Background
Meniscal lesions are common and knee meniscectomy is the most common procedure performed by orthopedic surgeons [
1]. They are typically categorized as traumatic or non-traumatic based on their etiology. Traumatic meniscal lesions most commonly occur in younger active people and are caused by serious traumatic injury [
2]. Non-traumatic lesions that result from repetitive stresses to the menisci over time often accompany knee osteoarthritis and are more common in middle-aged and older individuals [
3,
4].
Meniscal lesions can lead to unspecific symptoms like pain and swelling accompanied by a locking or catching sensation in the knee [
5]. However, structural damages in particular need not to correlate with the presence of pain [
6] and often (52.1–78.1%) occur without symptoms [
3,
7,
8]. Thus, they are challenging to assess, and incidence might be underreported. In a recent systematic review, Culvenor et al. [
9] investigated the prevalence of meniscal damage in asymptomatic uninjured knees in adults based on magnetic resonance imaging (MRI) findings. The overall pooled prevalence of meniscal tears was 10%, with higher prevalence in individuals ≥ 40 years of age (19%). In this group, medial meniscal tears (14% (95% CI 8–20%)) were statistically significantly more common than lateral meniscal tears (5% (95% CI 2–8%)). The prevalence of meniscal injuries in asymptomatic athletes was even higher with changes of meniscal tissue in 31% [
10].
There are indications that meniscal lesions are associated with the development of knee osteoarthritis [
11,
12]. Total meniscectomy and partial lateral meniscectomy are risk factors for osteoarthritis of the knee [
1]. There is some evidence that meniscus repair is associated with a lower risk for osteoarthritis of the knee compared with partial meniscectomy [
13]. The risk of partial medial meniscectomy compared with conservative treatment for the future risk of osteoarthritis is not known. There is little research about risk factors for meniscal lesions. In a systematic review, Snoeker et al. [
14] identified sex and age to be major risk factors for non-traumatic meniscal lesions and sports participation (playing rugby or football) to be associated with a high risk for acute meniscal tears. Further known risk factors for meniscal lesions are overweight, generalized joint hypermobility and time from anterior cruciate ligament (ACL) injury to reconstruction [
14,
15]. High biomechanical stresses during knee-straining working positions may affect intra- and periarticular knee structures (e.g. cartilage, menisci, cruciate and collateral ligaments, bursae and patella tendon) and can lead to acute or chronic injuries [
16,
17]. Snoeker et al. [
14] indicated that there is also an association between occupational kneeling, squatting and frequent stair climbing and the development of meniscal lesions. However, these findings were based on only a few studies and no dose-response relationship was reported.
Meniscal lesions resulting from extended periods of work in a kneeling or squatting position are part of the European schedule of occupational diseases directly related to occupation [
18]. They were accepted as occupational diseases in several EU member states. However, information regarding the required duration of exposure is rare.
Furthermore, identifying occupational risk factors is important in the development of prevention strategies for meniscal lesions at worksites. Therefore, a systematic review was conducted (1) to summarize the evidence on the potential relationship between occupational risk factors and the development of meniscal lesions, (2) to identify specific occupations or occupational activities at risk of meniscal lesions and (3) to assess whether a positive dose-response relationship is present.
Methods
The study protocol was registered in the International prospective register of systematic reviews (PROSPERO) under record number CRD42020196279 and is available online at
https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42020196279. The systematic review with meta-analysis was performed in accordance with the criteria of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [
19] and the guidelines for conducting and reporting meta-analyses of observational studies in epidemiology (MOOSE) [
20]. The systematic review also meets all criteria of AMSTAR 2 [
21].
Search strategy
We conducted a systematic literature search on Medline (via the Ovid interface), Embase (via the Elsevier interface) and Web of Science until 21th of August 2021 (search update; first search on 28th of February 2020). The research question was specified based on the Population, Intervention (Exposure), Control/Comparison, Outcome (PICO) scheme [
22]. The search strategy combined a broad range of Medical Subject Headings (MeSH) and keywords describing the exposure (knee-loading exposure at work) and the outcome (meniscal lesion) to gain a highly sensitive search. Search terms for the exposure included knee-straining activities, occupations at risk for the development of knee disorders, and occupational determinants that were used in the search filter of Mattioli et al. [
23]. No date or language restrictions were applied. A priori defined key articles [
8,
24‐
27] identified through preliminary search for existing reviews were used to validate the search string. The search strategy was modified for each database accordingly and is described in Additional file
1. In addition, we conducted a manual search on grey literature (e.g. thesis, research reports, unpublished manuscripts) and used the “citation tracking function” by Web of Science to supplement the electronic search. Further, the reference lists of all included studies and related key reviews were reviewed manually.
Eligibility criteria
We searched for epidemiological observational studies on the association between occupational risk factors and meniscal lesions. The following inclusion criteria were applied: (a) the study had a cohort, case-control, case-cohort or cross-sectional design with a response of at least 10%, (b) the study examined working population or retired workers (male and female, 16–75 years old), (c) the exposure was described as work-related knee-loading activities and positions or employments in specific occupational groups with intensive physical activities, (d) general population (16 years and older) or non-exposed workers served as comparison, (e) reported outcome was meniscal lesions diagnosed arthroscopically, by MRI, open surgery, clinical examination, diagnose codes (e.g. ICD-10) or self-reported. For the assessment of prevalence regarding meniscal lesions in specific occupational groups and exposure groups, cross-sectional studies without comparison group were included as well. Studies investigating injury-related meniscal lesions or secondary complaints after osteoarthritis or ACL-injury were excluded.
Study selection
All citations were exported to EndNote X9.1 and duplicates were removed. Two reviewers (CB and UBA) independently screened the titles and abstracts of the remaining studies against the described in- and exclusion criteria. Subsequently, the same two reviewers checked full texts for eligibility. For excluded full text reports, the reasons for exclusion were recorded. Any disagreements during the selection process were resolved by discussion or, if needed, a third reviewer was consulted.
From identified studies the following data were independently extracted by two reviewers (CB and UBA): study characteristics (authors, year of publication, country of origin, study design), study population (setting, sample size, demographics, response), occupational exposure (definition, job title, method used to identify the exposure), outcome (definition, assessment, localization of meniscal damage), and study results (number of participants analyzed, prevalence or incidence of the outcome in exposed and comparison subjects, relative risk measures, data indicating dose relationship, confounders). Discrepancies were resolved through discussion.
Quality assessment
The same two reviewers (CB and UBA) independently assessed the risk of bias for each included study using a modified set of predefined criteria according to Ijaz et al. [
28] and Kuijer et al. [
29] (see Additional file
2). The following items were considered as major domains: (i) recruitment procedure and follow-up, (ii) exposure definition and measurement, (iii) outcome source and validation, (iv) confounding and effect modifications, (v) analysis method (methods to reduce research bias), (vi) chronology. The items (vii) blinding of assessors, (viii) funding and (ix) conflict of interest were considered as minor domains. Each item was categorized as either high risk, low risk or unclear risk of bias. Disagreements were discussed in consensus meetings moderated by the principal investigator. Studies were classified as low risk of bias if all major domains scored low risk. In other cases, studies were considered as high risk of bias.
Data synthesis
Meta-analyses were conducted to pool the results from included studies regarding different occupational exposures as risk factors for the development of meniscal lesions. When available, we used the fully adjusted risk estimates of the individual studies. Unadjusted prevalence ratios were manually calculated if they were not reported in the studies, but the necessary information on frequency distributions was available. Because meniscal lesions are common in the general population [
9] and odds ratios (ORs) tend to overestimate the relative risk when the prevalence of the outcome of interest is high, we converted the ORs to prevalence ratios for studies with a prevalence higher than 10%, according to the methods of Zhang and Yu [
30]. The pooled risk of occupational exposure to meniscal lesions was estimated using a random effects model for the meta-analysis. If at least two primary studies which were comparable in terms of exposure and outcome were included, the meta-analysis was performed. The I
2 value was used as a measure of heterogeneity. The occurrence of publication bias was determined using funnel plots and Egger’s tests, if at least three studies were included in the meta-analysis. Synthesis calculations were conducted using Stata Version 14.2 (StataCorp, College Station, TX).
Data from studies that were not eligible for meta-analysis was summarized qualitatively.
Assessment of evidence
The Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) approach [
31] was used to assess the quality of the total body of evidence, following the example of Hulshof et al. [
32] with modifications [
33,
34]. We considered three levels of quality: high, moderate, and low, with an initial “high” level indicating the presence of randomized studies. However, as only observational studies were included, the starting level was set to “moderate”. The quality of evidence was downgraded based on five factors: quality of study limitations, indirectness, inconsistency, imprecision (range of the CI of studies > 2.0), and publication bias. An upgrade would follow if the study findings had large effect sizes (ES) (an effect estimate > 2.0), a dose-response relationship, and the presence of residual confounding (which would increase confidence in the association). If a pooled ES larger than 5.0 was present, the quality of evidence was upgraded twice.
Discussion
This systematic review evaluated the possible relationship between occupational knee-straining exposures and the development of meniscal lesions. Twenty-two studies met our inclusion criteria of which nine studies were eligible for meta-analysis. Significant associations between occupational risk factors and the development of meniscal lesions were found for kneeling, squatting, climbing stairs, lifting and carrying weights ≥ 10 kg, lifting and carrying weights ≥ 25 kg and specific occupational groups (professional football players, miners and floor layers). The overall quality of evidence according GRADE was moderate for kneeling, squatting, climbing stairs, lifting and carrying weights ≥ 10 kg, playing football and working as a hard coal miner, and low for standing or walking, lifting or carrying weights ≥ 25 kg and working as a floor layer.
The findings of our review are in line with a previous review by Snoeker et al. [
14] that also identified kneeling and squatting, climbing stairs, lifting and carrying weights and playing football to be associated with meniscal lesions. But in contrast, we did not find a statistically significant association in walking > 2 miles per day and standing or walking > 2 h per day, as we used the adjusted risk estimates to reduce bias due to confounding. Reid et al. [
55] concluded that squatting should be considered an occupational risk factor, which is consistent with our findings. Based on a meta-analysis including two studies [
26,
42], we found moderate evidence that working as a hard coal miner is associated with the development of meniscal lesions, as previously suggested by Mc Millan and Nichols [
56].
Due to insufficient data, we did not conduct a dose-response analysis. However, we investigated the association between lifting and carrying weights and the development of meniscal lesions, but no considerable differences were found between individuals exposed to weights ≥ 10 kg (ES 1.63, 95% CI 1.35–1.96) and ≥ 25 kg (ES 1.56, 95% 1.08–2.24). But since exposure is defined not only by intensity, future studies should focus on frequency and duration of lifting and carrying weights. In both case-control studies [
24,
25] the exposure frequency was kept very low with lifting or carrying weights at least ten times a week and information on exposure duration was lacking.
We identified three studies that investigated the localisation of structural changes in former elite footballers’ knee joints, reporting a high prevalence rate of meniscal lesions. In workers exposed to kneeling activities the medial meniscus is more frequently affected, whereas in professional football players meniscal lesions occurred equally in the lateral and medial meniscus. A possible explanation could be the different forces at the knee joint that result from different exposures. While workers exposed to kneeling or squatting activities, e.g. miners or floor layers, use to work in an awkward position that lead to high biomechanical stresses on the joint structures, athletes in high-speed contact sport as football, basketball or handball are exposed to highest intensity of joint impact with twisting and torsional loading. However, professional basketball players did not show the same prevalence of meniscal lesions in lateral meniscus as professional football players, but findings were based on only one study.
We performed an extensive literature review using a comprehensive search string in three databases and even included unpublished studies (grey literature). There were no language or time restrictions to ensure the inclusion of as many relevant studies as possible. A strength of our research methods was that the appraisal of titles, abstracts and full texts, data extraction and the assessment of study quality were carried out independently by two researchers. We used strict selection criteria and studies with no information on the response were excluded due to the potential for selection bias. However, there was a large heterogeneity among the included studies and only nine studies were eligible for meta-analysis, with a low number of studies within each exposure category. Although our results provide evidence of an association between occupational risk factors and the development of meniscal lesions, some limitations of our meta-analysis must be addressed.
The precision of the effect estimators is reduced by the heterogeneity of the exposure definition and measurement in individual studies included in the meta-analysis. The exposure definition varied using either job titles or the description of specific working tasks. We included populations from different occupations and occupational sectors. As even the spectrum of daily exposure within a single job can vary greatly due to different work content, specific characteristics of workplaces and individual preferences of working postures [
57], a large heterogeneity in described exposure can be assumed. Moreover, an adequate reporting of exposure duration, frequency and intensity was lacking in most studies. Information on the exposure was predominantly assessed via self-report using questionnaires or interviews. Although this method of measurement is a low cost and easy way to assess especially retrospective exposures of work shifts decades ago, the validity is low. Ditchen et al. [
58] stated that self-report showed good to acceptable quality in identifying knee postures but mostly poor to very poor quality in quantifying the load. More objectively methods for exposure assessment are workplace observations or video-recordings as used in the study from Kivimäki et al. [
27]. But since only specific working sequences are filmed and the duration of knee-straining postures is extrapolated to an entire work shift, there is a risk of overestimation. The use of task based measurement data in combination with self-reported diary information may be a cost efficient and valid alternative [
57]. Another promising approach for long-term technical measurement of occupational knee-straining activities is the use of wireless accelerometers that provide valid information on kneeling and squatting under laboratory conditions, and for kneeling as well under normal working conditions [
59].
This systematic review only examined the relationship between occupational physical activities and meniscal lesions; other potential risk factors such as knee-straining leisure-time physical activity were not studied. The possible bias induced by confounding factors of the associations examined was reduced by using the fully adjusted risk estimations of the individual studies. However, only three of the nine studies included in the meta-analyses had adjusted for leisure-time physical activities and the possibility of residual confounding cannot be excluded. It is therefore recommended that future studies assess both occupational and leisure-time activities, so that independent relationships to both can be examined.
Moreover, the overall methodological quality in all but one of the included studies was low. Besides the insufficient reporting of exposure as described above, the chronology was the most important domain limiting study quality. Only Mikkelsen et al. [
36] reduced risk of bias from existing meniscal lesions at baseline by excluding participants from the basic cohort with an outcome before first date of employment. There is little research determining the prevalence of meniscal lesions in asymptomatic, unexposed individuals at the beginning of employment. Two studies [
60,
61] that investigated the knee joints of adolescent volunteers (average age < 20 years) not exposed to regular sporting activities did not find any meniscal lesion in participants, whereas Jerosch et al. [
62] reported grade 2 meniscal lesions in 19.4% of unexposed volunteers under the age of 16, according to the classification of Glashow et al. [
63]. However, meniscal lesions that reached the upper or lower articular surface or led to fissuration or fragmentation of the meniscus (grade 3 and 4) were not found. Ludman et al. [
64] described Grad 3 meniscal lesions according to the classification of Stoller et al. [
65] in 11.5% of the posterior horns of the medial meniscus in 26 knees of unexposed volunteers aged 18–23 years. Although in all studies the number of participants was low and selection bias due to unreported response may have existed, these findings indicated that meniscal lesions can be present in asymptomatic unexposed individuals already at the beginning of employment. Thus, chronology was considered important.
These limitations could have affected the results of our review and limit the generalizability of our findings. To prevent recall bias, future studies should use more objective measurements of exposure and provide detailed information on intensity, duration, and frequency of risk factors.
However, our results indicate an association between occupational activities and the development of meniscal lesions. Prevention of knee disorders at work may be beneficial to reduce sickness absence and work-related health care costs. Until today, there is little research about how to prevent work-related knee disorders. Redistributing mechanical loads and to minimize the time spend in kneeling working positions are important strategies to reduce structural knee damages. Porter et al. [
66] indicated that kneepads could decrease the pressure on the bony structures of the knee by distributing the forces across more surface area. However, peak pressures over key anatomic structures of the knee (e.g. bursa sac) have still been reported, and new kneepad designs that redistribute the pressure across a greater surface area are needed. Another approach to prevent discomfort and pain related to occupational squatting is the use of leg support exoskeletons, which reduce worker’s muscle activity around the knee and thus, is expected to reduce the compressive loading at the joint [
67]. In the floor-laying trade, different new methods and tools (e.g. electrical screed levelling machines) have been introduced to carry out many of the job tasks from an upright work position and thus, reducing physical demands at work [
68,
69]. According to Jensen and Friche [
68], workers who had used the new working methods more than 1 year were less likely to report severe knee complaints compared with floor layers who had used the new working methods less than 1 year (knee complaints > 30 days during the previous 12 months: OR 1.49, 95% CI 1.0–2.23; locking of the knee: OR 1.23, 95% CI 0.88–1.71; moderate-to-severe knee pain: OR 1.42, 95% CI 0.93–2.16). But implementation of new methods is difficult and long-lasting, and a participatory approach is recommended.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.