Is exercise duration more important than intensity in endometriosis?

This comparative cross-sectional observational study sought to investigate the associations between exercise habits in women with endometriosis, systemic biomarkers, quality of life and mental health. It was hypothesised that those who engage in greater volumes of exercise will experience less adverse symptoms and improved quality of life. Women (assigned female at birth; AFAB) with clinically confirmed endometriosis (via laparoscopy or imaging) ( n  = 21) and healthy matched controls ( n  = 13; Fig.  1 ) were recruited via media advertisements and word of mouth, from the general community. Participants were aged between 18 and 51 years and were not currently pregnant, breastfeeding or undergoing fertility treatment. Participants completed an online survey (Supplementary File 1) to ascertain demographic information, current exercise behaviours (IPAQ) and preferences, scores of depression, anxiety and stress (DASS-21), and health-related quality of life (EHP-30; for endometriosis group only). Participants then attended a laboratory session to assess anthropometric variables, and salivary samples were self-collected. Prior to the commencement of the study, all participants were required to provide written and verbal consent following an outline of all procedures and measures. This study conformed to the Declaration of Helsinki and was approved by the Research in Human Ethics Committee at Charles Sturt University ( H23854 ). Fig. 1 Flow chart of participants Flow chart of participants Physical activity levels were assessed through the International Physical Activity Questionnaire - Short Form (IPAQ - SF) instrument [ 23 ]. The IPAQ-SF asks participants to estimate minutes spent walking, completing moderate-intensity activities and vigorous intensity activities over a 7-day period. The IPAQ-SF data was used to calculate: total volume (sum of total minutes of walking, moderate, and vigorous intensity PA); total metabolic equivalent of task (MET) minutes (sum of walking + moderate + vigorous MET minutes/week) and; vigorous MET minutes (calculated by multiplying the MET score of vigorous activity by the minutes performed (8.0 x vigorous-intensity activity minutes x vigorous-intensity days) as per the Guidelines for Data Processing and Analysis of the IPAQ. Moderate and high (vigorous) MET minutes were defined as accumulating at least 3000 or 600 MET-minutes/week respectively. The IPAQ-SF demonstrates acceptable reliability and validity for physical activity monitoring among 18–65-year-old adults, including women with chronic pelvic pain conditions [ 24 ]. The survey also included several bespoke questions related to the exercise habits (defined as planned, structured, and repetitive body movements, which are undertaken to improve one or more components of physical fitness) of participants, including how many sessions of exercise they typically participate in each week and for how long (minutes) they exercise per session. The 21-item Depression, Anxiety and Stress Scale (DASS-21) questionnaire [ 25 ] was used to measure negative emotional states including depression, anxiety, and stress over the past 7 days. Responses for each item are rated on a 4-point scale, with a total score for each subscale (depression, anxiety, and stress) being calculated by multiplying the sum of each by two. Higher scores equate to greater symptom severity. The DASS-21 demonstrates acceptable reliability and validity. The 30-item Endometriosis Health Profile (EHP-30) is a validated disease-specific questionnaire was used to measure the impact endometriosis had on participant’s health related quality of life [ 26 ]. The EHP-30 consists of 30-items across five sub domains including pain (11 items), control and powerlessness (6 items), emotions (6 items), social support (4 items) and self-image (3 items). Participant responses to each item were rated on a five-point scale (0–4). Raw scores for each domain were summed, then divided by the maximum possible raw score for that scale and multiplied by 100. Scores for each domain ranged from 0 to 100, where 0 represented the best health status and 100 the worst. Anthropometric variables were obtained using standardised techniques: stature (Stadiometer: Custom CSU, Bathurst, Australia), body mass (HW 150 K, A & D, Bradford, MA, USA), and waist and hip girths (Steel tape, EC P3 metric graduation, Australia). Participants underwent a dual energy x- ray absorptiometry (DEXA) scan for the determination of total body fat mass, body fat percentage, fat distribution, and lean muscle mass (GE Lunar Prodigy, GE Healthcare, Madison, WI, USA) and resting measures of blood pressure (BP) and heart rate (HR) were also collected. BP was measured according to standardised techniques; following five minutes of seated rest, two blood pressure measurements were taken 60 s apart and recorded as the average [ 27 ] (Welch Allyn, Arden, North Carolina, USA; Polar Team 2, Polar Electro Oy, Kempele, USA). Saliva samples were collected via passive drool into a specimen jar (Eppendorf microtube). Participants were instructed to collect the sample on day 12 of their menstrual cycle and within 45–60 min upon waking. Participants were instructed to abstain from eating, drinking or brushing their teeth prior to collection, with water swilling recommended ten minutes prior to collection (Salimetrics, Stratech Scientific APAC, PTY, LTD). Upon collection, samples were stored at −20 °C. Concentrations of estradiol, cortisol and high-sensitivity c-reactive protein (CRP) were analysed with a sandwich enzyme immunoassay technique (Salimetrics, Stratech Scientific APAC, PTY, LTD), according to manufacturer’s instructions. Intraassay coefficients of variation for all markers were 20%. Normal distribution of data was determined by the application of a Shapiro- Wilk’s test. Where data was normally distributed, independent samples t -test were applied to determine significance between groups. Where data was not normally distributed, Mann-Whitney U was applied to determine between group differences. Hedges g with a correction for small sample size was calculated with 95% confidence intervals (CI), where 0·2 indicates small, 0·5 indicates medium and 0·8 inductees a large effect. Pearson’s Correlation Coefficients were also performed to determine relationships amongst variables, to control for multiple comparisons and false discovery rate, the Benjamini-Hochberg procedure was applied (Q = 0·05), all correlations remained significant. Significance was set at p  < 0·05. Statistical procedures were performed using Predictive Analytic Software (PASW) (Statistical Package for the Social Sciences for Windows version 29·0 Chicago, IL, USA). An a-priori power analysis was completed based on data obtained from previous similar studies. The output parameters demonstrate a sample size of 32 to provide actual power of 0·70.

Anthropometric and descriptive variables and exercise data can be found in Tables  1 and 2 respectively. There were no between group differences for body composition, depression, anxiety, estradiol, cortisol, CRP or minutes of exercise per session ( p  > 0·05). However, the control group completed significantly more exercise, as represented by sessions per week ( p  = 0·007; Hedges g = 0·43, 95% CI −0·27 − 1·13) total volume of exercise ( p  = 0·007; Hedges g = 0·79, 95% CI 0·07 − 1·51), moderate metabolic equivalent of task (MET) minutes ( p  = 0·02; Hedges g = 0·48, 95% CI −0·22 − 1·18), vigorous MET minutes ( p  = 0·04; Hedges g = 0·43, 95% CI −0·27 − 1·13) walking MET minutes ( p  = 0·04; Hedges g = 0·53, 95% CI −0·18 − 1·23) and total MET minutes ( p  < 0·001; Hedges g = 0·55, 95% CI −0·15 − 1·26). Furthermore, the endometriosis group had higher levels of stress compared to the control group ( p  = 0·002; Hedges g = 1·34, 95% CI 0·58 − 2·11). Table 1 Mean ± SD descriptive and anthropometric variables within the endometriosis ( n  = 21) and control ( n  = 13) groups Endometriosis Control Age (years) 30.63 ± 7.87 29.83 ± 8.10 Height (m) 170.03 ± 6.66 169.72 ± 5.85 Weight (kg) 77.17 ± 15.59 74.89 ± 11.73 Systolic blood pressure (mmHg) 114.57 ± 8.27 118 ± 6.98 Diastolic blood pressure (mmHg) 72.86 ± 11.43 68.20 ± 10.56 Heart rate (bpm) 75 ± 12.55 73.53 ± 14.46 Fat mass (%) 39.52 ± 6.93 35.33 ± 7.65 Lean mass (kg) 45.06 ± 6.92 46.17 ± 4.35 Fat mass (kg) 30.81 ± 10.64 26.02 ± 9.72 Waist circumference (cm) 85.03 ± 13.76 83.37 ± 10.70 Hip circumference (cm) 107.69 ± 10.09 105.35 ± 8.94 Waist to hip ratio (WHR) 0.79 ± 0.07 0.79 ± 0.05 Depression 7.26 ± 8.67 9.45 ± 11.14 Anxiety 8.63 ± 8.43 3.64 ± 8.33 Stress 15.68 ± 10.14* 3.45 ± 6.33 Pain 47.49 ± 20.47 - Control & Powerless 53.07 ± 29.29 - Emotional Well-being 39.91 ± 20.05 - Self-image 52.96 ± 25.72 - Social support 58.33 ± 31.18 - * denotes statistically significant from the control group p  < 0·05 Table 2 Mean ± SD for exercise data within the endometriosis ( n  = 21) and control ( n  = 13) groups Endometriosis Control Minutes of exercise per individual session 50.79 ± 19.77 57.73 ± 16.88 Sessions per week 3.26 ± 1.58 * 5.18 ± 2.03 Total volume (minutes x sessions) 186.05 ± 127.81* 314.55 ± 200.44 Moderate MET minutes 537.89 ± 784.19* 898.46 ± 645.11 Vigorous MET minutes 1228.42 ± 2107.14* 2166.15 ± 2134.28 MET minutes walking 907.30 ± 1238.75* 1543.38 ± 1073.49 Total MET minutes 2661.18 ± 3694.31* 4608.00 ± 2937.01 * denotes statistically significant from the control group p  < 0·05 Mean ± SD descriptive and anthropometric variables within the endometriosis ( n  = 21) and control ( n  = 13) groups * denotes statistically significant from the control group p  < 0·05 Mean ± SD for exercise data within the endometriosis ( n  = 21) and control ( n  = 13) groups * denotes statistically significant from the control group p  < 0·05 The main finding from this study is that for women with endometriosis, there was an inverse relationship between minutes completed per individual exercise session and levels of CRP ( p  = 0·02; r = −0·500), estradiol ( p  = 0·04; r =−0·444), android ( p  = 0·001; r = −0·668) and gynoid fat mass ( p  = 0·02; r = −0·487), and pain ( p  = 0·02; r = −0·502) associated with endometriosis. Further, minutes of exercise per individual session were correlated with the levels of cortisol ( p  = 0·01; r  = 0·516). Interestingly, there were no significant relationships between total volume, moderate MET minutes or vigorous MET minutes ( p  > 0·05). In the endometriosis group, levels of CRP were correlated with gynoid fat mass ( p  = 0·006; r  = 0·577), whereas higher levels of pain were associated with android fat mass ( p  = 0·01; r  = 0·534), furthermore android fat mass was correlated with poor self-image ( p  = 0·01; r  = 0·534), and lower levels of control and increased powerlessness ( p  = 0·02; r  = 0·481). This relationship held true with total body fat mass which was positively correlated with poor control and powerlessness and poor self-image ( p  = 0·03; r  = 0·474 and p  = 0·002; r  = 0·636). Scores of depression were positively correlated with stress ( p  = 0·002; r  = 0·644) and anxiety ( p  < 0·001; r  = 0·745). Depression was also correlated with poorer emotional well-being ( p  = 0·005; r  = 0·59), poorer self-image ( p  = 0·027; r  = 0·481) and poorer social support ( p  < 0·001; r  = 0·699). Similar, stress was correlated with poorer emotional well-being ( p  = 0·001; r  = 0·649) poorer self-image ( p  = 0·02; r  = 0·494) and poorer social support ( p  = 0·04; r  = 0·442). Anxiety was also correlated with poor self-image ( p  = 0·01; r  = 0·535) and levels of pain were correlated with poorer control and powerless ( p  < 0·001; r  = 0.981) poorer emotional well-being ( p  = 0·01; r  = 0·516) poorer self-image ( p  = 0·002; r  = 0·625) and less social support ( p  = 0·001; r  = 0·651) in women with endometriosis. For the control group, a greater total volume of exercise and greater number of sessions per week were inversely associated to fat mass (p = < 0·001; r=−0·840 & p = 0·008; r=−0·701) this relationship was similar for android fat mass (p < 0·001; r=−0·887 & p = 0·002; r=−0·778). Moderate MET minutes were inversely associated with systolic blood pressure (p = 0·03; r=−0·598). Scores of stress were inversely associated with fat mass (p = 0·04’ r =−0·554) and android fat mass ( p  = 0·02; r =−0·623) and further inverse associations were observed between cortisol and minutes of exercise per session ( p  = 0·03; r =−0·585). Positive correlations were observed between minutes of exercise per session, total volume and number of sessions per week and scores of depression, anxiety and stress ( p  = 0·00- p  = 0·04; r  = 0·615-0.931.931).

Endometriosis is an estrogen-dependent, inflammatory condition, associated with defective immunosurveillance of autologous tissue found in the pelvic peritoneum, ovaries and fallopian tubes [ 1 ]. In some circumstances, these endometrial lesions can be found in the pericardium, brain, pleura and lung parenchyma [ 2 , 3 ]. While the cause of endometriosis remains elusive; genetic, environmental, hormonal, immune, and epigenetic factors are implicated [ 4 ]. Given the complexity of clinical presentations, no single theory or biological process can be attributed to the development of this condition [ 4 , 5 ]. Regardless, the role of the immune system, specifically, immune dysfunction is suggested to be a key contributor in the pathogenesis of endometriosis [ 1 ]. While it is unknown whether the dysfunctional immune pathways are involved in disease initiation or if immune dysfunction results from the development of endometrial lesions [ 1 ], the associated widespread inflammatory environment may contribute to symptom severity, decreased quality of life and play a role in the development of comorbidities [ 1 , 6 ]. In addition to chronic inflammation, abnormally high levels of estrogen are associated with endometriosis [ 1 ]. Estradiol, the biologically active form of estrogen, may aggravate the pathological processes by elevating inflammation, facilitating endometrial growth thereby increasing symptom severity [ 7 ]. Generally, estrogen is produced from three sites within the body, the ovaries, peripheral tissues and endometriotic tissues [ 7 ] and as such there is a complex interaction between estrogen, immune function and adipose tissue. Specifically, estrogen may influence immune function, as demonstrated through data regarding autoimmune diseases in humans and animal models, with the severity of such diseases correlated with circulating levels of estrogens [ 8 , 9 ]. Specifically, estradiol in endometriosis may induce a pro-inflammatory environment, where in circumstances without disease, estrogens generally exhibit anti-inflammatory actions [ 9 ]. Moreover, estrogens regulate metabolism by opposing excess fat accumulation through regulation of leptin signalling [ 10 ]. Estrogens are associated with peripheral (gynoid) fat distribution [ 11 ], with visceral fat classified as a highly active metabolic tissue associated with chronic low-grade inflammation [ 12 ]. While no curative treatment option exists, pain can be managed via the pharmacological inhibition of ovulation and menstruation, surgical removal of endometrial lesions, hormonal therapy, wholistic interventions and hysterectomy. Given current pharmacological interventions are symptomatic, and often temporary [ 13 ], it is imperative to look toward non-pharmacological strategies, such as exercise, to manage symptom severity and reduce the risk of comorbidities. In other chronic pain conditions, such as chronic lower back pain, exercise has been identified as effective in reducing pain, in addition to improving mood, sleep quality, and quality of life [ 14 ], all of which are commonly reported symptoms associated with endometriosis. Additionally, there is a large body of evidence to suggest exercise exerts anti-inflammatory effects [ 15 , 16 ] and can be effective in reducing estrogen levels [ 17 ]. However, current literature reports conflicting findings regarding the role of exercise on endometrioses-associated symptoms, with some reporting no effect of exercise on pain [ 18 ] and others reporting significant improvement in pain and quality of life [ 2 , 18 ]. Furthermore, while previous studies have primarily focused on exercise history and the risk of developing endometriosis, few have reported the impact on symptomology [ 19 ] and immune-endocrine variables. Moreover, despite increasing prevalence of endometriosis, there are no specific guidelines surrounding exercise and endometriosis, however there is evidence to suggest that women are self-managing with exercise [ 20 , 21 ]. The most recent guidelines for endometriosis management highlight the need for future investigations to focus on better understanding symptom management options beyond surgery and pharmacology [ 22 ].

The complexity of the symptomology of endometriosis in undeniable, however, data from this study suggest exercise requires further investigation as a useful adjunct management strategy to modulate inflammation, estradiol and mediate symptoms associated with the condition. Specifically, exercise duration may be more important than intensity to induce changes in markers of symptom severity. As symptoms fluctuate with hormones, inflammation and with the development of lesions, future research should examine the hormone and immune cells fluctuations across the menstrual cycle, and varying levels of exercise in terms of modality and intensity, duration and total dose.

To our knowledge this is the first paper to examine the relationship amongst exercise, CRP, estradiol, cortisol and health-related quality of life in women with endometriosis. The novel finding of this study were the inverse relationships between the duration of individual exercise sessions and CRP, estradiol, pain, and fat distribution, suggesting that with a greater duration of an acute bout of exercise, regardless of intensity and total volume, biological and psychological variables are improved in women with endometriosis. Secondary findings were the associations between body fat distribution, android fat was associated with increased pain, whereas gynoid fat is associated with increased inflammation, as indicated by CRP levels. Emerging literature suggests women who are living with endometriosis are using exercise to self-manage symptoms, improve quality of life and indices of mental health [ 20 , 21 ]. Furthermore, investigations into exercise for pain management have suggested that women with endometriosis who engage in exercise three time per week are less likely to report pain symptoms [ 28 ]. Despite the growing body of literature, there are no guidelines for exercise prescription for women with endometriosis, therefore, to understand how exercise can be therapeutically applied in endometriosis, it is important to understand the physiological outcomes of both acute and long-term exercise. The outcomes from this study are amongst the first to report a dose response with exercise and variables that may influence symptom severity. Our data suggests individual session duration may be more effective than the intensity to reduce CRP, as a significant relationship presented with session duration, however not with vigorous MET minutes. This is supported by Rose and colleagues [ 29 ] who reported no significant effect of exercise intensity on CRP, interleukin (IL)−6, IL-10 or tumor necrosis factor (TNF). Furthermore, the beneficial anti-inflammatory effects of exercise stem from the culminative effect of repeated acute bouts [ 25 ]. However, myokine expression is also influenced by duration [ 30 ], so perhaps in individuals with endometriosis, where symptoms fluctuate with hormones, exercise prescription should focus on inducing myokine expression with duration, rather than intensity initially, given the transient pro-inflammatory response to vigorous intensity exercise [ 31 ]. Similarly, a murine model on endometriosis and exercise indicated that regardless of exercise frequency, exercise had a significant impact on reducing the size of endometrial lesions. Moreover, moderate and vigorous intensity exercise demonstrated the greatest effects, which was coupled with as a reduction in oxidative stress with exercise [ 32 ]. Given lesions are a significant contributor to inflammation and inflammation contributes to symptom severity and comorbidity risk [ 33 ], exercise as a means to reduce lesion size and associated inflammation should be considered as a viable adjunct management strategy. Long term decreases in systemic inflammation are suggested to be mediated by reductions in adipose tissue. A systematic review and meta-analysis from Fedewa, Hathaway, & Ward-Ritacco [ 34 ] suggests regular exercise training is associated with lower levels of CRP, and greater improvements occur with reductions in body fat. This relationship held true for the present study, the duration of exercise (per individual session) was inversely associated with CRP and android and gynoid fat mass, suggesting longer acute exercise sessions may promote an anti-inflammatory environment through actions of myokines. Further, greater energy expenditure, results in decreases in fat mass, and subsequently downregulating metabolic inflammation. The effects of exercise on estrogen levels have been reported in post-menopausal women, suggesting exercise can decrease circulating levels of estrogen, thereby reducing the risk of breast cancer [ 17 , 35 ]. In endometriosis, estrogen has been associated with increased cell proliferation and disease progression [ 36 ] which significantly impacts quality of life and psychological well-being. The findings of this study are in alignment with the work of Kossman and colleagues [ 37 ] suggesting regular exercise may decrease estrogen, and as such, may result in improvements in symptoms. However, there are complex interactions with estrogen, immune function and the distribution of adipose tissue that must be further explored in the context of exercise in endometriosis. While previous research generally suggests an inverse relationship between body mass index (BMI) and endometriosis [ 38 ], there are several limitations associated with using BMI. Specifically, BMI is not a direct measure of body fat, and it does not indicate fat distribution [ 39 ]. Visceral fat is a highly active metabolic tissue which may play a role in inflammation and the development of comorbidities such as cardiometabolic disease [ 40 , 41 ]. Our study utilized DEXA to analysis body composition as the gold standard for indirect assessment [ 42 ], an important finding of this study is; where fat was more centrally mediated (android fat) in women with endometriosis, higher pain levels resulted. This is important as visceral fat is a site of IL-6 production [ 43 ], high levels of IL-6 are associated with endometriosis related infertility [ 44 ] and concentrations of IL-6 may affect pain sensitization [ 45 ]. Further, there are significant psychosocial implications of higher levels of pain, affecting self-image, control and feelings of social support, therefore, investigations into strategies to not only assist with pain management, but to improve psychosocial well-being in women with endometriosis are warranted., This study reported higher rates of depression which were associated anxiety and stress, and depression, anxiety and stress had significant impacts on emotional well-being, self-image and social support. While these findings are not uncommon amongst this population [ 46 ], they are important as negative emotional states may increase pain [ 47 ] and influence immune-endocrine variables [ 48 ]. However, regular physical activity is well-reported to decrease pain [ 49 ] and regulate emotions [ 50 ]. Moreover, women with endometriosis observed a correlation between cortisol and minutes of exercise per individual session. While hypercortisolism is reported in this population, the nature of this relationship and the physical and psychological interactions are unclear. Accordingly, investigations into the menstrual cycle, symptoms, affect, immune-endocrine variables and exercise dose are required to further understand the mechanisms associated with such changes. This study is not without limitations, firstly it must be acknowledged that the sample size is relatively small with an actual statistical power of 0.7 and the groups are unequal, therefore the study is slightly underpowered. Future studies should expand the sample size to validate the associations reported in this study [ 51 ]. Secondly, this study used self-reported exercise, as such there is the possibility exercise behaviours may be under- or over-reported, and recall bias may exist [ 52 ]. Finally, the coefficient of variation of the biomarker assay was 20%, which is considered high, therefore, results must be interpreted with caution.

Supplementary Material 1. Supplementary Material 1.