http://orcid.org/0000-0001-5226-1972
Figures
Abstract
Objective
To determine factors associated with differential changes in body fat, insulin resistance and aerobic capacity following a 12-week exercise intervention in overweight and obese women with and without polycystic ovary syndrome (PCOS).
Methods
16 overweight and obese women (9 PCOS; 7 without PCOS) completed a supervised progressive 12-week exercise program. Primary outcomes included changes in indicators of insulin sensitivity (including glucose infusion rate relative to fat-free mass [GIR/FFM]), body composition, and aerobic capacity (VO2 peak; 12 participants only). Comparisons were made between women with and without PCOS, and between participants who lost ≥5% (classified as exercise responders) and <5% (non-responders) in body fat (assessed by dual-energy X-ray absorptiometry).
Results
Training decreased body fat percentage by (mean; 95% CI) -2.3%; -5.3, 0.7% in women with PCOS and by -6.4%; -10.9, -1.9% in women without PCOS (P = 0.08). Ten women (7 PCOS; 3 without PCOS) did not reduce body fat by ≥5%. All participants improved VO2 peak (mean change 27%; 16–39%) but four (2 PCOS; 2 without PCOS) demonstrated decreases in GIR/FFM (mean change for whole cohort: 37%; 3–71%). Android-gynoid fat ratio (0.58; 0.51, 0.66 vs 0.46; 0.40, 0.51; P<0.01) was significantly higher and GIR/FFM (6.69; 3.49, 9.90 vs 11.44; 9.15, 13.72 mg/kg/min; P = 0.01) was significantly lower in non-responders compared with responders at baseline, but non-responders had significant post-training decreases in android-gynoid ratio (-0.02; -0.04, -0.01; P = 0.03), and increases in VO2 peak (7.24; 2.28, 12.21 mL/kg/min; P = 0.01) and GIR/FFM (1.44; 0.27, 2.61 mg/kg/min; P = 0.02). In women with PCOS, pre-training VO2 peak was significantly negatively correlated with change in total body fat (r = -0.75; P = 0.02), and pre-training fasting glucose negatively correlated with changes in VO2 peak (r = -0.76; P = 0.04), but positively correlated with changes in GIR (r = 0.67; P = 0.046).
Citation: Scott D, Harrison CL, Hutchison S, de Courten B, Stepto NK (2017) Exploring factors related to changes in body composition, insulin sensitivity and aerobic capacity in response to a 12-week exercise intervention in overweight and obese women with and without polycystic ovary syndrome. PLoS ONE 12(8): e0182412. https://doi.org/10.1371/journal.pone.0182412
Editor: Conrad P. Earnest, Texas A&M University, UNITED STATES
Received: February 8, 2017; Accepted: July 7, 2017; Published: August 3, 2017
Copyright: © 2017 Scott et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This study was supported by a National Health and Medical Research Council Project Grant (APP606553) awarded to NKS. DS is supported by a National Health and Medical Research Council RD Wright Biomedical Career Development Fellowship (APP1123014). BdC is supported by a National Heart Foundation Fellowship (100864). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Polycystic ovary syndrome (PCOS) affects up to one in five reproductive-aged women in Australia and is associated with reproductive dysfunction, as well as metabolic consequences including obesity, type 2 diabetes and cardiovascular disease [1]. Given the known benefits of exercise targeting weight loss for preventing and reversing cardiometabolic conditions, exercise is recommended as an important therapy for women with PCOS [2, 3]. However, we previously reported that a 12-week moderate to vigorous intensity exercise program failed to normalise insulin sensitivity in overweight women with PCOS compared to overweight women without PCOS [4]. Women with PCOS have also been observed to have significantly smaller increases in growth hormone in response to exercise than controls [5].
Poor exercise responsiveness appears highly heritable [6] and may affect around 10% of exercise participants in the general population [7]. However, modifiable factors, including higher body fat and insulin resistance are common in women with PCOS [8, 9], and may increase risk for poor exercise responsiveness in this population. It has been reported that 15–20% of individuals with type 2 diabetes [10], and as many as 40% of middle-aged individuals at high risk of type 2 diabetes [11], experience minimal improvements in metabolic health following exercise. Furthermore, higher total body fat mass is associated with smaller improvements in physical function following an exercise intervention, irrespective of exercise adherence in overweight and obese older adults [12]. It has been observed in overweight postmenopausal women however that aerobic capacity significantly increases in response to exercise even in those who do not reduce fat mass substantially [13]. Thus, although PCOS women tend to have higher amounts of body fat [14], it is unclear whether this may also attenuate exercise benefits and whether differences in body fat distribution are more important than overall body fat for exercise responsiveness.
The aims of this secondary analysis of a 12-week exercise intervention in overweight women with and without PCOS to determine factors associated with differential changes in body fat, insulin resistance and aerobic capacity following a 12-week exercise intervention in overweight and obese women with and without PCOS.
Material and methods
Study design and participants
This study was a secondary analysis of a 12-week exercise intervention for overweight and obese (BMI >25 kg/m2) and inactive (<100 minutes per week of self-reported moderate and vigorous physical activity) women. Participants constituted a subset of a cross-sectional study of pre-menopausal women with and without PCOS [4, 9, 15] and were recruited through community advertisements. The Monash Health Research Advisory and Ethics Committee approved the study and participants gave written informed consent. All investigations were conducted according to the principles expressed in the Declaration of Helsinki. The trial was registered previously (ISRCTN84763265).
As described in Fig 1, 117 women were initially screened of whom 34 (n = 20 with PCOS; n = 14 without PCOS) were eligible and consented to the exercise intervention. Thirteen PCOS and eight non-PCOS women completed the study, and amongst these 16 (9 with PCOS and 7 without PCOS) had complete body composition and insulin sensitivity data for this secondary analysis. Diagnosis of PCOS was undertaken by endocrinologists (S.K.H.) based on meeting NIH criteria with demonstrating irregular menstrual cycles (21 or 35 days) and clinical (hirsutism, acne) or biochemical (elevation of at least one circulating ovarian androgen) hyperandrogenism. As these women were not assessed for polycystic ovaries on ultrasound, they still meet the Rotterdam criteria but exact phenotype was unknown [16]. Hyperprolactinemia, thyroid dysfunction and specific adrenal disorders were excluded clinically and where indicated, biochemically. All women without PCOS had regular menses and no evidence of clinical or biochemical hyperandrogenism. Exclusion criteria were age <20 or >40 years, lipid-lowering agents, smoking, diabetes, recent weight change of 5 kilograms or more in the previous six months, actively trying to lose weight and pregnancy.
At screening, standard diet and lifestyle advice were delivered based on recommendations provided by the National Heart Foundation of Australia (www.heartfoundation.com.au) [17]. Medications known to affect end-points, including insulin sensitisers, anti- androgens and hormonal contraceptives were ceased three-months prior to baseline. All data were collected in the follicular phase for women without PCOS in order to minimise potential impact of female reproductive hormones on study outcomes. However, as women with PCOS had irregular menses they were measured at times convenient to participants.
Exercise intervention
Participants undertook 12 weeks of supervised, progressive, moderate and vigorous exercise training on a motorised treadmill [4, 15, 18]. Participants attended three 45–60 min sessions supervised by study staff each week, which sequentially alternated between moderate-intensity (20–60 minutes of walking or jogging at 75–80% of maximal heart rate [HRmax]) and high-intensity interval training (six to eight five-minute intervals at ∼95–100% HRmax with two-minute passive recovery periods). Participants progressed to eight repetitions by week four and reduced recovery time to 1 minute by week eight. Target exercise heart rates were achieved by altering speed and incline on the treadmill according to individual fitness [15]. Exercise adherence was assessed by recording the number of sessions attended, and indicators of exercise compliance were recorded, including total duration of exercise, total distance covered and average heart rate assessed by heart rate monitors (Polar Electro Oy, Kempele, Finland).
Measurements
All measurements were obtained by study staff blinded to the PCOS status of participants. Body weight and height were assessed and body mass index (BMI) calculated as weight (kilograms) divided by height (metres) squared. Total body fat mass and fat-free mass, and android and gynoid, fat mass were estimated by a whole-body dual energy X-ray absorptiometry (DXA) scan (GE Lunar Prodigy, GE Lunar Corp, Madison WI, USA; operating system version 9). VO2 peak was assessed during a graded exercise test (GXT) on a motorised treadmill (Biodex RTM 500, New York, USA) treadmill until volitional fatigue. Expired gases were collected and analysed during the GXT using the MOXUS modular system (AEI Technologies, Pittsburgh, PA, USA). The Moxus was calibrated prior to each test with known gases (16% O2; 4% CO2; BOC gases Australia) and volume (3L Hans Rudolph syringe).
The primary outcome was insulin sensitivity as measured by glucose infusion rate (mg/m2/min) assessed by the euglycaemic–hyperinsulinaemic clamp technique as previously described [15]. Clamp timing was standardised to 48 h after exercise and following a standardised high-carbohydrate diet for 72 hours before an overnight fast. Insulin (Actrapid; Novo Nordisk, Bagsvaerd, Denmark) was infused at 40 mU/m2 min for 120 min, with plasma glucose maintained at ∼5 mmol/L using variable infusion rates of 25% glucose. Glucose infusion rate (GIR) was calculated during steady state, achieved in the last 30 minutes of the clamp and expressed as glucose (mg) per kg of fat free mass per min (GIR/FFM). Stored blood samples (-80°C) were batch analysed for plasma fasting glucose and insulin, homeostatic model assessment of insulin resistance (HOMA-IR) was calculated [9].
Statistical analyses
Prior to analyses, data were assessed for normality by Shapiro-Wilk tests stratified by PCOS status and fasting insulin and HOMA-IR were log transformed. Independent samples t-tests compared changes in total body fat and VO2 peak from pre- to post-training in women with and without PCOS. Independent samples t-tests and paired t-tests compared pre- and post-exercise values for body composition, VO2 peak, and indicators of insulin resistance and exercise compliance in women defined as exercise ‘responders’ (≥5% decrease in total body fat) versus ‘non-responders’ (<5% decrease in total body fat) following training [13], and a chi-square test compared proportions with and without PCOS. Pearson correlations stratified by PCOS status examined associations of pre-training participant characteristics with post-training changes in body composition, VO2 peak and indicators of insulin resistance. All statistical analyses were performed using SPSS version 23.0 (IBM, USA) and P-values <0.05 were considered statistically significant.
Results
Compared with the 18 participants who were excluded due to non-completion or missing data (Table 1), the 16 included participants were of similar age (mean; 95% CI) (32.1 years; 29.4, 34.8 vs 32.1 years; 28.9, 35.3; P = 0.99), BMI (34.9 kg/m2; 31.9, 37.9 vs 36.29 kg/m2; 32.9, 39.4; P = 0.55), and GIR/FFM (9.39; 7.02, 11.76 mg/kg/min vs 8.47; 6.13, 10.81 mg/kg/min; P = 0.56). The majority of excluded participants were lost to contact (could not be contacted by phone or email; 4) or refused to continue (due to time commitment; 1, vasovagal episode; 1, knee soreness; 1, or no reason given; 5) in the exercise intervention by week six of the intervention. One participant who completed the intervention was excluded due to a protocol violation (commenced significant sustained physical activity outside the intervention), and five had incomplete GIR/FFM data at baseline or follow-up (Fig 1). Amongst the included participants, four (3 with PCOS, 1 without PCOS) did not have exercise tests either at baseline or follow-up and so were excluded from analyses of change in VO2 peak. As demonstrated by the body fat, VO2 peak and GIR/FFM changes reported in Fig 2, there was a wide range of individual responses to the exercise program. Ten (7 with PCOS and 3 without PCOS; P = 0.2) participants did not reduce their body fat by ≥5%, and were classified as non-responders. Training decreased body fat percentage by -2.3%; -5.3, 0.7% in women with PCOS and by -6.4%; -10.9, -1.9% in women without PCOS (P-value for difference between groups = 0.08). Participants with PCOS lost less gynoid fat compared with participants without PCOS (-0.02; -0.29, 0.24 kg vs -0.62; -1.14, -0.09 kg; P = 0.02). Conversely, all participants improved VO2 peak (27%; 16–39%) with no significant differences between women with PCOS and those without (P = 0.21). Four (2 with PCOS; 2 without PCOS) participants demonstrated no improvement in GIR/FFM (mean change for whole cohort: 37%; 3–71%) with no significant difference between groups (P = 0.39).
Baseline participant characteristics for exercise responders and non-responders defined according to body fat changes are presented in Table 1. There were no differences in mean age or PCOS status between groups. There were also no differences in total body fat, android and gynoid fat mass, however android-gynoid ratio was significantly higher in non-responders. Measures of insulin sensitivity were generally similar but GIR/FFM was significantly lower in non-responders at baseline. Table 1 also presents post-training values for body composition and indicators of insulin sensitivity according to exercise responsiveness. Total body fat, android fat and gynoid fat were significantly decreased post-training in responders, although the decrease in android fat in non-responders was of borderline significance (P = 0.05). Non-responders had a significant decrease in android-gynoid ratio post-training. Both responders and non-responders demonstrated a significant increase in VO2 peak, while only non-responders had significant increases in insulin sensitivity as measured by GIR.
As demonstrated in Table 2, there were no differences between responders and non-responders for exercise compliance or indicators of exercise workload. Women with PCOS covered significantly higher mean distance than women without PCOS for the duration of the intervention, however exercise workload (calculated as heart rate divided by distance covered (bpm/km), was significantly higher in women without PCOS.
Pearson correlations were performed to examine associations of baseline characteristics and exercise adherence/compliance in women with and without PCOS who completed the intervention, and changes in body composition, VO2 peak and indicators of insulin sensitivity from pre- to post-training. In women without PCOS, age was significantly positively correlated with changes in fasting insulin in response to training (r = 0.76; P = 0.047) while higher total body fat (r = -0.82; P = 0.03) and gynoid fat (r = -0.82; P = 0.03) at baseline were negatively correlated with changes in fasting glucose, and gynoid fat was also negatively correlated with change in gynoid fat (r = -0.77; P = 0.04). Baseline VO2 peak was negatively correlated with changes in android-gynoid fat ratio in women without PCOS (r = -0.97; P = 0.01). None of the baseline characteristics or exercise adherence/compliance indicators were associated with change in VO2 peak in women without PCOS (all P>0.05). Total number of exercise sessions was negatively correlated with change in fasting glucose (-0.76; P = 0.049).
As demonstrated in Fig 3A, baseline VO2 peak was significantly negatively correlated with change in total body fat in women with PCOS, but not those without PCOS (two without PCOS had missing VO2 peak data at baseline). Baseline VO2 peak was also negatively correlated with change in gynoid fat (r = -0.69; P = 0.04) in women with PCOS. Baseline total body (r = -0.67; P = 0.048) and gynoid (r = -0.73; P = 0.03) fat were negatively correlated with change in android-gynoid ratio. Conversely, baseline android fat (r = 0.70; P = 0.04) and android-gynoid ratio (r = 0.72; P = 0.03) were positively correlated with change in gynoid fat. Baseline fasting glucose was negatively correlated with change in VO2 peak (r-0.76; P = 0.046) but also negatively correlated with change in glucose (r = -0.76; P = 0.02), and positively correlated with change in GIR/FFM (Fig 3B). The only association observed between exercise adherence and compliance indicators and exercise responsiveness in women with PCOS was a positive correlation between total exercise distance and change in VO2 peak (0.82; P = 0.044).
Discussion
The primary findings of this secondary analysis of a 12-week intensified exercise intervention in overweight and obese women with and without PCOS was a substantial variation in insulin sensitivity, body fat and aerobic capacity changes in response to exercise, and although all participants improved VO2 peak (indicating improved fitness), approximately one–quarter did not improve insulin sensitivity and two-thirds did not reduce body fat by ≥5%. Although previous research has suggested that exercise non-responsiveness can be overcome by increasing exercise dose [19], we found few associations between indicators of exercise compliance and workload with study outcomes. Rather, high fasting glucose was associated with poorer improvements in VO2 peak, while lower baseline VO2 peak, higher relative central adiposity and lower insulin sensitivity at baseline were all associated with smaller fat losses. Nevertheless, despite smaller declines in total body fat, exercise non-responders significantly reduced relative central adiposity and improved aerobic capacity and insulin sensitivity. These findings suggest that overweight and obese women with and without PCOS can obtain health benefits from short-term exercise interventions even in the absence of significant body fat losses.
Women who failed to significantly reduce body fat in this study had significantly higher baseline android-gynoid ratios and lower GIR/FFM pre-training than exercise responders, and these associations are likely to be inter-related. In a recent study of normal weight PCOS women, intra-, but not subcutaneous, abdominal fat was positively associated with fasting insulin [20], supporting the concept that ectopic fat depots are more closely associated with cardiometabolic health than subcutaneous depots [21]. A systematic review has demonstrated that 15–20% of individuals with type 2 diabetes mellitus fail to improve metabolic health in response to supervised exercise programs [10], suggesting that dysregulation of glucose homeostasis (i.e high blood sugar levels, low insulin secretion and/or reduced insulin sensitivity) may be associated with poor exercise responsiveness in terms of body composition changes. Individuals at high risk of developing type 2 diabetes who fail to improve insulin sensitivity in response to an exercise program have increased transforming growth factor (TGF)-β activity in skeletal muscle, suggesting inflammation may contribute to poorer improvements on mitochondrial fuel oxidation and insulin sensitivity in this population [11]. Resistance to the beneficial effects of exercise has been shown to be highly heritable, with heritability estimates of almost 50% for VO2 peak in responses to exercise training [6]. However, the findings of the present study suggest adjunct therapies which improve insulin sensitivity and body composition, such as caloric restriction, may enhance metabolic exercise responses in overweight and obese women.
Despite losing small amounts of body fat following the 12-week exercise intervention, non-responders demonstrated significant decreases in android-gynoid ratio. This highlights that women who lose small amounts of fat in response to exercise may still undergo a shift to a more metabolically healthy body composition, and is supported by the finding that VO2 peak and GIR/FFM significantly improved in this group. An aerobic exercise and resistance training intervention in 31 overweight and obese postmenopausal women had similar numbers of non-responders (55%) as observed in our study (63%), despite its longer duration of 12 months [13]. Also similar to our study, the previous study observed higher BMI to be a predictor of poor responsiveness, although they did not assess differences in body fat deposition. Nevertheless, the investigators reported that primary outcomes including muscle strength and VO2 peak increased similarly for responders and non-responders [13]. These findings indicate that loss of body fat is not a limiting factor in improving aerobic capacity and insulin sensitivity in women in response to exercise, and highlight the need to focus on end-points other than body weight in interventions targeting overweight and obese individuals.
Several baseline characteristics were observed to be correlated with exercise-induced changes in body composition and indicators of insulin sensitivity in women with PCOS, yet these associations were somewhat controversial. For example, higher baseline total body and gynoid fat were associated with greater declines in android-gynoid ratio, but higher baseline android fat and android-gynoid ratio were associated with smaller declines in gynoid fat. These conflicting findings are consistent with the concept that not all individuals are similarly responsive to exercise training and potentially indicate that multi-component lifestyle modification programs may be most beneficial for women with PCOS. For example, a 16-week program of combined caloric restriction, behavioural modification, physical activity, and (for obese women) weight loss medication, resulted in weight loss of over 6% in women with PCOS [22], which is greater than the 2% weight loss observed in our intervention involving exercise only [4]. Nevertheless, our findings have demonstrated that insulin resistance and aerobic capacity can improve in response to exercise even when fat does not substantially decrease.
Higher baseline VO2 peak was associated with greater declines in total and gynoid fat in women with PCOS, which was somewhat surprising given improvements might be expected to be greater in those with the poorest cardiorespiratory fitness at baseline. Somewhat controversially, higher baseline fasting glucose was associated with smaller increases in VO2 peak but also with greater declines in fasting glucose, and greater increases in insulin sensitivity. The association of higher fasting glucose with smaller increases in VO2 peak may be reflective of the fact that higher levels of inflammatory markers and visceral fat are associated with lower VO2 peak in women with PCOS [23], but is unexpected given that insulin sensitivity appeared to be more likely to improve in women with PCOS who had higher fasting glucose at baseline. This may demonstrate that improvements in insulin sensitivity in response to exercise in women with PCOS do not necessarily associate with significant improvements in aerobic capacity.
Only the number of exercise sessions attended and average heart rate achieved were predictors for improvements in insulin sensitivity in women without PCOS. On the other hand, total distance covered during the intervention was associated with improvements in aerobic capacity in women with PCOS. A previous six-month exercise intervention also demonstrated improvements in aerobic capacity in women with PCOS who were most compliant with training [24]. Interestingly, while we observed that women with PCOS covered significantly more distance than those without PCOS in this intervention, the actual intensity or workload of exercise appeared to be significantly lower. Given that a previous study suggests that increasing exercise dose can abolish non-responsiveness [19], these findings highlight the need to maximise lifestyle program compliance and retention in women with PCOS [25], and suggest that during training, motivational strategies that ensure higher levels of exertion are likely to be beneficial.
A key limitation of this study was that there was substantial loss to follow-up resulting in small numbers for comparison of women with and without PCOS and responders and non-responders and therefore we were unable to control for potential confounders in multivariable analyses. There is also potential for bias in our results from this loss to follow-up although we did not identify any differences in characteristics of included and excluded participants. The sample size had only 79% power to detect a 10% difference in the primary outcome of insulin sensitivity between exercise responders and non-responders (based on body fat change) suggesting that it may have been inadequate to detect significant differences in other outcomes. However, we observed 90% adherence amongst those who completed the exercise intervention with no significant differences between women with and without PCOS [4], or between exercise responders and non-responders. Furthermore, participants who completed the intervention demonstrated 95% compliance to prescribed progressions in the exercise. We conducted a short-term exercise intervention and it is unclear whether interventions of longer duration would demonstrate similar associations between baseline characteristics and exercise non-responsiveness. Nevertheless, an analysis of six exercise interventions with a minimum duration of four months reported substantial proportions of metabolic non-responders in all studies, suggesting increasing the duration of an exercise intervention is not sufficient to overcome exercise non-responsiveness [7]. The strengths of our study include the gold-standard assessments for body composition and insulin resistance, as well as objective assessments of exercise adherence and compliance.
Conclusions
Although a high proportion of overweight and obese women with PCOS had small reductions in total body fat following a 12-week exercise intervention, these exercise non-responders significantly reduced relative central adiposity and improved aerobic capacity and insulin sensitivity. Women with PCOS should be encouraged that participation in exercise, regardless of the effect on body composition, it can result in significant improvements in metabolic health.
Supporting information
S1 File. Study protocol.
Description of the protocol for this study.
https://doi.org/10.1371/journal.pone.0182412.s001
(PDF)
S2 File. TREND checklist.
Transparent Reporting of Evaluations with Nonrandomized Designs (TREND) Statement and Checklist for the current study.
https://doi.org/10.1371/journal.pone.0182412.s002
(PDF)
S3 File. Database.
Pre- and post-training data for included study participants.
https://doi.org/10.1371/journal.pone.0182412.s003
(XLSX)
Acknowledgments
We thank Prof Boyd Strauss and staff for providing body composition measures and analysis, and the study volunteers for their contribution to this research.
References
- 1. Teede HJ, Misso ML, Deeks AA, Moran LJ, Stuckey B, Wong J, et al. Assessment and management of polycystic ovary syndrome: summary of an evidence-based guideline. Med J Aust. 2011;195(6):S65. pmid:21929505
- 2. Moran LJ, Hutchison SK, Norman RJ, Teede HJ. Lifestyle changes in women with polycystic ovary syndrome. Cochrane Database Syst Rev. 2011;(2).
- 3. Harrison CL, Lombard CB, Moran LJ, Teede HJ. Exercise therapy in polycystic ovary syndrome: a systematic review. Hum Reprod Update. 2011;17(2):171–83. pmid:20833639
- 4. Harrison CL, Stepto NK, Hutchison SK, Teede HJ. The impact of intensified exercise training on insulin resistance and fitness in overweight and obese women with and without polycystic ovary syndrome. Clin Endocrinol (Oxf). 2012;76(3):351–7.
- 5. Jaatinen TA, Anttila L, Erkkola R, Koskinen P, Laippala P, Ruutiainen K, et al. Hormonal responses to physical exercise in patients with polycystic ovarian syndrome. Fertil Steril. 1993;60(2):262–7. pmid:8339821.
- 6. Bouchard C, An P, Rice T, Skinner JS, Wilmore JH, Gagnon J, et al. Familial aggregation of VO2 max response to exercise training: results from the HERITAGE Family Study. J Appl Physiol. 1999;87(3):1003–8. pmid:10484570
- 7. Bouchard C, Blair SN, Church TS, Earnest CP, Hagberg JM, Häkkinen K, et al. Adverse metabolic response to regular exercise: is it a rare or common occurrence? PLoS ONE. 2012;7(5):e37887. pmid:22666405
- 8. Cassar S, Misso ML, Hopkins WG, Shaw CS, Teede HJ, Stepto NK. Insulin resistance in polycystic ovary syndrome: a systematic review and meta-analysis of euglycaemic—hyperinsulinaemic clamp studies. Human Reproduction. 2016;31(11):2619–31. pmid:27907900
- 9. Stepto NK, Cassar S, Joham AE, Hutchison SK, Harrison CL, Goldstein RF, et al. Women with polycystic ovary syndrome have intrinsic insulin resistance on euglycaemic—hyperinsulaemic clamp. Hum Reprod. 2013;28(3):777–84. pmid:23315061
- 10. Stephens NA, Sparks LM. Resistance to the beneficial effects of exercise in Type 2 diabetes: Are some individuals programmed to fail? J Clin Endocrinol Metab. 2014;100(1):43–52.
- 11. Böhm A, Hoffmann C, Irmler M, Schneeweiss P, Schnauder G, Sailer C, et al. TGF-β Contributes to Impaired Exercise Response by Suppression of Mitochondrial Key Regulators in Skeletal Muscle. Diabetes. 2016;65(10):2849–61. pmid:27358493
- 12. Chmelo EA, Crotts CI, Newman JC, Brinkley TE, Lyles MF, Leng X, et al. Heterogeneity of physical function responses to exercise training in older adults. J Am Geriatr Soc. 2015;63(3):462–9. pmid:25752778
- 13. Myette-Côté É, Archambault-Therrien C, Brochu M, Dionne IJ, Riesco E. Physical fitness improvement in overweight postmenopausal women who do not lose fat mass in response to exercise training. Menopause (New York, NY). 2016.
- 14. Douchi T, Ijuin H, Nakamura S, Oki T, Yamamoto S, Nagata Y. Body fat distribution in women with polycystic ovary syndrome. Obstet Gynecol. 1995;86(4 Pt 1):516–9. Epub 1995/10/01. pmid:7675372.
- 15. Hutchison SK, Stepto NK, Harrison CL, Moran LJ, Strauss BJ, Teede HJ. Effects of exercise on insulin resistance and body composition in overweight and obese women with and without polycystic ovary syndrome. The Journal of Clinical Endocrinology & Metabolism. 2011;96(1):E48–E56.
- 16. Revised 2003 consensus on diagnostic criteria and long-term health risks related to polycystic ovary syndrome. Fertil Steril. 2004;81(1):19–25. pmid:14711538
- 17. Meyer C, McGrath BP, Teede HJ. Effects of medical therapy on insulin resistance and the cardiovascular system in polycystic ovary syndrome. Diabetes Care. 2007;30(3):471–8. pmid:17327307
- 18. Hutchison SK, Teede HJ, Rachon D, Harrison CL, Strauss BJ, Stepto NK. Effect of exercise training on insulin sensitivity, mitochondria and computed tomography muscle attenuation in overweight women with and without polycystic ovary syndrome. Diabetologia. 2012;55(5):1424–34. pmid:22246378.
- 19. Montero D, Lundby C. Refuting the myth of non-response to exercise training: ‘non-responders’ do respond to higher dose of training. J Physiol. 2017; pmid:28133739
- 20. Dumesic DA, Akopians AL, Madrigal VK, Ramirez E, Margolis DJ, Sarma MK, et al. Hyperandrogenism Accompanies Increased Intra-Abdominal Fat Storage In Normal Weight Polycystic Ovary Syndrome Women. J Clin Endocrinol Metab. 2016:jc. 2016–586.
- 21. Britton KA, Fox CS. Ectopic fat depots and cardiovascular disease. Circulation. 2011;124(24):e837–e41. pmid:22156000
- 22. Legro RS, Dodson WC, Kris-Etherton PM, Kunselman AR, Stetter CM, Williams NI, et al. Randomized Controlled Trial of Preconception Interventions in Infertile Women With Polycystic Ovary Syndrome. J Clin Endocrinol Metab. 2015;100(11):4048–58. pmid:26401593
- 23. Giallauria F, Palomba S, De Sio I, Maresca L, Vuolo L, Savastano S, et al. Inflammatory markers and visceral fat are inversely associated with maximal oxygen consumption in women with polycystic ovary syndrome (PCOS). Clin Endocrinol (Oxf). 2009;70(3):394–400.
- 24. Randeva HS, Lewandowski KC, Drzewoski J, Brooke-Wavell K, O’Callaghan C, Czupryniak L, et al. Exercise decreases plasma total homocysteine in overweight young women with polycystic ovary syndrome. J Clin Endocrinol Metab. 2002;87(10):4496–501. pmid:12364425
- 25. Moran LJ, Lombard CB, Lim S, Noakes M, Teede HJ. Polycystic ovary syndrome and weight management. Womens Health. 2010;6:271.