Lifestyle and environmental contributions to ovulatory dysfunction in women of polycystic ovary syndrome

Polycystic ovary syndrome (PCOS) is a heterogeneous endocrinopathy with a prevalence of approximately 5.5–19.9% [1]. It is characterized by irregular cycles, clinical or biochemical hyperandrogenism, and polycystic ovarian morphology (PCOM). Moreover, it is the most common cause of chronic anovulatory infertility [2]. As reported previously, 90.30–95.28% of PCOS women diagnosed by Rotterdam criteria were characterized with ovulatory dysfunction [3, 4]. Additionally, oligo−/anovulatory (OA) subjects showed more severe hormonal and metabolic derangements compared to PCOS women with normal ovulation [4, 5]. According to Chinese diagnostic criteria, ovulatory dysfunction was even considered as necessity for the diagnosis of the syndrome [6]. However, the pathogenesis has not been fully understood.

Environmental factors and genetic variants are both well accepted as main etiologic factors of PCOS. Several susceptibility genes have been reported via candidate gene screening and GWAS [7‐9]. However, environmental components, being another key factor in the pathophysiology, have not been well documented. It was shown that environmental endocrine disruptors (EEDs) could perturb the hormonal regulation of the hypothalamic pituitary ovarian axis. Hence, EEDs are supposed to act as steroid-agonists and/or antagonists [10‐12]. Previous studies also indicated a correlation between PCOS and contact history of EEDs, as well as changes of environment and lifestyle [13, 14]. However, the exact effect of environmental factors on ovulatory dysfunction in PCOS women remains unclear.

In our study, lifestyles and EED exposure of PCOS women with and without OA were analyzed in order to determine the association between environmental factors and ovulatory dysfunction of PCOS.

Basic characters were shown in Table 1. PCOS women were older (PCOS-OA group, PCOS-non-OA group, Control group: 31.11 ± 3.71 vs 31.56 ± 3.49 vs 29.81 ± 3.77 yrs., P<0.01) and had a higher BMI (24.90 ± 4.09 vs 25.31 ± 4.39 vs 22.93 ± 3.86 kg/m2, P<0.01) compared to controls. Results of univariate analysis were shown in Table 2. Frequency of snoring (29.30% vs 18.10% vs11.50%, P < 0.01) and smoking (37.70% vs 28.10% vs 12.20%, P < 0.01) were different among three groups and were highest in the PCOS-OA group (Table 2, Fig. 1). After multivariate adjustment of age and BMI, these differences remained statistically significant except for the comparison of snoring between PCOS-non-OA and control group (Table 3). PCOS women, both with OA and without OA, preferred to drink tea more often than control women, even after adjustment (Table 3). Diet composition was only recorded in PCOS women. The comparison showed that PCOS-OA women had higher rate of meat favorable diet than their non-OA counterparts (54.60% vs 41.30%, P < 0.01; 1.69(1.28, 2.23), P < 0.01 after adjustment). No differences were found in exercise duration (P = 0.08), frequency of insomnia (P = 0.18), alcohol consumption (P = 0.92) amongst the three groups.

The higher usage of plastic tableware (38.30% vs 28.10% vs 25.40%, P < 0.01) and indoor decoration (32.10% vs 24.80% vs 16.80%, P < 0.01) were found in PCOS-OA women compared with either PCOS-non-OA or control women (Table 2, Fig. 1). Similar trends were found after adjusting age and BMI. Furthermore, indoor decoration showed a gradient trend among three groups [OR: 1.40, 95%CI: 1.03–1.91, P = 0.03 for PCOS-OA vs PCOS-non-OA, OR: 2.53, .95%CI: 1.78 to 3.57), P < 0.01 for PCOS-OA vs Control, OR: 1.86, 95%CI: 1.19 to 2.92, P < 0.01 PCOS-non-OA vs control, Table 3]. Air refresher usage (15.60% vs 9.30% vs 12.90%, P < 0.01), and cooking oil fume contact (53.00% vs 43.00% vs 32.30%, P < 0.01) were more frequent in PCOS-OA women compared to their non-OA counterparts. But these differences were non-significant after adjustment (Table 3). We also compared education, occupation, sleep duration, source of drinking water, pesticide free of fruits and vegetables, and microwave usage among three groups, but found no difference (data not shown).

In the present study, smoking, snoring, meat favorable diet, and usage of plastic tableware and indoor decoration were found to be associated with oligo or anovulation in PCOS women. Moreover, smoking and indoor decoration contact seemed to have a dose-dependent effect. Tea drinking was positively associated with PCOS but not ovulatory dysfunction. These factors indicated that lifestyle and environmental endocrine disruptors may associated with the pathophysiology of PCOS.

Smoking was found to be associated with ovulatory dysfunction in a dose-dependent way in our study. Several studies have observed that smoke toxicants can disrupt folliculogenesis, leading to premature luteinization of preovulatory follicle. These toxicants can also decrease oocytes maturation and ultimately accelerated depletion of the primordial follicle pool [20, 21] . However, other studies found there was no correlation between smoking and OA in PCOS women [22, 23]. The main difference between previous studies and our study is the definition of smoking. In our study, exposure to smoking included both active and passive smoking, whereas other studies only included active smoking.

Snoring was found to be another lifestyle indicator associated with OA in PCOS. It is one of the earliest symptoms of obstructive sleep apnea (OSA), which is a complex disorder characterized by repetitive collapse of the pharyngeal airway during sleep [24]. Previous studies have reported that PCOS is associated with a reduction in REM sleep stage time and increased risk for OSA [25‐27]. Alerted reproductive hormone secretion (i.e. high androgen and low estrogen levels) might contribute to the higher prevalence of OSA [28]. Furthermore, it was indicated that low estradiol-to-testosterone ratio was associated with chronic oligo-anovulatory cycles in PCOS [29].

The present study also confirmed the correlation of meat favorable diet and the risk of ovulatory dysfunction, which was consisted with findings in animal studies. Previous studies showed that a decrease of primordial and Graafian follicles in high-fat feeding rat [30]. Insulin resistance (IR) was supposed as one of the potential mechanisms. Studies in both animal and human beings demonstrated that hyper-caloric diet will induce IR and β-cell dysfunction [31, 32]. IR was found to interrupt follicle development through inhibiting hypothalamic positive feedback to estradiol (E2). Besides, the direct lesion on ovulation through kisspeptin down-expression and granulosa cell apoptosis were also indicated as underlying mechanism in high-fat diet feeding mice [33, 34].

Nowadays, more and more people pay attention to the impact of EEDs on female reproductive health. We found plastic tableware usage and indoor decoration were associated with PCOS ovulatory dysfunction in the present study. It was demonstrated that Bisphenol A (BPA), the main component of plastic containers, had a positive association with endocrine disturbances in PCOS [35, 36]. It acted as a potent binder of sex hormone-binding globulin [37] and had a bidirectional interaction effect with androgens [38, 39], which may result in perturbed ovarian steroidogenesis and folliculogenesis. However, recent studies indicated that the BPA exposure didn’t alter ovulation in mice [40, 41]. Hence, more in vivo studies were needed to replicate the association of BPA exposure and ovulatory dysfunction, and to elucidate the mechanism. Except for BPA, the plasticizer and its substitute diisononyl phthalate (DiNP), were other EED’s found in many consumers. DEHP or DiNP exposure was demonstrated to accelerate primordial follicle recruitment by up-regulating the PI3K pathway, and lead to prolonged estrous cyclicity and subfertility in female mouse. The pathological effect could last even long after cessation of exposure [42, 43] .

We also found a positive association between indoor decoration and PCOS ovulatory dysfunction. However, there were limited studies focusing on the association between indoor decoration and PCOS [13]. It was found that organic solvents, the most important constituents for indoor decoration, had a negative influence on glucose metabolism impairment [44, 45]. Hundreds of chemicals were detected in organic solvents that could activate tumor necrosis factor α (TNFα), one of the most famous proinflammatory cytokines, which resulted excessive hepatic glucose formation, inhibited muscular glucose uptake, and impaired insulin sensitivity [46, 47] through different pathways [47‐49]. There was a definite association between insulin resistance and interrupted follicle development. This may explain the correlation between indoor decoration and PCOS.

The strength of the present study was that it has large samples. Besides, it evaluated the association of lifestyle and exposure to environmental pollutants with ovulatory dysfunction in PCOS women systemically, which would provide important indication on next-step etiological study of impacts of environment factors on ovulation dysfunction and PCOS. However, it still had several limitations. Firstly, as a self-report questionnaire, social desirability cannot be eliminated. Secondly, only a minority of Chinese women smoke due to cultural habits. Thus, the sample size is too small to conduct further subgroup analysis to examine the contribution of active or passive smoking. Besides, the exact caloric intake per day was not recorded in our data. Comparison of diet composition only cannot provide further does-dependent evidence. Specific design studies involving caloric intake were needed to explore the effect of diet on oligo−/anovulatory in PCOS.

Ovulatory dysfunction in PCOS is related to unhealthy lifestyle and environmental pollutants exposure. Hence, lifestyle modification as the first-line therapy for PCOS especially women with OA should be promoted more vigorously.