The researchers also replicated the findings in mice, compressing the animals’ sciatic nerves to produce back and leg pain or injecting the sciatic nerves with an irritant. When they blocked the animals’ immune response with dexamethasone, a steroid commonly used to treat back pain, the pain became chronic.

Then, the group questioned whether chronic pain resulted from pain suppression or from suppression of inflammation. So they gave some mice a prescription anti-inflammatory, diclofenac. Other mice got one of three other analgesic, or pain-relieving, drugs — gabapentin, morphine and lidocaine.

Only with diclofenac did the pain persist, becoming chronic.

Those results led them to ask: Were patients who took nonsteroidal anti-inflammatories like ibuprofen or steroids like dexamethasone to relieve their back pain also more likely to develop chronic pain?

The researchers turned to data from the UK Biobank, a repository with information about half a million patients’ medical conditions and drug use. They studied 2,163 people with acute back pain, 461 of whom went on to have chronic pain. Those taking a nonsteroidal anti-inflammatory had nearly double the chance of developing chronic back pain as those taking other drugs or no drugs, the researchers found.

Dr. Diatchenko said she does not think her findings bear on the issue of opioid addiction. In fact, she said, “to avoid opioids, clinicians started to prescribe more nonsteroidal anti-inflammatory drugs.”

“We need to think further about how to treat our patients,” she said.

The tendency to use nonsteroidal anti-inflammatories persists despite their unimpressive performance. An analysis of randomized clinical trials found that these drugs had almost no benefit over placeboes in reducing low back pain.

Dr. Atlas says that short term use of nonsteroidal anti-inflammatories probably is not harmful, but the new study, he adds, while not proving long term use is harmful, “at least gives a biological mechanism that says short term use is not the same as long term.”