Beyond Birth Control: How Oral Contraceptives Impact Skeletal Muscle Health - a Systematic Review and Meta-Analysis

Abstract

Oral contraceptive pills (OCPs) are one of the most prescribed medications, yet we lack an understanding of if and how OCPs affect non-reproductive tissues. Given the well-documented effects of sex hormones on skeletal muscle, including on muscle mass, regeneration, and recovery, our review was aimed at assessing the impact of OCPs on skeletal muscle physiology in female humans and animals. We performed a literature search, title through full text screening, and citation search in accordance with PROSPERO guidelines. Rigor and reproducibility were assessed using the Modified Downs and Black Checklist. Meta-analyses were performed to assess the impact of OCPs on skeletal muscle outcomes. Although our search included both clinical and pre-clinical studies, the forty included studies were all clinical with no identified preclinical studies. Studies focused on young (20–30 y.o.) sedentary to active females with a healthy BMI (18–27 kg/m²) and included primarily strength and serum-based outcomes. All studies were retrospective and level III evidence. Notably, despite this literature spanning from the 1990’s to 2025, rigor was in the 69^th ± 6.5 percentile, and there was no correlation with rigor and year of publication. Meta-analyses did not detect an effect of OCPs on examined outcome measures; however, heterogeneity was high suggesting the lack of rejection of the null hypothesis may be driven by variations in studies, making it challenging to draw conclusions. Taken together, we recommend prospective preclinical and well-controlled clinical studies to examine the impact of OCPs on skeletal muscle in the setting of injury, disease, and varying demographics.

Graphical Abstract

Introduction

As a scientific and medical community, we have a blind spot with regards to how oral contraceptive pills (OCPs), which is the number one medication prescribed to non-pregnant women aged 15–44 in the United States,(1) affect non-reproductive tissues. The US Food and Drug Administration approved the first OCP, Enovid, in 1960,(2) and since then, over 100 different formulations have been developed and approved.(3) Much of this innovation has stemmed from addressing adverse effects with lesser attention given to additional outcomes or health effects.(4) For instance, first generation OCPs had high estrogen concentrations, which resulted in nausea, breast tenderness, and increased risk for venous thromboembolism (VTE).(5) Second generation OCPs lowered the estrogen concentration and added a progesterone compound, thereby addressing these side effects but worsened androgenic side effects, like acne and weight gain.(6) Third and fourth generation OCPs have fewer metabolic (e.g., not associated with weight gain) and mood symptoms (e.g., not associated with depression) but still carry an increased risk of VTEs, albeit not as high as first generation OCPs.(3) Thus, despite the many successes and gains provided by OCPs, there remain areas of optimization and improvement when considering the systemic impact of OCPs.

Another challenge that presents with current OCP usage is the fact that very few studies specifically focus on OCP effects on non-reproductive organ systems, outside of emergent adverse effects, including skeletal muscle. Healthy skeletal muscle is important throughout the female lifespan and contributes to maintenance of bone mineral density, metabolism, and independence. Sarcopenia is one of the most common contributors to loss of independence and poor quality of life and is the result of the steady decline in muscle mass over time, with women disproportionately affected by higher incidence of sarcopenia than men (10). Ample studies have demonstrated that sex hormones impact skeletal muscle, including the promotion and maintenance of muscle mass,(7–10) and it stands to reason that a medication that alters said sex hormones may impact these tissues. Indeed, several studies have documented OCP-related alterations in skeletal muscle health, performance, and related injury risks.(11, 12) For example, muscle size and type I fiber proportionality increase with OCP use,(11) suggesting OCP use may positively impact skeletal muscle. One important question that has arisen is, do OCPs impact muscular strength, performance, and power? Understanding this question and the relationship between OCPs and skeletal muscle could not only be informative to current OCP users on potential benefits and risks but could also expand the usage of OCPs to post-partum individuals returning to activity, cis-gender women recovering from muscular injuries, as well as support further research into the effects of hormone replacement therapy in perimenopausal and postmenopausal individuals experiencing sarcopenia. Therefore, the purpose of this systematic review and meta-analysis was to compare study characteristics, analyze rigor and reproducibility, and assess the effect of OCPs on skeletal muscle strength, performance, and hypertrophy in females. Our aim is due in part to expand on Nolan et al.’s review by using a broader inclusion criterion of outcome variables and study designs in addition to addressing rigor and reproducibility concerns in the scientific community at large. Specifically, we hypothesized that rigor and reproducibility would improve with time given the publication of well-established clinical and preclinical guidelines. Meta-analyses were performed on OCP effects on knee extension strength, knee flexion strength, skeletal muscle cross-sectional area (CSA), jump height, and grip strength.

Methods

Systematic Review

A systematic review of the literature aimed at understanding the effects of OCPs on skeletal muscle physiology was performed following the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) protocol (Appendix S1),(13) Meta-Analysis of Observational Studies in Epidemiology (MOOSE) checklist (Appendix S2),(14) and Cochrane Handbook for Systematic Reviews of Interventions.(15) This systematic review was not registered a priori on PROSPERO.(16) We identified eligible manuscripts using the following PECO:(17) Female participants (human or animal model) (Population), OCP use (no limitation on duration of treatment or type of OCP) (Exposure), no OCP use (relative age matched control (e.g., for animal studies, exact age-matched and for humans, a non-statistical significant difference in age)) (Comparison), metrics of skeletal muscle physiology (Outcome, i.e., any metric where the tissue being studied was skeletal muscle)(18, 19). Studies were excluded if they were written in a non-English language, were not peer reviewed, used non-OCP concentrations of hormone (e.g., the study was focused on hormone replacement in the setting of menopause), were not original research articles, and if there was not an appropriate control group. We note that throughout this review, we use the term “female” to refer to humans and animals assigned female at birth and as defined by each individual study’s inclusion criteria.

On October 28, 2025, we performed a literature search on PubMed, Physiotherapy Evidence Database (PEDro), Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Cochrane Central Register of Controlled Trials. At least two independent authors (AH, GG, AS, KK, SB, CKK) screened the titles, abstracts, and full texts of studies identified via our search based on the aforementioned PECO and Cochran Handbook recommendations.(15) Specific search terms are listed in Supplementary Table 1. Following full text screening, a citation search was performed via PubMed. Screening was completed in pre-designed spreadsheets, and disagreements between authors were adjudicated by the opinion of a third author.

Four authors (AH, GG, SB, SOE) collected study demographic information from included articles, such as location of study, age range of participants, inclusion and exclusion criteria, if any comorbid reproductive disorders were present in the study sample, reporting of why the participant were on OCPs in the case of a non-randomized control trial, baseline activity level, objective baseline fitness metrics (e.g., VO₂ max), OCP type, time in menstrual cycle measurements were taken, outcome variables, if a muscle biopsy was collected, if participants were injured or uninjured, if participants had previous pregnancies, how long participants have been on OCPs, race, BMI, markers of socioeconomic status, sample size, study type (e.g., cohort, case control), and level of evidence.(20) Metrics were categorized as molecular markers, performance, histology, serum marker, and strength. Of note, we defined “molecular markers” as any molecules measured from a muscle biopsy or muscle sample directly, “performance” as a non-strength based metric of muscle force (e.g., jump height), “histology” as any histological evaluation of the muscle itself, “serum markers” as any molecule measured via serum collection from participants or animals, and “strength-based metrics” as a quantitative or semi-quantitative measure used to assess the maximum voluntary force or torque of a muscle or muscle group. One author (GG) extracted data for use in meta-analyses. When primary numerical data was not available, graphical data was transcribed to numerical data using a digital ruler that has been previously utilized and validated.(21) Intra-rater reliability for graphical to numerical data conversion was previously reported by this author as ‘excellent’ (GG): (n = 5, intraclass correlation coefficient: 0.9752, 95% CI: [0.9413, 1.0000]).(22)

Rigor and reproducibility assessment

Three independent authors (KK, CKK, SB) assessed the rigor and reproducibility of the included clinical studies using the Modified Downs and Black Checklist (MDBC) (23) and included animal studies using the ARRIVE Guidelines.(24) The categories were ranked as either “clearly insufficient” (0), “unclear if sufficient” (1), or “clearly sufficient” (2). A maximum score of 28 for MDBC or 42 for ARRIVE indicates an extremely rigorous study, while a 0 indicates no rigor, with scores between the authors being averaged. Spearman’s correlation analysis between scores with year of article publication was calculated via SPSS Statistics for Windows, version 31 (IBM Corp., Armonk, NY, USA).

Meta-analyses

We performed meta-analyses on the effects of OCPs on knee extension strength, knee flexion strength, skeletal muscle cross-sectional area (CSA), jump height, and grip strength. Of note, we were unable to perform meta-analyses on serum markers or molecular markers due to too few studies having overlapping outcome variables, and these results were summarized qualitatively. We calculated the standardized mean differences (SMD) and pooled standard deviations (SD_pool) of these variables from included studies via the DerSimonian-Laird method. Cochran's Q test was utilized to assess heterogeneity, and statistical significance was defined as an alpha level of 0.05. Meta-analyses were also performed via SPSS Statistics for Windows, version 31 (IBM Corp., Armonk, NY, USA). Figures were generated in GraphPad Prism V10 and Microsoft PowerPoint.

Results

This body of literature examines a homogeneous population, with heterogeneous study designs

Our literature search identified 459 articles to be screened for our review. After title, abstract, and full text screening as well as a citation search, 40 articles were eligible for inclusion (Figure 1).(25–65) The reason for exclusion was documented at the full text level in Supplementary Table 2. Demographic information from included studies is listed in Supplementary Table 3 and summarized in Table 1.

Figure 1: Systematic review workflow.

Systematic search of the literature, followed by title, abstract, and full text screening as well as a citation search yielded 40 articles to be included in our review. Reason for exclusion at the full text level are shown in Table S2, and details about included articles are listed in Table S3.

Table 1:

Summary of study demographic information. Full details are listed in Table S3.

PMID	Location of study	Age Range (years)	Comorbid Conditions	Baseline Activity Level	OCP type	Phase of menstrual cycle	Sample Size	Study Type
8735711	UK	20.7 ± 1.4	Not reported	Sedentary	2nd or greater	Early Follicular, Mid cycle, Late Follicular, Mid Luteal, Late Luteal	10	Cross sectional
8910237	UK	17–39	Not reported	Active	Monophasic	Not taken into account	27	Case control
9315413	US	23.5 ± 4.96	None reported	Active	Mixed	Mid Luteal phase	7	Cohort
11680291	US	23.8 ± 4	Not reported	Not reported	Mixed	Late Follicular	17	Cross sectional
12169383	US	22.3 ± 1.2	Not reported	Sedentary	Monophasic	Late Follicular	14	Cohort
12706609	US	24 ± 1	Sedentary	Not reported	Not reported	Early Follicular, Late Follicular	7	Cross sectional
15618333	UK	22 ± 4	Sedentary	Sedentary	Monophasic	Mid-Luteal & Early Follicular	14	Cross sectional
18845777	Denmark	24 ± 4	None	Active	Mixed	Early Follicular	12	Cohort
19883384	Denmark	24 ± 4	None	Active	Mixed	Early Follicular	12	Cohort
19887010	Morocco	54.3 ± 11.2	Not reported	Not reported	2nd or greater	Not taken into account	410	Cross sectional
22008484	US	20.4 ± 2	Not reported	Active	Monophasic	Late Follicular	15	Cross sectional
22948447	Sweden	24.4 ± 4.8	None reported	Active	Mixed	Early Follicular phase, Ovulation phase, Luteal phase	9	Cross sectional
23445919	UK	20–25	None	Active	Monophasic	Not taken into account	24	Cohort
24091990	US	18–35	None	Active	Not reported	Not taken into account	19	Case control
24240566	US	18–30	Not reported	Sedentary	Mixed	Not taken into account	19	Cohort
24504652	Australia	21.6 ± 3	Not reported	Active	2nd or greater	Early Follicular	9	Cohort
25694209	Australia	22 ± 3	Not reported	Active	2nd or greater	Early Follicular	8	Cohort
28388333	US	21–30	Not reported	Sedentary	Not reported	Mid-Luteal	9	Cohort
28497386	UK	23.4 ± 2.4	Not reported	Active	Monophasic	Late Follicular	18	Cohort
30502591	Finland	20–40	None	Sedentary	Mixed	Not taken into account	400	Cross sectional
31312144	Denmark	23–25	Not reported	Sedentary	Mixed	Not taken into account	14	Case control
31526145	Germany	27.8 ± 6	None	Not reported	2nd or greater	Not taken into account	26	Cohort
31686212	Chile	26 ± 3.59	Not reported	Sedentary	2nd or greater	Late Follicular	30	Cross sectional
32595523	Australia	25.5 ± 6	Not reported	Active	Monophasic	Not taken into account	25	Cross sectional
32694286	Denmark	18–30	None	Sedentary	2nd or greater	Not taken into account	18	Cohort
33054662	Denmark	24 ± 3	None	Sedentary	2nd or greater	Not taken into account	20	Cohort
33692699	Norway	21.2 ± 3.3	None	athletes	Mixed	Not taken into account	29	Cohort
33993156	US	18–29	None	Sedentary	Mixed	Not taken into account	38	Cohort
34276392	Austria	23.3 ± 3.1	None	athletes	Monophasic	Late Follicular	16	Cross sectional
34682310	Australia	18–30	Not reported	Active	Monophasic	Early Follicular, Late Follicular, Mid-Luteal	30	Case control
35538569	Germany	25 ± 4.56	None (specifically endocrine disorders)	Sedentary	2nd or greater	Early Follicular	40	Cohort
36011425	Germany	20–30	None	Sedentary	Monophasic	Not taken into account	58	Case control
37453973	Germany	20–28	None	Active	Monophasic	Not taken into account	24	Case control
38356164	Denmark	18–30	None	Sedentary	2nd or greater	Late Follicular	20	Cohort
39052822	Finland	18–40	None reported	Active	2nd or greater	Not taken into account	21	Cohort
39964410	Finland	22–30	None	Active	Monophasic	Early Follicular, Mid Luteal, Late Follicular	21	Case control
40088271	Norway	18–40	None	Sedentary	2nd or greater	Not taken into account	39	Case control
40514618	Turkey	20.5 ± 5	PCOS	Not reported	Mixed	Early Follicular	20	Case control
40219704	Norway	18–40	healthy participants	Sedentary	Mixed	Not taken into account	17	Cohort
39915172	France	18–35	healthy participants	athletes	Mixed	Not taken into account	26	Cross sectional

All identified studies were clinical research, and we identified no pre-clinical studies that met our inclusion criteria. Most studies included females who were between the ages of 18–30 years old, however, one study included females in the menopause transition (Figure 2A).(26) Most studies included participants whose BMI was between 18–27 kg/m² (Figure 2B), and seven studies did not report BMI. There were a variety of inclusion factors featured in this body of literature, with the most common ones being age (n = 23, 56%), eumenorrheic for the non-OCP group (n = 19, 48%), being healthy (n = 25, 63%), and being un-injured (n = 17, 43%) (Figure 2C). About one fourth of studies took place in the United States (US, n=9, 25%), and the remaining studies took place in a variety of other countries, including the United Kingdom, Denmark, and Australia (Figure 2D). For the participants on OCPs, most studies included either only cyclic monophasic (n = 14, 35%) or a variety of other OCPs, including combined, progesterone only, and some did not clearly report the type (n = 12, 30%) (Figure 2H). Notably, only one study reported why participants were on OCPs, and about half of studies only required participants to be on OCPs for at least 6 months prior to the onset of the study (Figure 2G). While most studies reported that their participants were either sedentary (n = 16, 40%) or active (n = 20, 50%) at baseline (Figure 2I), many studies did not report an objective measure of fitness (n = 16, 40%) (Figure 2E). About half of the studies did not control for phase of the menstrual cycle in controls, with the other studies investigating metrics during the early follicular, late follicular, or mid-luteal phases (Figure 2F). A variety of outcome measures were investigated, with strength-based metrics (n = 21, 53%) and serum markers (n = 20, 50%) being the most common (Figure 2J). Most studies were cohort studies by design (n = 30, 75%), and all studies are considered Level III Evidence. The average sample size was 24.4±41.8 participants per group, with the median sample size being 14 (Figure 2K). Only one study reported race,(26) and no studies reported metrics of socioeconomic status. Notably, only one study considered OCP use within the context of disease, specifically assessing muscle outcomes within patients with poly-cystic ovarian syndrome (PCOS).(25)

Figure 2: Demographic information about included studies.

A) Age range of participants. Solid line is the mean from each study, and the dashed line is the standard deviation. B) BMI range of participants. Solid line is the mean from each study, and the dashed line is the standard deviation. C) Explicitly stated inclusion criteria from each study. D) Country that study took place in. E) Metric to determine the baseline fitness of participants. F) Phase of the menstrual cycle that outcome variables were collected at. G) Length of time the OCP group had to be on OCPs. H) OCP type that was utilized. I) Qualitatively described baseline fitness of participants. J) Outcome measures reported. K) Sample size.

Rigor & reproducibility has not changed significantly in 30 years of published literature

In quantifying the rigor and reproducibility of the MDBC categories, rigor was in the 69^th ± 6.5 percentile. Most studies clearly stated their hypothesis/objective/aim, main outcome measures, patient characteristics, outcome data, variability, statistical tests, and method validity. However, there is a need to improve reporting of power analysis, recruitment period, blinding, recruitment representation, probability values, and loss to follow up (Figure 3A, Supplementary Table 4). We found no relationship between publication year and overall rigor score (Figure 3B), with our expectation being that rigor would improve. Given that this body of literature spans from the late 1990’s to 2025 and remained in the 69^th percentile, this finding suggests a lack of change in rigor over this 30-year timeline.

Figure 3: Rigor and reproducibility analysis of included studies.

A) Modified Downs & Black Checklist categories averaged across studies. B) Relationship between year of publication and Modified Downs & Black Checklist total score

OCPs have no measurable effect on muscle physiology outcome variables in healthy, young women

When assessing strength, we found no effect of OCPs on knee extension, knee flexion, or grip strength (Figure 4). Similarly, we found no effect of OCPs on jump height or muscle CSA (Figure 5). Given that knee extension strength and grip strength had one outlier study,(56) we repeated analyses without this study, as a means of sensitivity, and still did not detect an effect (KE: Effect Size = 0.148, 95% CI = [−0.247, 0.542], p = 0.464; GS: Effect Size = 0.706, 95% CI = [−0.660, 2.072], p = 0.311). Meta-analyses for knee extension strength, grip strength, and CSA also had high heterogeneity and loss of homogeneity, suggesting intra-study variability (e.g., different types of OCPs, duration of use, baseline activity level, outcomes measurements) may be contributing to the null finding. Therefore, we repeated meta-analyses for knee extension with sub-group analyses for phase of the menstrual cycle and oral contraceptive type and again detected no effect, but with low heterogeneity and non-significant homogeneity (Appendix Figure 1).

Figure 4: Meta-analysis on the impact of OCPs on strength.

A) Knee extension strength. B) Knee flexion strength. C) Grip strength. Squares represent the effect size of the individual study, the size of the square represents the weight of that study, lines with no caps represent the 95% confidence interval (CI) of individual studies, the diamond represents the overall effect size, and the line with caps represents the overall 95% CI.

Figure 5: Meta-analysis on the impact of OCPs on jump height and muscle cross-sectional area.

A) Jump height. B) Muscle cross-sectional area. Squares represent the effect size of the individual study, the size of the square represents the weight of that study, lines with no caps represent the 95% confidence interval (CI) of individual studies, the diamond represents the overall effect size, and the line with caps represents the overall 95% CI.

Serum markers, molecular markers, and histology

Serum marker data — while too limited for a meta-analysis — suggests that OCPs raise several markers, including cortisol,(50) creatinine kinase,(47) fatty acid binding protein,(48) growth hormone,(66) interleukin (IL)-8,(50) and triglycerides (25). Conversely, OCP usage was associated with reduced insulin.(66) OCPs showed no effect on markers such as calcium,(26) follistatin,(65)(65) glucose,(66) insulin like growth factor 1,(67) IL-10,(50) IL-1β,(50) IL-6,(50) lactate,(37) osteocalcin,(26) phosphorus,(26) total cholesterol,(25) or vitamin D (26).

Histologically, OCP use was associated with no effect on collagen synthesis,(42) c-terminal cross-linking telopeptide of type I collagen,(26) or myofibrillar protein fractional synthesis rate (42). When examining the transcriptome, mechanistic targets of rapamycin (mTOR) (51) and myogenic regulatory factor 4 (51) expression levels displayed consistent increases with OCP usage. However, androgen receptors,(51) muscle RING finger 1,(51) myoblast determination protein 1,(51) myoglobin,(48) Pax7,(51) and tumor necrosis factor-α (51) were not affected by OCP use. (Table 2)

Table 2.

Qualitative summary of OCP effects on serum markers, molecular markers, and histological outcomes

Serum markers, molecular markers, or histological metric	OCP Effect	Effect Size [95% CI]	Reference
Androgen receptors	≈	−0.61 [−1.27, 0.05]	51
Calcium	≈	0.12 [−0.30, 0.54]	26
Collagen Synthesis	≈	−0.29 [−1.15, 0.58]	42
Cortisol	↑	3.87 [2.38, 5.35]	50
Creatinine Kinase	↑	2.62 [1.42, 3.82]	47
C-terminal cross-linking telopeptide of type I collagen	≈	0.21 [−0.21, 0.63]	26
Fatty Acid Binding Protein	↑	1.76 [0.61, 2.91]	48
Follistatin	≈	0.41 [−0.26, 1.08]	65
GH	↑	6.67 [4.57, 8.75]	66
Glucose	≈	0.72 [−0.12, 1.56]	66
IGF1	≈	−0.67 [−1.62, 0.28]	67
IL-10	≈	0.56 [−0.34, 1.45]	50
IL-1β	≈	−0.38 [−1.26, 0.50]	50
IL-6	≈	0.50 [−0.39, 1.39]	50
IL-8	↑	1.57 [0.57, 2.57]	50
Insulin	↓	−0.94 [−1.81, −0.08]	66
Lactate	≈	0.20 [−0.60, 1.02]	37
mTOR	↑	7.78 [5.89, 9.66]	51
Muscle RING finger 1	≈	−0.65 [−1.31, 0.02]	51
Myoblast determination protein 1	≈	0.46 [−0.19, 1.11]	51
Myofibrillar protein fractional synthesis rate	≈	−0.85 [−1.71, 0.03]	42
Myogenic regulatory factor 4	↑	2.45 [1.59, 3.30]	51
Myoglobin	≈	−0.13 [−1.11, 0.84]	48
Osteocalcin	≈	0.20 [−0.22, 0.62]	26
Pax7	≈	−0.47 [−1.12, 0.19]	51
Phosphorus	≈	−0.13 [−0.55, 0.29]	26
Total Cholesterol	≈	0.32 [−0.30, 0.94]	25
Triglycerides	↑	1.12 [0.45, 1.79]	25
Tumor necrosis factor-α	≈	0.57 [−0.09, 1.23]	51
Vitamin D	≈	−0.18 [−0.60, 0.25]	26

Discussion

In this systematic review, we aimed to compile the science studying the impact of OCPs on skeletal muscle in female humans and animals. A prior meta-analysis by Nolan et al. has compiled similar literature with a focus purely on strength outcomes with no demonstrated impact of OCPs on strength.(68) Here, we included research reporting non-strength related outcomes, resulting in 34 more articles being analyzed. We identified no preclinical studies that met our inclusion criteria, and the clinical studies were focused on a homogenous population (young, low BMI, healthy, sedentary to active females) but with high heterogeneity in study design (different types of OCP, lack of control for menstrual cycle or energetic status). Preclinical studies were largely excluded due to studying hormone replacement therapy, rather than OCPs, or not having skeletal muscle outcome variables. Additionally, all of the clinical studies were retrospective cohort or cross-sectional designs, lacked reporting of key control information (i.e., why were the participants on OCPs), and had relatively small sample sizes (e.g., in our meta-analyses, most studies had <10 participants per group). Most of the studies focused on strength and serum-based biomarkers, with very few studies examining markers within the skeletal muscle itself or mechanisms of action. We did not detect a relationship between overall MDBC score and year of publication, suggesting no significant improvements in rigor and reproducibility between 1990 and 2025. Although we did not detect any effects of OCPs on knee extension strength, knee flexion strength, grip strength, CSA, or jump height, given the loss of homogeneity in many of our analyses, it is unclear if these are true null findings or results of underlying heterogeneity between studies.

One possible explanation for true null findings amongst this body of literature is the study population. Specifically, nearly every study included in this review examined the impacts of OCPs on skeletal muscle in young, nulliparous individuals, with no comorbidities (e.g., no PCOS), while the population of individuals taking OCPs includes individuals who are middle-aged, post-partum, and have potential comorbidities. OCPs administer ethinyl estradiol, a synthetic estrogen structurally distinct from endogenous 17β-estradiol, at doses higher than physiological levels; however, circulating 17β-estradiol concentrations in OCP users (20–50 pg/ml)(69) remain within a range comparable to eumenorrheic women (5–100 pg/ml), (70) largely due to first-pass metabolism by the liver and gastrointestinal tract.(71) This contrasts with conditions like pregnancy, where 17β-estradiol concentrations are 10,000–40,000 pg/ml,(72, 73) or menopause, where 17β-estradiol concentrations are consistently less than 10 pg/ml.(74, 75) Impacts of OCPs on skeletal muscle health may be better appreciated, and ultimately more clinically impactful, in populations with irregular menstrual cycles and aberrant sex hormone profiles, such as individuals with PCOS (76) or functional hypothalamic amenorrhea secondary to energy deficiency (e.g., Female Athlete Triad and relative energy deficiency in sport (REDs) models).(77) Therefore, we recommend future prospective clinical and preclinical studies investigating these populations and developing relevant animal models.

An alternative explanation for why there was no detectable effect of OCPs on skeletal muscle is because of study variability. It is well established that the dose, route of administration, and formulation of hormone replacement therapy (HRT) impacts its efficacy,(66, 67, 78) and the same has been shown for OCPs.(79) There was high inter-study variability as well as variability between studies on the type of OCP utilized, and this may have affected the effect sizes. Most studies required participants to be on OCPs for 6 months and OCP related skeletal muscle changes may not be appreciated unless administered over a longer period of time. Surprisingly, many studies did not control for the menstrual cycle of control subjects, and of those that did, there was variability in the phases under study. It is also possible that OCP effects are only appreciable when compared to specific menstrual cycle phases. Thus, we recommend future studies more carefully control these factors such that the OCP-skeletal muscle relationship can be further understood.

Although we did not detect a significant effect of OCPs on skeletal muscle, this should not be interpreted as evidence that OCP- or sex-hormone based influences on skeletal muscle are negligible or unworthy of further investigation. For instance, most of the included studies stated inclusion criteria related to participants being either uninjured or healthy, suggesting that the impact of OCPs on muscle regeneration, repair, and healing is a nearly completely unexplored area of research. Collins et al. has established that 17β-estradiol is critical for maintaining the muscle stem cell pool in the setting of ovariectomy, with estrogen signaling preventing stem cell apoptosis and facilitating the cell cycle.(80, 81) Indeed, when estrogen receptor alpha was knocked out in muscle stem cells, female mice were unable to regenerate muscle following injury.(80, 82) OCPs have the potential to serve as an additional tool for enhancing muscle regeneration during an acute injury as well as in chronic conditions such as sarcopenia, as has been suggested previously.(83) Thus, we recommend future studies focus on these relatively unexplored therapeutic potentials.

The lack of significant improvement in rigor and reproducibility across a 30+ year period, with the rigor of the studies included being in the 69^th percentile, is concerning. Reporting guidelines, such as CONSORT,(84) STROBE,(85) and ARRIVE,(24), were generated with the aim of them being 100% reported in every publication. One contributing factor for this finding is the chronic lack of funding for women’s health research, and we hypothesize that most of the authors performing these studies were working with minimal fiscal and institutional resources. Indeed, the National Institutes of Health (NIH) in the United States has historically dedicated about 7–9% of funding specifically to issues disproportionately affecting women,(86) and globally, funding towards women’s health research ranges from 1–3% of annual budgets.(87) Moreover, of the 34 diseases that disproportionately affect one gender, 74% of the funding are male-favored, in that they are either female-dominant and underfunded or male-dominant and overfunded.(88) Importantly, a recent meta-analysis demonstrated that degree of funding is associated with higher quality research,(89) and we encourage future work to quantify how the systemic lack of funding has affected the quality of women’s health research. Given the priority to develop and fund human-centered research models, it will be important for researchers to design clinical studies that follow well-established reporting guidelines for transparency and completeness. This will ensure enhanced rigor and reproducibility and provide a well-researched rationale to utilize pre-clinical animal models to understand the mechanistic impacts of OCPs on skeletal muscle in a manner that is impossible to study in humans. We urge funding institutions to follow suit with organizations, like the Gates Foundation, Women's Health Access Matters Foundation and others, to expand prioritization of women’s health funding.

Limitations of the study

Although this systematic review adds to the growing body of literature aimed at understanding how OCPs impact skeletal muscle, it does exist with limitations. First and foremost, due to the aforementioned heterogeneity in study design, it is challenging to interpret the null findings. Additionally, the results of these analyses are limited to a small, healthy, homogeneous population with a lack of generalizability to all individuals taking OCPs, and we have no data on the race or ethnicity of included participants. Due to the lack of prospective clinical and basic science studies, we were unable to discern any mechanisms by which OCPs may be impacting skeletal muscle.

Concluding thoughts

Understanding the impact of OCPs on skeletal muscle health is an important human-centered research problem. Our ultimate goal is to understand sex-based differences in skeletal muscle physiology and empower females with reliable information to inform decisions about their health. By understanding the underlying physiology behind exogenous sex hormones such as hormonal contraception and their impact on muscle health, we hope to not only broaden treatment options in the clinic but also reinforce preventative strategies starting from a young age and extending throughout the lifespan. This systematic review and meta-analysis were a step towards the goal of understanding the role of OCPs beyond birth control by focusing on skeletal muscle health. Surprisingly, no basic science studies were identified. We urge basic scientists to investigate the impact of OCPs, as well as individual sex hormones, on skeletal muscle in the setting of health, diseases, and injury, like PCOS, as well as to expand outcome measures to examine the muscle itself, not just strength-based outcomes or serum biomarkers. Given that clinical studies were focused only on young, low BMI, healthy, sedentary to active females, we urge clinical scientists to expand population inclusion to those who are middle aged and aged, injured, higher BMI, athletes, individuals with endocrine, reproductive, and/or metabolic disorders, as well as trans-gender individuals. Additionally, clinical studies should improve reporting of power analysis, recruitment period, blinding, recruitment representation, probability values, loss to follow up, reason for being on OCPs, race, socioeconomic status, and comorbid health conditions with the goal of improving reproducibility. The concerning outcome of this review is that that rigor and reproducibility have not changed significantly over the last 30 years. Hence, we encourage funding agencies to further prioritize women’s health research in order to support large, high quality, and appropriately powered studies. Lastly, although we did not detect any effects of OCPs on the knee extension strength, knee flexion strength, grip strength, CSA, or jump height, it is unclear if these are true null findings or results of underlying heterogeneity between studies. Addressing these methodological areas of growth will help elucidate the true impact, if any, of OCPs on skeletal muscle.

Supplementary Material

Link to Supplemetary Tables:

https://doi.org/10.6084/m9.figshare.32302392

Appendix Figure 1: Knee extension strength sub-group meta-analyses.

A) Knee extension strength with menstrual cycle phase as the sub-group. B) Knee extension strength with type of OCP as the sub-group. Squares represent the effect size of the individual study, the size of the square represents the weight of that study, lines with no caps represent the 95% confidence interval (CI) of individual studies, the diamond represents the overall effect size, and the line with caps represents the overall 95% CI.

Data Availability Statement.

All data used in this study are available at the request of the corresponding author of this paper or the individual papers with the original data.