Correspondence: Daichi Morioka, M.D., Ph.D., Department of Plastic and Reconstructive Surgery, Showa University, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142-8866, Japan. Email: firstname.lastname@example.org
Axillary osmidrosis often disturbs a person's social life, particularly in Asian countries. However, the clinical aspects of this condition have not been well documented in the English-language published work. This study aimed to provide information on the features of axillary osmidrosis, with a particular focus on sex differences. A retrospective review was made of the charts for 723 Japanese patients (492 female, 231 male). The mean age at initial presentation (29.1 years) was nearly the same for males and females. Almost all patients (96.1%) had wet earwax, which was extremely high compared to its frequency in the general Japanese population. An association with hyperhidrosis was seen in 61.8% of these patients. Subjective odor levels in female patients were significantly lower than those in males (P <0.001). A positive family history was more frequent for females than for males (P <0.001), and prior treatment history was also more frequent for females than for males (P <0.015). Most patients (86.6%) had received some treatments in our clinic. There were significantly fewer females who underwent surgical treatments compared to males (P =0.026), as females preferred less invasive techniques (P <0.001). Several features, including male/female ratios, and associations of wet earwax and hyperhidrosis, corresponded to previously reported data on axillary osmidrosis. Female patients were more concerned with axillary odor than males, and females had a tendency for polysurgery.
Axillary osmidrosis (AO) is a distressing condition that can have a profoundly detrimental effect on patients' social lives due to the unpleasant odor and yellowish staining of clothing. Particularly in Asian nations, including Japan, Korea and China, patients are often left out of social activities or teased by colleagues and friends, and sometimes become social dropouts.[1-3] This condition was described in the oldest anthology of Japanese poems, Manyoshu (8th century), in which a person with AO was teased.
Many patients who suffer from axillary odor feel alienated, are not satisfied with “drug store grade” deodrants and seek a physician's help. There is a wide range of non-surgical and surgical treatments available to solve these problems.[5-7] To the best of our knowledge, at least 55 reports on AO have been published in the English-language published work. Most of these reported on the surgical techniques used and the outcomes, while others have reported on conservative treatments or AO genetics. However, the clinical aspects of this condition have not been well documented.
A few older studies in the Japanese-language published work reported that the majority of patients were female, onset was earlier in females than in males, and there was a strong association between AO and wet earwax, which was higher in females than in males.[10, 11]
It is very important to understand the recent information available for AO patients for clinical consultations, differentiation from olfactory reference syndrome and for determining surgical indications. This study provides the first large analysis on the clinical features of AO in the English-language published work, with a particular focus on sex differences. In addition, we compared the clinical features of patients with wet earwax and those with dry earwax.
From November 2007 to June 2011, over 900 patients came to our private clinic, the Nihonbashi Plastic Surgery and Skin Care Clinic, with complaints of axillary odor.
During their first interview with the authors, patients placed a small piece of gauze into both armpits. Using modifications of methods in previous studies,[12, 13] the gauzes were removed 15 min later and odor levels were scored by sniffing by two of the authors (D. M. and Y. A.): 1, no odor; 2, faint; 3, moderate; 4, severe; and 5, very severe. When they scored different levels, the highest was recorded. For this subjective odor rating scale, there was good agreement between the scorers; interobserver variability assessed by a kappa statistic was 0.954. Patients who had used deodorants within 24 h, received botulinum toxin injection within the previous 12 months or had any other previous treatments were excluded from the gauze test.
In addition to odor levels, earwax type and associated hyperhidrosis were evaluated. A diagnosis of hyperhidrosis was made as in a previous study. Patients who had olfactory reference syndrome and patients who complained of axillary hyperhidrosis but without osmidrosis were excluded. Finally, a retrospective medical chart review of 723 patients with AO was conducted for this study (Fig. 1).
Demographic information included age, sex and family history. When a patient had a family history of axillary osmidrosis among second-degree relatives, the history was recorded as positive. Referral sources were categorized into Internet referrals, personal referrals, professional referrals and others. Clinical findings recorded included odor levels, type of earwax, association with hyperhidrosis and past treatment history.
Treatments provided in our clinic were also analyzed as of 1 August 2011. Our clinic provides two different surgical procedures. First, the open method, involves subcutaneous tissue excision with a modification described by Rigg. It requires that patients have long-term (5–7 days) compression dressings. This procedure has a slightly higher rate of complications, such as flap necrosis and hematoma. It is covered by the national heatlth insurance system of Japan. Second, the closed method, involves a suction-assisted cartilage shaver system, is less invasive, has fewer complications and only requires a short term (1–2 days) of compression dressings. However, it is not covered by the national heatlth insurance system.
Our statistical analyses focused on possible sex differences. Mean ages of female and male patients at initial presentation were compared by Student's t-test. Odor levels between females and males, and odor levels between those with dry earwax and wet earwax were compared by Mann–Whitney U-tests. Other categorical variables were compared by a χ2-test -squared test or by Fisher's exact test when the expected values were too small to be analyzed by a χ2-test. Female–male comparisons for wet/dry earwax ratios, consultations/treatments performed in our clinic and closed methods/open methods for surgical treatments were made using 2 × 2 contingency tables. P < 0.05 was considered statistically significant. Statistical analysis was performed using Microsoft Excel 2010 (Microsoft, Redmond, WA, USA).
Among the total 723 patients with AO, 492 were female (68.0%) and 231 were male (32.0%). Figure 1 shows the age distributions for female and male patients, and their demographic data and clinical findings are shown in Table 1. There were no statistically significant differences in the age distribution patterns for females and males and the mean age at initial presentation (29.1 years for females and 29.2 years for males).
Table 1. Demographic data and clinical findings
*Internet also included mobile phone network services. Personal referrals included word of mouth from family members, friends and colleagues. Professional referrals included referrals by another plastic surgeon, dermatologist, general practitioner and other medical professionals. **Patients who had used deodorants within 24 h, received botulinum toxin injection within the previous 12 months or had any other previous treatments were excluded. ***Conservative treatments included medical deodorant (excluding drug store grade), laser depilation for the purpose of reducing odor, iontophoresis, electric cauterization and botulinum toxin injection.
Age (mean ± SD)
29.1 ± 7.97
29.2 ± 8.25
29.1 ± 7.84
Source of referral, n (%)*
Earwax, n (%)
Hyperhidrosis, n (%)
Odor levels, n (%)**
Positive family history, n (%)
Past treatment history, n (%)
Most patients (92.5%) had referred themselves via the Internet. There were no significant differences in sex or age groups with regard to the referral source (P =0.896).
Almost all of these patients (96.1%) had wet earwax, which exhibited significant sex differences (P =0.037); however, this was considerably higher compared to the frequency in the general Japanese population.
An association with hyperhidrosis was observed in 61.8% of the patients, which did not show significant sex differences (P =0.429). This frequency corresponded to previously reported data on AO. A Mann–Whitney U-test was used to compare subjective odor levels between sexes; the levels in females were significantly higher than those in males (P <0.001). A positive family history was recorded for 75.5% of these patients and was more frequent for females than for males (P <0.001).
A previous history of some medical treatment was recorded for 18.5% of patients; female patients reported this history significantly more often than males (P =0.015). Both prior history of conservative treatments and a previous surgical history were more frequent for females than for males (P =0.018 and =0.039, respectively).
Table 2 shows the treatments performed for the patients in this study; 510 of 589 patients (86.6%) had undergone at least one of these treatments in our clinic as of 1 August 2011. Those patients who had a prior treatment history were excluded from this analysis (n = 134). Although significantly more females had conservative treatments (P =0.040), this was due to significant differences in laser depilation (P =0.007). There were no significant sex differences for other treatments including medical deodorant (P =0.139) and botulinum toxin injection (P =0.474). In contrast, significantly fewer females had surgical treatments as compared to males (P =0.026). Furthermore, a 2 × 2 contingency table analysis showed that female patients preferred a less invasive closed method, whereas male patients preferred an open method.
Table 2. Treatments performed in our clinic
We excluded 134 patients who had prior treatment history. *Some patients had more than two treatments. **Compared using 2 × 2 contingency table. ***Medical deodorant prescribed in our clinic consists of a 20% aluminum chloride solution without alcohol. Laser depilation for esthetic reasons only was excluded.
Consultation only, n (%)
Total treatments, n (%)*
Botulinum toxin injection
The clinical features of AO patients with wet earwax and those with dry earwax were also compared (Table 3). As was shown in Table 1, only 3.9% of these patients had dry earwax. An association with hyperhidrosis was not significantly different between these earwax types (P =0.110). Subjective odor levels and a positive family history for patients with wet earwax were significantly higher than those with dry earwax (P <0.001 for both). In our clinic, patients with dry earwax more frequently had conservative treatments (P =0.001), whereas patients with wet earwax more frequently had surgery (P =0.024).
Table 3. Comparisons of clinical features for patients with wet earwax and those with dry earwax
*Patients who had used deodorants within 24 h, received botulinum toxin injection within the previous 12 months or had any other previous treatments were excluded. **Some patients had both treatments.
Total patients, n (%)
Hyperhidrosis, n (%)
Odor levels, n (%)*
Positive family history, n (%)
Treatments in our clinic, n (%)**
It has been well known for more than 75 years that AO is strongly correlated with the type of earwax.[10, 11] Both of these characteristics originate from the apocrine glands and exhibit a dominant pattern.[1, 10] The frequency of wet earwax in the general population of Japan is thought to be approximately 12.6–22.4%, while the frequency is reportedly 78–97% in AO patients. This high frequency corresponded to what we observed (96.1%) and what was observed in a large study done in Korea (96.0%).
Recently, the ABCC11 gene on chromosome 16 was found to determine the type of earwax and its association with AO. Either the GG or GA genotype of ABCC11 determines the phenotypic expression of wet earwax; it is an autosomal dominant trait, and is found in 98.7% of AO patients. In contrast, the AA genotype is associated with a considerably smaller number of apocrine glands, which results in dry earwax. A recent large study using genetic analysis showed that wet earwax (GG or GA genotype) was found in 22.9% of 1963 Japanese students.
Although the reason why females had a significantly higher frequency of wet earwax compared to males in this study is unknown, a few other studies have shown the same pattern.[1, 10, 11] Moreover, although the great majority of patients were female in almost all previous studies that reported surgery for AO,[3, 5, 6, 18, 20] the reasons for sex differences were not discussed. There are probably some cultural, psychological and/or hormonal factors that make females more sensitive to axillary odor.[1, 13] Many olfactory function tests have shown that females performed better than males. It has also been reported that olfactory sensitivity to a major component of odorous sweat, isovaleric acid, was higher for Japanese females than for Japanese males. In addition to these male–female differences, our study found that female patients exhibited significantly lower odor levels by the gauze test and were more aware of their family histories than were males. It is plausible that these findings may be due to sex differences in olfactory sensitivity.
A wide variety of surgical treatments for AO have been developed, including endoscopic transthoracic sympathectomy, superficial liposuction, ultrasound-assisted aspiration, subcutaneous tissue excision/curretage/shaving and total excision of axillar skin. Several papers have discussed treatment indications based on the odor level. In general, conservative treatments were recommended for patients with faint odor and surgical treatments were recommended for patients with moderate to severe odor.[2, 12, 13] Yoo et al. proposed a simple treatment algorithm based on the type of earwax and family history. In addition, we propose that an indication for treatments should be made based on a patient's age, because odor may only be temporarily strong during adolescence or it may lessen with older age.[1, 11] Indeed, our clinic primarily recommends conservative treatments for patients as follows: faint odor (level 1–2); moderate odor (level 3) with dry earwax; younger than 18 years of age; and older than 45 years of age. However, we have performed surgery for some of these patients for psychosocial reasons, such as being teased at school.
Differences in the clinical features of AO patients with wet earwax and those with dry earwax have not been previously discussed in the published work. Based on experience, patients with dry earwax probably have less odor and are less likely to have a positive family history compared to patients with wet earwax. Those observations were statistically validated in our study. Indeed, as shown in Table 3, we primarily recommended conservative treatments rather than surgery for those AO patients with dry earwax.
Thus, understanding the clinical aspects of AO provides an important source of information for consultations and for determining surgical indications. In addition, it may be helpful in the differential diagnosis of olfactory reference syndrome. If a patient has dry earwax and scores at level 1 in the gauze test, then olfactory reference syndrome should be suspected.
In conclusion, our results for male/female ratio, frequency of wet earwax and the rate of association with hyperhidrosis were nearly identical to previously reported results. When sex differences were analyzed, our results suggested that females were more sensitive to axillary odor and were more concerned with treatments. When they had surgery, females had a significantly higher likelihood of preferring a less invasive method, even though this is much more expensive than other treatments that are covered by the national health insurance system of Japan. However, females were sometimes unsatisfied with their first treatment and tended to have polysurgery.
Sources of referral for AO have not been reported in any of the published work, regardless of language. This study found that, irrespective of age and sex, more than 90% of these patients obtained information about axillary odor from the Internet, and much of this information is provided by online advertisements and social network services. Due to the rapid access provided by the Internet, the body odor market in Japan has increased by $US 1 billion (personal communication). However, some deodorants, supplements, alternative therapies or even medical treatments appear to be unreliable. Because patients are confused by this flood of uncertain information, as physicians and surgeons we should be responsible for providing more precise information and correct treatment options.