Background. According to several studies, the autoimmune response may lead to osteoarthritis and dyslipidemia and may affect the homeostasis of the human body’s internal environment and then cause its own immune regulation. Consequently, the risk of osteoarthritis might be increased by dyslipidemia, but this association is not universally acknowledged. Therefore, a systematic review and meta-analysis was conducted to study the relationship between dyslipidemia and the risk of osteoarthritis. Methods. In this study, PubMed, EMBASE, and the ISI Web of Science were used to identify related studies published before July 2018. The relationship between dyslipidemia and the risk of osteoarthritis was evaluated on the basis of relative risk (RR) values and the corresponding 95% confidence intervals (CIs). To further investigate this relationship, we also employed the random effects model proposed by DerSimonian and Laird. Results. A total of nine studies were included to study the effect of dyslipidemia on the risk of osteoarthritis, including four cohort, three case-control, and two cross-sectional studies. Among these studies, six stated data for knee osteoarthritis, two reported on hand osteoarthritis, and one reported on hip osteoarthritis. A total of 53,955 participants were included in the meta-analysis, comprising 22,501 patients with OA (19,733 hand OA, 2,679 knee OA, and 89 hip OA). Based on the meta-analysis of case-control and cross-sectional studies, osteoarthritis was clearly higher in those with dyslipidemia compared to those who did not suffer from dyslipidemia (case-control: ; –1.46; cross-sectional: ; -1.46). In addition, the meta-analysis of cohort studies did not present any relationship between dyslipidemia and OA (; –1.14). Conclusions. Even though our meta-analysis of case-control and cross-sectional studies suggested a strong relationship between dyslipidemia and osteoarthritis; this relationship was not validated by our meta-analysis of only cohort studies. As a result, further investigation needs to be conducted on the relationship between dyslipidemia and osteoarthritis, considering the significant public health relevance of the topic.
1. Introduction
Osteoarthritis (OA) refers to a chronic degenerative disease that involves the cartilage, as well as its surrounding tissues [1]. OA is considered the most common joint disease, and nearly 10-12% of the population suffers from OA [2]. In addition, it is expected that this number will increase dramatically due to the quickly increasing aging population combined with the growing prevalence of obesity [3]. Consequently, osteoarthritis is considered to have a negative influence on the health economy [4]. It can be forecast that by the year 2032, an additional 26,000 per million patients over the age of 45 will present to their general practitioner with osteoarthritis compared to 2012 [5]. OA is associated with age, female gender, obesity, joint injury, and career, as well as a high level of physical activity [5]. In addition, the autoimmune response of the synovium plays an important role in rheumatoid arthritis. In recent years, the immunological pathogenesis of synovium in osteoarthritis has attracted the attention of many researchers. Whether the immune mechanism and inflammatory mediators are involved in the occurrence and development of osteoarthritis deserves further discussion. This may provide a new research idea for the pathogenesis of osteoarthritis, to improve our understanding of the development of this disease and change the way of treatment. In recent years, research has shown that metabolic syndrome is closely associated with OA, which is even a part of generalized metabolic disorder. Metabolic syndrome is composed of a bundle of interrelated metabolic risk factors, including diabetes, obesity, dyslipoproteinemia, and hypertension [6]. Furthermore, the incidence of metabolic syndrome is very high; it has been estimated to be as high as 26.7% in industrialized countries [6]. In the context of musculoskeletal disorders, metabolic syndrome has increasingly gained more attention because of its relationship with knee OA [7]. Obesity, the main feature in metabolic syndrome, is overwhelmingly related to degenerative joint changes in regard to mechanical load [8]. Alternatively, obesity-related OA can afflict nonweight-bearing joints (e.g., the hands), signifying a role of adipokines (circulating mediators released by adipose tissue), such as leptin. Thus, OA may have a systemic metabolic element [9]. In addition, OA can be categorized into three phenotypes: metabolic OA, age-related OA, and injury-related OA [10]. Nevertheless, as one of the components of metabolic syndrome, the role of dyslipidemia in the pathogenesis of OA is not completely understood. Dyslipidemia may affect the homeostasis of the human body’s internal environment and then cause its own immune regulation. Dyslipidemia is related to chronic low-grade inflammation and oxidative stress, likely increasing the development of OA [11, 12]. A survey carried out by Ghandehari concluded that approximately 51.4 million US adults presented with high cholesterol and triglycerides, in addition to 36.1 million with elevated low-density lipoproteins [13]. As a result, we chose to conduct a systematic review and meta-analysis of the published observational studies to better comprehend the relationship between dyslipidemia and the risk of OA.
2. Materials and Methods
This research was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement [14] and the Meta-analysis of Observational Studies in Epidemiology (MOOSE) guidelines [15].
2.1. Data Sources and Search Strategy
Published studies in PubMed, EMBASE, and the Web of Science were searched based on the following keywords: (“hyperlipidaemia” OR “dyslipidemia” OR “triglyceride” OR “cholesterol” OR “lipoprotein” OR “lipid” OR “metabolic syndrome”) and (“OA”). No restrictions on language or the date of publication were placed. Additionally, this study also searched the reference lists. Unpublished studies and original data were not included.
2.2. Eligibility Criteria for Study Selection
The eligibility criteria were as follows: study design (randomized controlled trials and cohort, case-control, or cross-sectional studies); an exposure factor of blood lipid levels and an outcome of OA; availability of the odds ratio (OR)/risk ratio (RR) values and corresponding 95% confidence intervals (CIs) for dyslipidemia patients and the general population; or the availability of sufficient information to measure these variables. The most recent all-inclusive study was searched under the condition that two studies used the same population. The definition of dyslipidemia was in line with the US National Cholesterol Education Program Adult Treatment Panel III guidelines. In accordance with the National Cholesterol Education Program, the definition of dyslipidemia was high-density lipoprotein cholesterol , as well as total cholesterol, low-density lipoprotein cholesterol (LDL-C), and TG levels of ≥200, ≥130, and ≥130 mg/dL, respectively [5]. The definition of osteoarthritis was in line with the American College of Rheumatology (ACR) clinical and clinical plus radiographic criteria [6]. The ACR classification criteria for (OA) permits the categorization of individuals for hand, knee, and hip OA [6]. We strictly abided by this classification standard.
2.3. Data Abstraction and Quality Assessment
Two scholars (J.X. and J.L.) obtained the essential information from the chosen studies according to the standard. The following information was gathered: name of the first author, publication year, country in which the research was carried out, study design, number of participants, period of follow-up, sources of controls, potential adjusted confounding variables, OR/RR values, and 95% CIs.
To date, no available common scale has been proposed to evaluate the quality of all kinds of observational studies. As a result, two authors individually employ the modified Newcastle-Ottawa Scale (NOS) [16] as reported by Zhu et al. [17] to assess the quality of the included studies. Quality types were allocated in accordance with the scores of each study, consisting of high quality (score 7-9), medium quality (score 4-6), and low quality (score less than 4) [18]. The maximum total score could reach 9 points, and discrepancies were solved by mutual agreement.
2.4. Statistical Analysis
The random effects model put forward by DerSimonian and Laird was applied to investigate the relationship between dyslipidemia and the risk of OA among the cohort studies [19]. The statistic was employed to evaluate heterogeneity between the studies. Low, medium, and high heterogeneities were categorized as 25%, 50%, and 75%, respectively [20]. Definite heterogeneity was assumed if the value was less than 0.1. Sensitivity analyses were conducted by altering the pooling model [21]. In addition, a sensitivity analysis was carried out to evaluate the influence of each individual study on the summarized estimate by means of successively excluding one research study at a time. Publication bias was assessed using Begg’s [22] and Egger’s [23] tests. No testing for funnel plot asymmetry was carried out due to the limited number of studies included in the analysis () [24].
Furthermore, we conducted a meta-analysis of the case-control and cross-sectional studies in regard to the influence of dyslipidemia on the risk of OA and expressed the results as pooled risk ratios with 95% CIs with the application of a random effects model.
STATA version 12.0 (Stata) was carried out to perform all statistical analyses.
3. Results
3.1. Study Selection and Study Characteristics
The process of study selection for the meta-analysis can be found in Figure 1. In total, 1,917 articles were obtained through the initial search, and 502 were duplicates. An additional 1,266 studies were removed based on the title and abstract. Eventually, after evaluation of the full texts, eight studies were excluded for the reason that they did not satisfy our inclusion criteria: three studies offered inadequate information [25–27], three studies did not offer ORs or RRs for OA or adequate information to calculate these variables [28–30], and two studies were removed that either did not have dyslipidemia as an exposure or did not have OA as an outcome [31, 32]. Ultimately, nine available observational articles were recognized for our meta-analysis [33–41].