Coffee consumption and the risk of polycystic ovary syndrome: evidence from a case-control study

Abstract


Introduction
Polycystic ovary syndrome (PCOS) is the most common reproductive endocrine disorder in women during reproductive ages and one of the main causes of female infertility and menstrual dysfunction. PCOS affects about 5-20% of women in reproductive age according to Rotterdam criteria (1). The clinical manifestation of PCOS is highly heterogeneous and mainly characterized by hyperandrogenism, ovarian dysfunction, and polycystic ovarian changes (2). However, the underlying etiology of the disease remains not fully understood. At present, it is believed that PCOS may be the result of genetic and environmental factors, including dietary and lifestyle factors (3).
In addition to infertility and irregular menstruation, the majority of women with PCOS may also present with a series of metabolic abnormalities, including obesity, hyperandrogenism, dyslipidemia, insulin resistance/hyperinsulinemia, lower sex hormone-binding globulin (SHBG) levels, which are causal risk factors for PCOS and harmful to their long-term health (4)(5)(6)(7). Among them, hyperandrogenism and insulin resistance, which are regarded as the primary endocrine characteristics of PCOS, have a critical role in the occurrence and development of PCOS(8). In addition, the SHBG level is known to decrease in patients with PCOS and is considered an indicator of hyperandrogenemia and abnormal metabolism(6). Due to these metabolic dysfunctions, these women also have an increased risk of developing type 2 diabetes, endometrial cancer, cardiovascular disease, and ovarian cancer (9)(10)(11).
Coffee has become one of the most common beverages worldwide. It contains more than 1000 components in addition to caffeine and is rich in many phytochemicals with various physiological effects (12). Some biological studies have shown that coffee consumption may be associated with PCOS due to the potential effect of caffeine or other components on hormonal regulation. Caffeine and chlorogenic acid in coffee, have been recognized to improve insulin sensitivity (13,14). It was reported that increasing caffeine intake was inversely associated with bioavailable testosterone while positively correlated with SHBG (15,16). From the point of view that coffee intake affects the regulation of hormones related to PCOS, there might also be a relation between coffee consumption and the risk of PCOS. However, there is still a lack of relevant epidemiological studies.
In this present study, we sought to investigate whether associations exist between coffee consumption and the risk of PCOS, independent of possible confounding factors, including sociodemographic characteristics, lifestyle factors, and other relevant variables.

Methods
We conducted a hospital-based case-control study to investigate PCOS-related risk factors in Chongqing from July 2018 and January 2020. All women with PCOS diagnosed for the rst time by endocrine gynecology in Chongqing Hospital of Traditional Chinese Medicine were invited to participate in the study. As for the control group, healthy women with non-PCOS who underwent a physical examination at the hospital during the same period were included. Besides, due to the di culty of collecting controls in the hospital, we also collected a small amount of data from healthy female graduate students in Chongqing Medical University as part of the control group.
Patients were diagnosed with PCOS if they have at least two of the following three characteristics Patients were excluded from the study based on the presence of any of the following: congenital adrenal cortical hyperplasia, Cushing's syndrome, thyroid dysfunction, ovarian tumors or adrenal gland disorders that could ultimately cause ovulation disorders and abnormal androgen secretion, any other malignancy, current usage of any hormone therapy and pregnancy history in the past three months, lactation.
The research protocol was approved by the Ethics Committee of Chongqing Hospital of Traditional Chinese Medicine and all participants gave informed consent before participation in the study.

Data Collection
A structured questionnaire was administered and eligible participants completed the questionnaire. The questionnaire contained the following information: (1)sociodemographic characteristics, such as age, height, weight, monthly per-capita income, educational level, occupation type, ethnic group, residence area, marital status; (2) lifestyle of the individual including smoking and drinking habits, dietary habits; (3) sleep quality of the participants, assessed by the Pittsburgh Sleep Quality Index (PSQI) which included 19 items. Information on coffee consumption was obtained by a closed-ended question with four frequency categories: never, ≤ 1 cup/week, 2-3 cups/week and > 3cups/ week.

Statistical analysis
The IBM SPSS version 26.0 was employed for all statistical analyses. Shapiro-Wilk test was used to determine whether continuous data conform to normal distribution. All continuous variables following normal distribution were expressed as mean ± standard deviation, and categorical variables were presented as frequency and percentages. The Student's t-test and chi-square test were used to determine differences in sociodemographic variables between PCOS and control groups. Logistic regression analysis model was used to identify the association between coffee consumption and the risk of PCOS and to calculate the odds ratios (ORs) and 95% con dence intervals (CIs) with the rst category as the reference group. Models were adjusted for potential confounding factors, including age, BMI, marital status, other dietary habits, occupation type, night shift status, and PSQI total score. These variables are either clinically relevant or signi cantly associated with PCOS or suspected of affecting coffee consumption. A two-sided P-value of < 0.05 was considered to be statistically signi cant.

Sociodemographic characteristics of the study subjects
A total of 206 PCOS cases and 206 controls were enrolled in the study. Table 1 shows the distribution of sociodemographic characteristics of the study subjects. Most of the participants in both groups were of Han ethnicity and lived in urban areas, with no signi cant difference between the two groups. There were also no signi cant differences in monthly income level, occupation type and night shift status between PCOS cases and the control group. Compared with the control group, the mean age was signi cantly lower (30.09 vs 28.56, P = 0.001) and the mean BMI was higher (21.25 vs 22.53, P < 0.001) in PCOS cases. The majority of the study participants in both groups were married and the proportion of married women in the PCOS was signi cantly higher than that in the control group (82.5% vs 68.4%, P < 0.001). Additionally, a higher proportion of the control group was found with university or higher university degree in comparison to PCOS (62.1% vs 42.2%, P < 0.001). Further,

Discussion
Our study provides epidemiological evidence on the inverse association between coffee consumption and PCOS in the Chinese population. The risk of PCOS had a decreasing trend with increased coffee consumption. Evidence from this dose-response relationship supports the biological reasonableness of our ndings that higher coffee consumption may be a protective factor for PCOS.
So far, only a few studies have involved the relationship between coffee or caffeine intake and PCOS. In a randomized clinical trial involving 34 patients with PCOS investigating the effect of green coffee supplementation on androgen levels, supplementation of 400mg green coffee a day for 6 weeks was determined to be signi cantly associated with a reduction in free testosterone level, triglyceride, and cholesterol levels compared with the control group(18). Two case-control studies conducted in New York state and Spanish using food frequency questionnaire (FFQ) found that the caffeine intake of patients with PCOS was lower than that of the control group, but the difference was not statistically signi cant between PCOS and controls (19,20). Nevertheless, these results should be interpreted with caution, as the authors of both studies mentioned that the sample size they included was relatively small, which may lead to a decline in the power to detect any association. Moreover, in addition to coffee, caffeine also exists naturally in tea leaves and is often added to other beverages as a food additive (21). It should be noted that coffee contains thousands of components, not just caffeine. Therefore, similar caffeine intake between the two groups did not indicate that there was no association between coffee and the risk of PCOS. Their results also provided the possibility that substances other than caffeine might play a vital role in the pathogenesis of PCOS. Further research is needed in the future to determine how coffee and its components affect PCOS.
Although it was previously believed that coffee should be consumed less or not at all for the possible adverse effects on human health, recent researches have shown that coffee may be bene cial because of its bioactive components (14,22). Caffeine, along with chlorogenic acid(CGA), a mainly phenolic component in coffee, has been shown to be bene cial in maintaining metabolic homeostasis (23)(24)(25). Ma et al(26)studied the effect of CGA on obesity and insulin resistance in C57BL/6 mice. Obese mice were treated by 100 mg/kg CGA intra-peritoneal for 21 weeks. CGA suppressed diet-induced obesity and hyperinsulinemia, and improved insulin sensitivity when compared with the control group. Two prospective cohort studies found that not only caffeinated coffee but also decaffeinated coffee consumption was associated with body weight reduction (27,28). A meta-analysis concluded that both caffeinated and decaffeinated coffee had protective effects against Type-2 diabetes and improved glucose metabolism (29). SHBG is a transport carrier protein that binds estrogen and androgen. Lower concentrations of SHBG are related to hyperandrogenism and insulin resistance, and affects long-term prognosis in PCOS patients (6, 30).Data from a large cohort study indicated that women who drank ≥ 4 cups of coffee a day had higher concentrations of SHBG (31). Taken together, our study and these studies have shown that coffee could play a role in improving metabolic disorders and reducing the risk associated with PCOS.
However, there were some limitations in our study. First, due to the nature of case-control study, we could not determine a causal relationship between coffee consumption and PCOS. Secondly, since some of the control data were collected in the university, this could skew the distribution of educational levels towards one group. Although we have adjusted the educational level as a confounding factor in the multivariate logistic regression model, there may still be some deviations to some extent in our results. We recommend a more thorough and fair distribution of data in future studies. Moreover, the identi cation of coffee consumption was evaluated from a self-administered questionnaire, and not by a veri ed standardized questionnaire such as food frequency questionnaires, which may lead to some measurement errors due to the differences among people in the size of self-reported coffee cups. Additionally, we did not collect information about coffee types (caffeinated or decaffeinated), speci c composition and content, which limited our preliminary exploration of whether caffeine or other substances in coffee play a critical role in the development of PCOS.

Conclusion
In conclusion, our study provided epidemiological evidence that higher coffee consumption reduces the risk of polycystic ovary syndrome. However, considering the scarcity of studies on coffee consumption and PCOS, as well as the inherent limitations of our study, more prospective cohort studies in the future are needed to verify our results.

Declarations
Funding This work was supported by funds from Natural Science Foundation of Chongqing (nos: cstc2019jcyj-msxmX0749,cstc2019jxjl130030,cstc2018jxjl130065) and Intelligent Medicine Research Project of Chongqing Medical University (NO.YJSZHYX202010).
Availability of data and materials The datasets used and/or analyzed in the present study are available from the corresponding author on reasonable request.
Ethics approval and consent to participate

Consent for publication
All co-authors have seen and approved the nal version of the paper and have agreed to its submission for publication.