Androgens During the Reproductive Years: What Is Normal for Women?

Marina A. Skiba; Robin J. Bell; Rakibul M. Islam; David J. Handelsman; Reena Desai; Susan R. Davis

J Clin Endocrinol Metab. 2019;104(11):5382-5392.

Abstract and Introduction

Abstract

Objective: Whether serum androgen levels can identify women with "androgen insufficiency" or "androgen excess" is unresolved; thus, what constitutes "normal" remains uncertain. We sought to determine whether androgens, including 11-oxygenated C19 steroids, vary with age, menstrual cycle, or body mass index (BMI), during the reproductive years.

Design and Setting: Cross-sectional study recruited from eastern Australian states.

Participants: A total of 588 women, aged 18 to 39 years, who were not pregnant, lactating, or using systemic hormone therapy, with regular menstrual cycles and no previous diagnosis of polycystic ovarian syndrome.

Main Outcome Measures: Sex steroids measured using liquid chromatography-tandem mass spectrometry.

Results: Testosterone and androstenedione concentrations were significantly higher during the menstrual cycle mid- and luteal phases than in the early follicular phase, with median values across the cycle of 0.34 nmol/L (range, 0.04 to 1.01) and 1.97 nmol/L (range, 0.53 to 7.89), respectively. No cyclical variations were found in dehydroepiandrosterone (DHEA; 4.91 nmol/L; range, 0.08 to 23.51), 11-ketoandrostenedione (11KA; 7.99 nmol/L; range, 0.07 to 31.67), or 11-ketotestosterone (11KT; 1.27 nmol/L; range, 0.03 to 7.61). Overweight women had lower median testosterone (P < 0.05), DHEA (P < 0.05), and 11KA (P < 0.01) levels than normal-weight women. All C19 steroids were significantly lower (P < 0.01) in those aged 35 to 39 years than in those aged 18 to 25 years. The median 11KA/androstenedione (4.3:1) and 11KT/testosterone (3.9:1) ratios did not change with age, after adjustment for BMI and cycle stage.

Conclusions: We have demonstrated that 11KA and 11KT are stable across the menstrual cycle and make major quantitative contributions to the circulating androgen pool. All C19 androgens declined with age before menopause; hence, age-specific reference ranges are required for the interpretation of androgen levels in premenopausal women.

Introduction

Despite evidence that androgens are essential for female reproductive function^[1] and health,^[2] numerous obstacles to elucidating androgen physiology in women have been present. Among these have been the imprecision of assays for the measurement of low circulating testosterone concentrations in women relative to men,^[3,4] and the complexity of androgen action, which is mediated through intracellular aromatization and reduction to the biologically potent metabolite estradiol (E₂) and the major androgens testosterone and dihydrotestosterone (DHT). Moreover, the apparent disconnect between the clinical indicators of androgen excess and insufficiency in women and the blood levels of testosterone and the classic preandrogens, androstenedione and dehydroepiandrosterone (DHEA),^[5–7] might result from the presence of other circulating androgens that, until recently, have been overlooked. These include the 11-oxygenated steroids, 11-ketotestosterone (11KT) and 11-ketodihydrotestosterone (11KDHT), which are made in peripheral tissues from their abundant adrenal precursors, 11-hydroxyandrostenedione, 11-hydroxytestosterone, and 11-ketoandrostenedione (11KA).^[8] 11KT and 11KDHT have been reported to bind and activate androgen receptors and to exhibit similar androgenic potency to testosterone and DHT.^[9] Whether these steroids account for a meaningful proportion of the clinically active circulating androgen pool in healthy women is uncertain. Steroid measurement using liquid chromatography–tandem mass spectrometry (LC-MS/MS) enables accurate quantification of these steroids. To the best of our knowledge, the present study is the first to report the physiological concentrations of these steroids by menstrual cycle stage, age, and body mass index (BMI) and the concentrations relative to testosterone and its precursors androstenedione and DHEA, and metabolites E₂, estrone (E₁), and DHT. We have provided age-specific validated reference ranges for the main androgens and estrogens in women and, hence, a foundation for the exploration of the consequences of a woman having circulating androgen levels outside the age-specific ranges.

1 2 3 4

Next Section

Abstract and Introduction
Methods
Results
Discussion
References

Table 1. Sociodemographic Characteristics Stratified by Cycle Phase
Characteristic	All Participants (n = 588)	Early Follicular (n = 163)	Midcycle (n = 184)	Luteal (n = 241)
Age, y
18–<25	84 (14.3)	26 (16.0)	26 (14.1)	32 (13.3)
25–<30	106 (18.0)	27 (16.6)	39 (21.2)	40 (16.6)
30–<35	173 (29.4)	47 (28.8)	48 (26.1)	78 (32.4)
35–<40	225 (38.3)	63 (38.7)	71 (38.6)	91 (37.8)
BMI, kg/m²
<18.5 (underweight)	23 (3.9)	4 (2.5)	8 (4.3)	11 (4.6)
18.5–<25 (normal weight)	261 (44.4)	71 (43.6)	84 (45.7)	106 (44.0)
25–<30 (overweight)	161 (27.4)	49 (30.1)	47 (25.5)	65 (27.0)
≥30 (obese)	143 (24.3)	39 (23.9)	45 (24.5)	59 (24.5)
Ethnicity
Caucasian/white	433 (73.6)	123 (75.5)	130 (70.7)	180 (74.7)
Aboriginal/Torres Strait Islander	9 (1.5)	2 (1.2)	5 (2.7)	2 (0.8)
Asian	96 (16.3)	25 (15.3)	35 (19.0)	36 (14.9)
Polynesian/Melanesian/Maori	6 (1.0)	1 (0.6)	2 (1.1)	3 (1.2)
North African/Middle Eastern	6 (1.0)	3 (1.8)	2 (1.1)	1 (0.4)
Other	38 (6.5)	9 (5.5)	10 (5.4)	19 (7.9)
Relationship status (n = 587)
Married/de facto	358 (60.9)	98 (60.1)	107 (56.5)	153 (63.5)
Boyfriend/girlfriend or single	229 (38.9)	64 (39.3)	77 (41.8)	88 (36.5)
Parity
No children	320 (54.4)	91 (55.8)	104 (56.5)	125 (51.9)
≥1 child	268 (45.6)	72 (44.2)	80 (43.5)	116 (48.1)

Data presented as n (%).

Table 2. Hormonal Levels Stratified by Menstrual Cycle Phase Adjusted for Age and BMI
Hormone	Early Follicular (n = 163)	Midcycle (n = 184)	Luteal (n = 241)	Overall (n = 588)
E₂, pmol/L	194.20 (5.51–2301.72)	453.37^a (5.51–3582.90)	466.22^a (5.51–1997.02)	372.61 (5.51–3582.90)
E₁, pmol/L	190.13 (1.85–761.99)	290.37^a (13.21–1908.68)	314.05^a (33.03–1413.02)	257.45 (1.85–1908.68)
Testosterone, nmol/L	0.32 (0.04–0.85)	0.35^b (0.09–1.01)	0.35^c (0.07–0.89)	0.34 (0.04–1.01)
11KT, nmol/L	1.30 (0.33–7.61)	1.34 (0.03–5.79)	1.23 (0.09–4.57)	1.27 (0.03–7.61)
Androstenedione, nmol/L	1.82 (0.63–5.37)	2.07^b (0.63–7.89)	2.00 (0.53–6.84)	1.97 (0.53–7.89)
11KA, nmol/L	8.80 (0.07–28.84)	7.71 (1.10–31.67)	7.73 (0.57–27.97)	7.99 (0.07–31.67)
DHEA, nmol/L	5.27 (1.01–21.25)	4.91 (0.44–23.51)	4.78 (0.08–17.86)	4.92 (0.08–23.51)

Data presented as median (range).
E₁ and 11KT had one participant each with an unmeasurable level; E₂ had seven participants (1.1%) with unmeasurable levels; all other hormones had measurable results for all participants.
^a P < 0.001; linear regression of hormone levels stratified by cycle performed on log-transformed data, with early follicular phase as reference.
^b P < 0.01; linear regression of hormone levels stratified by cycle performed on log-transformed data, with early follicular phase as reference.
^c P < 0.05; linear regression of hormone levels stratified by cycle performed on log-transformed data, with early follicular phase as reference.

Table 3. Hormone Levels for Women in Day 1 to 35 of Cycle by BMI
Hormone	Underweight (<18.5 kg/m²; n = 23)	Normal (18.5–24.9 kg/m²; n = 261)	Overweight (25–29.9 kg/m²; n = 161)	Obese (>30 kg/m²; n = 143)	Overall (n = 588)
Testosterone, nmol/L	0.34 (0.18–1.01)	0.36 (0.04–1.00)	0.31^a (0.07–0.89)	0.34 (0.07–0.85)	0.34 (0.04–1.01)
11KT, nmol/L	1.09 (0.53–2.64)	1.30 (0.09–5.79)	1.24 (0.37–3.61)	1.43 (0.03–7.61)	1.27 (0.03–7.61)
Androstenedione, nmol/L	2.16 (0.87–4.89)	2.08 (0.53–5.34)	1.83 (0.63–7.89)	1.87 (0.68–6.84)	1.97 (0.53–7.89)
11KA, nmol/L	9.45 (1.68–24.14)	8.79 (0.69–31.67)	7.76^b (0.57–28.84)	6.68^c (0.07–27.47)	7.99 (0.07–31.67)
DHEA, nmol/L	5.55 (2.04–23.51)	5.37 (0.08–19.83)	4.40^a (1.01–21.25)	4.47 (0.28–17.86)	4.92 (0.08–23.51)

Data presented as median (range).
11KT had one participant with an unmeasurable level; all other hormones had measurable results for all participants.
^a P < 0.05; regression results (adjusted for age and cycle stage) with normal weight as the reference.
^b P < 0.01; regression results (adjusted for age and cycle stage) with normal weight as the reference.
^c P < 0.001; regression results (adjusted for age and cycle stage) with normal weight as the reference.

Table 4. Linear Regression Analysis of Hormone Levels by Age
Characteristic	Age, y	β-Coefficient (95% CI)
Characteristic	Age, y	Unadjusted	Adjusted for Cycle Stage and BMI
Testosterone, nmol/L	18–<25	1.0	1.0
	25–<30	−0.12 (−0.24 to −0.01); P = 0.071	−0.13 (−0.25 to −0.00); P = 0.050
	30–<35	−0.17 (−0.28 to −0.05); P = 0.005	−0.17 (−0.28 to −0.06); P = 0.004
	35–<40	−0.30 (−0.41 to −0.19); P < 0.0001	−0.30 (−0.41 to −0.19); P < 0.0001
11KT, nmol/L	18–<25	1.0	1.0
	25–<30	−0.12 (−0.27 to 0.03); P = 0.116	NS
	30–<35	−0.17 (−0.31 to −0.03); P = 0.014	−0.18 (−0.32 to −0.04); P = 0.010
	35–<40	−0.16 (−0.29 to −0.02); P = 0.021	−0.18 (−0.31 to −0.04); P = 0.010
Androstenedione, nmol/L	18–<25	1.0	1.0
	25–<30	−0.14 (−0.26 to −0.02); P = 0.025	−0.15 (−0.27 to −0.03); P = 0.016
	30–<35	−0.21 (−0.32 to −0.10); P < 0.0001	−0.22 (−0.32 to −0.10); P < 0.0001
	35–<40	−0.36 (−0.46 to 0.25); P < 0.0001	−0.36 (−0.46 to −0.25); P < 0.0001
11KA, nmol/L	18–<25	1.0	1.0
	25–<30	−0.15 (−0.34 to 0.03); P = 0.107	NS
	30–<35	−0.23 (−0.40 to −0.06); P = 0.010	−0.19 (−0.36 to −0.02); P = 0.029
	35–<40	−0.31 (−0.47 to −0.14); P < 0.0001	−0.23 (−0.40 to 0.07); P = 0.006
DHEA, nmol/L	18–<25	1.0	1.0
	25–<30	−0.21 (−0.39 to −0.04); P = 0.016	−0.21 (−0.38 to −0.04); P = 0.019
	30–<35	−0.38 (−0.54 to −0.22); P < 0.0001	−0.36 (−0.52 to −0.21); P < 0.0001
	35–<40	−0.49 (−0.66 to −0.34); P < 0.0001	−0.47 (−0.62 to −0.31); P < 0.0001
11KT/testosterone ratio	18–<25	1.0	NS
	25–<30	−0.01 (−0.17 to 0.16); P = 0.954
	30–<35	−0.01 (−0.16 to 0.15); P = 0.919
	35–<40	0.14 (−0.01 to 0.29); P = 0.064
11KA/androstenedione ratio	18–<25	1.0	NS
	25–<30	−0.02 (−0.21 to 0.17); P = 0.871
	30–<35	−0.13 (−0.19 to 0.16); P = 0.884
	35–<40	0.05 (−0.12 to 0.22); P = 0.548

11KT had one participant with an unmeasurable level; all other hormones had measurable results for all participants.
Abbreviation: NS, not significant.

Table 5. Sex Steroid Levels by Age
Sex Steroid	Age, y				Overall (n = 588)
Sex Steroid	18–<25 (n = 84)	25–<30 (n = 106)	30–<35 (n = 176)	35–<40 (n = 227)	Overall (n = 588)
Testosterone, nmol/L	0.40 (0.10–0.89)	0.35 (0.14–1.01)	0.35 (0.04–0.79)	0.30 (0.07–1.00)	0.34 (0.04–1.01)
11KT, nmol/L	1.45 (0.61–5.79)	1.29 (0.09–4.47)	1.23 (0.22–3.34)	1.27 (0.03–7.61)	1.27 (0.03–7.61)
Androstenedione, nmol/L	2.44 (0.89–5.97)	2.24 (0.53–4.89)	1.91 (0.68–7.89)	1.69 (0.63–6.84)	1.97 (0.53–7.89)
11KA, nmol/L	10.41 (1.10–31.67)	8.46 (0.69–25.11)	7.66 (1.57–28.84)	7.46 (0.07–27.97)	7.99 (0.07–31.67)
DHEA, nmol/L	6.86 (2.96–23.51)	6.01 (0.08–21.25)	4.78 (1.01–16.57)	4.40 (0.28–19.69)	4.92 (0.08–23.51)
11KT/testosterone ratio	3.68 (1.11–12.50)	3.61 (0.36–12.49)	3.36 (0.92–19.15)	4.23 (0.06–20.14)	3.88 (0.06–20.14)
11KA/androstenedione ratio	4.13 (0.48–14.18)	4.42 (0.28–19.08)	4.10 (0.54–19.08)	4.47 (0.05–23.72)	4.29 (0.05–23.72)

Data presented as median (range).

processing....

Androgens During the Reproductive Years: What Is Normal for Women?

Abstract and Introduction

Abstract

Introduction

Tables

Tables

Tables

Tables

Tables

References

Authors and Disclosures

Authors and Disclosures

Email This

Feedback