The first clinical manifestations of PCOS typically appear in adolescence as irregular periods, excessive hair growth, acne, weight gain, and other health issues. When diagnosing PCOS, it is important to rule out any other origins of the phenotype, such as adult-onset congenital adrenal hyperplasia, hyperprolactinemia, and androgen-secreting neoplasms. When combined with other symptoms, hyperandrogenism is recognized as a critical diagnostic factor (Azziz et al., 2006). Globally, there are differences in the methods and criteria used for diagnosis. The National Institutes of Health (NIH) criteria identify PCOS in about 6% of women of reproductive age if they exhibit both hyperandrogenism and oligo/amenorrhea (McCartney and Marshall, 2016). For PCOS diagnosis, the Rotterdam criteria demands two out of the three criteria (Rotterdam ESHRE/ASRM-Sponsored PCOS Consensus Workshop Group, 2004), and the diagnosis can be made with just one polycystic ovary present (Balen et al., 2003). Metabolic risk, specifically insulin resistance, is significantly higher in the NIH group compared to the Rotterdam-defined group. Using the Rotterdam definition raises the prevalence of PCOS in women to up to 20% from 6% using the NIH definition (Broekmans et al., 2006). Laboratory tests measuring free and bioavailable testosterone and androstenedione are used to diagnose hyperandrogenism. Ultrasound scans, blood testing, and clinical symptoms contribute to PCOS diagnosis. These techniques, however, can be expensive, time-consuming, and imprecise. Additionally, a number of PCOS symptoms, such as irregular menstruation and acne, coincide with the typical pubertal stage (Kiconco et al., 2023). The ovaries of women with PCOS are often bigger and contain more follicles. The most popular radiological technique for detecting PCOS is ultrasound; however, the diagnostic cut-offs have been an ongoing debate and have recently been revised. Additionally, multi-follicular ovaries are a common feature around menarche, which limits the utility of ultrasonography to diagnose PCOS for early diagnosis. Currently, the diagnostic cut-offs are dependent on fluctuating laboratory ranges specified by test manufacturers and are based on arbitrary percentiles, frequently from poorly characterized cohorts. This limits the precision of the diagnostic process. A recent study demonstrated that the PCOS diagnostic cut-offs correspond to lower percentiles than conventional PCOS diagnostic criteria (Kiconco et al., 2023). Thus, both physicians and patients are unsatisfied with available diagnosis and treatment options due to diagnostic challenges, delayed and unsatisfactory diagnosis experiences, and less-than-optimal treatment plans (Hoeger et al., 2021). This underscores the pertinent need for precision diagnostics using a combination of novel biomarkers and machine learning methods, with rigorous validation in larger, multi-ethnic, and well-characterized adolescent cohorts. The development of ovarian follicles is regulated by the anti-Müllerian hormone (AMH) during the menstrual cycle. The number of ovarian antral follicles is closely correlated with AMH levels, making a high level of AMH a potential PCOS biomarker (Butt et al., 2022). The broad application of such novel biomarkers is hampered by the absence of an international standard for blood hormone tests, making it challenging to establish consensus criteria (Dumont et al., 2015). Approaches utilizing machine learning have the potential for the identification of PCOS. An ensemble machine learning system was shown by Danaei Mehr and Polat (2022) to be highly sensitive and accurate in PCOS prediction. Using image data from ovarian ultrasound scans, a machine learning algorithm could accurately identify PCOS (Suha and Islam, 2022).

Individual circumstances determine the course of treatment for PCOS; similar to diagnosis, there is no single treatment option. As lifestyle changes are shown to improve both the reproductive and metabolic features of PCOS, they are used as the first line of ovulation induction in PCOS patients (Karimzadeh and Javedani, 2010). While rigorous aerobic exercise has been proven to improve body composition, cardiorespiratory fitness, and insulin resistance, no particular diet composition has shown promise in treating the symptoms of PCOS (Cowan et al., 2023). Medical interventions are directed primarily to treat core PCOS symptoms—irregular cycles, hirsutism, and anovulation. Hormonal birth control pills are frequent clinical medications that help regulate menstrual periods and lower testosterone levels. Metformin is advised for obese PCOS women to enhance insulin sensitivity and reduce insulin and blood glucose levels. Studies on the effects of inositol supplementation on metabolic profiles and reduction in hyperandrogenism have demonstrated promise (Cowan et al., 2023). Other pharmaceutical drugs, including ovulatory stimulants like letrozole or clomiphene citrate, are also applied; they induce ovulation by blocking estrogen activity and stimulating the pituitary gland to produce more luteinizing hormone (LH) and follicle-stimulating hormone (FSH), which, in turn, triggers ovulation. When other therapies do not work or are not effective, more intrusive procedures such as bariatric surgery or laparoscopic ovarian drilling are used, which involve destroying a portion of the tissue that produces androgens in an effort to restore ovulation (Hoeger et al., 2021).

A systematic classification of PCOS may enable better diagnosis, treatment, and management of the condition based on each woman’s unique needs. Two of the three are suggested by the Rotterdam criteria: polycystic ovarian morphology, hyperandrogenism, and oligo- or anovulation. This results in four unique phenotypes: 1) complete, i.e., all three requirements are met; 2) classic, i.e., anovulation and hyperandrogenism are present; 3) ovulatory, characterized by polycystic ovarian shape and hyperandrogenism; and 4) non-androgenic, defined by the existence of polycystic ovarian morphology and anovulation (Broekmans et al. 2006). A recent study distinguished between two primary PCOS subgroups using hormonal and metabolic markers. The reproductive subtype was identified by comparatively low body mass index (BMI) and insulin levels, as well as greater levels of LH and sex hormone-binding globulin (SHBG) and more severe infertility and irregular menstruation. In addition to having lower levels of SHBG and LH and higher levels of BMI, glucose, and insulin, the metabolic subtype was also linked to an increased risk of androgen excess symptoms, including hirsutism, acne, and hair loss. Significantly, distinct genetic connections were found for these subgroups, indicating potential differences in their origins. Variants in ovarian function-related genes, including PRDM2, IQCA1, BMPR1B, and CDH10, were linked to the reproductive subtype. Variants in glucose metabolism-related genes, including KCNH7 and FIGN, were linked to the metabolic subtype (Dapas et al., 2020). Thus, the primary cause of anovulation in obese PCOS women is likely hyperinsulinemia, whereas increased blood LH concentrations are more likely to produce anovulation in lean PCOS women (Barber, 2022).

Comparing PCOS women to BMI-matched controls, more accumulation of abdominal fat is observed. Nonetheless, imaging studies have revealed that the distribution of fat was comparable in PCOS and control groups, indicating that central obesity might exist independently of PCOS (Zhu et al., 2021). Visceral fat has high metabolic activity and secretes a variety of hormones and cytokines that can disrupt insulin signaling, induce inflammation, and result in a chronic low-grade inflammatory and insulin-resistant state (Zhao et al., 2023). Adipose tissue secretes leptin, which acts on the hypothalamus to inhibit appetite and increase energy expenditure. Higher amounts of leptin in the bloodstream in obese women may cause a long-term downregulation of LEPR in the brain. Research conducted in vitro has demonstrated that leptin impacts steroidogenic pathways in granulosa cells and reduces the synthesis of progesterone and estrogen in a dose-dependent manner (Brannian et al., 1999). A total of 65%–95% of women with PCOS, including the majority of overweight and obese women and more than half of women of normal weight, have compensatory hyperinsulinemia and insulin resistance. It indicates that insulin resistance in PCOS is tissue-specific rather than a general phenomenon. While the ovary, adrenal glands, and liver continue to be insulin-sensitive, skeletal muscle and adipose tissue develop insulin resistance, leading to decreased glucose absorption and increased lipolysis, respectively. Hyperinsulinemia is a compensatory reaction to insulin resistance common in PCOS. The ovaries and adrenal glands are stimulated indirectly by hyperinsulinemia, which leads to an increase in androgen production. More precisely, when LH is stimulated, excess insulin increases the production of androgen in the ovarian theca cells, leading to follicular arrest and, thereby, anovulation. Nevertheless, not all obese women with oligomenorrhea exhibit hyperandrogenism, but they have increased LH pulse frequency, similar to PCOS women (Yoo et al., 2006). SHBG secreted by the liver, the main protein that binds testosterone in the serum, is suppressed by hyperinsulinemia (Silvestris et al., 2018). The National Institute for Health and Care Excellence states that since there is a substantial inverse relationship between SHBG levels and fasting serum insulin in obese women, measuring SHBG levels can serve as a proxy measure for the severity of hyperinsulinemia in PCOS-affected women (Daka et al., 2013). Consequently, following weight loss, there was a substantial negative correlation between SHBG concentrations and insulin levels. Remarkably, insulin did not suppress the SHBG level in cases of extreme insulin resistance (Akin et al., 2007).

PCOS symptoms can have a significant influence on a person’s life and may affect their mood or general well-being. Accordingly, it is associated with a higher risk of several mental health disorders (Table 1), including depression, anxiety, eating disorders, low self-esteem, and negative body image. In women with PCOS, hirsutism, i.e., excess terminal hair development in male-typical body areas, had the greatest effect on the quality of life metrics relating to health (Khomami et al., 2015).

Endometrial carcinoma is the sixth most common cancer in women, and because of their higher frequency of obesity and extended oligomenorrhea or amenorrhea, women with PCOS appear to have a three- to four-fold increased risk of endometrial cancer. Thinning or irregular menstruation can lead to endometrial accumulation and thickening. In fact, a prognostic signature for endometrial cancer was shown to be derived from genes involved in the production of steroid hormones (Zhang et al., 2023).

Non-alcoholic fatty liver disease (NAFLD) is more common in PCOS patients (odds ratio 2.54). Even after controlling for confounding variables, the prevalence of NAFLD is greater in PCOS women with hyperandrogenism (classic phenotype) than in women with PCOS without hyperandrogenism. In women with PCOS, serum androgens were independent predictors of NAFLD (Rocha et al., 2017). Both PCOS and NAFLD have enhanced pathways linked to immunity and inflammation (Chen et al., 2022). A different study that used Biobank data discovered that there was less evidence linking genetically predicted NAFLD to a higher risk of PCOS development. The relationship between NAFLD and PCOS may be mediated by sex hormones and fasting insulin (Liu et al., 2023).

A systematic review and meta-analysis showed a clear link between PCOS and obstructive sleep apnea, particularly in obese individuals (Vgontzas et al., 2001). By increasing upper airway collapsibility and/or reducing the sensitivity and responsiveness of the ventilatory chemo-receptors, hyperandrogenism, and low progesterone levels may contribute to the pathophysiology of obstructive sleep apnea. It is unclear, nevertheless, how PCOS and obesity interact, leading to sleep apnea (Kahal et al., 2017).