8–12]. In the carpet and floor laying industries, ‘‘carpet-layer’s knee’’ is found in the literature [8, 11]. And for house cleaning or custodial businesses, ‘‘housemaid’s knee’’ is commonly spoken [11]. An investigation performed by Watkins et al. [12] discovered that knee bursitis incidents were not limited to a single occurrence in some cases. Of the 899 coal mining participants, 10% had recurring cases of the disorder. The impact of the disorder on the participant group was reported to have an average of 5.7 lost work shifts. One study performed in the carpet laying industry found knee bursitis in 8– 10% of their participants [13]. Two other studies noted 19% [14] and 20% [11] of carpet layers were diagnosed with knee bursitis. Additional industries affected by this knee disorder include manufacturing [15] and fishing [16]. Two primary types of extrinsic risk factors are noted by studies to be related to knee bursitis; using the knee as a hammer (sudden impact stress) and kneeling on or leaning against the knee (prolonged contact stress). Both types of stresses involve distributing forces through the knee to the knee bursae. Traumatic impact on the knee can occur by using devices such as knee kickers. Thun et al. [11] state that knee kickers are tools that utilize one of the knees as a hammer while kneeling. Knee kickers are notoriously unique to the carpet laying industry where they are used for stretching a carpet smoothly over a floor surface while getting it flush against a wall. The contact area for the kicking knee is allocated across its suprapatellar region. Bhattacharya et al. [17] determined that use of a knee kicker produces a range of 2,469–3,019 N of force against the knee, which was considered to be close to four times the body weight of the participants themselves. A later study carried out by Village et al. [18] had similar results with peak impact forces for seven male participants at 2,933 N (SD ±397) over 39 trial periods. In addition, the supporting non-kicking knee was noted to be at 894 N during the time of a knee kick. Enough exposure to this type of risk over time was confirmed by Thun et al. [11] to be correlated to the development of knee bursitis in carpet layers (Odds Ratio [OR] = 5.3, 90% CI: 2.8, 10.3).

491

While frequently using the knee as a hammer will indisputably speed up the development period of knee bursitis, it is less likely to be commonplace across industries. Prolonged contact stress on the other hand, from kneeling postures for example, is more likely to be used and pervade across multiple industries. Thun et al. [11] found from their investigation that carpet layers had a high prevalence towards knee bursitis (Prevalence Ratio [PR] = 3.2, 90% CI: 1.9, 5.4) and attributed this to high exposures of kneeling and using knee kickers. Physical exams of bursitis cases from their study noticed that 62% of those examined referred to the infrapatellar bursa as the most affected versus the prepatellar bursa region. Meticulous attention has also been given to kneeling and crawling in the coal mining industry where these postures and activities are known to be prevalent due to the restrictive work environment [9, 10, 12]. In contrast to Thun’s et al. findings [11], Sharrard’s [10] kneeling study for this industry found that prepatellar bursitis occurred twice as often as infrapatellar bursitis for the coal mining industry. Moore et al. [9] noted that employees who torso twist while kneeling add a supplementary affect on the knee cap and prepatellar bursa due to high torsional forces. Additionally, prepatellar bursitis was still noted to occur even when wearing knee pads for protection in the mining environment [12]. Their study recorded that 91% of their participants wore knee pads and used them throughout the work day. When the authors inspected further, they found that while kneeling, the body’s weight is distributed down the femur through the anterior knee surface (tibial tuberosity). When doing this with the constraints of knee pads, they suggested that forces might be redistributed back onto the prepatellar bursa region. Based on these remarks, further investigation into knee pad development should be conducted as they are not only for comfort and protection from debris but should also look at preventing knee bursitis. Compressive forces against the anterior of the knee have been noted for postures other than kneeling and squatting. Leaning against the knee while standing has also been noted as a risk factor. Torner et al. [16] found that fishermen on fishing vessels at times utilize their work environment for stabilization while working. Fishermen would balance themselves against railing or work surfaces using their knees and legs so that their hands could be free to perform other tasks. Prolonged exposure to this over time was found to lead to prepatellar bursitis. Personal hygiene is considered to be intrinsic to individuals. Moore et al. [9] found that hair follicles on the skin of the knee that become infected can also possibly lead to knee bursitis. Proper hygiene by workers (such as cleaning and disinfecting knee pads as well as replacing worn ones) is a plausible mitigation strategies noted by the investigators.

123

492

J Occup Rehabil (2010) 20:489–501

Meniscal Disorders Torn menisci are not just synonymous to athletic events, but are also known to occur in occupational settings. The lateral and medial menisci in the knee add structural integrity for alignment of the femur’s lateral and medial epicondyles (rounded end of the femur) during flexion. The menisci also provide cushioning and reduce friction during the joint’s articulation phase [9]. When a meniscal lesion does take place, Baker et al. [5] and Moore et al. [9] mention that pain and knee stiffness are frequently noted as symptoms followed by locking, swelling, and laxity. A meniscal lesion that takes place on the job has likely been building up for some time, though. Sharrard and Liddell [19] propose that cumulative knee laxity can be considered as a predisposition to a future meniscal disorder. Both of the studies performed by Sharrard and Liddell [19] and Atkins [20] identify the anterior cruciate ligament (ACL) as stretching or acquiring micro-tears over time. They add that this is likely to happen through frequent and prolonged episodes of kneeling activities where the lower legs are often rotated internally or externally and where jerking movements might possibly occur during work activities. With newfound knee laxity, the knee is now structurally more unstable and susceptible to injury during dynamic abrupt movements such as during a slip, twist, stagger, or fall [9, 10, 19]. Industries that are recognized as having meniscal disorder incidents include mining and carpet laying [19–23]. It is conceivable that these appearances are due to the high utilization of kneeling and crawling postural activities. Aside from these, athletic activities such as soccer [5, 24] and rugby [5] are shown to have a significant association towards this type of disorder (soccer [5]: OR = 6.9, 95% CI: 3.5,13.3; [24]: OR = 3.7, 95% CI: 2.1,6.6; rugby [5]: OR = 3.4, 95% CI: 1.5,7.8;). Outside of professional

athletes that are paid to participate in these events, voluntary participation for fun or as a hobby can label these activities as intrinsic to an individual. While kneeling has been revealed as a risk factor for meniscal disorders in several studies [5, 10, 19, 20, 24], few studies speak of additional risk factors to be aware of. Moore et al. [9] mentions that crawling on hands and knees or deep squatting should also be considered. The investigations by Baker et al. [5, 24] concurred with Moore et al. [9] that squatting should be considered as an occupationally based risk factor. In fact, Baker’s two studies even went as far as to begin expressing squatting risk exposure limits and their statistical associations (Baker [24]: OR = 1.8, 95% CI: 1.1, 3.0; Baker [5]: OR = 2.5, 95% CI: 1.2, 4.9). Other possible risk factors reviewed by the Baker et al. [5, 24] studies included the postures of sitting (while driving) and standing, and the activities of stair climbing, walking, and lifting or carrying of heavy objects (MMH). Although chair sitting, stair climbing and MMH tasks were considered significant risks in one study [24], standing and walking were not. Standing up from a kneel or squat position is another factor considered by these two studies to be a risk due to the stresses that are put on the knee during that type of movement. Only one of these studies found it to be a significant risk though (OR = 1.9, 95% CI: 1.2, 3.1) when performed more than 30 times per day [24]. All of the meniscal disorder extrinsic risk factors and their associated exposure levels and statistical significances are listed in Table 1. Knee Osteoarthritis One of the most debilitating occupational knee disorders is knee osteoarthritis (OA). This degenerative disease causes inflammation within the knee joint and is accompanied by cartilage rigidity and atrophy. The deformation and

Table 1 Extrinsic knee risk variables associated to knee meniscal disorders Occupational risk type Posture

Activity

Posture or activity

Statistical association

Source

Squatting

[1 h/day

OR = 1.8, 95% CI: 1.1,3.0

Baker et al. [24]

Squatting

[1 h/day

OR = 2.5, 95% CI: 1.2,4.9

Baker et al. [5]

Kneeling

[1 h/day

OR = 2.2, 95% CI: 1.3,3.6

Baker et al. [24]

Kneeling

[1 h/day

OR = 2.5, 95% CI: 1.3,4.8

Baker et al. [5]

Chair sitting (while driving)

[4 h/day

OR = 2.3, 95% CI: 1.4,4.0

Baker et al. [24]

Standing up from kneel or squat position

[30 times/day

OR = 1.9, 95% CI: 1.2,3.1

Baker et al. [24]

Stair climbing

[30 flights/day

OR = 2.4, 95% CI: 1.6,3.8

Baker et al. [24]

Lifting items C22 lbs

[10 times/week

OR = 1.9, 95% CI: 1.2,2.9

Baker et al. [24]

Lifting items C55 lbs

[10 times/week

OR = 1.7, 95% CI: 1.1,2.7

Baker et al. [24]

Lifting items C110 lbs

[10 times/week

OR = 2.4, 95% CI: 1.4,4.2

Baker et al. [24]

OR odds ratio, CI confidence interval

123

Exposure quantity

J Occup Rehabil (2010) 20:489–501

damage is exacerbated during movement and weight bearing periods which in turn affects additional joint structures within proximity such as the ligaments and tendons. Inflammation, bone spurs, cartilage wear, narrowed joint space between the femur and tibia, and bone to bone contact are all signs of knee OA [25, 26]. Symptoms of the disorder include pain, stiffness after sitting or sleeping, and joint weakness. Severity levels of the disorder are given during the diagnosis stage. These levels range from 0 to 4, with 0 being normal and 4 being the most severe signs of damage [27]. There are a broad range of industries whose workers become afflicted with knee OA. The published occupational listings include carpet and floor layers [6, 13, 28], construction workers [29, 30], mining workers [20, 23], and dock workers [31] to name a few (see Table 2). What stands out as connections between these industries are the intrinsic and extrinsic risk factors. Intrinsic risk variables for knee OA consist of the influence of obesity, age, past injuries and surgeries for the knee, and the physical activities and hobbies employees participate in outside of their work environments. Obesity is acknowledged by several studies as having an influence towards the development of knee OA [32–36]. Obesity itself is actually a category within the Body Mass Index (BMI) scale. Scores and categories within the BMI scale are determined by an individual’s height and weight. The other categories included are underweight, normal weight, and overweight [37]. The CDC [37] attaches BMI score ranges to each of these categories with underweight being considered as less than 18.5, normal weight between

Table 2 Occupations affected by knee osteoarthritis Occupation

Source

Carpenter

Jensen et al. [13, 28]

Carpet/floor layer

Jensen et al. [13, 28], Kivimaki et al. [6]

Civil servants

Partridge and Duthie [31]

Construction worker

Sandmark et al. [29], Vingard et al. [30], D’Souza et al. [39]

Custodial workers (female)

Rossignol et al. [55], Vingard et al. [30]

Dock worker

Partridge and Duthie [31]

Farm worker Firefighter

Sandmark et al. [29], Vingard et al. [30], D’Souza et al. [39] Vingard et al. [30]

Fishing workers

Lau et al. [36]

Forestry worker

Sandmark et al. [29]

Millwrights and Bricklayers

Thun et al. [11]

Mine workers

Atkins [20], McMillan and Nichols [23]

Tilesetter

Thun et al. [11]

493

18.5 and 24.9, overweight between 25 and 29.9, and obese being greater than 30. Anderson and Felson [32] mention an additional category in their study for very obese that carries with it a BMI score of greater than 35. The studies by Cooper et al. [34], Anderson and Felson [32], and Coggon et al. [33] all point out that individuals who are considered as overweight are already at risk for knee OA. This risk factor to disorder relationship continues for obese and very obese as well (Table 3). Some studies have looked to see if an association exists between combined risk factors such as BMI or age with knee bending postural activities (see Table 3) [32–34]. For the BMI and knee bending combination, the result seems to have made the association between BMI and knee OA even more sensitive. Normal weight categories (BMI: 18.5–24.9) when combined with those postural activities are now a risk factor where as when separated, BMI as a risk factor needs to be at least 25. Age ([54 years old) is another category that is also brought into consideration when combined with kneeling, squatting, or stair climbing (see Table 3). Anderson and Felson’s [32] investigation noted that although women in particular were recorded to have an increased prevalence towards knee OA at a younger age bracket than did (45–54 years old) than did their male counterparts, it was not found to be a statistically significant factor when combined with knee bending activities. Another set of intrinsic risk factors mentioned by the literature includes whether or not individuals may have had previously diagnosed knee injuries that affected the knee joint itself (see Table 3). This also includes whether or not they had surgery on those joints [33, 34, 36, 38]. Lau’s et al. [36] Hong Kong population study for example, found this relationship to be significant amongst both men and women (male: OR = 12.1, 95% CI: 3.4,42.5; women: OR = 7.6, 95% CI: 3.8,15.2). Coggon et al. [33] and Manninen et al. [38] (for men) too acknowledged that previous injury or surgery can be a risk in itself for knee OA ([33]: OR = 4.5, 95% CI: 3.0, 6.8; Males [38]: OR = 3.44, 95% CI: 1.74, 6.80). Cooper’s et al. [34] study regarded both past knee injury as a risk by itself and in combination with knee bending activities. Both conclusions were found to be significant factors with past injury itself having an OR of 7.8 (95% CI: 3.0, 20.2) and that an individual’s injury history combined with frequent and or prolonged knee bending postural activities to be an OR of 7.6 (95% CI: 2.1, 26.9). Some of these past injuries may have been related to sports incidents. Therefore, just like when assessing for meniscal disorder risk factors, sports, hobbies, and activities outside of work should be reflected on by practitioners as well. Lau’s et al. [36] analysis results indicated that women who practiced kung fu or gymnastics showed a significant association to knee OA (kung fu:

123

494

J Occup Rehabil (2010) 20:489–501

Table 3 Intrinsic knee risk variables associated with knee osteoarthritis Intrinsic risk type Injury history

Body mass index (overweight)

Body mass index (obese)

Body mass index (very obese)

Intrinsic risk

Statistical association

Source

Past injury or surgery*

OR = 12.1, 95% CI: 3.4,42.5

Lau et al. [36]

Past injury or surgery**

OR = 7.6, 95% CI: 3.8,15.2

Lau et al. [36]

Past injury or surgery*

OR = 3.44, 95% CI: 1.74,6.80

Manninen et al. [38]

Past injury or surgery

OR = 7.8, 95% CI: 3.0,20.0

Cooper et al. [34]

Past injury or surgery

OR = 4.5, 95% CI: 3.0,6.8

Coggon et al. [33]

BMI [25.3

OR = 3.6, 95% CI: 1.7,7.5

Cooper et al. [34]

BMI 25–29.9*

OR = 1.69, 95% CI: 1.03,2.80

Anderson and Felson [32]

BMI 25–29.9**

OR = 1.89, 95% CI: 1.24,2.87

Anderson and Felson [32]

BMI 25–29.9

OR = 3.2, 95% CI: 2.2,4.7

Coggon et al. [33]

BMI 30–35*

OR = 4.78, 95% CI: 2.77,8.27

Anderson and Felson [32]

BMI 30–35** BMI C30

OR = 3.87, 95% CI: 2.63,5.68 OR = 8.3, 95% CI: 5.2,13.4

Anderson and Felson [32] Coggon et al. [33]

BMI [35*

OR = 4.45, 95% CI: 1.77,11.18

Anderson and Felson [32]

BMI [35**

OR = 7.37, 95% CI: 5.15,10.53

Anderson and Felson [32]

Risk factors combined with kneeling, squatting, or stair climbing postural activities Age

Age C55–64*

OR = 2.45, 95% CI: 1.21,4.97

Anderson and Felson [32]

Age C55–64**

OR = 3.49, 95% CI: 1.22,10.52

Anderson and Felson [32]

Injury history

Past injury or surgery

OR = 7.6, 95% CI: 2.1,26.9

Cooper et al. [34]

Body mass index (normal weight)

BMI 18.5–24.9

OR = 2.2, 95% CI: 1.1,4.5

Coggon et al. [33]

Body mass index (overweight)

BMI 25–29.9

OR = 6.1, 95% CI: 3.4,10.9

Coggon et al. [33]

Body mass index (obese)

BMI C30

OR = 14.7, 95% CI: 7.2,30.2

Coggon et al. [33]

OR odds ratio, CI confidence interval * Males, ** Females

OR = 22.5, 95% CI: 2.5, 199; gymnastics: OR = 7.4, 95% CI: 2.6, 20.8). As stated previously, knee bending postures and activities are considered as extrinsic risk factors for knee OA. Several studies mention kneeling and squatting, hence they can be considered as the most popular risk factors to look for in the work setting [13, 14, 33, 34, 38, 39]. A brief overview of Table 4 allows one to see that there are a range of different risk exposure quantities for the kneeling and squatting variables. This minimum threshold range can go from as little as 30 min in the study by Cooper et al. [34], to as much as 2 h (for women) during an 8 h work day as stated by Manninen et al. [38]. In their review of the Third National Health and Nutrition Survey (NHANES III), D’Souza et al. [39] find that kneeling can be associated to a percentage of a work day’s daily activities (14% or 1.12 h in the case of kneeling for men). In fact, the study noticed the existence of a positive trend amongst the odds as kneeling exposure duration increased. Standing was also noted to be a risk amongst women in this study when they were exposed to it 30–36% of their work day. It was not included in the guideline though due to its statistical insignificance when exposure duration is increased past this point. Many of the studies reviewed referred to physical workload as a means of defining the extent of physical

123

exertion and type of physical activities that are performed. Five descriptions have been defined by the U.S. Department of Labor as a narrative for occupational job titles [40]. When physical workload refers to ‘‘sedentary’’, they mean to describe work that has no more than 10 lbs of lifting per item and that little walking or standing is performed. ‘‘Light’’ workload increases maximum lifting weight to 20 lbs per item and if this includes repetitive lifting, then the limit is 10 lbs per item. ‘‘Medium’’ physical workload has a maximum lift of 50 lbs and a recurring lift of 25 lbs per item. ‘‘Heavy’’ workloads have a maximum lift of 100 lbs per item and 50 lbs per item when recurring. The last category of ‘‘very heavy’’ includes items that exceed the two limitations given for ‘‘heavy’’ workload (maximum: [100 lbs; recurring: [50 lbs). Understanding these types of physical workloads helps clarify the studies that include lifting or carrying activities as knee OA risk factors [29, 33, 35, 36, 38, 39]. The studies listed in Table 4 for lifting by Coggon et al. [33] and Lau et al. [36] depict a lifting range that based on weight alone, would have a physical workload range from ‘‘light’’ to ‘‘very heavy’’ (depending on whether the frequency of exposure is a single lift or repetitive). One thing that is interesting to note about these two studies is that they only look at an exposure frequency of 10 lifts or carries per

J Occup Rehabil (2010) 20:489–501

495

Table 4 Extrinsic knee risk variables associated to knee osteoarthritis Occupational risk type Posture or activity Posture

Activity

Exposure quantity

Statistical association

Source

Squatting

[30 min/day

OR = 6.9, 95% CI: 1.8,26.4

Cooper et al. [34]

Squatting

[1 h/day

OR = 2.3, 95% CI: 1.3,4.1

Coggon et al. [33]

Kneeling

[30 min/day

OR = 3.4, 95% CI: 1.3,9.1

Cooper et al. [34]

Kneeling

[1 h/day

OR = 1.8, 95% CI: 1.2,2.6

Coggon et al. [33]

Kneeling*

[14% (1.12 h)/day OR = 3.08, 95% CI: 1.31,7.21 D’ Souza et al. [39]

Kneeling or squatting**

C2 h/day

OR = 1.81, 95% CI: 1.11,0.95 Manninen et al. [38]

Standing up from kneel or squat position Stair climbing

[30 times/day

OR = 1.7, 95% CI: 1.2,2.6

[10 flights/day

OR = 2.7, 95% CI: 1.2,6.1

Cooper et al. [34]

Stair climbing**

C15 flights/day

OR = 5.1, 95% CI: 2.5,10.2

Lau et al. [36]

Stair/ladder climbing*

[30 flights/day

OR = 2.3, 95% CI: 1.3,4.0

Coggon et al. [33]

Lifting items C22 lbs*

C10 times/week

OR = 5.4, 95% CI: 2.4,12.4

Lau et al. [36]

Cooper et al. [34]

Lifting items C22 lbs**

C10 times/week

OR = 2.0, 95% CI: 1.2,3.1

Lau et al. [36]

Lifting items C22 lbs

[10 times/week

OR = 1.7, 95% CI: 1.2,2.4

Coggon et al. [33]

Lifting items [22 lbs*

[20% (1.6 h)/day

OR = 2.72, 95% CI: 1.14,6.50 D’ Souza et al. [39]

Lifting items C55 lbs

[10 times/week

OR = 1.7, 95% CI: 1.2,2.6

Coggon et al. [33]

Heavy lifting combined with kneeling, squatting, or stair climbing

[55 lbs/item

OR = 5.4, 95% CI: 1.4,21.0

Cooper et al. [34]

Lifting/carrying combined with kneeling, C25–50 lbs/item squatting, crouching or crawling*

OR = 2.22, 95% CI: 1.38,3.58 Felson et al. [35]

Heavy lifting combined with kneeling or [55 lbs/item squatting

OR = 3.0, 95% CI: 1.7,5.4

Coggon et al. [33]

[2 miles/day

OR = 2.1, 95% CI: 1.4,3.2

Coggon et al. [33]

Walking** OR odds ratio, CI confidence interval * Males, ** Females

week [33, 36]. The D’Souza et al. [39] study looked at lifting or carrying exposure thresholds as percentages of the work day. In this case, lifting objects weighing more than 22 lbs for at least 20% of the work day was considered as a risk for men (OR = 2.72, 95% CI: 1.14,6.50). In their review of the literature, Reid et al. [41] noted similarities in dose exposure–response quantities between discomfort and disorder. An example case can be seen from Sobti et al. [42] who mention an association between climbing more than 30 flights of stairs per day and knee discomfort (Relative Risk [RR] = 1.17, 95% CI: 0.99,1.38). Cases from the Norwegian Petroleum Safety Authority seem to also show this relationship between knee disorders and the acts of climbing stairs or ladders or walking on hard surfaces [3]. Stair climbing is revealed to be a significant risk factor activity in a couple other studies as well [33, 34, 36]. Similar to Sobti’s et al. [42] knee discomfort affiliation, results of these studies depict that initial risk exposures can vary from 10 to 30 stair flights per work day. Coggon et al. [33] also make note of standing up from a kneel or squat and walking as another set of risk factors in their research. Greater than 30 exposures per day

was considered one risk to employees (OR = 1.7, 95% CI: 1.2, 2.6) while walking more than 2 miles per work day was associated to knee OA for women (OR = 2.1, 95% CI: 1.4, 3.2). One last activity suspected in the Lau et al. [36] Hong Kong Chinese survey proposed that vibration from more than 1 h of tool usage could be a knee OA risk. The result of this association was not found to be statistically significant though. Further detail such as types of tools or vibration frequencies were not given in the investigation. This gives opportunity for future risk association research in the occupationally related vibration field. Similar to the combination of risk factors seen for intrinsic variables (Table 3), combining multiple extrinsic variables such as lifting and carrying with knee bending activities (e.g., carrying a heavy object while stair climbing) has also been reviewed from retrospective studies. Coggon et al. [33], Cooper et al. [34], and Felson et al. [35] noticed that the existence of this combinational effect is illustrated when people lift or carry items weighing at least 25 lbs (Table 4). What remains unknown based on the literature reviewed is the risk dosage to the frequency or to the possible durations of exposure.

123

496

J Occup Rehabil (2010) 20:489–501

Discussion To date, epidemiology has captured occupational lower extremity (in this case the knee) risk exposure data through the quantification of frequencies and durations. These exposure levels associate to the minimum level of a cause– effect relationship as was verified by each study’s statistical significance. Exposure quantities in this review are based on either a national standard set by a government institution (such as a job title) or by the study investigators themselves. Although the latter method can lead to recall bias [43], investigators utilizing this option find that it allows a more complete occupational history for establishing a risk threshold [33, 34, 36]. Many exposure thresholds decided on by these same studies were done based on previous studies that initiated an exposure quantity for these thresholds. In the study by D’Souza et al. [39], the investigators utilized ergonomic subject matter experts to develop their risk factor exposure thresholds into proportions of a work day. They made a point to note that they could not say with complete certainty to what extent these thresholds were valid. To that extent, one cannot accurately say where and when hazard exposure brings about morbidity symptoms with these current observational design methods of risk assessment either. One can though use the results of these thresholds to act as an initial set of guidelines to allow experienced occupational safety and ergonomic practitioners a tangible concept of risk identification. With that, the three knee disorders discussed in this review are not only related through the body part affected, but also through the relation of the risk factors that are associated between them. Reference to the MSD causation Table 5 Fourteen risk variables that are associated to knee disorders

Risk variables

Extrinsic risk

Kneeling Squatting or crouching Crawling Stair/ladder climbing

Intrinsic risk

OA

Meniscal disorder

Knee Bursitis

X

X

X

X

X

X

X

X X

X X

X X X

Lifting/carrying/moving

X

X

Walking

X

X

Standing up from a kneel/squat/crawl

X

X

Chair sitting (while driving)

X

X X

BMI

X

Past knee injury/surgery

X

X

Age

X

X

X

Using the knee as a hammer (impact stress)

X

X

Prolonged contact stress against the patella bone other than when kneeling

X

X

Physically intensive habits/hobbies that could affect the knee

123

model ties both personal and occupational risk factors together as complimentary to each other for affecting the onset of low back and upper extremity disorders [7]. It is plausible that a similar approach can be examined when referring to knee disorders. A list of the 14 risk factors discussed throughout this review is shown in Table 5 below. Of the knee-related risk factors, kneeling is the most common to all of the listed knee disorders. A review of the references exposed those studies that resulted in not only a correlation of risk factor to disorder but also included an exposure quantity. Examples of these are shown for all three disorders except for knee bursitis, whose studies did not include an exposure quantity. Knee meniscal and OA disorders were also the only ones of the three to have risk factors listed that are relative to intrinsic nature. The nine studies mentioned in the guideline tables [1, 2, 4, 5] for meniscal and OA knee disorders and the single study about knee bursitis are core studies that look at the exposure quantity associations between physical stresses and knee morbidity. In addition, research on knee OA has shown to also have personal attributions. Aside from their results, these studies vary from each other in the nature of their research design. The most popular type of observational study noted amongst these is the retrospective case– control [24, 33, 34, 36, 38]. Four other studies were retrospective cross-sectional [5, 11, 32, 39] with one of them having the addition of a nested case control [5]. Only one of these ten studies was considered as a prospective cohort study [35]. While this review is not intended to investigate which type of observational study should be used over others, it does however, look into the strengths and weaknesses of

X

X

X

J Occup Rehabil (2010) 20:489–501

each of these studies in order to help legitimize the recommendations in the given guidelines. Historically, prospective study results have had preference over retrospective studies due to their ability to control for biases, exposures, confounders, and endpoints. Prospective studies are known however, to consist of a lengthy data collection period which can lead them to become expensive to perform. On the other hand, retrospective studies have the positive light of collecting past data rather easily and with less cost. Its negative aspects though are the inverse of the prospective design’s highlights. This leads to loss of the overall strength and validity of the study’s results [44]. Cohort (both retrospective and prospective), case–control, and cross-sectional observational studies all have the ability to establish cause–effect relationships (although the latter is noted to have less potential) [45]. Noordzij et al. [45] mention that case–control studies additionally have the strength of considering the multiple exposures and their outcomes. Specific biases for all three types of studies include selection bias (such as choosing a case or control group), recall bias (for case–control) and survival bias (for cross-sectional). In addition to these, some of this review’s studies also mention referral bias which may raise the prevalence of more severe cases. Many of these studies note the possibility of confounding. Examples include BMI, past injury or surgery, age, and participation in sporting activities (such as soccer). Nevertheless, if it was or was not possible to account for these biases and confounders, then they were typically noted by investigators throughout the studies of the guidelines. Details of each of the ten studies are given in Table 6.

Conclusion Current Methods of Risk Exposure Containment and Avoidance In their analysis of the NHANES III, D’Souza et al. [39] made mention that just about 20.7% of knee OA cases can be associated to male workers kneeling more than 14% of their workday. Due to that risk and other risks associated to the disorders mentioned in this review, a set of risk association tables were created as an initial set of guidelines to help raise knee disorder awareness amongst practitioners and researchers. While personal risk factors cannot be controlled by employers, influence is possible for occupationally related ones. Administrative controls for instance, can be considered by employers to help limit exposure to these risk factors. Examples of these types of controls include increasing awareness of the knee disorders and their risk factors through employee training programs. Implementing proper hygiene on a daily basis for work

497

clothing and personal protective equipment (PPE) [9] or taking into account work-related risk exposure restrictions (such as those mentioned in this review) are other considerations. Moore et al. [9] also suggests that workers exposed to prolonged episodes of kneeling or squatting consider flushing of the knee. This is conducted simply by flexing and extending the leg at the knee. Proactively, employers can also offer stretching and strengthening classes [9] such as through industrial athletic programs to assist employees in being better prepared for their work environment. Other common administrative suggestions to reduce employee exposure when applicable are job rotation, change in work procedure, or change in work technique. An example of the latter suggestion was an alternative in one study where a kneeling support device was compared against squatting and stooping techniques for vine crop picking agricultural workers [46]. The investigators found that for tasks such as pepper picking, an effective technique reduces time while increasing harvest productivity. The use of a kneeling support device would reduce harvest productivity by increasing the time needed to change positions and locations. So relevance of the proposed solution is something that employers would have to contemplate. Engineering solutions can also be considered for ergonomic strategies. This is typically the primary focus noted in industrial ergonomic mitigations. Lifting-related issues could be engineered out by introducing lifting aids, automation, or altering object weights to make them either too light or too heavy to safely lift. Workers in the carpet laying industries can also choose alternatives to the knee kicking device used for carpet stretching such as the crab re-stretcher, power stretcher [18], or carpet air stretcher [47]. These devices harness the necessary amount of force through mechanical and air pressure sources instead of from knee impact. In the results of their study, Village et al. [18] mention that kneeling force on the supporting knee was reduced to 476 N for the crab re-stretcher and 368 N for the power stretcher from the 894 N as was previously mentioned for a knee kicking device. Another fairly new thought is the implementation of exoskeletons that augment human capacities by taking on the weight of individuals or the materials they are handling. Systems such as that designed by Honda [48] are unique in that while the Bodyweight Support Assist device is attached to the wearer’s LE, it helps support the weight of the wearer during prolonged standing or squatting episodes as well as while walking and during stair ascent or descent. Expanding on this concept, many other exoskeleton systems such as the Berkeley Lower Extremity Exoskeleton (BLEEX) [49], Human Universal Load Carrier (HULC) [50], and Hybrid Assistive Limb (HAL) [51] additionally aid wearers with the excessive lifting and prolonged

123

123 Questionnaire

Knee meniscectomy Squatting

Kneeling;

Squatting

Lifting objects; Stair climbing; Standing up from kneel/ squat;

Interview; Medical exam

Medical exam; Job Title (Dep. Of Labor Questionnaire Dictionary of Occup. Titles [US] [40])

Knee OA diagnosis (radiograph)

Knee radiographs; Provided continuous occupational status information

Knee OA Cross-sectional (NHANES III)

Knee OA ProspectiveCohort (Framingham Study)

D’Souza et al. [39]

Felson et al. [35]

Kneeling; Lifting Objects; Squatting

Defined by investigators (using Kneeling; Lifting Objects; ergonomic subject matter Standing expert opinion)

Squatting

Standing up from kneel/squat;

Defined by investigators (based BMI; on previous study [58]) Kneeling; Lifting Objects; Past Injury/Surgery; Stair climbing;

Interview; Medical exam

Knee OA diagnosis (radiograph)

Knee OA Case–control

Cooper et al., 1994 [34]

Squatting; Walking

Stair climbing;

55–90

1,054 males; 916 females

79 female cases; 158 female controls

30 male cases; 60 male controls;

205 male cases; 205 controls; 313 female cases; 313 female controls

335 male controls

67 male cases;

Mean = 73 569 male; 807 female

60–90 Referral bias (cases); Exposure assessment bias

Recall bias

NA (none Age; BMI; mentioned by Education level; study) Past injury/ surgery; Smoking

Age; Current BMI; BMI at 25 years old; Gender; Smoking history

BMI; Heberden’s nodes

47–93

20–59 Referral bias (cases); Selection bias (cases/ controls); Recall bias

Selection bias BMI; Heberden’s (cases/controls); nodes; Past injury/ Recall bias surgery; Smoking

Sports participation

461 controls

47 female cases;

196 male cases;

20–59

Selection bias (cases/controls); Recall bias

BMI; Joint laxity; Sports participation; Social class

Chair sitting (while driving); Kneeling;

2,428 male; 2,765 female

35–74

NA (none mentioned by study)

Age; BMI; Education level; Race

Age; BMI

Sample size

Participant ages

Biases considered by study

Confounders considered

Risk factors considered for review guidelines

Defined by investigators (based BMI; Kneeling; on previous study [57]) Lifting Objects; Past Injury/ Surgery;

Knee OA diagnosis (radiograph)

Knee OA Case–control

Interview; Medical exam

Job title (Office of Pop. Censuses and Surveys—std. occupational classification scheme [London] [56])

Job title (office of pop. censuses and surveys—std. occupational classification scheme [London] [56])

Coggon et al. [33]

Baker Meniscal Cross-sectional et al. [5] disorder with nested case control

Interview; Medical exam

Meniscal tear diagnosis (arthroscopy)

Meniscal Case–control disorder

Baker et al. [24]

Medical exam; Job Title (Dep. Of Labor Questionnaire Dictionary of Occup. Titles [US] [40])

Knee OA diagnosis (radiograph)

Knee OA Cross-sectional (NHANES I)

Data collection Exposure threshold estimation method method

Anderson and Felson [32]

Nature of study Case definition

Disorder type

Study

Table 6 Knee risk guideline study comparison chart

498 J Occup Rehabil (2010) 20:489–501

55 male cases; 140 male controls; 226 female cases; 384 female controls

397 males

55–75

19–87

Recall bias

NA (none mentioned by study)

BMI; Past knee injury/surgery; Sports participation

carrying durations noticed amongst many industrial and military tasks. Lastly, PPE such as knee pads should be administered by employers and used by employees to help prevent future knee morbidity. Manufacturers of knee pads should also consider the results of bursitis-related research suggested by both past [12] and present studies [52]. Results of Porter’s et al. [52] study noted that although current knee pads sufficiently distribute the pressures of body weight across the knee and upper tibia surfaces, a significant quantity of compression is still transmitted to bursa sacs ([25 psi). The authors suggest that future knee pad design focus on reducing this pressure to the bursa areas along the patellar tendon and tibial tubercle. Further investigations along these focuses can contribute to allowing science to continue to push the boundaries of discovering where inference of these risk factor exposures ends and quantification of them begins.

NA not mentioned by the study

Past injury/surgery

Knee Impact (from Age knee kicker) Medical exam; NA (No exposure Questionnaire quantity defined) Thun et al. Knee CrossKnee disorder [11] Bursitis sectional symptoms

Defined by investigators First knee arthroplastic surgery Knee OA Case– control Manninen et al. [38]

Interview

Kneeling; Squatting

Referral bias NA (no age 166 male cases; 166 male (cases) given) controls; 492 female cases; 492 female controls NA — (Performed multiple logistic regression to account for confounding between variables) Lifting objects; Past injury/ surgery; Stair climbing Defined by investigators (based on previous studies [34, 58]) Interview; Medical exam Knee OA diagnosis (radiograph) Knee OA Case– control Lau et al. [36]

Nature of study Disorder type

Table 6 continued

499

Future Aims of Data Collection Research

Study

Case definition Data collection Exposure threshold method estimation method

Participant ages

Sample size Biases considered by study Confounders considered Risk factors considered for review guidelines

J Occup Rehabil (2010) 20:489–501

More epidemiological studies should be performed using prospective cohort methods. Although this method is expensive and lengthy, it can lead to more conclusive temporal evidence of the risk factor relationships mentioned throughout this review. Additionally, more studies should look at defining exposure thresholds as well. Is a person at risk if their daily work exposures are 30 min, versus, 1 h, versus 2 h? While what has been done so far is a good start for establishing guidelines and preliminary lower extremity risk models [53], it still depends on existing nomenclature from either occupational job titles or the decisions of study investigators. Studies such as Lau et al. [36] looked at vibration from tools as a risk factor towards knee OA. Further study of vibration’s influence on knee morbidity from tools and work surfaces needs to be done to see if this is truly significant and at what dose–response quantity. Another variable to consider for a future study would be standing. Although Baker et al. [5, 24] did not find that it had a significant statistical association to knee meniscal disorders, D’Souza et al. [39] did notice that it associated to knee OA for women. Further study is needed to define it as either a protective or a risk variable. Sitting while driving is another risk factor (for meniscal disorders) needing further investigation. Baker et al. [24] question its significance as they wonder if it occurred by chance. In their follow up study [5], this factor was shown as insignificant. It is due to its initial statistical significance and its mention in a study centered on knee pain from long driving durations [54] that this factor has been left in this review’s guidelines. Future studies for this exposure–result relationship should reveal more. Meanwhile, it is widely

123

500

known that exercising leads to a healthy life, but to what extent should exercising be done when it may lead to knee disorders? Examples of this are seen from the results of Coggon’s et al. [33] study showing an association between walking and knee OA for women and those studies showing athletic activity’s connection to meniscal and OA problems [5, 24, 36]. Understanding these physiological risks may also lead researchers to look beyond epidemiology alone and tie in other physiological evidence such as those found biomechanically or otherwise.

J Occup Rehabil (2010) 20:489–501

17.

18. 19. 20. 21. 22.

References 1. National Safety Council. Injury facts 2008 edition. Itasca: National Safety Council; 2008. 2. Bureau of Labor Statistics. Workplace injuries and illnesses in 2006. Annual Report. Washington D.C.: United States Department of Labor; 2007. Report No.: 07-1562. 3. Morken T, Mehlum IS, Moen BE. Work-related musculoskeletal disorders in Norway’s offshore petroleum industry. Occup Med. 2007;57(2):112–7. 4. National Safety Council. Injury facts. 2007 edition ed. Itasca: National Safety Council; 2007. 5. Baker P, Reading I, Cooper C, Coggon D. Knee disorders in the general population and their relation to occupation. Occup Environ Med. 2003;60:794–7. 6. Kivimaki J, Riihimaki H, Hanninen K. Knee disorders in carpet and floor layers and painters: part II: knee symptoms and patellofemoral indices. Scand J Work Environ Health. 1994;26: 97–101. 7. National Research Council. Musculoskeletal disorders and the workplace: low back and upper extremities. Washington, D. C.: National Academy Press; 2001. 8. Myllymaki T, Tikkakoski T, Typpo T, Kivimaki J, Suramo I. Carpet-layer’s knee: an ultrasonographic study. Acta Radiol. 1993;34:496–9. 9. Moore SM, Steiner ME, Mayton AG, Fitzgerald GK, Hubert JP. Keeping knees healthy in restricted work spaces: applications in low-seam mining. CDC Information Circular 9504. Pittsburgh: Department of Health and Human Services (National Institute for Occupational Safety and Health); 2008. Report No.: 2008-130. 10. Sharrard WJW. In: Pressure effects on the knee in kneeling miners Joseph henry lecture May 15, 1964; Sheffield, UK.; 1965. p. 309–24. 11. Thun M, Tanaka S, Smith AB, Halperin WE, Lee ST, Luggen ME, et al. Morbidity from repetitive knee trauma in carpet and floor layers. British J Ind Med. 1987;44:611–20. 12. Watkins JT, Hunt TA, Fernandez RHP, Edmonds OP. A clinical study of beat knee. British J Ind Med. 1958;15:105–9. 13. Jensen LK, Mikkelsen S, Loft IP, Eenberg W. Work-related knee disorders in floor layers and carpenters. J Occup Environ Med. 2000;42(8):835–42. 14. Kivimaki J, Riihimaki H, Hanninen K. Knee disorders in carpet and floor layers and painters. Scand J Work Environ Health. 1992;18:310–6. 15. Bruchal LC. Occupational knee disorders: an overview. In: Brittner AC, Champney PC, editors. Advances in industrial ergonomics and safety. 7th ed. London: Taylor & Francis; 1995. 16. Torner M, Almstrom C, Karlsson R, Kadefors R. Working on a moving surface-a biomechanical analysis of musculoskeletal load

123

23.

24.

25. 26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

due to ship motions in combination with work. Ergonomics. 1994;37(2):345–62. Bhattacharya A, Mueller M, Putz-Anderson V. Traumatogenic factors affecting the knees of carpet installers. Appl Ergon. 1985; 16(4):243–50. Village J, Morrison JB, Leyland A. Biomechanical comparison of carpet-stretching devices. Ergonomics. 1993;36(8):899–909. Sharrard WJW, Liddell FDK. Injuries to the semilunar cartilages of the knee in miners. British J Ind Med. 1962;19:195–202. Atkins JB. Internal derangement of knee joint in miners. British J Ind Med. 1957;14:121–6. Jensen LK, Eenberg W. Occupation as a risk factor for knee disorders. Scand J Work Environ Health. 1996;22:165–75. Kivimaki J. Occupationally related ultrasonic findings in carpet and floor layers’ knees. Scand J Work Environ Health. 1992; 18(6):400–2. McMillan G, Nichols L. Osteoarthritis and meniscus disorders of the knee as occupational diseases of miners. Occup Environ Med. 2005;62:567–75. Baker P, Coggon D, Reading I, Barrett D, McLaren M, Cooper C. Sports injury, occupational physical activity, joint laxity, and meniscal damage. J Rheumatol. 2002;29(3):557–63. Felson DT. Osteoarthritis of the knee. New Engl J Med. 2006; 354:841–8. What is osteoarthritis? [homepage on the Internet]. http:// www.webmd.com/osteoarthritis/guide/osteoarthritis-basics. Accessed 14 Jan 2010. Kellgren JH, Lawrence JS. Atlas of standard radiographs, vol 2: the epidemiology of chronic rheumatism. Oxford: Blackwells Scientific Publications; 1963. Jensen LK, Mikkelsen S, Loft IP, Eenberg W, Bergmann I, Logager V. Radiographic knee osteoarthritis in floorlayers and carpenters. Scand J Work Environ Health. 2000;26:257–62. Sandmark H, Hogstedt C, Vingard E. Primary osteoarthritis of the knee in men and women as a result of lifelong physical load from work. Scand J Work Environ Health. 2000;26(1): 20–5. Vingard E, Alfredsson L, Goldie I, Hogstedt C. Occupation and osteoarthritis of the hip and knee: a register-based cohort study. Int J Epidemiol. 1991;20(4):1025–31. Partridge REH, Duthie JJR. Rheumatism in dockers and civil servants: a comparison of heavy manual and sedentary workers. Ann Rheum Dis. 1968;27:559–68. Anderson JJ, Felson DT. Factors associated with osteoarthritis of the knee in the first national health and nutrition examination survey (hanes I). Am J Epidemiol. 1988;128(1):179–89. Coggon D, Croft P, Kellingray S, Barrett D, McLaren M, Cooper C. Occupational physical activities and osteoarthritis of the knee. Arthritis Rheum. 2000;43(7):1443–9. Cooper C, McAlindon T, Coggon D, Egger P, Dieppe P. Occupational activity and osteoarthritis of the knee. Ann Rheum Dis. 1994;53:90–3. Felson DT, Hannan MT, Naimark A, Berkeley J, Gordon G, Wilson PWF, et al. Occupational physical demands, knee bending, and knee osteoarthritis: results from the framingham study. J Rheumatol. 1991;18(10):1587–92. Lau EC, Cooper C, Lam D, Chan VNH, Tsang KK, Sham A. Factors associated with osteoarthritis of the hip and knee in Hong Kong Chinese: obesity, joint injury and occupational activities. Am J Epidemiol. 2000;152:855–62. BMI—body mass index: BMI for adults [homepage on the Internet]. 2007. http://www.cdc.gov/nccdphp/dnpa/bmi/adult_BMI/ about_adult_BMI.htm. Accessed 27 July 2009. Manninen P, Heliovaara M, Riihimaki H, Suomalainen O. Physical workload and the risk of severe knee osteoarthritis. Scand J Work Environ Health. 2002;28(1):25–32.

J Occup Rehabil (2010) 20:489–501 39. D’Souza JC, Werner RA, Keyserling WM, Gillespie B, Rabourn R, Ulin S, et al. Analysis of the third national health and nutrition examination survey (NHANES III) using expert ratings of job categories. Am J Ind Med. 2008;51:37–46. 40. US Department of Labor. Dictionary of occupational titles (DOT). 4th ed. Washington: US Government Printing Office; 1977. 41. Reid CR, Bush PM, Karwowski W, Durrani SK. Occupational postural activity and lower extremity discomfort: a review. Int J Ind Ergon. 2010;40(3):247–56. 42. Sobti A, Cooper C, Inskip H, Searle S, Coggon D. Occupational physical activity and long-term risk of musculoskeletal symptoms: a national survey of post office pensioners. Am J Ind Med. 1997;32:76–83. 43. D’Souza JC, Franzblau A, Werner RA. Review of epidemiologic studies on occupational factors and lower extremity musculoskeletal and vascular disorders and symptoms. J Occup Rehabil. 2005;15(2):129–65. 44. Euser AM, Zoccali C, Jager KJ, Dekker FW. Cohort studies: prospective versus retrospective. Nephron Clin Pract. 2009; 113(3):c214–7. 45. Noordzij M, Dekker FW, Zoccali C, Jager KJ. Study designs in clinical research. Nephron Clin Pract. 2009;113:c218–21. 46. Jin S, McCulloch R, Mirka GA. Biomechanical evaluation of postures assumed when harvesting from bush crops. Int J Ind Ergon. 2009;39(2):347–52. 47. Air stretcher (AS-1000 rev. 1) [homepage on the Internet]. 2009. http://www.carpetairstretcher.net/index.php?option=com_content& view=article&id=5&Itemid=13. Accessed 14 Apr 2010. 48. Honda—walk assist and mobility devices [homepage on the Internet]. 2009. http://corporate.honda.com/innovation/walk-assist/. Accessed 13 Jan 2010.

501 49. UC berkeley human engineering laboratory [homepage on the Internet]. http://bleex.me.berkeley.edu/bleex.htm. Accessed 13 Jan 2010. 50. HULC [homepage on the Internet]. 2010. http://www.lockheed martin.com/products/hulc/index.html. Accessed 13 Jan 2010. 51. What’s ‘‘HAL’’ (hybrid assistive limbÒ)? [homepage on the Internet]. 2009. http://www.cyberdyne.jp/english/robotsuithal/ index.html. Accessed 13 Jan 2010. 52. Porter W, Mayton A, Moore S. Pressure distribution on the anatomic landmarks of the knee and the effect of kneepads. Appl Ergon. (Forthcoming) 2010. 53. Reid CR, Bush PM, Karwowski W, McMullin DL. In: The need for a lower extremity risk assessment model. Proceedings of the human factors and ergonomics society 53rd annual meeting, San Antonio; 2009. p. 887–91. 54. Chen JC, Dennerlein JT, Shih TS, Chen CJ, Cheng Y, Chang WP, et al. Knee pain and driving duration: a secondary analysis of taxi drivers’ health study. Am J Public Health. 2004;94(4):575–81. 55. Rossignol M, Leclerc A, Allaert FA, Rozenberg S, Valat JP, Avouac B, et al. Primary osteoarthritis of hip, knee, and hand in relation to occupational exposure. Occup Environ Med. 2005;62: 772–7. 56. Office of Population Censuses and Surveys. Standard occupational classification. London: HMSO; 1991. 57. Campbell L, Pannett B, Egger P, Cooper C, Coggon D. Validity of a questionnaire for assessing occupational activities. Am J Ind Med. 1997;31(4):422–6. 58. Croft P, Cooper C, Wickham C, Coggon D. Osteoarthritis of the hip and occupational activity. Scand J Work Environ Health. 1992;18(1):59–63.

123