Introduction

Sex steroid hormones are important not only for reproduction but also for many aspects of women’s health. Levels of these hormones are associated with the risk of cardiovascular diseases, osteoporosis, mood disorders and depression [1–7]. Importantly, lowering the levels of sex hormones has also a significant impact on reducing the risk of breast cancer in women [8–10].

Levels of sex hormones change in response to many factors, and it is well documented that high energy expenditure has a suppressive effect on the ovarian function. Many studies have shown a negative relationship between energy expenditure resulting from sports activities and ovarian function [11–14]. Energy expenditure during exercise often stimulates compensatory mechanisms such as weight loss or energy conservation, resulting in inhibition of reproductive function, including reduced sex steroid levels [15]. In many studies, participants who engaged in high-intensity physical activity experienced significant reduction in estrogens and progesterone levels, sometimes leading to amenorrhea [16, 17]. For this reason, menstrual disorders are common in women professionally involved in sports [18, 19].

However, not every type of exercise leads to easily detectable changes in regularity of cycles. Moderate-intensity exercise has been shown to reduce levels of ovarian steroid hormones without visible influence on a cycle length or occurrence of menstruation [20, 21]. Women with high leisure time physical activity had lower estradiol levels [22, 23] and in premenopausal women estradiol levels decreased with increasing physical activity [24]. Inhibition of reproductive function has been also described in studies on populations where high levels of energy expenditure result from occupational physical work, for example related to seasonal agricultural work [25]. Importantly, physical work can have an inhibitory effect on levels of progesterone, regardless of the good nutritional status and positive energy balance of women [25].

Relationship between physical activity and sex hormone levels has been a matter of interest, due to a role of these hormones in initiation and progression of breast cancer [26, 27]. Although many researchers have found a link between physical activity and hormone levels in postmenopausal women [26, 28–33], fewer studies have looked at estrogen and progesterone levels in premenopausal women who are not professional athletes [12, 25, 34–36]. Due to the variability of hormone levels during the menstrual cycle, hormonal studies in reproductive-age women are methodologically challenging to perform at least 7–8 days per cycle in the follicular and luteal phases should be sampled (depending on a hormone studied) to reliably characterize hormone levels in a particular cycle [37].

Most studies on physical activity and sex hormones relied on a self-reported, questionnaire data about levels of physical activity of women, without actual measurements of energy expenditure. There are several methods to measure energy expenditure, but most of them are either not possible to use outside the laboratory or are too expensive for studies with a larger number of participants (e.g., doubly labelled water). Therefore, commonly used devices such as activity trackers (smartwatches, pedometers) are often used to assess physical activity by calculating a number of steps that a person takes during a day [38, 39]. This method is also potentially of great practical value given that most individuals monitoring their physical activity use similar devices

Walking is the most basic form of human physical activity [40]. There is a statistically significant inverse relationship between daily step count and the all-cause mortality [41]. Moreover, there has been a strong evolutionary selection for increased walking, as measured by the number of daily steps. Human hunter-gatherer ancestors walked 10,000 to 18,000 steps a day in search of food [42]. Only recently, especially in high-income industrial environments, has this most basic form of human physical activity declined [40, 42].

While the information about number of steps taken is easy to obtain with the use of step counting devices it is not clear how many steps per day are needed for reproductive age women to reduce levels of sex hormones, and thus reduce the risk of breast cancer. In order to address this question, we investigated whether the average number of steps taken per day is related to the levels of estradiol and progesterone in 85 healthy, non-smoking women of reproductive age who performed moderate physical activity for 2 complete menstrual cycles. We hypothesize that a greater number of steps taken per day will be associated with lower levels of sex hormones. We have also explored whether the widely recommended minimum number of steps (10,000 steps a day) was sufficient to have a reducing effect on sex hormone levels and, thus, is beneficial for the potentially lowering the breast cancer risk.

Materials and methods

Results

Descriptive statistics for the study group are presented in Table 1. There was no statistically significant differences between the groups of women who took less than 10,000 steps and more than 10,000 steps in mean age, cycle length, body height, body weight and BMI (Table 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Baseline characteristics of study participants, their physical activity and sex hormone levels and results of independent samples t-tests.

https://doi.org/10.1371/journal.pone.0299580.t001

There was a significant negative linear association between physical activity (i.e., average number of steps taken daily) and salivary progesterone levels (p = 0.014), after adjusting for potential confounding factors (e.g. age and BMI) (Table 2). The association between physical activity and estradiol levels was not statistically significant (p = 0.688). Age and BMI were not significantly related to levels of sex steroid hormones.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. The association between physical activity and salivary estradiol (E2) and progesterone (P) levels.

https://doi.org/10.1371/journal.pone.0299580.t002

Women who took more than 10,000 steps a day had significantly lower progesterone levels compared to women who took less than 10,000 steps (p = 0.043), after adjusting for age and BMI (Table 2 and Fig 1). We found no statistically significant differences in estradiol levels in these activity groups. The model with division of the study sample into tertiles showed a statistically significant difference in the levels of progesterone among the examined women for the highest tertile compared to the lowest tertile, after controlling for age and BMI (p = 0.01). No statistically significant differences were observed for estradiol levels (Table 2 and Fig 1).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1.

Differences between groups of women who took more or less than 10,000 steps in levels of estradiol (A) and progesterone (B), and between groups of women based on tertiles of the number of steps for estradiol (C) and progesterone (D). Lines and bars represent 95% confidence interval (CI) for predicted mean.

https://doi.org/10.1371/journal.pone.0299580.g001

Next, we rerun our analyses replacing BMI with the body fat percentage (S1 Table) or muscle mass (S2 Table). There was a significant negative linear association between the average number of steps taken daily and salivary progesterone levels, after adjusting for body fat percentage and muscle mass. The model with division of the study sample into tertiles showed a statistically significant difference in the levels of progesterone among the examined women for the highest tertile compared to the lowest tertile, after controlling for body fat percentage and muscle mass. No statistically significant differences were observed in the estradiol levels (S1 and S2 Tables).

Discussion

In this study we observed a significant, negative relationship between moderate physical activity, expressed as daily step count, and salivary progesterone levels. Women who took more steps per day (thus were more physically active) had lower progesterone levels compared to women who took fewer steps. We did not observe statistically significant relationship between physical activity and estradiol levels.

The epidemiological studies show that physically active women have a lower risk of developing breast cancer [46, 47]. Not only sports and exercise, but also occupational work, domestic work and total physical activity are associated with a reduced risk of breast cancer [48]. The main physiological mechanism that is responsible for the relationship between physical activity and breast cancer risk is related to the effect of physical activity on ovarian function and its ability to reduce levels of ovarian steroid hormones [49–51].

It has been assumed that the main risk factor for breast cancer in women is exposure to high levels of estrogens, while the role of progesterone in the development of breast cancer is less clear. However, a recent review suggested that elevated levels of both hormones are in fact important risk factors [10]. According to a new hypothesis, progesterone stimulates proliferation of normal breast epithelium and also upregulates the expression of one of the DNA mutator enzymes. Estrogens, in turn, stimulate growth of already existing mutated, precancerous cells. Others suggested that progesterone may potentiate the mitogenic effects of estradiol [52] and animal models indicate that progesterone contributes to the proliferation of the endothelial gland and sensitizes cancer cells to growth factors [53–55]. In participants of our study only reduction of progesterone, but not estradiol, was observed in relation to physical activity. In the light of above described hypothesis, a reduction in progesterone could be sufficient to lower lifetime risk of breast cancer, even when physical activity does not change estradiol levels produced during menstrual cycles. The fact that we observe a response of ovarian function to physical activity in a form of lower progesterone levels, but not estradiol levels, is consistent with a classical model of ovarian suppression proposed by Ellison [56]. According to this model, ovarian function changes gradually in response to various stressors, especially those associated with energetics, such as physical activity. The various forms of ovarian suppression are not independent conditions but occur gradually. Between a fully fertile ovarian cycle and complete cessation of menstruation (i.e., amenorrhea), there are subtle degrees of variability in ovarian function. The mildest form of ovarian suppression consists of lowering progesterone secretion in the luteal phase of the cycle. The next stages may be inhibition of follicular development in the follicular phase and a lack of ovulation. A lack of menstrual bleeding is only the extreme end of a gradual continuum of ovarian responses to a changing environment [56].

In our analyzes, we decided to take the daily step count as the estimate of the level of physical activity, since walking is the most basic form of physical activity [40]. It is also an effective and simple way to increase level of physical activity. Taking 10,000 steps a day is equivalent to 30 minutes of moderate-to-vigorous physical activity [57–59] and this corresponds to the World Health Organization’s recommendation of at least 150 minutes of moderate exercise per week [60]. In addition, daily step count is accessible information that people can use when they are concern about health recommendations and when they attempt to increase physical activity. While some women are aware that physical activity reduces the risk of breast cancer, they do not know if their own levels of activity are sufficient to make a change. Our results suggests that physical activity for premenopausal women, assessed by number of steps, should not be lower than 10,000 steps per day to reduce levels of progesterone, and thus, potentially reduce a risk of breast cancer.

It should be also emphasized that sex hormones, while related to the increased risk of breast cancer, are crucial for many aspects of healthy physiology [61, 62]. Their beneficial role is well established for fertility, cardiovascular function, bone health and mental health. However, insufficient levels of physical activity may lead to changes in the body’s delicate hormonal balance contributing to altered hormone levels and adverse health outcomes. Thus, it is important for physically active women to discuss these issues with their physicians.

The strengths and limitations of our study should be taken into account. An important strength was that the hormonal analyzes were based on reliable assessment of hormone levels, as several saliva samples were analyzed for each participant, whereas still many studies rely only on a single or few measurements. Further, ovulation was determined based on LH ovulatory tests instead of a less reliable day-counting method. The limitation of our study could be the fact that number of steps per day was measured using the Fitbit wristband. Some previous studies have shown that Fitbit devices reliably estimate the number of steps and energy expenditure during basic physical activities such as walking and running [63]. Yet, this device gained a recent critique of not fully-reflecting the undertaken physical activity by an individual, both in laboratory and free-living settings (for a systematic review see: [44]). Future studies in this area involving a highly-accurate methods of physical activity measurement (e.g., doubly labelled water) are needed.

Conclusions

In the quest to increase physical activity, daily step tracking is a valuable element of health promotion. In addition, it is one of the main and easily accessible functions supported by technology. Our results indicate that taking at least 10,000 steps a day may reduce levels of progesterone, but this intensity of physical activity may not be high enough to influence levels of estradiol. Potential health implications of exercise-induced changes in levels of sex hormones observed in this study are not well understood. However, currently recommended levels of physical activity may not be high enough to reduce levels of both sex hormones involved in development and progression of breast cancer and thus may not be sufficiently high for premenopausal healthy women to significantly reduce their risk of postmenopausal breast cancer.

Supporting information

Acknowledgments

We are grateful to the women who participated in this study. This work could not have been completed without their involvement. We also thank all the study assistants who have helped with data collection.