Based on our current understanding, this study represents a pioneering effort in examining and dissecting the repercussions of six gene targets associated with HbA1c (GCK, HK1, HKDC1,
TCF7L, INS, and HFE), which are susceptible to activation by antidiabetic medications. Employing MR methods, we aimed to scrutinize the plausible risk factors for the development of OA. This research endeavors to fill a critical gap in the existing literature, shedding light on the interplay between these specific gene targets and the potential onset of OA in the context of antidiabetic drug interventions. The analysis results indicate that glucokinase activators, metformin, and iron chelators can reduce the risk of osteoarthritis, while insulin increases the risk of osteoarthritis. We discuss the following aspects regarding these results.
GCK and HK1 were chosen as targets for glucokinase activators[35]. However, the effects of these two genes differ: GCK activation reduces the incidence of hip OA but does not affect knee OA or OA at any site. HK1 activation, on the other hand, decreases the incidence of knee OA and OA at any sites but not hip OA. We attribute these differences to the following reasons: firstly, GCK and HK1 may operate through different mechanisms. GCK has a 20-fold lower affinity for glucose compared to the second-ranked hexokinase 2 (HK2)[36]. GCK can regulate its activity through slow conformational dynamics, influenced by various protein interactions and post-translational modifications, leading to downstream physiological consequences[37]. Studies suggest that increased cytoplasmic HK1 content and decreased GAPDH activity in macrophages of diabetic patients may affect glucose metabolism by modulating GAPDH through HK1 mitochondrial binding[38]. The mitochondrial binding of HK1 allows preferential entry into the mitochondrial ATP pool, enhancing tricarboxylic acid cycle activity and increasing protein translation[39]. However, research on HK1 is currently limited. Secondly, GCK mainly regulates glucose in the liver and pancreatic beta cells, while HK1 is primarily associated with glucose metabolism in brain tissues[38, 40, 41].
HFE serves as a gene target for iron chelator drugs. Our data analysis indicates that activation of the HFE gene target can reduce the risk of hip osteoarthritis. It is noteworthy that, although iron chelators are not antidiabetic medications, research suggests that they can activate the HFE gene target, leading to an increase in HbA1c expression[42, 43]. The development of osteoarthritis may be likewise correlated with mutations of the HFE gene[24]. Therefore, we included HFE in this analysis, and the results show a negative correlation between HFE and hip osteoarthritis. Through our discussion, we propose that HFE may influence the relationship between osteoarthritis and HbA1c concentration by regulating HbA1c levels. However, further research is needed to explore whether HFE affects osteoarthritis by influencing liver or other tissue iron concentrations due to the lack of detailed data on iron levels.
Analysis of the TCF7L2 gene revealed that, before removing confounding factors, activation of the TCF7L2 gene increased the risk of hip OA (OR 3.548, 95% CI 1.203–10.462), but the risk of knee OA (OR 0.160, 95% CI 0.063–0.408) decreased. After removing confounding factors related to BMI using the PhenoScanner database, TCF7L2 lost its association with hip and knee osteoarthritis. Therefore, we speculate that reducing BMI may be the mechanism by which antidiabetic drugs targeting TCF7L2 (rs7903146) decrease the risk of osteoarthritis[44, 45]. (Supplementary Table S2)
Notably, activation of INS, as a target for insulin, increased the risk of knee OA and OA at any site. For this result, we list two possibilities: firstly, some research has shown that cellular homeostasis and the metabolic balance of an organization can be maintained through a mechanism such as autophagy. However, insulin can decrease the level of autophagy, leading to cartilage degradation and ultimately an increased risk of osteoarthritis[46]. Second, after we removed confounders, only four instrumental variables were ultimately included in the analysis, and the results of the data analysis showed wide confidence intervals, suggesting that there may be uncertainty in the conclusions. Further validation of this result requires additional randomized controlled trials characterized by high quality and large sample sizes.
In previous analyses of the relationship between related glycaemic indicators and osteoarthritis, LR Chen demonstrated a positive correlation between HbA1C and knee osteoarthritis[47]. ZY Cui's study concluded that there was no causal relationship between FG and OA[48]. These two results are different from the conclusions drawn in the present study, which may be related to differences in the experimental inclusion of the data, which were obtained using data published by Chen J et al. in the MAGIC database in 2021 published in the MAGIC database, which included newer data and a larger sample size than the previous study.
Study Strengths
First, the Mendelian Randomization analysis minimizes the impact of survival bias and reverse causation that may occur in conventional randomized controlled trials. Second, compared to typical randomized controlled trials, MR analysis benefits from an ample sample size and abundant data. Third, to minimize the influence of different populations, this analysis included only participants of European ancestry. Fourth, to ensure the reliability of the experimental results, we conducted heterogeneity and horizontal pleiotropy analyses.
Study Limitations
Our study has several limitations. First, we used GWAS data from European ancestry populations, so the generalizability of the results to other populations remains to be verified. Second, despite conducting analyses for horizontal pleiotropy and heterogeneity, we cannot completely eliminate the influence of confounding factors and horizontal pleiotropy. Third, we cannot rule out the possibility of false positives in the ability of antidiabetic drugs to reduce the risk of osteoarthritis. Fourth, we selected six target points for antidiabetic drug action, but there may be other gene targets. Fifth, we were unable to obtain information on drug dosage and duration of use through MR analysis for their impact on osteoarthritis.
In conclusion, the results of this MR study suggest that GCK, HFE, HK1, and HKDC1 may be potential therapeutic targets for osteoarthritis, and glucokinase activators and metformin, among other antidiabetic drugs, may benefit osteoarthritis patients. It is essential to underscore that this study can only provide a possible clinical direction, indicating that the use of antidiabetic drugs may help reduce the risk of osteoarthritis. However, whether antidiabetic drugs can be repurposed for osteoarthritis treatment requires validation through numerous randomized controlled trials. If glucokinase activators and metformin, among other antidiabetic drugs, are confirmed to be effective for osteoarthritis treatment in clinical practice, this would contribute to improving early symptoms, alleviating disease progression, reducing disability rates, and significantly decreasing the economic burden on patients and society.