Why did we conduct this study?
Osteoporosis is a common disease where bones become weaker, increasing significantly the risk of fractures, even from minor falls or bumps1-3. It is estimated that around 30 million people in the European Union have osteoporosis, being particularly common among postmenopausal women and the elderly. Several effective treatments have been developed during the past years. Guidelines generally recommend antiresorptive treatments, which focus on slowing down bone breakdown, a process accelerated in people with osteoporosis. For more severe cases, treatments that stimulate bone formation are recommended. Sclerostin inhibitors are part of this family of anabolic agents, with Romosozumab being the only medicine in this class currently approved for the treatment of osteoporosis.
Sclerostin is a protein made by bone cells. By blocking it, bones can grow stronger and hence, the risk of fractures can be drastically reduced4,5. After four large studies6-10, Romosozumab was approved as a medication to treat severe osteoporosis in most countries worldwide. However, the previous studies raised some concerns about possible heart-related side effects, so the European Medicines Agency (EMA) added restrictions to reduce any potential risks11. Nonetheless, even with further research performed after Romosozumab was approved, the conclusions about its heart safety remain unclear12-14.
To assess the safety of medicines, genetic data is increasingly being used through an approach known as “triangulation of evidence”. This means using different methods, study designs, and/or populations to answer the same research question. If all these various studies lead to similar findings, we can feel more confident in our conclusions. A popular approach is Mendelian Randomisation (MR), which uses genetic variation to simulate a drug’s effect and estimate its impact on specific health outcomes, like heart-related events. Thereby, we conducted an MR study to estimate how different levels of sclerostin might influence heart health risk factors.
How did we do this?
Imagine we want to calculate the age of a tree. Unless the tree decides to let us know its age directly -which is quite unlikely-, this can be tricky to estimate directly. However, we know that each ring in a tree trunk represents one year of growth, so counting these rings gives us an accurate estimate of the tree’s age. In this case, we are using the number of rings as a measurable indicator (or “instrument”) for the tree’s age.
Mendelian randomisation uses a similar approach, but in this case we use genetic variations as instruments to help us calculate certain traits that might be difficult otherwise. What makes genetic variants especially powerful as instruments is that they are randomly distributed in the population. Let us go back to our study, where we were interested in studying the causal effect of sclerostin levels on cardiovascular disease and risk factors. We first started by identifying which genetic variants make people genetically predisposed to express low levels of sclerostin. For example, one group of people might carry a genetic variant (we can call it Variant A) that tends to make people have lower sclerostin levels, while another group of people might have a different variant (Variant B) which predisposes them to have higher levels of sclerostin.
By examining these variants in a large sample of people, we can identify patterns: people with Variant A may consistently have lower sclerostin levels, whereas those with Variant B may have higher levels. And, because these groups are formed randomly (as genetic variants are assigned randomly), we can be confident that any differences between them are likely due to the specific genetic variant they carry and hence, due to sclerostin levels.
Once we found these genetic determinants of sclerostin levels, the next step is to determine if they are associated with differences in the risk of a disease of interest. That means, do people with genetic variant A (associated with lower levels of sclerostin) tend to have a higher frequency of cardiovascular disease compared to those with variant B (associated with lower levels)?
What did we find?
We first explored the effect of sclerostin levels on bone mineral density and fracture risk. As we already know (based on previous evidence), that lower sclerostin levels will increase bone mineral density and decrease fracture risk, we used these outcomes to validate our approach and the reliability of our genetic instruments. Our findings were aligned with previous knowledge: lower sclerostin levels were associated with a significant increase in bone mineral density and an 85% reduction in hip fracture risk.
Our study also found that lower levels of sclerostin were causally associated with a 25-85% increased risk of coronary artery disease, and a 40% to 60% increased risk of type 2 diabetes. Our study also showed that lower levels of sclerostin increased levels of triglycerides in blood, known to increase cardiovascular risk.
Does this mean that taking a drug to lower sclerostin levels will increase cardiovascular disease risk? The short answer is “We still do not know”. The longer answer is a bit more complex. Genetic variants are fixed from birth, so in this study, we explored the lifetime effects of having lower sclerostin levels due to genetics. In contrast, sclerostin inhibitors are typically given for months and prescribed at a specific point in time. Additionally, sclerostin inhibitors are prescribed to people with severe osteoporosis, while in our study we included a general population. There are other aspects to consider when interpreting our findings. For a deeper understanding, please refer to the full publication.
What is the take home message?
Our study supports previous research suggesting that lower levels of sclerostin may increase heart-related disease risk. Notably, while earlier studies were conducted across different populations and used different methods -each one with its own limitations-, we have pointed out to the same direction. Further research is essential to understand the biological mechanisms and pathways underlying the association between sclerostin inhibition and cardiovascular disease.
References
1. Compston JE, McClung MR, Leslie WD. Osteoporosis. Lancet. 2019;393(10169):364-376.