Obesity and polycystic ovary syndrome

Obesity is a common feature of PCOS. A meta-analysis of cross-sectional and retrospective studies illustrated there is a greater prevalence of obesity in PCOS; BMI >30 kg/m² (relative risk 2·77; 95 % CI 1·88, 4·10; PCOS n 4165; controls n 4885) compared with healthy women⁽ Reference Lim, Davies and Norman ^{20 )}. The variability of data reporting the prevalence of obesity in women with PCOS is likely to reflect geographical, environmental and population variability between studies. There is also uncertainty as to whether PCOS causes weight gain and obesity or if obesity is linked to the development of PCOS⁽ Reference Hoeger and Oberfield ^{21 ,} Reference Teede, Joham and Paul ^{22 )}. Furthermore, clarification of the role of adiposity in PCOS entails measurements other than the measurement of BMI.

A meta-analysis clearly indicated that overweight or obese women with PCOS had significantly worse metabolic and reproductive outcomes when compared with healthy-weight women with PCOS⁽ Reference Lim, Norman and Davies ^{23 )}. Palomba et al.⁽ Reference Palomba, de Wilde and Falbo ^{24 )} confirmed that obesity has a bidirectional relationship with PCOS and noted that excessive weight gain could unmask a latent PCOS condition. IR promotes the development of visceral adiposity in women with PCOS, whereas the role of androgens is less clear⁽ Reference Spritzer, Lecke and Satler ^{25 )}. Tosi et al. ⁽ Reference Tosi, Di Sarra and Kaufman ^{26 )} have recently shown adiposity to be correlated with IR and free testosterone levels (fat mass determined by dual-energy X-ray absorptiometry (DXA), IR with euglycaemic clamp and serum free testosterone by liquid chromatography–MS; n 100 women with PCOS). However, insulin sensitivity, but not adiposity, was found to be an independent predictor of free testosterone concentrations. The central:peripheral fat ratio identified by DXA was shown to be related to androgen levels in twenty-four women with PCOS⁽ Reference Godoy-Matos, Vaisman and Pedrosa ^{27 )}. Details of the complex relationship between androgens and obesity are gradually unfolding⁽ Reference Rojas, Chávez and Olivar ^{28 )}; adipocytes seem to be prone to hypertrophy when exposed to androgen excess, and both adipose tissue hypertrophy and hyperandrogenism are related to IR⁽ Reference Barber and Franks ^{29 )}. What is known is that weight reduction, even as little as 5 %, improves PCOS symptoms in overweight or obese women with PCOS⁽ Reference Moran, Pasquali and Teede ^{30 ,} Reference Haqq, McFarlane and Dieberg ^{31 )}.

Whether women with PCOS present with increased abdominal adiposity compared with women in the general population continues to be debated. A diagnostic method of abdominal obesity has not yet been defined and thus a range of methods is used. Visceral adiposity index (VAI) is a simple surrogate marker of visceral adiposity which utilises BMI, waist circumference, TAG and HDL measurements and has been increasingly used since it was established in 2010⁽ Reference Amato, Giordano and Galia ^{32 )}. Anthropometric studies that include waist circumference measurements have shown that, regardless of BMI, women with PCOS typically have an increased abdominal fat distribution⁽ Reference Yucel, Noyan and Sagsoz ^{33 –} Reference Herriot, Whitcroft and Jeanes ^{36 )}. A meta-analysis by Lim et al. ⁽ Reference Lim, Davies and Norman ^{20 )} also reported that women with PCOS had increased prevalence of central obesity (risk ratio 1·73; 95 % CI 1·31, 2·30; PCOS n 1191; controls n 2396) (Fig. 1).

Fig. 1 Meta-analysis⁽ Reference Amato, Galluzzo and Finocchiaro ^{113 –} Reference Hahn, Tan and Sack ^{118 )} of the prevalence of central obesity in women with and without polycystic ovary syndrome (PCOS) (taken from Lim et al. ⁽ Reference Lim, Davies and Norman ^{20 )}). M-H, Mantel–Haenszel.

When body composition has been assessed by MRI, amounts of visceral adipose tissue have been found to be similar in women with PCOS and controls with comparable BMI (n 31), despite increased waist:hip ratios⁽ Reference Manneras-Holm, Leonhardt and Kullberg ^{37 )} in the women with PCOS. However, it is worth noting that in this study a number of participants were only 1 week after ceasing the insulin-sensitising agent, metformin, which has been shown to reduce visceral adiposity in women with PCOS⁽ Reference Velazquez, Mendoza and Hamer ^{38 ,} Reference Pasquali, Gambineri and Biscotti ^{39 )}. Until relatively recently, studies infrequently reported PCOS phenotypes, which are known to have an impact on metabolic risk. Panidis et al.⁽ Reference Panidis, Tziomalos and Misichronis ^{40 )}, in a relatively small study (n 100 PCOS), reported significantly higher VAI in the hyperandrogenism and oligo/anovulation phenotype, when compared with other phenotypes. Additionally, Androulakis et al.⁽ Reference Androulakis, Kandaraki and Christakou ^{41 )} reported that women with PCOS with menstrual disorders had significantly higher VAI when compared with PCOS without menstrual disorders. A more recent, albeit relatively small, study⁽ Reference Bil, Dilbaz and Cirik ^{15 )} reported a higher VAI in PCOS women with hyperandrogenism and oligo/annovulation compared with other phenotypes. Further research is clearly warranted with more accurate assessments of adiposity for different PCOS phenotypes to help clarify the link between obesity and PCOS.

Visceral adipose tissue is an important endocrine organ, producing and releasing several adipokines; any specific differences in resistin, leptin and visfatin in PCOS women compared with controls remain uncertain; however, a clear role for adiponectin in PCOS has been reported. Adiponectin is down-regulated in obesity and is associated with IR; it also has strong insulin-sensitising, anti-inflammatory and anti-diabetic functions. Adiponectin circulates in different polymer complexes; it has been suggested that its actions are mediated primarily by the high-molecular-weight form of adiponectin⁽ Reference Liu and Liu ^{42 )}. In 2014⁽ Reference Li, Huang and Zhong ^{43 )}, a meta-analysis based on thirty-eight trials (n 1944 PCOS; n 1654 healthy controls) highlighted significantly lower total adiponectin levels in PCOS women compared with healthy controls, independent of BMI (weighted mean difference −2·67; 95 % CI −3·22, −2·13). Visceral fat secretes less adiponectin compared with subcutaneous fat⁽ Reference Motoshima, Wu and Sinha ^{44 )}, so the proposed increased visceral adiposity in women with PCOS could be a reason behind decreased high-molecular-weight adiponectin levels. Metformin, a common treatment option for PCOS, was associated with significantly elevated serum adiponectin concentrations in women with PCOS (standard mean differences −0·43, 95 % CI −0·75, −0·11)⁽ Reference Kong, Niu and Zeng ^{45 )}. In addition, it is likely that adiponectin levels are different in the PCOS phenotypes as suggested by Cankaya et al.⁽ Reference Cankaya, Demir and Aksakal ^{46 )}; in a small study (n 40), adolescents with androgen excess PCOS phenotype had significantly lower high-molecular-weight adiponectin levels compared with non-hyperandrogenic women. Testosterone has been found to contribute to high-molecular-weight adiponectin variance⁽ Reference O’Connor, Phelan and Tun ^{47 )}; however, in a 2009 meta-analysis testosterone levels were not associated with adiponectin levels⁽ Reference Toulis, Goulis and Farmakiotis ^{48 )}. Further high-quality studies are needed to help clarify the role of testosterone and adiponectin in PCOS.

Obesity is evidently complex and the location of excess adiposity is key and all too rarely measured in combination with presenting PCOS phenotypes. In particular, visceral fat plays a key role in generating the insulin-resistant state which leads to impaired glucose tolerance (IGT) and is a significant and independent risk factor for T2DM⁽ Reference Barber, McCarthy and Wass ^{49 )}.

Insulin resistance in polycystic ovary syndrome

A link between PCOS and IR was first highlighted in 1980, whereby obese women with PCOS were shown to have an increased insulin response to an oral glucose tolerance test compared with obese controls⁽ Reference Burghen, Givens and Kitabchi ^{50 )}. Prevalence rates of IR have been reported between 44 and 85 %⁽ Reference Stepto, Cassar and Joham ^{10 ,} Reference Diamanti-Kandarakis and Dunaif ^{51 )}. This variability is in part due to differences in PCOS phenotype and ethnicity⁽ Reference Bil, Dilbaz and Cirik ^{15 ,} Reference Daan, Louwers and Koster ^{17 )}. Reports of the prevalence of IR are limited by the methods used to determine insulin sensitivity; the euglycaemic insulin clamp technique is the most accurate method to diagnose IR, but its high cost limits use; thus the surrogate markers, fasting insulin and homeostasis model assessment of insulin resistance (HOMA-IR)⁽ Reference Levy, Matthews and Hermans ^{52 )}, are often reported.

IR and compensatory hyperinsulinaemia are fundamental to the exacerbation of PCOS symptoms and metabolic complications⁽ Reference Moran and Norman ^{53 )}. Several studies have shown that the majority of women with PCOS, both lean and overweight, have a form of IR that is intrinsic to the syndrome⁽ Reference Stepto, Cassar and Joham ^{10 )}. Researchers have shown that obesity and PCOS have a separate and synergistic relationship with IR⁽ Reference Dunaif and Thomas ^{54 )}. Defects in both insulin sensitivity and secretion are present in lean women with PCOS⁽ Reference Dunaif, Segal and Futterweit ^{55 )}. The pathogenesis of IR in PCOS is complex and incompletely understood, though it is agreed there is a post-binding defect in insulin receptor signalling⁽ Reference Diamanti-Kandarakis and Dunaif ^{51 )}. There is an increasing body of evidence to suggest a genetic susceptibility to PCOS and its associated IR as well as evidence to suggest tha PCOS may also have fetal origins due to androgen exposure at critical periods of growth or due to intra-uterine growth restrictions⁽ Reference Obermayer-Pietsch, Trummer and Schwetz ^{4 )}.

IR is a common feature in annovulatory women with PCOS; however, it is not as frequently observed in ovulatory women with PCOS, even when hyperandrogenism is present⁽ Reference Baptiste, Battista and Trottier ^{56 )}. Bil et al.⁽ Reference Bil, Dilbaz and Cirik ^{15 )} reported significantly greater HOMA-IR in the androgen excess PCOS phenotype compared with non-hyperandrogenic women (n 100). Some older studies have failed to demonstrate IR in lean women with PCOS⁽ Reference Nestler, Jakubowicz and de Vargas ^{57 ,} Reference Wijeyaratne, Balen and Barth ^{58 )}, although this may be accounted for by the inability to distinguish between the different PCOS phenotypes which resulted in the inclusion of women who were ovulatory.

Chronic inflammation and its role in IR in PCOS continues to be investigated⁽ Reference Victor, Rovira-Llopis and Bañuls ^{59 )}. The increase in multiple markers of inflammation such as leucocyte count, C-reactive protein, IL-18, protein-1, and monocyte chemo-attractant in addition to increased oxidative stress and endothelial dysfunction are evidence that PCOS is commonly coupled with low-level systemic inflammation⁽ Reference Escobar-Morreale, Luque-Ramírez and González ^{60 )}. However, there are limited studies that assess the presence of chronic inflammation independent of adiposity in women with PCOS. Circulating markers of oxidative stress are abnormal in women with PCOS independent of weight excess and IR⁽ Reference Murri, Luque-Ramírez and Insenser ^{61 )}. The close association between oxidative stress and inflammation make it difficult to determine their individual contribution to PCOS and the mechanism of oxidative stress-induced IR remains unclear⁽ Reference Zuo, Zhu and Xu ^{62 )}. Oxidative stress and inflammation markers are positively correlated with androgen levels in PCOS patients⁽ Reference González, Minium and Rote ^{63 )}; thus they seem to contribute to hyperandrogenaemia, but detailed interactions still need to be identified, and as of yet few investigations have been done.

Insulin acts directly on the ovary by stimulating the thecal cells to increase androgen production and by activation of the cytochrome P450c17α⁽ Reference Nestler, Jakubowicz and de Vargas ^{57 ,} Reference Conway, Avy and Rumsby ^{64 )}. Hyperinsulinaemia further exacerbates the pathogenesis of PCOS by inhibiting the production of insulin-like growth factor-1 (IGF-1) binding protein in the liver, leading to elevated circulating levels of IGF-1, which in turn stimulates ovarian thecal cell androgen production⁽ Reference Hopkinson, Sattar and Fleming ^{65 )}. Hyperinsulinaemia reduces hepatic production of sex hormone-binding globulin, thus increasing free testosterone levels⁽ Reference Cassar, Misso and Hopkins ^{66 )}. In combination, hyperinsulinaemia and hyperandrogenaemia can disrupt follicle growth; this is accompanied by menstrual irregularity, anovulatory sub-fertility and accumulation of immature follicles⁽ Reference Diamanti-Kandarakis and Dunaif ^{51 )}. Diamanti-Kandarakis & Dunaif⁽ Reference Diamanti-Kandarakis and Dunaif ^{51 )} in a review described the modulating action of insulin on ovarian steroidogenesis as well as the importance of the insulin signalling pathway in the control of ovulation.

Women with PCOS commonly have postprandial dysglycaemia⁽ Reference Legro, Kunselman and Dodson ^{67 ,} Reference Ehrmann, Kasza and Azziz ^{68 )}, which reflects peripheral, primarily skeletal muscle, IR. In lean women with PCOS hyperinsulinaemia is often evident postprandially but not in the fasted state⁽ Reference Morales, Laughlin and Butzow ^{69 )}. IR and compensatory hyperinsulinaemia have also been proposed as a cause of reactive hypoglycaemia in women with PCOS. A cross-sectional study of sixty-four lean young women with PCOS reported the rate of reactive hypoglycaema to be 50 %⁽ Reference Altuntas, Bilir and Ucak ^{70 )}. A survey of clinical practice (n 88) also indicated a high prevalence of hypoglycaemia; self-reported in 71 % of women with PCOS⁽ Reference Herriot, Whitcroft and Jeanes ^{36 )}. A recent study also indicated a higher prevalence of reactive hypoglycaemia in PCOS (n 88) compared with age- and BMI-matched controls (n 34), though a much lower prevalence of 17 %⁽ Reference Mumm, Altinok and Henriksen ^{71 )}. Robust studies investigating hypoglycaemia are sparse.

Impaired glucose tolerance and type 2 diabetes mellitus

The prevalence of IGT and T2DM in women with PCOS is substantially higher compared with age- and weight-matched healthy women; a meta-analysis of cross-sectional and cohort studies reported an OR of 2·54 (95 % CI 1·44, 4·47) for OGT and 4·00 (95 % CI 1·97, 8·10) for T2DM⁽ Reference Moran, Misso and Wold ^{72 )}. Differences in the diagnostic criteria for IGT and PCOS were evident, and there were inadequate data to allow for PCOS phenotype sub-analysis; additionally, differences in ethnicity and BMI have resulted in variable risk estimates between PCOS studies. Studies in the USA have indicated a prevalence of 23–35 % for IGT and 4–10 % for T2DM⁽ Reference Legro, Kunselman and Dodson ^{67 ,} Reference Ehrmann, Kasza and Azziz ^{68 )}, whereas in non-US studies prevalence rates in women with PCOS are elevated although not all to the same extent, with prevalence rates of 16 % for IGT and 2·5 % for T2DM described in an Italian cohort⁽ Reference Gambineri, Pelus and Manicardi ^{73 )}. In the US studies, even lean women with PCOS had increased rates of IGT and T2DM compared with healthy controls⁽ Reference Legro, Kunselman and Dodson ^{67 )}, the onset of IGT occurring in their thirties and forties which is earlier than the normal population⁽ Reference Pelusi, Gambineri and Pasquali ^{74 )}. A more recent longitudinal study following women for at least 10 years reported that the age-standardised prevalence of diabetes at the end of follow-up was 39·3 %, significantly higher than in healthy women (prevalence of 5·8 %)⁽ Reference Gambineri, Patton and Altieri ^{75 )}.

The Royal College of Obstetrics and Gynaecology 2014 guidelines^{( 76 )} state that women presenting with PCOS with a BMI ≥ 25 kg/m² and lean women with PCOS, but who have additional risk factors such as advanced age (> 40 years), personal history of gestational diabetes or family history of T2DM, should have a 2-h post-75 g oral glucose tolerance test.

Metabolic syndrome and non-alcoholic liver disease in polycystic ovary syndrome

PCOS commonly includes many metabolic syndrome components, such as abdominal obesity, dyslipidaemia, hyperglycaemia and hypertension, which predispose women with PCOS to the development of T2DM and CVD. There is a higher prevalence of the metabolic syndrome in women with PCOS compared with women in the general population⁽ Reference Moran, Misso and Wild ^{77 )}, estimated to be between 34 and 46 %, based on studies of Caucasian women with PCOS from the USA using the National Cholesterol Education Program Adult Treatment Panel III criteria⁽ Reference Ehrmann, Liljenquist and Kasza ^{78 –} Reference Apridonidze, Essah and Iuorno ^{80 )}. Women diagnosed with PCOS using National Institutes of Health criteria are more likely to have a higher prevalence of the metabolic syndrome compared with controls, compared with women diagnosed with PCOS using the ESHRE/ASRM guidelines⁽ Reference Moran, Misso and Wold ^{72 ,} Reference Panidis, Macut and Tziomalos ^{81 )}. Studies that clearly indicate the prevalence of the metabolic syndrome in differing PCOS phenotypes are lacking.

IR and obesity are strong predictors of non-alcoholic fatty liver disease (NAFLD)⁽ Reference Marchesini, Brizi and Morselli-Labate ^{82 ,} Reference Bugianesi, Gastaldelli and Vanni ^{83 )}; thus it is not surprising that there is a higher prevalence of NAFLD in women with PCOS compared with age- and BMI-matched controls; prevalence rates ranging from 27 to 62 %⁽ Reference Cerda, Perez-Ayuso and Riqelme ^{84 –} Reference Ramezani-Binabaj, Motalebi and Karimi-Sari ^{87 )}. This variation can be attributed to differences in laboratory diagnostic criteria (for example, liver function tests, liver imaging studies, liver biopsy) and differences in study populations relating to ethnicity, age, BMI and PCOS diagnostic criteria⁽ Reference Vassilatou ^{88 )}. Vassilatou⁽ Reference Vassilatou ^{88 )} has highlighted the increasing evidence that hyperandrogenism is related to NAFLD in PCOS and suggested that hyperandrogensim should be considered as an additional link in the synergy with obesity and IR for the development of NAFLD in PCOS. However, sub-analyses with the PCOS phenotypes were not reported and would seem prudent for further evaluation.

Of clinical importance, the prevalence of non-alcoholic steatohepatitis (NASH) in women with PCOS is particularly high⁽ Reference Koroli, Fatima and Chandra ^{89 )}. Bariatric surgery has shown beneficial outcomes on NASH⁽ Reference Weiner ^{90 )}. With the exception of weight loss through lifestyle modification and encouraging exercise in obese women, few evidence-based effective treatments specifically target NAFLD/NASH in women with PCOS⁽ Reference Chen and Ho ^{91 )}.

Dietary influences of the metabolic aspects of polycystic ovary syndrome

With no cure for PCOS, clinical management focuses on treating the presenting symptoms. Lifestyle management is advocated as the primary therapy in overweight and obese women with PCOS since even a modest weight loss of just 5–10 %, without medical intervention, improves many of the symptoms associated with PCOS⁽ Reference Moran, Pasquali and Teede ^{30 ,} Reference Moran, Hutchison and Norman ^{92 ,} Reference Marzouk and Sayed Ahmed ^{93 )}.

The optimal method of achieving sustainable weight loss is under constant debate; studies demonstrating the clinical benefits of weight loss in women with PCOS incorporate a variety of methods to reduce energy intake and achieve a negative energy balance⁽ Reference Moran, Ko and Misso ^{94 )}. Behaviour-change strategies have long been recognised as being instrumental in managing chronic conditions, particularly weight management. However, lifestyle modification programmes are often reported to have low compliance and high dropout rates⁽ Reference Domecq, Prutsky and Mullan ^{95 )}. Furthermore, many overweight or obese women with PCOS are not following a diet that would promote weight loss⁽ Reference Jeanes, Reeves and Gibson ^{96 )}. Bariatric surgery, on the other hand, in women with PCOS has resulted in substantial weight loss and marked improvement of multiple biochemical abnormalities⁽ Reference Jamal, Gunay and Capper ^{97 ,} Reference Eid, Cottam and Velcu ^{98 )}. A recent meta-analysis of thirteen studies (2130 women) reported a reduction in the incidence of PCOS preoperatively from 45·6 %, to 6·8 % at 12-month follow-up⁽ Reference Skubleny, Switzer and Gill ^{99 )}.

Dietary modification to improve IR may produce benefits greater than those achieved by modest weight loss and would also be suitable for lean IR women with PCOS. Modifying the carbohydrate content and fatty acid content of the diet have both been proposed as methods to improve insulin sensitivity. The concept of a low-glycaemic index (GI) diet aims to reduce the glycaemic load (GL) and hence insulin response to ingested foods and drinks. Greater improvement in insulin sensitivity and regular menses were reported after weight loss with a low dietary GI approach (n 29) compared with weight loss through conventional healthy eating (n 20)⁽ Reference Marsh, Steinbeck and Atkinson ^{100 )}. In this study, women were followed for 12 months or until they achieved 7 % weight loss. Barr et al. ⁽ Reference Barr, Reeves and Sharp ^{101 )} also demonstrated an improvement in insulin sensitivity after following an isoenergetic diet with a reduction in dietary GI in twenty-one women with PCOS over a 12-week intervention period. Furthermore, one study⁽ Reference Mehrabani, Salehpour and Amiri ^{102 )} reported a reduction in IR after a 12-week hypoenergetic diet, high in protein (30 % protein: n 23), with a low GL compared with a conventional hypoenergetic healthy eating intervention (15 % protein: n 26). A low-GL diet in lean women with PCOS (n 41) resulted in a reduction in waist circumference⁽ Reference Herriot, Whitcroft and Jeanes ^{36 )}. These small studies suggest a benefit for promoting lower GI dietary advice to lean women with PCOS, many of whom are also insulin resistant. However, exploring the mechanisms behind the intervention and assessing IR with the euglycaemic clamp technique and evaluating the effect of modifying dietary GI in a subgroup of IR women with PCOS is prudent to improving the evidence base for dietary recommendations.

Inflammation is implicated in the pathogenesis of PCOS; dietary constituents such as SFA can influence inflammatory mediators, such as TNF-α and IL-1β and disrupt insulin signalling⁽ Reference Lyons, Kennedy and Roche ^{103 ,} Reference Gonzales ^{104 )}. Conversely, MUFA and PUFA do not interrupt metabolism and inflammation to the same extent. SFA intake has been shown to influence glucose metabolism through the altering of insulin signalling and cell membrane function, with a diet high in SFA associated with a decrease in insulin sensitivity when compared with a high-MUFA diet⁽ Reference Lyons, Kennedy and Roche ^{103 )}. Of interest, it has been reported that 68 % of women with PCOS had a total fat intake greater than a 35 % contribution to total energy, with SFA intake accounting for 12 % of total energy, substantially exceeding the reference nutrient intake (<10 % energy)⁽ Reference Barr, Hart and Reeves ^{105 ,} Reference Hart, Barr and Reeves ^{106 )}. Thus, a reduction of total fat intake should focus on lowering SFA intake concurrently with the encouragement of the maintenance of MUFA and PUFA in women with PCOS. There is some evidence to show that by reducing the consumption of saturated fat⁽ Reference Turner, Cooney and Kraegen ^{107 )} and reducing the GI and GL of the diet⁽ Reference Turner, Cooney and Kraegen ^{107 )}, insulin sensitivity can be improved in insulin-resistant populations; however, there remains a paucity of studies in women with PCOS⁽ Reference Schwingshackl and Hoffmann ^{108 )}.

Advanced glycation end products (AGEs) are elevated in PCOS⁽ Reference Diamanti-Kandarakis, Piperi and Kalofoutis ^{109 )}; these pro-inflammatory molecules trigger a state of intracellular oxidative stress and inflammation after binding to their cell membrane receptors (RAGE). They are highly reactive molecules formed by non-enzymic reactions of carbonyl group of carbohydrates with free amino groups of proteins, nucleic acids or lipids. AGEs within the body are from endogenous production or ingested in high amounts from cooked fast food⁽ Reference Goldberg, Cai and Peppa ^{110 )}. A small study⁽ Reference Tantalaki, Piperi and Livadas ^{111 )} (twenty-three PCOS participants) has shown that a 2-month intervention of a low dietary intake of AGEs is associated with improved levels of serum testosterone, oxidative stress and HOMA-IR in PCOS. The role of dietary AGEs as mediators of metabolic and reproductive alterations in PCOS is an exciting and emerging area of dietary modification for PCOS.

An area yet to be explored in PCOS is the role of the gut microbiota⁽ Reference Canfora, Jocken and Blaak ^{112 )}, which could influence glucose homeostasis and insulin sensitivity by fermenting indigestible dietary components to produce SCFA. These SCFA have been shown to beneficially modulate adipose tissue, skeletal muscle and liver tissue function. Furthermore, research, including well-designed robust studies, is needed to evaluate the mechanisms behind the improvements observed with dietary modification and inherent with all weight-reduction interventions to improve adherence to interventions and weight maintenance.