Evidence for the Clinical Association between Demodex and Rosacea: A Review

Rosacea is a chronic inflammatory condition that manifests primarily on the skin of the central face and is characterized by persistent facial erythema, papules, pustules, telangiectasia, and flushing [1, 2]. Rosacea affects 5.5% of the global population, and its prevalence in the white population is up to 10% or higher [2]. Unfortunately, current therapies are not curative, and the pathogenesis remains uncertain.

The Demodex mites are some of the most widespread but often overlooked human commensals. There are more than 100 species of Demodex mite, many of which are commensal organisms of the pilosebaceous unit of mammals, including cats, dogs, sheep, cattle, pigs, etc. Currently, two species of Demodex mites are identified as being present on human skin: Demodex folliculorum and Demodex brevis. The former is longer and primarily parasitic in hair follicles, while the latter is shorter and tends to reside in sebaceous glands [3] (shown in Fig. 1). Demodex is observed in normal skin with a prevalence of 100% and a density of ≤5 D/cm² in the adult population [4]. Most humans are carriers of Demodex and do not develop clinical symptoms. The pathogenicity of Demodex remains questioned by clinicians due to the high infection but low morbidity and an incomplete correlation between mite loads and the severity of clinical symptoms.

Demodicosis and rosacea are common skin disorders encountered in dermatology practice. While demodicosis is the result of infestation by the Demodex mites, the etiology of rosacea remains unclear [5]. Although papulopustular rosacea (PPR) and rosacea-like demodicosis have numerous similarities, they are generally considered two distinct entities [6]. On the other hand, rosacea appears to have a close relationship with Demodex. A systematic review and meta-analysis of twenty-three case-control studies concluded that patients with rosacea had significantly higher prevalence and degrees of Demodex mite infestation compared with control patients [7]. Forton et al. [4] analyzed the direction of the causal relationship at several levels (molecular, histologic, clinic) and concluded that PPR is caused by the Demodex mite and not the inverse. In addition, she suggested that erythematotelangiectatic rosacea (ETR) may be related to a subclinical stage of demodicosis and that rosacea which manifests with papulopustules should be considered a demodicosis [8, 9].

All of these indicate that Demodex mites may be involved in the development of rosacea. However, more direct proof is needed to confirm the pathogenic role of Demodex in rosacea. This evidence has not been established and might not be easily available in the near future because Demodex is a host-specific, obligate commensal that is currently difficult to be cultured in vitro to parasitize other animal hosts [10]. However, there are promising case reports currently available: Demodex canis infestation has been found in a ferret [11], canine Demodex mites were cultured on SCID mice [12, 13], and the first ex vivo cultivation study of Demodex was conducted by Clanner-Engelshofen et al. [14]. Still, there is a lack of direct evidence for the relationship between Demodex and rosacea. Therefore, we explored the relationship between Demodex and rosacea through a review of the existing literature in order to orient future research and suggest clinical treatments.

Methods for detecting the density of Demodex mite infestation in skin are categorized as non-invasive and invasive. In recent years, non-invasive detection methods have been widely used due to their advantages of non-invasiveness and simplicity, which make them more easily accepted by patients. Invasive methods, while valuable, are less utilized. Although direct evidence for the pathogenic role of Demodex mites in rosacea is not available, indirect evidence provided by skin testing techniques shows significantly higher Demodex density (Dd) in rosacea patients than in controls.

Standardized superficial skin biopsies and reflection confocal microscopy are currently the most widely utilized non-invasive skin detection methods for Demodex infection.

The standardized superficial skin biopsy is a common non-invasive method to determine Dd which was developed by Forton et al. [15] in 1993 to detect Dd on a facial surface of 1 cm², with a normal value being <5 D/cm². Briefly, the method requires three steps. First, a drop of cyanoacrylate glue is placed in the center of a 1 cm² grid on a glass slide. Second, the slide is adhered to a designated area of the patient's forehead for 1 min. Third, the slide is gently removed from the skin, extracting the follicular contents [16]. During a clinical trial among 39 patients with mild-to-severe bilateral PPR, 35 patients were determined to carry ≥5 D/cm² on both sides of their faces by standardized superficial skin biopsy [17], which is higher than the normal value of <5 D/cm². A case-control study demonstrated via two consecutive standardized skin surface biopsies that Dd was higher in both ETR patients and PPR patients when compared to healthy controls and patients with other skin conditions [8]. The aberrant increase in Dd in rosacea patients indicates an association of Demodex with rosacea.

Reflectance confocal microscopy (RCM; also called confocal laser scanning microscopy) is an imaging technique that has emerged in recent years which demonstrates utility in diagnosing and monitoring rosacea [18‒25]. Under RCM, Demodex appears as small, round bodies with a hyperreflective contour in the horizontal section [26]. RCM also identifies adult mites with a worm-like appearance and four pairs of short legs within the hair follicle in the longitudinal section [23]. Substantial evidence provided via RCM indicates that rosacea patients have significantly higher Dd on the face than healthy controls [19, 25, 27]. According to a study that used RCM to investigate 30 PPR patients, 30 ETR patients, and 40 healthy controls, Dd was significantly higher in ETR and PPR patients compared to healthy controls and higher in PPR patients than in ETR patients [28]. RCM was used to examine 25 patients with facial rosacea before and after treatment in a prospective study. Demodex mites were significantly reduced after treatment with metronidazole 2% cream or gel in combination with systemic tetracyclines (doxycycline or minocycline) [19]. Another study found that residual mites detected by RCM were altered in number and appearance after topical ivermectin treatment in rosacea patients [21]. The appearance of Demodex mites decreased in brightness and the definition of their bright contours disappeared. Because these structural changes of the Demodex mites coincided with a significant clinical improvement in the affected patients, Demodex infection may have a role in the progression of rosacea.

Other methods for measuring Demodex include Tape method, Fluorescence Advanced Video Dermatoscopy (FAV), PCR for D. folliculorum, and so on. In 79 rosacea patients and 87 healthy controls detected by the tape method, the numbers of Demodex infections were 64 (81%) and 11 (13%), respectively [29]. FAV is a novel non-invasive method for in vivo detection of Demodex mites. Like RCM, FAV also detected an abnormally high density of Demodex mites in rosacea patients [24, 30]. Quantitative PCR was used to assess Demodex infestation in 50 people with facial rosacea and 48 healthy participants. Demodex was found to be present more frequently in rosacea patients than in healthy controls. The density of Demodex in rosacea patients was 5.7 times higher than in healthy participants [31]. All of these methods demonstrate abnormally elevated Dd in rosacea patients.

The invasive method for Dd detection generally refers to skin biopsy. One study compared the prevalence of Demodex mites on the faces of 80 patients with rosacea and 80 patients with other skin diseases via skin biopsy, and the results showed that the prevalence of mites in the rosacea group was significantly higher than that in the control group [32]. According to the results of another study, the facial biopsies of 75 patients with rosacea, 75 patients with discoid lupus erythematosus, and 75 patients with actinic lichen planus demonstrated that the prevalence of Demodex mites was significantly higher in patients with rosacea than in the other two skin disease patients [33]. Biopsies from several other studies have shown abnormally high densities of Demodex mites in patients with rosacea [34‒36].

The etiology and pathogenesis of rosacea remain uncertain. However, the possible role of Demodex mites in the pathogenesis of rosacea has been previously reported. Researchers co-cultured live Demodex mites with human sebocytes and found that Toll-like receptor 2 (TLR2) immune responses were down-regulated when mite numbers were low, but pro-inflammatory responses were activated when mite numbers increased [16]. It is speculated that in normal skin, low numbers of Demodex mites may downregulate the host immune TLR signaling pathway to favor their survival, whereas a large mite burden may trigger a host immune reaction via the TLR2 pathway, leading to the inflammatory skin changes typical of rosacea.

Furthermore, low doses of Demodex limited cytokine production, whereas high doses of Demodex promoted the release of inflammatory cytokines such as IL-10, IL-8, and IL-12p70 [16]. Elevated levels of the inhibitory inflammatory factor IL-10 can protect Demodex mites from excessive inflammatory responses, which may be another survival strategy. IL-8 can chemotacticize neutrophils and lead to a septic inflammatory response. Only the highest number of mite challenges (30) induced a significant increase in pro-inflammatory factor, IL-8 [16], consistent with papular pustules only occurring in severe cases of rosacea. IL-12p70 can induce a marked infiltration of CD4⁺ T helper cells, Langerhans cells, and macrophages, which also characterizes the histological findings in rosacea [37]. Furthermore, IL-17, a cytokine released by CD4⁺ T helper cells, is increased in Demodex blepharitis patients. IL-17 can induce the production of vascular endothelial growth factor-A (VEGF-A), promote angiogenesis, and cause rosacea-like lesions with telangiectasias [38]. These findings suggest that Demodex plays an important role in the pathogenesis of rosacea.

The treatment of rosacea also provides some evidence for the pathogenic role of Demodex in rosacea. The first-line treatments recommended by N Engl J Med in 2017 for PPR were topical ivermectin 1% cream, metronidazole 0.75% or 1% cream or gel for mildly affected patients, and oral tetracycline or doxycycline for moderately affected patients [39]. Established rosacea treatments primarily utilize direct acaricidal effects (such as permethrin) or anti-inflammatory properties (antibiotics such as metronidazole) or both (ivermectin). Twenty patients with mild-to-severe bilateral PPR were found to carry ≥5 Demodex mites per cm² on both sides of their faces before therapy. After treatment with permethrin 5% topical gel, Dd and the severity of clinical manifestations were greatly reduced compared to placebo [17].

Although Demodex mites were significantly reduced after treatment with metronidazole 2% cream or gel in combination with systemic tetracyclines (doxycycline or minocycline), the post-treatment mite levels were still higher than the average number of mites in healthy controls. Papules and pustular lesions resolved in most patients, but erythema was only partially reduced and telangiectasias persisted [19, 40]. Besides the anti-inflammatory effects of the antibiotics, such therapies for treating rosacea might also act on the mites’ endobacteria, as Clanner-Engelshofen et al. showed [41]. Forton et al. [42] found that metronidazole showed no acaricidal effect on Demodex, while benzyl benzoate induced complete elimination of Demodex. Topical administration of benzyl benzoate was effective in lowering Dd and improving clinical symptoms, both in demodicosis and PPR [43]. Due to the lack of anti-inflammatory effect, the therapeutic effect of benzyl benzoate is solely attributed to its acaricidal effect.

Ivermectin acts on rosacea in a dual anti-inflammatory and anti-parasitic manner, reducing Demodex burden and significantly improving clinical signs and symptoms [25]. In a monocentric pilot study, twenty Caucasian patients with moderate to severe rosacea were treated with topical ivermectin 1% cream once daily for 12 weeks. At weeks 6 and 12, the mean mite density was significantly decreased (p < 0.001), and IL-8, LL-37, HBD3, TLR4, and TNF-α gene expression were all downregulated at both time periods [44]. In addition, Ebbelaar et al. conducted a systematic review of both medical literature and clinical guideline recommendations, including three randomized trials, three extension studies, and two meta-analyses. Their results demonstrated that topical ivermectin was more effective than topical metronidazole in the treatment of moderate to severe PPR [45], suggesting that eliminating Demodex mites is beneficial for treating rosacea.

The debate regarding the pathogenic role of Demodex in rosacea is ongoing. Forton et al. analyzed the causal relationship direction between Demodex and rosacea at different levels (molecular, histologic, clinic), and concluded that PPR is caused by Demodex and not the inverse. Furthermore, another paper suggested that pityriasis folliculorum (PF) is the link between Demodex and PPR [4]. Finally, another study proposed that ETR is a subclinical stage of demodicosis that should be treated with acaricidal creams [5]. Hypothyroidism, pregnancy, corticosteroid administration, Staphylococcus epidermidis, contagiousness, and genetic background were proposed to influence Demodex proliferation via another study [46].

However, some phenomena remain unanswered: (i) ETR patients have been shown to have a milder inflammatory response and lower Dd than PPR patients [8, 28]. Furthermore, it is proposed that the more intense the inflammatory reaction, the lower the Dd. For this reason, the Dd in patients with PF was significantly higher than in patients with PPR [4]. However, if this theory is valid, then how is it that the Dd of ETR – which results in a lesser inflammatory response – is lower than that of PPR patients? (ii) What is the relationship between PF and ETR? It has been shown that PF is potentially misdiagnosed as ETR, as the two often coincide [5]. However, the Dd in ETR patients was lower than that in PPR patients, while the Dd in PF patients was significantly higher than that in PPR patients [4].

Gazi et al. collected human whole blood samples and analyzed skin-homing CD4⁺ T-cell subsets. They found that, compared with controls, non-rosacea patients with Demodex infection (Demodex group) had higher levels of Treg and Th9 cells, while PPR patients without Demodex infection (Rosacea group) showed increased levels of Th1 cells. PPR patients with Demodex infestation exhibited lower Treg but higher Th1 and Th22 cell levels than the Demodex group and higher Th2 cell levels than the rosacea group [47]. It appears that the rosacea disease process reduced the number of Treg cells and increased the number of Th1 cells in patients with Demodex infection. Similarly, Demodex infection increased the number of Th2 cells in patients with rosacea. Th2 immunity is activated in response to both allergens and parasitic infections [48]. In recent years, cases of rosacea-like lesions have been reported in patients following the use of dupilumab [49, 50]. Live Demodex mites’ colonization of the face was discovered after further study by direct scraping of pustules [50]. Th2 immune responses are essential in the control of extracellular parasitic infections [47]. Dupilumab is a Th2 pathway inhibitor, which may boost Demodex proliferation [50]. The elevated Th1 skin-homing cells in rosacea [47] can also inhibit the proliferation of Th2 cells [51]. These may explain the massive proliferation of Demodex in rosacea. A low number of Demodex can proliferate under the protection of Treg cells [47] and IL-10 [16] without triggering an inflammatory response. A high load of Demodex can trigger an immune response. The factors affecting the proliferation of Demodex mites are complex [46], which is why most patients with Demodex infections are asymptomatic while some patients show a robust inflammatory response.

Therefore, a hypothesis is proposed based on all of the above. Usually, Demodex mites can proliferate under the protection of local Treg cells, and proliferation of Demodex presents as PF. If a patient develops rosacea and a Demodex infection, Th1 immunity is triggered and Th2 immunity is inhibited [47]. The immune balance between Th2 and Treg is disrupted. Without the suppression of Th2 immunity, Demodex takes advantage of the opportunity to proliferate rapidly during the ETR phase. This is consistent with Forton et al.’s [8] assertion that ETR is a subclinical stage of demodicosis. Overproliferative Demodex mites trigger an increase in Th2 cells and restore them to their previous levels [47]. Due to the decrease in Treg cells and the increase in Th22 cells, large numbers of Demodex mites cause severe inflammatory responses that manifest clinically as papular pustules. This is in agreement with Forton et al. that PPR is indeed caused by Demodex [4]. Therefore, both rosacea treatment and acaricidal treatment are necessary. A combination of topical Th1 inhibitor and ivermectin is recommended for the treatment of rosacea patients with Demodex infection. Downregulation of Th1 immunity can not only inhibit the progression of rosacea but also promote the upregulation of Th2 immunity, thereby inhibiting the overgrowth of Demodex. The administration of ivermectin may alleviate symptoms caused by the excessive proliferation of Demodex mites (shown in Fig. 2).

Rosacea favors the proliferation of Demodex, and the over-proliferation of Demodex aggravates the symptoms of rosacea, thus forming a vicious circle. This cycle is interrupted when pharmacological interventions are used. However, if favorable factors persist, the vicious cycle tends to repeat again. This is likely why the use of the acaricides such as ivermectin, benzyl benzoate, and metronidazole, which has only anti-inflammatory effects, can reduce the number of Demodex mites and improve the patient’s symptoms but cannot eliminate telangiectasia. This is also likely why acaricidal and anti-inflammatory treatments are prone to relapse [19, 25, 43]. Therefore, early use of ivermectin, which has acaricidal and anti-inflammatory properties, in combination with topical Th1 inhibitors to block the progression of the inflammatory response which facilitates Demodex survival, can break this vicious cycle and stop disease progression.

Rosacea favors the growth of Demodex, and overgrowth of Demodex exacerbates the symptoms of rosacea, creating a vicious cycle. The changes in immune interactions when PF encounters ETR and throughout the course of the disease need further exploration. The early use of topical ivermectin in combination with topical Th1 inhibitors early should be considered for the treatment of patients with rosacea and Demodex infection.

Rosacea appears to have a close bond with Demodex. Recent studies have implicated that Demodex mites are involved in the development of rosacea. Based on a comprehensive review of the current literature, we have cited substantial evidence supporting this relationship, proposed a possible vicious cycle between Demodex and rosacea, and illustrated the changes in immune responses during this process. In addition, we recommend that patients with both rosacea and Demodex infection be treated early with a combination of topical ivermectin and topical Th1 inhibitors to improve treatment outcomes.

All authors have no conflict of interest to declare.

This work was supported by the National Natural Science Foundation of China (No. 81974475; No. 82173423).

Fen Wei conceived the idea for this review, performed the literature search, and wrote the first draft of the manuscript. Li Li drafted the visualization. Yi Kong, Song Zhang, and Jian Jiang provided critical feedback. Xiaofeng Yan and Kevin Varghese participated in manuscript writing and editing. Bao Chai and Hongxiang Chen supervised the work. All authors commented on previous versions of the manuscript and critically revised the draft. All authors have read and approved the final version of the manuscript.