What does an MRI Scan Say about Future Back Pain?
If the MRI scan findings are important in low back pain, we could expect them to predict how our back pain is going to be in the future. That doesn’t seem to be the case.
A group of people were given MRI scans and then followed up 10 years later.
This showed the findings on MRI scans (disc bulges, high intensity zones, modic changes and spondylolisthesis – all presumed indicators of back pain) did not predict the back pain the patients had in the following ten years.
The frequency of disc degeneration increased from 50% to 85% across the ten years, which was expected. And 77% of people who had no back pain during that time also showed disc degeneration. Similar results were found for the high intensity zones, disc bulges and modic changes.
Overall, the back pain over the following ten years was not predictable by looking at the initial MRI scans. And many people had “abnormalities” on their MRI scans but had no back pain.
This questions the idea that MRI scanning gives us objective information about our backs. And that the results mean anything for our future pain or disability. And that “abnormalities” condemn us to suffer back pain or neck pain in the future. They don’t seem to.
Many of the findings seem to be incidental, perhaps indicators of how long we have been alive, rather than a pain-producing pathology.
This also makes it hard to connect MRI scan results with a person’s back pain. If disc bulges and all the other changes are present in people without back pain and don’t mean anything for future back pain, how can we be reasonably certain they are relevant?
And even if the changes are relevant to the pain, how will they alter the treatment? If they don’t change the treatment, this may be an example of spurious diagnostic accuracy, an accurate diagnosis that means little.
Tonusu, J., et al. (2017). The associations between magnetic resonance imaging findings and low back pain: A 10-year longitudinal analysis. PLoS ONE 12(11): e0188057. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0188057