Current streams of research in the Musculoskeletal Health Research Group

​​​​​​​What contributes to musculoskeletal disease and can it be targeted with specific intervention?

This aim of this stream is to use big data and data science techniques to improve the management of musculoskeletal conditions. Our vision is to create targetted diagnostics for people with muscle and joint pain to then enable individualised, precision, treatments. 

We secured funding to evaluate, using machine learning methods, a larger sample of people with (300 people) and without (100 people) back pain to see which of a number of body systems, including intervertebral disc and muscle tissue, pain thresholds, brain MRI, mental health and social factors, muscle strength and endurance, contribute the most to back pain (study website LINK, study protocol LINK). 

Along this path, we published a series a systematic reviews (LINK) examining the state of the science on artificial intelligence and machine learning in back pain to identify what needs to be done. We also evaluated, via systematic review and meta-analysis, how much psychosocial, brain and central nervous system and spine tissues contribute to back pain (LINK). Further, we utilised data from the UKBiobank to see if sub-group (phenotypes) exist in people with back pain (LINK, LINK) and conducted a pilot study to see how much of a contribution is made to back pain by changes in the brain and spine (via MRI), psychological factors, and spine function (LINK). 

Aligned with this work, we have an ongoing project measuring astronauts before and after spaceflight to see what changes happen in the body that might be behind the increased risk of neck disc herniations that the astronauts have (study protocol LINK, study website LINK). 

Optimising and personalising management of musculoskeletal disease

The aim of this stream is to evaluate effective treatments for musculoskeletal conditions. This includes the identification of optimal modalities, as well as how these modalities are best personalised at the individual patient level. We consider systematic reviews and meta-analyses, as well as randomised controlled trials (RCTs), that can inform evidence-based clinical practice and the guidelines within these settings. Currently we focus more on evaluating medical (e.g. invasive treatments, medication) versus conservative non-pharmacological appraoches as well as critically evaluating how well different treatments work and why.

Some examples of this include:

• How much exercise dose do I need for back pain? (LINK and an ongoing project)  
• How placebo effects influence how well treatments work (LINK, LINK, LINK, LINK, LINK)
• A network meta-analysis that identified optimal modes of exercise training for low back pain (LINK)
• Running exercise for treating chronic back pain (RCT) (LINK, LINK)
• Exercise for neck pain (LINK) or after hip replacement (LINK)
• Telemedicine for patients with musculoskeletal pain (LINK)
• Does adding exercise to pharmacological treatment for osteoporosis help? (LINK)
• Implementation of evidence based guidelines in clinical practice (LINK, LINK, LINK)
• Randomised control trial in back pain of showing two types of commonly performed, but different modes of exercise had a similar impact on clinical outcomes (LINK), and depressive symptoms (LINK), but may differentially impact the tissues in the spine, such as the musculature (LINK), intervertebral discs (LINK), and bone marrow (LINK)
• Whether current treatments that rely on classifying patients with back pain work (or not) (LINK)
• Surgery (or not) for tears of the anterior cruciate ligament in the knee (LINK)