Maternal polycystic ovarian syndrome and offspring psychopathology and neurodevelopment

Abstract

Polycystic ovarian syndrome (PCOS) is a common female endocrinologic condition that affects both the metabolic and reproductive systems and is the most frequent cause of anovulatory infertility. It is also associated with a range of psychiatric outcomes in individuals, including bulimia nervosa, schizophrenia, bipolar disorder, depression, anxiety, and personality disorders. At the same time, evidence suggests that hyperandrogenism, the characteristic trait of PCOS, may impair fetal neurodevelopment. Epidemiological studies have linked maternal PCOS with a variety of behavioral and psychiatric conditions in offspring including autism spectrum disorder and attention deficit hyperactivity disorder. In this review, we explore evidence for potential underlying biological mechanisms that might explain these observed associations, discuss the complex interplay between genetics and various environmental factors across generations, and highlight avenues for future research.

Graphical Abstract

Graphical Abstract.

Relevant biological mechanisms underlying associations between maternal PCOS and child psychopathology and neurodevelopment. PCOS, polycystic ovarian syndrome; EDC, endocrine-disrupting chemical.

Introduction

Polycystic ovarian syndrome (PCOS), a condition characterized by hyperandrogenism, insulin resistance, and polycystic ovaries, affects between 4% and 20% of women of reproductive age worldwide (Deswal et al., 2020) and is the most common cause of anovulatory infertility (Teede et al., 2010). Depending on which of the Rotterdam criteria are met, PCOS may be classified into four phenotypes, which manifest as three subtypes: ‘classic’ PCOS (phenotypes A and B) accounts for two-thirds of PCOS cases and is characterized by high BMI, high anti-Mullerian hormone, insulin resistance, and dyslipidemia, while ‘ovulatory’ PCOS, which is characterized by high LH:FSH ratio (phenotype C), and ‘nonhyperandrogenic’ or ‘indeterminant’ PCOS (phenotype D) have approximately equal prevalence (Lizneva et al., 2016). While generally considered to be an endocrinologic and metabolic disorder, PCOS is also associated with a range of psychiatric outcomes in individuals, including bulimia nervosa, schizophrenia, bipolar disorder, depression, anxiety, and personality disorders (Cesta et al., 2016). Increasing evidence suggests that offspring of women with PCOS may also be at elevated risk of psychopathology and neurodevelopmental problems, including behavioral problems, attention deficit hyperactivity disorder (ADHD), autism spectrum disorder (ASD), schizophrenia, and chronic tic disorder (Katsigianni et al., 2019; Dubey et al., 2021).

Potential explanations for cross-generation similarities in PCOS-related neurodevelopmental and psychopathological conditions include genetic inheritance of psychiatric illness with or without endocrine dysregulation, as well as nongenetic transmission based on factors associated with the familial environment. In addition, consistent with the theory of prenatal programming (Gillman, 2005), in utero exposure to elevated androgen levels characteristic of maternal PCOS may affect the organization of the developing fetal brain, with potentially long-term effects on offspring behavior, cognitive, and emotional development. Animal models of induced maternal prenatal hyperandrogenism have demonstrated effects on offspring gene expression in the amygdala, hippocampal neurogenesis, and behavior (Cheng et al., 2019; Risal et al., 2021). Because sex hormones have a broad influence on brain development (Hines, 2008) and some studies have found sex-specific effects of prenatal exposure to endocrine-disrupting chemicals (EDCs) on fetal neurodevelopment (Toledano et al., 2024), offspring outcomes may also manifest differently according to sex.

Figure 1 illustrates multiple potential causal pathways that link PCOS with adverse offspring psychopathology and neurodevelopment, some of which may originate in prior generations. The co-occurrence of hyperandrogenism and psychopathology observed in women with PCOS suggests a possible common biological etiology. Supporting this hypothesis, siblings of women with PCOS have been shown to have both higher androgen levels and higher rates of psychiatric disorders compared with siblings of women without PCOS (Cesta et al., 2016). Both phenomena may result from shared genetics and/or similar intrauterine environment, including the siblings’ exposure to their own mother’s endogenous hormone levels and/or to endocrine disruptors during gestation. However, twin studies have yet to test such a hypothesis.

Figure 1.

Hypothesized intergenerational mechanisms linking maternal PCOS with child psychopathology and neurodevelopment. PCOS, polycystic ovarian syndrome; EDC, endocrine-disrupting chemical; LBW, low birth weight; PTB, preterm birth.

Likewise, the intergenerational transmission of psychopathology and neurodevelopmental problems from mothers with PCOS to their children could be genetic, environmental, or a combination thereof. Both PCOS and the types of psychiatric problems observed in women with the condition have genetic components (Bray and O’Donovan, 2019; Welt, 2021). Characteristics of PCOS such as anovulation, obesity, and insulin resistance may also program fetal brain development via independent pathways (Bliddal et al., 2014; Xiang et al., 2015; Veena et al., 2016; Wang et al., 2016b). Furthermore, offspring behavior may be influenced by caregiving factors as well as by direct exposure to maternal psychiatric symptoms (Phua et al., 2020).

The epidemiologic literature linking PCOS to offspring psychopathology and neurodevelopment is robust and remarkably consistent. With few exceptions, positive associations have been reported for maternal PCOS and offspring ASD, ADHD, anxiety, behavioral problems, communication problems, inattention, chronic tic disorder, hyperactivity/impulsivity, sleeping problems, eating disorder, problem-solving, and intellectual disabilities (Katsigianni et al., 2019; Dubey et al., 2021; Abu-Zaid et al., 2022; Maleki et al., 2022). Some studies have noted sex-specific effects (Cherskov et al., 2018; Dalgaard et al., 2021; Risal et al., 2021; Zhang et al., 2023). In this mini-review, we present evidence for several relevant biological mechanisms that may underlie these associations, discuss the complex interplay between genetics and various environmental factors across generations, and highlight avenues for future research.

Evidence supporting a genetic link

Studies of twins, siblings, and other first-degree relatives demonstrate the presence of a genetic component in PCOS, as rates of hyperandrogenism among family members related to women with PCOS are many times higher than in the general population (Legro et al., 1998). The strongest evidence implicates polymorphisms in the insulin-receptor gene (Shaaban et al., 2021; Velmurugan et al., 2024). Genome-wide association studies (GWAS) have identified associations with single-nucleotide polymorphisms in genes coding for other hormone receptors, such as luteinizing hormone receptor and FSH receptor, as well as for gonadotropin secretion, androgen biosynthesis, metabolic regulation, and follicle development (Welt, 2021; Dapas and Dunaif, 2022). GWAS have shown that PCOS subtypes are associated with distinct genetic variants and that these subtypes tend to run in families (Dapas et al., 2020).

Neuropsychiatric disorders are also heritable, although the number and complexity of these conditions and the likelihood that they are highly polygenic have precluded identification of a consistent set of candidate genes (Bray and O’Donovan, 2019), particularly genes that overlap with those linked to PCOS.

The impact of androgens on the fetal brain

One argument in support of the biological plausibility of intrauterine effects of PCOS on offspring psychopathology and neurodevelopment is built on the role of hormones in fetal brain development. Pregnancy is a critical period characterized by hormonal crosstalk between the mother, placenta, and fetus. Imbalances in maternal hormones, such as sex, stress, and thyroid hormones, can affect fetal brain development, with subsequent influences on psychopathology and neurodevelopment (Hines, 2011; Faa et al., 2014; Miranda and Sousa, 2018). The fetal brain is particularly sensitive to sex hormones, as gonadal steroids direct sexual differentiation of the brain during the prenatal period by influencing neuronal proliferation, maturation, and migration, as well as neurite growth and synaptogenesis (Forest, 1983; Garcia-Segura and Melcangi, 2006; Haraguchi et al., 2012). Cortical regions have high concentrations of sex steroid receptors during the prenatal period that reduce substantially in the postpartum period, supporting the importance of gestation as a critical period for the organizational effects of sex hormones on the brain.

Androgen receptors are found not only in the cortex but also the hypothalamus, hippocampus, media amygdala, and many other brain regions where changes in circulating androgen levels may lead to altered expression of genes that influence neurogenesis, differentiation, and apoptosis (Lombardo et al., 2020). Animal studies have established that sex steroid receptor concentrations are highest in brain regions that are sexually dimorphic in neuroimaging studies (e.g. the amygdala) (Morse et al., 1986; Clark et al., 1988; Goldstein et al., 2001), suggesting that these brain regions might be most sensitive to disruption of sex hormone pathways (Morse et al., 1986; Clark et al., 1988; Goldstein et al., 2001). Some of these brain regions are implicated in ASD and other neurodevelopmental disorders (van Rooij et al., 2018; Boedhoe et al., 2020). Fetal testosterone levels (as measured through amniocentesis) and proxy measures of in utero androgen exposure (e.g. the ratio of the second to fourth digit) have been associated with sex-specific development of gray matter in infancy and childhood (Lombardo et al., 2012; Knickmeyer et al., 2014). Excess fetal androgen exposure can be due to high production by the fetus, for example in cases of congenital adrenal hyperplasia, placental aromatase deficiency, exposure to maternal smoking, or androgen excess from maternal sources, for example in cases of ovarian or adrenal tumors, hyperreactio luteinalis, and PCOS (Hakim et al., 2017). Experimental studies in animals confirm that the latter influences fetal brain development (Hu et al., 2015), but human epidemiological studies that examine maternal androgen excess and brain measures in offspring are sparse. Neuroimaging studies that examine associations of maternal PCOS and androgen levels with offspring brain structure and function are needed to further elucidate potential links.

Thyroid hormones and fetal brain development

Thyroid hormones provide another potential link between PCOS and offspring psychopathology and neurodevelopment. PCOS increases the incidence of some thyroid diseases, including autoimmune thyroiditis and subclinical hypothyroidism (Gaberšček et al., 2015; Kowalczyk et al., 2017; Zhao et al., 2021). Furthermore, some thyroid diseases such as subclinical hypothyroidism increase the incidence of PCOS and exacerbate metabolic and reproductive disorders (Fan et al., 2023).

Since thyroid hormones regulate brain processes, including neuronal migration, differentiation, and myelination, maternal thyroid dysfunction in mothers with PCOS may lead to impaired neurodevelopment in offspring (Jansen et al., 2019). A growing evidence base suggests that maternal thyroid dysfunction, mainly hypothyroidism, is associated with delayed offspring cognitive development and with increased risk of neurodevelopmental disorders, such as ADHD and ASD (Wang et al., 2016a; Miranda and Sousa, 2018; Jansen et al., 2019). Although biologically plausible, this potential role of thyroid hormones in the relation between PCOS and neurodevelopment has not been studied and may be a fruitful avenue for future research.

Maternal obesity/insulin resistance and offspring psychopathology and neurodevelopment

Another possible pathway connecting PCOS to offspring neurodevelopment involves obesity and insulin resistance. PCOS is strongly linked to both conditions, with ∼50% of women worldwide affected by PCOS also experiencing obesity, and an even higher percentage in the USA (Yildiz et al., 2008; Hirschberg, 2009). Epidemiologic research has shown that pre-pregnancy BMI and/or obesity are associated with offspring cognitive function and mental health, and with an increased risk of adverse neurodevelopmental outcomes such as ASD (Kosidou et al., 2017; Shook et al., 2020), ADHD (Rodriguez et al., 2008; Chen et al., 2014; Shook et al., 2020), inattention, negative emotionality (Rodriguez, 2010), and emotional disorders (Shook et al., 2020; Basak et al., 2022). Obesity during pregnancy can induce placental and intrauterine inflammation and oxidative stress, which may be partially mediated by alterations in the gut microbiota (Basak et al., 2022). This state of heightened prenatal inflammation has been associated with adverse offspring neurodevelopmental outcomes in both human and animal studies (Kwon et al., 2022). Obesity can also lead to changes in maternal hormone levels, particularly thyroid hormones, which—as discussed above—are crucial for fetal brain development (Basak et al., 2022).

Insulin resistance and resulting hyperinsulinemia also affect a significant portion of women with PCOS, leading to a higher risk of type 2 diabetes mellitus and gestational diabetes mellitus (Legro et al., 1999; Yildiz et al., 2008; Moran et al., 2010; Choudhury and Rajeswari, 2022). It has been established that intrauterine exposure to high glucose levels can lead to neurocognitive and behavioral problems in offspring, possibly due to increased inflammation in the fetal brain (Rodolaki et al., 2023; Marra et al., 2024). Hyperinsulinemia also increases hyperandrogenism in women with PCOS (Poretsky et al., 1999; Hirschberg, 2009). Insulin stimulates androgen production by interacting with LH and inhibits hepatic synthesis of sex hormone-binding globulin, leading to elevated testosterone levels (Poretsky et al., 1999; Hirschberg, 2009). Taken together, these findings suggest that maternal insulin resistance may indirectly impact offspring neurodevelopment by contributing to increased intrauterine androgen levels.

Whether obesity and insulin resistance are causes of PCOS or mediate the association between PCOS and offspring psychopathology and neurodevelopment is unclear; likely both pathways are active to differing degrees depending on the individual. Epidemiologic studies that have controlled for BMI or obesity have still found significant associations between PCOS and neurodevelopmental problems, however (Kosidou et al., 2017; Bell et al., 2018; Hisle-Gorman et al., 2018), suggesting that additional factors are at play.

Subfecundity, infertility treatment, and related birth outcomes

While not all women with PCOS have trouble conceiving, many do, especially those with PCOS phenotypes characterized by chronic or oligo-anovulation (Guastella et al., 2010). Indeed, four out of five cases of anovulatory infertility are attributed to PCOS (Melo et al., 2015), and both the treatments these patients receive and birth outcomes related to those treatments may explain some of the associations between PCOS and offspring neurodevelopment. The initial medical approach to address anovulatory infertility is the prescription of a selective estrogen receptor modulator, most commonly clomiphene citrate, followed by ovarian drilling, and lastly IVF (Melo et al., 2015). Clomiphene citrate and IVF are associated with increased rates of both monozygotic and dizygotic twinning (Dawson et al., 2015). While multiple gestation is a known risk factor for preterm birth and low birth weight, the odds of these outcomes are even stronger among twins conceived via IVF compared with those that are spontaneously conceived (McDonald et al., 2010). A robust literature documents the relationship of both preterm birth and low birth weight with risk of adverse child neurodevelopmental outcomes, including ASD and ADHD (Hee Chung et al., 2020; Cortese et al., 2021; Song, 2023).

Regardless of whether or not they use ART, women with anovulatory PCOS are likely to take longer to become pregnant, increasing their own and their partners’ age at conception. Advanced paternal age at conception is an independent risk factor for offspring neurodevelopmental problems and psychopathology, including ASD, ADHD, schizophrenia, bipolar disorder, anxiety, depression, and eating disorders (Zweifel and Woodward, 2022). Proposed mechanisms include de novo genetic mutations, epigenetic alterations, and hereditary psychopathology that might select into late fatherhood (Frans et al., 2015), although the latter would not necessarily be relevant in the case of a couple that has been having difficulty conceiving because of PCOS.

Multi-faceted risks associated with maternal psychopathology/stress

Maternal psychiatric and personality disorders are well-established risk factors for a wide range of psychopathology types and neurodevelopmental problems in offspring (Breaux et al., 2014; Middeldorp et al., 2016; Joelsson et al., 2017). For example, evidence has consistently linked maternal depression to emotional and behavioral problems in children, although effect sizes are often small, and maternal eating disorders are associated with higher levels of multiple types of child psychopathology (e.g. hyperactivity, anxiety, depression, and personality problems) (Martini et al., 2020). Studies have also shown greater risks among offspring of parents experiencing more than one type of psychopathology. Hypothesized intergenerational transmission of psychopathology includes a broad array of mechanisms such as shared genetics; disruptions to caregiving behaviors; exposure to parents’ symptoms such as maladaptive cognitions, depressed mood, or dysregulated behavior; and exposure to stressful environments such as negative life events and low levels of support and resources that often co-occur with psychiatric morbidity (Goodman and Gotlib, 1999; McCarty and McMahon, 2003).

During pregnancy, maternal exposure to stress impacts offspring neurodevelopment in part via physiological systems such as the hypothalamic–pituitary–adrenal (HPA) axis. Maternal activation of the HPA axis releases glucocorticoids (i.e. cortisol) that impact the fetus directly via transportation across the placenta, altering fetal brain development (Barrett and Swan, 2015). While experiences with PCOS are varied, prior literature has shown that individuals with PCOS have significantly higher depression and anxiety scores compared to individuals without PCOS (Barnard et al., 2007; Podfigurna-Stopa et al., 2015; Li et al., 2022). Stress and depression related to a PCOS diagnosis may be particularly high in the preconception period, as many individuals with PCOS report anxieties about infertility and may have used infertility treatment to conceive, as described above (Barnard et al., 2007). Chronic prenatal stress exposure is associated with changes in the fetal hippocampus, amygdala and prefrontal cortex, areas involved with emotional regulation, memory, and executive function (Welberg and Seckl, 2001; Buss et al., 2012). A body of literature supports that chronic prenatal stress exposure is associated with increased risk of ADHD, ASD, anxiety, depression, and aggressive behavior in offspring (Van den Bergh et al., 2005; Manzari et al., 2019; Lautarescu et al., 2020). Previous studies have reported sex differences, although results have been inconsistent (Buss et al., 2012; Lombardo et al., 2020).

Although neurodevelopmental disorders in offspring may result from genetic and/or epigenetic susceptibility, maternal physiology, and biological events during gestation, subsequent exposure to maternal PCOS-related psychiatric disorders and associated risk factors (e.g. disruptions to family life, stress) during childhood likely contribute, as well. Additionally, there may be reciprocal influences between child neuropsychiatric problems and maternal mental health, maternal stress, and family functioning that are likely to compound and escalate problems over time (Mackler et al., 2015; Baker et al., 2020). While associations between maternal PCOS and offspring psychopathology and neurodevelopment have been observed even when controlling for maternal psychopathology, these analyses have adjusted for a history of psychiatric diagnoses (i.e. genetics), but not for symptomatology or for behavior during prenatal and postnatal periods. Longitudinal analyses examining maternal mental health and the home environment during pregnancy and throughout childhood are needed to better interrogate the contribution of maternal psychopathology to the association between PCOS and child outcomes, to identify potential critical periods of exposure, and to examine potential gene–environment interactions.

The potential role of endocrine-disrupting chemicals

Mounting evidence suggests that PCOS may be linked to exposure to EDCs, which are ubiquitous in consumer products and the built environment, and that these associations may be bidirectional. The strongest evidence is for bisphenol A (BPA), an estrogenic compound used in the manufacture of polycarbonate plastics and epoxy resins. In rodent studies in which pregnant dams are exposed to BPA, pups develop a PCOS-like phenotype, including advanced onset of first estrous and a shortened estrous period, decreased number of primordial follicles, increased incidence of meiotic aberrations in oocytes, and impaired ovulation. The biological mechanisms underlying these associations have yet to be elucidated, but may include potentially heritable epigenetic effects on fetal granulosa cells (Khaghani et al., 2021; Sabry et al., 2024).

Epidemiological studies have consistently shown associations of BPA with PCOS and with individual Rotterdam criteria (Kahn et al., 2020; Srnovršnik et al., 2023). All of these studies have been cross-sectional, however, and because BPA has a short half-life, exposure measurements likely do not reflect concentrations during critical periods for PCOS development, including gestation, early childhood, and puberty. Rather, these associations between current BPA exposure and PCOS may be due to reverse causation: the majority of cases of PCOS are the ‘classic’ phenotype characterized by obesity, and BPA is commonly found in food packaging materials. It is therefore plausible that women with PCOS are exposed to higher concentrations of BPA through increased food consumption, resulting in higher in utero exposure of their own children. Epidemiological evidence supports associations of other EDCs present in food packaging, such as phthalates and per- and polyfluoroalkyl substances (PFAS), with PCOS as well (Kahn et al., 2020; Neuvonen et al., 2023).

In the case of BPA and child neurodevelopment, exposure data during the critical period of fetal gestation is commonly available from pregnancy cohorts that collect maternal prenatal urine samples. These studies have linked prenatal exposure to BPA with behavioral problems in children (Braun et al., 2009, 2011; Perera et al., 2012; Evans et al., 2014; Roen et al., 2015), and some have observed differing results according to sex (Ejaredar et al., 2017; Kahn et al., 2020). Prenatal exposure to other EDCs, such as PFAS, phthalates, polybrominated flame retardants, organophosphate pesticides, organophosphate esters, and pyrethroid insecticides has also been associated with behavioral problems in children (Gore et al., 2015; Kahn et al., 2020; Ghassabian et al., 2022). A number of these flame retardants, plasticizers, and pesticides have been associated with ASD, as well (Braun et al., 2014; Marí-Bauset et al., 2022).

Likely mechanisms by which EDCs may affect both PCOS and fetal neurodevelopment include direct interaction with androgen receptors as well as dysregulation of steroidogenesis. In vivo and in vitro studies have shown that organochlorine pesticides, polychlorinated biphenyls, brominated flame retardants, organotin compounds, and a variety of fungicides and herbicides interfere with production of sex hormones by the gonads and adrenal glands as well as function of CYP450 enzymes necessary for the synthesis and conversion of steroid hormones in the liver, kidneys, and other tissues (Sanderson, 2006).

Conclusion

While our review confirmed the consistency of the epidemiological evidence linking maternal PCOS to offspring psychopathology and neurodevelopmental outcomes, the underlying mechanisms, especially those that could be targeted for intervention, remain unclear. Because mechanisms may vary according to PCOS phenotype, wherever possible, analyses should be stratified by phenotype and consideration should be given to the potential impact of PCOS diagnosis bias on results. Future studies that include information on both genetics and environmental exposures, as well as potential mediators such as maternal sex hormone profile during the prenatal period and maternal mental health both prenatally and postnatally could further unravel the mechanisms by which PCOS leads to neurodevelopmental disorders such as ASD or ADHD in offspring. Examples might include studies focused on identification of genetic polymorphisms that are associated with both PCOS and ASD/ADHD or fetal and child neuroimaging studies that highlight brain circuits affected by androgen exposure and their connection with neurodevelopmental outcomes. The USA Environmental influences on Child Health Outcomes (ECHO) consortium, a nationwide longitudinal study with an intensive data and biospecimen collection protocol that begins in pregnancy and follows offspring from infancy through adolescence (Knapp et al., 2023), provides a unique opportunity to advance knowledge on the complex interplay between genetics and various environmental factors across generations in shaping child health and development.

Data availability

No new data were generated or analyzed in support of this research.

Authors’ roles

L.G.K., A.E.H., and A.G. conceptualized and outlined the manuscript. All authors contributed to the research and drafting of the manuscript, and provided critical feedback. L.G.K. was responsible for editing and revising.

Funding

During the creation of this manuscript, LGK received support from NIH grant R00ES030403; AEH received support from NIH grants R01HL157787, R01MH129308, and R01MH134010; AG received support from NIH grant R01ES032826; MC received support from National Science Foundation grant 20-A0-00-1005789; and LGK, AG, and KNC-T received support form NIH grant UG3OD023305.

Conflict of interest

The authors have no conflicts to declare.