Perioral Dermatitis
Perioral dermatitis (POD) is an inflammatory papulopustular eruption around the mouth, nose, and eyes — predominantly affecting young women — that is strongly associated with topical corticosteroid use and is treated by stopping all steroids and using topical or oral antibiotics.
Table of Contents
1. Epidemiology and Risk Factors
Perioral dermatitis is a common inflammatory skin condition seen worldwide, with an estimated prevalence of 0.5–1% in dermatology populations. It predominantly affects young women between the ages of 16 and 45, with a female-to-male ratio of 5–10:1. While men and children can develop POD, peak incidence occurs among reproductive-age women — a pattern that has led researchers to investigate hormonal contributions to disease pathogenesis.
The single most important and well-documented trigger is topical corticosteroid (TCS) use. Fluorinated corticosteroids — including triamcinolone, fluocinonide, and clobetasol — carry a higher risk than non-fluorinated preparations, but prolonged or repetitive application of any potency class, including over-the-counter hydrocortisone, can precipitate or perpetuate POD. Intranasal corticosteroid sprays and inhaled steroids (whose swallowed particles contact the perioral skin) are also implicated, particularly when use is chronic.
Secondary triggers include:
- Fluorinated toothpaste — sodium fluoride in standard toothpaste formulations is a well-recognized perioral irritant in susceptible individuals
- Heavy occlusive cosmetics — thick foundation, heavy moisturizers, and barrier-disrupting emollients contribute via follicular occlusion and alteration of the perioral microenvironment
- Oral contraceptive pills — hormonal modulation of sebaceous gland activity and local immune responses may increase susceptibility
- Moisturizers and face creams — especially thick, occlusive formulations applied repeatedly to the perioral area
- Inhaled or intranasal corticosteroids — particles from sprays deposit on the perioral skin and upper lip
A familial tendency has been observed in some case series, suggesting a genetic predisposition to inflammatory skin reactivity. Notably, the COVID-19 pandemic era brought a documented surge in POD cases, attributed to both increased mask-wearing (creating a warm, occlusive perioral microenvironment) and increased topical corticosteroid use for mask-related irritant dermatitis — a cycle that reinforced and worsened underlying POD in many patients.
2. Pathogenesis
The exact pathogenesis of perioral dermatitis remains incompletely understood, but current evidence points to a multifactorial process involving skin barrier disruption, microbiome dysbiosis, and an aberrant inflammatory response — often initiated or perpetuated by topical corticosteroid use.
Skin barrier disruption: Topical corticosteroids thin the epidermis through suppression of keratinocyte proliferation and collagen synthesis. Chronic application reduces ceramide production and disrupts the intercellular lipid matrix of the stratum corneum, impairing barrier integrity. This creates a permeable perioral environment susceptible to colonization by commensal and pathogenic microorganisms.
Microbiome dysbiosis: Several organisms have been implicated in POD pathogenesis. Fusiform bacteria (Fusobacterium species) have been identified in lesional biopsies in some studies. Demodex folliculorum mites are found at increased density in POD lesions compared to normal perioral skin, and their role — either as direct triggers of inflammation or secondary colonizers taking advantage of a disrupted barrier — remains debated. Candida species may contribute in some patients, particularly those who have used topical corticosteroids extensively.
The steroid paradox: Topical corticosteroids initially suppress the redness and irritation of early POD, providing apparent relief. This encourages continued or escalating use, further disrupting the skin barrier and altering the local immune environment. When the steroid is withdrawn, a pronounced rebound flare occurs — erythema intensifies, lesions multiply, and burning worsens — mimicking a severe inflammatory episode and compelling many patients to restart topical steroids. This self-perpetuating cycle is central to the chronicity of steroid-induced POD.
Fluoride hypothesis: Sodium fluoride in toothpaste may elicit a perioral inflammatory reaction in susceptible individuals. The mechanism is believed to involve direct irritation of the follicular epithelium and perifollicular connective tissue. Switching to non-fluoride toothpaste is a common part of treatment and is reported to help many patients, though it is not universally effective, suggesting that fluoride sensitivity is one contributing factor rather than the sole cause.
Occlusive cosmetics: Heavy emollients and occlusive cosmetics can obstruct pilosebaceous units around the mouth, creating a comedone-like plugging effect that initiates papule formation — a mechanism similar to acne mechanica. The occlusive environment also promotes bacterial overgrowth within follicular units.
Hormonal contribution: The strong female predominance and the clinical observation that POD frequently worsens premenstrually and may develop or exacerbate with oral contraceptive use suggest that sex hormones influence sebaceous gland activity and local inflammatory pathways — possibly through androgen receptor modulation of sebocyte proliferation and lipid production.
3. Clinical Features
Perioral dermatitis has a characteristic distribution and morphology that distinguish it from other common facial dermatoses. The term periorificial dermatitis is increasingly preferred, as the eruption affects three orificial areas: perioral (most common), perinasal, and periocular.
Distribution: The chin and upper lip area are most commonly affected, followed by the nasolabial folds. Perinasal involvement produces a similar papulopustular pattern flanking the nostrils. Periocular (periorbital) POD — affecting the skin around the eyelids — may occur in isolation or together with perioral lesions and is especially common in children. Bilateral but asymmetric involvement is typical.
The pathognomonic hallmark of POD is sparing of the vermilion border — a 1–3 mm pale ring of clinically normal skin immediately adjacent to the lip margin. This characteristic sparing reliably distinguishes POD from acne vulgaris (which has no predilection for the perioral zone) and from rosacea (which involves the central face but not in this specific perioral-sparing pattern).
Lesion morphology: The primary lesions are erythematous papules measuring 1–3 mm in diameter, often with superimposed micropustules. The lesions are non-comedonal — there are no blackheads or whiteheads (comedones), which is a critical distinguishing feature from acne. Papules cluster in groups on an erythematous background, giving the skin a rough, bumpy texture.
Scaling and texture: Fine scaling is common, and patients frequently report that affected skin feels tight, dry, and uncomfortable. Unlike the greasy scale of seborrhoeic dermatitis, the scale in POD is dry and superficial.
Symptoms: Mild burning, stinging, or itching is frequently reported. Some patients experience only cosmetic concerns without significant discomfort. The condition is not typically painful. Psychological distress related to the facial location of the eruption is common and can be considerable.
Natural history without treatment: Untreated POD follows a chronic, waxing-and-waning course. In patients continuing topical corticosteroid use, the condition typically worsens over months to years. Spontaneous remission can occur but may take many months. Relapses are common, particularly if the precipitating factors (steroid use, heavy cosmetics, fluoride toothpaste) are not addressed.
4. Diagnosis
Perioral dermatitis is a clinical diagnosis, established by recognizing the characteristic distribution (perioral, perinasal, or periocular), the pathognomonic sparing of the vermilion border, the non-comedonal papulopustular morphology, and a history of topical corticosteroid use or other recognized triggers. No laboratory tests or imaging are required for the diagnosis in typical cases.
Skin biopsy is reserved for atypical presentations or when the diagnosis is uncertain. Histopathological findings include a perifollicular and subepidermal lymphohistiocytic infiltrate, spongiosis, sebaceous gland dilation, and variable degrees of epidermal disruption. Non-caseating granulomas are the defining feature of the granulomatous variant. Biopsy does not show the caseating necrosis of cutaneous tuberculosis or the elastolytic changes of granuloma annulare.
The differential diagnosis includes several important conditions:
- Acne vulgaris: Comedones (blackheads and whiteheads) are present and are diagnostic of acne; POD is strictly non-comedonal. Acne has a sebaceous distribution (cheeks, forehead, nose, upper chest, back) and does not spare the vermilion border or concentrate in the strict perioral pattern of POD.
- Rosacea: Rosacea involves flushing, persistent central facial erythema, dilated telangiectatic vessels, and papulopustules concentrated over the nose and cheeks in older patients. The age of onset is typically later (30s–60s), and the vermilion border sparing seen in POD is not a feature of rosacea. Overlap between POD and rosacea exists, however, and the two conditions may coexist.
- Seborrhoeic dermatitis: Characterized by greasy, yellowish scale in sebaceous distribution zones (nasolabial folds, scalp, eyebrows, glabella). Responds to antifungal agents (ketoconazole, selenium sulfide). Less papulopustular than POD; not concentrated in the strict perioral zone.
- Allergic contact dermatitis: Eczematous rather than papulopustular morphology; acute onset after specific allergen exposure (toothpaste ingredients, cosmetic preservatives, fragrances); positive patch testing identifies causative allergens. The distribution follows the pattern of allergen contact and is not strictly perioral.
- Impetigo: Bacterial skin infection (Staphylococcus aureus, Streptococcus pyogenes) producing honey-colored crusted erosions. Responds rapidly to topical or oral antibiotics. Not restricted to the perioral zone; not papulopustular without crusting; can be distinguished by the characteristic crusting and rapid treatment response.
- Milia: Small, firm, white epidermal inclusion cysts that may cluster under the eyes or on the cheeks. Hard and white on palpation; no erythematous base; no pustules; not associated with topical steroid use.
5. Treatment: Steroid Withdrawal
The single most important step in managing perioral dermatitis is stopping all topical corticosteroids. Continuing even a low-potency preparation perpetuates the cycle of barrier disruption and inflammatory rebound that drives the disease. Stopping covers:
- All topical corticosteroid creams, ointments, and gels applied to the face
- Nasal corticosteroid sprays, where they can be stopped (they are often used for allergic rhinitis)
- Inhaled corticosteroids where particles may be depositing on perioral skin
- Over-the-counter hydrocortisone preparations (many people do not think of these as "steroids")
The steroid withdrawal phenomenon is the most important concept to understand. Within 2–3 weeks of stopping topical corticosteroids, erythema worsens, lesions multiply, burning and stinging intensify, and the overall appearance deteriorates significantly. This rebound flare is expected and temporary — it is not a sign of treatment failure or disease progression. Many people restart their topical steroid at this point, which keeps the cycle going.
What typically happens after the steroid is stopped:
- The skin worsens significantly for 2–4 weeks before improving
- This worsening is a predictable rebound from stopping the steroid, not an allergy or an infection
- Typical timeline: peak rebound at 2–4 weeks; beginning of improvement at 4–8 weeks; resolution at 8–12 weeks with treatment
- Restarting the steroid during the rebound brings the cycle back
Some dermatologists advocate a "zero therapy" approach in the first 2–4 weeks: stopping not only all topical corticosteroids but also all cosmetics, moisturizers, and non-essential topical products. This complete elimination of potential irritants and occlusive agents allows the skin barrier to begin recovering without interference. While initially uncomfortable, zero therapy is described by its advocates as speeding the return of normal perioral skin.
The evidence behind these approaches is thin. A 2005 critical review of 30 treatment reports found only 2 therapeutic trials of medium quality; the other studies were of low quality. The reports agreed most consistently on oral tetracycline, on stopping topical corticosteroids and cosmetics and, to a lesser extent, on no treatment at all ("zero therapy"). They did not agree on topical treatments.
Treatment plans often also include switching to a non-fluoridated toothpaste and stopping heavy, occlusive moisturizers and cosmetics for the duration of treatment.
6. Treatment: Topical Therapies
Topical therapies are the foundation of treatment for mild-to-moderate perioral dermatitis and are used as adjuncts to oral antibiotics in more severe cases. They are applied after stopping all topical corticosteroids and are typically continued for 8–12 weeks for full effect.
Metronidazole 0.75–1% gel or cream is the first-line topical agent for perioral dermatitis. Applied twice daily to affected areas, metronidazole exerts both antimicrobial effects (against anaerobic bacteria and Demodex mites) and direct anti-inflammatory effects through inhibition of reactive oxygen species and cytokine production. It is well tolerated, with minimal systemic absorption and a favorable safety profile for prolonged use. Mild initial stinging or dryness is common but usually transient.
Erythromycin 2% gel is an alternative topical antibiotic useful when metronidazole is unavailable, not tolerated, or in patients with a history of sensitivity. Applied twice daily, it provides antimicrobial coverage with some anti-inflammatory benefit. Increasing bacterial resistance to erythromycin means it is generally considered second-line to metronidazole.
Azelaic acid 15–20% gel or cream is described as having a dual mechanism: it is both antimicrobial (including activity against Demodex and anaerobes) and directly anti-inflammatory through inhibition of reactive oxygen species. It is particularly useful in patients with overlapping POD and rosacea. Applied twice daily, azelaic acid causes mild stinging and transient skin irritation in the first 1–2 weeks of use, which typically resolves. It has an excellent long-term safety profile and is safe in pregnancy. However, a 2022 systematic review of 11 studies (733 people) found that azelaic acid gel may make no difference to severity, as judged by doctors or by patients, after 6 weeks of treatment (low certainty evidence).
Pimecrolimus 1% cream is a calcineurin inhibitor that reduces perifollicular and dermal inflammation without the atrophogenic effects of topical corticosteroids. It is particularly effective for steroid-induced POD and for patients in whom the barrier is severely compromised by prior corticosteroid use. Applied twice daily, it does not cause skin thinning or telangiectasia formation. Initial mild burning is common. It is especially useful for periocular POD where thin eyelid skin makes other topicals poorly tolerated.
Two randomized, double-blind trials compared pimecrolimus cream with the same cream base without the drug (vehicle). In a multicentre trial in adults treated twice daily for up to 4 weeks, the average severity score fell to 2.6 with pimecrolimus versus 3.5 with the cream base (both groups started at 5.2), and people who had used topical steroids before benefited most. In a single-centre trial in 40 patients, severity scores were lower with pimecrolimus during the 4 weeks of treatment, but there was no significant difference at follow-up 4 weeks later.
Tacrolimus 0.03–0.1% ointment has a similar mechanism to pimecrolimus and may be more effective in refractory cases. The 0.03% concentration is preferred in children; the 0.1% formulation is available for adults with refractory disease.
Clindamycin 1% lotion or gel is widely used as a topical antibiotic for POD, though clinical evidence is less robust than for metronidazole. Applied twice daily, it is a reasonable alternative, particularly for patients with mild disease or when other agents are not available.
Measures commonly described alongside topical therapy:
- Non-fluoridated toothpaste (for example, fluoride-free natural toothpastes)
- No heavy moisturizers, greasy emollients, or occlusive cosmetics
- A gentle, fragrance-free cleanser, without scrubbing
- Little or no makeup during the acute phase
- A non-comedogenic, lightweight sunscreen when sun protection is needed
7. Treatment: Oral Antibiotics
Oral antibiotics are indicated for moderate-to-severe perioral dermatitis, widespread facial involvement, cases that have failed to respond adequately to topical therapy alone, and the granulomatous variant. They are often combined with topical agents (particularly metronidazole gel) for faster and more complete resolution.
A 2022 systematic review of 11 studies (733 people) found only low or very low certainty evidence for any medicine used for perioral dermatitis: oral tetracycline may improve severity from day 20 onward; pimecrolimus cream may improve severity slightly after 4 weeks; and azelaic acid gel may make no difference after 6 weeks. Side effects reported with oral tetracycline included stomach discomfort, facial dryness and itching.
Tetracyclines — first-line in adults:
Doxycycline at 50–100 mg/day for 8–12 weeks is the most widely used oral agent for POD. Its therapeutic benefit in POD — as in rosacea — operates primarily through anti-inflammatory mechanisms rather than purely antimicrobial ones: doxycycline inhibits matrix metalloproteinases (MMPs), suppresses pro-inflammatory cytokines (interleukin-1, interleukin-6 (IL-6), tumor necrosis factor-alpha), and reduces reactive oxygen species production by neutrophils. A sub-antimicrobial dose of 40 mg/day (modified-release formulation) is effective in some patients and minimizes resistance pressure. Taking doxycycline with food and a full glass of water reduces irritation of the esophagus (food pipe). Photosensitivity (sunburning more easily) is a clinically important side effect. Doxycycline is contraindicated in pregnancy and in children under 8 years (risk of permanent tooth staining and impaired bone growth).
Minocycline at 50–100 mg/day is an alternative tetracycline. It is more lipophilic than doxycycline and achieves higher tissue concentrations in skin. However, it carries a higher risk of blue-grey hyperpigmentation with prolonged use (minocycline-induced pigmentation), vestibular side effects (dizziness, vertigo), and rare but serious drug reactions (drug-induced lupus, autoimmune hepatitis). For these reasons, minocycline is generally considered second-line to doxycycline.
Erythromycin at 250–500 mg twice daily for 8–12 weeks is the preferred oral antibiotic for pregnant women and children under 8 years, in whom tetracyclines are contraindicated. Gastrointestinal side effects (nausea, cramping, diarrhea) are common and may limit adherence. Erythromycin is generally less effective than tetracyclines in adults but is a clinically important alternative when tetracyclines cannot be used.
Tapering rather than abrupt discontinuation of oral antibiotics at the end of the course is used to reduce the risk of rebound flare, typically as a gradual dose reduction over 2–4 weeks.
Isotretinoin (low-dose) is reserved for refractory perioral dermatitis that has failed multiple courses of conventional therapy, and for the granulomatous variant. Doses of 0.2–0.5 mg/kg/day for 4–6 months are typically effective. Standard isotretinoin precautions apply: it is a potent teratogen requiring strict contraception and iPLEDGE program enrollment in the United States; monthly liver function and lipid monitoring; and screening for depression and mood changes. Despite these requirements, low-dose isotretinoin is highly effective for refractory POD and offers the prospect of long-term remission.
Practical combination approach: For moderate-to-severe POD, initiating oral doxycycline together with topical metronidazole gel at the time of steroid cessation provides the fastest response and highest rate of complete clearance. Topical therapy is continued throughout the oral antibiotic course and for 4–8 weeks after oral antibiotics are discontinued.
8. Granulomatous Perioral Dermatitis
Granulomatous perioral dermatitis is a rare but clinically distinct variant of POD characterized by firm, indurated papules with histological evidence of non-caseating granulomas. In children, this variant is known as facial Afro-Caribbean childhood eruption (FACE), though it occurs in children of all ethnicities and racial backgrounds.
Epidemiology: Granulomatous POD primarily affects prepubertal children, with a roughly equal sex distribution — in contrast to the strong female predominance of classic adult POD. Adult cases occur but are less common. The condition may affect any race or ethnicity.
Clinical features: The lesions are firm, indurated papules 1–4 mm in diameter, appearing skin-colored to yellow-brown. Unlike classic POD, which features erythematous micropustules, granulomatous POD papules are more solid and dome-shaped. Diascopy (pressing a glass slide against the lesion) reveals the characteristic yellowish-brown "apple-jelly" translucency typical of granulomatous inflammation. Lesions cluster around the perioral, perinasal, and periocular zones — the same distribution as classic POD. The vermilion border sparing is typically preserved.
Corticosteroid history: Unlike adult POD, many children with granulomatous POD have no history of prior topical corticosteroid use. This suggests that other factors — possibly infectious or immunological — drive granuloma formation in some cases. However, steroid-exposed cases do occur in this variant as well.
Histopathology: Biopsy shows perifollicular non-caseating granulomas with lymphohistiocytic infiltrates and foreign body giant cells. The absence of caseating necrosis distinguishes granulomatous POD from cutaneous tuberculosis (lupus vulgaris) — a critical distinction, particularly in high-prevalence tuberculosis (TB) regions. The granulomas are typically perifollicular, centred on the hair follicle unit.
Differential diagnosis:
- Cutaneous sarcoidosis: Systemic involvement (lymphadenopathy, pulmonary sarcoidosis, elevated serum ACE); angiolupoid sarcoid affects the nose differently; biopsy may be identical, requiring clinical correlation and systemic evaluation
- Lupus miliaris disseminatus faciei (LMDF): Confined to the face; central caseating necrosis on diascopy (distinguishes from the non-caseating granulomas of granulomatous POD); larger papules; responds to tetracyclines
- Acne agminata: An older term for LMDF; same entity
- Cutaneous tuberculosis (lupus vulgaris): Caseating granulomas; tuberculosis contact history; responds to antituberculous therapy; biopsy with acid-fast staining is diagnostic
Treatment: Low-dose oral erythromycin (for children) or azithromycin is the preferred first-line treatment for granulomatous POD. Topical metronidazole gel may provide adjunctive benefit. The condition may undergo spontaneous resolution over months to years, particularly in children. For adult granulomatous POD resistant to antibiotics, low-dose oral isotretinoin has demonstrated efficacy and is the preferred second-line option.
Key Research Papers
- Wollenberg A, Bieber T, Dirschka T, et al. "Perioral dermatitis." J Dtsch Dermatol Ges. 2011;9(5):422–427. — Guideline of the German Society of Dermatology on perioral dermatitis. — PubMed PMID: 21040469
- Nguyen V, Eichenfield LF. "Periorificial dermatitis in children and adolescents." J Am Acad Dermatol. 2006;55(5):781–785. — Chart review of 79 children and adolescents: 72% had prior steroid exposure, and topical metronidazole was associated with clearing. — PubMed PMID: 17052482
- Hafeez ZH. "Perioral dermatitis: an update." Int J Dermatol. 2003;42(7):514–517. — Review of pathogenic theories, clinical variants, and treatment approaches including metronidazole and tetracyclines. — PubMed PMID: 12839598
- Cotterill JA. "Perioral dermatitis." Br J Dermatol. 1979;101(3):259–262. — Describes 43 patients and the link between perioral dermatitis and topical steroid use on the face. — PubMed PMID: 159712
- Oppel T, Pavicic T, Kamann S, et al. "Pimecrolimus cream (1%) efficacy in perioral dermatitis — results of a randomized, double-blind, vehicle-controlled study in 40 patients." J Eur Acad Dermatol Venereol. 2007;21(9):1175–1180. — Severity scores were lower with pimecrolimus during 4 weeks of treatment, with no significant difference at follow-up. — PubMed PMID: 17894701
- Schwarz T, Kreiselmaier I, Bieber T, et al. "A randomized, double-blind, vehicle-controlled study of 1% pimecrolimus cream in adult patients with perioral dermatitis." J Am Acad Dermatol. 2008;59(1):34–40. — Multicentre trial: average severity score 2.6 with pimecrolimus versus 3.5 with vehicle; most benefit after prior topical steroid use. — PubMed PMID: 18462835
- Lipozencic J, Ljubojevic S. "Perioral dermatitis." Clin Dermatol. 2011;29(2):157–161. — Clinical dermatology review covering etiology, differential diagnosis, and treatment options with a focus on practical management. — PubMed PMID: 21396555
- Tempark T, Shwayder TA. "Perioral dermatitis: a review of the condition with special attention to treatment options." Am J Clin Dermatol. 2014;15(2):101–113. — Literature review grading treatments by evidence quality; oral tetracycline had the best evidence, and topical corticosteroids carry a risk of rebound when stopped. — PubMed PMID: 24623018
- Gray NA, Tod B, Rohwer A, et al. "Pharmacological interventions for periorificial (perioral) dermatitis in children and adults: a systematic review." J Eur Acad Dermatol Venereol. 2022;36(3):380–390. — 11 studies, 733 participants; all evidence of low or very low certainty. — PubMed PMID: 34779023
- Weber K, Thurmayr R. "Critical appraisal of reports on the treatment of perioral dermatitis." Dermatology. 2005;210(4):300–307. — 30 treatment reports; only 2 trials of medium quality. — PubMed PMID: 15942216
- Dirschka T, Szliska C, Jackowski J, et al. "Impaired skin barrier and atopic diathesis in perioral dermatitis." J Dtsch Dermatol Ges. 2003;1(3):199–203. — 40 patients and 62 controls: water loss through the skin of the face was significantly higher in patients, pointing to an impaired skin barrier. — PubMed PMID: 16285495
10. PubMed Searches
11. Connections
- Dermatology
- Rosacea — close clinical overlap; both involve central facial papulopustules; TCS worsens both
- Acne — key differential diagnosis; distinguished by comedones and sebaceous distribution
- Seborrheic Dermatitis — overlapping nasolabial fold involvement; different scale quality and antifungal response
- Contact Dermatitis — fluoride toothpaste contact trigger; patch testing relevant in atypical cases
- Eczema — topical corticosteroid connection; steroid-induced POD common complication of eczema treatment
- Impetigo — differential diagnosis; honey-colored crusted erosions vs. non-crusted papulopustules of POD
- Folliculitis — perifollicular inflammation shared; distinguishable by distribution and clinical context