How hairy are hirsute women?

Transcription

1 Clinical Endocrinology (1997) 47, Review How hairy are hirsute women? Julian H. Barth Institute of Pathology, Leeds General Infirmary, Leeds, UK (Received 26 November 1996; returned for revision 7 February 1997; finally revised 24 March 1997; accepted 7 May 1997) Why is there so much variation in studies of hirsuties? There are clearly a number of reasons. First, hirsuties is an end organ response to androgenic stimulation and there is enormous individual variation in both the local response of individual hair follicles as well as the overall pattern of hair. Secondly, there are a number of causes of increased androgen production but they rejoice under a number of different names and definitions. A further factor that does not appear to have been considered is the definition of hirsutism. Is a woman hirsute because she thinks she is or because her physician thinks she is, and on which definition are hirsuties studies based? This paper will address this issue and also the implicit question of how hirsuties should be measured. What is hirsuties? Hirsuties is the growth of terminal hair on the body of a woman in the same pattern as that which develops in the normal postpubertal male. This pattern of hair growth should be differentiated from hypertrichosis, which is characterized by a diffuse fine growth over the body. Hypertrichosis may be congenital or acquired due to drug therapy and some metabolic disorders. Since the latter is not hormonally mediated, it does not respond to anti-androgen therapy. When is a woman hirsute? Who decides which women are hirsute? The majority of hirsute women will seek advice from a beautician prior to seeking medical advice. However, since there is a wide spectrum in the tolerance of facial and body hair due to personal, social and cultural factors, only a proportion of hirsute women will present for medical advice. It is important to consider the issue of whether the definition is made by all hirsute women, those hirsute women who attend for medical advice or their physicians, because we have found that there is little, if any, Correspondence Dr J.H. Barth, Department of Chemical Pathology and Immunology, Institute of Pathology, Leeds General Infirmary, Great George Street, Leeds LS1 3EX, UK. j.h.barth@leeds.ac.uk 1997 Blackwell Science Ltd correlation in body hair scores made by physician and patient after effective therapy (Barth et al., 1989). The prevalence of hirsuties, as defined by physicians examining hospital patients, has been reported at %. Ferriman and Gallwey (1961), who developed the most widely used scoring system for body hair, examined women presenting to hospital for non-endocrine reasons and proposed that 1. 2% of females in London were hirsute using a score > 1 (Ferriman & Gallwey units). Lunde and Grottum (1984) and Lunde (1984) in Norway, using a more sophisticated scoring system than Ferriman and Gallwey, defined hirsuties by the hair scores obtained from women attending a hirsuties clinic. They then evaluated a control group of 1 women with regular menstrual cycles attending hospital for sterilization, termination of pregnancy or blood donation and noted that there was a marked overlap of hair scores between the control women and the hirsute women (see Fig. 1) but the controls were not asked whether they considered they had excessive hair. A similar overlap in hair scores was noted by Shah (1957) in India. Both these investigators had identified a qualitative difference between groups; despite having similar scores, women complaining of hirsuties are more likely to have significant hair growth on the upper lip, chin, chest and upper back. Only one study of a random population appears to have addressed the issue of how many women themselves feel they have too much hair. McKnight (1964) examined 4 European students, 6% of whom were Welsh. Nine percent were aware that they were particularly hairy although only 4% had themselves requested advice and treatment for their hirsutism. This difference highlights the difficulty in using physician vs patient definitions of excess hair. An alternative approach to the definition of hirsutism might be the consumer demand for treatment by electrologists. For example in Leeds, UK, there are 36 electrologists for a population of 2 women. This would represent about 1 electrologist for every 28 women in their reproductive years. A limited survey of local electrologists reveals an average workload of approximately 85 regular clients, of whom about half are over 5 years (unpublished data). This could be extrapolated to 3 hirsute women treated by the 36 electrologists, or 1. 5% of the total female population. Interestingly, this figure is similar to Ferriman and Gallwey s (1961) estimate. In contrast, the Institute of Electrology state in their information pamphlet that 8% of all women would benefit from removal of superfluous hair. Clearly, not all hirsute women present for medical help. Shah (1957) interviewed 34 women in Bombay, India regarding their 255

2 256 J.H. Barth 12 1 Number of women Hair score Fig. 1 Hair scores in 113 women referred to a gynaecologist with hirsuties (g) compared with 1 control women (A). The latter had regular menses, mean age 29 yr (range yrs) and were seen prior to termination of pregnancy or sterilisation. Hair scores were as described by Lunde (1984). Data from Lunde & Grottum (1984) The data appear to demonstrate two apparently distinct normal distributions. motivation for visiting a physician. Their responses in order of frequency were (1) they or their relatives felt the prospects for marriage were jeopardized by their excess hair; (2) they thought they might be changing gender; (3) their husbands had begun to dislike them since they had become hirsute; and (4) they were concerned of the risk of their children becoming hirsute. In our experience, the effect of male partners has been a minimal factor but the issue of body image has been more important. Effect of age on hair growth The epidemiology of hair growth in women has been addressed by a number of investigators (Pedersen, 1943; Beek, 195; Garn, 1951, Thomas & Ferriman, 1957), all of whom have demonstrated the gradual increase in facial hair and reduction in body hair after the menopause as illustrated in Fig. 2. Although many young women present for medical advice with hirsuties, it would appear that most post-menopausal women rely on beauticians and electrologists (see above). Body hair on the breasts, lower abdomen and limbs is seen in women after the onset of puberty (> 15 years) with a peak prevalence in the third decade (2 29 years) and a gradual reduction thereafter (Pedersen, 1943; Beek, 195). This pattern might give a physiological rather than social explanation for a greater rate of presentation of young women with hirsutism. It is unknown whether this difference in hair might influence the effects of therapeutic intervention. Racial factors The pattern of hair growth in populations not complaining of hirsuties has been studied in European and Indian subjects (see Table 1). The hair is similar in the Europeans. The study of Indian women, however, shows a lack of facial and abdominal hair despite using a similar definition (terminal hair shafts greater than. 5 cm). A comparison of women from Italy, Japan and the United States with polycystic ovaries reported that the Italian and American women were equally hirsute but that the Japanese women had a lighter covering of body hair (Carmina et al., 1992). There has been no formal study of hair patterns in hirsute women from different racial groups. The data in Table 1 describing women not complaining of excessive hair suggests that the pattern is largely similar, although Indian women appear to have less facial hair. Social, cultural and commercial factors Since a degree of body hair is present on essentially all women, one of the most important influences on individual women s 1997 Blackwell Science Ltd, Clinical Endocrinology, 47,

3 How hairy are hirsute women? Fig. 2 Differences in facial and abdominal hair in women throughout the age range years. W Chin; A Lip; K abdomen. Data were collected on 584 women admitted to hospital. Hair was graded 4 at each site, scores 3 4 were considered to represent significant hair growth (Thomas & Ferriman (1957)). This data is similar to that reported by Beek (195) and Pedersen (1943). Women with hair at specific sites (%) Age (years) body image will be the effects of images in cosmetic advertisements. These are designed to make every woman feel inadequate in order to promote sales; and in the context of hirsuties, any body or facial hair is portrayed as excessive and is described as unsightly or disfiguring. A typical advertisement for a depilatory cream for bikini lines use uses the caption Go as bare as you dare ; and in the medical press an advertisement for an HRT preparation uses an image of a young woman shaving with the caption Why treat her like a man?. Local cultural factors are also important as seen in the variation in social norms for women to shave their axillae and legs. This is more frequently practised by women in the United Kingdom and United States than in Germany. Another interesting cultural influence can be seen by the outpouring of hirsuties research from northern Italy. This might be interpreted as a response to the social conflicts between the relatively bodily bald northern European and the hirsute Mediterranean. In our practice, many Asian women claim that they only Table 1 Qualitative (but not quantitative) presence (%) of terminal hair at different body sites in women not complaining of hirsuties. Hair growth is defined as the presence of terminal hairs greater than. 5 mm in length Country and age range surveyed England (a) Holland (b) India (c) Norway (d) Wales (e) USA (f) years years years years 18 years years Site (n ¼ 257) (n ¼ 81) (n ¼ 1) (n ¼ 1) (n ¼ 4) (n ¼ 35) Lip Chin Upper arm Lower arm g Chest o Upper abdomen 6 4 Lower abdomen Upper back <1 6 3 Lower back Thigh Leg (a) Ferriman & Gallwey (1961); (b) Derksen et al., (1993); (c) Shah (1957); (d) Lunde & Grottum (1984); (e) McKnight (1964); (f) Danforth & Trotter (1922) Blackwell Science Ltd, Clinical Endocrinology, 47,

4 258 J.H. Barth became concerned with their hair growth after migrating to the United Kingdom, although Hochberg et al. (1996) state that middle eastern Arab women regard hairlessness as an ideal. Why is hair measured? If one considers that a woman is hirsute if she considers herself hirsute, then the purpose of scoring facial and body hair should be to determine whether or not such women should be investigated for an underlying endocrine disorder, whether oral or topical therapy would be most appropriate and, probably most importantly, how the efficacy of these agents should be monitored. Whatever the reason, it is clearly essential to ensure that the methods are standardized and reproducible. How is body hair measured? Body hair can be assessed by subjective clinical scoring systems or by direct measurement of individual hair shafts. Subjective scoring schemes currently used for evaluating hirsuties were originally devised by anthropologists. The methods were produced to study overall hair patterns and distribution of body hair. They subdivide the body into zones, each of which is individually scored, and the total aggregated. The score for each zone may be based on the presence alone of terminal hairs, or extra sophistication may be introduced by evaluation of the extent of hair growth or hair density within each site. Although the scoring schemes are used for the evaluation of hirsuties and its treatment, none has been subjected to any attempt to standardization or evaluation of sensitivity. The various grading systems are described in Table 2. The scoring system developed by Ferriman and Gallwey (1961) is by far the most widely used by hirsuties investigators as it is the least complex and easiest to use. Despite this ease of use, there appears to be considerable observer variation in hirsuties scores by various investigators (see Fig. 3). The lack of observer conformity is compounded by the lack of method standardization and estimates of reproducibility in publications by investigators of hirsute women. Since the subjective Ferriman and Gallwey score is relatively insensitive, even in the hands of meticulous investigators, there must be doubt in the results of studies which do not report details of methodology (Barth, 1996). Direct measurements of human hair growth have been thoroughly evaluated for the scalp (Rushton et al., 1993) largely fuelled by the interest in Minoxidil; and the physiology of hair growth on all body sites has been studied methodically by Hamilton (1958) and by Pecoraro and Astore (199). There has, however, been no formal evaluation of the most appropriate methods for objective measurement of hair in hirsute women. How is hair currently measured in hirsute women? The most frequently reported measurement methods are the subjective systems as described above. A relatively small number of studies have made direct measurements of small numbers of hair shafts including diameter, daily growth rate, density, presence or absence of a central shaft medulla and weight of hair clippings. The most important aspect of most hirsutes studies is the apparent lack of use of accepted clinical trial methodologies, e.g. double-blind or placebo-controlled study design, and an almost complete lack of method evaluation (Barth, 1996). Moreover, some investigators have declared a complete lack of Table 2 Summary of subjective methods for the evaluation of hair growth on the face, trunk and limbs Number of sites Investigators evaluated Specific feature evaluated Pedersen (1943) 12 Evaluation only of presence of hair at each site but not quantitative assessment Beek (195) 19 Evaluation only of presence of hair at each site but no quantitative assessment Garn (1951) 11 Evaluation of patterns of hair growth at each site and form of hair shaft present (e.g. curly or straight). No quantitative assessment Shah (1957) 9 Quantitative assessment of terminal hairs >. 5 cm. (Q)uality (scored 1 3), (D)ensity ( 3) & (P)roportion of zone covered with hair ( 1); total score ¼ Q D P Ferriman & Gallwey (1961) 11 Quantitative score based upon distribution of hair on each site Lorenzo (197) 7 sub-zones Quantitative score based upon density and extent of involvement Hatch et al. (1981) 9 Qunatitative score based on distribution of hair on each site Lunde (1984) 18 Quantitative assessment of length & density of terminal hairs >. 5cm Adaptation of Ferriman & Gallwey (1961) and Lorenzo (197) Blackwell Science Ltd, Clinical Endocrinology, 47,

5 How hairy are hirsute women? Fig. 3 Comparison of the hirsutism scores in 11 studies all of which used the Ferriman & Gallwey (1961) scoring system (the mean and standard deviation of each study is drawn). This indicates that mutli-centre comparisons of therapies for hirsuties cannot rely on subjective grading and must be supplemented with critera which can be objectively compared. If it is assumed that the population of hirsute women presenting to any physician is similar, the wide scatter of values obtained in these studies would indicate marked inter-observer variation. An alternative interpretation might be that there was selection bias in choosing hirsute women but this, too, would mitigate against either multi-centre studies or comparisons between studies. Ferriman & Gallwey score knowledge of hair physiology by measuring hairs from separate sites, e.g. face and abdomen, and subsequently combining the data for statistical analysis. How should hair be measured? Women seeking treatment for hirsuties want a global reduction in facial and body hair and for clinical purposes the overall subjective scoring system is the most practical method, provided that this can be performed reliably. In practise this means a single observer who performs regular blinded repeat measurements, although paired investigators can be trained to give similar scores (Derksen et al., 1993). This is important as patients examine themselves regularly and their recollections compare poorly with a physician s grading (Barth et al., 1989). Therefore, treatment may be continued or terminated on uncertain grounds. A further factor supporting subjective scoring is that there appears to be a close correlation between these scores and the measurement of hair shaft diameter on the forearm, lower abdomen and anterior thigh (Barth, 1997). There is a degree of imprecision inherent with the subjective scoring methods, but in careful hands this is likely to be less than any acceptable degree of improvement. For more precision, it may be necessary to supplement with direct measurements of hair shafts. Previous studies have measured diameter and linear growth rates but these may not be the most appropriate variables as there appear to be conflicting data on the value of both variables. Careful studies of hair on the upper arm and thigh by Seago and Ebling (1985) have shown that, although men have thicker hair shafts, the critical difference in hair growth between men and women is the duration of the anagen (growth) phase of the hair cycle which affects the overall length of the hair shaft; whereas density and daily growth rate are not significantly different and probably do not merit measurement. The length of hair shafts as a measure of the length of anagen can be measured by plucking or trimming hair shafts at the skin surface, placing them in a straight line on microscope slides using double-sided adhesive tape and measuring them with a ruler. This method does rely on patients not using any cosmetic procedures at that site for the previous 3 months (duration of the hair cycle) and restricts its use to converted parts of the body. It is a time-consuming process and clearly not suitable for routine clinical practice. Whichever measurement of individual hairs is made, there is a need for the sampling and measurement of the hairs to be made by blinded observers; there is also an imperative to ensure that hairs are sampled from the same site each time. Sufficient hairs need to be taken to ensure a constant measurement and that these measurements are repeatable. Madanes and Novotny (1987) reported a novel method. Hairs are shaved from a site and the shafts are examined microscopically and scored as the ratio of vellus (fine, nonpigmented shafts without a central medulla) to terminal hairs (pigmented shafts with central medulla). This may be suitable for monitoring therapy in hirsuties, since it is easy to perform and is unlikely to be significantly affected by bias in sampling as the absolute number of shafts is not important. Further work on this method is required. Is measurement of hair necessary for routine clinical practice? If hirsuties is considered to be a subjective disorder defined by women, it could be argued that no formal measurement of the 1997 Blackwell Science Ltd, Clinical Endocrinology, 47,

6 26 J.H. Barth degree of hair is necessary for the purposes of clinical management, particularly since the requirement to investigate hirsute women is based on the relative speed of onset of hirsuties rather than its severity (Crosignani & Rubin, 1989). The choice of the most appropriate form of therapy is probably made on the site of the excess hair and the physician s experience with that therapy. Furthermore, since the decision to treat the hirsute women has been based on her concern of her body image, it would be logical to continue treatment until she feels that it is no longer worthwhile. However, most women notice a return in their hirsuties a few months after stopping systemic anti-androgen therapy, even if they had previously believed that therapy had been ineffective. Since anti-androgen drugs have potentially toxic side-effects, any decision about recommencing therapy will therefore need to be made on some objective grounds, such as any change in hair growth scores during previous therapy. Conclusion Hair on the face and body in women is a highly subjective phenomenon and it is unlikely that there will ever be a consensus over its definition. It is possible that variation in the findings of both endocrine investigations and therapy for hirsute women is due to a lack of standardization and consequent selection bias of subjects. Secondly, there is a need for hirsuties investigators to use clinical trial methodology for therapeutic trials and to study clinical measurements with the same rigour as laboratory methods. Finally, there is still a need to formulate the optimal method for measurement of body hair growth. Acknowledgements I am grateful to Drs P. J. Hammond and D. Williamson for their helpful comments. References Barth, J.H. (1996) How robust is the methodology for trails of therapy in hirsute women? Clinical Endocrinology, 45, Barth, J.H. (1997) Subjective measurements of body hair in hirsute women compare well with direct diameter measurements of hair shafts. Acta Dermato-Venereologica (Stockholm), 77, Barth, J.H., Cherry, C.A., Wojnarowska, F. & Dawber, R.P.R. (1989) Spironolactone is an effective and well tolerated systemic antiandrogen therapy for hirsute women. Journal of Clinical Endocrinology and Metabolism, 68, Beek, C.H. (195) A study on the extension and distribution of the human body hair. Dermatologica, 11, Carmina, E., Koyama, T., Chang, L., Stancyk, F.Z. & Lobo, R.A. (1992) Does ethnicity influence the prevalence of adrenal hyperandrogenism and insulin resistance in polycystic ovary syndrome? American Journal of Obstetrics and Gynecology, 167, Crosignani, P.G. & Rubin, B. (1989) Strategies for the treatment of hirsutism. Hormone Research, 4, Derksen, J., Moolenaar, A.J., van Seters, A.P. & Kock, D.F.M. (1993) Semiquantitative assessment of hirsutism in Dutch women. British Journal of Dermatology, 128, Ferriman, D. & Gallwey, J.D. (1961) Clinical assessment of body hair growth in women. Journal of Clinical Endocrinology, 21, Garn, S.M. (1951) Types and distribution of the hair in man. Annals of the New York Academy of Science, 53, Hamilton, J.B. (1958) Age, sex and genetic factors in the regulation of hair growth in man: a comparison of Caucasian and Japanese populations. In The Biology of Hair Growth (eds W. Montagna & R.A. Ellis), pp Academic Press, New York. Hatch, R., Rosenfield, R.L., Kim, M.H. & Tredway, D. (1981) Hirsutism: implications, etiology and management. American Journal of Obstetrics and Gynecology, 14, Hochberg, Z., Chayen, R., Reiss, N., Falik, Z., Makler, A., Munichor, M., Farkas, A., Goldfarb, H., Ohana, N. & Hiort, O. (1996) Clinical, biochemical, and genetic findings in a large pedigree of male and female patients with 5a-reductase deficiency. Journal of Clinical Endocrinology and Metabolism, 81, Lorenzo, E.M. (197) Familial study of hirsutism. Journal of Clinical Endocrinology and Metabolism, 31, Lunde, O. (1984) A study of body hair density and distribution in normal women. American Journal of Physical Anthropology, 64, Lunde, O. & Grottum, P. (1984) Body hair growth in women: normal or hirsute. American Journal of Physical Anthropology, 64, McKnight, E. (1964) The prevalence of hirsutism in young women. Lancet, i, Madanes, A.E. & Novotny, M. (1987) The vellus index: a new method for assessing hair growth. Fertility and Sterility, 48, Pecoraro, V. & Astore, I.P.L. (199) Measurement of hair growth under physiological conditions. In Hair and Hair Diseases (eds C.E. Orfanos & R. Happle), pp Springer-Verlag, Berlin. Pedersen, J. (1943) Hypertrichosis in women. Acta Dermato-Venereologica (Stockholm), 12, Rushton, D.H., de Brouwer, B., de Coster, W. & van Neste, D.J.J. (1993) Compariative evaluation of scalp hair by phototrichogram and unit area trichogram analysis within the same subject. Acta Dermato- Venereologica (Stockholm), 73, Seago, S.V. & Ebling, F.J.G (1985) The hair cycle on the human thigh and upper arm. British Journal of Dermatology, 113, Shah, P.N. (1957) Human body hair a quantitative study. American Journal of Obstetrics and Gynecology, 73, Thomas, P.K. & Ferriman, D.G. (1957) Variation in facial and pubic hair growth in white women. American Journal of Physical Anthropology, 15, Blackwell Science Ltd, Clinical Endocrinology, 47,