Findings from a Mendelian randomization (MR) study have shown that 2 metformin targets, AMP-activated protein kinase (AMPK) and growth differentiation factor 15 (GDF-15), can be potential therapeutic targets for osteoarthritis (OA), particularly hip OA. Results of the study were published in Osteoarthritis and Cartilage.

The researchers sought to provide causal evidence on the effects of metformin on OA via the use of AMPK and GDF-15. These 2 metformin targets were used to mimic the effect of metformin use on OA outcomes. Utilizing this 2-sample MR design, they constructed 44 AMPK-related variants genetically predicted in hemoglobin A1c as instruments for AMPK and 5 single-nucleotide polymorphisms that were significantly associated with GDF-15.

A total of 455,211 participants of European descent were included in the study. Summary-level data obtained from the largest genome-wide meta-analysis across the UK Biobank and arcOGEN were used for 3 phenotypes of OA, including OA at any site (n=77,052), knee OA (24,955), and hip OA (n=15,704), as well as for 378,169 healthy control participants.

Continue Reading

Results of the study showed that genetically predicted AMPK was negatively associated with OA at any site (odds ratio [OR], 0.60; 95% CI, 0.43-0.83; P =.002) and with hip OA (OR, 0.42; 95% CI, 0.22-0.80; P =.008), but not with knee OA (OR, 0.85; 95% CI, 0.49-1.50; P =.585).

Higher levels of genetically predicted GDF-15 significantly decreased the risk for hip OA (OR, 0.95; 95% CI, 0.90-0.99: P =.039), but not the risk for OA at any site (OR, 1.00; 95% CI, 0.98-1.02; P =.904) or the risk for knee OA (OR, 1.02; 95% CI, 0.98-1.07; P =.310).

Limitations of the study included the fact that although the investigators used 2 targets of metformin, they were “unable to assess the overall effect of metformin…because metformin may [have influenced] OA via multiple protein targets, or even through the gut microbiome as reported in other diseases.” The summary statistic of the study was based on European populations, thus limiting the generalizability of the results.

The study authors concluded that the findings from the current study can provide only a hypothesis that metformin use might be useful in reducing an individual’s risk for OA. Randomized controlled trials are warranted in order to explore whether metformin can be repurposed for OA therapy.


Zhang Y, Li D, Zhu Z, et al.Evaluating the impact of metformin targets on the risk of osteoarthritis: a Mendelian randomization study. Osteoarthritis Cartilage. Published online July 6, 2022. doi:10.1016/j.joca.2022.06.010