The hormonal profile in women using combined monophasic oral contraceptive pills varies across the pill cycle: a temporal analysis of serum endogenous and exogenous hormones using liquid chromatography with tandem mass spectroscopy

Abstract

Oral contraceptive pills, of all types, are used by approximately 151 million women worldwide; however, a clear understanding of the concentrations of endogenous and exogenous hormones across a 28-day combination monophasic oral contraceptive pill pack is not well described. In our study of 14 female participants taking various combination monophasic oral contraceptive pills, we found significant fluctuations in endogenous and exogenous hormone levels throughout the pill cycle. Our analysis revealed significantly greater levels of ethinyl estradiol on the 20th and 21st days of active pill ingestion, compared with days 1–2 (active) and days 27–28 (inactive pill ingestion). Conversely, estradiol concentrations decreased during active pill consumption, while progestin and progesterone levels remained stable. During the 7 days of inactive pill ingestion, estradiol levels rose sharply and were significantly higher at days 27–28 compared with the mid and late active phase time points, while ethinyl estradiol declined and progestin did not change. These findings challenge the previous assumption that endogenous and exogenous hormones are stable throughout the 28-day pill cycle.

NEW & NOTEWORTHY The results from this study have wide-ranging implications for research and treatment in women’s health including considerations in research design and interpretation for studies including women taking oral contraceptives, the potential for more precise and personalized methods of dosing to reduce unwanted side effects and adverse events, and the potential treatment of a variety of disorders ranging from musculoskeletal to neurological with exogenous hormones.

INTRODUCTION

Girls and women are underrepresented in biomedical research (1–3). This lack of inclusion has resulted in profound knowledge gaps and negative health impacts, partly due to the extrapolation of data from men to women (4), and increasing the likelihood of misdiagnosis and mistreatment in women compared with men (5–7). Cited reasons for the lack of inclusion of female subjects in both animal and human biomedical research are concerns that the cyclic fluctuation of hormones across the menstrual cycle could confound results and that research participation could decrease fertility or harm pregnancy (2, 8). To mitigate the potential impact of menstrual cycle hormones in female participants, one approach is to limit the inclusion to women who are using hormonal contraceptives. The rationale is that women chronically exposed to exogenous hormones in contraceptives will have stable and similar concentrations of both endogenous and exogenous hormones (9, 10).

Hormonal contraceptives are available in variable modes of delivery (e.g., oral, injectable, intrauterine), and oral contraceptives have been one of the most commonly used types since their creation in 1960 (11). Combination oral contraceptives contain ethinyl estradiol (EE) and synthetic progestin (SP), which can be classified by generation (e.g., first: i.e., norethindrone, second: i.e., levonorgestrel, third: i.e., norgestimate, fourth: i.e., drospirenone), with the typical pack containing 21 pills with active hormones and seven placebo pills. The exogenous hormones prevent ovulation by suppressing the hypothalamic-pituitary-ovarian axis, which leads to suppression of endogenous estradiol (E2) and progesterone (P) (12). There are a wide variety of combination oral contraceptives depending on type (monophasic: same dosage of EE and SP in the 21 active pills vs. triphasic: gradually increasing dose of progestin during the 21 active pills) and composition [range of EE dose: 10–50 mcg (13) and progestins with varying affinity for the progesterone and androgen receptors] (14–16). Although all oral contraceptives have similar efficacy rates in preventing pregnancy, the variability in type and composition influences the secondary effects of oral contraceptive use. For example, oral contraceptives containing progestins with high affinity for androgen receptors may result in acne, weight gain, and facial hair (17). The role of progestin androgenicity may also play a role in the risk of gestational diabetes (18), ovarian cancer (14), and cardiovascular disease (19). However, our understanding of which women will experience adverse events due to oral contraceptive exposure is limited. It is possible that the risk of secondary events is related not only to the oral contraceptives type but also to how an individual metabolizes the medication and the concentration of exogenous hormones to which tissues are exposed.

Understanding the basic biology of how exogenous hormones in oral contraceptive pills influence tissues is critical as approximately 151 million women worldwide are taking them (20), and the Food and Drug Administration has just approved an over-the-counter OC (21). At present, women on hormonal contraceptives, particularly combination monophasic oral contraceptive (CMOC), are considered a “control” group in many studies based on the assumption that their hormonal milieu is comparable and consistent (10, 22, 23). However, there remains a notable gap in our understanding of the concentrations of endogenous and exogenous hormones throughout the 28-day oral contraceptive CMOC pill pack. These data are crucial for guiding research design and accurately assessing the efficacy of interventions in women using CMOCs. The challenge in discerning endogenous and exogenous hormone data from existing studies stems from the methodologies employed for hormone measurement and the specific focus of experiments conducted. Initially, early investigations into both endogenous and exogenous hormones in oral contraceptive users relied on radioimmunoassays (24–29). Although these immunoassays were the “gold” standard at the time, they suffer from inherent limitations including antibody availability/selectivity, requirements for large sample volumes, limited multiplexing capabilities, extended assay run times, and issues with reproducibility (30) which prompted a push, by The Endocrine Society, to establish liquid chromatography/tandem mass spectrometry (LC-MS/MS) as the standard method for measuring sex hormones (31).

In addition, although some studies do track hormone concentrations over time, they often fail to report both endogenous and exogenous values (24, 25, 32–34). Moreover, many studies focus solely on pharmacokinetics hours after ingestion of a CMOC pill (33, 35–38), rather than everyday measurements over the pill treatment cycle.

Therefore, the objective of this study was twofold: first, to characterize the every-other-day concentrations of E2, P, EE, and SP in women using CM monophasic OCs using LC-MS/MS and to determine the intersubject variability of these hormones.

MATERIALS AND METHODS

Written informed consent was received from 14 healthy women (age: 28 ± 4 yr; body mass index: 22 ± 2 kg/m²; means ± SD) who had been on a stable regimen of CMOCs for at least 6 mo before the study. Of the 14 subjects, one used a CMOC containing EE and drospirenone (DRSP), three used a CMOC containing EE and levonorgestrel (LNG), and 10 used a CMOC containing EE and norethindrone (NET)/norethindrone acetate (NETA). Table 1 contains a detailed description including brand name and dosages of the exogenous hormones. These subjects were part of a larger study which had the following exclusion criteria: history of musculoskeletal or orthopedic injury of the spine, hip, knee, ankle, or foot; history of neurological injury or disease of the peripheral or central nervous system; current smoking habit; history of disordered eating, stress fracture, connective tissue disorder (Marfan syndrome, Ehlers–Danlos disease), or menstrual dysfunction (primary or secondary amenorrhea, oligomenorrhea, anovulatory cycles, polycystic ovarian disease); current or prior pregnancy; starting or stopping CMOCs within the previous 6 mo; use of an extended-cycle oral contraceptive that eliminates placebo pills; or use of a non-oral hormonal contraceptive. We also excluded those women exercising more than 7 h per week or participating in competitive-level sports because of the high rate of undiagnosed menstrual dysfunction in females in this population (14). This study was approved by the Northwestern University (STU00203492) and University of Texas Southwestern Medical Center (STU-2018-0192) institutional review boards.

Table 1.

Summary of oral contraceptives used by study participants

Subject No.	Brand Name (Active/Inactive Pills)	Ethinyl Estradiol Dose, μg	Progestin	Progestin Dose, mg
1	Zarah (21/7)	30	Drospirenone	3
2	Portia (21/7)	30	Levonorgestrel	0.15
3	Aviane (21/7)	20	Levonorgestrel	0.1
4	Altavera (21/7)	30	Levonorgestrel	0.15
5	Microgestin (21/7)	20	Norethindrone	1
6	Microgestin (21/7)	30	Norethindrone	1.5
7	Microgestin (21/7)	20	Norethindrone	1
8	Microgestin (21/7)	20	Norethindrone	1
9	Junel (21/7)	20	Norethindrone acetate	1
10	Larin (21/7)	20	Norethindrone acetate	1
11	Junel (21/7)	20	Norethindrone acetate	1
12	Vyfemla (21/7)	35	Norethindrone	0.4
13	Aurovela (21/7)	20	Norethindrone	1
14	Loestrin (21/7)	20	Norethindrone	1

General Procedures

Subjects participated in venipuncture every other day for one CMOC pill cycle. If the subject was unable to attend a venipuncture appointment, it was performed a day earlier or later, if this was not possible the session was considered a missed visit and data were not available for that day. The record of each subject’s visits is summarized in Supplemental Tables S6–S9 (all Supplemental material are available at https://doi.org/10.17632/65jcdgrc7m.1). The starting date was randomly assigned across subjects. The venipuncture was conducted at the same time within the day to minimize the potential effects of diurnal fluctuation of endogenous hormone levels and the time difference between CMOC ingestion and blood draw (33, 39, 40). The subjects were asked to take their CMOC pill at the same time each day, based on their usual time of ingestion before joining the study. The average time between ingestion and venipuncture was 16.2 ± 5.2 h for 11 subjects; however, these data were not recorded for three subjects. A table of subject-specific average time to venipuncture can be found in Supplemental Table S5. Subjects were also instructed to maintain stable exposure to caffeine, alcohol, and exercise for at least 12 h before each testing session.

Determination of Endogenous Hormone Concentrations

Systemic estradiol (E2) and progesterone (P) concentrations were measured in serum. Venipuncture was performed on the antecubital area at each testing session using a 5 mL vacutainer serum separator tube (BD, Franklin Lakes, New Jersey). Once the tube was filled, the sample was mixed well, allowed to clot based on the manufacturer’s guidelines, and analyzed on the day of testing. The samples obtained at Northwestern University were processed and analyzed in the Shirley Ryan Ability Lab outpatient laboratory. The samples obtained at UT Southwestern Medical Center were sent to MedFusion (Lewisville, TX) for processing and analysis on the same day of testing. At MedFusion, the samples were centrifuged at room temperature at 4,000 RPM for 10 min and then tested for estradiol and progesterone using a competitive, enzyme immunoassay with no replicates.

Exogenous Hormone Concentration Measurement

Systemic concentrations of ethinyl estradiol (EE) and of a variety of progestins (SP) were measured in serum. A second 5 mL vacutainer serum separator tube (BD, Franklin Lakes, New Jersey) was collected and allowed to clot for 30 min at room temperature. Subsequently, the tube was centrifuged at 1,600 g for 10 min at room temperature, aliquoted, and stored at −80°C for later batched analysis (two batches) using liquid chromatography-mass spectrometry.

Liquid Chromatography-Mass Spectrometry Analysis

Standard curve generation.

Hormone standards included drospirenone, levonorgestrel, norgestimate, 19-norethindrone, norethindrone acetate, norelgestromin, etonogestrel, and ethinyl estradiol. The internal standards (ISs) were progesterone-2,2,4,6,6,17A,21,21-D9 and ethinyl estradiol-2,4,16,16-d4. Commercially available human plasma (BioIVT, Westbury, NY, K2EDTA, pooled female, 0.2 µm, filtered, 3X charcoal stripped from at least five females verbally screened to be medication-free at random points in their menstrual cycle) was used for the preparation of standard curves. To identify transitions, an individual solution was made for each of the compounds in methanol and infused into the mass spectrometer. Mass spectrum was acquired for each compound, daughter ions were chosen, and voltages were optimized using automated compound optimization feature of the acquisition software (Analyst 1.7.2). Results were reviewed, optimized, and a liquid chromatography method was developed as described, which allowed for resolution of each analyte of interest. Supplemental Table S1 contains pertinent information about the materials used for the standard curve generation.

Liquid chromatography-mass spectrometry.

Compounds were subsequently detected with the mass spectrometer in positive multiple reaction monitoring mode (MRM). The following transitions were monitored for each hormone: drospirenone 366.9/348.7; levonorgestrel 313.0/244.9; norgestimate 370.3/310.0; 19-norethindrone 299.0/230.9; norethindrone acetate 341.0/281.0; norelgestromin 328.0/260.0; etonogestrel 324.8/269.9; progesterone-d9 324.1/100.0; ethinyl estradiol (dansyl) 530.4/171.2; and ethinyl estradiol-d4 (dansyl) 534.2/171.2. An ACE (Advanced Chromatography Technologies, Aberdeen, UK) Excel C18-PFP column (1.7 µm, 10 cm, 2.1 mm) was used for all hormones except ethinyl estradiol and the corresponding IS with the following conditions: Buffer A: dH₂O containing 0.1% formic acid and 2 mM NH4 acetate; Buffer B: methanol containing 0.1% formic acid and 2 mM NH4 acetate, 0–1.0 min 70% B, 1.0–2.0 min gradient to 85% B, 2.0–4.0 gradient to 100% B, 4.0–6.0 min 100% B, 6.0–6.1 min gradient to 70% B, 6.1–.0 min 70% B; flow rate 0.4 mL/min. Ethinyl estradiol (EE) and the corresponding IS ethinyl estradiol-d4 (EE-d4) were detected using a Phenomenex (Torrence, CA) Kinetex C18 column (2.6 µm, 150 × 4.6 mm) after derivatization with dansyl chloride with the following conditions: Buffer A: dH₂O with 0.1% formic acid; Buffer B: acetonitrile, 0 min 0% B, 0.0–5.0 min gradient to 100% B, 5.0–7.0 100% B, 7.0–7.1 min gradient to 60% B, 7.1–8.0 min 60% B. Non-EE standards were dissolved in DMSO, pooled, and serial dilutions were prepared at various concentrations. Charcoal-stripped female human plasma (0.1 mL) was spiked with 1 µL of each standard pool to create a standard curve. Each 0.1-mL standard and 0.1-mL patient sample were then mixed with a 2X volume of acetonitrile, vortexed for 15 s, and incubated for 10 min at RT with rocking. To each sample and standard, 0.6 mL of dH₂O containing 250 pg internal standard progesterone-d9 was added, which were vortexed briefly and then centrifuged for 5 min at 4°C at 16,100 g. The supernatant was then passed over a Phenomenex Strata C-18E column (0.1 mL, 50 mg) preconditioned with 0.3 mL methanol followed by 0.3 mL dH₂O. The column was washed with 0.8 mL 30% methanol and then samples and standards were eluted with 0.8-mL methanol. Samples were dried under medium heat in a Thermo Scientific Savant SpeedVac and resuspended in 0.05 mL 50:50 methanol:H₂O containing 0.1% formic acid and 2 mM NH4 acetate. Samples were centrifuged for 5 min at 4°C at 16,100 g to remove any particulate material and evaluated as described earlier in this section by LC-MS/MS.

Ethinyl estradiol derivatization.

Ethinyl estradiol required derivatization with dansyl chloride for detection by electrospray mass spectrometry (Supplemental Fig. S1). EE standards were dissolved in DMSO, and serial dilutions were prepared at various concentrations. Charcoal-stripped female human plasma (0.25 mL) was spiked with 2.5 µL of each standard to create a standard curve. Each 0.25 mL of standard and 0.25 mL patient sample were then spiked with 10 µL of EE-d4 in H₂O (10 pg/µL). Samples and standards were vortexed for 5 s and 1 mL hexane:ethyl acetate (3:1) was added, and the samples and standards were vortexed a second time for 15 s and then centrifuged at 16,100 g for 5 min at 4°C. The upper layer (0.9 mL) was transferred to a glass screw cap tube and dried under nitrogen at 37°C for 10 min. Samples and standards were resuspended in 200 µL of a 50:50 mixture of acetone:50 mM carbonate buffer, pH 10.5 containing 1 mg/mL dansyl chloride, and incubated for 10 min at 60°C. After the samples were cooled to RT, 1 mL of hexane:ethyl acetate (3:1) was added, and the samples were vortexed for 15 s and then centrifuged for 5 min at 812 g. The upper layer (0.9 mL) was transferred to a microcentrifuge tube and dried in a SpeedVac without heat. Samples and standards were resuspended in 0.05 mL of 50:50 dH₂O:acetonitrile containing 0.05% formic acid and analyzed as described above in Liquid chromatography-mass spectrometry section by LC-MS/MS.

Quantitation of samples.

The linearity and sensitivity of the assay are reported in the Supplemental Information. The samples were quantitated in comparison to standard curves consisting of 5 to 14 points with quality control samples evaluated for accuracy at up to four concentrations, each in duplicate within the relevant standard curve range. A limit of detection (LOD) was set at a value threefold above the blank charcoal-stripped plasma matrix, and the lower limit of quantitation (LLOQ) was set at the lowest point on the standard curve that, upon back-calculation, was within 20% of theoretical and was above the LOD. The upper limit of quantitation (ULOQ) was set by the highest standard which quantitated to within 15% of theoretical. Samples were analyzed in two sets. Detailed information pertaining to the linearity and sensitivity, precision and accuracy, and the matrix effects and recovery of the assay can be found in Supplementary Tables S2–S5.

Data Analysis

To determine the variability of hormone concentrations, discrete time points were selected to compare across. These time points were days 1–2 (active), days 6–7 (active), days 13–14 (active), days 20–21 (active), days 22–23 (inactive/placebo), and days 27–28 (inactive/placebo) and coincide with the start and end of each week of the active pill phase and the beginning and end of the placebo phase. Furthermore, to control for differences in drug metabolism between subjects, hormone concentrations were normalized for each individual. For each subject, each day’s hormone concentrations were divided by the maximum concentration measured over the treatment cycle of the hormone of interest. Absolute concentrations of endogenous and exogenous hormones were also reported. Summary statistics were generated and include the absolute mean concentration, standard deviation, minimum and maximum concentrations, and number of subjects measured on each day. An additional sub-analysis was conducted to characterize hormone concentration profiles based on exogenous hormone type and dose. Those data were fitted with a fifth degree polynomial function with 95% confidence intervals. Concentrations that were below the limit of detection (LOD) or did not have an identifiable peak were considered as zero concentration. Those samples with a concentration below the lower limit of quantification (LLOQ) but above the LOD were considered as half of the LLOQ concentration value, a record of which samples can be found in Supplemental Tables S6–S9. All data fitting, statistical analysis, and figure generation were done with GraphPad Prism (V. 10.0.2). The normalized concentrations of E2, EE, P, and SP at days 1–2, 6–7, 13–14, 20–21, 22–23, and 27–28 underwent analysis for variations using a one-way analysis of variance (ANOVA) test and subsequent Tukey’s post hoc test. These analyses aimed to elucidate fluctuations in hormone concentrations within each week of the treatment cycle. Exact P values and the F value [F (DFn, DFd)] for ANOVA tests are reported. In addition, we used a fifth degree polynomial regression model to depict the mean absolute endogenous and exogenous hormone profiles, supplemented by 95% confidence intervals. Outliers were identified and removed using the ROUT method and excluded from analysis. Outliers are reported in Supplemental Tables S6–S9.

RESULTS

The number of subjects and samples in each time point can be found in Tables 2 and 3. A more detailed accounting of subject visits and samples can be found in Supplemental Tables S6–S9.

Table 2.

Summary data of estradiol and progesterone concentrations during the pill treatment cycle

	Estradiol					Progesterone
Day	Avg., pg/mL	Stdev	Min. Conc., pg/mL	Max. Conc., pg/mL	n	Avg., ng/mL	Stdev	Min. Conc., ng/mL	Max. Conc., ng/mL	n
1	30.22	13.98	12.00	59.00	9	0.71	1.02	0.10	2.88	9
2	59.00	67.59	22.00	178.00	5	0.67	0.48	0.03	1.34	5
3	23.00	13.04	10.00	50.00	8	0.64	0.68	0.10	2.19	8
4	38.25	21.36	16.00	58.00	4	1.07	1.17	0.10	2.66	4
5	21.00	9.26	10.00	39.00	9	0.39	0.32	0.10	0.90	8
6	18.00	1.41	17.00	20.00	4	1.03	1.06	0.10	2.52	4
7	23.33	13.86	10.00	52.00	9	0.68	0.63	0.20	2.17	9
8	16.50	4.95	13.00	20.00	2	0.58	0.29	0.37	0.78	2
9	18.70	6.36	10.00	29.00	10	0.58	0.66	0.10	2.25	10
10	17.00	4.76	10.00	20.00	4	0.37	0.42	0.10	0.99	4
11	21.33	7.12	11.00	29.00	9	0.70	0.57	0.10	2.02	9
12	40.80	47.71	15.00	126.00	5	0.36	0.46	0.10	1.04	4
13	20.00	7.93	12.00	31.00	8	0.74	0.74	0.10	2.44	8
14	29.80	26.00	13.00	76.00	5	1.06	1.54	0.10	3.73	5
15	20.29	9.25	10.00	35.00	7	0.80	0.73	0.10	2.38	7
16	20.50	7.55	10.00	26.00	4	0.48	0.37	0.10	0.91	4
17	16.89	4.68	10.00	23.00	9	0.74	0.90	0.10	2.87	8
18	17.40	5.59	10.00	23.00	5	0.35	0.23	0.10	0.59	4
19	16.89	4.62	10.00	22.00	9	0.65	0.66	0.10	2.30	10
20	20.80	6.83	13.00	30.00	5	0.25	0.11	0.10	0.34	4
21	20.00	4.57	14.00	28.00	8	0.61	0.73	0.10	2.43	9
22	17.00	11.37	10.00	34.00	4	0.35	0.22	0.10	0.53	3
23	32.00	27.81	10.00	92.00	8	0.76	0.71	0.10	2.19	7
24	23.50	8.66	12.00	33.00	4	0.34	0.27	0.00	0.62	5
25	29.11	10.93	16.00	44.00	9	0.58	0.62	0.10	2.14	9
26	54.50	33.52	20.00	88.00	4	0.27	0.28	0.00	0.62	4
27	38.78	21.95	14.00	71.00	9	0.55	0.59	0.10	1.96	9
28	62.33	31.77	43.00	99.00	3	0.50	0.49	0.01	0.99	3

Avg., average; Conc., concentration; Max., maximum; Min., minimum; n, number of subjects; Stdev, standard deviation.

Table 3.

Summary data of ethinyl estradiol and progestin concentrations during over-the-pill treatment cycle

	Ethinyl Estradiol					Progestin
Day	Avg., pg/mL	Stdev	Min. Conc., pg/mL	Max. Conc., pg/mL	n	Avg., ng/mL	Stdev	Min. Conc., ng/mL	Max. Conc., ng/mL	n
1	7.74	10.00	0.00	21.70	7	1.77	1.86	0.00	5.26	7
2	11.74	12.91	0.00	35.90	6	2.06	2.74	0.00	7.42	6
3	44.63	38.28	0.00	93.64	6	2.93	1.58	1.10	4.94	6
4	47.24	30.08	15.21	84.54	4	2.21	1.70	0.00	4.12	4
5	18.97	18.07	0.00	61.53	9	2.02	1.40	0.44	4.91	9
6	26.13	22.86	0.00	42.40	3	1.69	1.57	0.25	3.36	3
7	42.51	26.30	8.62	72.35	7	2.65	3.20	0.43	9.50	7
8	20.55	19.45	6.80	34.30	2	2.34	1.46	1.31	3.37	2
9	16.18	9.40	0.00	35.26	9	3.65	4.52	0.84	15.27	9
10	31.63	18.54	13.30	50.38	3	4.07	1.43	2.57	5.42	3
11	43.95	31.64	0.00	91.14	9	3.54	2.85	0.90	9.97	9
12	37.08	21.43	19.03	66.44	4	3.05	1.94	0.90	5.52	4
13	11.87	8.81	0.00	24.24	8	3.70	6.07	0.00	18.38	8
14	21.56	9.97	4.70	29.01	5	4.11	2.95	1.06	7.71	5
15	47.05	27.54	17.04	80.29	7	4.68	5.16	0.71	15.38	7
16	43.42	23.25	20.55	69.30	4	3.74	1.87	1.24	5.74	4
17	17.08	4.77	8.00	23.65	8	4.96	7.75	1.45	24.05	8
18	20.34	13.79	4.30	41.10	5	3.09	1.41	1.06	4.42	5
19	29.14	20.91	0.00	61.35	9	3.37	2.25	0.96	7.69	9
20	37.85	23.91	15.84	62.12	4	1.81	1.27	0.95	3.68	4
21	39.35	36.84	0.00	111.14	7	4.05	5.91	0.00	17.06	7
22	12.47	9.97	0.00	21.37	4	1.36	1.41	0.00	3.31	4
23	16.44	34.14	0.00	85.60	6	0.81	1.07	0.00	3.06	7
24	30.08	32.81	0.00	73.81	5	2.12	4.33	0.00	9.85	5
25	14.84	31.69	0.00	96.68	9	1.57	4.19	0.00	12.72	9
26	0.00	0.00	0.00	0.00	4	0.00	0.00	0.00	0.00	4
27	12.66	19.19	0.00	54.05	9	1.92	3.69	0.00	11.29	9
28	11.43	22.86	0.00	45.72	4	0.00	0.00	0.00	0.00	4

Avg., average; Conc., concentration; Max., maximum; Min., minimum; n, number of subjects; Stdev, standard deviation.

Concentrations of Serum Estradiol

Qualitative observations showed that E2 concentrations decreased from day 1 (30.22 ± 13.98 pg/mL) to day 7 (23.33 ± 13.86 pg/mL) of the active pill period and remained similarly low until day 21 (20.00 ± 5.57 pg/mL), then increased between day 22 (17.00 ± 11.37 pg/mL) until day 28 (62.33 ± 31.77 pg/mL) during the withdrawal week (Fig. 2). ANOVA testing revealed a significant difference [P = 0.002, F(5, 71) = 4.241] in normalized E2 concentration between the selected discrete time points. Normalized E2 concentrations were significantly higher during days 27–28 (withdrawal) compared with days 6–7 (P = 0.022), days 13–14 (P = 0.045), and days 20–21 (P = 0.012), and similar to days 1–2 (P = 0.137) and days 22–23 (withdrawal, P = 0.060, Fig. 1). The absolute concentrations for the aforementioned days can be found in Table 2. The full cycle, raw, concentration profiles for E2 can be found in Fig. 2A. The fitted E2 profile over the entire pill cycle is shown in Fig. 3.

Figure 2.

Absolute mean values of endogenous and exogenous hormones during the treatment course of monophasic oral contraceptives. Absolute concentrations for estradiol, progesterone, ethinyl estradiol, and progestin were included for this figure. These data are in tabular form in Tables 2 (endogenous hormones) and 3 (exogenous hormones) and include average, standard deviation, number of subjects, and the range of the concentrations for each time point. Error bars represent the standard error of the mean.

Figure 1.

Comparison of normalized endogenous and exogenous hormones concentrations at discrete time points. The normalized concentrations of E2, EE, P, and SP at days 1–2, 6–7, 13–14, 20–21, 22–23, and 27–28 underwent analysis for variations using a one-way analysis of variance (ANOVA) test and subsequent Tukey’s post hoc test. A: normalized estradiol concentrations were significantly higher during days 27–28 compared with days 6–7 (P = 0.022), days 13–14 (P = 0.045), and days 20–21 (P = 0.012), and similar to days 1–2 (P = 0.137) and days 22–23 (P = 0.060). B: normalized progesterone concentrations were similar at each time point. C: normalized ethinyl estradiol concentrations were significantly higher during days 20–21 of active pill period compared with days 1–2 (P = 0.006) of the active pill phase and days 22–23 (P = 0.016) and days 27–28 of the placebo phase (P = 0.003). Normalized Days 6–7 ethinyl estradiol levels were significantly higher than days 27–28 of the placebo phase (P = 0.042). D: normalized progestin concentrations were similar at each time point. Error bars represent the standard error of the mean. *P < 0.05 and **P < 0.01. n represents the number of subjects with data at each timepoint.

Figure 3.

The combined circulating endogenous and exogenous hormone profiles of all subjects. Subjects using a 21/7 combination monophasic oral contraceptive format were included in this study. Data were fit with a fifth degree polynomial with 95% confidence intervals. n represents the number of subjects contributing data to the fitted hormone profiles.

During the active pill period, E2 concentrations for all subjects ranged from 10 to 178 pg/mL and during the inactive week, E2 ranged from 10 to 99 pg/mL. Individual ranges of E2 concentrations during the active weeks are reported in Table 2.

Concentrations of Serum Ethinyl Estradiol

Qualitatively, EE concentrations increased from day 1 (7.74 ± 10.00 pg/mL) up to day 12 (37.08 ± 21.43 pg/mL) and sharply declined from day 21 (39.35 ± 36.84 pg/mL) until day 28 (11.43 ± 22.86 pg/mL) (Fig. 2). ANOVA testing revealed a significant difference [P = 0.005, F(5, 65) =5.117] in normalized EE concentrations over the pill cycle. Normalized ethinyl estradiol concentrations were significantly higher during days 20–21 of active pill period compared with days 1–2 (P = 0.006) of the active pill phase and days 22–23 (P = 0.016) and days 27–28 of the placebo phase (P = 0.003). Normalized EE levels on days 6–7 were significantly higher than days 27–28 of the placebo phase (P = 0.042). The absolute concentrations for the aforementioned days can be found in Table 3. The full cycle, raw, concentration profiles for EE can be found in Fig. 2C. The fitted profile for all days is shown in Fig. 3.

During active pill ingestion, EE ranged from 0 to 111.14 pg/mL and 0 to 96.68 pg/mL during the placebo week. Individual ranges of EE concentrations during the active weeks are reported in Table 3.

Concentrations of Serum Progesterone

Qualitative observations revealed that normalized P levels were low and had small variation compared with the other hormones during the active pill phase (day 1: 0.71 ± 1.02 ng/mL, day 7: 0.68 ± 0.63 ng/mL, day 14 1.06 ± 1.54 ng/mL) and did not increase over the withdrawal phase (day 22: 0.35 ± 0.22 ng/mL, day 28: 0.5 ± 0.49 ng/mL).

ANOVA testing revealed no statistically significant changes in normalized P concentrations [P = 0.33, F(5, 65)=1.187] at any of the predefined time points. The full cycle, raw, concentration profiles for P and SP are found in Fig. 2, B and D. The fitted profiles are shown in Fig. 3. During the active pills, the range of P for all subjects was 0.1 to 8 ng/mL. During the inactive week, the range of P for all subjects was 0.1 to 2.2 ng/mL. Individual ranges of P over the entire pill cycle are reported in Table 3.

Concentrations of Serum Progestins

Qualitatively, progestins increased from day 1 (1.77 ± 1.86 ng/mL) through day 9 (3.65 ± 1.86 ng/mL) and remained at similar levels above 3 ng/mL until day 21. Progestin concentrations dropped sharply during the placebo phase from day 22 (1.36 ± 1.41 ng/mL) through day 28 (0.00 ± 0.00 ng/mL).

ANOVA testing revealed no statistically significant changes in SP [P = 0.33, F(5, 65)=1.187] at the predefined discrete time points. The full cycle, raw, concentration profiles for SP are found in Fig. 2, B and D. The fitted profile is shown in Fig. 3. During active pill ingestion, the concentrations of SP for all subjects ranged from 0 to 28.9 ng/mL and 0 to 12.7 ng/mL during the inactive week.

Qualitative Subcharacterization of Type and Dose of Exogenous Hormones

Hormone profiles were broken down by EE dosage and progestin type and dose (when n > 3). Fitted hormone profiles of subjects using 20 μg (n = 9) and 30 μg doses (n = 4) of EE are found in Figs. 4A and 3B. Qualitatively, subjects using 20 μg of EE showed an increase in E2 after withdrawal with a simultaneous decrease of EE, while subjects using 30 μg of EE had more variable EE concentrations compared with the 20 μg subjects and showed a smaller increase of estradiol following withdrawal.

Figure 4.

Resulting circulating endogenous and exogenous hormone profiles based on oral contraceptive dosage. Subjects using a 21/7 oral contraceptive format and with at least three subjects with the same dosage were included. Data were fit with a fifth degree polynomial with the 95% confidence interval. A: subjects using 20 μg ethinyl estradiol showed an increase in estradiol after withdrawal with a simultaneous decrease of ethinlyestradiol. B: subjects using 30 μg of ethinyl estradiol had more variable ethinyl estradiol concentrations compared with the 20 μg and showed a smaller increase of estradiol following withdrawal. C: subjects using levonorgestrel (n = 2, 0.15 mg) showed a decrease of progesterone in the early active pill phase and a decrease in levonorgestrel following the last ingestion of the active pill. D: subjects using 1 mg of norethindrone/acetate showed a decrease in norethindrone/acetate concentrations following withdrawal. n represents the number of subjects contributing data to the fitted hormone profiles.

Subsequently, the hormone profiles of a subset of subjects using 0.1–0.15 mg of levonorgestrel (n = 3) and 1 mg of norethindrone/acetate (n = 8) were assessed. Subjects using levonorgestrel (n = 2, 0.15 mg, n = 1, 0.1 mg) showed a decrease of progesterone in the early active pill phase and a decrease in levonorgestrel following the last ingestion of the active pill (Fig. 4C). After qualitative assessment, these norethindrone/acetate subjects showed a decrease in serum concentrations following withdrawal (Fig. 4D).

DISCUSSION

In our manuscript, we present the endogenous and exogenous hormone profiles throughout the oral contraceptive pill cycle of 14 women taking various 28-day CMOCs. Our analysis revealed significantly greater levels of EE at days 20–21 of active pill ingestion compared with days 1–2 and days 27–28. Conversely, E2 concentrations decreased during the active pill phase, while SP and P levels remained low and stable. During the 7 days of inactive pills, E2 levels rose sharply, while EE declined and SP levels remained low and stable. To our knowledge, ours is the first study to investigate both endogenous and exogenous hormone concentrations, at such a high resolution women throughout the CMOC pill cycle in women. Our findings support previous literature that show E2 rises during the placebo/pill-free phase (41, 42), EE increases from day 1 to 21 (32, 42), and SP increases from day 1 to 8 and maintains over day 21, although not the same magnitude (32, 33). However, our data did not support the reported rise in progesterone over the placebo time period reported (41, 42).

Since more than 150 million women worldwide are taking some form of oral contraceptive pill, this new knowledge of hormonal variability may provide an important context for understanding the secondary effects of CMOCs as well as inform methodology and data interpretation in studies including women on CMOCs.

A limited number of prior studies have investigated the influence of CMOC’s on endogenous concentrations of E2 and Elliot et al. (43) measured serum concentrations of E2 and P on days 5 and 14 of the active pills and day 5 of the inactive pills and found no significant changes in the concentrations of either hormone during one OC cycle. Our study included more frequent, every-other-day sampling, which enabled us to observe the decline of E2 during the active pills and rise during the inactive pills. Elliot-Sale et al. (23) conducted another investigation of a one-time measurement of serum E2 and P in women taking CMOCs, phasic combined, and progestin-only pills. They found that the type and composition of medication influenced the mean hormone concentrations, with wide variations observed when different brands were analyzed as a group. In terms of exogenous hormone concentrations, Morris et al. (24) measured serum EE and levonorgestrel every 4 days in 10 women who were exposed to 30 µg EE and 150 µg levonorgestrel. They found that it took 3–4 days for the exogenous hormones to reach maximal values and that EE ranged from 1 to 3 nmol/L and levonorgestrel from 6 to 24 nmol/L (24). Krattenmacher’s (44) work also demonstrated a ramp-up period of 7 days to reach the maximal concentration of drospirenone and Sarkar et al. (25) found that serum concentrations of norethindrone increased over 21 days from approximately 1 ng/mL to 3.5 ng/mL. Further comparisons between our data and the literature (27, 28, 45, 46) can be found in Table 4. Collectively, these results suggest that there is variability in concentrations of E2 and EE throughout the OC pill cycle and that both OC type and composition might contribute to this variability.

Table 4.

A comparison of the concentrations found in the current study compared with package inserts provided with oral contraceptives

	Cycle_Day	1_1	1_6	1_21	3_21	6_21	9_21	13_21
Aviane (45)	EE (0.02 mg) average Cmax, pg/mL	62 ± 20.5	76.7 ± 29.9	82.3 ± 33.2
Zarah (46)	EE (0.03 mg) average Cmax, pg/mL	53.5 ± 23		92.1 ± 32.2		99.1 ± 44.6	87 ± 37.4	90.5 ± 40.7
Current study	EE (0.02 mg) average Cmax, pg/mL	59.6 ± 23.1 (n = 9)
	EE (0.03 mg) average Cmax, pg/mL	49.7 ± 36.0 (n = 4)
	EE (0.035 mg) Cmax, pg/mL	50.38 (n = 1)
Aviane (45)	Levonorgestrel (0.1 mg) average Cmax, ng/mL	2.75 ± 0.88	4.52 ± 4.52	6 ± 2.65
Kuhnz et al. (28)	Levonorgestrel (0.15 mg) average concentration, ng/mL			1.4 ± 0.5			1.7 ± 0.4
Zarah (46)	Drospirenone (3 mg) average Cmax, ng/mL	36.9 ± 4.8		87.5 ± 51.6		84.2 ± 16.0	81.3 ± 15.4	78.7 ± 14.2
Back et al. (27)	Norethindrone/NA concentration, ng/mL	4.7 ± 1
Current study	Drospirenone (3 mg) Cmax, ng/mL	28.8 (n = 1)
	Levonorgestrel average (0.1–1.5 mg) Cmax, ng/mL	7.3 ± 7.7 (n = 3)
	Norethindrone/NA (0.4–1.5 mg) Cmax, ng/mL	7.9 ± 7.6 (n = 10)

Data are presented as means ± SD. Cycle_Day corresponds to the time point of interest where the first number is the current treatment cycle and the second number is the day of the treatment cycle. Cmax, maximum concentration; EE, ethinyl estradiol; NA, norethindrone acetate.

In the past few years, there has been considerable interest in the secondary effects of oral contraceptives on non-reproductive tissues (e.g., bone and ligament) and sports/exercise performance (10, 47). A better understanding of the MSK effects of oral contraceptives is paramount, as 66% of female collegiate athletes regularly take hormonal contraceptives, and the majority of these are combination oral contraceptives (48). Despite growing interest in this area of research, much controversy still exists around how exogenous hormones found in oral contraceptives influence tissues, injury, and performance (10, 47, 49–64) (Table 5). A recent systematic review and meta-analysis of the effects of oral contraceptives on performance demonstrated that exercise performance may be slightly reduced compared with eumenorrheic women, yet significant variability was seen limiting the clinical relevance and ability to make group generalizations (47). The failure to account for differences in OC type and composition likely explains some of the lack of conclusive findings, as most studies group women on various types of oral contraceptives.

Table 5.

Summary of off-target effects of oral contraceptives

Category	Oral Contraceptive Effect	References
Skeletal muscle	– Strength	10, 47
	↓Protein turnover	49
	↑Myo-proliferation	50, 51
	↓Muscle injury	64
Tendon/ligament	– ACL laxity	56
	↑ACL strength	53
	– ACL injury	54–56
	↑Achilles tendinopathy risk	57
Bone	↓Bone mass/mineral density	59–61
	↑Fracture risk	62
Performance	↓Aerobic capacity/VO2	63
	↓General performance	47

Where ↑ indicates an increase, ↓ indicates a decrease, and – indicates mixed findings in the literature surveyed.

Furthermore, not standardizing the day of testing adds another dimension of uncertainty when interpreting data. In studies we surveyed (43, 51, 59, 65–76), the most common testing dates included day 14 of active pill ingestion (43, 68, 71, 72, 74, 76), days 18–21 of active pill ingestion (65–67, 70, 76), and the second and third day (days 23–24) of the inactive/placebo week (51, 65, 68–70, 72, 75). Time points that were less common included days 2–13 (43, 68, 72, 74), 15–17 (65, 68, 70, 71), 25–28 (placebo) (43, 65, 68–70, 72, 74). Interestingly, studies that were concerned with the effects of chronic exposure over a multi-month training programs do not report the day of the treatment cycle when outcomes were measured, possibly introducing unexpected variability due to acute effects of the hormone levels at the time of testing (59, 73). Our findings suggest that differences in EE and E2 concentrations across the CMOC pill cycle might also be a factor that clouds data interpretation.

Our findings raise important considerations regarding methodological approaches when studying women exposed to CMOCs. Women taking CMOCs have often been utilized as a control group because it was presumed that they demonstrate stable, low levels of sex hormones (9). However, our data show that there is variability in endogenous E2 as well as exogenous EE, so women tested during the first few days of the CMOC cycle will have a lower EE, and potentially SP, and higher E2 than women tested in the last few days of the active pills. The fluctuating levels of endogenous and exogenous hormones across the CMOC pill cycle introduce a source of variability that can potentially mask any discernible changes between experimental groups. Elliot-Sale et al. (9) recently published a working guide for experimental design when including women as participants in sport and exercise science studies. The recommendations for studying women exposed to hormonal contraceptives include documentation of the phase type (mono vs. tri), the dose of EE and the name and dose of progestin, and only using one brand/type per group of participants to limit variation in progestin effects (9). Our data suggest that it would be ideal to measure serum concentrations of EE and SP to yield a complete understanding of the potential impact of both acute and chronic exposure to CMOCs on the measured outcomes in female sex hormone-related human studies.

Measuring exogenous hormones with LC-MS/MS could be cost, equipment, or technically prohibitive depending on the research group. At present, large clinical analytical laboratories do not offer exogenous steroid panels, further hindering researchers’ ability to characterize EE and SP levels in their subjects. Immunoassays have been the “gold standard” for measuring biomolecules, including exogenous hormones, due to their ease of use, scientific acceptance, and relatively low equipment costs (30). However, immunoassays do suffer from inherent limitations such as issues with antibody availability/selectivity, large sample volumes, limited multiplexing, long assay run time, and reproducibility (30). Although LC-MS/MS has high equipment costs and technical complexity, these barriers are offset by its strengths, including: high selectivity, sensitivity, and throughput, low sample volumes and cost per sample, high reproducibility, and its multiplexing capability (thousands of analytes from one sample) (30). To alleviate the high equipment costs and technical complexity of LC-MS/MS from researchers, we hope that clinical laboratories and university core facilities develop similar assays and offer their service to interested research groups.

If measuring exogenous hormones is not feasible for a research group, our research supports the notion recommended in Elliot-Sale 2021 (9) that researchers should aim to test in a narrow window of comparable times during the treatment cycle. However, the day/days of testing chosen may depend on the research question being asked. For example, studies investigating the effects of chronic exposure (multiple pill cycles) to exogenous hormones should measure outcomes between days 22 and 24 (inactive/placebo) as to minimize the concentrations of EE, E2, SP, and P, and decrease the likelihood that differences in the outcomes are due to acute effects from either endogenous or exogenous hormone levels. A second example would be a study interested in the acute effects of EE exposure over one treatment cycle. Based on our data, researchers should test subjects between days 19 and 21 (active pill ingestion) to maximize EE levels while SP, P, and E2 are low. However, researchers should also have the same outcomes tested between days 22 and 24 (placebo/inactive) to minimize all hormone levels (a within-subject control).

The assay we used to determine exogenous hormone concentrations could be used in future studies investigating personalized prescription of oral contraceptives to achieve the desired effect (e.g., birth control) while reducing unwanted secondary effects or adverse events. For example, physicians consider oral contraceptive type and composition when there are concerns for particular side effects such as acne, facial hair, and deep vein thrombosis (17, 77). However, the systemic concentrations of EE and SP, not just the hormone dose, may be a factor in the development of side effects and could be included in the decision-making algorithm for medication choice. Exogenous EE and SP are used in other circumstances beyond preventing pregnancy. For example, exogenous progestins can be used to treat neurological disorders, including seizures, multiple sclerosis, and spinal cord injury (78). Exogenous progestins can reduce seizures, but the dosing can be challenging since antiseizure medications can influence the metabolism of OCs in women taking them simultaneously (79). At present, the dosing of exogenous hormones in women with seizures is based on the type and composition of oral contraceptive, but titration based on systemic concentration would offer an additional and possibly more precise method of dosing to offer optimal seizure management. Finally, the assays we used to determine systemic concentrations of exogenous hormones may be useful in titrating the hormone dose in hormone replacement therapy prescription for postmenopausal women. Hormone replacement therapy prescription requires consideration of the benefits and risks, including risk of cardiovascular disease and cancers (56, 58, 61). Factors such as age, prior hysterectomy, oophorectomy, and hormonal combinations influence risk of serious adverse events, and knowledge of systemic concentrations of exogenous hormones might provide additional information for determining how to titrate and time hormone replacement therapy prescription.

The limitations of this study include a small sample size and the inclusion of only one CMOC pill cycle in the analysis. In addition, although we asked subjects to take their CMOC pills at the same time each day, the timing of pill consumption may have been inconsistent in relation to the blood sample collection. Finally, our subjects were taking multiple types of CMOCs and although were within a narrow dosage range of EE and progestin, the progestins were from different generations (first, second, and fourth) which may have contributed to the variability seen in our data.

Perspectives and Significance

Our study presents a method of measuring serum concentrations of both endogenous and exogenous sex hormones in women taking oral contraceptives. The variability in estradiol and ethinyl estradiol across the CMOC pill cycle was greater than previously reported, mainly because very few studies have measured exogenous hormone concentrations in women taking CMOCs. The hormonal variability in women taking CMOCs challenges the belief that this group can serve as a control in studies investigating the influence of hormonal fluctuations across the menstrual cycle on a particular outcome. Our findings underscore the importance of measuring and accounting for hormonal changes in women across the menstrual and oral contraceptive pill cycle and should inform future research design and data interpretation. When including women taking oral contraceptives in research studies where sex hormone exposure is being testing, it would be ideal to consider hormone compositions (hormone dose and type) (9), systemic hormonal concentrations, and timing of the contraceptive pill cycle on the outcome (9). If measurement of systemic concentrations is not possible, then women should be tested in consistent and narrow windows of the active pill cycle to attempt to reduce as much variability as possible in systemic concentrations (9). Future work could include characterizing both endogenous and exogenous hormones in users of hormonal contraceptives [in isolation and in accordance with Elliot-Sale 2021 recommendations (9)] such as bi-, tri-, and four-phasic oral contraceptives, progestin-only pills, vaginal rings, intrauterine devices, and injections. This would provide data for an experimental design framework in these populations, adding to the current methodological guidance (9).

DATA AVAILABILITY

The data that support the findings of this study are available from the corresponding author, Dr. Yasin Y. Dhaher, upon reasonable request.

SUPPLEMENTAL DATA

Supplemental Tables S1–S9 and Supplemental Fig. S1: https://doi.org/10.17632/65jcdgrc7m.1.

GRANTS

This work was supported by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (1R01AR069176-01A1 and 1R01AR069176-03S1).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

E. Casey and Y.Y.D. conceived and designed research; E. Crossley and N.W. performed experiments; L.A.R., E. Crossley, and N.W. analyzed data; L.A.R., E. Casey, and Y.Y.D. interpreted results of experiments; L.A.R., E. Crossley, and N.W. prepared figures; L.A.R., E. Casey, E. Crossley, and N.W. drafted manuscript; L.A.R., E. Casey, E. Crossley, N.W., and Y.Y.D. edited and revised manuscript; L.A.R., E. Casey, E. Crossley, N.W., and Y.Y.D. approved final version of manuscript.