Prednisone

Prednisone is a powerful anti-inflammatory that works by suppressing the immune system. It can help quickly, but the evidence shows it comes with serious, dose-dependent risks that are often downplayed.

The evidence here is a mixed bag. One case report shows a patient with cancer-related arthritis who got a "good clinical response" from prednisone, but his symptoms came back as soon as the dose was lowered [1]. That is the pattern: prednisone treats symptoms, not causes, and the body can become dependent on it. The real question is what it costs you.

The most concrete risk numbers come from a 2022 retrospective review of over 113,000 adults who received oral corticosteroids [2]. Among the 217 confirmed cases of avascular necrosis (AVN) — bone death from disrupted blood flow — the average cumulative prednisone dose was 3,314 mg. That is roughly 10 mg per day for a year. The average time to AVN diagnosis was over three years. AVN is not rare at higher doses: 81.6% of cases involved the femoral head (the hip joint), and 22.6% of those patients had a pathologic fracture [2]. A separate study in children with lupus found that a maximum daily dose of about 1.25 mg per kg of body weight was linked to AVN [3].

The risk of blood clots is also substantial. A 2021 self-controlled case-series study found that during glucocorticoid treatment, the risk of a first venous thromboembolism (VTE) was about 3.5 times higher than normal [10]. In the first week of treatment, that risk jumped to over 5 times higher [10]. For people who already had a clot, the risk of a second one during steroid treatment was 7.5% per year, compared to 2.8% per year without steroids [10].

The evidence also shows that even short courses of prednisone suppress your adrenal glands. A 1979 study of cancer patients found that after just one week of high-dose prednisone, most had suppressed adrenal function for at least 24 hours, and some took up to a week to recover [8]. This means your body cannot produce its own stress hormones, which can be dangerous if you get sick or injured.

The studies here are mostly observational, not randomized trials. The AVN study is a single-institution retrospective chart review [2]. The VTE study used a self-controlled design, which is stronger but still observational [10]. No study here compared prednisone against a placebo for the conditions most people take it for. The evidence is enough to say the risks are real and dose-dependent, but not precise enough to say exactly where the line is for any individual.

My call: prednisone is a short-term rescue drug, not a long-term solution. The risks of bone death, blood clots, and adrenal suppression are real and increase with dose and duration. Use the lowest dose for the shortest time possible, and have a plan to taper off. Confidence: moderate.