Steroid side effects in cats: diabetes, heart and liver, and when to check what
Cats tolerate steroids fairly well (article 1), but that does not mean no side effects. The side effects that matter in cats are different from those in people or dogs. Blood glucose and the heart come before bones or stomach ulcers. This article sets out what is common, when it happens, and which numbers at home can give you an early warning.
- When steroids are first prescribed: what they are and why cats tolerate them
- Dosing and tapering: from tablet maths to the last dose
- Side effects: diabetes, heart and liver, and when to check what (you are here)
- Injections, tablets, inhalers and creams: the form is half the story
- When there is another disease: heart, diabetes, kidneys, infection
- Diabetes: 9.7% of 143 cats on high doses (1.9 mg/kg or more) for over 3 weeks. 12 of 14 developed it within 3 months (Nerhagen 2021). Watch urine glucose and fructosamine in the first 3 months.
- Heart failure: driven by more water in the blood vessels as glucose rises. After a long-acting injection, plasma volume rose 13.4% (over 40% in 3 cats, Ployngam 2006). After 5-7 days of oral steroids the left atrium also enlarged (Block 2020). Count the sleeping breathing rate daily for the first 2 weeks.
- With a healthy heart and normal glucose, risk is low. 1-2 mg/kg for 7 days caused no meaningful change in blood pressure, echocardiography or glucose (Khelik 2019).
- Blood test numbers shift: liver enzymes (ALP), cholesterol, triglycerides, albumin and glucose go up, lymphocytes and eosinophils go down, and creatinine can fall (Sharkey 2007). The kidneys may not actually be better.
- Urinary bacterial infection did not increase, unlike in dogs (0 of 32 cats on long-term treatment, Lockwood 2018).
- At equal potency, dexamethasone leans more towards diabetes than prednisolone (Lowe 2009).
| Who | Decides |
|---|---|
| At home | Baseline records before starting, daily sleeping breathing rate for 2 weeks, urine glucose strips for 3 months, water intake and urine clump size, weight |
| Together | Dates for fructosamine and glucose, whether to do heart tests (NT-proBNP, echo) before starting, whether to reduce or switch when side effects appear |
| The clinic | Diagnosing and treating diabetes or heart failure, choosing alternatives to steroids |
1. Steroid side effects in cats at a glance
| Side effect | How common | When | What you see first at home |
|---|---|---|---|
| High glucose / diabetes | 9.7% on high doses for 3+ weeks (Nerhagen 2021), 3.8% with long-acting injections (Dutch 2023) | Mostly within 3 months | More drinking and urine, positive urine strip, weight loss despite good appetite |
| Heart failure | 0.8% with long-acting injections (Dutch 2023); far higher with hidden heart disease | 3-6 days after injection, about a week on tablets | Sleeping breathing rate up (over 30), panting at rest |
| Liver change (steroid hepatopathy) | All 14 cats on high doses for 56 days had liver glycogen (Lowe 2008) | Weeks | Usually no signs; raised ALP on tests |
| Appetite and weight gain | Common | Within days | Begging for food, weight gain |
| Skin and muscle wasting (iatrogenic Cushing's) | Very rare; a series of 12 cats (Lien 2006) | Long-term | Thin or tearing skin, curled ear tips, thigh muscle loss |
| Infection | Urinary bacteria did not increase (Lockwood 2018); viruses are a theoretical risk | - | Sneezing and eye discharge (herpes flare), slow wound healing |
| Raised eye pressure (eye drops) | 8 of 11 glaucomatous cats, 2 of 5 normal cats (minor) (Gosling 2016) | 1-5 weeks | Red or enlarging eye (glaucoma-prone cats) |
2. Diabetes: catching it early with numbers
Steroids make the liver produce glucose and make the body respond less to insulin. Just 8 days of 2 mg/kg in healthy cats reduced glucose tolerance in most of 6 cats, and 3 developed fasting hyperglycaemia (Middleton 1985). Most cats compensate by producing more insulin. Those that cannot become diabetic.
We still cannot predict which cats will fail. In the 143-cat study, starting dose, weight, age and breed were not significantly linked to diabetes. But glucose in the urine before treatment was found in 44% of cats that became diabetic vs 15% of those that did not (P=0.056, a trend, Nerhagen 2021). So checking urine glucose before starting is worthwhile.
- Before starting: glucose, fructosamine and urine glucose (baseline).
- Every 1-2 weeks for the first 3 months: urine glucose strips at home (urine on litter or non-absorbent litter). Glucose spills into cat urine only above about 250-290 mg/dL (14-16 mmol/L), so a positive strip means blood glucose is above that.
- At 1 and 3 months: fructosamine (average glucose over 2-3 weeks). A single clinic glucose rises easily with stress, so fructosamine is more reliable (reading diabetes numbers).
- At anti-inflammatory doses (0.5-1 mg/kg) in a cat that is not obese or pre-diabetic, a looser schedule is reasonable.
- One positive urine strip: recheck within 2-3 days. Two positives: blood glucose and fructosamine.
- Diabetes diagnosed: the first question is whether the steroid can be reduced or replaced (inhaled, ciclosporin, chlorambucil). In the 143-cat study, 65% of the diabetic cats (9 of 14) started insulin (Nerhagen 2021). Diabetes caused by long-term steroids sometimes resolved when the drug was withdrawn (transient diabetes in 4 of 12 cats, Lien 2006).
- On dexamethasone? At equal potency, urine glucose at day 56 was significantly more common than with prednisolone (P=0.027, 14 cats, Lowe 2009). If diabetes is a concern, prednisolone is the better choice.
3. The heart: why steroids can cause heart failure
The sequence is: glucose rises, water is drawn from outside the vessels into the blood, plasma volume increases. A healthy heart passes it out as urine, but a thickened heart (HCM) cannot cope with the extra volume, and fluid backs up into the lungs.
| Study | Subjects | Results |
|---|---|---|
| Ployngam 2006 | 12 cats, long-acting methylprednisolone acetate 5 mg/kg | Sharp glucose rise at 3-6 days, plasma volume +13.4% (over 40% in 3). Back to baseline by 16-24 days |
| Block 2020 | 10 cats, oral steroids for 5-7 days | Left atrium 1.26 to 1.38 cm, LV internal diameter 1.55 to 1.72 cm. NT-proBNP rose more than 60% in 6 of 10 |
| Khelik 2019 | 10 allergic cats with normal hearts, 1 and 2 mg/kg for 7 days | No meaningful change in blood pressure, echo or cardiac biomarkers |
| Smith 2004 (not available online) | Case series of 12 cats with sudden heart failure after steroids | Acute heart failure after relatively short courses. The starting point for later plasma-volume studies |
| Dutch 2023 | 732 cats given long-acting injections vs 310 never treated | Heart failure 0.82% vs 1.90%: not higher in the treated group |
In short, cats with healthy hearts are at low risk, and the risk sits with hidden heart disease. The catch is that HCM often has no signs.
- Every day for the first 2 weeks after starting (or after an injection), count chest rises for 30 seconds while the cat sleeps, then double it.
- Under 30 is normal. Over 30 for several days, or 5-10 above the cat's usual: call the clinic. 40 or more, or open-mouth breathing: emergency (resting breathing rate).
- If a murmur or gallop has ever been heard, or the breed is HCM-prone (Maine Coon, Ragdoll and others), discuss NT-proBNP or an echo before starting. Block 2020 suggested that the percentage change in the same cat's NT-proBNP may help judge risk.
4. How blood test numbers change
Blood tests taken during steroid treatment need to be read as "is the disease worse, or is this the drug?"
| Test | Change on steroids | Evidence | How to read it |
|---|---|---|---|
| ALP, ALT | Up | Sharkey 2007 (ALP at 16-24 days), Lowe 2008 | May be a drug effect, not liver failure. Look separately if jaundice or poor appetite come with it |
| Cholesterol, triglycerides | Up | Lowe 2008, Khelik 2019 | Consider fasting status |
| Albumin | Up | Sharkey 2007, Khelik 2019 | Distinguish from dehydration |
| Glucose | Up | Ployngam 2006, Leal 2022 | Fructosamine over a single reading |
| Neutrophils up, lymphocytes and eosinophils down | "Stress leukogram" | Lowe 2008, Leal 2022 | Do not mistake for infection |
| Creatinine | Can fall | Sharkey 2007, Leal 2022 | The kidneys may not be better. Read with SDMA and urine specific gravity from the same time |
Logging the steroid in the app keeps this context next to every blood test.
5. The research so far, in one table
How to read it: (1) is there a control group (without one, it is hard to say the drug caused it)? (2) are the subjects healthy cats or real patients? (3) is the dose anti-inflammatory or immunosuppressive, tablets or a long-acting injection?
5-1. Glucose and diabetes
| Study | Subjects and design | Results (numbers) | Notes |
|---|---|---|---|
| Middleton 1985 | 6 healthy cats, 2 mg/kg for 8 days, before vs after | Reduced glucose tolerance in most, fasting hyperglycaemia in 3 | Glucose handling changes even with very short courses |
| Lowe 2009 | 14 cats randomised, prednisolone 4.4 vs dexamethasone 0.55 mg/kg, 56 days | Urine glucose at day 56 significantly more common with dexamethasone (P=0.027) | Pilot study. Source of "dexamethasone is more diabetogenic" |
| Nerhagen 2021 | 143 cats, referral hospital, retrospective, 1.9 mg/kg or more for 3+ weeks | Diabetes 9.7%, 12/14 within 3 months. No link to dose, weight or age | No control group. Source of "the first 3 months" |
| Leal 2022 | 28 patients randomised: single 20 mg long-acting injection plus metformin, diabetic diet or nothing | Glucose and insulin up at day 15 in all groups; neither metformin nor diet prevented it | Evidence that it cannot be prevented with a drug or diet |
| Dutch 2023 | 1,042 cats, cat-only practice, retrospective: long-acting injections (non-immunosuppressive) vs untreated | Diabetes 3.83% vs 7.10%; repeated injections did not raise risk | Controls not matched for age etc. Single practice |
5-2. Heart and plasma volume
| Study | Subjects and design | Results (numbers) | Notes |
|---|---|---|---|
| Ployngam 2006 | 12 dermatology cats, long-acting injection 5 mg/kg, before vs after | Plasma volume +13.4% (40%+ in 3), peak at 3-6 days, recovered by 16-24 days | Key evidence for the heart failure mechanism |
| Block 2020 | 10 patients, oral steroids for 5-7 days, before vs after | Left atrium 1.26 to 1.38 cm (P=0.02), NT-proBNP up 60%+ in 6/10 | Oral steroids also expand plasma volume. One author funded by a diagnostics company |
| Khelik 2019 | 10 allergic + 10 healthy control cats, 1 and 2 mg/kg for 7 days, crossover | No clinically meaningful change in BP, echo, glucose or NT-proBNP | Short anti-inflammatory courses look safe with a normal heart |
| Dutch 2023 | As above | Heart failure 0.82% vs 1.90% | Unknown whether hidden heart disease was screened |
5-3. Liver, blood, infection and long-term use
| Study | Subjects and design | Results (numbers) | Notes |
|---|---|---|---|
| Lowe 2008 | 14 cats, immunosuppressive doses for 56 days, liver biopsy | Liver glycogen in all, glucose, albumin and lipids up, 1 cat with jaundice and curled ears | Liver changes with high-dose, long courses |
| Sharkey 2007 | 11 dermatology cats, single long-acting injection | ALP, calcium and albumin up, creatinine down | Basis for the "kidney values look better" illusion |
| Lockwood 2018 | 32 cats on long-term steroids/ciclosporin vs 34 healthy, cystocentesis culture | Positive culture 0/32 vs 1/34 | Unlike dogs, no evidence of more urinary infection |
| Rojko 1979 | Healthy research cats, long-acting injection then leukaemia virus challenge | Persistent infection 82% (18/22) vs 11% (1/9) | Experimental. Evidence that antiviral defence can weaken |
| Lien 2006 | 12 cats with iatrogenic Cushing's, retrospective | Skin lesions, muscle wasting, transient diabetes in 4, resolved a mean 4.9 months after withdrawal | Rare, at the far end of long-term use |
Put together, the direction is consistent. Glucose moves even with short courses, and plasma volume rises within days. But the step to real disease (diabetes, heart failure) comes when high doses, long courses and hidden heart disease overlap. Short anti-inflammatory courses in healthy cats caused no major problems in any study.
What research has not answered yet
- How to predict which cats will develop steroid diabetes (dose, weight and age were all unrelated).
- How often heart failure actually happens in cats on tablets (the data are from injections).
- Whether stomach protectants are needed. We found no feline study of ulcers from steroids alone; the problem is overlap with NSAIDs (article 5).
- The "daily breathing rate for 2 weeks" and "urine glucose every 1-2 weeks for 3 months" in this article are management starting points built from the timings in the studies (plasma volume at 3-7 days, diabetes within 3 months). No study has tested the schedule itself.
6. Where opinions differ
Why it divides people: the 1,042-cat study found that diabetes (3.83%) and heart failure (0.82%) were lower in cats given long-acting injections than in untreated cats, and that repeated injections did not raise the risk (Dutch 2023). Plasma volume and glucose studies (Ployngam 2006, Leal 2022) show that even one injection clearly changes the body.
Our view: the two do not necessarily conflict. The 1,042-cat study used non-immunosuppressive doses, its controls were not matched for age or weight, and single-practice records can miss diagnoses. We read it as "low dose, short course, healthy cat: low risk", and keep the monitoring schedule above for cats with high doses, long courses, obesity or heart disease.
Checking it in your cat: keep baseline values (glucose, fructosamine, sleeping breathing rate) from before the steroid, so later changes can be compared.
7. What to bring to the clinic
- Baseline values: glucose, fructosamine, urine glucose, sleeping breathing rate.
- Sleeping breathing rate log (daily for the first 2 weeks), with dates.
- Urine strip results, with dates and colours.
- Changes in water and urine: ml gone from the water bowl per day, or number and size of litter clumps.
- For blood tests, the fact that the cat is on steroids, and the time of the last dose.
8. Read next
- Article 1 · When steroids are first prescribed
- Article 2 · Dosing and tapering
- Article 4 · Injections, tablets, inhalers and creams
- Article 5 · When there is another disease
- Resting breathing rate: the cheapest heart monitor
- Reading diabetes numbers
📌 One-line tip: on the day the steroid starts, measure the sleeping breathing rate and urine glucose once. A single baseline makes every later decision easier.