Corticosteroid Adverse Effects, Organised by Body System

Most drug adverse-effect lists are a soup of maybes. The corticosteroid list is unusual, because several entries arrive with a published dose or duration threshold attached, and a threshold is what turns a warning into an answerable question.

Sort them by body system and the list shrinks to something you can hold.

The mechanism underneath all four systems

One drug producing bone loss, ulcers, cataracts and psychosis looks arbitrary until you remember what a corticosteroid is imitating. It stands in for a hormone with receptors almost everywhere, so the adverse effects are distributed the same way.

That is why the list sorts by system rather than by frequency. The body system does the organising work, and the drug is simply present in all of them.

Systemic hormone, systemic consequences.

Bone: the risk with a number attached

Long-term use carries fracture and bone-loss risk, and StatPearls puts that risk in up to 40% of long-term users. That is a large share, and it is the figure worth carrying.

Trabecular bone goes first, and the same source describes that loss beginning within 6 to 12 months of therapy. So this is not a decade-long problem. It is a first-year problem.

Avascular necrosis carries its own threshold in that source: risk rises with prednisone above 20 mg/day. Hip or groin pain in someone on long-term steroid is not a musculoskeletal footnote.

Steroid-related bone loss is also silent until it is not. Nothing announces trabecular thinning, which is why the question gets asked on the basis of exposure rather than on the basis of complaint.

Ask about it directly.

Gut: the risk that lives in a combination

On its own, steroid-related gastritis, ulceration and bleeding is a real but background concern. Add an NSAID and the same source puts the risk roughly four times higher.

That is the exam-relevant fact, because a stem will hand you both drugs. A patient on prednisone who takes ibuprofen for the same joint pain is a combination, not a coincidence.

Reported risk here is comparative rather than absolute, and four times higher than what depends on the baseline patient. Comparative figures still change practice. They just do not belong in a sentence pretending to be a rate.

Over-the-counter analgesia is the blind spot. Patients do not report what nobody prescribed, so the question has to be asked as a list rather than as a yes or no.

Eyes: two thresholds and a condition nobody names

Cataracts come with the most specific threshold in the set, which is prednisone above 10 mg/day taken for at least a year. Glaucoma is named as well, and so is central serous chorioretinopathy, which is the one most students have never heard of.

Notice how precise the ophthalmic thresholds are compared with the rest of the list. Dose and duration both appear, which means an item can test either half.

Vision change in a long-term steroid patient earns a referral, not reassurance.

Mind: quote the hedge rather than flattening it

Psychiatric effects run from euphoria at one end to frank psychosis at the other. The source is careful about the severe end: psychosis occurs almost exclusively at prednisone above 20 mg/day, taken for prolonged periods.

Keep the words almost exclusively. Dropping them turns a careful observation into an absolute, and the absolute is what produces a wrong answer when a stem describes a low-dose patient with a mood change.

The euphoria end matters too. Someone who feels unusually good on a new steroid course is showing you a drug effect, not a personality.

Mood change at any dose is worth reporting. Psychosis at any dose is worth escalating. A threshold describes where risk concentrates, not where it begins.

Dose and duration are two separate questions

Read any of these thresholds carefully and two variables appear. Prednisone above 20 mg/day is a dose statement. For at least a year is a duration statement. Cataract risk needs both.

Stems exploit that. Somebody on a high dose for a week and somebody on a low dose for two years are different risk profiles, and only one of them is the answer.

Ask both questions every time. How much, and how long.

Why these thresholds are worth memorising when most are not

We tell students to stop memorising numbers they cannot cite. This list is the exception, because each figure traces to one verified source and each one changes an action.

A threshold that changes nothing is trivia. A threshold that decides whether you ask about hip pain, eye appointments or an over-the-counter painkiller is a nursing assessment in disguise.

Learn what the number does. The digits follow.

A monitoring plan you can build straight from the list

Building a monitoring plan out of the mechanism instead of out of a list is the wider method set out in drug class questions start with the mechanism.

What is not in this post

Stopping the drug is the other half of steroid safety, and it runs on a different mechanism entirely, which is the suppressed axis rather than accumulated exposure. That case is argued in corticosteroids and why tapering is not optional.

Worth separating too: a steroid-induced psychiatric effect is not the same animal as a drug-induced movement syndrome. If a stem gives you fever with rigidity you are in different territory, described in neuroleptic malignant syndrome and its classic tetrad.

And when the drug in the stem is an antipsychotic rather than a steroid, the class-wide warning has its own careful wording and its own pair of dates, covered in antipsychotics and the dementia boxed warning.

One caution about all of these figures. They describe risk in populations on long-term therapy. Long-term is doing quiet work in every sentence above, and none of them predicts what happens to the one patient in front of you.