Background
Polycystic ovary syndrome (PCOS) is a common yet complex clinical entity that affects 6–10% of reproductive-aged women, and is characterized by ovulatory dysfunction, hyperandrogenism and polycystic ovaries [
1]. Beyond anovulatory infertility, PCOS has important implications for general health as these patients are often affected by other conditions, such as depression [
2], and have an increased cardiovascular risk due to associations with obesity [
3], lipid abnormalities [
4], insulin resistance [
5] and high blood pressure [
6].
Lifestyle modifications, especially physical exercise, have been proposed to improve metabolic profiles, reproductive trajectories, as well as depressive symptoms, for patients with PCOS [
2,
7‐
9]. It is also well recognized that physical inactivity is closely linked with the cumulative burden from chronic diseases and contributes significantly to mortality worldwide [
10]. In 2008, the Department of Health and Human Services (DHHS) published updated guidelines for physical activity intended for adult Americans [
11]. According to these guidelines, adults should perform at least 150 min of moderate-intensity exercise per week, or 75 min of vigorous-intensity exercise per week, or an equivalent combination of moderate- and vigorous-intensity activity in order to achieve substantial health benefits [
11]. While the optimal exercise regimen for PCOS patients has not yet been determined, it has been shown that PCOS patients who are physically active demonstrate superior metabolic health parameters, especially those who perform vigorous-intensity exercise [
12].
Previous studies have shown that a significant portion of women with PCOS do not conduct adequate physical exercise, and that those of non-White ethnicity was associated with a higher risk of inadequate physical activity [
2,
12]. As the United States population continues to become more ethnically and socially diverse, efforts to better understand contributors to health disparities are needed. The objectives of this study were (1) to quantify exercise behaviors by metabolic equivalents and (2) to identify predictors of adequate physical activity within a multiethnic PCOS patient population in order to direct efforts in counseling high risk patients to improve their overall health outcomes.
Methods
Study population
We performed a cross-sectional assessment of patients diagnosed with Rotterdam-PCOS at a multidisciplinary PCOS clinic [
13]. Patients were either self-referred or referred by outside providers for assistance in diagnosing and/or in management of PCOS. Patients were recruited consecutively between 2006 and 2019 to participate in a research cohort at the University of California, San Francisco. Cohort participants consented to systematic clinical and questionnaire-based data collection. Institutional Review Board approval was obtained.
Diagnosis of PCOS was made according to the Rotterdam ESHRE/ASRM criteria [
13]. Evidence of clinical hyperandrogenism was defined by the presence of hirsutism using the modified Ferriman-Gallwey score ≥ 8 and/or severe androgen-dependent acne. Biochemical hyperandrogenism was defined by elevations in serum androgen levels (free testosterone, total testosterone, dehydroepiandrosterone sulfate, and/or androstenedione) above the resulting laboratory’s reference range. Participants were instructed to abstain from hormonally active medications for at least one month prior to serum testing. Exclusion of other endocrine disorders, such as hyperprolactinemia, thyroid dysfunction, and late-onset congenital adrenal hyperplasia, was also performed. Ovarian morphology was assessed via endovaginal ultrasound by a Reproductive Endocrinology and Infertility specialist, and was defined to meet polycystic ovarian morphology by the presence of at least 1 ovary with a volume greater than 10 cm
3 and/or with ≥ 12 follicles that measured between 2 and 9 mm in diameter.
Data collection
Before the initial clinic visit, patients were asked to complete a series of self-administered questionnaires. Demographic information gathered from these questionnaires pertinent to this study included self-reported race/ethnicity, menstrual history, age, income range, educational attainment, parity status, and place of birth. Anthropometric data including body mass index (BMI) and waist circumference were collected at the initial clinic visit.
Exercise data were collected prior to the initial visit using the International Physical Activity Questionnaire (IPAQ), which has been extensively validated across 12 countries as an international tool to measure physical activity [
14]. The IPAQ consists of 27 questions that ask participants to report the amount of moderate and vigorous physical activity performed in the past 7 days related to work, house chores, transportation, and recreation/sport/leisure-time physical activity. In this questionnaire, moderate physical activity refers to those that take moderate physical effort and make one breathe somewhat harder than normal. Vigorous physical activity refers to those that take hard physical effort and make one breathe much harder than normal. Patients were evaluated for adequate physical activity as published by the DHHS. Adequate physical activity (APA) was defined as at least 150 min of moderate-intensity, or 75 min of vigorous-intensity, or an equivalent combination of moderate- and vigorous-intensity activity per week [
11].
Exercise data were also analyzed by metabolic expenditure according to the amount of physical activity reported on the IPAQ. Metabolic expenditure was determined using metabolic equivalents (METs) as the unit, in which one MET is defined as the amount of oxygen consumed while sitting at rest. Its use represents a simplified and standardized way to express the energy cost of physical activity as a multiple of the resting metabolic rate [
15]. According to the IPAQ scoring protocol, each minute of moderate physical activity was assigned four METs and each minute of vigorous physical activity was assigned eight METs. The weekly metabolic equivalents from moderate physical activity, vigorous physical activity, and total (moderate plus vigorous) physical activity were determined for each patient [
14].
Data analysis
Statistical analyses were performed using SAS v9.4, 32-bit edition. For sample size calculation, we performed a post-hoc power analysis. Given our population of 465 patients and our approximately 2:1 ratio of White to non-White subjects, we had adequate power to detect a difference of 15 percentage points in likelihood of achieving APA. Age and anthropometric data are reported by their median and 25th/75th percentile values. Baseline characteristics and demographic data were compared across racial/ethnic groups using Kruskal–Wallis test, Fisher’s exact test, or chi-square test as appropriate. Metabolic expenditure from moderate-intensity exercise, vigorous-intensity exercise, and total (moderate plus vigorous-intensity) exercise among ethnic groups was analyzed using Kruskal–Wallis test. Logistic regression was used to identify whether ethnicity, income level, educational attainment, parity status, and place of birth (U.S. or non-U.S. born) were potentially relevant predictors for adequate physical activity. All testing was performed at the 0.05 level of significance.
Discussion
In the U.S., there is an emphasis on recognizing and addressing health disparities among Americans given our increasingly diverse population. In the field of obstetrics and gynecology, racial/ethnic disparities are well recognized, such as a higher risk of maternal mortality in Black women [
16,
17] and worse IVF outcomes in patients on non-White ethnicity [
18,
19]. Highlighting and understanding health disparities is an important first step to address contributing factors, which can subsequently be used to translate into improvement of health outcomes for at-risk groups. We sought to characterize exercise behaviors of women with PCOS from various racial/ethnic backgrounds, and to identify risk factors for inadequate physical activity within this patient group. We found that ethnicity, education level, and parity status are predictors of exercise behavior. Importantly, we identified South Asian PCOS patients specifically as a high-risk group for low metabolic expenditure and rates of APA. Women with PCOS who did not obtain a college degree and those who are parous also exhibited lower rates of APA.
To further elucidate the differences in frequency of APA seen among racial/ethnic groups, we performed logistic regression analyses controlling for multiple factors. We showed that the differences seen across ethnic groups are influenced by BMI and education level. As shown in Additional file
1: Table C, the relationship between ethnicity and APA was attenuated after both age and BMI were controlled (
p = 0.05); this effect appeared to be mediated by differences in BMI (
p = 0.02). After controlling for age, BMI, education level, and parity, ethnicity itself did not appear to significantly predict for adequate physical activity; this finding appeared to be driven by education level (
p < 0.05). These findings suggest that sociodemographic and anthropometric factors play an important role in explaining disparities in exercise adequacy across racial/ethnic groups.
Overall, our findings suggest that South Asian PCOS patients as a group demonstrate less physical activity, both from a threshold perspective as well as by absolute amount of metabolic expenditure. This highlights South Asian patients with PCOS as a group that is at high risk for worse health outcomes given the association of PCOS and sedentary lifestyle with adverse metabolic parameters. Specifically, vigorous-intensity activity has been associated with improved metabolic parameters in PCOS patients and was also performed the least by South Asian patients in our study cohort [
12]. Our findings supplement previous studies that assessed cardiovascular risk factors among Asian Americans compared to non-Hispanic Whites in the U.S. and found that Asian Indians demonstrate significantly less physical activity when compared to Whites in general [
20]. A prior assessment focusing on metabolic profiles in PCOS patients also noted a high prevalence of metabolic syndrome in Indian women [
21]. South Asian Americans also exhibit a significantly higher prevalence of type 2 diabetes mellitus than non-Hispanic Whites, Blacks, and Hispanics [
22,
23]. Therefore, our findings emphasize the importance of addressing physical exercise with South Asian PCOS patients, as it is such a cornerstone for treatment of PCOS [
7‐
9] and improvement in cardiovascular health [
24].
We postulate that barriers to physical activity seen among South Asian PCOS patients are different from other ethnic minority groups in the U.S., as South Asian patients had the lowest mean BMI, highest proportion of achieving a college degree, and a high percentage being of nulliparous status in our study. A possible explanation is that social and cultural norms within the South Asian American community differ in their approach to promoting physical activity. For instance, a prior community-based study interviewing South Asian women regarding their perspectives on physical activity found that there was a notion in which physical activity is unnecessary if one is “skinny”, and a general lack of awareness for the benefits associated with exercise was observed [
25]. Further, family disapproval was found to be a major barrier in this study, as South Asian women in the study were told that time should be spent on cooking and house chores, instead of on physical activity [
25]. Gender-norms related to modesty are also of concern in some South Asian women, noting that the presence of men in facilities and the pressure for culturally appropriate gym attire were barriers to physical activity in this study [
25]. Further research is needed to understand the unique challenges faced by South Asian PCOS patients in obtaining adequate physical exercise.
To our knowledge, this is the only observational study that describes exercise behaviors of U.S. PCOS patients with a specific focus on racial/ethnic variation. We recognize that our study has several limitations. Given the nature of a cross-sectional assessment, it is hard to ascertain whether the reported exercise behaviors represent the true exercise habits of these patients over a longitudinal period. Further, given the use of questionnaires for our data collection, it is possible that the patient’s self-reported ethnicity could be inaccurate or does not perfectly fit within the categories provided. The exercise data gathered from questionnaires may also introduce recall bias and potential inaccuracies. However, as compared to being interviewed directly by a researcher in person, the use of a survey potentially reduces the patient’s pressure to report an exercise amount that is not representative of their true behavior. We were also unable to capture which patients had previously received exercise counseling, potentially contributing to the differences seen between cohorts. Further, as these patients are all motivated in obtaining help with their PCOS condition, as evidenced by their presentation to the specialty clinic, our findings may be overestimated compared to the general PCOS population. The exercise trends of our Bay Area-specific patient population may also not be applicable to the general PCOS population. Lastly, while this is the largest observational study of exercise trends in women with PCOS, we also acknowledge that our findings are drawn upon relatively small sample sizes in certain racial/ethnic subgroups, which may limit the power of the logistic analyses. The relative contributions of different variables for each racial/ethnic group could be investigated in a future study once a large number of individuals were enrolled for each group.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.