Zhihao Diao
Danyang Guo2†- 1School of Acupuncture and Tuina, Shandong University of Traditional Chinese Medicine, Jinan, China
- 2The First Clinical Medical College, Shandong University of Traditional Chinese Medicine, Jinan, China
- 3Complutense University of Madrid, Madrid, Spain
Background: While several risk factors for knee osteoarthritis (KOA) have been recognized, the pathogenesis of KOA and the causal relationship between modifiable risk factors and KOA in genetic epidemiology remain unclear. This study aimed to determine the causal relationship between KOA and its risk factors.
Methods: Data were obtained from published Genome-Wide Association study (GWAS) databases. A two-sample Mendelian randomization (MR) analysis was performed with genetic variants associated with risk factors as instrumental variables and KOA as outcome. First, inverse variance weighting was used as the main MR analysis method, and then a series of sensitivity analyses were conducted to comprehensively evaluate the causal relationship between them.
Results: Univariate forward MR analysis revealed that genetically predicted hypothyroidism, hyperthyroidism/thyrotoxicosis, educational level, income level, metabolic syndrome (MS), essential hypertension, height, hot drink temperature, diet (abstaining from sugar-sweetened or wheat products), and psychological and psychiatric disorders (stress, depression, and anxiety) were causally associated with KOA. Reverse MR exhibits a causal association between KOA and educational attainment. Multivariate MR analysis adjusted for the inclusion of potential mediators, such as body mass index (BMI), smoking, alcohol consumption, and sex, exhibited some variation in causal effects. However, hyperthyroidism/thyrotoxicosis had a significant causal effect on KOA, and there was good evidence that height, hypothyroidism, educational level, psychological and psychiatric disorders (stress, depression, and anxiety), and abstaining from wheat products had an independent causal relationship. The mediating effect of BMI as a mediator was also identified.
Conclusion: This study used MR to validate the causal relationship between KOA and its risk factors, providing new insights for preventing and treating KOA in clinical practice and for developing public health policies.
1 Introduction
Knee osteoarthritis (KOA) is a common type of osteoarthritis (1), and its clinical manifestations include chronic knee pain, limited activity, and dysfunction (2). KOA affects approximately 16% of the population worldwide, and the number of new cases of KOA reached 86.7 million in 2020 (3, 4). Therefore, KOA significantly contributes to disability (5), imposing substantial social and economic burdens and presenting a significant challenge to global public health (6, 7). The pathogenesis of knee osteoarthritis remains unclear. Previous observational studies have found increasing evidence that risk factors such as educational level, economic level (8, 9), metabolic syndrome (MS), essential hypertension (10), thyroid dysfunction (11), diet (12), hot drink temperature (13), height (14), and psychological and mental diseases (such as stress, depression, and anxiety) (15) are associated with the pathogenesis of KOA. Despite this relationship, the causal association obtained from observational studies of traditional epidemiology may be challenging and cannot be used as reliable evidence because residual confounding factors and reverse causality may lead to bias in traditional observational studies (16). Concurrently, large-sample randomized controlled studies are expensive and time-consuming, making them difficult to use in practice.
Mendelian randomization (MR) is an emerging method for inferring causal associations in genetic epidemiology that provides reliable evidence for inferring causal effects between risk factors and outcomes using genetic variants from genome-wide association studies (GWAS) as instrumental variables (IV) (17). Because human genetic variation is characterized by random allocation, irreversibility, and fixity, MR can be used to effectively avoid the influence of confounding factors and reverse causality (18). Thus, univariate MR could be used to estimate the direct causal association between each risk factor and KOA. Multivariate MR is an extension of MR and has great advantages in avoiding unobserved confounding factors and collider biases. It can assess direct effects even when a single nucleotide polymorphism (SNP) is associated with multiple exposures. Multivariate MR can also be used to infer the potentially causal relationship between risk factors and KOA and the effect of mediating factors (19–21). Therefore, this study aimed to investigate the relationship between KOA and thyroid dysfunction, educational level, economic income, MS, essential hypertension, diet, height, hot drink temperature, and psychological and mental disorders (tension, depression, and anxiety) using MR, from the perspective of genetic inheritance.
2 Methods
2.1 Study design
This study follows the Strengthening the Reporting of Observational Studies in Epidemiology using MR (STROBE-MR) reporting guidelines (22, 23) (Supplementary Table S1). In this study, genetic variants significantly associated with exposures were selected as IV to infer causal relationships between exposures and outcomes, and we chose SNPs significantly associated with each risk factor as IVs for the MR analyses. To ensure the validity of IVs, we needed to satisfy the three assumptions of relevance, independence, and exclusivity for MR. Firstly, the genetic variants should be directly associated with the exposures in question. Second, genetic variants should be uncorrelated with any of the other confounders. Third, genetic variants should not be directly related to the outcome (KOA), but should only influence the outcome through exposure (16). In this study, SNPs were obtained from published GWAS data, and causal and sensitivity analyses were performed. Figure 1 shows a schematic of this process.
Figure 1. Schematic diagram of Mendel randomization research design process. (A) Flow chart of Mendelian randomization. (B) Univariate two-sample Mendelian randomization analysis of risk factors and KOA. (C) Multivariate Mendelian randomization analysis of risk factors and KOA. SNP, single nucleotide polymorphism; KOA, knee osteoarthritis; MR, Mendelian randomization.
All data in this study were obtained from published data, and ethical approval for each study was obtained from the appropriate ethics committee. The current study was a secondary analysis of the data and did not therefore require ethical approval. In this study, three types of analyses–namely, two-sample bidirectional MR, multivariate MR, and mediation analyses–were conducted as research methods.
2.2 Data sources
All GWAS data used in this study are available from the IEU OpenGWAS project1 and FinnGen alliance.2 The pooled KOA data used in this study were obtained from a large GWAS conducted by Tachmazidou et al. (24), which included 24,955 cases and 378,169 controls from England, Scotland, Wales, and Northern Ireland. Additionally, meta-analysis was performed using GWAS data from the United Kingdom Biobank and arcOGEN database. Based on a comprehensive consideration of previous literature review and clinical observation, the risk factors for KOA included in this study were hyperthyroidism, hypothyroidism, economic income, diet (abstaining from sugar-sweetened or wheat products), height, essential hypertension, educational level, hot drink temperature, psychological and mental disorders (stress, depression, and anxiety), and MS. Educational level was included in both subtypes. Diseases associated with these risk factors were diagnosed according to the International Classification of Diseases-10th Revision criteria. Hot drink temperature was derived from a questionnaire containing the following main question: ‘What hot drinks do you prefer (e.g., coffee or tea)?’, with the response options being ‘very hot’, ‘hot’, ‘mildly hot’, and ‘non-hot’ drinks (13). Additionally, genetic variation data on risk factors were provided by the FinnGen Consortium, Medical Research Council Integrative Epidemiology Unit, United Kingdom Biobank, Neale Lab, and Science Genetic Association Consortium (Supplementary Table S2). The present study also included potentially relevant mediators such as BMI, alcohol consumption, smoking, and sex (25–27), with BMI and smoking coming from the United Kingdom Biobank’s big data analysis and with alcohol consumption from the GWAS and Sequencing Consortium of Alcohol and Nicotine use. Pooled data on sex were from the combined analysis data of five different cohorts (23andMe, United Kingdom Biobank, iPSYCH, FinnGen, and Biobank Japan), totalling 3,309,398 samples by Pirastu et al. (28).
2.3 Genetic instrument selection
Independent SNPs with thresholds smaller than the genome-wide significance threshold (5 × 10−8) were selected as IV to ensure the authenticity and reliability of the causal relationship between risk factors and KOA. Simultaneously, to avoid bias caused by linkage disequilibrium among IVs, we used the TwoSampleMR package in R to remove linkage disequilibrium (R2 < 0.001 and clumping distance = 10,000 kb) (29), resulting in a total of 2039 SNPs. The F statistic was employed to evaluate the effects of weak instrumental variables. When F > 10, there is no bias caused by the influence of weak IV (30). The F values of the SNPs in this study were > 29.69, providing sufficient evidence that the strong association will not introduce bias. SNP summary data for risk factors are presented in Supplementary Table S3. To ensure that the SNP effects on exposure and outcome correspond to the same alleles and to avoid any distortions in strand orientation or allele coding (31), we excluded SNPs with incompatible alleles and those exhibiting palindromic structures with intermediate allele frequencies. Finally, 14 incompatible alleles and 57 SNPs with palindromic structures were excluded (Supplementary Table S4).
2.4 Statistical analysis
2.4.1 Univariate and multivariate MR analyses
A univariate two-sample bidirectional MR analysis was performed to clarify the direct genetic causality between KOA and its risk factors. In this study, the Wald ratio method was first used to test the causal effect size of each IV on KOA and provide evidence for the identified association (32). Multiple independent and valid IVs were included, and inverse variance weighting (IVW) was used as the main analysis method to test causality. Additionally, to prevent the influence of unknown confounding factors and ensure the reliability of the results, four supplementary analysis methods—weighted median estimation, MR-Egger regression, simple mode, and weighted mode—were used for verification (33). The premise of IVW analysis is that there is no pleiotropy in the IV. Fixed effect IVW is usually used when all included SNPs are valid, and random effect IVW is used only when there is obvious heterogeneity among the SNPs (34–36). When the instrumental strength meets the requirement of independence from the direct effect, the MR-Egger regression test can provide stronger proof for the causal estimate. Additionally, when the intercept of the MR-Egger regression is infinitely close to zero, the result is infinitely close to the IVW (37). Even if up to 50% of the IVs are invalid, the weighted median estimation can still provide robust analytical results (38). Simple and weighted modes ensure that when some IVs are invalid, the results will not be distorted by the influence of bias (39). To further assess the direct causality, mediating effects, and potential horizontal multidirectionality between KOA and thyroid dysfunction, income, educational level, diet, mental disorders, height, essential hypertension, multiple sclerosis, and other risk factors, all factors (including the four mediators of BMI, smoking, drinking, and sex) were incorporated in the same model, and multivariate MR analysis was conducted based on linear weighted regression IVW and MR-Egger methods to jointly estimate their causal effects on the risk of KOA. However, because more exposures are likely to cause collinearity, the mv-lasso function was applied to remove unnecessary exposures and correct the results. Additionally, the formula was used to determine whether a weak instrument bias influenced the F-value measurement.
2.4.2 Mediation analysis
Multivariate MR results indicated that BMI might be a mediator; hence, mediation analysis was performed using R software in order to further evaluate the mediating effect and mediator share of BMI in the causal relationship between KOA and the risk factors. First, two-sample MR was conducted to ensure that there was a causal relationship between exposure and BMI and between BMI and KOA, and the effect value between exposure and BMI (BetaXZ) was then calculated. Second, multivariate MR was used to examine the effect of BMI as a potential mediator on KOA (BetaZY) and the adjusted effect value between exposure and KOA (BetaXY1). Finally, univariate MR was performed to evaluate the causal effect between exposure and KOA (BetaXY) and to calculate the mediating effect (BetaXZ×BetaZY) and the share of the mediating effect (BetaXZ×BetaZY/BetaXY). Additionally, the coefficient product test was used to calculate the mediating effect and its confidence interval (CI).
2.4.3 Sensitivity analyses
Sensitivity analyses can be used to test whether the causal associations from MR analyses are robust. First, heterogeneity among the IVs was assessed by calculating Cochran’s Q value; if significant heterogeneity existed, a random-effects model was used (29). Second, MR-Egger regression was performed, and the p value of its intercept was used to test horizontal pleiotropy. Leave-one-out tests were performed to assess whether causal associations were driven by a specific SNP (40). Finally, using MR-PRESSO detection, we re-examined whether level pleiotropy existed, removed the outliers, and corrected the causal effect values (41). We also used R language to visualize the MR analysis results, including scatter, forest, and funnel plots and sensitivity analysis. All statistical analyses were performed using the ‘TwoSampleMR’, ‘MR-PRESSO’, ‘MendelianRandomization’, and ‘MVMR’ packages of R software version 4.2.2 (The R Foundation for Statistical Computing, Vienna, Austria), with two-sided p values <0.05 being considered statistically significant.
3 Results
3.1 Causal effects of risk factors and KOA
Univariate MR analysis revealed that genetically predicted hypothyroidism, hyperthyroidism/thyrotoxicosis, economic income, diet (abstaining from sugar-sweetened or wheat products), height, essential hypertension, educational level, hot drink temperature, mental disorders (stress, depression, and anxiety), and MS were causally associated with KOA (Supplementary Figure S1). The results indicated that hypothyroidism (odds ratio (OR): 5.56, 95% CI: 1.22–25.32, p = 0.026), hyperthyroidism/thyrotoxicosis (OR: 711.17, 95% CI: 49.38–10242.78, p = 1.40E-06), abstaining from wheat products (OR: 23.89, 95% CI: 5.13–111.22, p = 5.256E-05), abstaining from sugar-sweetened products (OR: 6.95, 95% CI: 1.83–26.42, p = 0.004), height (OR: 1.09, 95% CI: 1.03–1.17, p = 0.006), primary hypertension (OR: 2.11, 95% CI: 1.11–4.02, p = 0.023), and MS (OR: 1.13, 95% CI: 1.04–1.22, p = 0.003) increased the risk of KOA, whereas income (OR: 0.69, 95% CI: 0.58–0.82, p = 4.062E-05), educational level (OR: 0.52, 95% CI: 0.38–0.71, p = 2.829E-5), years of schooling(OR: 0.58, 95% CI: 0.52–0.64, p = 3.254E-24), hot drink temperature (OR: 0.55, 95% CI: 0.36–0.85, p = 0.007), and psychiatric disorders such as stress, depression, and anxiety (OR: 0.37, 95% CI: 0.003–0.47, p = 0.011) reduced the risk of KOA (Figure 2; Supplementary Table S5; Supplementary Figure S2). In the reverse MR analysis, KOA was only associated with educational level (OR: 0.96, 95% CI: 0.93–0.99, p = 0.011), and no significant evidence was obtained for the causal relationship or association of KOA with hypothyroidism (p = 0.362), hyperthyroidism/thyrotoxicosis (p = 0.662), income (p = 0.171), abstinence from sugar-sweetened products (p = 0.228), abstaining from wheat products (p = 0.651), essential hypertension (p = 0.533), MS (p = 0.675), height (p = 0.148), hot drink temperature (p = 0.823), and mental disorders (stress, depression, and anxiety) (p = 0.447).
Figure 2. Univariate Mendelian randomization results for the causal relationship between KOA and its risk factors. The blue diamond and error line indicate the odds ratio and 95% confidence interval, respectively. OR, odds ratio.
3.2 Multivariate MR analysis
In the multivariate MR model, owing to the potential collinearity problem, we used the mv-lasso function to perform a lasso test (collinearity correction) to correct outliers (34). After incorporating the mediators of BMI, smoking, alcohol consumption, and sex, the adjusted results indicated that hyperthyroidism/thyrotoxicosis (OR: 39.18, 95% CI: 1.85–825.95) had a significant causal relationship with KOA. Sufficient evidence also suggested that hypothyroidism (OR: 10.37, 95% CI: 2.19–48.97), abstaining from wheat products (OR: 18.18, 95% CI: 3.35–98.45), mental disorders (stress, depression, and anxiety) (OR: 3.71, 95% CI: 1.01–13.75), height (OR: 1.22, 95% CI: 1.14–1.31), and educational level (OR: 0.58, 95% CI: 0.42–0.80; OR: 0.67, 95% CI: 0.58–0.78) were casually associated with KOA, whereas essential hypertension (OR: 1.89, 95% CI: 0.92–3.88), MS (OR: 0.99, 95% CI: 0.77–1.27), income (OR: 0.82, 95% CI: 0.65–1.02), hot drink temperature (OR: 0.71, 95% CI: 0.46–1.07), and abstaining from sugar-sweetened products (OR: 1.7, 95% CI: 0.65–4.42) were not statistically significant. Therefore, there was no evidence of a direct effect on the incidence of KOA (Figure 3). In addition, we found in a multivariate MR study that adjusted for smoking, alcohol consumption, and sex were not mediators between risk factors and KOA, and that the true mediator of significance was BMI.
Figure 3. Multivariate Mendelian randomization results for the direct causal relationship between KOA and its risk factors. The blue diamond and error line indicate the odds ratio and 95% confidence interval, respectively. OR, odds ratio.
3.3 Sensitivity analyses
The F-statistic for the genetic instrument ranged from 29.69 to 1976.82 (Supplementary Table S3). Therefore, there was sufficient evidence to suggest that a weak instrument bias was unlikely. Heterogeneity tests revealed potential heterogeneity in the causal effect estimates between KOA and hypothyroidism, abstaining from wheat and sugar-sweetened products, height, essential hypertension, educational level, hot drink temperature, MS, and psychiatric disorders (Supplementary Figure S3). A random-effects model was used to estimate the MR effect size, and the results suggested a causal relationship (p < 0.05). Based on the evidence of the pleiotropic effect of essential hypertension on KOA, determined by means of the MR-Egger’s intercept test, we repeated the MR-PRESSO test after deleting the outliers in the IV, and no significant pleiotropic effect was observed (Supplementary Table S6). Finally, leave-one-out analysis showed that the causal effect between risk factors and KOA was robust (Supplementary Figure S4).
3.4 BMI mediated The genetic predictive effect of risk factors On KOA
Mediation analysis revealed that genetically predicted BMI mediated income (mediation effect = −0.2736; 95% CI: −0.3138, −0.2334; mediated proportion = 73.99%), abstaining from wheat products (mediation effect = 0.9789; 95% CI: 0.6204, 1.3374; mediated proportion = 30.85%), abstinence from sugar-sweetened products (mediation effect = 1.1101; 95% CI: 0.5556, 1.6646; mediated proportion = 57.27%), height (mediation effect = −0.0717; 95% CI: −0.1142, −0.0292; mediated proportion = 99.18%), educational level (mediation effect = −0.2774; 95% CI: −0.3172, −0.2376; mediated proportion = 42.66%; mediation effect = −0.2442; 95% CI: −0.2755, −0.2129; mediated proportion = 44.18%), mental disorders (nerves, anxiety, tension, or depression) (mediation effect = −0.675; 95% CI: −1.0436, −0.3064; mediated proportion = 20.47%), and KOA.(Table 1).
Table 1. BMI mediated the genetic predictive effect of risk factors on KOA.
4 Discussion
Currently, the prevention, diagnosis, and treatment of KOA are major challenges for the public health system. In this study, based on large GWAS data, we used an MR analysis system to verify the causal association between 12 KOA-related risk factors and KOA and to determine whether this relationship could be trusted. Single-sample MR analysis based on genetic predictions revealed that hypothyroidism, hyperthyroidism/thyrotoxicosis, income, educational level, height, hot drink temperature, diet (abstaining from wheat and sugar-sweetened products), psychiatric disorders (stress, depression, and anxiety), and KOA risk were causal factors. MVMR analyses showed that the causal effects somewhat changed after the inclusion of BMI, sex, smoking, and alcohol consumption and that hyperthyroidism/thyrotoxicosis, essential hypertension, educational level, abstaining from wheat products, and income remained to be significantly causally associated with KOA and could directly influence KOA, whereas height, hot drink temperature, hypothyroidism, abstinence from sugar-sweetened products, depression, and other psychological factors were not directly causally associated with KOA. A mediating factor, BMI, was also identified.
4.1 Thyroid function and KOA
Few observational studies have examined the association between thyroid function and the incidence of KOA. In the Framingham Osteoarthritis Study of 1996, involving 798 women and 577 men, no association was observed between serum thyroid-stimulating hormone (TSH) concentration and KOA. Therefore, there was no evidence that thyroid function was associated with KOA. However, this study only considered serum TSH concentration over a certain period and ignored the development and changes in TSH levels (42, 43). Our MR study made important adjustments, and the results showed that there was a significant positive correlation between hyperthyroidism or hypothyroidism and KOA. A recent retrospective study of 109 patients with thyroid dysfunction who underwent musculoskeletal ultrasound (MSUS) examination and clinical evaluation showed that the knee joint effusion rate, overall MSUS severity score, Visual Analog Scale score, and abnormal frequency of MSUS were significantly higher in patients with thyroid dysfunction, and the imaging examination results were consistent with the clinical evaluation. Therefore, hypothyroidism and hyperthyroidism are causally associated with KOA and could increase the risk of KOA (11). The potential mechanism may be that TSH affects the synthesis of hyaluronic acid and proteoglycans, and increases the viscosity of the knee synovial fluid, leading to KOA-related symptoms (44, 45). Another prospective cohort study also demonstrated that hypothyroidism could cause knee joint degeneration through chondrocalcinosis (46), which is consistent with the results of our MVMR analysis.
4.2 Hypertension, MS, and KOA
Hypertension is an important risk factor for KOA (47, 48), and some recent studies have identified MS as a new risk factor for KOA, which has increasingly gained attention (49, 50). Univariate MR results indicated a causal relationship between MS, hypertension, and KOA; in contrast, multivariate MR analysis showed no statistical significance, which was a somewhat surprising result. We randomly performed mediation analysis and found that BMI mediated the association between MS, essential hypertension, and KOA. A large meta-analysis of four databases–EMBASE, PubMed, Cochrane Library, and MEDLINE–and conference materials, including 1,609 articles, showed that KOA was positively correlated with MS (OR: 1.41, 95% CI: 1.16–1.73) and hypertension (OR: 1.70, 95% CI: 1.411–2.052) in radiology studies and with MS (OR: 1.17, 95% CI: 1.03–1.33) and hypertension (OR: 1.32, 95% CI: 1.18–1.47) in symptomatic studies (10). A previous meta-analysis combined with MR analysis highlighted that hypertension increased the incidence of KOA by 62% and that the association between hypertension and KOA persisted even when the association strength decreased from 3.06 to 1.42 after adjustment for body mass index (BMI) (51). The underlying mechanism may be that hypertension may cause intraosseous hypertension, leading to arterial and venous blockage, reduced bone blood flow, subchondral bone ischaemia, and bone cell apoptosis. Therefore, osteoclasts mediate bone resorption, destroy the mechanical support of the covering cartilage, interfere with the exchange of gasses and metabolites in the bone-cartilage functional unit, and thus induce KOA (52, 53). Concurrently, hypertension can also damage vascular endothelial cells and promote the secretion of prostaglandins, which leads to joint inflammation and cartilage damage (54). A cohort study based on baseline data and controlling for covariates such as sex, race, and BMI not only demonstrated a causal relationship between hypertension and KOA but also demonstrated an association between hypertension and pain in patients with KOA (55). MS is a series of diseases caused by abnormal human metabolism, including central obesity, dyslipidaemia, and insulin resistance (56). However, the mechanisms underlying the relationship between MS and KOA remain unclear. Inflammatory mechanisms, oxidative stress, the accumulation of advanced glycation end products (AGE), and ectopic lipid deposition in chondrocytes caused by abnormal lipid metabolism can cause KOA (57). Recent studies have shown that macrophages play a key role in this process. On the one hand, MS can promote the polarization of M1 macrophages by acting on AGE and free fatty acids (FFAs) of macrophages, and AGE can increase the transcription of interleukin (IL)-1β by regulating the NF-κB pathway. FFAs bind to toll-like receptor 4 to release pro-inflammatory factors (58, 59), whereas the adipokine leptin activates the JAK2-STAT3 and PI3K-AKT–mTOR pathways in macrophages, promoting a pro-inflammatory phenotype through the secretion of tumor necrosis factor (TNF)-α and IL-1β (60, 61), resulting in chronic inflammatory hyporesponse and cartilage destruction or deformation. On the other hand, MS can increase chondrocyte degradation through AMPK activity inhibition, whereas excessive mTOR activation inhibits chondrocyte autophagy, preventing self-repair and causing knee chondrocyte damage and KOA (62–64).
4.3 Educational attainment, income, and KOA
With respect to the social economy, educational level and income are closely related to KOA onset (9). This is supported by the results of the univariate MR analysis in the present study, which indicated that educational level and income were negatively associated with KOA onset. Nevertheless, no independent effect of income on KOA was observed in multivariate MR, which may be mediated by BMI. A study based on data from the Korean National Health and Nutrition Examination Survey (KNHANES V) found that the risk of KOA was 1.5 times higher for people with low income than for those with high income, and 2.6 times higher for those with a low educational level than for those with college or higher education. These findings did not vary based on sex. The results remained significant after adjusting for confounding factors such as age and BMI, and the imaging findings provided evidence supporting this idea (65). This point of view has also been confirmed in relevant studies in China, the United States, Denmark, and Japan (66–69). The underlying reason may be that people with a lower relative income engage in relatively heavy physical labor, experience more serious wear on the knee joint, and are less willing to seek medical treatment when they experience discomfort in the early stage (66). Educational level was generally positively correlated with income. However, this correlation was not absolute. People with low educational levels are less aware of KOA and related health policies, which cannot be effectively prevented. This is also a potential cause of KOA development and aggravation (70).
4.4 Depression and KOA
Previous studies have found that mental disorders such as depression increase the risk of KOA (71, 72). This was supported by a recent national cohort study, which showed that depression increased the risk of KOA at baseline over a 4-year follow-up and found a bidirectional association between depression and KOA (73). The association between depression and KOA may be explained by the inflammatory immune mechanism, which is related to the production and release of TNF-α, IL-1β, IL-6, and IL-8 (74, 75). Additionally, the proposed bone-brain axis may provide new insights into this association (76). Our MR results showed that psychological factors such as depression were negatively associated with KOA. We also performed reverse MR; however, this was contrary to the above clinical observations and did not achieve the expected results. Therefore, more rigorous clinical trials in genetics should be conducted to verify this association, which is a new direction for future in-depth research.
4.5 Diet and KOA
In this study, we mainly studied the consumption of whole wheat and its products and the consumption of sugar-sweetened products. Univariate and multivariate MR revealed a significant positive correlation between abstaining from whole wheat and its products and the risk of KOA. In other words, the intake of wheat and its derivative products was negatively correlated with the risk of KOA. The results of previous studies were consistent with our MR analysis results (77). Wheat is an important source of dietary grain fiber (78, 79), and it is also part of the grains included in the Mediterranean diet, which can reduce the risk of KOA owing to its high dietary fiber content (80, 81). Presently, there is limited research on the mechanism linking wheat and its products to KOA; however, the whole wheat diet may be associated with a reduced inflammatory response (82), and clinical trials have demonstrated that it can lower blood lipid and cholesterol levels (83), which could impact KOA. The univariate MR analysis of genetic prediction also showed a causal association between sugar deprivation and KOA; however, the multivariate MR analysis showed that this association was not obvious. Therefore, we believe that this association can be due to the existence of mediating factors. The mechanism underlying the relationship between glucose deprivation and KOA remains unclear. Current studies have shown that articular cartilage is an indispensable part of the knee joint, where only chondrocytes reside. The metabolic homeostasis of chondrocytes is related to the structure and function of cartilage tissue. Chondrocytes are required to obtain glucose and oxygen from the subchondral valley and synovial fluid for glucose metabolism (84). If this process leads to an insufficient nutrient supply, the cartilage will be damaged. Approximately 95% of adenosine triphosphate in chondrocytes is produced by glucose metabolism. We hypothesized that prolonged fasting from carbohydrate products may restrict chondrocyte glucose intake, leading to inhibited mitochondrial respiration, overactive or impaired glycolysis, and reduced total adenosine triphosphate production, assuming other factors remain constant (85–87). Therefore, it leads to a glucose metabolism disorder in chondrocytes and destroys the stability of the structure and function of cartilage tissue, leading to joint degeneration (88, 89). Additionally, current experiments have demonstrated that a high-sugar diet can increase the risk of KOA (90). A high-sugar diet can increase the release of free radicals, accelerate the process of degeneration, and increase the production of pro-inflammatory factors, leading to the formation of a local pro-inflammatory environment and increased risk or aggravation of KOA (91). Therefore, sugar should be moderately consumed in daily life.
4.6 Height and KOA
We observed a possible association between height and genetically predicted KOA. A Finnish population-based cohort study that adjusted for confounding factors yielded consistent results (92). Current research has focused on obesity and KOA. Particularly, most studies have focused on the associations among BMI, weight, and KOA. Interestingly, few studies have focused on the causal relationship between height and KOA. Hart et al. incidentally found a positive correlation between height and KOA in knee joint radiological observations in middle-aged British women (93). Presently, the basic index for determining obesity is BMI; however, most researchers consider weight, waist circumference, and fat (94) and ignore the effect of height, especially in patients with KOA having a normal BMI. From a mechanical perspective, leg length is directly proportional to the pressure generated by the knee. In addition, some studies have found that height and upper body weight are closely related to knee cartilage compression (92). Concurrently, some studies have shown that height is closely related to bone morphology and cartilage thickness, which may be a potential factor influencing the relationship between the two (95).
4.7 Hot drink temperature, balneotherapy, and KOA
In daily life, drinking hot drinks helps attenuate the risk of KOA, as compared to cold drinks. Nonetheless, very little research has been conducted on this topic, with a greater focus on balneotherapy and KOA. We found a genetic link between hot drink temperature and KOA, whereas previous studies reported that balneotherapy effectively relieved the symptoms of KOA (96, 97). As a non-drug complementary therapy for KOA, the mechanism of balneotherapy remains unclear; however, balneotherapy has been verified to affect the progression of KOA through thermal, chemical, and mechanical pathways, among which thermal stimulation plays a more important role (98). Matsumoto et al. found that balneotherapy can significantly improve pain, stiffness, and functional limitations of patients and greatly improve their quality of life (99). Moreover, soaking the knee joint in warm water causes a neuroendocrine response through overheating stimulation, which increases the concentration of serum opioid peptides (such as enkephin) to achieve analgesic and sedative effects (100, 101). Heat stimulation also relieves muscle spasms, inflammation, and oxidative stress. A hot mud bath may reduce the release of pro-inflammatory factors IL-1β, TNF-α, and IL-8, increase the level of anti-inflammatory transforming growth factor-β, and increase the level of cortisol to reduce the inflammatory response. Promoting the reduction in serum extracellular heat shock protein 72 kDa levels can also reduce the release of pro-inflammatory factors (102, 103). Cells and microRNAs (miRNAs) can regulate each other. miRNAs play an important role in the pathogenesis of osteoarthritis and have been detected in human synovial fluid. Therefore, miRNAs can be used as both diagnostic markers and therapeutic targets (104, 105). miRNA-181a has a positive correlation with KOA owing to its involvement in cartilage degradation (106), and the upregulation of miRNA-155, miRNA-223, miRNA-181a, and miRNA-146a levels all play a role in the pathogenesis of cartilage damage and synovitis (107). A previous trial found that after mud bath treatment, miRNA-155, miRNA-223, miRNA-181a, and miRNA146a levels in patients with KOA were significantly decreased, suggesting that balneotherapy has a unique therapeutic effect on KOA (108). In our MR study, hydrotherapy was the main focus of balneotherapy, and the multivariate MR results suggest a potential mediating effect. Through literature review, we found that balneotherapy also includes mud therapy; therefore, confounding factors may affect the MR results. Consequently, we included mud therapy in this study. Bath therapy has the advantages of safety and convenience, and as an alternative therapy for KOA, it has a certain effect, which can appropriately reduce the use of non-steroidal anti-inflammatory drugs, as well as reduce pain and certain economic burden for patients with KOA (109), with good economic and social benefits (110).
4.8 Advantages and limitations
This study has several advantages. First, the MR method used was novel and could avoid bias caused by traditional epidemiological observational studies. In this study, we used univariate and multivariate MR analyses to evaluate the causal relationship between risk factors and KOA from genetics and conducted multiple sensitivity analyses to eliminate outliers. A comprehensive assessment of causality makes the results more reliable. Second, we used publicly available large GWAS data to ensure sufficient sample size and reliable results. We also used independent data for validation to avoid an overlap between the exposure and outcome samples (111). Third, the data were obtained from a European population to avoid bias caused by different populations. We used stringent standard screening tool variables to ensure that extreme values did not affect the results.
However, this study had some limitations. First, there was significant heterogeneity between some risk factors and KOA. Although the evaluation using the random-effects model showed that the results were robust, it inevitably posed some challenges to causality. Second, our use of rigorous criteria for the selection of IV might have missed some outcomes. Third, because of the excessive number of risk factors studied, the results from the multivariate MR analysis were inaccurate owing to bias caused by the existence of collinear problems; therefore, our application of the mv-lasso function to solve collinear problems corrected some outliers. Fourth, this study was based on a European population, which limits the generalizability of the results because of genetic differences among ethnic groups. Finally, the results of the MR analysis suggested a causal association; however, further clinical trials are needed to investigate the underlying mechanisms.
4.9 Clinical implications
This study provides an in-depth causal exploration of several potential risk factors for KOA. We found that BMI plays an irreplaceable role as a mediator between risk factors and KOA, which is in line with previous MR and meta-analysis studies on the causal relationship (112–114) between BMI and KOA that found that BMI increases the risk of KOA in line with our-study results. In clinical settings, physicians can emphasize the importance of weight management in KOA prevention and can encourage healthy eating (advising them to avoid prolonged abstinence from wheat products) and regular exercise to reduce obesity and related health problems. By identifying factors that have a clear causal relationship with KOA (hyperthyroidism/thyrotoxicosis, essential hypertension, MS), physicians can be more precise in assessing a patient’s risk, and concurrently treating and managing his or her thyroid disease, hypertension, and MS may help to reduce the risk of KOA. Public health policymakers can build on these findings to develop more targeted prevention and intervention measures, and by improving public health education, the associations between educational attainment and income and KOA risk found in the study remind public health policymakers of the need to consider the impact of socioeconomic factors on health. Policies can target low-income and less-educated populations with more health support and resources. The findings could also be used to develop health education policies for the general public, particularly in the areas of diet and lifestyle, and consideration could be given to incorporating screening and management of thyroid disorders and hypertension into public health programs, as well as raising public awareness of KOA risk factors.
5 Conclusion
In summary, our study systematically analyzed the causal relationships between genetically predicted KOA and thyroid dysfunction, MS, essential hypertension, educational level, income, dietary factors, height, balneotherapy, and psychophysiological factors. This study highlights the impact of modifiable risk factors on KOA and suggests that the use of drugs that interfere with risk factors can also affect the progression of KOA. Therefore, this study provides a new research direction for the prevention and treatment of KOA and information for the formulation of public health policies.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author/s.
Ethics statement
This study was based on publicly available datasets. Ethical review and approval was not required for the study, in accordance with the local legislation and institutional requirements.
Author contributions
ZD: Conceptualization, Methodology, Software, Visualization, Writing – original draft, Writing – review, editing. DG: Formal analysis, Methodology, Resources, Validation, Visualization, Writing – review, editing. JZ: Data curation, Formal analysis, Writing – review, editing. RZ: Validation, Writing – review, editing. CL: Resources, Validation, Writing – review, editing. HC: Data curation, Methodology, Writing.YM: Conceptualization, Funding acquisition, Methodology, Supervision, Writing–review and editing.
Funding
The author(s) declare that financial support was received for the research, authorship, and/or publication of this article. This work was supported by the fifth batch of special funds for the Training Program for Outstanding Clinical Talents of Traditional Chinese Medicine [National Letter of TCM Education (2022) No. 1]; the Natural Science Foundation of Shandong Province (ZR2021MH373); High Level Traditional Chinese Medicine Key Disciplines of the State Administration of Traditional Chinese Medicine-External Treatment of Traditional Chinese Medicine(zyyzdxk-2023116); Shandong Traditional Chinese Medicine Science and Technology Project (2020Q002); and the Jinan University 20 Funding Project (2020GXRC005).
Acknowledgments
Data for this study were obtained from the FinnGen Consortium, Medical Research Council Integrated Epidemiology Unit (MRC-IEU), United Kingdom Biobank, Neale Lab, Science Genetic Association Consortium, and the IEU OpenGWAS project. The authors acknowledge the efforts of their staff.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmed.2024.1405188/full#supplementary-material
Footnotes
References
1. Bortoluzzi, A, Furini, F, and Scirè, CA. Osteoarthritis and its management - epidemiology, nutritional aspects and environmental factors. Autoimmun Rev. (2018) 17:1097–104. doi: 10.1016/j.autrev.2018.06.002
2. Glyn-Jones, S, Palmer, AJ, Agricola, R, Price, AJ, Vincent, TL, Weinans, H, et al. Osteoarthritis. Lancet. (2015) 386:376–87. doi: 10.1016/S0140-6736(14)60802-3
3. Cui, A, Li, H, Wang, D, Zhong, J, Chen, Y, and Lu, H. Global, regional prevalence, incidence and risk factors of knee osteoarthritis in population-based studies. EClinicalMedicine. (2020) 29-30:100587. doi: 10.1016/j.eclinm.2020.100587
4. Aw, NM, Yeo, SJ, Wylde, V, Wong, SB, Chan, D, Thumboo, J, et al. Impact of pain sensitisation on the quality of life of patients with knee osteoarthritis. RMD Open. (2022) 8:e001938. doi: 10.1136/rmdopen-2021-001938
5. Vos, T, Flaxman, AD, Naghavi, M, Lozano, R, Michaud, C, Ezzati, M, et al. Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990-2010: a systematic analysis for the global burden of disease study 2010. Lancet. (2012) 380:2163–96. doi: 10.1016/S0140-6736(12)61729-2
6. Vina, ER, and Kwoh, CK. Epidemiology of osteoarthritis: literature update. Curr Opin Rheumatol. (2018) 30:160–7. doi: 10.1097/BOR.0000000000000479
7. Safiri, S, Kolahi, AA, Smith, E, Hill, C, Bettampadi, D, Mansournia, MA, et al. Global, regional and national burden of osteoarthritis 1990-2017: a systematic analysis of the global burden of disease study 2017. Ann Rheum Dis. (2020) 79:819–28. doi: 10.1136/annrheumdis-2019-216515
8. Kang, X, Fransen, M, Zhang, Y, Li, H, Ke, Y, Lu, M, et al. The high prevalence of knee osteoarthritis in a rural Chinese population: the Wuchuan osteoarthritis study. Arthritis Rheum. (2009) 61:641–7. doi: 10.1002/art.24464
9. Callahan, LF, Cleveland, RJ, Shreffler, J, Schwartz, TA, Schoster, B, Randolph, R, et al. Associations of educational attainment, occupation and community poverty with knee osteoarthritis in the Johnston County (North Carolina) osteoarthritis project. Arthritis Res Ther. (2011) 13:R169. doi: 10.1186/ar3492
10. Xie, Y, Zhou, W, Zhong, Z, Zhao, Z, Yu, H, Huang, Y, et al. Metabolic syndrome, hypertension, and hyperglycemia were positively associated with knee osteoarthritis, while dyslipidemia showed no association with knee osteoarthritis. Clin Rheumatol. (2021) 40:711–24. doi: 10.1007/s10067-020-05216-y
11. Kim, BY, Kim, SS, Park, HK, and Kim, HS. Assessment of the relationship between knee ultrasound and clinical symptoms in patients with thyroid dysfunction. J Int Med Res. (2020) 48:7701. doi: 10.1177/0300060519897701
12. Wang, H, and Ma, B. Healthcare and scientific treatment of knee osteoarthritis. J Healthc Eng. (2022) 2022:5919686. doi: 10.1155/2022/5919686
13. Dashti, HS, Chen, A, Daghlas, I, and Saxena, R. Morning diurnal preference and food intake: a Mendelian randomization study. Am J Clin Nutr. (2020) 112:1348–57. doi: 10.1093/ajcn/nqaa219
14. Hunter, DJ, Niu, J, Zhang, Y, Nevitt, MC, Xu, L, Lui, LY, et al. Knee height, knee pain, and knee osteoarthritis: the Beijing osteoarthritis study. Arthritis Rheum. (2005) 52:1418–23. doi: 10.1002/art.21017
15. Deveza, LA, Melo, L, Yamato, TP, Mills, K, Ravi, V, and Hunter, DJ. Knee osteoarthritis phenotypes and their relevance for outcomes: a systematic review. Osteoarthr Cartil. (2017) 25:1926–41. doi: 10.1016/j.joca.2017.08.009
16. Davies, NM, Holmes, MV, and Davey Smith, G. Reading Mendelian randomisation studies: a guide, glossary, and checklist for clinicians. BMJ. (2018) 362:k601. doi: 10.1136/bmj.k601
17. Bowden, J, and Holmes, MV. Meta-analysis and Mendelian randomization: a review. Res Synth Methods. (2019) 10:486–96. doi: 10.1002/jrsm.1346
18. Emdin, CA, Khera, AV, and Kathiresan, S. Mendelian randomization. JAMA. (2017) 318:1925–6. doi: 10.1001/jama.2017.17219
19. Sanderson, E . Multivariable Mendelian randomization and mediation. Cold Spring Harb Perspect Med. (2021) 11:a038984. doi: 10.1101/cshperspect.a038984
20. Carter, AR, Sanderson, E, Hammerton, G, Richmond, RC, Davey Smith, G, Heron, J, et al. Mendelian randomisation for mediation analysis: current methods and challenges for implementation. Eur J Epidemiol. (2021) 36:465–78. doi: 10.1007/s10654-021-00757-1
21. Sanderson, E, Davey Smith, G, Windmeijer, F, and Bowden, J. An examination of multivariable Mendelian randomization in the single-sample and two-sample summary data settings. Int J Epidemiol. (2019) 48:713–27. doi: 10.1093/ije/dyy262
22. Skrivankova, VW, Richmond, RC, Woolf, BAR, Davies, NM, Swanson, SA, VanderWeele, TJ, et al. Strengthening the reporting of observational studies in epidemiology using Mendelian randomisation (STROBE-MR): explanation and elaboration. BMJ. (2021) 375:n2233. doi: 10.1136/bmj.n2233
23. Skrivankova, VW, Richmond, RC, Woolf, BAR, Yarmolinsky, J, Davies, NM, Swanson, SA, et al. Strengthening the reporting of observational studies in epidemiology using Mendelian randomization: the STROBE-MR statement. JAMA. (2021) 326:1614–21. doi: 10.1001/jama.2021.18236
24. Tachmazidou, I, Hatzikotoulas, K, Southam, L, Esparza-Gordillo, J, Haberland, V, Zheng, J, et al. Identification of new therapeutic targets for osteoarthritis through genome-wide analyses of UK biobank data. Nat Genet. (2019) 51:230–6. doi: 10.1038/s41588-018-0327-1
25. Hartley, A, Sanderson, E, Granell, R, Paternoster, L, Zheng, J, Smith, GD, et al. Using multivariable Mendelian randomization to estimate the causal effect of bone mineral density on osteoarthritis risk, independently of body mass index. Int J Epidemiol. (2022) 51:1254–67. doi: 10.1093/ije/dyab251
26. Larsson, SC, and Burgess, S. Appraising the causal role of smoking in multiple diseases: a systematic review and meta-analysis of Mendelian randomization studies. EBioMedicine. (2022) 82:104154. doi: 10.1016/j.ebiom.2022.104154
27. Gill, D, Karhunen, V, Malik, R, Dichgans, M, and Sofat, N. Cardiometabolic traits mediating the effect of education on osteoarthritis risk: a Mendelian randomization study. Osteoarthr Cartil. (2021) 29:365–71. doi: 10.1016/j.joca.2020.12.015
28. Pirastu, N, Cordioli, M, Nandakumar, P, Mignogna, G, Abdellaoui, A, Hollis, B, et al. Genetic analyses identify widespread sex-differential participation bias. Nat Genet. (2021) 53:663–71. doi: 10.1038/s41588-021-00846-7
29. Hemani, G, Zheng, J, Elsworth, B, Wade, KH, Haberland, V, Baird, D, et al. The MR-base platform supports systematic causal inference across the human phenome. eLife. (2018) 7:e34408. doi: 10.7554/eLife.34408
30. Maciejewski, ML, Dowd, BE, and Norton, EC. Instrumental variables and heterogeneous treatment effects. JAMA. (2022) 327:1177–8. doi: 10.1001/jama.2022.2505
31. Long, Y, Tang, L, Zhou, Y, Zhao, S, and Zhu, H. Causal relationship between gut microbiota and cancers: a two-sample Mendelian randomisation study. BMC Med. (2023) 21:66. doi: 10.1186/s12916-023-02761-6
32. Burgess, S, Small, DS, and Thompson, SG. A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res. (2017) 26:2333–55. doi: 10.1177/0962280215597579
33. Liu, S, Feng, S, Du, F, Zhang, K, and Shen, Y. Association of smoking, alcohol, and coffee consumption with the risk of ovarian cancer and prognosis: a Mendelian randomization study. BMC Cancer. (2023) 23:256. doi: 10.1186/s12885-023-10737-1
34. Bowden, J, Hemani, G, and Davey Smith, G. Invited commentary: detecting individual and global horizontal pleiotropy in mendelian randomization-a job for the humble heterogeneity statistic. Am J Epidemiol. (2018) 187:2681–5. doi: 10.1093/aje/kwy185
35. Bowden, J, del Greco M, F, Minelli, C, Davey Smith, G, Sheehan, N, and Thompson, J. A framework for the investigation of pleiotropy in two-sample summary data Mendelian randomization. Stat Med. (2017) 36:1783–802. doi: 10.1002/sim.7221
36. Labrecque, J, and Swanson, SA. Understanding the assumptions underlying instrumental variable analyses: a brief review of falsification strategies and related tools. Curr Epidemiol Rep. (2018) 5:214–20. doi: 10.1007/s40471-018-0152-1
37. Bowden, J, Davey Smith, G, and Burgess, S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol. (2015) 44:512–25. doi: 10.1093/ije/dyv080
38. Bowden, J, Davey Smith, G, Haycock, PC, and Burgess, S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. (2016) 40:304–14. doi: 10.1002/gepi.21965
39. Hartwig, FP, Davey Smith, G, and Bowden, J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. Int J Epidemiol. (2017) 46:1985–98. doi: 10.1093/ije/dyx102
40. Burgess, S, Bowden, J, Fall, T, Ingelsson, E, and Thompson, SG. Sensitivity analyses for robust causal inference from Mendelian randomization analyses with multiple genetic variants. Epidemiology. (2017) 28:30–42. doi: 10.1097/EDE.0000000000000559
41. Verbanck, M, Chen, CY, Neale, B, and Do, R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. (2018) 50:693–8. doi: 10.1038/s41588-018-0099-7
42. Chaisson, CE, McAlindon, TE, Felson, DT, Naimark, A, Wilson, PW, and Sawin, CT. Lack of association between thyroid status and chondrocalcinosis or osteoarthritis: the Framingham osteoarthritis study. J Rheumatol. (1996) 23:711–5.
43. Hellevik, AI, Johnsen, MB, Langhammer, A, Fenstad, AM, Furnes, O, Storheim, K, et al. Incidence of total hip or knee replacement due to osteoarthritis in relation to thyroid function: a prospective cohort study (the Nord-Trøndelag health study). BMC Musculoskelet Disord. (2017) 18:201. doi: 10.1186/s12891-017-1565-6
44. Bland, JH, Frymoyer, JW, Newberg, AH, Revers, R, and Norman, RJ. Rheumatic syndromes in endocrine disease. Semin Arthritis Rheum. (1979) 9:23–65. doi: 10.1016/0049-0172(79)90002-7
45. Newcombe, DS, Ortel, RW, and Levey, GS. Activation of human synovial membrane adenylate cyclase by thyroid stimulating hormone (TSH). Biochem Biophys Res Commun. (1972) 48:201–4. doi: 10.1016/0006-291X(72)90363-4
46. Tagoe, CE, Wang, W, Wang, S, and Barbour, KE. Association of anti-thyroid antibodies with radiographic knee osteoarthritis and chondrocalcinosis: a NHANES III study. Ther Adv Musculoskelet Dis. (2021) 13:5199. doi: 10.1177/1759720X211035199
47. Solak, Y, Afsar, B, Vaziri, ND, Aslan, G, Yalcin, CE, Covic, A, et al. Hypertension as an autoimmune and inflammatory disease. Hypertens Res. (2016) 39:567–73. doi: 10.1038/hr.2016.35
48. Zhang, YM, Wang, J, and Liu, XG. Association between hypertension and risk of knee osteoarthritis: a meta-analysis of observational studies. Medicine (Baltimore). (2017) 96:e7584. doi: 10.1097/MD.0000000000007584
49. Li, H, George, DM, Jaarsma, RL, and Mao, X. Metabolic syndrome and components exacerbate osteoarthritis symptoms of pain, depression and reduced knee function. Ann Transl Med. (2016) 4:133. doi: 10.21037/atm.2016.03.48
50. Zhuo, Q, Yang, W, Chen, J, and Wang, Y. Metabolic syndrome meets osteoarthritis. Nat Rev Rheumatol. (2012) 8:729–37. doi: 10.1038/nrrheum.2012.135
51. Lo, K, Au, M, Ni, J, and Wen, C. Association between hypertension and osteoarthritis: a systematic review and meta-analysis of observational studies. J Orthop Translat. (2022) 32:12–20. doi: 10.1016/j.jot.2021.05.003
52. Chan, PMB, Wen, C, Yang, WC, Yan, C, and Chiu, K. Is subchondral bone cyst formation in non-load-bearing region of osteoarthritic knee a vascular problem. Med Hypotheses. (2017) 109:80–3. doi: 10.1016/j.mehy.2017.09.027
53. Zhen, G, Wen, C, Jia, X, Li, Y, Crane, JL, Mears, SC, et al. Inhibition of TGF-β signaling in mesenchymal stem cells of subchondral bone attenuates osteoarthritis. Nat Med. (2013) 19:704–12. doi: 10.1038/nm.3143
54. Velasquez, MT, and Katz, JD. Osteoarthritis: another component of metabolic syndrome. Metab Syndr Relat Disord. (2010) 8:295–305. doi: 10.1089/met.2009.0110
55. Shi, X, and Schlenk, EA. Association of hypertension with knee pain severity among people with knee osteoarthritis. Pain Manag Nurs. (2022) 23:135–41. doi: 10.1016/j.pmn.2021.08.002
56. Ambroselli, D, Masciulli, F, Romano, E, Catanzaro, G, Besharat, ZM, Massari, MC, et al. New advances in metabolic syndrome, from prevention to treatment: the role of diet and food. Nutrients. (2023) 15:640. doi: 10.3390/nu15030640
57. Gao, YH, Zhao, CW, Liu, B, Dong, N, Ding, L, Li, YR, et al. An update on the association between metabolic syndrome and osteoarthritis and on the potential role of leptin in osteoarthritis. Cytokine. (2020) 129:155043. doi: 10.1016/j.cyto.2020.155043
58. Kierdorf, K, and Fritz, G. RAGE regulation and signaling in inflammation and beyond. J Leukoc Biol. (2013) 94:55–68. doi: 10.1189/jlb.1012519
59. Rogero, MM, and Calder, PC. Obesity, inflammation, toll-like receptor 4 and fatty acids. Nutrients. (2018) 10:432. doi: 10.3390/nu10040432
60. Monteiro, L, Pereira, JADS, Palhinha, L, and Moraes-Vieira, PMM. Leptin in the regulation of the immunometabolism of adipose tissue-macrophages. J Leukoc Biol. (2019) 106:703–16. doi: 10.1002/JLB.MR1218-478R
61. Zegeye, MM, Lindkvist, M, Fälker, K, Kumawat, AK, Paramel, G, Grenegård, M, et al. Activation of the JAK/STAT3 and PI3K/AKT pathways are crucial for IL-6 trans-signaling-mediated pro-inflammatory response in human vascular endothelial cells. Cell Commun Signal. (2018) 16:55. doi: 10.1186/s12964-018-0268-4
62. Zhou, S, Lu, W, Chen, L, Ge, Q, Chen, D, Xu, Z, et al. AMPK deficiency in chondrocytes accelerated the progression of instability-induced and ageing-associated osteoarthritis in adult mice. Sci Rep. (2017) 7:43245. doi: 10.1038/srep43245
63. Terkeltaub, R, Yang, B, Lotz, M, and Liu-Bryan, R. Chondrocyte AMP-activated protein kinase activity suppresses matrix degradation responses to proinflammatory cytokines interleukin-1β and tumor necrosis factor α. Arthritis Rheum. (2011) 63:1928–37. doi: 10.1002/art.30333
64. Dickson, BM, Roelofs, AJ, Rochford, JJ, Wilson, HM, and De Bari, C. The burden of metabolic syndrome on osteoarthritic joints. Arthritis Res Ther. (2019) 21:289. doi: 10.1186/s13075-019-2081-x
65. Lee, JY, Han, K, Park, YG, and Park, SH. Effects of education, income, and occupation on prevalence and symptoms of knee osteoarthritis. Sci Rep. (2021) 11:13983. doi: 10.1038/s41598-021-93394-3
66. Liu, Y, Zhang, H, Liang, N, Fan, W, Li, J, Huang, Z, et al. Prevalence and associated factors of knee osteoarthritis in a rural Chinese adult population: an epidemiological survey. BMC Public Health. (2016) 16:94. doi: 10.1186/s12889-016-2782-x
67. Vennu, V, Abdulrahman, TA, Alenazi, AM, and Bindawas, SM. Associations between social determinants and the presence of chronic diseases: data from the osteoarthritis initiative. BMC Public Health. (2020) 20:1323. doi: 10.1186/s12889-020-09451-5
68. Jørgensen, KT, Pedersen, BV, Nielsen, NM, Hansen, AV, Jacobsen, S, and Frisch, M. Socio-demographic factors, reproductive history and risk of osteoarthritis in a cohort of 4.6 million Danish women and men. Osteoarthr Cartil. (2011) 19:1176–82. doi: 10.1016/j.joca.2011.07.009
69. Ikeda, T, Aida, J, Tsuboya, T, Sugiyama, K, Kondo, K, and Osaka, K. Psychosocial factors and knee pain among older people in Japan: the JAGES Cross-sectional study. Clin J Pain. (2019) 35:983–8. doi: 10.1097/AJP.0000000000000761
70. Aily, JB, De Almeida, AC, Ramírez, PC, Da Silva Alexandre, T, and Mattiello, SM. Lower education is an associated factor with the combination of pain catastrophizing and kinesiophobia in patients with knee osteoarthritis. Clin Rheumatol. (2021) 40:2361–7. doi: 10.1007/s10067-020-05518-1
71. Swain, S, Coupland, C, Mallen, C, Kuo, CF, Sarmanova, A, Bierma-Zeinstra, SMA, et al. Temporal relationship between osteoarthritis and comorbidities: a combined case control and cohort study in the UK primary care setting. Rheumatology (Oxford). (2021) 60:4327–39. doi: 10.1093/rheumatology/keab067
72. Vennu, V, Misra, H, and Misra, A. Depressive symptoms and the risk of arthritis: a survival analysis using data from the osteoarthritis initiative. Indian J Psychiatry. (2019) 61:444–50. doi: 10.4103/psychiatry.IndianJPsychiatry_241_18
73. Lu, H, Wang, L, Zhou, W, Jin, S, Chen, H, Su, Y, et al. Bidirectional association between knee osteoarthritis and depressive symptoms: evidence from a nationwide population-based cohort. BMC Musculoskelet Disord. (2022) 23:213. doi: 10.1186/s12891-022-05137-8
74. Mathiessen, A, and Conaghan, PG. Synovitis in osteoarthritis: current understanding with therapeutic implications. Arthritis Res Ther. (2017) 19:18. doi: 10.1186/s13075-017-1229-9
75. Da Silva, MR, Linhares, D, Vasconcelos, DM, Alves, CJ, Neves, N, Costa, G, et al. Neuroimmune expression in hip osteoarthritis: a systematic review. BMC Musculoskelet Disord. (2017) 18:394. doi: 10.1186/s12891-017-1755-2
76. Zhao, Y, Suo, Y, Yang, Z, Hao, Y, Li, W, Su, Y, et al. Inspiration for the prevention and treatment of neuropsychiatric disorders: new insight from the bone-brain-axis. Brain Res Bull. (2021) 177:263–72. doi: 10.1016/j.brainresbull.2021.10.009
77. Zertuche, JP, Rabasa, G, Lichtenstein, AH, Matthan, NR, Nevitt, M, Torner, J, et al. Alkylresorcinol, a biomarker for whole grain intake, and its association with osteoarthritis: the MOST study. Osteoarthr Cartil. (2022) 30:1337–43. doi: 10.1016/j.joca.2022.07.004
78. Slavin, JL, Martini, MC, Jacobs, DR, and Marquart, L. Plausible mechanisms for the protectiveness of whole grains. Am J Clin Nutr. (1999) 70:459S–63S. doi: 10.1093/ajcn/70.3.459s
79. Liu, S . Intake of refined carbohydrates and whole grain foods in relation to risk of type 2 diabetes mellitus and coronary heart disease. J Am Coll Nutr. (2002) 21:298–306. doi: 10.1080/07315724.2002.10719227
80. Dai, Z, Niu, J, Zhang, Y, Jacques, P, and Felson, DT. Dietary intake of fibre and risk of knee osteoarthritis in two US prospective cohorts. Ann Rheum Dis. (2017) 76:1411–9. doi: 10.1136/annrheumdis-2016-210810
81. Veronese, N, Koyanagi, A, Stubbs, B, Cooper, C, Guglielmi, G, Rizzoli, R, et al. Mediterranean diet and knee osteoarthritis outcomes: a longitudinal cohort study. Clin Nutr. (2019) 38:2735–9. doi: 10.1016/j.clnu.2018.11.032
82. Xu, Y, Wan, Q, Feng, J, Du, L, Li, K, and Zhou, Y. Whole grain diet reduces systemic inflammation: a meta-analysis of 9 randomized trials. Medicine (Baltimore). (2018) 97:e12995. doi: 10.1097/MD.0000000000012995
83. Sang, S, and Chu, Y. Whole grain oats, more than just a fiber: role of unique phytochemicals. Mol Nutr Food Res. (2017) 61:715. doi: 10.1002/mnfr.201600715
84. Marconi, A, Hancock-Ronemus, A, and Gillis, JA. Adult chondrogenesis and spontaneous cartilage repair in the skate, Leucoraja erinacea. Elife. (2020) 9:e53414. doi: 10.7554/eLife.53414
85. Terabe, K, Ohashi, Y, Tsuchiya, S, Ishizuka, S, Knudson, CB, and Knudson, W. Chondroprotective effects of 4-methylumbelliferone and hyaluronan synthase-2 overexpression involve changes in chondrocyte energy metabolism. J Biol Chem. (2019) 294:17799–817. doi: 10.1074/jbc.RA119.009556
86. Ohashi, Y, Takahashi, N, Terabe, K, Tsuchiya, S, Kojima, T, Knudson, CB, et al. Metabolic reprogramming in chondrocytes to promote mitochondrial respiration reduces downstream features of osteoarthritis. Sci Rep. (2021) 11:15131. doi: 10.1038/s41598-021-94611-9
87. Kong, P, Chen, R, Zou, FQ, Wang, Y, Liu, MC, and Wang, WG. HIF-1α repairs degenerative chondrocyte glycolytic metabolism by the transcriptional regulation of Runx2. Eur Rev Med Pharmacol Sci. (2021) 25:1206–14. doi: 10.26355/eurrev_202102_24823
88. Anderson, JR, Chokesuwattanaskul, S, Phelan, MM, Welting, TJM, Lian, LY, Peffers, MJ, et al. 1H NMR metabolomics identifies underlying inflammatory pathology in osteoarthritis and rheumatoid arthritis synovial joints. J Proteome Res. (2018) 17:3780–90. doi: 10.1021/acs.jproteome.8b00455
89. Wei, JY, Zhang, DM, Xie, J, and Zhou, XD. Research progress in glucose metabolism of chondrocytes. Sichuan Da Xue Xue Bao Yi Xue Ban. (2021) 52:923–8. doi: 10.12182/20211160206
90. Min, Y, Ahn, D, Truong, TMT, Kim, M, Heo, Y, Jee, Y, et al. Excessive sucrose exacerbates high fat diet-induced hepatic inflammation and fibrosis promoting osteoarthritis in mice model. J Nutr Biochem. (2023) 112:109223. doi: 10.1016/j.jnutbio.2022.109223
91. Ricker, MA, and Haas, WC. Anti-inflammatory diet in clinical practice: a review. Nutr Clin Pract. (2017) 32:318–25. doi: 10.1177/0884533617700353
92. Welling, M, Auvinen, J, Lehenkari, P, Männikkö, M, Karppinen, J, and Eskola, PJ. Association between height and osteoarthritis of the knee and hip: the northern Finland birth cohort 1966 study. Int J Rheum Dis. (2017) 20:1095–104. doi: 10.1111/1756-185X.13059
93. Hart, DJ, Doyle, DV, and Spector, TD. Incidence and risk factors for radiographic knee osteoarthritis in middle-aged women: the Chingford study. Arthritis Rheum. (1999) 42:17–24. doi: 10.1002/1529-0131(199901)42:1<17::AID-ANR2>3.0.CO;2-E
94. Sun, Y, Li, Y, Yu, T, and Zhang, J. Causal associations of anthropometric measurements with osteoarthritis: a Mendelian randomization study. PLoS One. (2023) 18:e0279198. doi: 10.1371/journal.pone.0279198
95. Schneider, MT, Rooks, N, and Besier, T. Cartilage thickness and bone shape variations as a function of sex, height, body mass, and age in young adult knees. Sci Rep. (2022) 12:11707. doi: 10.1038/s41598-022-15585-w
96. Varzaityte, L, Kubilius, R, Rapoliene, L, Bartuseviciute, R, Balcius, A, Ramanauskas, K, et al. The effect of balneotherapy and peloid therapy on changes in the functional state of patients with knee joint osteoarthritis: a randomized, controlled, single-blind pilot study. Int J Biometeorol. (2020) 64:955–64. doi: 10.1007/s00484-019-01785-z
97. Antonelli, M, Donelli, D, and Fioravanti, A. Effects of balneotherapy and spa therapy on quality of life of patients with knee osteoarthritis: a systematic review and meta-analysis. Rheumatol Int. (2018) 38:1807–24. doi: 10.1007/s00296-018-4081-6
98. Fioravanti, A, Karagülle, M, Bender, T, and Karagülle, MZ. Balneotherapy in osteoarthritis: facts, fiction and gaps in knowledge. Eur J Integr Med. (2017) 9:148–50. doi: 10.1016/j.eujim.2017.01.001
99. Matsumoto, H, Hagino, H, Hayashi, K, Ideno, Y, Wada, T, Ogata, T, et al. The effect of balneotherapy on pain relief, stiffness, and physical function in patients with osteoarthritis of the knee: a meta-analysis. Clin Rheumatol. (2017) 36:1839–47. doi: 10.1007/s10067-017-3592-y
100. Karagülle, M, Karagülle, MZ, Karagülle, O, Dönmez, A, and Turan, M. A 10-day course of SPA therapy is beneficial for people with severe knee osteoarthritis. A 24-week randomised, controlled pilot study. Clin Rheumatol. (2007) 26:2063–71. doi: 10.1007/s10067-007-0618-x
101. Cozzi, F, Lazzarin, P, Todesco, S, and Cima, L. Hypothalamic-pituitary-adrenal axis dysregulation in healthy subjects undergoing mud-bath applications. Arthritis Rheum. (1995) 38:724–5. doi: 10.1002/art.1780380530
102. Ortega, E, Gálvez, I, Hinchado, MD, Guerrero, J, Martín-Cordero, L, and Torres-Piles, S. Anti-inflammatory effect as a mechanism of effectiveness underlying the clinical benefits of pelotherapy in osteoarthritis patients: regulation of the altered inflammatory and stress feedback response. Int J Biometeorol. (2017) 61:1777–85. doi: 10.1007/s00484-017-1361-x
103. Hou, C, Liang, L, Chu, X, Qin, W, Li, Y, and Zhao, Y. The short-term efficacy of mud therapy for knee osteoarthritis: a meta-analysis. Medicine (Baltimore). (2020) 99:e19761. doi: 10.1097/MD.0000000000019761
104. De Palma, A, Cheleschi, S, Pascarelli, NA, Tenti, S, Galeazzi, M, and Fioravanti, A. Do MicroRNAs have a key epigenetic role in osteoarthritis and in mechanotransduction. Clin Exp Rheumatol. (2017) 35:518–26. doi: 10.1016/j.molmed.2011.11.005
105. Murata, K, Yoshitomi, H, Tanida, S, Ishikawa, M, Nishitani, K, Ito, H, et al. Plasma and synovial fluid microRNAs as potential biomarkers of rheumatoid arthritis and osteoarthritis. Arthritis Res Ther. (2010) 12:R86. doi: 10.1186/ar3013
106. Nakamura, A, Rampersaud, YR, Sharma, A, Lewis, SJ, Wu, B, Datta, P, et al. Identification of microRNA-181a-5p and microRNA-4454 as mediators of facet cartilage degeneration. JCI Insight. (2016) 1:e86820. doi: 10.1172/jci.insight.86820
107. Okuhara, A, Nakasa, T, Shibuya, H, Niimoto, T, Adachi, N, Deie, M, et al. Changes in microRNA expression in peripheral mononuclear cells according to the progression of osteoarthritis. Mod Rheumatol. (2012) 22:446–57. doi: 10.3109/s10165-011-0536-2
108. Giannitti, C, de Palma, A, Pascarelli, NA, Cheleschi, S, Giordano, N, Galeazzi, M, et al. Can balneotherapy modify microRNA expression levels in osteoarthritis? A comparative study in patients with knee osteoarthritis. Int J Biometeorol. (2017) 61:2153–8. doi: 10.1007/s00484-017-1420-3
109. Fraioli, A, Mennuni, G, Fontana, M, Nocchi, S, Ceccarelli, F, Perricone, C, et al. Efficacy of spa therapy, mud-pack therapy, balneotherapy, and mud-bath therapy in the management of knee osteoarthritis. A systematic review. Biomed Res Int. (2018) 2018:1042576. doi: 10.1155/2018/1042576
110. Ciani, O, Pascarelli, NA, Giannitti, C, Galeazzi, M, Meregaglia, M, Fattore, G, et al. Mud-bath therapy in addition to usual care in bilateral knee osteoarthritis: An economic evaluation alongside a randomized controlled trial. Arthritis Care Res. (2017) 69:966–72. doi: 10.1002/acr.23116
111. Burgess, S, Davies, NM, and Thompson, SG. Bias due to participant overlap in two-sample Mendelian randomization. Genet Epidemiol. (2016) 40:597–608. doi: 10.1002/gepi.21998
112. He, Y, Zheng, C, He, MH, and Huang, JR. The causal relationship between body mass index and the risk of osteoarthritis. Int J Gen Med. (2021) 14:2227–37. doi: 10.2147/IJGM.S314180
113. Ho, J, Mak, CCH, Sharma, V, To, K, and Khan, W. Mendelian randomization studies of lifestyle-related risk factors for osteoarthritis: a PRISMA review and Meta-analysis. Int J Mol Sci. (2022) 23:11906. doi: 10.3390/ijms231911906
Keywords: Mendelian randomization, knee osteoarthritis, risk factors, causal relationship, genetic variants
Citation: Diao Z, Guo D, Zhang J, Zhang R, Li C, Chen H and Ma Y (2024) Causal relationship between modifiable risk factors and knee osteoarthritis: a Mendelian randomization study. Front. Med. 11:1405188. doi: 10.3389/fmed.2024.1405188
Edited by:
Carlos Goicoechea, Rey Juan Carlos University, SpainReviewed by:
Yang Sun, The First Hospital of Jilin University, ChinaM. Carmen Ruiz-Cantero, University of Barcelona, Spain
Copyright © 2024 Diao, Guo, Zhang, Zhang, Li, Chen and Ma. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Yuxia Ma, ZHpoOTcxMjExQDE2My5jb20=
†These authors have contributed equally to this work