The aim of the present study was to investigate a possible correlation between decreased androgen levels and female sexual function index (FSFI) in women with low libido and compare these findings with normal age-matched subjects. In total, 20 premenopausal women with low libido (mean age 36.7; range 24–51 y) and 20 postmenopausal women with low libido (mean age 54; 45–70 y), and 20 premenopausal healthy women (mean age 32.2; range 21–51 y) and 20 postmenopausal healthy women (mean age 53.5; range 48–60 y) as controls were enrolled in the current study. Women with low libido had symptoms for at least 6 months and were in stable relationships. All premenopausal patients had regular menstrual cycles and all postmenopausal patients and controls were on estrogen replacement therapy. None of the patients were taking birth control pills, corticosteroids or had a history of chronic medical illnesses. All completed the FSFI and Beck's Depression Inventory (BDI) questionnaires. Hormones measured included: cortisol; T3, T4 and TSH; estradiol; total and free testosterone; dehydroepiandrosterone sulfate (DHEA-S); sex hormone binding globulin (SHBG). We performed statistical analysis by parametric and nonparametric comparisons and correlations, as appropriate. We found significant differences between the women with low libido and the controls in total testosterone, free testosterone and DHEA-S levels and full-scale FSFI score for both pre- and postmenopausal women (P<0.05). In addition, decreased total testosterone, free testosterone and DHEA-S levels positively correlated with full-scale FSFI score and FSFI-desire, FSFI-arousal, FSFI-lubrication and FSFI-orgasm scores (P<0.05). Our data suggest that women with low libido have lower androgen levels compared to age-matched normal control groups and their decreased androgen levels correlate positively with female sexual function index domains.
Increasing interest is being shown in the role of androgens in female health and well being; however, the presence of androgen deficiency states and their clinical diagnosis and management is still a controversial matter.
For the last two decades, low desire has been the most common sexual problem of women. Female sexual dysfunction studies have started to focus on biological and classification aspects of this problem.1
Several studies have shown a decrease in sexual desire symptoms with androgen replacement therapy where adequate levels of estrogen are present.2, 3 For several decades, testosterone has increasingly been given to naturally postmenopausal or castrated women when estrogen replacement did not change decreased libido caused by a decrease in ovarian steroid production.4, 5, 6 A recent study has shown that premenopausal women could have decreased testosterone levels and still have regular menstrual cycles.7 Riley and Riley8 have also shown a correlation between low testosterone levels and decreased libido in younger women who complain of a lack of sexual desire.
Since Rosen et al9 developed and validated a brief and reliable self-report measure of sexual functioning in women, this questionnaire has been popular in evaluating the key dimensions of female sexual dysfunction.
In fact, little information is available about what constitutes an abnormal androgen level in pre- and postmenopausal women and whether those values directly correlate with sexual function index domains.
The aim of the present study was to establish a possible relation between serum androgen hormone levels (total testosterone (total T), free testosterone (free T), dehydroepiandrosterone sulfate (DHEA-S)) and Female Sexual Function Index (FSFI) domains and full-scale scores in pre- and postmenopausal women with low libido. In addition to this, all these findings were compared with pre- and postmenopausal healthy controls.
Materials and methods
Between October 2002 and March 2003, 107 women from Ege University Urology and Gynecology & Menopause Clinics were evaluated during initial consultation of sexual dysfunction. A total of 40 women with a chief complaint of decreased sexual desire were then selected for this study. A total of 20 premenopausal women with low libido (mean age 36.7; range 24–51 y) and 20 postmenopausal women with low libido (mean age 54; 45–70 y) were enrolled in the current study. Women with low libido had symptoms for at least 6 months and were in stable relationships. All premenopausal patients had regular menstrual cycles and all postmenopausal patients and controls were on estrogen replacement therapy (ERT). Five postmenopausal patients had a history of total abdominal hysterectomy with bilateral salpingo-oophorectomy (TAH and BSO), whereas 15 postmenopausal patients had natural menopause. Patients with surgical menopause were on 0.625 mg conjugated estrogen and patients with natural menopause were on 0.625 mg conjugated estrogen and 2.5 mg medroxyprogesterone acetate treatment. None of the patients were taking birth control pills, corticosteroids or had a history of chronic medical illnesses. Patients with a history of classic adrenal insufficiency, thyroid disease, major depression, chronic fatigue syndrome, chronic metabolic disease, radiotherapy and chemotherapy history were excluded from the study. All the participants indicated their weight in kilograms and height in centimeters. All women who participated in the study signed an informed consent.
All women completed Beck's Depression Inventory (BDI) and FSFI (score range 4–95). Patients with total BDI score over 21 were excluded (BDI score range: 0–63). We used the FSFI questionnaire for the study, as FSFI is a validated and reliable 19 item questionnaire with six domains (desire (1–2), arousal (3–6), lubrication (7–10), orgasm (11–13), satisfaction (14–16), pain (17–19)), which has been developed as a brief, multidimensional, self-report instrument for assessing the key dimensions of sexual function in women.
Hormones measured were cortisol; T3, T4 and TSH; estradiol; sex hormone binding globulin (SHBG); total and free testosterone; DHEA-S. This was carried out in the middle of the menstrual cycle (days 8–15) for all premenopausal women. Blood was drawn from all women between 0800 and 1200 and stored at −20°C for a period of 1–30 days until assayed. All tests were performed in the Ege University Endocrinology Laboratory in a single continuous hormone assay for each hormone.
Hormones were measured by commercially available RIA kits. Total testosterone analysis was performed by direct Automated Chemiluminescence Sytem (Bayer Diagnostics, New York) with a normal range of 14–76 ng/ml. Free testosterone was determined by direct analog free testosterone radioimmunoassay (DSL-4900, TX, USA) with a normal range of 0.6–3 pg/ml. DHEA-S analysis was carried out by direct RIA (DSL-3500, TX, USA) with a normal range of 35–450 μg/dl. SHBG analysis was determined by immunoassay (DPC, Los Angeles) with a normal range of 18–114 nmol/l. The sensitivity of the assays was expressed as a minimal amount of the hormones distinguishable from the zero sample with 95% probability and the intra-assay and interassay coefficients of variation for total T, free T, DHEA-S and SHBG were 5.8 and 6.1, 5.3 and 6, 4.5 and 5.5, and 2.8 and 6.9%, respectively.
In total, 40 healthy women from Ege University staff were included in the study as volunteers. Of them, 20 premenopausal (mean age 32.2; range 21–51 y) and 20 postmenopausal (mean age 53.5; range 48–60 y) healthy women formed the control group. The total FSFI score for each patient in the control group was above 60. None of the control patients had a history of adrenal insufficiency, thyroid disease, major depression, radiotherapy, chemotherapy or any systemic chronic diseases. In addition to this, none of the postmenopausal controls had a history of TAH and BSO and they were all on 0.625 mg of conjugated estrogen and 2.5 mg medroxyprogesterone acetate treatment.
Student's t-test, Two-way analysis of variance, Pearson correlation analysis and Spearman rank correlation analysis were used for statistical analysis.
To show the statistically significant difference between groups, a minimum of 40 patients and 40 controls were required to achieve 80% power for the current study.
The mean ages of the patients population and control group were 45.3±10.9 y (range 24–70 y) and 42.6±12.6 y (range 21–60 y), respectively. The mean age of the 20 premenopausal patients was 36.7±6.9 y (range 24–51 y), whereas the mean age of the postmenopausal patients was 54±6.4 y (range 45–70 y). The mean age of the premenopausal healthy women was 32.2±9 y (range 21–51 y) and the mean age of the 20 postmenopausal healthy women was 53.5±4.6 y (range 48–60 y). Mean ages between the controls and patients did not show statistical difference (P>0.05). All of the patients had experienced decreased libido for a period of more than 6 months.
The mean BMI values for premenopausal controls, premenopausal patients, postmenopausal controls and postmenopausal patients were 22.2, 23.1, 24.9 and 25.6 kg/m2, respectively. The BMI between premenopausal controls and patients did not show statistical difference, and similarly the BMI between postmenopausal participants did not show any statistical difference (P>0.05). All the patients' and controls' SHBG levels were within normal levels (range 36–99 nmol/l).
The means and standard deviations (s.d.) for total testosterone, free testosterone and DHEA-S are listed in Table 1, separated by menopausal status as well as decade for both the patients and the control group.
The androgen levels in our patients were significantly lower than our control population (P<0.05). In Figure 1, the mean total testosterone (ng/ml) levels for all patients and controls are given according to menopause status and decade. There was no statistical difference in both patients and controls in their menopause status. Therefore, statistical analysis was performed for patients and controls. There was a statistical difference in both groups for mean total testosterone levels (P<0.05).
Figure 2 shows the mean free testosterone (pg/ml) levels for patients and controls separated according to menopause status and decade. There was a statistical difference for menopause status for both patients and controls (P<0.05). However, this difference had similar effects in both groups so there was no statistical interaction. As a result, there was a significant difference in both groups for mean free testosterone levels (P<0.05).
The mean DHEA-S levels for patients and controls, separated by menopause status and decade, are given in Figure 3. Statistical difference was significant for menopausal status for both patients and healthy women (P<0.05). However, this difference did not have a similar effect in both groups; therefore, there was a statistical interaction. In addition, there was no statistical difference for mean DHEA-S levels for pre- and postmenopausal patients (P>0.05). In contrast, there was a statistical difference in mean DHEA-S levels for pre- and postmenopausal women (P<0.05).
When age and androgen hormones were compared via statistical correlation analysis, there was no correlation between age and total testosterone levels, whereas there was a negative correlation between free testosterone and DHEA-S levels (P<0.05).
The comparison for total FSFI scores between patients and controls showed statistical difference (P<0.05). The mean (±s.d.) total FSFI score for patients was 63.4 (±11.4), but the mean (±s.d.) total FSFI score for healthy women was 80.6 (±7.7). When FSFI domains (desire, arousal, lubrication, orgasm, satisfaction, pain) were compared separately for both patients and healthy women, there was a significant difference between the two groups in all the components of FSFI except pain (P<0.05). In our study population, mean (±s.d.), desire domain scores were 5.2 (±2.2), 8 (±3.1), 7.5 (±2.4), 4.7 (±1.9), for premenopausal patients, premenopausal controls, postmenopausal patients and postmenopausal controls, respectively, and there was a significant difference between patients and healthy women (P<0.05).
We found a significant positive correlation between serum androgen levels and each FSFI domain, including the full-scale score with the exception of pain. In the correlation analysis of total FSFI scores and androgen hormones, the levels of total testosterone, free testosterone and DHEA-S showed a positive correlation with total FSFI scores (P<0.05) (Figure 4a–c). Moreover, there was a positive correlation between total testosterone levels and all FSFI domains except pain (P<0.05). Similarly, there was a positive correlation between DHEA-S levels and all FSFI domains except pain (P<0.05). Free testosterone levels also had a positive correlation with all FSFI domains except satisfaction and pain (P<0.05).
Our results have primarily shown that both pre- women and postmenopausal women with low libido have decreased levels of total testosterone, free testosterone and DHEA-S levels compared to age-matched healthy volunteers with comparable BMI and menopause status. Secondly, healthy women reported higher FSFI scores compared to the women with low libido. Lastly, all androgen levels positively correlated with FSFI domains except pain.
It is worth noting that the figures in this study do not show standard deviations.
Tazuke et al10 found no change in total testosterone in postmenopausal women on estrogen therapy but found lower free testosterone and DHEA-S levels. In contrast, Guay and Jacobson11 did not find any significant differences between the free T and DHEA-S levels in postmenopausal women on ERT or not. As this is a controversial issue, we selected postmenopausal patients on ERT so that possible hormonal imbalances could not be attributed to estrogen alone. It is important to note that postmenopausal patients in our study are a heterogeneous group, as five (one quarter) of the postmenopausal cases had a history of TAH and BSO. Therefore, this fact might impinge on androgen levels independently, as will the taking of estrogen alone vs estrogen and progesterone hormone therapy.
The second group of patients we studied were otherwise healthy premenopausal women with low sexual desire. As low sexual desire might be caused by other factors such as oral contraceptives, corticosteroids, depression and chronic medical illnesses, we ensured our selected patient and control groups did not have any of these conditions.12
Laumann et al13 evaluated the incidence of sexual dysfunction both in men and women and found that the mean incidence of decreased sexual desire in women was 32%. Guay and Jacobson11 then established that 70% of women complaining of decreased libido have lower testosterone levels. These results indicate that this problem is not rare. In fact, Munarriz et al14 have shown that treatment for androgen insufficiency and female sexual dysfunction by androgen replacement therapy is safe and effective.
The absence of precise definitions of androgen deficiency, unknown normal ranges for androgens and nuances in the sensitivity of the various assays have made research on androgens and libido in women difficult to interpret. For this reason, we compared our findings with a similar healthy group. In addition, establishing normal ranges of testosterone in women is difficult, and also free and bioavailable testosterone assays are new and there are numerous variations in technique. The normal ranges created by the companies producing the RIA kits are often made using a limited number of patients. Although a larger patient and control population must be enrolled to try and obtain ultimate conclusions, we found significant differences on androgens between the two groups (P<0.05).
As described earlier, blood was withdrawn in the midcycle between 0800 and 1200. The reason for this is that most researchers feel that androgens elevate to a peak near midcycle during this morning period.15
Several reports seem to underline a significant decline in androgens throughout the life span and a significant impact of the body weight over the hormonal profile and sexual activity; our patients and controls were accurately matched for both age and BMI.16, 17
Androgens in the female are derived from the adrenal gland (25%), the ovary (25%) and peripheral conversion of precursor hormones (50%).18 As suggested by Guay and Jacobson,11 the fact that both testosterone and DHEA-S are decreased in these patients brings up the possibility that there may be a decreased conversion in DHEA-S from its precursor steroid.
FSFI was designed to be a clinical trials assessment instrument that addresses the multidimensional nature of female sexual function. The comparison for total FSFI scores between patients and controls showed statistical difference (P<0.05). In the comparison of patients and healthy women, there was a significant difference between the two groups for all the components of FSFI except pain (P<0.05). One possible criticism of our study might be that postmenopausal patients were heterogeneous (constituted of both surgical and natural menopause patients) to allow direct comparison of the patients and controls.
Recently, Nappi et al19 showed the positive relationship between serum allopregnanolone and free T levels and each FSFI domain including the full-scale score except pain.
Similarly in our study, FSFI scores and androgen hormone correlation analysis showed that total testosterone, free testosterone and DHEA-S levels had positive correlation with total FSFI scores and each FSFI domain except pain.
In conclusion, our data suggest that women with low libido have lower androgen levels compared to age-matched normal control groups. We found that this problem also occurs in younger regularly menstruating women who are not on oral contraceptives. As all postmenopausal women in our study were on ERT, the low libido in these patients could well be attributed to decreased androgens. These women with apparently normal ovarian function might be suffering from adrenal deficiency as DHEA-S is produced from the adrenal gland. Lastly, it is possible to evaluate decreased androgen levels using self-report questionnaires.
Hartmann U, Heiser K, Rüffer-Hesse C, Kloth G . Female sexual desire disorders: subtypes, classification, personality factors and new directions for treatment. World J Urol 2002; 20: 79–88.
Davis S, McCloud P, Strauss B, Burger H . Testosterone enhances estradiol's effects on postmenopausal bone density and sexuality. Maturitas 1996; 21: 227–236.
Davis S . Androgen replacement in women: a commentary. J Clin Endocr Metab 1999; 84: 1886–1891.
Davis S . Testosterone and sexual desire in women. J Sex Educ Ther 2000; 25: 25–32.
Sarrel P, Dobay B, Wiita B . Estrogen and estrogen–androgen replacement in postmenopausal women dissatisfied with estrogen–only therapy: sexual behavior and neuroendocrine responses. J Reprod Med 1998; 43: 847–856.
Sarrel P . Effect of hormone replacement therapy on sexual psychophysiology and behavior in postmenopause. J Women Health Gender Based Med 2000; 9(Suppl 1): 25–32.
Guay AT . Decreased testosterone in regularly menstruating women with decreased libido: a clinical observation. J Sex Marital Ther 2001; 27: 513–519.
Riley A, Riley E . Controlled studies on women presenting with sexual disorders: I. Endocrine status. J Sex Marital Ther 2000; 26: 269–283.
Rosen R et al. The female sexual function index (FSFI): a multidimensional self-report instrument for the assessment of female sexual function. J Sex Marital Ther 2000; 26: 191–208.
Tazuke S, Khaw K, Barrett-Connor E . Exogenous estrogen and endogenous sex hormones. Medicine 1992; 71: 44–51.
Guay AT, Jacobson J . Decreased free testosterone and dehydroepiandrosterone-sulfate (DHEA-S) levels in women with decreased libido. J Sex Marital Ther 2002; 28: 129–142.
Bachmann G et al. Female androgen insufficiency: The Princeton consensus statement on definition, classification, and assessment. Fertil Steril 2002; 4: 660–665.
Laumann EO, Paik A, Rosen RC . Sexual dysfunction in the United States: prevalence and predictors. JAMA 1999; 281: 537–544.
Munarriz R et al. Androgen replacement therapy with dehydroepiandrosterone for androgen insufficiency and female sexual dysfunction: androgen and questionnaire results. J Sex Marital Ther 2002; 28: 165–173.
Massafra C et al. Androgens and osteocalcin during the menstrual cycle. J Clin Endoc Met 1999; 84: 971–974.
Barnhart KT et al. The effect of dehydroepiandrosterone supplementation to symptomatic perimenopausal women on serum endocrine profiles, lipid parameters, and health-related quality of life. J Clin Endocr Metab 1999; 84: 3896–3902.
Parker Jr CR et al. Effects of aging on adrenal function in the human: responsiveness and sensitivity of adrenal androgens and cortisol to adrenocorticotropin in premenopausal and postmenopausal women. J Clin Endocr Metab 2000; 85: 48–54.
Longcope C . Adrenal and gonadal androgen secretion in normal females. Clin Endocr Metab 1986; 15: 213–228.
Nappi RE et al. Serum allopregnanolone relate to FSFI score during the menstrual cycle. J Sex Marital Ther 2003; 29: 95–102.
About this article
Cite this article
Turna, B., Apaydin, E., Semerci, B. et al. Women with low libido: correlation of decreased androgen levels with female sexual function index. Int J Impot Res 17, 148–153 (2005) doi:10.1038/sj.ijir.3901294
- low libido
- androgen level
- female sexual desire disorder
- female sexual function index
Do Psychosocial Factors Moderate the Relation between Testosterone and Female Sexual Desire? The Role of Interoception, Alexithymia, Defense Mechanisms, and Relationship Status
Adaptive Human Behavior and Physiology (2019)
Archives of Gynecology and Obstetrics (2019)
Effects of dehydroepiandrosterone (DHEA) supplementation on sexual function in premenopausal infertile women
Effect of a multi-ingredient based food supplement on sexual function in women with low sexual desire
BMC Women's Health (2019)
Archives of Sexual Behavior (2019)