Given that approximately one-third of patients with HF also have diabetes mellitus [30, 31], it becomes crucial to make informed decisions in selecting hypoglycemic agents to delay the progression of HF in these patients. To our knowledge, this is the first MR study investigating the potential effects of metformin on HF outcomes, using genetic instrumental variants from the two pharmacological targets (AMPK and GDF-15). The study findings do not provide convincing evidence that metformin reduces the risk of HF by activating the AMPK pathway or increasing GDF-15 expression. Further research is needed to explore whether metformin may mitigate the risk of HF through other biological mechanisms.
Cardiac remodeling is a critical process in the progression of HF, involving myocardial cell hypertrophy, inflammatory response, and cardiac fibrosis[32]. AMPK has emerged as a significant cellular energy sensor and regulator, drawing attention for its potential therapeutic effects on HF. Animal studies have shown that metformin activates the AMPK pathway, leading to the mitigation of cardiac remodeling and the prevention of HF progression[12, 33, 34]. Moreover, research by Yen-Chun Lai et al. showed that metformin improved HF-related pulmonary hypertension by activating the AMPK pathway[35]. Additionally, Inflammation and stress are closely related to HF occurrence, with GDF-15, a stress-responsive cytokine, significantly increased in HF patients [36–38]. Animal studies have found that metformin promotes the expression of GDF-15, which controls triglyceride metabolism through liver and kidney adrenaline signal transduction to maintain cardiac function under stress or inflammatory states, while also inhibiting myocardial cell hypertrophy[39]. Hence, AMPK and GDF-15 are promising therapeutic targets for HF.
As an activator of AMPK and GDF-15, metformin appears to have potential biological effects that may reduce the risk of HF. However, conflicting results have been reported in clinical studies regarding the relationship between metformin and HF risk. A meta-analysis study found that metformin could reduce hospitalization and mortality rates in HF patients (HR: 0.90; 95% CI: 0.87–0.94) [16], and a real-world study also supported its beneficial effects on reducing HF risk (HR: 0.61; 95% CI: 0.52–0.73) [40]. Nevertheless, a prospective study involving 12,156 participants suggested no improvement in heart function, hospitalization, or mortality rates in HF patients after adjusting for relevant variables[40]. More recently, a randomized controlled trial with 3,234 patients showed that metformin did not reduce the incidence of cardiovascular events in individuals with impaired glucose tolerance[41].
To explore the causal relationship between AMPK and GDF-15, two potential pharmacological targets of metformin for treating HF, and the risk of HF, we conducted a two-sample MR study. MR employs genetic variation as instrumental variants to assess causality between exposure and outcome. Since genetic variation is independently distributed, randomly assigned, and not affected by time or disease progression, MR analysis can avoid confounding bias and reverse causality prevalent in observational studies[42]. Despite this advantage, potential pleiotropy of instrumental variants remains a concern[20]. To address this, we used the Egger-intercept analysis method to detect pleiotropy and excluded instrumental variants associated with potential confounders. Heterogeneity tests were conducted on instrumental variants. Employing five sensitivity analysis methods (IVW, weighted median, MR-Egger, simple mode, and weighted mode), all results consistently indicated that AMPK and GDF-15 are not causally related to the risk of HF. These findings inconsistent with previous animal experimental results[12, 35]. Possible reasons for this discrepancy include (1) Although animals such as dogs and rabbits share genetic similarities with humans, they may exhibit variations in terms of their pathological conditions, physiological needs, and behavioral patterns[43, 44]. (2) The inconsistent dosage of metformin across studies is a significant factor contributing to the discrepant effects observed. The mechanisms of action that have been widely studied, such as complex I inhibition leading to AMPK activation, have only been observed at supra-pharmacological concentrations (> 0.01mol/L) of metformin, which are not typically achieved in clinical settings[45].Therefore, the AMPK and GDF-15 pathways activated by metformin may be dose-dependent, and future research is needed to explore the relationship between metformin dosage and the risk of HF.
As the inaugural MR analysis exploring the causal link between metformin and HF outcome through AMPK and GDF-15, two pharmacological targets of metformin, our study acknowledges several limitations. Firstly, due to database constraints, bidirectional analyses were unattainable, precluding us from establishing whether HF occurrence leads to AMPK pathway activation and increased GDF-15 expression. Secondly, the study population comprised individuals of European descent, limiting the generalizability of our findings to other ethnic groups. Furthermore, our study did not investigate potential dose-response effects of metformin on the causal association between its usage and HF outcome. Lastly, while AMPK and GDF-15 are promising metformin targets for HF treatment, other unexplored pathways or biological mechanisms may also contribute to reducing HF-related risks.