Relationship between obesity and musculoskeletal disorders: systematic review and meta-analysis

Introduction: Obesity is considered a public health problem and it is associated with various musculoskeletal disorders, that haven ́t been reviewed systematically. Therefore, the objective of this study was to describe the relationship between these changes in individuals with obesity. Methods: Lilacs, PubMed, Scielo, Scopus and Web of Science electronic databases were systematically reviewed until July 2018, conducted according to Cochrane guidelines for the development of meta-analysis. Eligibility criteria: inclusion (studies with humans; with a BMI greater than 30 Kg/m2 or above the 85 and 95 percentile for children and adolescents; articles published in Portuguese, English, and Spanish; articles with validated evaluation methods); exclusion (duplicate, non-indexed articles; special populations; meeting summaries, editorials, letters to the editor and reviews). Results: Based on moderate evidence, knee postural changes in patients with obesity were found in two studies (n=158) (OR 14,26, IC 95% 4,98-40,80), I = 0% (p = 0,000), foot postural changes were found in 3 studies (n =338) (OR 3,20, IC 95% 2,084,94), I=0% (p=0,000). Two studies were identified for arthritis (n=1331) (OR 3,38, IC 95% 0,72-1582), I=38,95% (p=0,122). Pain was reported in the lower limb and the spine. Four studies were found for the former (n=1560) (OR 5,51 IC 95% 2,07-14,61), I=81,02% (p=0,001), and three for the latter (n=499) (OR 9,78 IC 95% 2,59-36,93), I=86,58% (p=0,000). Conclusion: Results show that obesity has a negative impact on musculoskeletal health, with results statistically significant for knee and foot postural changes, arthritis, and pain in the spine and lower limbs.


Introduction
Obesity represents a global challenge due to the growing number of individuals with this disease (Aeseth, et al., 2019). Its incidence has increased in recent decades and obesity has become a major public health problem (Calenzani, et al., 2017).
Obesity is considered not only the oldest metabolic disorder, but also the most important nutritional disorder in developed countries presently. According to the 2015 World Health Organization (WHO) report, approximately 52% of the individuals aged 18 years and over are already overweight (Shan, et al., 2019).
The prevalence of obesity is equivalent a cross races and sexes, especially in the age group between 15 and 44 years.
It is believed that 10% of the Brazilian population is overweight (Mello;Luft;Meyer, 2004).
In addition to the metabolic complications present in obesity, other orthopedic problems can also be observed, which are less studied (Brandelize & Leite, 2010). However, the increase in the Body Mass Index (BMI) has been identified and considered as a risk factor for the presence of musculoskeletal symptoms (Onyemaechi, et al., 2016).
Disorders in the locomotor system can trigger not only pain, discomfort and postural changes, but also degenerative diseases, depending on the stage of bone maturation in which the individual is at (Enes & Slater, 2010). Excess weight is the main cause for musculoskeletal discomfort, directly interfering with the structures responsible for joint movements (Rasia, et al., 2007).
These changes, evident in overweight individuals, can affect joint mobility, contributing to physical incapacities and reducing the quality of movements, and consequently generating a considerable economic burden for the health care system (Brach, et al., 2011;Lehnert, et al., 2013). It is noteworthy that musculoskeletal disorders are the main causes of disability and pain worldwide, especially in industrialized countries (Malik, et al., 2018).
Despite the limited scientific evidence available on this topic, the review of the scientific literature enables the observation of several changes in the musculoskeletal system at different stages of bone growth, demonstrating the importance of early therapeutic interventions based on lifestyle changes and physical activity (Batista, et al., 2003).
As the number of publications on this subject is limited, there is a need for scientific studies that address this topic.

Methods
This study was a systematic review of the literature, conducted according to the Cochrane guidelines for the development of meta-analysis. The review procedures were guided by the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) and the review protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) under the reference number: CRD42018102078.

Data Source and Search Strategies
The search was conducted using Lilacs, Scielo, PubMed, Scopus and Web of Science electronic databases, from September 2017 to July 2018, updated in February 2020and the search strategies were defined by two researchers. The strategy PICO was used in this study. PICO represents an acronym for Patient, Intervention, Comparison and Outcome.
The National of Medicine publishes the Medical Subject Headings Sections (MeSH) in order to index articles in PubMed -MEDLINE. Bireme created DeCS (BIREME health sciences descriptors) from MeSH. The search headings used and combined were: "postural change", musculoskeletal disorders", orthopaedic alterations", obesity", "overweight" (Table 1).

Eligibility Criteria
For the selection of articles, inclusion and exclusion criteria were established. The inclusion criteria were: a) studies with humans, without any restrictions for sample size, b) sample with a BMI greater than 30 Kg/m2 for adults and equal to or greater than the 85 th and 95 th percentile for age and sex respectively for children and adolescents, c) articles published in Portuguese, English and Spanish, d) articles with observational design, e) articles that used valid methods of assessment. The exclusion criteria were: a) duplicated articles, b) non-indexed articles, c) special populations, such as those with movement disorders or pain, d) meeting summaries, editorials, letter to the editor, and reviews. Functionality assessments of the locomotor system due to musculoskeletal disorders were beyond the scope of this review and were also not included.

Study Selection
In the first step selection, potential studies identified through the database search were inserted in the web-based systematic review software package "Covidence" (Veritas Health Innovation), which detected duplicate articles. After duplicates were excluded from the database, titles and abstracts were examined in order to identify potential articles for inclusion. In the second step, the full article text was examined for final inclusion/exclusion decision.
All stages of the selection process were conducted by two independent researchers (LMF/TK), and the few discrepancies were decided by consensus.

Data Extraction
Data were extracted in a standardized manner, according to author and year of publication, type of study, journal where it was published, title, language and purpose of the study. The extraction continued identifying the sample size, and the predictor variables (age, gender, race, and mean BMI). Subsequently, the assessment methods and the outcomes were obtained.

Quality Assessment and Evidence Synthesis
The methodological quality of the included studies was analysed through STROBE -The Strengthening Reporting of Observational Studies in Epidemiology. A 22-item instrument, pertaining to title and abstract (item 1), introduction (items 2 and 3), methods (4 to 12), results (13 to 17), discussion (18 to 21) and other information on funding (item 22) (Elm, et al., 2017).
In order to assess study evidence, the adapted Levels of Evidence proposed by the Oxford Center for Evidence-Based Medicine was used. The Levels of Evidence are directly related to the research design, classifying and validating the study according to the focus: diagnostic, etiological, prognostic or therapeutic.

Statistical Analysis
The coefficient was conducted using Comprehensive Meta-Analysis 3.0 software. For the coefficient, the 95% confidence interval odds ratio of each study were combined in order to estimate the effect of obesity on musculoskeletal disorders. The sample size and the occurrence of results were extracted from the data extraction form elaborated during this study.
The Mantel-Haenszel method was used. The technique assumes the effect of the randomized effect model, with a 95% CI to perform the coefficient. Data heterogeneity was assessed using the inconsistency indicator. The Kappa coefficient was utilized in order to assess the interrater reliability.

Results
The database search identified a total of 5782 articles, of which 2012 were duplicates. The abstract titles of the remaining articles were independently selected by the two reviewers, including 789 articles that required a full-text screening.
After screening, 16 observational articles that met the inclusion criteria were considered and included in this review, and seven were included in the meta-analysis (Figure 1).

Musculoskeletal disorders in patients with obesity
Seven studies were included in the meta-analysis. There was a statistically significant association between obesity and postural changes in the knee and foot, as well as pain in the spine and lower limbs.
As for level of evidence, 1 study presents moderate evidence (consistent level 2 or 3 studies or extrapolations from level 1 studies) and 6 studies (85,1%) present weak evidence (level 4 or extrapolations from level 2 or 3 studies).

Discussion
The purpose of this study was to examine the relationship between obesity and musculoskeletal disorders of the locomotor system. It was found that these changes manifest at different stages of bone maturation, in both sexes.
The association between these disorders and obesity is not yet clear. It is hypothesized that excess weight generates an overload which may interfere with bone remodelling, manifesting as deformities in the muscular and articular systems (Janini, et al., 2011).
One study observed that children with excess weight or obesity, regardless of the age group, report more musculoskeletal symptoms compared to normal weight children, and these changes commonly occur in the lower limbs, especially the ankle and the foot (Krul, el al., 2009).
Similarly, the prevalence of orthopaedic conditions in normal weight and overweight children was quantified using paediatric clinical records. Results showed that the most common self-reported joint complain in overweight children was knee pain, compared to the normal weight children (Taylor, et al., 2006).
Some studies show a higher prevalence of musculoskeletal pain in adolescents with obesity, especially in the lower limbs and lumbar spine compared to adolescents of normal weight (Stovitz, et al., 2008). Orthopedic changes was compared in overweight children and adolescents, as well as children and adolescents with obesity and found that this group has more genu valgum, recurvatum and quadriceps contracture compared to eutrophic children and adolescents (Janini et al., 2011).
Foot morphology of 2887 children between the ages of 2 and 14 years was compared. It was found that normal weight children have approximately equal distribution in all foot types for almost the entire childhood. Slender, robust and flat feet were more common in overweight children (Much, et al., 2008).
In the adult population with obesity, pain is more common in the lower limbs and the lumbar spine. One of the most common complaints reported among the adult population is chronic low back pain, leading to absence from work, decrease in physical function and disability (Malta, et al., 2017).
Similarly, another study examined the correlation between obesity and foot type among those with chronic heel pain in a sample of 33 men and 47 women. It was found that in the chronic heel pain group, BMI was higher and foot was more pronated compared to the group without chronic heel pain (Irving et al., 2007).
The relationship between obesity and foot pain in a sample of 136 individuals aged 25 to 62 years was investigated using the Manchester Foot Pain and Disability Index (MFPDI). Results showed that increasing BMI was associated with foot pain and disability (Vismara, et al., 2010).
Among the musculoskeletal changes that commonly occur in patients with obesity are spinal impairment, particularly disc degeneration and functional impairment of the joints (Adebipme, 2019).
A study examined the effect of obesity and low back pain on the spine, in a cross-sectional design including 13 individuals with obesity, 13 individuals with obesity and low back pain, and 11 healthy individuals. It was found that obesity was associated with reduced range of motion of the spine due to a reduced mobility at the pelvic and thoracic levels, and increased lumbar lordosis (Karlson et al., 2003). Overweight negatively affects common daily movements such as rising, lateral trunk inclination, lateral flexion, as well as walking (Bliddal, et al., 2016).
Risk factors for osteoarthritis, including age and obesity were examined. Results showed that higher body mass index and advanced age increased significantly the risk for total hip arthroplasty due to osteoarthritis (Tanamas et al., 2012).
It is important to detect these changes at an early age, thus children may be treated in a specialized manner, including the practice of guided physical activity. As a consequence, they may improve their pre-existing physical condition and may prevent future complications. Physical activity is considered an important primary prevention initiative to prevent obesity (Aaseath, et al., 2019).

Limitations of the Study
Limitations of this study included the predictive characteristics of the sample, regarding mean BMI values, which provided results for obese samples. However, these values were not accredited according to the World Health Organization (WHO). Therefore, they were not eligible for inclusion, since this study compared results of obese samples with control groups comprised of overweight or eutrophic individuals. Several studies had a cross-sectional design, therefore results from this study are based on hypothesis, due to absence of longitudinal studies.