The influence of sex on cognition in psychosis Authors: Bodyl A. Brand, Robert A. McCutcheon Published: 2026 Introduction Cognitive impairment is a core feature of psychosis, with cognitive performance in people with psychosis being on average two standard deviations below that in healthy controls (Keefe et al., 2011). Deficits in memory, processing speed, attention and executive function predate illness onset, persist across all phases of the disorder, and strongly contribute to poor clinical outcomes such as unemployment and inability to live independently (Green, 1996). Sex differences in cognitive functioning in psychosis largely mirror those seen in the general population, with women outperforming men on verbal memory and processing speed, and men outperforming women on visuospatial tasks (Mendrek and Mancini-Marïe, 2016; Verdaguer-Rodríguez et al., 2025; Buck et al., 2020; Pu et al., 2019; Smelror et al., 2021). This apparent similarity has led to the assumption that sex may not be a particularly informative variable when it comes to understanding cognitive impairment in psychosis. Yet cognitive function is not static across the lifespan. While age-related decline is its most familiar driver, hormonal transitions also exert an important influence on cognition in the general population. The menopausal transition is one of the clearest examples (Weber et al., 2014; Wood Alexander et al., 2024; Kuh et al., 2018), but there is also evidence that hormonal changes during the menstrual cycle and perinatal period impact cognition (Pletzer et al., 2019; Scheuringer and Pletzer, 2017; Pletzer et al., 2018; Younis et al., 2025; Qiu et al., 2021). Women with psychosis navigate these same transitions while already carrying a substantial cognitive burden, raising the question of whether hormonal state is an underappreciated determinant of cognitive outcomes in this population. In the current article, we examine how hormonal transitions across the female lifespan shape cognitive function relevant to psychosis, and propose that hormonally driven changes in cognition may represent a sex-specific pathway to clinical vulnerability that is worth exploring. Oestrogen as a modulator of cognition Among the ovarian hormones, oestrogen in particular emerges as a candidate modulator of cognition, consistent with its well-characterised neurobiological effects on the brain systems underlying cognitive functioning. Oestrogen interacts with several neurotransmitter systems, impacting dopaminergic, glutamatergic, and cholinergic signalling (Nathan et al., 2026; Woolley et al., 1990; Brinton, 2008; Gibbs, 2010). Recent preclinical work has shown that oestrogen plays a key role in regulating hippocampal synaptic spine density across the oestrous cycle (the rodent equivalent for the menstrual cycle), with direct consequences for learning and memory (Wolcott et al., 2025). Oestrogen fluctuations across the oestrous cycle have further been shown to modulate reward prediction errors and reinforcement learning, supporting a role for oestrogen in dopamine-dependent cognitive processes (Golden et al., 2025). Oestrogen also enhances cholinergic function by increasing choline acetyltransferase activity and basal forebrain cholinergic signalling, mechanisms that are closely linked to attention and memory (Gibbs, 2010). Together, these findings point to oestrogen as a dynamic modulator of the neural circuits underlying learning, memory, and reward. In line with this, hormonal fluctuations across the menstrual cycle appear to be associated with measurable changes in brain structure and function in healthy populations. For example, EEG studies show increased alpha and decreased theta activity during the high-oestrogen (late-follicular) phase of the menstrual cycle, suggesting greater cortical activation and attentional readiness (García-Granados et al., 2026). Additionally, fMRI studies demonstrate cycle-related variation in hippocampal and fronto-striatal activation during cognitive tasks (Pletzer et al., 2019). Structural changes have also been reported, with ovarian hormone levels linked to variation the parahippocampal cortex and hippocampus (Zsido et al., 2023). Conversely, a recent meta-analysis showed no effect of menstrual cycle phase on objective cognitive performance in healthy individuals (Jang et al., 2025), suggesting that neurobiological changes do not necessarily translate into detectable differences on standard cognitive tasks. This may reflect limited sensitivity of commonly used cognitive paradigms to capture subtle, state-dependent effects. A similar dissociation between measurable brain changes and relatively modest cognitive effects is observed in the peripartum period, which is characterised by some of the most pronounced endocrine fluctuations across the female lifespan. Neuroimaging studies indicate substantial structural brain changes during pregnancy and early postpartum, including widespread reductions in cortical grey matter volume that follow dynamic, hormone-linked trajectories (Hoekzema et al., 2016; Luders et al., 2020; Pritschet et al., 2024; Servin-Barthet et al., 2025). Despite these marked neurobiological changes, cognitive effects in healthy populations are typically subtle and domain-specific, most consistently affecting memory and executive functioning (Davies et al., 2018; Anderson and Rutherford, 2012). Although less well studied, the menopause transition has also been linked to measurable changes in brain structure and function (Mosconi et al., 2021). Subjective cognitive functioning appears to be consistently affected (Weber et al., 2014; Greendale et al., 2020), with approximately two-thirds of women going through menopause reporting cognitive complaints, whereas evidence for objective cognitive decline during the menopause transition is again less consistent, though subtle changes have been reported (Greendale et al., 2020). While oestrogen-based menopause hormone therapy does not appear to have a consistent effect on cognitive function, small effects may be present depending on timing of initiation, formulation, and menopause type (i.e., surgical or natural) (Andy et al., 2024). Yet, long-term follow-up data indicate that menopausal hormone therapy may not have lasting effects on cognitive function (Gleason et al., 2024). Taken together, these findings indicate that hormonal transitions across the lifespan are associated with measurable changes in brain structure and function, but only subtle changes in cognitive performance in the general population. This raises the possibility that in psychosis, where cognitive systems are already compromised, these same hormonal shifts may translate into periods of heightened vulnerability, both cognitively and clinically. Hormonal transitions and psychotic and cognitive vulnerability Women show elevated risk of psychosis onset and relapse at times of oestrogen decline (Sommer et al., 2023; Reilly et al., 2020; Bergink et al., 2016). The general period of reproductive life, when oestrogen levels are relatively high, is associated with relative protection against psychosis in women compared to men (Sommer et al., 2023; Brand et al., 2022a; González-Rodríguez et al., 2021). Nevertheless, a systematic review and meta-analysis showed that the period of the menstrual cycle when oestrogen levels decline (i.e., the perimenstrual phase) is associated with higher rates of hospitalisations and more severe symptom severity (Reilly et al., 2020). Whether a similar pattern for cognitive symptoms exists is less clear. Menstrual cycle studies in women with psychosis provide some evidence to support an impact of oestrogen fluctuations on cognitive functioning. One study showed that while both women with and without psychosis performed better on several cognitive tasks during the first half of the menstrual cycle, when oestrogen increases, performances on a psychomotor task specifically worsened in women with psychosis in the second half of the cycle, which coincides with the period during which they are more vulnerable to relapse (Thompson et al., 2000). Further, a small study in 22 women with schizophrenia, including users of oestrogen-based hormone therapy or oral contraceptives and nonovulatory participants, reported positive associations between oestrogen levels and cognitive performance (Hoff et al., 2001). Similar associations between oestrogen and cognition were reported in a separate study of 35 premenopausal patients, all assessed during the first half of the menstrual cycle, when oestrogen levels are gradually increasing (Ko et al., 2006a). Another study found no association, but did not assess the phase when oestrogen levels are highest and thought to most strongly benefit cognition, sampling instead the mid-luteal phase, when both oestrogen and progesterone are high, and the early follicular phase, when both hormones are low (Rubin et al., 2015). Menstrual irregularity is common in women with psychosis and is often induced by prolactin-raising antipsychotics (Riecher-Rössler, 2017). This irregularity reflects suppression of oestrogen, which may have particular consequences for cognitive function. A cross-sectional study found that associations between oestrogen levels and cognition were present in women treated with prolactin-raising antipsychotics, but not in those treated with prolactin-sparing antipsychotics, suggesting that cognitive function may be especially affected in those whose oestrogen levels are pharmacologically suppressed (Brand et al., 2024a). In line with this, another study showed that menstrual cycle irregularity predicted significantly poorer cognitive performance in the areas of psychomotor speed, verbal fluency, and verbal memory, as compared to women with a regular cycle (Gurvich et al., 2018). However, another study did not report any effects of menstrual cycle irregularity on global cognition, highlighting the need for further research in this area (Gleeson et al., 2016). The postpartum period represents another period of oestrogen decline and is one of the most acute examples of psychosis vulnerability. Postpartum psychosis affects approximately 1 to 2 per 1000 women after childbirth, coinciding with the most rapid decline in oestrogen levels that occurs across the female lifespan (Friedman et al., 2023). Research on cognitive functioning during this period is limited, but available studies suggest that cognitive differences may already be detectable during late pregnancy. Women at risk for postpartum psychosis show poorer executive functioning and processing speed in the third trimester compared to controls, a difference not seen in women at risk for postpartum depression, with impairments being most pronounced in those who developed postpartum psychosis within four weeks postpartum (Cattarinussi et al., 2025). Similarly, women at risk for postpartum psychosis showed poorer verbal learning and memory than healthy controls when assessed in the third trimester, alongside increased prefrontal connectivity within the executive network postpartum. This increased connectivity correlated with better cognitive performance and was most pronounced in those who remained well, suggesting a compensatory mechanism (Sambataro et al., 2021). After reaching menopausal age, women with schizophrenia-spectrum disorders are consistently hospitalised for psychosis more often than their male peers (Sommer et al., 2023; Monteiro Fernandes et al., 2025). The effectiveness of antipsychotic drugs in preventing rehospitalisation is substantially lower in women over 45 compared to younger women and to age-matched men, a divergence that coincides with the mean age of onset of perimenopause (Sommer et al., 2023). Yet, research into the effects of menopause on cognitive functioning in psychosis remains limited. One study reported poorer visuospatial performance in the post-menopausal period, as compared to the peri- and pre-menopausal period (Gurvich et al., 2018), yet it is unclear whether these findings reflect menopausal effects or age-related changes, as no male control group was included. Beyond naturally occurring hormonal transitions, a number of interventional studies have also examined whether oestrogen-based augmentation can improve outcomes in women with psychosis. These studies have shown consistent positive effects of oestrogen-based treatments on psychosis symptoms (Peng et al., 2025; Brand et al., 2024b; Li et al., 2023). Several randomized controlled trials have examined whether augmentation with oestrogen or the oestrogen-like compound raloxifene can improve cognitive outcomes in women with psychosis irrespective of menopausal status, with evidence of pro-cognitive effects particularly in the domains of memory and attention (Brand et al., 2023; Gurvich et al., 2019; Huerta-Ramos et al., 2014; Ko et al., 2006b; Weickert et al., 2015), even though not all studies have found benefits (Huerta-Ramos et al., 2020; Weiser et al., 2017; Weiser et al., 2019). An important caveat is that the majority of these trials were not designed with cognition as a primary outcome and may therefore have been underpowered to detect cognitive effects, limiting the conclusions that can be drawn from null findings. Neuroimaging studies further suggest that raloxifene, a selective oestrogen receptor modulator (SERM) that has oestrogenic effects in the brain, enhances prefrontal and hippocampal activation during cognitive tasks (Ji et al., 2016; Kindler et al., 2016), pointing to oestrogen receptor-mediated effects on the neural circuitry underlying cognition. Antipsychotics as a complicating factor The picture is further complicated by antipsychotic treatment, which may itself disrupt the hormonal environment on which cognitive function partly depends. Antipsychotics may disproportionately affect cognitive function in women via prolactin elevation and its downstream effects on oestrogen (Brand et al., 2024a). Potent D2 blockers such as risperidone, haloperidol, and paliperidone are particularly prone to causing hyperprolactinaemia, and women are more susceptible to this effect than men, both because of a higher baseline sensitivity to prolactin elevation and an elevated risk associated with higher plasma concentrations (Kaar et al., 2020; Grigg et al., 2017). Elevated prolactin inhibits the hypothalamic-pituitary-gonadal axis, thereby suppressing endogenous oestrogen production and effectively inducing a state of iatrogenic hypo-oestrogenism (Riecher-Rössler, 2017; Smith et al., 2002). This pathway may have direct cognitive consequences. In premenopausal women with psychosis taking prolactin-raising antipsychotics, higher prolactin levels have been associated with lower oestrogen and poorer cognition: prolactin correlated negatively with processing speed, and the accompanying oestrogen reduction with poorer verbal fluency (Brand et al., 2024a). This may create a compounding pharmacological mechanism: the same drug that elevates prolactin simultaneously suppresses the very hormone that would otherwise buffer cognitive vulnerability. Another critical issue is that current dosing guidelines for antipsychotics are sex-neutral, despite clear pharmacokinetic differences between men and women (Brand et al., 2026). Differences in CYP enzyme activity, body composition, and gastric motility result in higher antipsychotic plasma concentrations in women at equivalent doses, going hand in hand with higher rates of most side effects (Brand et al., 2022b; Jönsson et al., 2019; Castberg et al., 2017; Hoekstra et al., 2021; de Beer et al., 2026; Soldin and Mattison, 2009). This effective overdosing increases women's exposure to dose-dependent cognitive side effects (McCutcheon et al., 2025), including the anticholinergic burden associated with several commonly prescribed antipsychotics. Furthermore, given oestrogen's pro-cholinergic properties (Gibbs, 2010), periods of low oestrogen may render the cholinergic system particularly vulnerable to anticholinergic challenge, creating a biological context in which the cognitive impact of these drugs is amplified. This is particularly relevant for clozapine, olanzapine, and quetiapine, which carry the highest anticholinergic load and have been most consistently linked to cognitive impairment (Mancini et al., 2025; McCutcheon et al., 2023). For clozapine and olanzapine, this concern is compounded by the fact that their metabolism is slowed by oestrogens via CYP enzyme inhibition, placing premenopausal women at particular risk of elevated plasma concentrations and associate side effects (Brand et al., 2022b; de Beer et al., 2026). Quetiapine presents a somewhat different pattern: plasma concentrations are more comparable between the sexes in younger adults but show a relative increase in women at older ages, potentially reflecting age-related decline in CYP3A4 activity (Castberg et al., 2017), though another study found no sex-specific effect of age (de Beer et al., 2026). This suggests that for quetiapine specifically, the risk of pharmacokinetic overdosing and anticholinergic cognitive burden may potentially increase with age. Towards a lifespan-sensitive and sex-informed understanding of cognitive dysfunction in psychosis Taken together, these observations point to cognitive dysfunction in psychosis as a dynamic rather than static phenomenon which is in part shaped by the cognitive effects of naturally occurring hormonal transitions, and the hormonal and pharmacological side effects of antipsychotic treatment. Yet this complexity is rarely reflected in how cognitive outcomes in psychosis are studied. Most research treats cognitive performance as a trait-like variable, assessed at a single timepoint without reference to hormonal status, reproductive stage, or treatment-related endocrine effects, and thereby potentially precluding the detection of sex related effects. A lifespan-sensitive and sex-informed approach to cognitive research in psychosis would involve routinely recording hormonal status and reproductive stage in studies of cognitive outcomes, rather than restricting this to specialised designs. This can be implemented with minimal effort and burden by incorporating a small number of targeted questions (e.g., menstrual regularity, phase, contraceptive use, menopausal status). In addition, there is a need to design longitudinal studies that specifically track cognitive functioning across hormonal transitions, including the menstrual cycle, pregnancy and the postpartum period, and the menopausal transition. This also means reconsidering how antipsychotic treatment is evaluated in women, accounting not only for sex differences in plasma concentrations but also the downstream endocrine effects of treatment. Hormonal transitions constitute repeated natural experiments in modulated brain function and recent preclinical and neuroimaging evidence suggests that the mechanisms underlying this modulation are within reach. Studying cognition across hormonal transitions in women with psychosis therefore offers not only a window into sex-specific vulnerability but also offers a broader opportunity to illuminate the neural mechanisms of cognitive dysfunction in psychosis itself. CRediT authorship contribution statement Bodyl A. Brand: Writing – review & editing, Writing – original draft, Conceptualization. Robert A. McCutcheon: Writing – review & editing, Conceptualization. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.