Attention to the need to consider biological sex differences across the lifespan has increased in recent years, with the majority of neurobiological research involving models of the brain and aging being conducted using only males (Beery and Zucker, 2011; Klein et al., 2015; Beltz et al., 2020; Rechlin et al., 2022). Differences in affective and cognitive processing between biological sex is well documented, with males generally performing better at spatial-related cognitive processing, and females having a higher prevalence of stress, anxiety and fear-related affective disorders (Bauer, 2023; Yagi and Galea, 2019). At the neural level, molecular mechanisms within the central amygdala and hippocampus have been shown to differentially regulate fear memory consolidation and extinction in young male and female mice (Florido et al., 2024; Matsuda et al., 2015; Velasco et al., 2019). Morphologically, sex differences within the hippocampus have been observed in young rodents and regulated by estrous cycling, including gross hippocampal volume (Qiu et al., 2013), dendritic spine density (Woolley et al., 1990), and neural cell proliferation (Rummel et al., 2010). Confounding factors and variability associated with active estrous cycling in females is often given as the rationale for using exclusively male rodents in studies of cognition and physiology (Becegato and Silva, 2024; Kaluve et al., 2022), limiting and biasing the pool of studies of brain and behavior across the lifespan. A recent meta-analysis of studies assessing preclinical behavioral models of anxiety-like behavior found that just 21% of studies included both sexes and considered biological sex in their analyses, and that 61% of published studies used exclusively male subjects (Kaluve et al., 2022). While a growing number of studies are considering how environmental factors such as stress exposure during early-life and critical developmental periods differentially impact males and females (Bale and Epperson, 2015; Guadagno et al., 2018; Hodes and Epperson, 2019; LoPilato et al., 2020), direct investigations of sex differences in behavior and brain plasticity in response to modifiable environmental and lifestyle risk factors and protective factors active into late-adulthood are scarce (Bangasser and Cuarenta, 2021).

Adequately assessing differences in cognitive and affective behavior along with brain structure and function across the lifespan in humans exposed to different environmental factors remains a considerable challenge. Many studies tend to be correlational, rely on retrospective self-report, are limited to measures of observable forms of brain activity with limited spatial and temporal resolution, or are based on post-mortem examination of brain pathology (Grady, 2012; Palliyaguru et al., 2021). While these approaches provide valuable insight, they are limited in identifying underlying changes at the cellular level (Fischer, 2016). Rodent models of aging can overcome this limitation and provide critical information about differences in neuronal plasticity in the young and aged brain, while maintaining control over exposure to external environmental factors. However, due to cost-prohibitive limitations on aging research using animal models, few studies to date have directly investigated sex differences related to the aging brain and behavior (Palliyaguru et al., 2021; Özalay et al., 2024).

Chronic stress has been associated with detrimental effects to both physical and psychological health (McEwen and Gianaros, 2010; O’Connor et al., 2021; Sapolsky, 1996). These effects include memory impairment (Kim and Diamond, 2002; Luine et al., 2017) and increased levels of anxiety and depression (McEwen, 2004). Further, it has been shown that behavioral stress is associated with increased inflammatory glial responses (Frank et al., 2007; Bollinger et al., 2017), pro-inflammatory cytokine expression (Mograbi et al., 2020), depleted brain neurotrophic growth factors (Duman and Monteggia, 2006; Murakami et al., 2005; Notaras and van den Buuse, 2020), stunted prefrontal cortex dendritic growth (Bloss et al., 2010), and reduced hippocampal volume and neurogenesis (Duman, 2004; Jung et al., 2020; Schoenfeld et al., 2017). Many of the adverse effects of stress exposure compound with age and may contribute to an acceleration of brain aging (Bishop et al., 2010; Wolf et al., 2024; Yegorov et al., 2020).

Certain modifiable lifestyle and environmental factors may be protective against age-related brain and behavioral changes. Brief periods of exposure to voluntary aerobic exercise and environmental enrichment contribute to beneficial effects on neuronal development and cognition, including enhanced hippocampal-dependent memory and hippocampal neurogenesis (Epp et al., 2021; Kempermann et al., 1997, 2002; Scholz et al., 2015; van Praag et al., 1999a, 1999b, 2005), increases in endothelial growth factors (Cao et al., 2004; Fabel et al., 2003; Morland et al., 2017) and brain-derived neurotrophic factor (BDNF), particularly in young adults (Adlard et al., 2005; Cotman et al., 2007; Huang et al., 2014; Wrann et al., 2013; Maass et al., 2016). There is strong evidence that long-term engagement in voluntary aerobic exercise such as running is neuroprotective, and protective against cognitive decline with aging in humans (Chan et al., 2005; Duzel et al., 2016; Tyndall et al., 2018) and rodents (Alvarez-López et al., 2013; Diederich et al., 2017; Robison et al., 2018; Connolly et al., 2022). However, the degree to which running confers protection against cognitive impairment and promotes brain plasticity may differ in males and females (Triviño-Paredes et al., 2016; Barha et al., 2017a, 2017b; Barha and Liu-Ambrose, 2018; Loprinzi and Frith, 2018). One study exposing mice to daily voluntary running throughout young and middle adulthood did include both male and female subjects, but analyses considering biological sex were not performed (Robison et al., 2018). A systematic review of sex differences in aerobic exercise effects on brain and cognition using aged rodent models noted that most studies conducted to date used only a single sex, the majority of which were male, limiting any conclusions that can be directly drawn related to the effectiveness of exercise exposure in aged males and females (Barha et al., 2017b).

In male mice, 16 months of continuous exposure to voluntary wheel running enhanced synaptic plasticity-related gene expression in the hippocampus and improved spatial memory relative to sedentary sex-matched controls, suggesting a cumulative benefit of aerobic exercise throughout the lifespan in aged males (Stranahan et al., 2010). Aged male mice exposed to 2 months of late-life voluntary running exhibited enhanced spatial memory performance relative to socially isolated aged mice, although neither of these late-life environmental factors (running or isolation) significantly altered gene expression of microglial and astrocytic genes (i.e., Iba1, GFAP) or anti-inflammatory cytokines in the hippocampus (Ederer et al., 2022), suggesting that relatively brief environmental changes in late-life may not robustly alter brain physiology in males, despite being sufficient to impact spatial memory and behavior. Findings on the effects of environmental enrichment including voluntary running on brain and cognition in aged females are also mixed. Late-life exposure to 6 months of voluntary running decreased anxiety-like behavior in the open field, enhanced spatial memory, and increased rates of late-adult hippocampal neurogenesis and BDNF levels in aged female mice, suggesting that periods of prolonged exercise in late-adulthood are beneficial to affect, cognition, and brain plasticity in females (Marlatt et al., 2012). However, female mice continuously exposed to enrichment conditions, including access to voluntary wheel running for 11 months, showed enhancements in memory performance, despite no enhancements in neurogenesis (Hüttenrauch et al., 2016). The apparent inconsistency in brain and behavior between sexes in response to brief or prolonged enrichment and exercise into late-adulthood is striking, yet the methodological differences used across studies make it difficult to directly compare their findings.

Exposure to chronic mild stress has differential effects on males and females (Bremner and Narayan, 1998; Lupien et al., 2009; McLaughlin et al., 2009), although the majority of animal model investigations considering the role of sex in reactivity to chronic stress involve juvenile and young adult animals with stressor periods typically lasting only 1–8 weeks (Franceschelli et al., 2014). In response to 21 days of chronic restraint stress, young adult female rats showed decreased dendritic branching complexity and length within hippocampal CA3 subfield, and increased corticosterone relative to males experiencing the same restraint stress (Galea et al., 1997), suggesting that the female brain is particularly sensitive to the effects of chronic stress exposure in young adulthood. At the behavioral level, young adult female rats have shown greater resilience to chronic restraint stressors than young adult male rats, which show impairments in spatial and non-spatial memory (Bowman et al., 2003, 2009). The same sex distinction was not observed in aged rats exposed to 3 weeks of restraint stress in late-adulthood, with comparable memory performance in both males and females on the spatial recognition test that was unimpaired by stress (Bowman et al., 2006). Potential sex differences following chronic stress exposure throughout the entire adult lifespan are still unknown. These findings highlight the importance of identifying how environmental factors may differentially impact males and female behavior and neural physiology across the lifespan and between biological sexes.

As there are no studies contrasting behavior and brain plasticity in male and female mice with identical environmental rearing conditions into late-adulthood, it is not yet known the degree to which life-long chronic stress and exercise differentially impact cognition, affect, and brain plasticity between biological sexes. A central goal of this study is identifying how modifiable environmental factors (running, stress) may differentially influence affective and cognitive behavior and brain plasticity in males and females across the lifespan. Given the co-morbidity seen with late-life inactivity, depression, and dementia (Alexopoulos, 2005; Cunningham et al., 2020; Strawbridge et al., 2002), understanding the impact of environmental risk factors is critical to developing interventions aimed at protecting against age-related cognitive decline, anxiety and depression, neurodegeneration and other pathologies that may differentially affect males and females. Here, we investigated sex differences on affective behavior, memory, and hippocampal cell division related to the effects of long-term exposure to chronic mild stress or voluntary exercise in young and aged male and female mice.

In line with previous findings, we found that voluntary running increased hippocampal cell division in both young and aged mice (Epp et al., 2021; Pereira et al., 2007; van Praag et al., 1999a, 1999b, 2005). While previous analyses of sex differences in rates of cell division and survival in young rats identified rates of attrition of newborn hippocampal neurons are higher in males than in females (Yagi et al., 2020), here, no differences in rates of Brdu⁺ cell survival 1-month post-labelling were observed in young mice, although BrdU labelling alone cannot identify cell type (neuron or glial). However, sex differences were evident in aged mice, with females retaining slightly higher survival of late-adult newborn cells, regardless of environmental stressors or exercise. Sex-specific effects of stress were not evident in the BrdU⁺ cell survival of young or aged mice. Previous reports in young adult rats exposed to 2 weeks of restraint stress did show evidence of a sex-specific reduction of survival of BrdU⁺ hippocampal neurons in females (Hillerer et al., 2013). It may be that the high levels of restraint stress-induced corticosterone in females relative to males contributed to the sex-differences in cell survival (Hillerer et al., 2013). While corticosterone measures were not obtained in the present study, it is possible that the varied mild stressors employed for 1 month in young mice were not sufficient to induce corticosterone mediated suppression of hippocampal proliferation and survival. To thoroughly characterize the cumulative and progressive impact of chronic stress on the brain and behaviors into aging, it will be informative for future investigations to repeatedly assesses additional biochemical and physiological variables including hormone and corticosterone levels, as well as a longitudinal behavioral assessment rather than using only end-point behavior and BrdU measures. Given normal age-related declines in neural stem cells and hippocampal neurogenesis neurotrophic factors, and brain volume which may be moderated by running behavior into late-adulthood (Connolly et al., 2022; Gao et al., 2023; Gil-Mohapel et al., 2013; Zocher and Toda, 2023), repeated in-vivo assessments of brain volume and structural changes over time, as well as cell specific measures of neural plasticity (DCX, Ki67, NeuN, BDNF, GFAP, IBA1) will be essential to identifying the neural mechanisms mediating sex-specific differences in response to environmental factors across the lifespan. Previous investigation of granule cell and molecular layer volumes in the dentate gyrus also identified age-related reductions associated with a smaller pool of precursor cells (Olariu et al., 2007), and fewer synaptic inputs into the granule cells (Geinisman and Bondareff, 1976; Geinisman et al., 1992), rather than being attributable to a gross reduction in the number of granule cells or the dendritic complexity of granule cells (Chawla and Barnes, 2007; Morgenstern et al., 2008). The factors contributing to potential age-related differences in hippocampal cell layer composition and morphology have not yet been systematically investigated between sexes, and highlight an avenue for further investigation in order to identify the factors that are most susceptible to the cumulative effects of stress or exercise into late-adulthood.

Studies involving prolonged periods of voluntary wheel running using single housing introduce a confounding stressor associated with extended periods of social isolation (Barha et al., 2017b). The environmental enrichment associated with group exercise is an important mediator of running-related enhancements of hippocampal neurogenesis (Stranahan et al., 2006). Given the manipulation of periodic social isolation as one of the stressors in the stress rearing condition, we deliberately chose to group-house running-reared mice to avoid introducing confounding stress induced by single housing. Consistent with a landmark study using group-running into late-adulthood that identified increased hippocampal BrdU⁺ cells (Kronenberg et al., 2006), the use of group housing during the 15-month running period limited the ability to track individual running activity of each mouse. As a result, directly correlating running behavior with each animal’s results are not possible, limiting the identification of a dose-dependent relationship between rates of running behavior and brain and behavioral measures between sexes. Previous investigation in young male rats exposed to voluntary running for 30 days did find a relationship between rates and intensity of running behavior and BrdU⁺ cell survival, with moderate running activity (4–8 h/day) producing the greatest benefit to cell survival. Interestingly, they found a negative correlation between distance run and BrdU⁺ cell survival, suggesting that prolonged, excessive running behavior may be detrimental to the survival of newly generated hippocampal cells (Nguemeni et al., 2018). In the aged brain, there is evidence that longer durations of running are required to induce a neurogenic boost, with higher rates of hippocampal neurogenesis observed after 7, but not 2, weeks of running in aged male mice (Connolly et al., 2022). Given the prolonged testing period following rearing, it is possible that the rated of BrdU⁺ cell survival observed in the present study reflect a more cumulative response to the interventions, rather than the acute effects of stress or exercise which might have been captured by testing immediately following the rearing interventions.

We observed that prolonged exercise (15 months) was associated with quicker acquisition of a spatial learning task compared to stressed and sedentary control aged mice, while acute (1 month) mild stress or exercise was not sufficient to significantly influence memory performance in any task in young mice. Despite an enhancement of spatial memory performance in the water maze, no other enhancements of memory were observed in aged mice, suggesting that running throughout the lifespan induced only a modest cumulative protective effect on memory. Though no sex-specific differences in memory performance were observed in response to running or stress exposure in the present study, increasing stress during a spatial water maze task by using cold water identified sex-specific search strategies, in which females used more variable search strategies relative to males (O’Leary et al., 2022), suggesting that the timing of exposure and type of stress are important factors to consider when assessing stress-induced sex differences in brain and behavioral activity. In the present study, stress did not differentially affect the acquisition of spatial memory across training days in either young or aged mice, but a more nuanced evaluation of search strategies may have identified potential sex differences associated with acute or prolonged stress exposure and performance in the water maze.

Methodological variability in common chronic stressor models depends on whether the intent is to model depression, post-traumatic stress, or other affective and mood disorders (Duman and Monteggia, 2006; Schöner et al., 2017; Antoniuk et al., 2019). The majority of chronic stress models have been optimized using exclusively male mice, and resultantly may not adequately model stress-related behaviors experienced by females (Lopez and Bagot, 2021). The decision to use multiple mild stressors in the present study was motivated by the goal of mimicking relevant everyday stressors in mouse’s natural environment (occasional food or water deprivation, proximity to predator, periodic isolation, circadian disruption, soiled nesting area) to enhance the ecological validity of unpredictable daily stressors relevant to a rodent (Antoniuk et al., 2019).

Life-long running was associated with low body weight in aged males, but not females, while long-term stress exposure resulted in low body weights in both aged males and females. Low weight gain in aged male and female rodents has been previously observed following several weeks of restraint stress exposure (Bowman et al., 2006; Bloss et al., 2010). The behavioral manifestations of stress appear to vary across sex, with both young and aged females displaying a more anxious-like phenotype in the open field, relative to males. Previous findings following late-life restraint stress found that stress differentially increased anxiety-like behavior in the open field in males but not in females, while having no effect on overall activity levels (Bowman et al., 2006). Aged females also show an increased depressive-like phenotype compared to males, while no differences in sex were evident in younger mice. Hyperactivity and enhancements in anxiety-like behavior in aged mice were the most notable observations following prolonged stress exposure. In the open field, elevated plus maze, and Y-maze exploration tasks, aged stressed mice generally exhibited a hyperactive phenotype, which was not observed in young mice. This effect was especially pronounced in aged female stressed mice, consistent with prior observations of hyperactivity in aged females exposed to social isolation stress in late-adulthood (Sullens et al., 2021). Hyperactivity and impulsivity following prolonged exposure to stress has been commonly observed in young and aged rodents (Ashby et al., 2010; An et al., 2017; Strekalova et al., 2005; Schweizer et al., 2009), with some studies interpreting hyperlocomotion or hyperactivity as a symptom of chronic stress, potentially confounding typical behavioral measures such as the open field and the forced swim task in which low activity is thought to be indicative of anxiety-like and depressive-like behaviors (Strekalova et al., 2005, 2022). As noted by Strekalova et al. (2005), stress-induced hyperactivity may be exacerbated by strong illumination, as evident in the present study’s observed enhanced distance travelled in the open field by stressed mice, irrespective of age or sex, and high speed of exploration for aged males and females. In support of this interpretation, during the beam walk test where bright illumination is used at the starting point, the rapid escape latency observed in aged males and females may be reflective of a hyperactive phenotype following prolonged stress into late-adulthood. Interpreting the meaning behind the observed hyperactive behavior remains a significant challenge to rodent models of stress, anxiety-like and depression-like behaviors, with few studies directly considering the differential effects between biological sex (Lima et al., 2022).

In sum, we provide evidence supporting differential effects of exercise and chronic mild stress across the lifespan in male and female mice, and highlight the sensitivity of females to stress-induced increases in hyperactivity, anxiety-like and depressive-like behavior with age. A growing body of literature highlights the importance of considering sex as a biological variable when determining how healthy aging is differently impacted by modifiable lifestyle factors across the lifespan, and points to the need to develop models that explicitly include biological sex as a variable of interest using standardized models of chronic stress or enrichment and a standardized behavioral battery in order to improve the methodological rigor required to systematically investigate sex differences in aging models.