Many features of polycystic ovary syndrome (PCOS), such as elevated androgens, insulin resistance and inflammation are likely to affect cognition. However, the majority of research on the psychological effects of PCOS has focused on symptoms of anxiety and depression in relation to distress caused by living with symptoms of PCOS (Barry et al., 2018). As PCOS is the most common endocrine disorder affecting women of reproductive age, with prevalence ranging from 4 – 20% (Azziz et al., 2016, Diamanti-Kandarakis and Dunaif, 2012, March et al., 2010, Norman et al., 2007), knowledge of its effects on cognition would contribute to more informed treatment and increased quality of life of a large portion of women. In this review, we synthesize and evaluate the existing literature on PCOS and cognition, and highlight aspects of PCOS symptomatology that require further study with respect as to how they affect cognition. Whereas prior reviews on PCOS symptomatology have looked at clinical treatment of PCOS (Azziz et al., 2016), metabolic effects of oral contraceptives in PCOS (Mendoza, Simoncini & Genazzini, 2014), the role of central brain circuits in PCOS symptoms and etiology (Moore & Campbell, 2017), and knowledge of PCOS gained through animal models (Abbot et al., 2016; Roland and Moenter, 2014, Walters, 2016), this is the first review to examine studies on the cognitive function of women with PCOS.

PCOS is a complex and highly heterogenous condition, classified into four phenotypes depending on diagnostic criteria and variation in core diagnostic. The two main criteria used for the diagnosis of PCOS are the National Institutes of Health (NIH) criteria, which define PCOS in terms of oligomenorrhea and hyperandrogenism (Fauser et al., 2012) and the Rotterdam criteria, which offer a slightly broader definition, with two out of the following three symptoms being sufficient for a diagnosis: oligomenorrhea, hyperandrogenism, and polycystic ovaries on ultrasound (Rotterdam, 2004). The two most common phenotypes, present in two thirds of women with PCOS, involve either all three of the above listed criteria (phenotype A), or oligomenorrhea and hyperandrogenism (phenotype B). The remaining third of women with PCOS have either polycystic ovaries and hyperandrogenism (phenotype C) or oligomenorrhea and polycystic ovaries on ultrasounds (phenotype D) without the remaining third diagnostic characteristic (Lizneva et al., 2016, Moore and Campbell, 2017).

Additionally, PCOS is often comorbid with other metabolic conditions such as insulin resistance (El Hayek et al., 2016), obesity (Apridonidze et al., 2005), type 2 diabetes mellitus, cardiovascular disease and cancers of reproductive organs (Azziz et al., 2016), and chronic inflammation (Escobar-Morreale et al., 2011). Hormonal, metabolic and inflammatory symptoms of PCOS persist after the menopausal transition (Puurunen et al., 2011). Women with PCOS also have elevated rates of mental health concerns such as depression (Hollinrake et al., 2007), anxiety (Dokras, 2012) and eating disorders (Rassi et al., 2010).

Three key features of PCOS – elevated androgens, insulin resistance and inflammation – as well as their effects on cognition are briefly reviewed below in order to highlight the complexity of potential interplay between PCOS and cognition and to inform the subsequent summary and discussion of the current literature on PCOS and cognition.

Elevated androgens levels are one of the hallmarks of PCOS, observed in 80-90% of women with the syndrome (Azziz et al., 2009, Hull, 1987) and frequently discussed in terms of their potential effects on cognition in women with PCOS. Effects of testosterone on cognition in PCOS are suggested to be both organizational, due to elevated prenatal exposure to testosterone and anti-mullerian hormone (Barry et al., 2010, Tata et al., 2018) and activational due to elevated circulating levels of testosterone later in life (Barry et al., 2018). The organizational hypothesis is supported by the fact that intrauterine levels of androgens increase the chance of female offspring developing PCOS symptoms later in life (Cresswell et al., 1997, Xita and Tsatsoulis, 2006). Furthermore, female children of mothers with PCOS have elevated prenatal levels of testosterone (Barry et al., 2010, Blank et al., 2006) and are likely to develop PCOS-consistent symptoms during puberty such as hyperandrogenism, hyperinsulinemia, and increased ovarian volume (Maliqueo et al., 2009, Sir-Petermann et al., 2009).

Prenatal exposure to testosterone may affect cognitive abilities by acting on and altering cortical networks relevant to cognition (Barry et al., 2013; Schoning et al., 2007; Weiner et al., 2004). As well, prenatal exposure to elevated levels of androgens may lead to altered function of the hypothalamic–pituitary–gonadal axis later in life, causing an activational effect of testosterone with organizational origins (Barry et al., 2010, Barry et al., 2018). Further support for the potential activational effects of testosterone on cognition in women with PCOS comes from studies which show a positive correlation between circulating levels of testosterone and performance on visuospatial tasks in postmenopausal women (Stangl et al., 2011) as well as those demonstrating improvement on visuospatial tasks in young women who were administered testosterone (Aleman et al., 2004). Thus, there are grounds to expect a difference in cognitive performance between women with PCOS who have elevated androgens and women with lower/more typical testosterone levels.

Albeit not a part of the diagnostic criteria, another common feature of PCOS with likely effects on cognition is insulin resistance (IR). IR is estimated to affect between 44-70% of women with PCOS (Ciampelli et al., 2005, de Paula Martins et al., 2007, Deugarte et al., 2005, Ducluzeau et al., 2003) and women with PCOS have more severe IR than weight-matched women without PCOS (Dunaif et al., 1989). While it has been suggested that IR is primarily associated with obesity in PCOS (Diamanti-Kandarakis & Dunaif, 2012), studies show that IR is equally prevalent across PCOS phenotypes (Bannigida et al., 2020, Jamil et al., 2015); lean PCOS patients show IR both before the initial manifestation of symptoms during puberty and in adulthood (Dunaif et al., 1989, Dunaif et al., 1992, Sir-Petermann et al., 2009). In fact, there is strong support for IR and associated chronically high levels of insulin (hyperinsulinemia) as a cause of PCOS. Insulin promotes ovarian androgen production directly (Nestler et al., 1998), through interaction with luteinizing hormone (Barbieri et al., 1986), and through suppression of sex-hormone-binding globulin production (Galloway et al., 2001, Plymate et al., 1988). Further evidence for the role of IR and hyperinsulinemia in causing PCOS symptoms comes from trials of metformin, a medication that increases the body’s sensitivity to insulin. Taking metformin lowers androgen levels and increases frequency of ovulation (Baillargeon et al., 2004, Baillargeon and Nestler, 2006) in women with PCOS.

In addition to being linked to chronic conditions that increase risk of cognitive decline (i.e., diabetes, obesity, vascular impairment), hyperinsulinemia resulting from IR may have a direct negative effect on cognitive function (Bosco et al., 2011, Cholerton et al., 2013, Craft, 2009). Prolonged peripheral hyperinsulinemia reduces the number of insulin receptors at the blood-brain barrier, leading to downregulation of insulin transport into the brain (Schwartz et al., 1990, Wallum et al., 1987). IR is associated with worse verbal fluency performance in mixed-sex cohorts (Ekblad et al., 2017), and negatively correlated with executive function in women in their mid-30s (Bove et al., 2013) and mid-40s (Schuur et al., 2010). Furthermore, postmenopausal women with IR show reduced resting state functional connectivity between regions of the prefrontal cortex and the hippocampus, when compared to age-matched women with lower insulin levels (Kenna et al., 2013). This highlights the importance of studying IR in the context of PCOS and cognition, and suggests that the effects of IR on cognition in PCOS may extend past the reproductive years, into menopause.

Finally, women with PCOS tend to have chronic low-grade inflammation as evidenced by elevated levels of inflammatory markers such as C-reactive protein (CRP; Escobar-Morreale et al., 2011), interleukin-18 (IL-18; Escobar-Morreale et al., 2004, Yang et al., 2011), monocyte chemoattractant protein-1 (MCP-1; González et al., 2009, Hu et al., 2011) and white blood count (Herlihy et al., 2011, Orio et al., 2005, Ruan and Dai, 2009) – all of which are likely to affect cognitive performance (Au et al., 2016). A meta-analysis found that serum CRP levels were 96% higher in women with PCOS compared to healthy controls, or up to 102% higher in BMI-matched PCOS cohorts and controls (Escobar-Morreale et al., 2011).

This elevation in CRP in women with PCOS may be especially pertinent to cognition with increasing age. Higher CRP levels are associated with worse global cognition (Schram et al., 2007) and executive function (Jefferson et al., 2011) in mixed-sex cohorts of healthy older adults. Additionally, CRP is negatively associated with verbal memory performance in adults aged 30 – 80 (Teunissen et al., 2003) and episodic memory, verbal memory, attention, and executive function in adults aged 30 – 64 (Beydoun et al., 2018). As mentioned, IL-18 levels are also elevated in women with PCOS, and increased IL-18 production is associated with cognitive decline in patients with Alzheimer’s disease, as evidenced by worse global cognition scores (Bossù et al., 2008). Finally, MCP-1, elevated in women with PCOS (González et al., 2009) even when compared to BMI-matched controls (Hu et al., 2011) is likewise associated with cognitive decline. A recent longitudinal study found that levels of MCP-1 at baseline predicted accelerated cognitive decline in older adults with mild cognitive impairment and Alzheimer’s disease (Lee et al., 2018). Inflammatory markers may thus be important factors to consider when studying cognition in women with PCOS, particularly in midlife and in older cohorts.

With these connections between PCOS-associated symptoms and cognition in mind, we conducted a literature review in order to summarize current knowledge on PCOS and cognition and note the extent of PCOS symptomatology currently studied in relation to cognitive measures.