Corticosteroid harm by dose band: what starts at 2.5 mg and what waits for 7.5

It is not right, and the threshold is lower than most people expect. The 2022 American College of Rheumatology guideline on glucocorticoid-induced osteoporosis applies to adults taking 2.5 mg a day or more for longer than three months. You are at twice that dose. The guideline strongly recommends an initial fracture risk assessment as soon as possible after starting, and in practice within the first six months, consisting of three components: a history covering both symptomatic and unrecognised fractures, the FRAX tool if you are 40 or over, and bone mineral density measurement with vertebral fracture assessment or spine radiographs. There is a reason the trigger sits so low. Bone loss on glucocorticoids is fastest in the early months of treatment, and it happens through mechanisms that do not need a large dose - reduced bone formation, increased resorption, reduced calcium absorption and altered sex hormone production. Fracture risk rises before density falls far enough to look alarming on a scan, which is why the guideline asks for fracture history and FRAX alongside the density measurement rather than relying on the number alone. Two technical points are worth knowing before the appointment. FRAX is not validated below age 40, so if you are younger the assessment leans on history and density rather than a calculated score. And above 7.5 mg a day, the FRAX result needs a glucocorticoid dose correction, with the guideline noting that even the corrected figure may understate the real risk. At 5 mg you are below that correction threshold but well above the assessment one. What follows from the assessment depends on where you land. Everyone in this group is advised to optimise calcium and vitamin D intake, dietary or supplemented, alongside the usual lifestyle measures. For anyone assessed at medium, high or very high fracture risk, pharmacological treatment is strongly recommended, and the choice between oral or intravenous bisphosphonates, denosumab and parathyroid hormone analogues is made through shared decision-making rather than by protocol. For someone at low risk who is not started on treatment, the recommendation is reassessment every one to two years while the steroid continues. It is worth being specific with your prescriber rather than asking generally whether a scan is needed. Naming the guideline, the 2.5 mg threshold and the three-month duration usually changes the conversation, because the objection almost always comes from a mental model in which 5 mg is trivially low. It is not trivially low for bone. It sits inside the range the Endocrine Society describes as increasing cardiovascular disease, severe infection, hypertension, type 2 diabetes and overall mortality. One last thing to bring: the duration. If your lung condition has needed steroids on and off for years, the cumulative exposure is what matters, and nobody is keeping that running total unless you are. Take the dates and doses with you. Two things are worth separating out so the appointment does not collapse into one question. Assessment and treatment are different decisions: having a fracture risk assessment does not commit you to bisphosphonates, and plenty of people at 5 mg turn out to be at low risk and need nothing beyond calcium, vitamin D and a reassessment in a year or two. The assessment exists to tell you which group you are in, and at present nobody knows. Second, ask whether the lung condition itself contributes, because chronic inflammatory lung disease, reduced activity and low vitamin D all push in the same direction as the steroid, and the fracture risk calculation is meant to capture the whole picture rather than the tablet alone. If a scan does show low bone density, that finding is also an argument to revisit whether 5 mg is still the minimum effective dose for your lungs. Background on this page is on my reviewer page.

Because the fasting test is the measurement least likely to detect the problem you are being monitored for. Steroid hyperglycaemia from once-daily morning prednisolone has a characteristic daily shape: glucose climbs through the afternoon and evening, peaks some hours after the dose, and drifts back towards normal overnight. By the time you test before breakfast the following morning, the curve has come down. A run of normal fasting values is entirely compatible with afternoon readings well above target, and that is precisely the pattern the nurse is trying to catch. The Joint British Diabetes Societies guidance is unusually specific about where to look. Capillary glucose is checked once daily, either before lunch or the evening meal or one to two hours after them, because that is when the hyperglycaemic effect of morning steroid dosing is greatest. If a reading comes back above 12.0 mmol/L, testing increases to four times a day. If readings exceed 12.0 mmol/L on two occasions within 24 hours, treatment is considered, commonly gliclazide starting at 40 mg in the morning and titrated upward. If values stay consistently under 10 mmol/L, testing can stop. A hospital study looking at this specifically concluded that routine monitoring between 4 and 8 am adds little. HbA1c has the same blind spot for a different reason. It averages glucose over roughly three months, so it lags a rise that develops within days of starting steroids, and a normal result early in a course means very little. Where it is genuinely useful is before treatment starts, to identify people who already have undiagnosed type 2 diabetes and are therefore at higher risk. If you did not have one done beforehand and your course is a long one, it is a reasonable thing to ask about for after it finishes rather than during. The other half of this that patients are rarely warned about is the way down. Steroid-induced hyperglycaemia falls as the dose falls, and glucose-lowering treatment started at a high steroid dose becomes an overtreatment risk at a low one. Published guidance gives a rough guide of a 20 to 25 per cent insulin reduction, or a 40 mg gliclazide reduction, for each 5 mg drop in prednisolone from 20 mg. If your steroid dose is due to come down, ask at the same appointment what happens to any glucose treatment when it does, and keep testing through that period specifically to catch lows. One practical suggestion. Record the time of the reading alongside the value, not just the value. A logbook of numbers without times is close to useless in this situation, because interpretation depends entirely on where the reading sits relative to the morning dose and the last meal. Note the steroid dose on the same line if it is changing. That single habit turns a set of numbers into something a clinician can act on, and it makes the eventual conversation about reducing treatment much shorter. Two further points that patients on steroid courses ask about. Symptoms are an unreliable guide here: thirst, frequent urination and fatigue arrive late and are easily attributed to the illness being treated, so their absence is not evidence of normal glucose. And if you had no diabetes before the steroid started, the sensible time to check for it properly is about three months after the course finishes and readings have settled, when HbA1c becomes interpretable again. Steroid-induced hyperglycaemia usually resolves when the drug stops, but a proportion of people turn out to have had undiagnosed type 2 diabetes that the steroid merely unmasked, and that group needs following up rather than discharging. Ask now for that check to be diarised, because it is the step most often lost once the steroid course ends and everyone's attention moves on.

It depends on which shingles vaccine, and this is a distinction worth getting right because the two products behave completely differently. The threshold that governs live-virus vaccination is 20 mg or more of prednisone daily, given for 14 or more consecutive days - the CDC's definition of high-dose corticosteroid therapy, based on what most clinicians consider sufficiently immunosuppressive to raise safety concerns. At 15 mg you are below that line. Where the therapy does meet the threshold, live vaccination is deferred until at least a month after the steroid stops. The refusal may still have been reasonable for other reasons. Dose is not the only input: the underlying condition, other immunosuppressive drugs, and the total picture all matter, and 15 mg alongside methotrexate or a biologic is a different situation from 15 mg alone. A pharmacist working from a simple screening rule rather than the full picture will decline, and referring the question back to the prescriber is the correct next step rather than shopping for a different pharmacy. What frustrates me is the volume of unnecessary refusals at the other end of the scale. Steroid therapy is not considered sufficiently immunosuppressive to contraindicate live vaccines when it is short-term at under 14 days, when the dose is below 20 mg a day, when it is long-term alternate-day treatment with short-acting preparations, when it is physiological replacement, or when it is topical, inhaled, intra-articular, bursal or intra-tendinous. Inhaled steroids in particular get patients turned away every year with no basis whatsoever. Timing runs in both directions. If immunosuppressive treatment is about to start, the general advice is to wait four weeks after a live vaccine and two weeks after a non-live one before beginning - with the important caveat that treatment for an active inflammatory condition should not be delayed because of past vaccination. For non-live vaccines there is no safety issue at any steroid dose; the concern is reduced response, since glucocorticoids above physiological doses blunt immunogenicity. That affects how well the vaccine works, not whether it is safe. Two things to do with this. Ask your prescriber directly which product was being offered and whether it is live, because the answer changes everything and the record should reflect the reasoning rather than a bare refusal. And separately, make sure someone is thinking about infection risk more broadly, since corticosteroids mask the signs of infection as well as increasing susceptibility - the label warns that new infections may appear during use and that the usual signals can be blunted. On 15 mg, a fever that seems mild deserves more attention than it would otherwise get, not less. Timing is the other lever, and it is underused. If your steroid dose is expected to rise, or a course of immunosuppressive treatment is being planned, the window to vaccinate is before it starts rather than during. General guidance is to allow four weeks between a live vaccine and the start of immunosuppression, and two weeks for a non-live one. Where a dose is coming down instead, live vaccination is deferred until at least a month after high-dose therapy stops. Either way, somebody has to look at the calendar, and in practice nobody does unless the patient raises it. Bring the question to the appointment where the steroid plan is being discussed rather than to the vaccination appointment, since that is where the dates are actually decided. Ask as well what else is due. People on long-term steroids are usually candidates for the seasonal influenza and pneumococcal vaccines, neither of which is live and neither of which raises the safety question you ran into. Getting those arranged at the same visit is a better use of the appointment than a second argument about the shingles product, and it addresses the risk that matters most at 15 mg, which is infection rather than vaccination.

I would want an ophthalmological assessment before accepting a new prescription, and the symptom pattern you describe is the reason. Posterior subcapsular cataract is the lens change specifically associated with corticosteroids, and it affects near vision and glare disproportionately - haze around oncoming headlights, difficulty in bright sunlight, reading vision degrading faster than distance. It is easy to mistake for the ordinary presbyopia of middle age, and a stronger prescription can mask it for a while without addressing what is happening. The dose does not protect you. Cataract had the lowest threshold of any adverse effect measured in the 1066-patient German cohort, appearing below 5 mg a day. A later systematic review of glucocorticoid use in rheumatoid arthritis placed both cataract and glaucoma in the threshold category, with cataract at around 5 mg daily and glaucoma at around 7.5 mg in patients treated beyond six months. In older cross-sectional series, posterior subcapsular opacities were found almost exclusively in steroid-exposed patients. At 10 mg for eight months you are well inside the exposure range where this is expected rather than unusual. There is a second reason to see an ophthalmologist rather than an optician, and it is the one that actually worries me. The label lists raised intraocular pressure and glaucoma with possible damage to the optic nerve alongside cataract. Steroid-induced ocular hypertension is typically silent until damage has occurred, which means it does not present with symptoms you would notice and report. Someone already booked in for a vision complaint should have pressure measured as part of the same visit, and at 10 mg for eight months you are above the dose band where glaucoma frequency rises. Practical points for that appointment. Say clearly at the start that you are on long-term oral corticosteroids, give the dose and the duration, and mention any earlier courses, because cumulative exposure is what drives lens change and it is not visible in a current medication list. Ask specifically about pressure as well as about the lens. If a cataract is found, it does not necessarily mean stopping the steroid - that decision balances the eye against whatever the drug is treating - but the finding does belong in the conversation with your prescriber, since sight-threatening complications are among the reasons a more rapid reduction towards physiological dosing gets considered. One expectation to set honestly. Most steroid side effects settle after the drug stops. Lens opacities do not reverse. That is not a reason for alarm at this stage, since cataract is treatable surgically and very successfully, but it is a reason to have it looked at now rather than after another year of assuming it is age. One practical point about the referral itself. An optometrist can measure intraocular pressure and can see a posterior subcapsular opacity on examination, so the high-street appointment is not wasted - but the request needs to be explicit. Saying that you are on long-term oral steroids and would like the lens and the pressure checked, rather than presenting for a routine refraction, changes what gets done and what gets written in the report. Take that report to your prescriber afterwards rather than filing it. And if driving at night has become difficult, treat that as a safety matter in its own right while the assessment is arranged, since glare disability from a subcapsular cataract can affect night vision well before standard letter-chart acuity looks abnormal. Do not stop or reduce the steroid on your own account while this is being sorted out. Eight months of treatment at 10 mg means the adrenal axis is suppressed, and an abrupt stop at that exposure carries its own serious risk. If the eye findings do argue for coming down faster, that decision belongs with your prescriber and is carried out on a schedule rather than overnight. The reduction side of this is covered in the taper briefing.

There is, and it is the single most consequential interaction on the prednisone list for someone in your position. Corticosteroids on their own carry a modest ulcer risk. Corticosteroids taken with a non-steroidal anti-inflammatory carry a serious one, and the difference between those two statements is much larger than most patients realise. The numbers make the case better than adjectives. A nested case-control study in the Tennessee Medicaid population found a relative risk for peptic ulcer disease of 2.0 among current oral corticosteroid users overall - but 4.4 in those also taking NSAIDs, against 1.1 in those who were not. Put differently, people taking both drugs had about fifteen times the ulcer risk of people taking neither. A Swiss claims analysis reproduced the shape of that finding with an adjusted odds ratio of 1.63 for glucocorticoid alone and 4.80 for the combination. And in an analysis of seven European databases covering roughly 20 million people, concurrent non-selective NSAID and corticosteroid use produced the highest incidence rate ratio for upper gastrointestinal bleeding of any combination examined, at 12.8, with the largest excess risk beyond what either drug contributes alone. The reason your GP may not have raised it is worth saying plainly: over-the-counter ibuprofen does not appear on the prescribing record. From the practice computer's point of view you are on a steroid and nothing else. This is the most common way the interaction is missed, and it is entirely fixable by telling them. Take the actual packets to the appointment rather than describing them, since brand names for the same drug multiply and people frequently take two products containing the same ingredient without realising. On alternatives, paracetamol is the safer analgesic in this setting, though it is not free either - daily doses of 2 g and above have been associated with an increased risk of upper gastrointestinal complications in their own right, and the combination of high-dose paracetamol with a steroid has been measured too. Low-dose aspirin taken for cardiovascular protection adds risk as well and is often forgotten because patients class it as prevention rather than medication. Anticoagulants and SSRIs compound bleeding risk independently. What happens next is a prescriber decision rather than a pharmacy purchase. Gastroprotection is offered on the basis of the whole risk profile - the combination, previous ulcer or bleed, age, other drugs - rather than automatically to everyone on a steroid. The more important question is whether daily ibuprofen is the right treatment for your back at all, since chronic back pain is not a condition where continuous NSAID use performs particularly well, and the risk-benefit arithmetic changes completely once a steroid is in the picture. Raise it as a specific question rather than a general one, and mention the back pain duration - that combination usually produces a better plan than either issue discussed alone. In the meantime, know the symptoms that turn this from a discussion into an emergency. Black, tarry stools, vomiting blood or material resembling coffee grounds, sudden severe abdominal pain, or unexplained faintness and breathlessness all point at a bleed and need immediate assessment rather than a routine appointment. The reason to memorise them is that corticosteroids blunt the usual warning signs: the label notes that steroids may mask signs of infection, and the same masking applies to peritoneal irritation, so a perforated ulcer in someone on prednisone can present with far less pain and guarding than expected. Anyone in that position who feels suddenly and unaccountably unwell should be assessed on that basis alone, without waiting for the textbook presentation to appear. None of this means your back has to go untreated. Topical anti-inflammatory preparations carry far lower systemic exposure than tablets, and physiotherapy, graded activity and the other non-drug approaches to persistent back pain perform at least as well as continuous oral NSAIDs over months. Those options are worth putting on the table in the same conversation, because a plan that simply removes the ibuprofen without replacing it tends not to survive the first bad week.

Most of it will, and the two changes you describe have somewhat different timelines. Cushingoid redistribution - the rounding of the face, the fat pad at the back of the neck, the thinning of the limbs relative to the trunk - reverses over months once the glucocorticoid is reduced and stopped. It is slower than people hope and less complete in the first few months than they expect, but it is genuinely reversible. Weight gain is more of a shared responsibility between the drug's effect on appetite and fluid retention on one side and intake on the other, so it responds partly to stopping and partly to the same measures that work otherwise. Both effects belong in the linear group rather than the threshold one, which is why 7.5 mg was enough to produce them. In the German cohort of 1066 patients, cushingoid phenotype rose steadily with each milligram and showed no floor below which it stopped occurring. Weight gain appeared as a threshold effect from 5 mg a day. You are above both, and a year is ample time. Nothing about your experience is unusual or suggests something has gone wrong with the treatment. The parts of the harm list that do not fully reverse are worth knowing, because framing everything as temporary is a kindness that misleads. Lens opacities do not reverse; cataract, once formed, is treated surgically rather than by stopping the drug. Bone density recovers partially and slowly, which is why fracture assessment is recommended from 2.5 mg a day beyond three months rather than deferred until the course ends. Skin thinning and the tendency to bruise improve but can leave residual change in long-term users. Blood pressure, glucose, mood and sleep generally return to baseline. The practical question your description raises is whether 7.5 mg is still necessary, and that is a conversation worth having explicitly rather than waiting for the annual review to raise it. Reduction is only attempted when the underlying condition is controlled and the steroid is no longer required for it, so the answer depends entirely on what you are being treated for. But visible cushingoid change plus nine kilograms is exactly the sort of accumulating cost that should prompt someone to ask whether a steroid-sparing agent has a role, or whether the maintenance dose can come down. Note also that you sit at the edge of the band where glaucoma, low mood and rising blood pressure become more frequent. One thing not to do is reduce on your own account because of how you look. A year at 7.5 mg means your adrenal axis is suppressed, and stopping abruptly at that exposure carries a real risk of adrenal crisis. The mechanics of coming down safely, including why the last few milligrams take the longest, are set out in the taper briefing. Bring the weight and the facial change to the appointment as data rather than as a complaint - they are clinical findings and they carry weight in the decision. On timescale, set expectations low for the first stretch. Facial rounding typically takes several months to soften noticeably after the dose comes down, and it lags the dose reduction rather than tracking it, so the first eight weeks can feel like nothing is happening. Photographs taken at intervals are more useful than daily mirror checks for judging whether change is occurring. Weight responds to the same measures it would otherwise, with the caveat that appetite stimulation from the steroid is real and not a failure of willpower; people find it easier once the dose is falling. What I would not do is start a restrictive diet while still on a maintenance steroid dose with unassessed bone status, since adequate calcium, vitamin D and protein intake matter more in this group than in most.

Unlikely to be coincidence, and the mechanism is well described. Sodium retention, fluid retention, potassium loss and hypokalaemic alkalosis all appear in the prednisone label's adverse reaction list under fluid and electrolyte disturbances, with hypertension named directly. In the German cohort of 1066 patients, rising blood pressure was one of three effects that clustered as a threshold pattern above 7.5 mg a day, alongside glaucoma and depression. At 20 mg you are well above that band, and a month is a typical timeframe. A 22 mmHg rise is not trivial and it should not simply be absorbed. Two responses are reasonable and they are not mutually exclusive: treat the blood pressure, or reduce the steroid faster than originally planned. The second is a recognised option - guideline text specifically names uncontrolled hypertension alongside uncontrolled hyperglycaemia, steroid psychosis and herpetic keratitis as glucocorticoid complications that may justify a more rapid reduction towards physiological dosing. Whether that is possible depends on what the steroid is treating and how well controlled the underlying condition is. Before either decision, make sure the measurement is solid. A single clinic reading in someone who has just been told their steroid is causing problems is not a reliable basis for changing treatment. Home readings taken over a week, twice daily, seated and rested, at consistent times, give a far better picture and are what most guidelines now prefer for exactly this sort of decision. Record the steroid dose alongside the readings so that the two can be tracked together as the dose changes. Watch for the associated electrolyte problem as well. Potassium loss on corticosteroids is common enough to be worth a blood test at this dose, particularly if you are also on a diuretic, and hypokalaemia has its own symptoms - muscle weakness, cramps, palpitations - that patients tend to attribute to the underlying illness. Ankle swelling is another linear effect that rises with dose and often accompanies the pressure change. The thing to avoid is treating this as a permanent new diagnosis. Steroid-related hypertension typically improves as the dose comes down, so antihypertensive treatment started at 20 mg may need reviewing at 5 mg, in the same way that glucose-lowering treatment does. Patients who were started on a blood pressure tablet during a steroid course and never reviewed afterwards are a recognisable group, and the review is easy to arrange if somebody notes at the outset that it will be needed. Ask for that note to be made now rather than remembering it in a year. Ordinary lifestyle measures still apply and salt intake is worth attention, given that sodium retention is the mechanism in play. One drug-selection point to raise if treatment is started. Non-steroidal anti-inflammatory drugs raise blood pressure in their own right and carry a serious ulcer risk when combined with a corticosteroid, so if you have been taking ibuprofen or similar for anything, that is worth stopping first and seeing what the pressure does before a second drug is added. Decongestant cold remedies containing pseudoephedrine belong in the same category. Neither is likely to explain a 22 mmHg jump on its own, but both are common, both are bought without a prescription, and removing them costs nothing. Take the actual packets of everything you use to the appointment, prescribed or not, since that single habit resolves more of these puzzles than any additional test does. Do not wait for a routine review to raise this. A sustained systolic in the 150s is worth acting on within weeks rather than months, and the action may simply be a decision to monitor closely while the steroid dose comes down on the schedule already planned. What matters is that somebody makes that a decision rather than an omission.

Two and a half milligrams is low. It is not nothing, and three years is a long time. Two specific facts are worth knowing, because they contradict the reassurance you are being given. The first is bone. The 2022 American College of Rheumatology guideline on glucocorticoid-induced osteoporosis applies from 2.5 mg a day for more than three months. That is the threshold, not a level below it, and you have been above it for thirty-six times the qualifying duration. The guideline strongly recommends initial fracture risk assessment - fracture history, FRAX if you are 40 or over, and bone density with vertebral fracture assessment or spine radiographs - and, for anyone continuing between 2.5 and 7.5 mg a day at low or moderate risk who is not on treatment beyond calcium and vitamin D, reassessment every one to two years. If nobody has done that assessment in three years, it is overdue rather than optional. The second is the eye. Cataract had the lowest threshold of any adverse effect in the German cohort study of 1066 glucocorticoid-treated patients, appearing below 5 mg a day, and it belongs to the group of harms that track cumulative exposure rather than today's dose. Three years at 2.5 mg is meaningful cumulative exposure. Periodic ophthalmic review, including intraocular pressure, is reasonable at this dose and duration even though the daily figure looks trivial. There is a broader statement worth quoting because it is stronger than most patients have heard. The Endocrine Society notes that even low-dose glucocorticoid use, in the 2.5 to 7.5 mg a day range, increases cardiovascular disease, severe infections and hypertension, raises the incidence of type 2 diabetes and increases overall mortality. That is not an argument for stopping a drug you may well need. It is an argument for the dose being reviewed deliberately rather than renewed on repeat prescription indefinitely, and for the monitoring that goes with it being actually done. The adrenal question is separate and cuts the other way. Three years of continuous treatment means your axis is suppressed regardless of how small the tablet is, so you need sick-day instructions and you should not stop abruptly. The physiological equivalent of your own daily cortisol output is roughly 4 to 6 mg of prednisone, which means 2.5 mg is below replacement - you are dependent on the tablet without being fully covered by it. That combination is exactly why an illness with fever needs an increased dose. What I would ask for is a single appointment with three items on it: whether 2.5 mg is still needed, a fracture risk assessment, and written sick-day dosing instructions in the tablets you have. That is a twenty-minute conversation that has not happened in three years. It is also worth asking whether a reduction is realistic at all, because for some conditions the honest answer is no, and knowing that changes what you do next. If 2.5 mg is genuinely permanent, the plan stops being about stopping and becomes about managing a long exposure well: the bone assessment on a defined cycle, periodic eye checks, blood pressure and glucose monitored rather than assumed, and the dose confirmed as necessary at intervals rather than by inertia. That is a legitimate outcome and a much better one than years of vague reassurance. What I would resist is the framing you have been given, in which a small number is treated as equivalent to no drug at all. It is a real drug at a real dose with three years behind it, and it deserves the same attention as any other long-term prescription. The mechanics of any reduction, if one is agreed, are in the taper briefing, and my background is on the reviewer page.