TL;DR: Perioral dermatitis is one of the few conditions where the wrong treatment makes it dramatically worse, and topical steroids are the most common wrong treatment. The condition is caused or sustained by steroids in about a third of cases, which means the cream the doctor prescribed for the rash you were calling eczema is probably what made the rash perioral dermatitis. The actual treatment is the opposite of what intuition suggests: stop everything, then introduce specific antibiotics or azelaic acid. I went through this twice. The second time I got it right.
A reader described the pattern I recognised immediately. Small red bumps appeared around her mouth, sparing the thin border of skin right next to the lip line. Her GP prescribed hydrocortisone 1 percent. The rash cleared within two days. Two weeks later it came back, this time worse, with small pustules and dry scaling. The GP prescribed a stronger steroid. The rash cleared again. Two weeks later it came back worse still.
By the time she emailed me she had been on and off four different topical steroids for nine months and the rash had spread to her chin, the sides of her nose, and around her eyes. She was using makeup to cover it and avoiding photos.
This is the classic course of misdiagnosed perioral dermatitis. The steroid clears it temporarily. The condition rebounds worse each time the steroid is stopped. The cycle continues until either the patient stops applying anything at all (which initially makes the rash dramatically worse for two to four weeks, then resolves) or a doctor recognises the pattern and switches to the correct treatment.
I think this is one of the worst-handled conditions in dermatology and the reason is structural. The condition mimics eczema in early presentation, eczema is treated with steroids, steroids make it look better acutely, and the underlying condition gets worse. The patient and the prescriber both think they are doing the right thing.
This is what I have learned from going through it twice and from reading the actual evidence.
What perioral dermatitis is
Perioral dermatitis (also called periorificial dermatitis when it includes the eyes and nose, which it often does) is a chronic inflammatory facial rash that appears in a characteristic distribution: small papules and pustules on a background of red, often scaly skin, concentrated around the mouth, nose, and sometimes the eyes. The thin border of skin directly adjacent to the lips, called the vermillion border, is usually spared. This sparing is one of the diagnostic clues.
The condition affects mostly women between 20 and 45. The Tolaymat and Hall review put the male-female ratio at around 1:9 (Tolaymat and Hall, StatPearls 2023, PMID: 29763050). It is not rare. Dermatology clinics see it regularly. Primary care sees it more often and misdiagnoses it more often.
The cause is unclear in the formal sense. The Weber and Thurmayr critical appraisal listed several proposed mechanisms: topical or inhaled corticosteroids, fluorinated toothpaste, occlusive cosmetics, oral contraceptives, hormonal changes, demodex mites, Candida overgrowth, and barrier dysfunction (Weber and Thurmayr, Dermatology 2005, PMID: 15942215). The picture is probably multifactorial. The strongest single trigger in clinical practice is corticosteroid exposure, either topical to the face or inhaled (asthma inhalers) and deposited around the mouth.
The Hafeez review put corticosteroid use as the precipitating factor in roughly 30 to 50 percent of cases (Hafeez, Int J Dermatol 2003, PMID: 12780702). When the steroid is introduced the rash improves. When it is withdrawn the rash returns. This is the steroid-dependent course that turns a self-limited inflammatory event into a chronic condition.
Why steroids look like they work
Topical corticosteroids suppress the inflammatory response that produces the visible papules and erythema. The rash flattens within a few days. The patient sees clear skin. The prescriber sees a successful treatment. Everyone is satisfied.
The problem is what is happening underneath. The steroid is not addressing the cause, which is some combination of barrier dysfunction, follicular inflammation, and microbial imbalance depending on the specific case. The suppression masks the rash but does not resolve it. When the steroid is withdrawn the inflammation rebounds, often more aggressively than before, because the steroid has thinned the skin slightly, altered the local microbiome, and possibly upregulated the receptors that were being suppressed.
Each cycle of suppression and rebound is harder to resolve than the last. After six months of intermittent steroid use the rash typically requires several months of complete steroid avoidance plus active treatment to resolve. After a year the resolution can take six months or longer.
This is the trap. The early treatment is comfortable. The accumulating damage is invisible. By the time the trap is recognised the patient is in a worse position than when they started.
What the actual treatment is
The protocol is straightforward once the diagnosis is correct. The Hall and Reichenberg evidence-based review summarised the data (Hall and Reichenberg, G Ital Dermatol Venereol 2010, PMID: 20629476). The two interventions with the strongest evidence are oral tetracycline-class antibiotics (doxycycline 100 mg daily or minocycline 50 mg twice daily, typically for six to eight weeks) and topical azelaic acid (15 to 20 percent, applied twice daily for similar duration). Topical metronidazole 0.75 percent and topical erythromycin are second-line and weaker.
The other essential component is what is called “zero therapy” in the older European literature: the stopping of every topical product on the face except a basic non-medicated emollient. This means no cleanser other than water, no makeup, no sunscreen if possible, no toothpaste with fluoride contacting the perioral area, no occlusive lip products, and certainly no steroids.
The first two to four weeks of zero therapy are visually worse. The rash flares because the steroid suppression is being removed and the inflammation rebounds. Patients who do not understand this is the expected course almost always restart the steroid at this point. The flare can last three weeks. Pushing through it is the most difficult part of recovery.
After the flare phase, with consistent oral antibiotic or topical azelaic acid treatment and continued steroid avoidance, the rash resolves over six to twelve weeks. The skin returns to baseline. The recurrence rate is reasonable but not zero, and a future steroid exposure can restart the cycle.
What I did the first time and the second time
The first time I got perioral dermatitis I did not know what it was. I was 28. Small bumps appeared on my chin. I went to a GP who said it looked like eczema and prescribed hydrocortisone 1 percent. I used it for two weeks and stopped when the rash cleared. Two weeks later the rash was back. I went back. He gave me a stronger steroid. This pattern repeated for about six months. I gained 6 kg from inhaled budesonide for asthma that I had been on the whole time, which I suspect contributed.
Eventually a dermatologist recognised it. She prescribed doxycycline 100 mg daily and told me to stop everything I was putting on my face. The first three weeks were the worst the rash had ever been. I almost gave up. By week six the rash had cleared. I stayed on doxycycline for eight weeks total and the rash did not return for two years.
The second time I got it I was 32. The early signs were familiar. I did not see a GP. I went straight to azelaic acid 15 percent twice daily, stopped my retinoid for the duration, and used only water to cleanse. The rash cleared in five weeks without antibiotics. I have stayed on azelaic acid as a maintenance ever since (three years now) and have not had a third episode.
The difference between the two courses was the speed of correct diagnosis. The first time took eight months and a lot of damage. The second time took six weeks because I recognised the pattern.
The misdiagnosis problem
Perioral dermatitis is misdiagnosed as eczema, acne, rosacea, contact dermatitis, and seborrheic dermatitis. The distribution and sparing pattern are the key diagnostic features. Eczema does not typically spare the vermillion border so cleanly. Acne does not produce the diffuse erythema between the papules. Rosacea is concentrated on the central face and rarely produces the perioral concentration. Contact dermatitis presents more uniformly across the contact area. Seborrheic dermatitis follows the eyebrows and nasal folds more than the perioral skin.
The signal that matters most clinically is the steroid response. If a rash is being treated as eczema, clears on a steroid, and returns within two weeks of stopping the steroid, the diagnosis should be reconsidered. Eczema does not behave this way. Perioral dermatitis does.
I get an email about once a month from a reader on a steroid cycle who has not yet been told this. The cycle stops when somebody draws the line and says no more steroid, and that line is rarely drawn by the prescriber who started the steroid.
What I would tell my past self
The bumps around the mouth that look like eczema are probably not eczema if they spare the lip border. Do not start a steroid until you have ruled out perioral dermatitis. The cost of being wrong about this and using a steroid for a few weeks is months of recovery. The cost of being right and not using a steroid is mild discomfort while a non-steroid treatment takes effect.
If you have already been on the steroid cycle, the path out is the same regardless of how long you have been on it. Stop the steroid completely. Expect a flare. Use azelaic acid or oral tetracyclines under medical supervision. Use only water on the face for cleansing. Avoid all heavy lip and skin products. Push through three to six weeks of difficulty. The condition resolves.
The maintenance after resolution is to keep using a mild active (azelaic acid is what I use) and to be cautious with future steroid exposure on or near the face. Inhaled steroids deposited around the mouth count. Hydrocortisone for an unrelated rash on the cheek counts. The threshold for restarting the cycle can be low for the first year or two after recovery.
Frequently asked
Why does the rash get worse before it gets better? The steroid was suppressing the inflammation. When the steroid is removed the inflammation rebounds, often more severely than before because of the rebound effect on local immunity. The first two to four weeks of correct treatment look worse than the steroid years. This is the expected course and the most common reason people give up.
How long does treatment take? Six to twelve weeks for the visible rash to clear. Several more weeks for the skin texture to fully recover. People who have had multiple steroid cycles can take six months or longer. The longer the steroid exposure, the longer the recovery.
Can I use sunscreen during treatment? Most physical sunscreens with zinc or titanium are tolerated. Chemical sunscreens often are not. The simplest approach during the flare phase is sun avoidance and hats rather than sunscreen on the affected area. Sunscreen can be reintroduced after the rash has cleared.
Is fluoride toothpaste really a trigger? It is associated in some cases. The mechanism is not fully understood. Switching to a fluoride-free toothpaste during recovery is a reasonable trial, and some patients have clear recurrences linked to specific toothpaste formulations. I do not think it is a primary cause but it can sustain the rash in susceptible people.
Will it come back? It can. Recurrence rates after successful treatment vary by series but a meaningful percentage of patients have at least one recurrence within two years. Long-term maintenance with azelaic acid or careful steroid avoidance reduces the rate. Some people have a single episode and never recur. Others have multiple cycles over a decade.
Related Elelaf tools
Sources
- Tolaymat L, Hall MR. Perioral Dermatitis. StatPearls 2023. PMID: 29763050.
- Hafeez ZH. Perioral dermatitis: an update. Int J Dermatol 2003. PMID: 12780702.
- Hall CS, Reichenberg J. Evidence based review of perioral dermatitis therapy. G Ital Dermatol Venereol 2010. PMID: 20629476.
- Weber K, Thurmayr R. Critical appraisal of reports on the treatment of perioral dermatitis. Dermatology 2005. PMID: 15942215.