Rosacea is a chronic inflammatory disease of skin and it is commonly observed in middle-aged women1. It shows various manifestations such as intermittent or persistent erythema, telangiectasia, papules and pustules, ocular findings with or without cutaneous symptoms, and phymatous change. Although the pathogenesis of rosacea remains unclear, it has been suggested that recurrent dilatation of cutaneous vessels followed by impairment of their regulatory function, invasion of inflammatory cells, changes in the perivascular tissue and infestation by microorganisms may play critical roles1-3. Many trigger or aggravating factors have also been reported. It has generally been accepted that sun exposure may play a role in developing or aggravating rosacea, and especially in fair skinned people1,4-6. However, their relationships with the development or the progression of rosacea have not yet been evaluated in an Asian population.

There have been several attempts by western researchers and clinicians to establish a proper classification and severity grading system for rosacea; however, it was not easy to set up a unified standard classification and grading system due to this disease's variable clinical manifestations and the lack of definite diagnostic methods2,7,8. Therefore, the National Rosacea Society (NRS) Expert Committee suggested a new classification and severity grading system in 2002 and 2004, respectively9,10. To the best of our knowledge, there has been no prior clinical study in the medical English literature with using the new NRS classification and grading system for rosacea in any type of population. We surveyed the clinical features in Korean patients with rosacea and who visited our department during the recent 5 years and we classified them by the new classification and grading system. We also statistically determined if there was any correlation between each subtype or between the clinical features and the subtypes, and also between the degree of sun exposure and each subtype.

One hundred sixty eight patients who were diagnosed with rosacea from March 2002 through February 2007 were enrolled in this study. Each patient's Fitzpatrick's skin type was determined by their experience of skin reactions to sun exposure and the clinician's inspection. The patient's medical records and clinical photos were reviewed. Age, gender, the duration of symptoms, the past history, the triggering or aggravating factors, the relieving factors, the area of living and the degree of daily sun exposure were analyzed. The primary features of rosacea (flushing, nontransient erythema, papules, pustules and telangiectasia), as suggested by the NRS, were adopted as a diagnostic tool. The patients with at least one of the primary features were considered as having rosacea. The secondary features such as burning or stinging, a dry appearance, edema, ocular manifestations, a peripheral location and phymatous changes were also checked. The four subtypes of rosacea were classified by a physician's global assessment; the erythematotelangiectatic, papulopustular, phymatous and ocular types. The new classification system defines the potential overlap of subtypes; therefore, a patient may have two or more subtypes. The characteristics of each subtype are described in Table 1.

With using the standard grading system suggested by the NRS in 2004, the severity of the primary and secondary features and each subtype were graded as absent, mild, moderate or severe (0~3). The Berg-Liden classification system6 was used for the erythematotelangiectatic subtype: erythema on the nose and perinasal area or cheeks only was regarded as mild; erythema with or without telangiectasia on both the nose and cheeks with a relatively light hue was regarded as moderate; erythema with or without telangiectasia on almost the entire face or a deeper hue was regarded as severe. The extent or number of lesions and the confluence and grade of the inflammation were all considered in grading the papulopustular subtype. The ocular subtype was clinically graded by the severity of the subjective symptoms or the clinical findings. Finally, the phymatous subtype was clinically graded by the extent and severity of the thickening, irregularity and nodularity.

We statistically examined if there was any correlation between each subtype and the incidence or severity. Moreover, we found significantly frequent clinical features in the individual subtypes. Finally, the relationship between the degree of sun exposure and each subtype was analyzed.

In this study, we classified and graded Korean patients with rosacea by using the new NRS classification and grading system. We also derived the characteristics of each subtype and the relationships between them. Interestingly, we found that there was a close relationship between sun exposure and the erythematotelangiectatic subtype of rosacea.

The concept of rosacea has recently been changed so that it allows the overlap of variable manifestations with different severity. The new classification and severity grading system suggested by the NRS strongly reflects this tendency. Therefore, one patient may have more that one overlapping subtype simultaneously with different severity grades. The new system seems more proper and convenient when evaluating the patients with rosacea. It would also be helpful for treating rosacea patients because of the system's more realistic understanding of this disease.

The existence of a wide clinical spectrum and various subtypes causes clinicians to assume that many factors appear to contribute differently to the development or aggravation of rosacea. Impairment of vascular reactivity and homeostasis, changes or degenerations of dermal structures, inflammatory processes, pilosebaceous unit abnormalities and infestations of microorganisms such as Demodex and Helicobacter pyroli may directly or indirectly influence the development of rosacea2,3,11-14.

The incidence of rosacea varies among different studies. Although rosacea affects all races, it has generally been accepted that the disease more commonly develops in Caucasians and fair-skinned populations1,4,5,7. However, our study suggests that chronic sun exposure can also be a causative factor for rosacea in an Asian population.

The male:female ratio of 1:2.29 in our study was not much different from that of the other studies. The mean age of the patients was 47.8 years and the peak age of incidence was in the 40s to 60s, which was slightly older than what has been previously reported. As peoples' life expectancy is prolonged, the proportion of elderly patients with rosacea in our clinic may have increased. Another possibility may be the effect of long-term sun damage in elderly persons.

Though there are many factors in the development or aggravation of rosacea, sun exposure and hot bath/exercise were revealed in this study to be the two most common precipitating factors, followed by alcohol ingestion and a warm environment. Sun damage has been accepted to be a critical factor in the development of rosacea since the distribution of erythema is most prominent on the facial convexities5. In addition, solar elastosis is a frequent histologic finding in the lesions of rosacea12. In contrast, some studies have demonstrated that the incidence of rosacea was not higher for the groups of patients with chronic sun exposure. Some studies have also focused on the fact that sun exposure to non-facial areas did not induce the disease10,15,16. These findings could support the hypothesis that sun exposure may not be an independent factor for the development of this disease. From the assumption that chronic sun exposure may induce or aggravate rosacea, we analyzed the correlation between the degree of sun exposure and the development of each subtype. It's interesting that the severity of the erythematotelangiectatic subtype had correlation with the degree of sun exposure. However, such a relationship was not found for the papulopustular, ocular and phymatous subtypes. Altogether, our result suggests that sun exposure has a different influence on each subtype.

Hot baths, exercise and alcohol ingestion can induce rosacea, as these things all dilate cutaneous vessels due to an increased core temperature. This phenomenon is especially seen in the facial area, where the vasculature is prominent as well as having a shallow distribution in relation to the skin surface17,18.

The majority of our patients had the erythematotelangiectatic subtype, which has been considered to be an initial phase of rosacea. In our study, we found that grade of this subtype increased with the duration of disease. It was notable that the grade of erythematotelangiectatic rosacea had correlation with the development of ocular rosacea, while no clinical correlation was found between the grade of erythematotelangiectatic rosacea and the papulopustular and phymatous subtypes of rosacea. This result highlights the fact that ocular rosacea is an extension of the clinical spectrum of erythematotelangiectatic rosacea, while the other subtypes are not related to ocular rosacea.

About half of our patients had papulopustular rosacea. This result was relatively higher than the general consensus from other studies. Since the NRS system permits overlapping of the subtypes, clinicians do not need to select one particular subtype. Accordingly, this change in classification system is postulated as the cause for a higher rate of the incidence of the papulopustular subtype. In addition, our study revealed several characteristics of papulopustular rosacea. First, a burning or stinging sensation, which was the most common secondary feature, had no correlation with the development of the severity of this subtype, while a strong correlation was found with the erythematotelangiectatic and ocular rosacea. Second, edema (as a secondary feature) had correlation with the severity of the subtype. These results are in close agreement with those of several authors19,20.

Twenty four (14.3%) patients had ocular rosacea, including 21 cases with a mild grade. The reported incidence of ocular rosacea is variable according to the researcher, and it has ranged from 6% to 50%. This variable incidence of ocular rosacea is considered to be partly due to the condition's subjective nature, and it is partly due to the absence of definite diagnostic criteria. Some controversy exists concerning the casual relationship between the ocular and cutaneous manifestations21-23. In this study, we speculated that the development of ocular rosacea increased with the aggravation of cutaneous manifestations, and especially flushing and telangiectasia.

All eight patients diagnosed with the phymatous type had a mild degree of rhinophyma. Since this subtype is considered as late rosacea, it always accompanied the other subtypes in this study. Even though the nose was the only involved site in our study, lesions can occur on the cheeks (gnathophyma; zygophyma), forehead (metophyma), eyelids (blepharophyma) and ears (otophyma)24.

There have been few reported clinical studies that have used the new classification and grading system. There have been four previous studies in the Korean medical literature that have investigated the clinical spectrum of rosacea25-28. Among them, Shin and Kim28 adopted the new NRS classification and grading system. Although they reported overall results that were similar to our results, Shin and Kim's study did not accept the potential overlap of each subtype. This new system is useful in analyzing the subjective symptoms and cutaneous manifestations of each subtype. A well established consensus on rosacea in an Asian population is necessary since this disease has become more common. In addition, sun exposure should be considered as a critical factor in the development and aggravation of rosacea in Asian people.