- Top of page
- Subjects and methods
- Conflict of interest
Adequate nutritional status is a critical determinant of the onset and maintenance of normal reproductive function. Obesity is encountered in 30–70% of women affected by polycystic ovary syndrome (PCOS), one of the most common endocrinopathies in women of fertile age, and its presence significantly modifies both clinical and laboratory expression of the syndrome. Moreover, women with PCOS have been shown to have higher amounts of visceral fat compared to healthy controls, even when they are normal weight. The importance of obesity in the pathophysiology of PCOS is definitely confirmed by the fact that even modest to moderate weight loss may improve all features of PCOS, including insulin resistance, lipid abnormalities, androgen excess, menses and ovulatory rates, and favour the complete recovery of the PCOS phenotype in some of them.
The association between abnormal dietary habits and history and risk for PCOS has not been widely examined, and available data are still sparse and contradictory. Most researchers found that daily energy intake and diet composition did not differ in the majority of PCOS women compared to controls,[5-8] although minor discrepancies, regarding specific nutrients and food categories,[5-7, 9] or subgroups of women,[5, 7, 8] were reported. Interestingly, an increased risk of anovulatory infertility has been associated with a higher consumption of animal proteins, total carbohydrates and foods with a high glycaemic index, low-fat dairy foods and cola beverages in different reports from the Nurses’ Health Study cohort.[10-13] However, whether women with PCOS have different patterns of dietary intake and food preferences with respect to the unaffected population is far from being established.
Diet is also a major source of advanced glycosylated end products (AGEs) and other oxidants. Meals cooked at high temperatures and under dry conditions have the highest AGE content, especially if the fat content is high. Dietary AGEs contribute to a state of elevated oxidative stress and inflammation and have been shown to play a role in promoting diabetes, insulin resistance and atherosclerosis in mice.[16, 17] Recent data have shown that oxidative stress may also be involved in the pathophysiology of PCOS. Both AGEs and advanced oxidation protein products were in fact found to be higher in women with PCOS than in healthy controls,[19, 20] and a positive correlation between androgen and AGE levels was reported.
With this background, we performed a case–control study in overweight or obese women with PCOS to investigate their dietary intake (including AGEs) and habitual food choices, the relationship with hormones and metabolism and the potential interaction by PCOS or control status.
- Top of page
- Subjects and methods
- Conflict of interest
In this study, performed in overweight or obese women with PCOS, we found several marginal differences in diet composition with respect to age- and BMI-matched controls. Total energy and macronutrient intake was similar between the two groups, but a higher intake of fibres and a lower per cent energy intake from lipids were found in the PCOS group. Unexpectedly, no difference was found in AGE intake in two subgroups of the same cohorts, in spite of a series of studies suggesting a potential impact of dietary AGE intake on their tissue and blood levels. Taken together, our findings support the concept that the usual diet of adult overweight or obese women with PCOS may only marginally differ with respect to a comparable non-PCOS population.[6-13] Whether diet may differ in normal weight individuals with or without PCOS still remains poorly defined, although some studies found marginal but significant differences in daily energy intake in a small cohort of lean women with PCOS with respect to control women.
The higher fibre content in women with PCOS with respect to controls in our study contrasts with previous findings by Wild et al., reporting a lower fibre intake in women PCOS with respect to normoandrogenic women, and with those by Chavarro et al., reporting that fibre dietary content was totally unrelated to ovulatory infertility. We suggest that these data should be interpreted with caution, because of different dietary habits among different countries, the specific phenotype of the subjects investigated and the small differences found in all the studies.
By contrast, we suggest that the analysis of food choice may reveal some interesting findings. Compared to controls, the PCOS group showed in fact a significantly higher consumption of starchy sweets, cheese and oil, a lower intake of cooking fats and a tendency to a higher consumption of starchy foods with high-glycaemic index. The importance of specific foods in determining metabolic and hormonal patterns has been suggested by some studies. Specifically, two studies linked anovulatory infertility or PCOS to high-glycaemic index food intake, although a third study did not confirm the association. It is worth mentioning that the efficacy of diets including low-glycaemic index foods in improving insulin resistance and glucose control has been widely demonstrated. Very recently, the impact of specific dietary and lifestyle factors, rather than daily energy intake and macronutrient proportion, has been shown to be independently associated with long-term weight gain. Although we could not detect any significant intergroup difference in the amounts of dietary oligosaccharides, the available data, including our own presented here, support the need for additional longitudinal studies, reporting on quality, rather than quantity, of carbohydrates.
Our study also shows that cheese intake was higher in PCOS women, which is a relatively new finding, somewhat in contrast with previous data supporting an association between high-fat diary food consumption and reduced risk of anovulatory infertility.
In addition, we found that women with PCOS had a higher intake of oil (mostly olive oil) than controls, which in some way parallels their significant lower intake of cooking fats. Because of the well-defined beneficial properties of olive oil on lipid and carbohydrate metabolism, inflammation indexes and oxidative stress, all markers that are often altered in PCOS, one would expect an inverse association between olive oil intake and PCOS status. Therefore, a further evaluation by a more appropriate methodology should be performed.
The principal aim of the study was to investigate the relationship between hormonal and metabolic parameters and diet composition or food choices. Interestingly, we found that hormones (specifically, androstenedione, SHBG and FAI) and metabolic parameters (specifically, insulinAUC and HDL-cholesterol) significantly correlated with selected nutrients and food categories, particularly in the PCOS group, and to a lesser extent in controls, which suggests a potential role for PCOS or control status on such relationships. The statistical analysis we applied showed the existence of different types of behaviour of PCOS or control groups as regards the impact of diet composition on laboratory parameters, particularly for the relationship between FAI and carbohydrate per cent fraction and for that of HDL-cholesterol with carbohydrates (intake and per cent fraction) and protein per cent fraction. As far as food choice is concerned, disparate behaviour by PCOS vs control status was disclosed in the relationships of testosterone with legume and wine intakes, androstenedione with fruit intake, insulinAUC with fish, low-fat milk and chocolate intakes and HDL-cholesterol with fish intake. Although these data are relatively new and should be interpreted with caution, they support the concept that specific components of diet (such as carbohydrates), as well as specific food categories, may directly or indirectly influence metabolic parameters (such as glucose-stimulated insulin response and HDL-cholesterol) more in the PCOS than in the control group, which in turn suggests that PCOS women may be more sensitive to the dangerous effects of unhealthy dietary behaviour. Whether this may play a pathological role in PCOS, particularly at its onset in childhood and adolescence, represents an attractive speculative hypothesis.
The strength of this study is the inclusion of a homogeneous group of well-defined PCOS women and age- and BMI-matched women with normal menses and without any sign of androgen excess and the exclusion of possible interferences by metabolic affections or drugs, including oral contraceptives. Moreover, the assessment of dietary habits was performed by means of the 7 days diary, which is recognized as offering a good estimate of quantity and quality of food, and an interview with our dieticians was included, to minimize the risk of underreporting. Still, none of the traditional dietary assessment instruments avoids the risk of underestimate, and it is possible that patients may change their diet knowing that they are been evaluated. A weakness of this study is that the estimate of AGE content in foods was performed only in small subsets of PCOS and control women and was based on the available estimates by US researchers, which may have led to inaccuracies. In addition, we did not measure circulation blood AGEs. Moreover, the choice of overweight or obese subjects may have influenced our results, because most studies demonstrating a relationship between AGE levels and PCOS have been performed in normal weight women.[19, 20]
In conclusion, this study, conducted on a group of overweight/obese women with polycystic ovary syndrome compared with age- and body mass index-matched controls, showed that diet does not differ between the two groups as regards energy, macronutrient and advanced glycosylated end products intake, except for a lower percentage of lipids and a higher intake of fibre by polycystic ovary syndrome women. Polycystic ovary syndrome women were characterized by a higher consumption of cheese and high-glycaemic index starchy sweets with respect to controls and a preference for raw oil rather than other cooked fats. Overall, our findings give little support to the hypothesis of a strong dependence of polycystic ovary syndrome status on nutritional factors.