Portal de Periódicos da CAPES

Phototrichogram Analysis of Japanese Female Subjects with Chronic Diffuse Hair Loss

2003; Elsevier BV; Volume: 8; Issue: 1 Linguagem: Inglês

10.1046/j.1523-1747.2003.12184.x

1529-1774

Rie Ueki, Ryoji Tsuboi, Hideoki Ogawa, Yutaka Inaba,

melanin and skin pigmentation

Hair growth patterns of 101 Japanese female subjects with diffuse, chronic hair loss and 58 healthy Japanese female volunteers were categorized into subgroups using noninvasive quantitative methods after determining the key parameters of hair growth. Phototrichogram was performed at 0 and 48 h after clipping hairs in the parietal region of the scalp. Shaft diameters of the excised hairs were then measured. Multiple regression analysis indicated that hair densities, hair diameters, short hair ratios, and hair growth rates, but not anagen hair ratios, were significant, in order of decreasing importance, for grading female diffuse alopecia. Using cluster analysis, hair growth patterns among subjects complaining of diffuse hair loss were divided into six abnormal groups (n=60), two borderline groups (n=21), and one normal group (n=20). The control subjects judged to be normal by macroscopic observation, actually included two subjects with borderline hair growth patterns and one abnormal subject. Most of the abnormal groups shared features of female androgenetic alopecia. Hair patterns showing a decrease in hair density but without vellus hair change, however, emerged as the most prevalent and distinct pattern of chronic diffuse hair loss among the Japanese female subjects. The phototrichogram, combined with the measurement of hair diameters, is an accurate tool for assessing hair growth patterns, especially in detecting the slight changes indicative of the early phase of diffuse alopecia. Hair growth patterns of 101 Japanese female subjects with diffuse, chronic hair loss and 58 healthy Japanese female volunteers were categorized into subgroups using noninvasive quantitative methods after determining the key parameters of hair growth. Phototrichogram was performed at 0 and 48 h after clipping hairs in the parietal region of the scalp. Shaft diameters of the excised hairs were then measured. Multiple regression analysis indicated that hair densities, hair diameters, short hair ratios, and hair growth rates, but not anagen hair ratios, were significant, in order of decreasing importance, for grading female diffuse alopecia. Using cluster analysis, hair growth patterns among subjects complaining of diffuse hair loss were divided into six abnormal groups (n=60), two borderline groups (n=21), and one normal group (n=20). The control subjects judged to be normal by macroscopic observation, actually included two subjects with borderline hair growth patterns and one abnormal subject. Most of the abnormal groups shared features of female androgenetic alopecia. Hair patterns showing a decrease in hair density but without vellus hair change, however, emerged as the most prevalent and distinct pattern of chronic diffuse hair loss among the Japanese female subjects. The phototrichogram, combined with the measurement of hair diameters, is an accurate tool for assessing hair growth patterns, especially in detecting the slight changes indicative of the early phase of diffuse alopecia. female androgenetic alopecia chronic telogen effluvium Advances in understanding the pathogenesis of androgenetic alopecia and the development of new drugs to treat this condition have led to the recent publication of a series of reviews on hair biology and on the management of hair problems (Rebora, 1997Rebora A. Telogen effluvium.Dermatology. 1997; 195: 209-212Crossref PubMed Scopus (41) Google Scholar;Van Neste and Rushton, 1997Van Neste D.J.J. Rushton D.H. Hair problem in women.Clin Dermatol. 1997; 15: 113-125Abstract Full Text PDF PubMed Scopus (43) Google Scholar;Whiting, 1998Whiting D.A. Male pattern hair loss: current understanding.Int J Dermatol. 1998; 37: 561-566Crossref PubMed Scopus (46) Google Scholar;Price, 1999Price V. Treatment of hair loss.N Engl J Med. 1999; 341: 964-973Crossref PubMed Scopus (294) Google Scholar;Sinclair, 1999Sinclair R.D. Diffuse hair loss.Int J Dermatol. 1999; 38: 8-18Crossref PubMed Scopus (65) Google Scholar;Tosti et al., 1999Tosti A. Camacho-Martinez F. Dawber R. Management of androgenetic alopecia.J Eur Acad Dermatol Venereol. 1999; 12: 205-214Crossref PubMed Scopus (47) Google Scholar;Tran and Sinclair, 1999Tran D. Sinclair R.D. Understanding and managing common baldness.Aust Fam Physician. 1999; 28: 248-253PubMed Google Scholar;Sawaya and Shapiro, 2000Sawaya M.E. Shapiro J. Androgenetic alopecia; new approved and unapproved treatments.Dermatol Clin. 2000; 18: 47-61Abstract Full Text Full Text PDF PubMed Scopus (32) Google Scholar). Androgenetic alopecia in men is characterized by the thinning of hairs in the frontal and/or vertical areas, starting in mid-life and culminating in complete hair loss in the affected areas. On the other hand, androgenetic alopecia in women, also known as female pattern hair loss, has been recognized generally as diffuse thinning, often worse centrally, but not culminating in complete baldness. The clinical state of "chronic diffuse alopecia" in middle-aged women is caused by various factors, including female androgenetic alopecia (FAGA), chronic telogen effluvium (CTE), intake of chemical agents or drugs, nutritional deficiency, and systemic disorders such as endocrine diseases and serious chronic diseases (Dawber et al., 1998Dawber R.P.R. de Berker D. Wojnarowska F. Disorders of hair.in: Champion R.H. Burton J.L. Burns D.A. Breathnach S.M. Textbook of Dermatology 6th edn. London, Blackwell Science1998: 2869-2973Google Scholar). Among them, the majority of the idiopathic cases are thought to be FAGA, CTE, or a combination of both. Chronic telogen effluvium (CTE) is characterized by the abrupt onset of shedding and thinning of hair over the entire scalp, with fluctuations in severity throughout its course (Whiting, 1996Whiting D.A. Chronic telogen effluvium: increased scalp hair shedding in middle-aged women.J Am Acad Dermatol. 1996; 35: 899-906Abstract Full Text PDF PubMed Scopus (185) Google Scholar). The differential diagnosis of FAGA and CTE, however, especially in cases with mild clinical manifestations, is sometimes difficult, partly due to a lack of information about the mechanisms of these two types of hair loss. FAGA without the deep frontotemporal recessions has been graded, by macroscopic observation, into three levels byLudwig, 1977Ludwig E. Classification of the types of androgenetic alopecia (common baldness) occurring in the female sex.Br J Dermatol. 1977; 97: 247-254Crossref PubMed Scopus (572) Google Scholar; however, FAGA patients with deep frontotemporal recessions are sometimes encountered (Ludwig, 1968Ludwig E. The role of sexual hormones in pattern alopecia.in: Baaccaredda-Boy A. Moretti G. Frey J.R. Biopathology of Pattern Alopecia. New York, Karger1968: 50-60Google Scholar), indicating that the Ludwig classification is too simple to describe all of the features of this condition (Olsen, 1999Olsen E.A. The midline part. An important physical clue to the clinical diagnosis of androgenetic alopecia in women.J Am Acad Dermatol. 1999; 40: 106-109Abstract Full Text Full Text PDF PubMed Scopus (88) Google Scholar).Whiting, 1996Whiting D.A. Chronic telogen effluvium: increased scalp hair shedding in middle-aged women.J Am Acad Dermatol. 1996; 35: 899-906Abstract Full Text PDF PubMed Scopus (185) Google Scholar emphasized the utility of histologic findings in distinguishing CTE from FAGA. Taking biopsies from a large number of subjects is inadvisable, however, because of the invasive nature of the procedure. Recent reports (Rushton et al., 1993Rushton D.H. De Brouwer B. De Coster W. Van Neste D.J.J. Comparative evaluation of scalp hair by phototrichogram and unit area trichogram analysis with the same subjects.Acta Derm Venereol (Stockh). 1993; 73: 150-153PubMed Google Scholar;Van Neste et al., 1994Van Neste D.J.J. De Brouwer B. De Coster W. The phototrichogram: analysis of some technical factor of variation.Skin Pharmacol. 1994; 7: 67-72Crossref PubMed Scopus (35) Google Scholar;Van Neste and Rushton, 1997Van Neste D.J.J. Rushton D.H. Hair problem in women.Clin Dermatol. 1997; 15: 113-125Abstract Full Text PDF PubMed Scopus (43) Google Scholar) have indicated the advantages of the phototrichogram, a noninvasive quantitative method, over the trichogram and biopsy. The utility of the phototrichogram has been demonstrated in the assessment of hair growth in normal subjects and men with androgenetic alopecia (Hayashi et al., 1991Hayashi S. Miyamoto I. Takeda K. Measurement of human hair growth by optical microscopy and image analysis.Br J Dermatol. 1991; 125: 123-129Crossref PubMed Scopus (80) Google Scholar;Rushton et al., 1993Rushton D.H. De Brouwer B. De Coster W. Van Neste D.J.J. Comparative evaluation of scalp hair by phototrichogram and unit area trichogram analysis with the same subjects.Acta Derm Venereol (Stockh). 1993; 73: 150-153PubMed Google Scholar;Ishino et al., 1994Ishino A. Tsuji Y. Uzuka M. et al.Quantitative evaluation of male pattern baldness by measurements of anagen ratio and hair shaft diameter.Jpn J Dermatol. 1994; 104: 1737-1746Google Scholar). In this study, instead of collecting typical cases of FAGA and CTE, we examined hair growth patterns of Japanese female subjects complaining of diffuse hair loss and attempted to divide "chronic diffuse alopecia" into subgroups. The combination of the phototrichogram and measurements of mean hair diameter were used as a new diagnostic approach. Also addressed were issues of identifying the parameters of hair growth necessary for analyzing chronic diffuse alopecia, characteristics of each alopecia subgroup and commonalities in the patient histories. One hundred and eleven Japanese women (from 17 to 70 y old) complaining of a gradual shedding and thinning of scalp hair without frontotemporal recession were enrolled. These subjects had noticed changes in their hair over the course of several years to a decade prior to commencement of this study. Subjects with postpartum alopecia, alopecia areata, and alopecia induced by chemical agents or secondary to systemic diseases were excluded. Sixty healthy Japanese women (from 21 to 64 y old) with no symptoms of diffuse alopecia or history of diseases related to hair growth were used as controls. All subjects had dark brown or black hair. Informed consent was obtained from all subjects. The study was approved by the review board of Juntendo University Hospital. All subjects received a blood analysis, including a complete blood cell count with differentiation, as well as liver function, microsome, and thyroid tests. Values for creatinine, blood urea nitrogen, serum iron, total iron-binding capacity, anti-nuclear antibody, complement titer (CH50), free thyroxine, thyroid-stimulating hormone, serum testosterone, and serum estradiol were also obtained. Levels of prolactin and 11-hydroxycorticoid were measured in some subjects. Hairs from the upper parietal area (1.5×1.5 cm square) were clipped using scissors. The state of the hairs was recorded using an optical microscope (microscope with a built-in monitor VH, KEEYENCE) at the time of clipping (0 h) and 48 h afterwards. This procedure was performed once on each subject during the months from autumn to spring. In order to obtain an accurate measurement of hair length and number, a transparent glass slide was pressed to the skin surface. The resulting images were recorded on a still video recorder (MVR 5300 ProMavica, SONY) and analyzed by using a NIH image. After identifying the same hairs in the images taken at 0 h and 48 h following clipping, the length of each hair was measured and the number of hairs in a given unit area was counted from the image taken. Then the following parameters were calculated: (i) hair growth rate (mm per day); (ii) anagen hair ratio (%); (iii) hair density (number per cm2); (iv) hair length at 48 h (mm) – hair length at 0 h (mm)/2 d; and (v) the number of hairs per 1.0 cm2 with a growth rate ≧0.2mm per day per total number of hairs per 1.0 cm2×100. All of the hairs clipped for observation by phototrichogram were collected and dry mounted using double-sided adhesive tape attached to a microscope slide, with distal and proximal ends of the hairs arranged in a uniform direction. Three parameters were calculated using clipped hairs, namely: (i) mean diameter of hair shaft (μm) (the average of three measurements of the diameter of each hair shaft, at a length of 15 mm from the scalp; (ii) the nonvellus hair ratio (percentage of hairs with a diameter >40 μm); and (iii) the short hair ratio (percentage of hairs with a <3 cm length). Multiple regression analysis of multivariate analysis was performed. By refereeing results of the multiple regression analysis, hair growth patterns were then analyzed using Ward's method of cluster analysis, a form of multivariate analysis. In order to exclude aging changes, and due to the small pool of normal control subjects over 55 y old, data from 101 subjects complaining of hair loss (17–54 y old, mean: 34.1 y old) and 58 healthy female control subjects (21–53 y old, mean: 35.7 y old) were used for the statistical analyses. Macroscopic severity of diffuse alopecia was roughly estimated by applying the Ludwig classification (Table I). Subjects complaining of hair loss included 35 with Ludwig grade I and 16 with Ludwig grade II hair loss. None of the subjects showed signs of Ludwig grade III. Fifty subjects complaining of hair loss and all of the controls appeared to have normal hair patterns.Table 1Severity of diffuse alopecia in subjects as judged by the Ludwig classificationGroupTotal no.Mean ageNormal appearanceLudwig ILudwig IILudwig IIISubjects complaining of hair loss10134.7±10.15035160Normal control subjects5835.7±10.158000 Open table in a new tab Individual history of the subjects are summarized in Table II. No cases of seborrhea, acne, hirsutism, diffuse alopecia syndrome (SAHA syndrome), or polycystic ovarian disease were found. Only two postmenopausal subjects were included. None of the 89 (88.1%) subjects complaining of hair loss, nor any of the control subjects, had a history of diseases related to hair growth.Table IINotable factors in the histories of subjects with chronic diffuse hair loss (101 subjects)Acute body weight loss1Post-menopausal2Menstrual disturbance4Oophorectomy3Contraceptive use1HRTaHRT: female hormone replacement therapy.3Hyperthyroidism1Hirsutism0Severe acne0Polycystic ovarian disease0Nothing notable89a HRT: female hormone replacement therapy. Open table in a new tab Laboratory blood tests showed the following results (Table III): five subjects had anemia due to iron deficiency (about 10 g hemoglobin per dl) and nine subjects had low levels of serum iron (less than 60 μg per dl; normal range: 80–170 μg per dl) without anemia. Two subjects who were taking tranquilizers had hyperprolactinemia at the level of 60 ng per ml (normal range: 3.2–26.2 ng per ml), whereas two other subjects showed slightly elevated serum testosterone at 72 and 93 ng per dl, respectively (normal range: 10–60 ng per dl). Seven subjects were positive for the microsome test, but all of these had normal levels of free thyroxine and thyroid-stimulating hormone.Table IIIResult of laboratory tests (101 subjects complaining of hair loss)Low serum iron titer14Anemia (hemoglobin less than 10.0 g per dl)5High prolactin titer2High serum testosterone titer2Positive microsome test7None76 Open table in a new tab Data obtained from the phototrichogram and measurements of hair shaft diameter were analyzed statistically. Using multiple regression analysis of multivariate analysis, the relative importance of each parameter (age, growth rate, anagen hair ratio, hair density, mean hair diameter, nonvellus hair ratio, and short hair ratio) in grading hair patterns was assessed (Table IV). Hair density revealed the highest F-value (20.8), indicating this to be the most important parameter in female diffuse alopecia. Unexpectedly, the anagen hair ratio, nonvellus hair ratio and age showed low F-values, suggesting that these three parameters were of relatively little importance in evaluating female diffuse alopecia. From the results of multiple regression analysis, we selected four parameters, namely, hair density, mean hair diameter, short hair ratio, and growth rate, for further consideration.Table IVResult of multiple regression analysisVariableF-value (p-value)Partial regression coefficientAge2.710 (0.102)0.00549Growth rate3.730 (0.055)1.097Anagen hair ratio0.013 (0.909)-0.000412Hair density20.858 (0.00001)0.00443Mean hair diameter9.840 (0.002)0.0127Nonvellus hair ratio2.100 (0.149)-0.0102Short hair ratio4.154 (0.043)-0.0113 Open table in a new tab Ward's method of cluster analysis was then performed on all 159 subjects (101 subjects complaining of hair loss and 58 healthy controls). All data in each parameter were initially collected into three subgroups in order of the values. Statistical analysis showed significance in grouping the growth rate data into three subgroups and the hair density data, mean hair diameter, and short hair ratio into two subgroups (the abnormal and the normal value groups). Table V shows the threshold values for the groupings in each parameter. Thus, approximately 95% of the control subjects had a hair density of more than 120 strands per cm2, a mean hair diameter of more than 70 μm, and a short hair ratio of less than 16.1%.Table VParting point of each parameter for cluster analysisAbnormal value groupBorderline value groupNormal value groupHair density (n per cm2)45–119120≦Mean hair diameter (μm)20–6970≦Short hair ratio (%)16.1%≦≦16.1%Growth rate (mm per day)≦0.290.3–0.390.4≦ Open table in a new tab Table VI shows the results of cluster analysis using the previously mentioned threshold values. All the subjects, including the 58 healthy controls, were divided into nine hair growth patterns, including six abnormal value groups, two borderline value groups, and one normal value group. The abnormal value groups are: D–1 (low hair density), D-2 (thin hair and high short hair ratio), D-3 (low hair density, thin hair, and high short hair ratio), D-4 (low growth rate, thin hair, and high short hair ratio), D-5 (low growth rate and low hair density), and D-6 (all parameters abnormal). Any subject showing only one abnormality in any parameter besides hair density was placed into one of the two borderline groups, namely B-1 (high short hair ratio) or B-2 (low growth rate). Table VI was then reorganized into Table VII to indicate the correlation between grading of hair growth patterns by phototrichogram analysis and macroscopic observation (Ludwig classification). Of 58 healthy control subjects, 47 (81.0%) fell into the normal value group, 10 subjects (17.2%) fell into the borderline value groups and one subject (1.7%) fell into the low hair density group (D-1). Among the 101 subjects complaining of hair loss, 20 subjects (19.8%) showed normal hair growth patterns, 21 subjects (20.8%) were in the borderline groups, and 60 subjects (59.4%) showed abnormal hair growth patterns. From the 20 subjects with normal hair growth patterns evaluated by phototrichogram, three subjects were diagnosed with Ludwig grade I hair loss by macroscopic observation. Group D-6 was comprised of subjects with the most progressive, diffuse alopecia, and all 10 subjects showed either Ludwig grade I or II symptoms. Generally, grading of hair growth patterns provided the phototrichogram analysis correlated positively with the macroscopic observation method of the Ludwig classification, although the phototrichogram was more accurate. All subjects macroscopically diagnosed with Ludwig type II symptoms (16 subjects) were included in groups D-1 to D-6; however, subjects diagnosed macroscopically with normal or Ludwig type I symptoms varied in the results of their phototrichogram analysis from normal to D-6. No correlation was found between the severity or typing of hair growth patterns by phototrichogram and patient histories or abnormal laboratory data, although it should be noted that two young subjects (25 y and 31 y) with prolactinemia were included in the most progressive subgroup (D-6).Table VIHair growth pattern from cluster analysisGroupSubject (n)Mean age (y)Growth rate (mm per day)Hair density (n per cm2)Mean hair diameter (μm)Short hair ratio (%)Normal67 (47)33.70.40166.284.68.0B-117 (5)40.00.39144.279.522.3B-214 (5)39.70.25164.382.011.1D-116 (1)33.10.44104.383.39.6D-21328.60.37153.261.923.4D-3731.70.43108.657.725.3D-4838.80.25152.066.525.6D-5737.40.2385.786.312.1D-61039.90.2493.661.525.3Bold print indicates abnormal value.Number of subjects from healthy control group in parentheses. Open table in a new tab Table VIIComparison between phototrichogram analysis and Ludwig classification.58 control subjects101 subjects complaining of hair lossNumber of subjects Macroscopic observationNormal appearance 58Normal appearance 50Ludwig I 35Ludwig II 16Phototrichogram analysis. Normal471730B-1: high short hair ratio5930B-2: low growth rate5810D-1: low hair density1672D-2: thin hair and high short hair ratio0670D-3: low hair density, thin hair and high short hair ratio0142D-4: low growth rate, thin hair and high short hair ratio0242D-5: low growth rate and low hair density0124D-6: all parameters abnormal group0046Total58503516 Open table in a new tab Bold print indicates abnormal value.Number of subjects from healthy control group in parentheses. The perception that complaints by female patients of diffuse scalp hair loss is prompted by mere self-consciousness or vanity is belied by the results of macroscopic observation. Yet, unaided visual inspection using the Ludwig classification system is oversimplified and too vague to serve as an objective method of assessing diffuse alopecia (Olsen, 1999Olsen E.A. The midline part. An important physical clue to the clinical diagnosis of androgenetic alopecia in women.J Am Acad Dermatol. 1999; 40: 106-109Abstract Full Text Full Text PDF PubMed Scopus (88) Google Scholar). This study clearly demonstrates that the combination of phototrichogram and measurements of mean hair diameter is a useful, noninvasive, quantitative method by which to evaluate diffuse alopecia. During the process of determining the key parameters of hair growth, multiple regression analysis indicated that hair density, hair diameter, short hair ratio, and hair growth rate, in the order of decreasing importance, were most crucial. Hair density revealed a very high F-value (20.8), indicating this to be a more important parameter than hair diameter and short hair ratio in assessing chronic diffuse alopecia in Japanese female subjects. Unexpectedly, the anagen hair ratio and nonvellus hair ratio, which had been used frequently as hair growth parameters, appeared to be of little importance in evaluating female diffuse alopecia. Emphasis solely on these two parameters would fail to account for one type of alopecia characterized by reduction of hair density only. Grading of hair growth patterns, provided by phototrichogram analysis combined with the measurement of hair diameter was roughly compatible with the results of macroscopic observation based on the Ludwig classification system, but was more accurate in measuring slight changes of hair growth during the early phase of diffuse alopecia. For instance, 50 subjects complaining of diffuse alopecia but with a normal appearance were shown to include 17 subjects with borderline, and 16 subjects with abnormal, hair growth patterns (Table VII). Similarly, 10 borderline and one abnormal subjects were found among the control group evaluated as normal by macroscopic observation (Table VII). In contrast to the previous findings observed among Caucasian subjects (Venning and Dawber, 1988Venning V.A. Dawber M.A. Patterned androgenic alopecia in women.J Am Acad Dermatol. 1988; 18: 1073-1077Abstract Full Text PDF PubMed Scopus (122) Google Scholar), half of the Japanese female subjects complaining of diffuse hair loss were found, in our study, to have a normal appearance of scalp hair, and none of the subjects showed Ludwig grade III alopecia. Although straight, black, Mongoloid hairs are ideal specimens for accurate analysis by phototrichogram, it remains uncertain whether the type and severity of diffuse alopecia varies among the different races. Chronic diffuse alopecia in middle-aged women sometimes occurs secondary to systemic disorders such as thyroid dysfunction (Freinkel and Freinkel, 1972Freinkel R.K. Freinkel N. Hair growth and alopecia in hypothyroidism.Arch Dermatol. 1972; 106: 349-352Crossref PubMed Scopus (82) Google Scholar;Jackson et al., 1972Jackson D. Church R.E. Ebling F.J. Hair diameter in female baldness.Br J Dermatol. 1972; 87: 361-367Crossref PubMed Scopus (49) Google Scholar), iron deficiency (Hard, 1963Hard S. Non-anemic iron deficiency as an etiologic factor in diffuse loss of hair of the scalp in women.Acta Derm Venereol. 1963; 43: 562-569PubMed Google Scholar), prolactinemia (Rushton et al., 1990Rushton D.H. Ramsay I.D. James K.C. Norris M.J. Gilkes J.J.H. Biochemical and trichological characterization of diffuse alopecia in women.Br J Dermatol. 1990; 123: 187-197Crossref PubMed Scopus (114) Google Scholar), and high testosterone levels (Kasick et al., 1983Kasick J.M. Bergfeld W.F. Ateck W.D. Adrenal androgenic female-patterned alopecia: sex hormones and balding women.Cleveland Clin Q. 1983; 50: 111-122Crossref PubMed Scopus (45) Google Scholar;Futterweit et al., 1988Futterweit W. Dunaif A. Yeh H.C. Kingsley P. The prevalence of hyper-androgenism in 109 consecutive female patients with diffuse alopecia.J Am Acad Dermatol. 1988; 19: 831-836Abstract Full Text PDF PubMed Scopus (141) Google Scholar;Rushton et al., 1990Rushton D.H. Ramsay I.D. James K.C. Norris M.J. Gilkes J.J.H. Biochemical and trichological characterization of diffuse alopecia in women.Br J Dermatol. 1990; 123: 187-197Crossref PubMed Scopus (114) Google Scholar). In this study, some of the Japanese subjects complaining of diffuse hair loss had a history of problems related to hair growth (Table II) and showed mild abnormalities in laboratory tests (Table III); however, the involvement of factors in patient history and abnormal laboratory data did not correlate with the severity or typing of diffuse alopecia. The majority of the subjects had no history of problems related to hair growth, suggesting that chronic diffuse alopecia observed in Japanese female subjects is idiopathic in nature. In light of the previous definition of FAGA and CTE, most subjects complaining of diffuse hair loss in this study can be seen as exhibiting features of FAGA; however, the mild form and early phase of chronic diffuse alopecia cannot easily be subgrouped. This is partly because the biologic and endocrinologic mechanisms inducing FAGA-type alopecia and CTE-type alopecia have not yet been clarified. Our results in Table VII suggest three pathways in the progression of chronic diffuse alopecia culminating in D-6. FAGA, in which the primary change is an increase of vellus hairs, may progress from D-2 and D-3. CTE, which involves the thinning of whole scalp hairs, may progress from D-2 and D-4. The third pattern, which shows a decrease in hair density without vellus change, may progress from D-1 and D-5. Although the third pattern of alopecia has not been well described in previous literature, it appears to be a distinct form of chronic diffuse alopecia. It is not certain, at this moment, whether this type of alopecia is specific to Japanese female subjects or is a more universal phenomenon. The procedures used in this study were performed on each of the subjects at a single site in the parietal area of the scalp and monitored only for 48 h. In order to evaluate the hair pattern of each individual more precisely and to arrive at a clearer system of classification, repeated analyses of multiple areas on the scalp are required.