Susan R. Davis, MD, PhD,[1,2] Rebecca Goldstat, MPH,[1,2] Mary-Anne Papalia, MD,[1,2] Sonal Shah, BSci (Hon),[1,3] Jayashri Kulkarni, MD,[3] Susan Donath, MA,[4] and Robin J. Bell, MD, PhD[1,2]

From the [1]Women,s Health Program, Department of Medicine, Central and Eastern Clinical School, Monash University, Alfred Hospital, Prahran, Victoria, Australia; [2]The Jean Hailes Foundation Research Unit, Clayton, Victoria, Australia; the [3]School of Psychiatry, Monash University, Prahran, Victoria, Australia; and [4]Murdoch Children’s Research Institute and University of Melbourne, Department of Paediatrics, Melbourne, Australia. This research was funded by the NHMRC of Australia, Grant number 219279.

R. Davis, MD, PhD, Women’s Health Program, Department of Medicine, Central and Eastern Clinical School, Monash Medical School, Alfred Hospital, Commercial Road, Prahran, VIC 3181, Australia.; E-mail: susan.davis@med.monash.edu.au.

Abstract

Objective: The extent to which aromatization of testosterone (T) to estradiol is required for the observed effects of testosterone therapy on sexual function and well-being are not known. Therefore, the authors investigated the effects of aromatase enzyme inhibition on sexual function, well-being, and mood in estrogen- and T-replete postmenopausal women in a double-blind, randomized, placebo-controlled study.

Design: Postmenopausal women using transdermal estrogen therapy for at least 8 weeks and reporting low sexual satisfaction (score < 42 for the Sabbatsberg Sexual Self-rating Scale [SSS]) with a total T value of less than 1.2 nmol/L were treated with 400 μL of a 0.5% T gel (total dose 2 mg) and were randomly assigned to receive treatment with either 2.5 mg/day of letrozole or an identical placebo tablet. Women were assessed at baseline (week —2) and at 0, 4, 8, and 16 weeks. Sexual function was assessed with the SSS, well-being was assessed with the Psychological General Well-being Index, and mood was assessed with the Beck Depression Inventory at 0 and 16 weeks. Eighty-one women were screened, 76 were randomly assigned to a treatment group, and 30 in each group completed the study. Because this was a mechanistic study, only the 60 women who completed the study per protocol were included in the final analysis.

Results: Total T and calculated free T increased from baseline in both groups, with no difference between groups. At 16 weeks, estradiol, sex hormone-binding globulin, fasting lipids, lipoprotein(a), and C-reactive protein did not differ from baseline or between groups. Significant increases in total Sabbatsberg Sexual Self-rating Scale scores, total Psychological General Wellbeing Index scores, and a reduction in Beck Depression Inventory scores from baseline to 16 weeks was seen for both treatment groups, with no effect of treatment allocation. No adverse treatment effects were reported.

Conclusions: Increases in total and free T in the physiologic range in postmenopausal women were associated with improved sexual satisfaction, well-being, and mood. In this study, aromatase inhibition did not influence any of these outcomes. Short-term transdermal T therapy did not modify fasting lipids, lipoprotein(a), or C-reactive protein.

Key Words: Testosterone - Postmenopausal androgen therapy - Androgen action.

Introduction

Testosterone seems to play a major role in stimulating sexual motivational behaviors, maintaining optimal levels of sexual desire, and possibly contributing to sexual gratification.[1-3] Improvements in sexual function, mood, and well-being have been demonstrated in both premenopausal and postmenopausal women treated with testosterone.[4-12] Acutely, compared with placebo, testosterone therapy also results in increased vaginal vasocongestion, with increased subjective arousal in young eugonadal women.[13]

Transdermal testosterone replacement in surgically menopausal women significantly improves the Psy-chological General Well-being (PGWB) Index[14] score over placebo, with the greatest change being in improved general well-being and less depressed mood.[5,15] We have reported similar improvements in the total and all subscales of the PGWB Index with testosterone administration in premenopausal women.[4]

Whether the above effects are caused by testosterone acting directly via the androgen receptor or as a precursor for extragonadal estrogen biosynthesis and then via the estrogen receptor (ER) is not known. Testosterone may also act by decreasing sex hormone binding globulin (SHBG) and increasing free circulating levels of multiple sex steroids. There are both androgen receptors and ERs in the healthy female brain,[16] and high concentrations of estradiol (E₂) and testosterone have been demonstrated in the human female hypothalamus and preoptic area, the concentrations of testosterone being 10-fold greater than those of E₂.[17] These regions correlate with regions of high aromatase activity in animals.18 ERα knockout (KO) mice and aromatase KO mice, despite high circulating androgen levels, exhibit almost no sexual interest,[19,20] whereas ERβKO mice exhibit normal reproductive behavior.[19] In birds, visual exposure to sexual intercourse evokes increased aromatization of testosterone to E₂ in the preoptic area.[21] Therefore, findings from animal models imply that testosterone influences sexual behavior in the brain via conversion to E₂, which then acts through ERα.

To explore whether aromatization is required to achieve the reported sexual and mood effects of testos-terone in postmenopausal women, we evaluated the effects of aromatase enzyme inhibition in estrogen- replete postmenopausal women, free of vasomotor symptoms but presenting with low libido, treated with concurrent transdermal testosterone in a double-blind, placebo-controlled, randomized study. To maximize the opportunity to explore the mechanism of testosterone action, we monitored circulating levels during the study and adjusted doses to minimize the likelihood that testosterone levels did not exceed those considered physiologic for young women. Therefore, a placebo-placebo group was not included because this would have resulted in unblinding. Because all women were treated with testosterone, this study also provided the opportunity to examine the short-term effects of trans-dermal testosterone on established markers of cardio-vascular disease risk.

Methods

Study population

The study involved postmenopausal women (aged older than 55 years or bilateral oophorectomy or >12 months amenorrhea and FSH >20 IU/L, or treated for climacteric symptoms with postmenopausal estrogen therapy for >12 months) with the following character-istics: a body mass index between 18 and 34 kg/m², diminished sexuality [Sabbatsberg Sexual Self-rating Scale (SSS)[22] score less than 42 out of a possible 84[23]], no evidence of severe clinical depression on the Beck Depression Inventory (BDI) (score < 28),24 and an early morning serum total testosterone of less than 1.2 nmol/ L. Participants were required to have been stabilized for at least 8 weeks on standard transdermal estrogen therapy equivalent at least to 50 mg of 17β-E₂ by matrix patch or 1 mg/day by transdermal gel to eliminate any effect of estrogen therapy on SHBG, with appropriate cyclical or continuous-combined progestin for those with an intact uterus and to be free of vasomotor symptoms and vaginal dryness.

All women were in general good health on history and physical examination and had a normal cervical smear within the past year. Women who had relationship problems, poor feelings for their partner, or dyspareunia were excluded, as were women receiving pharmacotherapy for depression, those taking medica-tion known to interfere with normal sexual function (such as β-blockers and α-blockers), and those who had received tibolone in the past 4 weeks, transdermal testos-terone in the last 2 weeks, oral androgen therapy in the previous 3 months, or testosterone injections or implants in the previous 12 months. Women with a history of acne or hirsutism, estrogen-related cancer, confirmed thromboembolic disease, a previous cerebrovascular accident, uncontrolled hypertension (blood pressure >160/95 mm Hg), unstable cardiovascular disease (a myocardial infarction or coronary or peripheral angio-plasty within preceding 3 months), genital bleeding of unknown cause, alcohol intake greater than 30 g/day, insulin-dependent diabetes mellitus or unstable non-insulin-dependent diabetes mellitus, homozygous familial hypercholesterolemia, or abnormal liver func-tion test results or who were taking medications known to interfere with sex steroid metabolism, a neurologic disorder that would affect cognition, including head injury, epilepsy, history of seizures, or intellectual disability, were also excluded. The use of thyroid hormone was acceptable if the dose was expected to remain stable throughout the study.

Participants were recruited from the Jean Hailes Medical Centre, Melbourne, Australia, and from the general population via television and radio announ-cements over a 12-month period. The protocol was approved by the Human Research and Ethics Commit-tee, Southern Health (Clayton, Victoria, Australia), and all participants gave written informed consent.

Study design

The study was a randomized, placebo-controlled trial with a 16-week treatment period. All participants continued their transdermal E₂ therapy and in addition received testosterone in the form of 400 μL of a 0.5% testosterone gel (total dose 2 mg) administered to a skin surface area of approximately 60 cm² (the size of the average female palm) dispensed from a canister that releases 200 μL per depression, packaged and supplied by Cellergy Pty, Ltd. Women were randomly assigned to receive either the aromatase inhibitor, letrozole, 2.5 mg orally once daily (dose selected on the ability to inhibit in vivo aromatization by 98%25), or an identical placebo tablet, one tablet daily. Randomization sequences were generated centrally at Novartis (Basel, Switzerland), and Novartis also supplied and packaged the active and placebo tablets. Women were seen at baseline (week —2) and at 0, 4, 8, and 16 weeks to review therapy and document any adverse events. Women were also reviewed at 4 weeks to have blood drawn for measurement of free testosterone. If the value of the free testosterone, measured by the direct radioimmunoassay, was found to exceed 12.5 pmol/L (above the upper limit of normal provided by the manufacturers of the assay kit for a young woman), the dose was reduced to 200 mL/day and blood was drawn again 2 weeks later to determine the new level achieved. This was to prevent excessive testosterone exposure, both to protect participants and to limit excessive dosing overriding physiologic systems.

Serum hormone measurements

Serum total testosterone, SHBG, E₂, total cholesterol, triglycerides, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein (LDL) cholesterol, lipoprotein(a) [Lp(a)], and highly sensitive C-reactive protein (hsCRP) were measured at 0 and 16 weeks. Thyroid-stimulating hormone, prolactin, and follicle- stimulating hormone were measured at screening. Hormone analyses were performed by Mayne Health Dorevitch Pathology (Melbourne, Australia).

Total testosterone was measured by a validated direct radioimmunoassay method (Biosource Europe S.A., Belgium)[26] using antibody-coated tubes and iodine- labeled testosterone tracer. Mean between-batch coef-ficients of variation (CVs) were 12.8%, 9.7%, 8.8%, and 7.1% at 0.17, 0.61, 1.77, and 11.5 nmol/L (n = 100), respectively; mean within-batch CVs were 10.5%, 5.3%, 4.2%, and 4.7% at the same concentrations (n = 20). SHBG was measured by a solid-phase, two-site chemilu-minescent enzyme immunometric assay using the Immulite 2000 automated analyzer (Diagnostic Products Corporation, Los Angeles, CA). Intra-assay and inter-assay CVs for SHBG are 6.5% and 8.7%, respectively. During the conduct of the study, free testosterone was measured using a direct assay (Diagnostic Products Corporation) for the purpose of adjusting the testoster-one dose. The interassay CVs for this assay were 23.2%, 15.2%, 8.8%, and 9.3% at concentrations of 3.4, 9.7, 33.0, and 77.0 pmol/L, respectively. Calculated free testosterone (using the Sodergard equation[27]) was determined for reporting purposes but was not available to the investigators during the study.

Total cholesterol was determined by the CHOD-PAP method, and triglycerides were determined by the GPO- PAP method using a Boehringer Mannheim Systems Hitachi 747 machine. HDL cholesterol was also measured by an enzymatic colorimetric test on a Boehringer Mannheim/Hitachi 747 machine. Lp(a) was determined by a rate nephelometry method with a kit from Beckman Image Immunochemistry Systems (Beckman Instruments, Inc.). LDL cholesterol was calculated according to Friedewald et al.[28]

Measurement of hsCRP was by a particle-enhanced immunoturbidimetric assay performed on the Hitachi 917 analyzer. The assay range is 0.1 to 20 mg/L, intra-assay CVs are 1.34% at 0.55 mg/L and 0.28 at 12.36 mg/L, interassay CVs are 5.7% at 0.52 mg/L and 2.51% at 10.98 mg/L, and the detection limit is 0.03 mg/L.

Evaluation of well-being, mood, and sexual function

Overall general well-being was assessed with the PGWB Index, a validated 22-item multiple-choice questionnaire.[14] It includes subscales for anxiety, depressed mood, positive well-being, self-confidence, general health, and vitality. The composite score ranges from 0 (most negative affective experience) to 110 (most positive affective experience).

Depressed mood was assessed by the BDI, a widely used, validated questionnaire that contains 13 variables to assess mood and depression.[24] Possible scores ranged from 0 (minimal depression) to 63 (severe depression).

The SSS[22] is a multiple-choice questionnaire con-taining seven aspects of sexuality (sexual interest, sexual activity, satisfaction of sexual life, experience of sexual pleasure, sexual fantasy, orgasm capacity, and sexual relevancy). We have considerable experience with using this questionnaire.[4,11] Possible composite scores range from 0 (low sexuality) to 84 (high sexuality). Each of the three questionnaires was self- administered at screening, at baseline, and at each study visit.

Evaluation of safety

At each visit, hirsutism was evaluated using the Ferriman and Gallwey score[29] by a physician blinded to treatment allocation. Adverse events were recorded. Measurement of lipids, Lp(a), and hsCRP were performed as described.

Statistics

Sample Size

Calculation of the sample size was based on the assumptions that the baseline total score for the PGWB Index, the changes expected with treatment, and the within-group SD for the PGWB Index in postmenopausal women would be similar to those seen in the premenopausal women recently studied.[4] With these assumptions, a sample size of 30 women per group would be required to have an 80% power of detecting an effect size of at least 0.75 (alfa = 0.05). Anticipating a dropout rate of 33%, we decided to recruit 40 women per group (total of 80 women) to the study.

Statistical analyses

Because this was designed as a mechanistic study, only women who provided evaluable data at baseline and 16 weeks were included in the analysis.

All primary and secondary outcomes are continuous variables, although they are not all normally distributed. Where the distribution of the results approximates that of a normal distribution, results are presented as mean (SD), and differences between means are compared using t tests. Where the results do not approximate a normal distribution, the results are presented as medians, and the results are compared using nonparametric tests. Unpaired tests were used to assess the characteristics of the group of women excluded from the study compared with those who remained in the analysis, baseline differences between the intervention and control groups, differences between groups at 16 weeks, and change in results between week 0 and week 16 in the two treatment groups. Paired t tests were used to assess change from baseline to week 16 in the two treatment groups combined. The contribution of baseline value to the value of week 16 for the outcomes of SSS, BDI, and PGWB Index was assessed using linear regression, with the value at week 16 as the dependent variable and the value at week 0 and group allocation (letrozole or placebo) being the dependent variables. A P value of 0.05 was considered statistically significant, and all tests were two-tailed.

When biochemical results were reported as being below the level of detection, these results were ascribed the value of the level of detection. Where a considerable number of results were affected, such as for serum E₂, where 12 of the 60 results at baseline were reported as less than 100 pmol/L, results were compared using ranks rather than means.

Results

Eighty-one women were screened, and 76 were randomly assigned to groups. Reasons for exclusion included total testosterone greater than 1.2 nmol/L (n = 2), increased BDI score (n = 1), detection of a breast lump (n = 1), and failure to attend follow-up (n = 1). Sixty women completed the study according to protocol, with equal numbers in each group. Reasons for discontinuation are listed in Table 1. No significant differences were found for women who did and did not complete the study in relation to any of the parameters measured at screening and baseline. Characteristics of those who completed the study are listed in Table 2. There were no differences in biochemical measures at baseline between the two groups (Table 3). Per protocol reduction in testosterone dose based on increased free testosterone, measured by the direct assay, was made at week 4 in two women in the letrozole group.

Table 1. Reasons for study discontinuation.

Effects on circulating hormone levels and biochemical parameters

Table 2. Screening characteristics of women who completed the study.

The results ofbiochemical measures are summarized in Table 3. There were no statistically significant differences between treatment groups at week 16 for any biochemical parameter.

Table 3. Biochemical variables at baseline and week 16 for the two treatment groups.

The calculated free testosterone values were less than the 90th centile value for young women in 80% of the participants at week 16. Of the women who had values above 38.64 pmol/L at week 16, 7 were in the letrozole group and 5 were in the placebo group. Furthermore, 30% of women had free testosterone values below the 10th centile for young women at week 16.

There was no difference between the two groups for the change in total testosterone from week 0 to week 16 [letrozole group, mean ± SD, 1.54 ± 2.61 nmol/L, and placebo group, 1.38 ± 2.02 nmol/L; mean difference 0.16 nmol/L (95% CI, —1.04 to 1.37nmol/L)] orforthe change in calculated free testosterone from week 0 to week 16 [letrozole group, mean ± SD, 23.88 ±39.51 nmol/L, and placebo group, 18.34 ± 24.65 nmol/L; mean difference 5.54 nmol/L (95% CI, —11.48 to 22.56 nmol/L)].

Because the use of letrozole was not associated with any biochemical differences between the treatment groups, to examine the effects of transdermal testos-terone therapy on each of the biochemical end points, the two treatment groups were combined and considered as one group. Between week 0 and week 16, there were statistically significant increases in the mean levels of total and calculated free testosterone. Testosterone therapy was not associated with any statistically significant changes in SHBG, E₂, total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, Lp(a), or hsCRP (Table 3).

Sexual function, general well-being, and mood

The treatment groups were comparable in terms of baseline values for sexual function (SSS), mood (BDI), and well-being (PGWB Index) (Table 4). There was no difference between groups at week 16 for SSS, BDI, or PGWB Index.

Table 4. Effects of therapy on sexual function, well-being, and mood.

We again looked at the effects of testosterone treat-ment in the two groups combined. There was a highly significant increase in the mean scores for the SSS and the PGWB Index and a highly significant decrease in the mean scores for the BDI between weeks 0 and 16. In the regression models for each outcome, the baseline value for the PGWB Index and BDI made a statistically significant contribution (P < 0.001) to the value of the PGWB Index and the BDI at week 16. The value of SSS at baseline did not make a statistically significant contri-bution to the value for SSS at week 16. Group allocation did not contribute significantly to the determination of the value at week 16 for any of these parameters.

The changes in SSS domain scores for the two treatment groups are shown in Table 5. The change in the mean score for the domain of sexual activity was significantly greater in the placebo group than the letrozole group (P = 0.05), but a statistically significant difference was not seen for any other sexual domain. There were no significant differences between treatment groups seen for changes in any of the domains of the PGWB Index.

Table 5. Change between baseline and week 16 in Sabbatsberg Sexual Self-rating Scale domains, Psychological General Well-being Index domains, and Beck Depression Inventory.

Safety

Application of the transdermal gel was well tolerated, although 2 women (1 in each group) reported hair growth at the site of application, and one woman reported itchiness at the application site.

Hirsutism scores out of a total of 44 changed by up to 2 points in 10 women (a decrease in 4 women and an increase in 6 women), which we considered within observational error. One woman treated with testoster-one plus letrozole had an increase of 4 points at week 16, achieving a final score of 8 of 44. No woman had a score higher than 8 at week 16. One woman treated with testosterone plus letrozole and 2 women treated with testosterone reported acne. No serious adverse event occurred during the study.

Discussion

The aim of this study was to investigate whether aromatization plays a significant role in the therapeutic action of testosterone in women, as suggested by animal models. Significant and equivalent increases in the mean sexual function and well-being scores and a decrease in the mean BDI scores was seen in both the letrozole and placebo groups, indicating that inhibiting the aromatization of testosterone to E₂ does not impair the effects of testosterone on sexual function and mood in estrogenized postmenopausal women.

All women who participated were postmenopausal at entry into the study and were considered estrogen replete, being free of vasomotor symptoms and vaginal dryness on their ongoing transdermal oestrogen therapy. All received transdermal testosterone and achieved increased circulating total and free testosterone levels with this therapy.

The study design did not include a placebo-placebo treatment group because it was a mechanistic study, not an efficacy study. Furthermore, the inclusion of a placebo-placebo group would have meant that we could not monitor serum levels and include the opportunity for dose adjustment without unblinding the study. Not only did monitoring protect participants, but limiting excessive exposure was also intended to minimize the possibility of achieving supraphysiologic testosterone levels that would override the normal physiology we wished to study.

The mean and range of baseline total and calculated free testosterone levels were consistent with mean levels we have recently reported for Australian women aged 55 to 64 years.[30] Furthermore, the mean levels achieved in total and free testosterone were within the range reported for young healthy women[30] at week 16. However, 30% of women had low values and 80% had higher values at this time point despite attempted titration at week 4. This may reflect variations in timing of application of the final dose or possibly differences in transdermal absorption of the gel. Mean SHBG levels did not change during the study period. Therefore, we believe we achieved a sound mechanistic model in which to study the effects of aromatase enzyme inhibition.

There were no significant differences between letrozole and placebo treatment groups for any total score or domain score other than the sexual activity domain of the SSS. As an isolated finding and in the setting of multiple statistical comparisons, this is of questionable significance and is very likely to be a chance finding.

The improvements in sexual function and well-being over baseline were clinically significant and equal in magnitude to increases previously reported by us and others with exogenous testosterone therapy.[4,5,15] Such substantial increases in these parameters are highly unlikely to be attributable to a placebo effect and have not been seen with placebo therapy in our studies.[4,31] Although Shifren et al reported a significant placebo effect on sexual function, this did not apply to the PGWB Index in the same study.[5] Therefore, we do not believe the lack of an effect of aromatase inhibition in our study is simply due to a placebo effect of testosterone in both treatment groups. The baseline scores for the BDI and PGWB Index made highly significant contributions to each of these measures at week 16. That this did not hold true for the assessment of sexual function suggests that the determination of sexual function is complex, and its assessment remains problematic.

Letrozole is a nonsteroidal aromatase inhibitor. Clinical tracer studies in postmenopausal women show that letrozole inhibits peripheral aromatase by more than 98% and suppresses blood and urinary estrogen levels by more than 95% after 2 weeks of treatment.[25] Despite the possible contribution of a placebo effect, we conclude that near-complete blockade of aromatization of exogenous and endogenous testosterone does not impair improvements in sexual function and well-being with testosterone therapy. Therefore, our findings would support further evaluation of the use of poorly aromatizable androgens such as methyltestosterone and nonsteroidal selective androgen receptor modulators32 for the treatment of low libido in postmenopausal women.

In summary, aromatase inhibition did not seem to negate the favorable effect of testosterone therapy on sexual function and well-being. Short-term restoration of total and free testosterone to the physiologic range for young, healthy women did not adversely affect fasting lipids, Lp(a), or hsCRP.

References

Persky H, Dreisbach L, Miller WR, et al. The relation of plasma androgen levels to sexual behaviours and attitudes of women. Psychosom Med 1982;44:305-319. 

Sherwin BB, Gelfand MM. The role ofandrogen in the maintenance of sexual function in oophorectomized women. Psychosom Med 1987;49:397-409. 

Sherwin BB, Gelfand MM. The role ofandrogen in the maintenance of sexual functioning in oophorectomized women. Psychosom Med 1987;49:397-409. 

Goldstat R, Briganti E, Tran J, Wolfe R, Davis S. Transdermal testosterone improves mood, well being, and sexual function in premenopausal women. Menopause 2003;10:390-398. 

Shifren JL, Braunstein G, Simon J, et al. Transdermal testosterone treatment in women with impaired sexual function after oophorectomy. NEngUMed 2000;343:682-688. 

Sherwin BB. Affective changes with estrogen and androgen replacement therapy in surgically menopausal women. J Affect Disord 1988;14:177-187. 

Brincat M, Studd JWW, O’Dowd T, et al. Subcutaneous hormone implants for the control of climacteric symptoms. Lancet 1984; 1:16-18. 

Montgomery J, Brincat M, Appleby L, Versi E, Fenwick P, Studd JWW. Effect of oestrogen and testosterone implants on psychological disorders in the climacteric. Lancet 1987;1:297-299. 

Burger HG, Hailes J, Menelaus M. The management of persistent symptoms with estradiol-testosterone implants: clinical, lipid, and hormonal results. Maturitas 1984;6:351-358. 

Sarrel P, Dobay B, Wiita B. Estrogen and estrogen-androgen replacement in postmenopausal women dissatisfied with estrogen- only therapy: sexual behaviour and neuroendocrine response. JReprod Med 1998;43:847-856. 

Davis SR, McCloud PI, Strauss BJG, Burger HG. Testosterone enhances estradiol’s effects on postmenopausal bone density and sexuality. Maturitas 1995;21:227-236. 

Studd JWW, Colins WP, Chakravarti S. Estradiol and testosterone implants in the treatment of psychosexual problems in postmenopausal women. Br J Obstet Gynaecol 1977;84:314-315. 

Tuiten A, Von Honk J, Koppeschaar H, Bernaards C, Thijssen J, Verbaten R. Time course of effects of testosterone administration on sexual arousal in women. Arch Gen Psychiatry 2000;57:149-153. 

Dupuy H. The Psychological General Well-being (PGWB) Index. In: Wenger N, Mattson M, Furberg C, Elinson J, eds. Assessment of Quality of Life in Clinical Trials of Cardiovascular Therapies. New York: Le Jacq Publishing, 1984:170-183. 

Davis S, Rees M, Ribot J, Moufarage A, Rodenberg C, Purdie D. Efficacy and safety of testosterone patches for the treatment of low sexual desire in surgically menopausal women [Abstract]. Fertil Steril 2003;80(Suppl 3):76. 

McEwen BS. Sex differences in the brain: what are they and how do they arise? In: Notman MT, Nadeson CC, eds. Women and Men: New Perspectives on Gender Differences. Washington, DC: American Psychiatric, 1991:35-41. 

Bixo M, Backstrom T, Winblad B, Andersson A. Estradiol and testosterone in specific regions of the human female brain in differentendocrine states. JSteroidBiochemMolBiol 1995;55:297- 303. 

Roselli CE, Klosterman S, Resko J. Anatomic relationships between aromatase and androgen receptor mRNA expression in the hypothalamus and amygdala of adult male cynomolgus monkeys. J Compl Neurol 2001;439:208-223. 

Ogawa S, Chan J, Chester AE, Gustafsson J-A, Korach KS, Pfaff DW. Survival of reproductive behaviours in estrogen receptor beta gene-deficient male and female mice. Proc Natl Acad Sci USA 1999;96:12887-12892. 

Roberston K, Simpson E, Lacham-Caplan O, Jones MEE. Characterisation of the fertility and sexual behaviour of the male aromatase knockout (arKO) mouse. JAndrol 2001;22:825-830. 

Hutchison JB, Steimer TH. Prepoptic formation of 17β oestradiol is influenced by behavioural stimuli in the dove. Brain Res 1985; 360:366-369. 

Fedor-Freybergh P. The influence of estrogens on the wellbeing and mental performance in climacteric and postmenopausal women. Acta Obstet Gyaenecol Scand 1977;64(Suppl):1-68. 

Garratt A, Torgerson D, Wyness J, Hall M, Reid DM. Measuring sexual functio in pre menopausal women. Br J Obstet Gynaecol 1995;102:311-316. 

Beck AT, Steer R, Brown G. BDI Manual, 2nd ed. San Antonio, TX: Harcourt, Brace and Co., The Psychological Corporation, 1987. 

Lamb HM, Adkins JC. Letrozole: a review of its use in postmenopausal women with advanced breast cancer. Drugs 1998;56: 1125-1140. 

Vermeulen A, Verdonck L, Kaufman M. A critical evaluation of simple methods for the estimation of free testosterone in serum. J Clin Endocrinol Metab 1999;84:3666-3672. 

Sodergard R, Backstrom T, Shanhag V, Carstensen H. Calculation of free and bound fractions of testosterone and estradiol-17 beta to human plasma proteins at body temperature. J Steroid Biochem 1982;16:801-810. 

Friedewald WT, Levy RI, Frederickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18: 499-502. 

Ferriman D, Gallwey JD. Clinical assessment ofbody hair growth in women with hyperandrogenism. Endocrinol Metab 1961;21:1440- 1447. 

Davison S, Bell R, Donath S, Montalto J, Davis S. Androgen levels in adult females: changes with age, menopause and oophorectomy. J Clin Endocrinol Metab 2005;90:3847-3853. 

Davis S, Lucas J, Moufarage A, Braunstein G. Testosterone patch for treatment of low sexual desire in surgically menopausal women [Abstract]. Proceedings of the 7th Annual Congress of the Australasian Menopause Society. Hobart, Tasmania, Australia: 2003. 

Brown T. Nonsteroidal selective androgen receptors modulators (SARMs): designer androgens with flexible structures provide clinical promise. Endocrinology 2005;145:5420-5428. 

Lobo R, Rosen RC, Yang H-M, Block B, Van der Hoop R. Comparative effects of oral esterified estrogens with and without methyl testosterone on endocrine profiles and dimensions of sexual function in postmenopausal women with hypoactive sexual desire. Fertil Steril 2003;79:1341-1352. 

Hickok LR, Toomey C, Speroff L. A comparison of esterified estrogens with and without methyltestosterone: effects on endometrial histology and serum lipoproteins in postmenopausal women. Obstet Gynecol 1993;82:919-924. 

Ridker PM, Buring J, Shih J, Matias M, Hennekens CH. Prospective study of C-reactive protein and the risk of future cardiovascular events among apparently healthy women. Circulation 1998;98:731- 733. 

Cremer P, Nagel D, Labrot B, et al. Lipoprotein Lp(a) as predictor of myocardial infarction in comparison to fibrinogen, LDL cholesterol and other risk factors: results from the prospective Gottingen Risk Incidence and Prevalence Study (GRIPS). Eur J Clin Inv 1994; 24:444-453.