Adolescent alcohol and the spectrum of cognitive dysfunction in aging

Abstract

Among the many changes associated with aging, inflammation in the Central Nervous System (CNS) and throughout the body likely contributes to the constellation of health-related maladies associated with aging. Genetics, lifestyle factors and environmental experiences shape the trajectory of aging-associated inflammation, including the developmental timing, frequency, and intensity of alcohol consumption. This chapter posits that neuroinflammatory processes form a critical link between alcohol exposure and the trajectory of healthy aging, at least in part through direct or indirect interactions with cholinergic circuits that are crucial to cognitive integrity. In this review, we begin with a discussion of how inflammation changes from early development through late aging; discuss the role of inflammation and alcohol in the emergence of Mild Cognitive Impairment (MCI); elaborate critical findings on the contribution of alcohol-related thiamine deficiency to the loss of cholinergic function and subsequent development of Wernicke’s-Korsakoff Syndrome (WKS); and present emerging findings at the intersection of alcohol and Alzheimer’s Disease and Related Dementias (ADRD). In doing so, our analysis points toward inflammation-mediated compromise of basal forebrain cholinergic function as a key culprit in cognitive dysfunction associated with chronic alcohol exposure, effects that may be rescuable through either pharmacological or behavioral approaches. Furthermore, our review reveals an interesting dichotomy in the effects of alcohol on neuropathological markers of ADRD that depend upon both biological sex and genetic vulnerability.

1. Introduction

Excessive alcohol use is a major risk factor for early cognitive decline and dementia (Rehm et al. 2019), the emergence of neurodegenerative states (Crews and Nixon 2009), and all-cause mortality (Organization 2019). In particular, recent studies have shown that early onset dementia was especially prevalent among individuals diagnosed with Alcohol Use Disorder (AUD) (Schwarzinger et al. 2018). In contrast, systematic review of studies on light to moderate drinking reported a slightly reduced risk of dementia relative to individuals who largely abstained from alcohol consumption (Ilomaki et al. 2015). Despite these long-standing associations, alcohol misuse was historically an exclusion criterion for evaluating dementia prevalence world-wide but is now listed as one of the top 3 modifiable risk factors for dementia (Livingston et al. 2020). These findings have motivated extensive research into the mechanisms contributing to alcohol-mediated disruption in cognitive dysfunction across the lifespan, which can manifest in a wide spectrum of neurocompromised states.

The immune system transforms across early development and later aging, showing progressive deterioration throughout the lifespan (Lynch et al. 2010). Even in healthy individuals, aging produces sustained inflammation due to exacerbated inflammatory responses of immune cells and their eventual senescence (Rozovsky et al. 1998), which often takes the form of delayed or impaired “shutoff” by late-acting, anti-inflammatory mechanisms (Norden and Godbout 2013; Norden et al. 2015). Though this transformation occurs throughout the body, it is especially harmful in the central nervous system (CNS) as neuroinflammation is a key driver of age-related cognitive decline. While natural aging is inevitable, certain lifestyle choices, such as alcohol use, can exacerbate these sensitized immune responses (Carlson et al. 2023). Indeed, a substantial challenge for the field is to identify how the developmental timing, frequency, and intensity of alcohol consumption might influence the trajectory of healthy brain aging and cognitive integrity (Deak et al. 2022; Deak and Savage 2019; Nunes et al. 2019). The over-arching goal of this chapter is to integrate what is known about the relationship between binge-like alcohol consumption (or exposure models) and the spectrum of cognitive dysfunction that is commonly associated with prolonged alcohol misuse (see Figure 1). To do this, we will first discuss consequences of natural aging, specifically age-related impairments in neuroimmune function, and then highlight how alcohol can aggravate and advance such dysfunction. Subsequent sections will describe how cholinergic mechanisms by which alcohol can accelerate normal, age-related cognitive decline to progress into Mild Cognitive Impairment (MCI) and more severe pathological states such as Wernicke-Korsakoff Syndrome (WKS) and Alzheimer’s Disease and Related Dementias (ADRD).

Figure 1:

Cascading impact of chronic alcohol use across the lifespan.

2. Alcohol interactions with natural aging.

2.1. The neuroimmunology of aging

Natural aging is accompanied by a decline in the efficiency and accuracy of the neuroimmune system that first emerges in middle age and amplifies in later ages (Moca et al. 2022). Such impairments have been observed in a variety of immune-derived cells within the CNS, and extensive work has described age-related changes in microglia and astrocytes (Lynch et al. 2010; Fonken et al. 2016; Frank et al. 2016; Clarke et al. 2018). For example, multiple studies have suggested that aged microglia adopt a sensitized or “primed” state that is characterized by excessive and prolonged inflammatory responses (Fonken et al. 2016; Norden and Godbout 2013). Under normal conditions, microglia are extremely dynamic and rapidly transition from their homeostatic resting state to a reactive, pro-inflammatory state when stimulated by an injury or pathogen (Yirmiya et al. 2015). Reactive microglia secrete a host of inflammatory cytokines, chemokines, and reactive oxygen species (ROS) in order to remove the insult and limit damage to surrounding tissue (Safaiyan et al. 2016; Eggen et al. 2013). This process differs based on the type of insult, and activated microglia respond differently to an antigen compared to cerebral infarct or injury. When a pathogen-associated antigen is detected, activated microglia initiate a signaling cascade of inflammatory mediators that may recruit peripheral leukocytes to the CNS, which then remove the pathogen through phagocytosis (Rock et al. 2004). Microglia themselves are phagocytic, meaning they are also able to clear cellular waste and debris, including dead and dying cells (Sierra et al. 2010; Neumann et al. 2009). This is especially important after CNS injuries and infarcts as reactive microglia can remove damaged cells, including neurons that are secreting toxic amounts of excitatory signals, and limit damage to surrounding cells (Loane and Byrnes 2010). Importantly, healthy microglia can quickly transition back to their resting state when the threat is resolved and the anti-inflammatory response is initiated (Eggen et al. 2013).

Unlike activated microglia found in healthy adults, microglia in aged brains are much less efficient at transitioning back to a resting state, resulting in their sustained activation and release of pro-inflammatory mediators (Norden and Godbout 2013). Preclinical studies reported significant elevations in the pro-inflammatory cytokines interleukin (IL)-1β and IL-6 in the brains of aged rodents relative to adults after they were challenged with the endotoxin lipopolysaccharide (LPS) (Godbout et al. 2005; Henry et al. 2009). Furthermore, this heightened inflammation persisted in the aged brains while adult brains quickly resolved the immune challenge (O’Neil et al. 2022). Additional ex vivo studies suggested that aged microglia are primed for these heightened responses because even in the absence of a stimulus, they exhibit increased major histocompatibility complex II (MHCII) expression, which is indicative of a reactive, pro-inflammatory phenotype (Henry et al. 2009; Frank et al. 2010). As immune threats are a common occurrence throughout the lifespan, sensitized microglia likely contribute to age-related increases in basal inflammation. Indeed, basal expression of IL-1β, IL-6, and MHCII is significantly higher in the brains of aged rodents and humans (Gano et al. 2017; Frank et al. 2006; Streit et al. 2004). Similar age-related increases have been observed in the pro-inflammatory cytokine, high mobility group box 1 (HMGB1), and toll-like receptors (TLRs), which are also considered a result of primed microglia (Fonken et al. 2016). Along with aberrant inflammatory responses, aging also influences microglia’s phagocytic activity as aged microglia express less phagocytic receptors and exhibit reduced debris clearance (Thomas et al. 2022). It has been proposed that such dysfunction is due to increased lipid burden within aged microglia that inhibits phagocytic activity (Marschallinger et al. 2020). While the mechanisms driving this impairment have not been fully elucidated, the accumulation of waste within the CNS is known to produce cytotoxic amounts of inflammatory mediators and cause widespread cell death (Neumann et al. 2009). Thus, age-related increases in neuroinflammation are not solely due to microglial priming but are a product of multi-faceted alterations in the functional state of the neuroimmune defense network.

While dysfunctional microglia greatly contribute to the sustained neuroinflammation seen in aging, astrocytes undergo a similar deterioration that further drives this inflammatory state (Jyothi et al. 2015). Astrocytes regulate a number of processes within the CNS including neurotransmitter release and blood flow, as well as synapse formation and elimination (Sofroniew and Vinters 2010). Additionally, astrocytes play an important role in the glymphatic system, which clears cellular waste through fluid exchange between the brain’s interstitial fluid and cerebral spinal fluid (Iliff et al. 2012). Such actions are impaired in aging brains as there is an age-associated reduction in the expression of aquaporin 4 (AQP4), the astrocytic water channel responsible for this exchange (Kress et al. 2014). As a result, the accumulation of cellular debris and neuroinflammation produced by aged microglia and their impaired phagocytic activity was potentiated by aging astrocytes (Verkhratsky and Semyanov 2023). Lastly, much like microglia, astrocytes can adopt a pro-inflammatory phenotype to respond to an immune threat, and such responses can become dysregulated in aging (Clarke et al. 2018). Reactive astrocytes are often characterized by an upregulation in glial fibrillary acidic protein (GFAP) and numerous preclinical and clinical studies report elevated GFAP expression in aging brains (Verkhratsky and Semyanov 2023; Nichols et al. 1993; Robillard et al. 2016). Other studies suggest that like microglia, aging astrocytes adopt a primed state that sensitizes their inflammatory responses to immune threats and increases basal inflammation (Clarke et al. 2018).

Another shared consequence of aging that affects both microglia and astrocytes is that these cells become resistant to anti-inflammatory mechanisms (O’Neil et al. 2022). Normally, once an immune threat is contained, anti-inflammatory mediators regulate the transition of reactive microglia to their homeostatic resting states or an anti-inflammatory, reparative state (Eggen et al. 2013). However, activated microglia from aged brains show reduced sensitivity to these markers, especially the prominent anti-inflammatory mediators IL-4 and transforming growth factor beta (TGFβ) (Fenn et al. 2012; Fenn et al. 2014; Tichauer et al. 2014; Rozovsky et al. 1998). Reactive microglia are also regulated by astrocytes through their release of the anti-inflammatory cytokine IL-10 which stimulates the production of TGFβ and inhibits microglial activation (Norden et al. 2014). Unfortunately, aged astrocytes exhibit a similar decline in anti-inflammatory actions as reduced IL-10 expression was observed in astrocytes from aged rats both 4 and 24 hours after LPS challenge and they were unable to prevent microglial activation (O’Neil et al. 2022; Norden et al. 2016). These age-related resistant and erroneous immune responses are referred to as immunosenescence, which is thought to propagate neuroinflammation as it is no longer contained by critical regulatory feedback systems (O’Neil et al. 2022).

Interestingly, the priming and eventual immunosenescence of aging astrocytes and microglia appear to be region specific, with some areas showing extreme sensitivity to these age-related effects. One such region is the hippocampus, which plays an integral role in cognitive function, particularly learning and memory (Golomb et al. 1993). Aging studies consistently report exaggerated levels of pro-inflammatory mediators, including IL-1β and IL-6, in the hippocampus of aged rodents relative to adults (Gano et al. 2017; Sierra et al. 2007). There is extensive evidence that these elevations are due to increased microglial priming in this brain region as an LPS challenge elicited much larger inflammatory responses in the hippocampi of aged rats (Barrientos et al. 2006; Frank et al. 2010). Similarly, there is evidence from both rodent and human studies that hippocampal astrocytes are also more susceptible to age-related dysfunction as they show the largest upregulation of GFAP and undergo the most transcriptomic changes that promote a primed, pro-inflammatory state (Nichols et al. 1993; Soreq et al. 2017; Clarke et al. 2018; David et al. 1997). As the hippocampus is a critical integration center for learning and memory, it is unsurprising that sustained neuroinflammation in this region has been identified as a key contributor to age-related cognitive decline. Beyond the hippocampus, neuroinflammation in other regions such as the cortex, thalamus and cerebellum are thought to contribute to age-associated impairments in executive function, sensory processing, speech, and coordination (Carlson et al. 2023; Ownby 2010). Given the growing number of aged individuals in the global population, it is important to identify which lifestyle choices may accelerate or prevent such detrimental changes in neuroimmune function.

2.2. Interactive effects of alcohol and aging

One lifestyle/experiential variable that can greatly accelerate age-related inflammation, especially within the CNS, is alcohol exposure across the lifespan (Carlson et al. 2023; Deak et al. 2022). A major challenge facing the alcohol field is to better understand the relationship between alcohol consumption patterns and its influence on overall health outcomes, including and especially for CNS function. Not only is it difficult for individuals to accurately reconstruct their history of alcohol consumption across prolonged periods of time, but self-report of alcohol intake is also influenced by a variety of factors such as individual expectancies, positive self-presentation, and memory impairments associated with drinking (Maisto et al. 1995; Tevik et al. 2021). Although objective biomarkers of alcohol exposure such as phosphatidylethanol (PETH) concentrations offer some promise (Harris et al. 2021), PETH levels persist in blood and organs on the timescale of days to weeks, which is probably not adequate for realistic assessments of lifetime alcohol exposure. PETH is also not an effective measure in preclinical (rodent) models, further limiting its use in research and development (Aradóttir et al. 2004). For these reasons, many investigators have now changed tactics to incorporate objective measures of senescence markers as a means to determine the influence of alcohol exposure and/or consumption on the trajectory of aging, further cementing a transition in focus from chronological aging to biological aging (eg., (Zillich et al. 2024)).

A second significant challenge in evaluating the influence of alcohol consumption patterns across the lifespan on aging-related CNS function is the developmental period in which the alcohol exposure was incurred. As discussed in the chapter by Matthews and colleagues, developmental stage is a critical issue because alcohol exposure is not uniform across the lifespan, nor are the biological substrates on which alcohol acts. For instance, it is estimated that between 1–9% of individuals worldwide have been exposed to alcohol in utero, making Fetal Alcohol Spectrum Disorders (FASD) the leading cause of preventable intellectual disabilities (May et al. 2018). Prenatal development is undoubtedly a period during which the organism may be especially vulnerable to accelerated aging outcomes (eg., (Smith et al. 2022; Church et al. 1996)) as well as neurodegenerative diseases (Araujo et al. 2021), with many of these effects involving long-lasting changes in microglial activity and other features of neuroimmune signaling later in life (Walter et al. 2023; Bake et al. 2023; Reid et al. 2015; Boots et al. 2023; Pinson et al. 2021; Gano et al. 2020). These conclusions are built upon a long history and strong foundation of preclinical and clinical studies elaborating hippocampal and cortical deficits associated with Prenatal Alcohol Exposure (PAE) and form the basis for broader concerns about early alcohol effects on later cognitive function as a result.

Similarly, binge drinking peaks among adolescents and emerging adults, a developmental period marked by dynamic re-architecture of the CNS that involves substantial changes in glial cell activity and function (Doremus-Fitzwater and Deak 2022). For instance, synaptic pruning and paring mechanisms involve the active clearance of complement-tagged synapses through microglial phagocytic activity (Kopec et al. 2019; Germann et al. 2021), which thereby increases the strength of remaining synaptic contacts in an experience-dependent manner. Ongoing angiogenesis formalizes its delivery of blood perfusion to gross anatomical regions of the CNS (Rowan and Maxwell 1981; Ogunshola et al. 2000), while astrocytes envelop microcapillaries in a final sealing of the Blood-Brain Barrier (BBB) (Molofsky and Deneen 2015). Similarly, progressive myelination of axons by oligodendrocytes during adolescence is critical to improvements in processing speed and connectivity across distal regions of the neuroaxis (Jessen et al. 2015; Rice and Barone 2000). The extent to which these processes are immediately responsive to acute alcohol consumption/exposure during adolescence is a somewhat open question for the field because most studies fail to include critical tests of alcohol action at multiple ages. However, the few studies comparing adolescent to adult neuroimmune sensitivity suggest that induction of neuroimmune genes is severely muted in adolescents (relative to adults), regardless of whether they have been challenged with alcohol, LPS, or an acute stress challenge (Doremus-Fitzwater and Deak 2022; Doremus-Fitzwater et al. 2015; Marsland et al. 2022). Thus, it will be critical for the alcohol field to determine how and why the adolescent brain seems to show such categorically distinct sensitivity to alcohol relative to the mature adult brain.

Nevertheless, because binge drinking peaks during late adolescence and early adulthood, a considerable number of studies have examined long-lasting effects of adolescent alcohol exposure in preclinical models. By and large, studies suggest Adolescent Intermittent Ethanol (AIE) substantially influences neuroimmune reactivity even after a protracted period of alcohol abstinence. For instance, a multi-day binge alcohol exposure in rodents sent microglia into a dystrophic state (Marshall et al. 2020; McClain et al. 2011; Marshall et al. 2016). Consistent with this, several studies have shown prolonged microglial activation after AIE, with at least some of these studies showing that the adverse effects of AIE were prevented by CSF1 inhibitors (Coleman et al. 2020), minocycline (Khan et al. 2023; Hu et al. 2020; Barnett et al. 2022), or other drugs with anti-inflammatory properties such as indomethacin (Pascual et al. 2007; Vetreno and Crews 2018; Vetreno et al. 2018; Macht et al. 2023; Monleón et al. 2022; Monleón et al. 2020). Additionally, recent studies have shown sensitized neuroimmune gene induction in adult rats with a history of AIE (Vore et al. 2021) as well as a sensitized fever response to Poly I:C; a synthetic form of double-stranded RNA that mimics a viral infection (Gano et al. 2024). Additional findings on the neuroimmune consequences of adolescent alcohol are discussed in the chapter by Macht and coworkers. These findings need to be kept in perspective, however, as other studies have shown reduced extracellular cytokine concentrations using large molecule microdialysis (Gano et al. 2019) and impaired fever responses provoked by LPS in AIE-exposed males (Telles et al. 2017; Cruz et al. 2020). Together, these studies paint a more complex portrait of neuroimmune reorganization after adolescent alcohol, with many of these effects persisting for 30 days or more after the final ethanol exposure.

Historically, studies examining patterns of alcohol intake across the lifespan have shown that for most individuals, alcohol intake (and binge drinking in particular) peaks in adolescents and emerging adults, then decreases gradually as individuals advance in age from adulthood into later life (Deak et al. 2022). However, some studies have found that rates of alcohol consumption among elderly individuals are on the rise and at a historic high (Grant et al. 2017; Blazer and Wu 2009; Breslow et al. 2017). Although the reasons for this shift toward greater alcohol consumption later in life remain unclear, several factors may contribute. For instance, social isolation tends to increase for many aging individuals as they lose friends/companions and become less mobile (Perkins et al. 2019; Perkins et al. 2016; Ravenel et al. 2024). Social isolation is especially challenging, and even rodents will increase alcohol intake during prolonged periods of social isolation (Karkhanis et al. 2016; McCool and Chappell 2009). As social occasions become less frequent, especially through the recent pandemic, older adults tend to consume more alcohol during such functions and adopt binge-like drinking behaviors (Kuerbis et al. 2014). Related to this, many elderly individuals experience painful bereavement and consume larger amounts of alcohol for its euphoric, relaxing, and anxiolytic effects (Kuerbis et al. 2014). Alternatively, it is possible that the rise in alcohol consumption among aged individuals reflects overall better health and longevity compared to previous decades. Indeed, both the overall lifespan and quality of life in later aging (termed the “healthspan”) have increased substantially over the past 5 decades, perhaps setting the stage for alcohol consumption patterns to mirror what would previously have been observed at younger ages. Regardless, this troubling pattern of heightened alcohol intake among aging individuals requires additional studies to clarify the psychosocial factors that contribute.

As mentioned above, alcohol is a powerful toxicant, making increased alcohol consumption among older adults especially dangerous as alcohol produces many of the same neuroimmune deficits as aging (Carlson et al. 2023). In adult non-aged rodents, chronic ethanol has been shown to stimulate a pro-inflammatory state within the CNS by increasing levels of IL-1β, IL-6 and tumor necrosis factor alpha (TNF-α) (Qin et al. 2008; Lippai et al. 2013; Alfonso-Loeches et al. 2010). Similarly, elevated HMGB1 and TLR3 expression was observed in non-aged rodent brains following chronic ethanol exposure and in post-mortem brain tissue of individuals with alcohol use disorder (AUD) (Crews et al. 2013). Since these elevations in pro-inflammatory mediators in non-aged adults mirror the neuroimmune alterations seen in aging, it has been suggested that ethanol may produce a similar primed phenotype within microglia. For example, microglia isolated from non-aged rodents after a 4-day binge-like ethanol exposure displayed increased MHCII expression, a marker of activation, and similar findings have been reported in microglia isolated from aged rodents (Peng et al. 2017; Frank et al. 2006; Henry et al. 2009). Furthermore, mice exposed to chronic ethanol and then challenged with LPS showed exaggerated levels of brain IL-1β and TNF-α, which is reminiscent of the sensitized responses to LPS observed in the brains of aged rodents (Qin et al. 2008; Godbout et al. 2005; Henry et al. 2009). Beyond microglial priming, ethanol also mirrors age-related changes in the phagocytic activity of these cells as studies report that microglia isolated from mice exposed to chronic ethanol exhibit reduced expression of the phagocytic marker, CD68 (Lowe et al. 2020). Additionally, in vitro studies reported a decline in cultured microglia’s ability to phagocytose amyloid beta after being exposed to ethanol for 24 hours (Kalinin et al. 2018).

Though there is substantial evidence supporting the hypothesis that alcohol primes microglia and exaggerates immune responses, other studies have reported contrasting findings suggesting that alcohol may be better characterized as immunomodulatory, not purely neuroinflammatory. Specifically, studies have shown that relative to adult rats, adolescent rats display a blunted pro-inflammatory response within the CNS (IL-1β, TNF-α) after an acute ethanol challenge (Doremus-Fitzwater et al. 2015). The same group exposed adult rats to multiple ethanol challenges and again observed a suppression of IL-1β and TNF-α in the CNS (Gano et al. 2016). Proteomic studies support this immunomodulatory role of alcohol as cultured microglia exposed to ethanol express more anti-inflammatory mediators compared to microglia exposed to LPS which primarily express pro-inflammatory signals (Guergues et al. 2020). These findings suggest that instead of priming microglia, alcohol may be causing these cells to become senescent as they are not producing the expected immune responses. For example, young rats given an acute ethanol challenge show increases in hippocampal IL-6 expression, whereas aged rats exhibited exaggerated basal levels of hippocampal IL-6 but no change after ethanol administration (Gano et al. 2017). A similar debate has occurred over the effects of alcohol on astrocytes as some studies report increased GFAP expression and astrocyte reactivity in the hippocampus and prefrontal cortex after chronic ethanol exposure (Evrard et al. 2006; Vongvatcharanon et al. 2010; Dalçik et al. 2009), while others report reduced GFAP expression in these regions which is more indicative of senescent cells (Franke 1995; Rintala et al. 2001). It is important to note that these contrasting studies used various alcohol exposure procedures, highlighting that consequences of alcohol on neuroimmune function differ greatly based on the age of consumption, route of administration, and the length of exposure (Deak and Savage 2019). Thus, more consistent studies are needed before the consequences of alcohol within the CNS and how they mirror or potentially accelerate age-related impairments can be fully understood.

Regardless of the mechanisms driving alcohol’s deleterious actions in the CNS, aging can greatly increase an individual’s alcohol sensitivity, making older adults more susceptible to its neurotoxic effects. This heightened sensitivity and lower tolerance to alcohol is likely due, at least in part, to age-related reductions in alcohol pharmacokinetics, including changes in body composition and ethanol metabolism. Aged individuals show reduced activity of acetaldehyde dehydrogenase and cytochrome P-4502E1, two major enzymes responsible for metabolizing alcohol (Meier and Seitz 2008). Additionally, aging is often associated with changes in body composition and lower water content, creating less volume for alcohol distribution (Vestal et al. 1977). As a result, the blood ethanol content (BEC) of an elderly individual will likely be larger than the BEC of a younger adult who consumed an equal amount of alcohol (Cederbaum 2012). This enhanced sensitivity increases the likelihood that elderly individuals will experience impaired motor coordination and greater confusion while intoxicated which can be extremely dangerous (Kalant 1998). Such impairments have been recapitulated in rodent studies that report greater ethanol-induced impairments in motor function and Morris water maze performance in aged rodents relative to adults (Ornelas et al. 2015; Perkins et al. 2018; Novier et al. 2013).

Interestingly, the cognitive impairments observed after ethanol exposure in rodents and humans closely resemble those seen in aging, which suggests that certain brain regions may be more sensitive to both. The hippocampus is one region that is vulnerable to age-related neurodegeneration as well as alcohol toxicity and both are associated with cognitive decline (Beresford et al. 2006; Golomb et al. 1993; Wilson et al. 2017; Jack et al. 2000). Though the hippocampus is the most widely studied, similar effects of aging and alcohol have been implicated in the frontal cortex, parietal cortex, temporal cortex, and the hypothalamus, which are also important mediators of cognitive function (Toledo Nunes et al. 2019; Harper and Kril 1989). These studies highlight an intersection of aging and alcohol, as any alcohol-induced impairments in these regions, whether a result of primed or immunosenescent cells, are likely amplified in the aged brain and vice versa. In sum, alcohol and aging have synergistic effects on brain health, especially neuroimmune function, which likely accelerates age-related cognitive decline. The following sections will discuss how this acceleration creates a sensitized environment that can rapidly progress into more severe neurodegenerative pathologies.

3. Developmental alcohol exposure and the progression of pathological age-related cognitive impairment

3.1. Alcohol use disorder (AUD) is a risk factor for pathological aging.

Excessive alcohol use over the lifespan can increase the risk for alcohol-related brain damage (ARBD) and cognitive decline. Over 70% of individuals with chronic AUD display some degree of brain pathology (Goldstein and Shelly 1980; Harper 1998), and many of the brain regions implicated in AUDs mirror those vulnerable to degeneration during advanced aging and Alzheimer’s Disease and Related Dementias (ADRD). This includes reductions in brain volume in several regions critical for cognition, including the frontal, temporal, parietal, cingulate, and insular cortices, cerebellum, thalamus and hippocampus, and this loss can be more pronounced in adults with AUDs that are 65-years and older (Sullivan et al. 2018; Zahr et al. 2019).

Several cross-sectional human studies support the idea that chronic alcohol misuse alters normal aging as patients with AUD exhibit accelerated brain shrinkage by middle-age, and other longitudinal studies revealed age-AUD neuropathological interactions in the frontal cortex and hippocampus (Zahr et al. 2019; Sullivan et al. 2018). Such studies have developed the hypothesis that heavy alcohol use (greater than 5 drinks per day) is associated with accelerated cognitive aging (Woods et al. 2016) and can also increase an individual’s risk of developing Alzheimer’s Disease (AD) if they carry the ApoE e4 allele (Anttila et al. 2004; Kivipelto et al. 2008). Thus, AUD is an enduring, complex disease that continues to evolve during the aging process and likely exacerbates cognitive decline.

3.2. Mild Cognitive Impairment (MCI) as pathological aging

A diagnosis of Mild Cognitive Impairment (MCI) is often made when cognitive difficulties are significantly greater than typical age-related cognitive decline, often over the age of 65. Both cross-sectional and quasi-longitudinal studies indicate modest declines in memory and reasoning abilities until about age 65, after which the decline accelerates (Salthouse 2019). However, there is individual variability in cognitive aging in humans and other species, and different domains are affected by aging. Specifically, aging has been associated with declines in processing speed, short-term memory, language, visuospatial and executive functions (Harada et al. 2013).

The Mayo Clinic’s Petersen/Winblad criteria for MCI include four aspects (Petersen et al. 2014; Winblad et al. 2004): (1) Reported change in cognition; (2) impairment in one or more cognitive domains relative to a person’s age and education; (3) spared independence in functional abilities; (4) lack of dementia. In addition, MCI is classified into four subtypes: single-domain amnestic MCI, multi-domain amnestic MCI, single-domain non-amnestic MCI, and multi-domain non-amnestic MCI (Winblad et al. 2004). The amnestic subtypes are associated with higher risk of progression to AD, compared with non-amnestic subtypes

Hippocampal volume has been identified as a significant predictor of cognitive decline in many cases of MCI (Mieling et al. 2023; Morrison et al. 2023; Richter et al. 2022). Recent research has extended this predictive capacity to basal forebrain volume, particularly in patients undergoing pharmacotherapy with cholinergic drugs to treat cognitive and memory dysfunction. Consistent with this, larger basal forebrain volume was associated with slower global cognitive decline, while greater left hippocampus volume was linked to slower memory decline.

In the transition from prodromal MCI to AD, the volume of the basal forebrain region has been found to predict cognitive decline (Richter et al. 2022; Tiernan et al. 2018). Basal forebrain structural changes seem to precede cortical atrophy in the progression of AD (Schmitz and Nathan Spreng 2016). Thus, alterations in basal forebrain structure serve as pre-symptomatic biomarkers for MCI and progression to AD (Richter et al. 2022; Tiernan et al. 2018; Nicolas et al. 2020). Dysregulation of basal forebrain circuitry in MCI and AD is not limited to cognitive decline; it is also associated with dysregulation of the default mode network, which is critical for executive function and episodic memory (Nair et al. 2018). In addition, basal forebrain degeneration predicts AD-specific pathology. In vivo MRI studies have demonstrated a correlation between the degree of basal forebrain atrophy and amyloid-beta burden (Grothe et al. 2013; Kerbler et al. 2015). Notably, baseline volumes of the nucleus basalis of Meynert (NbM) predict progressive cortical degeneration and a trans-synaptic spread of amyloid-β that starts in the NbM (Kerbler et al. 2015; Schmitz and Nathan Spreng 2016). Furthermore, changes in cognition were associated with the development of pre-tangle markers in basal forebrain cholinergic neurons before frank tau deposition occurred throughout the brain (Vana et al. 2011). Such results suggest that cholinergic basal forebrain neurons are an early vulnerable target to AD-related pathology.

Lower cortical cholinergic innervation is also associated with cognitive decline and MCI (Xia et al. 2022). However, a higher educational level in MCI patients is linked to increased cholinergic activity, suggesting an initial compensatory effect, but it is unknown whether this prevails in later AD stages (Xia et al. 2022). Recent resting-state functional MRI studies have found reduced basal forebrain functional connectivity in several cholinergic projection sites (cortex, hippocampus, amygdala) during the early phases of AD (Li et al. 2017; Qi et al. 2021). NbM dysconnectivity in early AD specifically targets cortical regions enriched with astrocytes/microglia cells and/or immune process-related genes (Ren et al. 2023).

In addition, there is a significant reduction in the number of adult-born neurons within the hippocampus in both MCI and AD patients, compared to normal age-matched controls (Salta et al. 2023). There is also a significant reduction in the number of neural progenitor cells in MCI and AD (Moreno-Jiménez et al. 2019; Tobin et al. 2019). Importantly, the number of new astrocytes increased in MCI and AD, compared to NCI, suggesting changes in the cell fate for newly born cells within the hippocampus during pathological aging (Ginsberg et al. 2019).

Thus, degeneration within the septohippocampal pathway appears critical to the transition from healthy aging toward MCI and ultimately in the development of AD. There are several factors such as genetics, sex, and the environment, including drug and alcohol exposure, that can delay or accelerate the progression of pathological aging (McQuail et al. 2020). Next, we will review the role of early developmental exposure to alcohol in the progression of pathological aging and the use of animal models to reveal critical behavioral changes and neurobiological mechanisms, with a focus on the septohippocampal circuit.

3.3. Aging and the septohippocampal circuit

Rodent models have shown that the septohippocampal circuit is vulnerable to both aging and chronic ethanol exposure, in particular AIE. It is important to note that age-related cognitive decline is not absolute; only a subset of aged rats display cognitive impairment as a consequence of natural aging (Fischer et al. 1989; Gage et al. 1989). In the rodent, cognitive deficits emerge in middle-age (12–18 months), and impaired spatial memory begins to appear around 12 months of age, but cognitive performance can be highly variable (Guidi et al. 2014; Bizon et al. 2009). A longitudinal study also found impairments in episodic spatial memory using the water maze in middle-aged male rats (Febo et al. 2020).

Several studies have shown that hippocampal ACh efflux was reduced as a result of aging (Shao et al. 2019; Chang and Gold 2008; Stanley and Fadel 2012). Age-related declines in ACh efflux have been found as early as 10 months old (Chang and Gold 2008). Aged rats that are cognitively-impaired also show a significant loss of cholinergic neurons in the MS/DB compared to young rats, but show no change in the pontine reticular cholinergic neurons, implicating the basal forebrain as a key structure in age-associated memory impairment (Baskerville et al. 2006). There is also evidence that the age-related loss of integrity of basal forebrain cholinergic neurons occurs to a greater extent in aged rats with impaired spatial learning (Sugaya et al. 1998). Studies have also reported age-related reductions in the density of cholinergic fibers within the dentate gyrus, which was more profound in aged male rats relative to aged female rats (Lukoyanov et al. 1999). In addition, cortical cholinergic innervation became less dense as rats aged, and this was further augmented when the expression of tropomyosin-related kinase A receptors were suppressed by viral vector-based RNA interference (Parikh et al. 2013). This suppression also decreased the ability of cholinergic neurons to release ACh, more so in aged-suppressed rats compared to aged or young controls (Parikh et al. 2013), suggesting that cholinergic transmission declines slowly over the lifespan.

Drugs that enhance ACh levels increase hippocampal adult neurogenesis (Kotani et al. 2006) and the survival of newly born neurons is regulated in part by nicotinic α7 nicotinic ACh receptors (AChR α7) (Kita et al. 2014). However, the neurogenic effect of AChRα7 may be limited to males (Otto and Yakel 2019). In reciprocal fashion, neurogenesis appears to be critical for maintaining and stabilizing the cholinergic septohippocampal circuit, which is critical to successful spatial memory during normal aging (Kirshenbaum et al. 2023). Although hippocampal neurogenesis was decreased as rodents aged (as much as 80% by middle age) (Kuhn et al. 1996; Wu et al. 2023), there does not appear to be an association between decreased neurogenesis and degree of spatial memory impairment (Stepanichev et al. 2023). Although, some studies reported that increased neurogenesis led to improvements in age-related cognitive decline (Marlatt et al. 2012), other studies reported that the decreased hippocampal neurogenesis that occurs with age did not predict the presence or severity of cognitive impairment (Bizon et al. 2004). Neurogenesis is not just involved in initial learning and pattern separation but is also a mechanism for ameliorating proactive memory interference (Akers et al. 2014; Scott et al. 2021), so this issue needs to be considered carefully in aging studies.

3.4. Adolescent ethanol exposure and the septohippocampal circuit

Heavy intermittent alcohol exposure during early adolescence is associated with persistent changes in brain structure and connectivity (Spear 2018). Two inter-related pathologies that have been consistently observed in rodent models of adolescent binge ethanol exposure (AIE) are a suppression of hippocampal neurogenesis (30–60%) (Macht et al. 2021; Vetreno and Crews 2015) and a decrease in the number of cholinergic basal forebrain neurons (25–30%) (Fernandez and Savage 2017; Vetreno et al. 2020). The loss of hippocampal neurogenesis and the reduction in cholinergic basal forebrain neurons seen following AIE were persistent into advanced adulthood (Reitz et al. 2021), but were not evident if the alcohol exposure occurred during adulthood (Broadwater et al. 2014; Vetreno et al. 2014). These findings suggest that the adolescent brain may be especially sensitive and/or vulnerable to long-lasting effects of AIE on cholinergic function. Despite the evidence surrounding the loss of forebrain cholinergic phenotype and the decrease in hippocampal neurogenesis following AIE, hippocampal-dependent spatial behavioral deficits are not consistently observed (Vetreno and Crews 2012; Swartzwelder et al. 2015; Macht et al. 2021). Indeed, there are certain cognitive tasks that are more sensitive to AIE than others (Crews et al. 2019).

3.5. Interactions between developmental ethanol exposure and pathological aging

Some studies have revealed that heavy, but not light-to moderate alcohol consumption, increased the risk for dementia in humans after a diagnosis of MCI (Lao et al. 2021; Rehm et al. 2019; Xu et al. 2009). Thus, heavy alcohol use has emerged as a risk factor for progression from pathological aging to dementia. Recently, interest in the interactions between chronic alcohol exposure and pathological aging has intensified (White et al. 2023). This includes an interest in whether early developmental alcohol exposure alters the trajectory of healthy aging and animal models can be used to directly review the role of developmental alcohol exposure in the progression to pathological aging. An emerging hypothesis is that the septohippocampal pathway is a critical neural circuit that is especially sensitive to early developmental ethanol exposure and aging (Reitz et al. 2024).

An early study found that learning of place location was not impaired in young mice, but a deficit emerged when mice that underwent Prenatal Alcohol Exposure (PAE) reached middle age (White et al. 2023). Another PAE study in male rats found spatial navigation deficits in the Morris water maze in both young and middle-aged animals (Gabriel et al. 2002). However, a progressive change in age-related cognitive decline (spontaneous alternation, novel object recognition and fear conditioning) were not found in C57BL/6J mice exposed to PAE (Smith et al. 2022), and no aging effects were observed on these behaviors at 17 months of age. These data suggest that PAE may accelerate pathological aging, dependent on the PAE model and the behavioral paradigms used to assess cognitive function.

A few studies have examined the effects of adolescent ethanol exposure on successful cognitive aging. Intermittent alcohol exposure during adolescence (from PD 30 to PD 48) facilitated spatial memory impairments to acute ethanol challenges at 18 months of age (Matthews et al. 2017). Another study (Matthews et al. 2023) found that when rats exposed to AIE were tested for spatial memory at 19.5 months, male but not female rats exposed to AIE displayed heightened anxiety and significantly longer swim path lengths on several training days. In addition, changing the platform location to test behavioral flexibility following spatial learning revealed that AIE impaired performance in a sex-dependent manner (males not females) over age-matched controls. However, in a longitudinal design assessing cognition multiple times across 20 months following AIE, spatial memory in the water maze was not affected by alcohol, whereas reversal learning was impaired by AIE—but without alcohol-age-dependent interactions (Matthews et al. 2022). One concern with longitudinal studies is that repeated cognitive testing can serve to “boost” later cognitive performance as either a “learning to learn” or “practice effect” or as a form of enrichment (Cnops et al. 2022; Zheng et al. 2022). This needs to be balanced with the benefits of longitudinal aging studies, which can better identify the onset of cognitive impairment and control for cohort effects (McQuail et al. 2020).

In a cross-sectional study, male and female rats underwent AIE and were tested as either young adults (4 months old) or at middle-age (14 months old) on two hippocampal behavioral assays. These assays were concurrent with the assessment of behaviorally evoked Ach efflux and followed by histological assessment of cholinergic neurons in the medial septum-diagonal band and hippocampal neurogenesis (Reitz et al. 2021). We found that AIE-induced impairments on an object location task resembled age-related cognitive decline. Aging, regardless of AIE status, led to suppression of hippocampal cholinergic tone during the object location task. In contrast, hippocampal neurogenesis was sensitive to the synergistic interaction between AIE and aging, as young adult AIE rats had a higher number of doublecortin staining cells than middle-aged AIE rats. On a different hippocampal dependent task, spontaneous alternation, we found a sex-dependent effect: Middle-aged males were impaired relative to young males, but middle-aged females performed similar to young females. As in previous studies (Kipp et al. 2021a; Fernandez and Savage 2017), AIE had no effect on spontaneous alternation. However, regardless of age, male AIE rats had lower behaviorally evoked ACh during spontaneous alternation, compared to male control rats; an effect not seen in female rats and not observed previously when only young males were assessed (Kipp et al. 2021a; Fernandez and Savage 2017). Middle age did not lead to the suppression of the cholinergic phenotype, which other studies have not observed until advanced age (20–24 months; (Martínez-Serrano and Björklund 1998; Pitkin and Savage 2004)). Thus, our findings support the hypothesis that developmental ethanol exposure may lead to accelerated age-related cognitive decline in spatial location memory, and that this may be associated with reduced hippocampal neurogenesis and Ach deficits.

3.6. Conclusions: Developmental alcohol is a factor that drives pathological aging.

Emerging evidence suggests that binge-like alcohol exposure during adolescence contributes to accelerated pathological aging—particularly in male rodents starting at middle age and progressing into advanced age. A dysfunctional septohippocampal system may make an organism more vulnerable to normal and pathological aging processes, and whether adaptive aging responses are able to compensate under such disease states is unknown (Gray and Barnes 2015). In contrast, in rodents with AD transgenes, the progression of cognitive decline and AD-associated pathology occurred more in females than males, raising important questions about mechanisms contributing to sex-specific vulnerabilities to alcohol, aging, and the progression of ADRD (more on this below).

In humans, there is evidence that basal forebrain pathology is critical for the transition from healthy aging to pathological aging, including the development of MCI and ultimately AD. Degeneration of cholinergic neurons in the basal forebrain is a hallmark of pathological aging, contributing to the reduction of cortical innervation and ACh activity—thereby reducing activity of cortical regions involved in memory and cognition. Basal forebrain volume predicts longitudinal limbic cortical degeneration and the spread of amyloid biomarkers in MCI and AD (Schmitz and Nathan Spreng 2016). Cholinergic neurons in the basal forebrain accumulate both intraneuronal tau and Aβ, and this becomes more profound with the transition to MCI (Schmitz et al. 2018).

A key pathology that arises from adolescent alcohol exposure is loss of the basal forebrain cholinergic phenotype, suppression of behaviorally activated ACh activity in the hippocampus and frontal cortex and reduced hippocampal neurogenesis (Kipp et al. 2021b; Reitz et al. 2021). Given the overlap in septohippocampal pathology between pathological aging and AUD, it is critical to understand how these neurobiological changes associated with developmental exposure to alcohol drives the brain into pathological aging.

4. Thiamine deficiency as a driver of adult alcohol-related brain damage and memory impairment

4.1. Thiamine deficiency related and unrelated to alcohol use disorders in humans

Prolonged and excessive alcohol use has been related to different forms of dementia as well as structural and functional changes in the brain. One of the critical drivers of alcohol-related brain damage (ARBD) is thiamine deficiency (TD). Thiamine (vitamin B1) is an important nutrient required by all tissue with an essential role in the development and maintenance of brain function. During TD, metabolism of lipids, glucose, amino acids, and syntheses of neurotransmitters - aminobutyric acid (GABA), ACh, and glutamate - are affected both in neurons and glial cells (Abdou and Hazell 2015; Gibson et al. 2016). These effects in the brain (human and experimental models) have been associated with development of ARBD (Butterworth 2003; Nardone et al. 2013). It is estimated that TD is prevalent in 15–80% of patients with AUD (Li et al. 2008; Morgan 1982). Alcohol-related dementia (ARD), a chronic AUD in the absence of TD, or other complicating factors, has a less distinct pathophysiological profile than TD (Ridley et al. 2013). Furthermore, it has been questioned whether ARD is a distinct neurocognitive disorder, or has a multifactorial etiopathology that includes TD, head injury, or liver disease (Palm et al. 2022). Given this, we will focus on AUD with TD. Among individuals with AUD, TD occurs due to a combination of poor dietary intake and a decrease in gastrointestinal absorption of thiamine. The high calories and low nutrients in alcoholic beverages require more thiamine to maintain metabolic function, affecting glucose metabolism by inhibiting phosphorylation of thiamine and its incorporation into enzymes leading to TD (Butterworth et al. 1993; Chandrakumar et al. 2018; Donnino et al. 2007; Harper 2006). TD is common in developing countries due to the prevalence of acute malnutrition (Attias et al. 2012); however, TD is also reported in developed countries and occurs as a result of many health conditions, including diseases associated with aging (eg: Alzheimer’s Disease, Parkinson’s Disease), bariatric surgery, eating disorders, HIV, diabetes, and hyperemesis gravidarum (for review: (Dhir et al. 2019)).

If left untreated, TD can manifest into Wernicke’s encephalopathy (WE) and if thiamine is not restored the chronic neurological sequela, Wernicke-Korsakoff syndrome (WKS) may develop (Nardone et al. 2013). The classic triad of signs and symptoms associated with WE are abnormal eye movements, gait ataxia, and cognitive impairment (Harper et al. 1986). Studies have shown that if WE patients are treated with thiamine before the development of significant brain damage, the associated cognitive dysfunctions may be reversible. However, irreversible lesions can develop and persist if TD continues (Isenberg-Grzeda et al. 2012; Ogershok et al. 2002), and the disorder can progress to a long-lasting amnestic state, Korsakoff’s amnesic syndrome, or WKS. This progression happens in approximately 56%–84% of individuals with AUD, with less frequency in cases unrelated to alcohol misuse (Arts et al. 2017; Harper 2006; Kopelman et al. 2009; Victor et al. 1971).

The primary cognitive signs of WKS include psychosis, confabulation, memory loss and learning deficits (Marrs and Lonsdale 2021). Individuals with WKS have difficulty in forming new memories, as well as retaining or recalling new information, due to the profound anterograde amnestic state. Although, the anterograde memory is usually more affected than retrograde memory in WKS patients, retrograde amnesia is observed with deficits in temporal gradient, where the memory impairments could extend retrospectively for up to 30 years (Arts et al. 2017; Kopelman et al. 1999; Kopelman et al. 2009; Victor et al. 1971).

The neural underpinnings of the profound amnesia in WKS have been studied in both post-mortem studies and imaging studies with damage to the mammillary bodies and the thalamic nuclei as key regions of interest (Harding et al. 2000; Mayes et al. 1988; Kopelman 1995; Pitel et al. 2012; Sullivan and Pfefferbaum 2009). Although shrinkage of frontal cortical structures (gray and white matter), hippocampus, and cerebellar cell loss are seen in WKS, it occurs to a lesser degree and with less frequency than diencephalic damage (Nunes et al. 2019; Sambon et al. 2021; Sullivan and Pfefferbaum 2009). Importantly, pathology of the thalamus has been considered one of the critical predictors of memory dysfunction in alcohol related WKS (Harding et al. 2000; Pitel et al. 2012; Pitel et al. 2015); (for review: (Savage et al. 2021; Sullivan and Pfefferbaum 2009). The thalamus plays an important role as a node of two networks that are implicated in AUD patients with WKS: The Papez circuit that subserves episodic memory, and the frontocerebellar circuit involved in working memory and executive functions (Fadda and Rossetti 1998; Nunes et al. 2019; Pitel et al. 2011; Pitel et al. 2015). Segobin and colleagues (2019) showed that atrophy of different thalamic nuclei is dissociated in these two brain circuits where atrophy of the anterior thalamus was found uniquely impacted in WKS patients. Decreased connectivity between anterior thalamic nuclei and hippocampus was also observed in WKS (Segobin et al. 2019), likely a consequence of disruption of the fornix. This finding supports the hypothesis that amnesia in WKS is associated with a disrupted neural circuit involving medial temporal lobe and diencephalic regions (Nahum et al. 2015).

Investigating cognitive and brain changes in patients with WKS over months and up to 10 years after the diagnosis, Maillard and colleagues (Maillard et al. 2021) showed that even after years of alcohol abstinence, only a mild recovery of the volume in three brain regions of frontocerebellar circuit (cerebellum, pontine crossing tract, and middle cerebellar peduncle) was observed, and there was no significant improvement in cognitive performance (especially episodic memory). In addition, structural and metabolic alterations of the Papez circuit in WKS persisted: The thalamus, hypothalamus, and fornix were severely atrophied at all times (early, 1 year and 10 years after diagnosis; (Maillard et al. 2021)). In contrast, studies have shown that in individuals with AUD, without TD, prolonged abstinence led to the recovery of memory and restoration of brain volume (Segobin et al. 2014; Pfefferbaum et al. 1995). These results emphasize that irreversible anterograde amnesia is observed in WKS from TD and not the alcohol damage per se (Arts et al. 2017; Maillard et al. 2021).

Considering that WE is underdiagnosed and the existence of high mortality rates of untreated patients (20%; (Harper et al. 1986)), thiamine supplementation has been considered as a treatment for suspected WE. It should also be considered as a treatment for AUD patients with or without diagnosis of WE, to avoid the progression to WE (Pruckner et al. 2019). Praharaj and colleagues (2021) also suggested that other neuropsychiatric syndromes associated with TD-- in the context of AUD-- such as alcohol cerebellar syndrome, Marchiafava–Bignami syndrome, alcoholic polyneuropathy, and delirium tremens should also receive thiamine replacement therapy (Praharaj et al. 2021).

Oral thiamine supplementation has also been suggested to prevent the development of WE in AUD as well as to improve alcohol-induced cognitive impairment (Dhir et al. 2019; Lagercrantz et al. 1986; Smith et al. 2021). Examining alcohol-dependent persons during routine inpatient detoxification with oral thiamine supplementation (oral dose of 200 mg thiamine/day, scheduled for 14 days), a recent study showed that thiamine replacement together with alcohol abstinence improved thiamine blood levels and cognitive function (Bonnet et al. 2023). Listabarth and colleagues (2023) also revealed an association between improvements in memory and thiamine replacement, showing both oral and intravenous thiamine administration had equal efficiency in increasing the blood thiamine pyrophosphate levels (Listabarth et al. 2023). These finding suggest that thiamine supplementation should be implemented as a clinical practice to prevent cognitive impairment in AUD. Aside from the intravenous thiamine replacement therapy and oral supplements, there is no optimal treatment for AUD, WE or WKS since the treatment varies depending on symptoms and severity (The Oxford Handbook of Adult Cognitive Disorders 2019). Furthermore, the thiamine dose, duration of treatment, and type or frequency of administration are still unclear (Isenberg-Grzeda et al. 2012; Latt and Dore 2014; Pruckner et al. 2019).

4.2. Developmental thiamine is rare and understudied

As described earlier, TD is frequently associated with alcohol consumption, however TD can also occur during pregnancy and in the months following parturition (Fattal et al. 2011; Guerrini et al. 2007). Pregnant women are at risk of TD-development when there is severe malnutrition (especially reported in low- and middle-income countries), excessive alcohol use, or hyperemesis gravidarum. In a recent review, Kareem and colleagues (2023) presented a broad range of clinical manifestations of TD during pregnancy and their consequences to the mother and fetus, e.g: subclinical TD, wet beriberi, dry beriberi (WE and WKS) and infantile beriberi (Kareem et al. 2023).

Scattered clinical cases have reported WE’s symptoms and the risk of the WKS development in pregnant and lactating women due to hyperemesis gravidarum, characterized by persistent severe nausea and vomiting during pregnancy that causes a rapid loss of thiamine (Hillbom et al. 1999; Ismail and Kenny 2007; Oudman et al. 2019; Rane et al. 2022). Fetal and maternal mortality was reported in half of the patients diagnosed with WE from hyperemesis gravidarum; however, there was not a relation between fetus survival and thiamine dose used to treat WE symptoms (Oudman et al. 2019). Breastfed infants of TD mothers are at highest risk for developing TD, resulting in infantile beriberi. If left untreated, infantile beriberi can result in death or the emergence of neurodevelopmental problems later in life (Allen 2012; Mimouni-Bloch et al. 2014). For instance, in 2003, several infants were hospitalized in Israel with severe neurological symptoms, prolonged vomiting, nystagmus, seizures and coma. After investigation, they were diagnosed with infantile TD, which was caused by a non-dairy, soy-based infant formula that was not supplemented with thiamine. These TD-infants were responsive to thiamine treatment; however, years later, some children that had infantile TD had persistent cognitive and language impairments (Fattal-Valevski et al. 2005; Fattal-Valevski et al. 2009; Fattal et al. 2011).

In 1990, a clinical study showed that mothers with fetuses with severe intrauterine growth retardation exhibited lower blood cell thiamine content compared to women with normal pregnancies, suggesting that thiamine supplementation should be included during pregnancy (Heinze and Weber 1990). The thiamine-depleting effects of alcohol are expected to result in adverse health outcomes since thiamine levels are crucial for neurodevelopment and function, and alcohol is a potent teratogen in humans (Bâ 2017). Indeed, the high incidence of intrauterine growth retardation and other abnormalities observed in children born to alcohol-consuming mothers in the 1960s was later characterized as “fetal alcohol spectrum - FAS” (Butterworth 1993). Although both TD and early developmental alcohol exposure separately cause FAS characteristics such as intrauterine growth retardation, microcephaly, and language impairments, the lack of developmental data on TD-alcohol synergism during pregnancy and lactation in humans and how it can interfere in fetal brain development in human population is still unresolved ((Bâ 2011, 2017) for review see: (Kloss et al. 2018)).

Experiments with TD in female rats during pregnancy support the idea that adequate intake of thiamine during pregnancy is essential for successful development of offspring. Maternal thiamine deficiency induces delayed development of the fetus, decreased brain weight, myelination, and neurochemical changes in the CNS (Fournier and Butterworth 1990; Roecklein et al. 1985; Trostler et al. 1977). Additionally, Freitas-Silva and colleagues (2010) have shown that maternal thiamine restriction in the perinatal period induced spatial learning deficits that were observed in peri-adolescent but not adult rats (de Freitas-Silva et al. 2010). Extensive work with different patterns of maternal TD, with or without chronic ethanol intake, support cellular differentiation as the critical period for alcohol-thiamine synergism (Bâ 2009, 2011). This synergism also provokes extensive cellular death and tissue necrosis related to FAS. Thus, understanding how apoptotic mechanisms compromise the trajectory of neurodevelopment and brain aging in cases of ethanol exposure and TD will be crucial for developing therapeutic strategies for FAS treatment (Bâ 2017).

4.3. Animal models for thiamine deficiency and pathological aging

Rodent models of thiamine deficiency have been used to reproduce neurological, neuropathological, and neurochemical changes described in WE and WKS patients. TD is induced by feeding adult rats/mice with TD diet in combination with i.p. injections of pyrithiamine, a thiamine pyrophosphokinase inhibitor that accelerates thiamine depletion (review: (Savage et al. 2012; Vetreno et al. 2012). As such, pyrithiamine-induced thiamine depletion (PTD) provides a useful and highly tractable model for examining the neurological consequences of prolonged thiamine depletion in preclinical (rodent) models. A few studies have also used a prolonged ethanol consumption model with or without PTD to understand better how the treatments interact and the independent effects of each treatment (review: (Nunes et al. 2019; Vetreno et al. 2011)).

Over the years, studies have shown that TD in rodents using the PTD model can affect the expression of genes/proteins related to energy metabolism, neuroinflammation, neurotransmitter synthesis (cholinergic, GABAergic, glutamatergic systems), and myelin production in the brain. Additionally, PTD increased oxidative stress and mitochondrial dysfunction, contributing to neurological symptoms like impaired cognitive function and motor incoordination (Liu et al. 2017; Nardone et al. 2013; Nunes et al. 2018; Nunes et al. 2019; Vetreno et al. 2012). Given that molecular and behavioral changes related to TD have been observed in many aging-related neurodegenerative diseases, such as Alzheimer’s disease and Parkinson’s disease, animal models of TD are also being used in research of degenerative processes associated with aging (Ke and Gibson 2004; Nunes et al. 2019).

Significant changes in the thalamus observed in WKS patients have been confirmed by histological and immunohistochemical assessment in brains of PTD rats, with neuronal loss (44 to 83%) in many thalamic nuclei and thalamic mass loss (20–30%) (Anzalone et al. 2010; Hall and Savage 2016; Kipp et al. 2021b; Savage et al. 2021; Vemuganti et al. 2006). Although most studies using the PTD model have been conducted in young male rats (3 months), studies have shown that aging can potentiate the neuropathology associated with TD-- where thalamic lesions were more prevalent in middle-aged (10 months) and aged (22 – 23 months) rats exposed to PTD treatment compared to young rats of 2–3 months (Pitkin and Savage 2001, 2004). In addition, studies have shown that thalamic shrinkage/lesions were prominent in middle-aged rats (8–10 months) after months of recovery from PTD treatment (Kipp et al. 2021b; Vedder et al. 2015). Furthermore, middle-aged rats (9–10 months) treated with PTD in combination with prolonged chronic ethanol consumption also displayed thalamic shrinkage (Kipp et al. 2021b). However, chronic ethanol alone did not result in thalamic lesions (Kipp et al. 2021b; Vedder et al. 2015). The induction of thalamic lesions and the severity of the lesions may be explained by the possible contribution of rapid neuroimmune changes observed in response to TD. Our previous study showed that PTD treatment led to a profound increase of neuroinflammation markers within the thalamus of middle-aged rats (9 months) while chronic ethanol exposure showed small fluctuations in inflammatory genes regardless of brain region examined. In addition, fluctuations in neuroimmune genes varied as a function of vulnerability of brain regions by TD where thalamus showed distinct and rapid neuroinflammation profile during PTD treatment compared with other two regions of interest in ARBD (hippocampus and frontal cortex). The neuroimmune gene induction also varied significantly as a function of stage of TD and time of recovery (Nunes et al. 2019). These results suggested that severity of TD, the order of the TD that co-occurs with ethanol exposure, the age of the animals and mechanisms that contributes to ARBD, such as neuroinflammation, are important determinants for the extent of thalamic pathology associated with AUD (Nunes et al. 2019; Savage et al. 2021).

There is a close relationship between thalamic pathology, learning and memory impairments, and loss of cholinergic neurons in the medial septum/diagonal band (MS/DB) in the PTD rodent model (Hall and Savage 2016; Langlais and Savage 1995; Nardone et al. 2013). However, aging appears to cause greater dysfunction in the MS/DB cholinergic system following PTD treatment; this neuropathology was intensified by the duration of PTD treatment where aged PTD rats exhibited the greatest decline in cholinergic cell numbers with increasing duration of PTD treatment (Pitkin and Savage 2001, 2004). In addition, thalamic pathology found in rats co-exposed to PTD during chronic ethanol (over 9–10 months) correlated with suppression of ACh levels in the frontal cortex measured through in vivo microdialysis during spontaneous alternation but did not correlate with ACh levels in the hippocampus (Kipp et al. 2021b). These findings suggest that aging with TD and chronic ethanol disrupted the thalamocortical circuits to the greatest degree. In order to explore conditions that contribute to the development WKS and the mechanisms underlying the pathological changes, recent studies have explored the effects TD and chronic ethanol consumption alone or combined on behavior tasks dependent on the frontal cortex, cerebellum, and hippocampus. In a study by Moya and colleagues (Moya et al. 2022), behavioral tests were performed during TD and at the end of EtOH exposure, without a withdrawal period. Hyperactivity and deficits in recognition memory was observed in rats treated with EtOH independent of TD, however spatial memory was not affected by any treatments. The combination of chronic ethanol and PTD resulted in disinhibited-like behavior evidenced by more time spent in the inner zone of an open-field arena and in the open arm of the elevated plus maze, suggestive of heightened anxiety. The measures used on the disinhibition tests were correlated with changes in markers of lipid peroxidation (4-hydroxynonenal), apoptosis (caspase 9) and cell damage (HSP70 and HMGB1) in the frontal cortex, as well as a marker of nitrosative stress (nitrite) in the plasma. These changes suggest a synergism between chronic ethanol and PTD aggravates the molecular and behavior changes dependent of frontal cortex in middle-aged rats (Moya et al. 2022).

A recent study (Kipp et al. 2021b), that included both sexes, also found spatial working memory impairments chronic ethanol with or without PTD in male and female rats as they approached middle age (10 months of age). The memory deficit was correlated with reductions in acetylcholine efflux in the prefrontal cortex and hippocampus. The combination of TD and ethanol appeared to affect frontal cortical activity, assessed by acetylcholine efflux, and decision times during attention set shifting. Although there was no sex difference on behavioral and neurochemical disruption, female rats displayed similar impairments even with lower BECs than males, suggesting that female rats may become more sensitive to ethanol toxicity as they advance into middle age.

In summary, data across several studies confirm that TD causes severe neurobehavioral impairments, including deficits in spatial and working memory, and behavioral inflexibility when compared with ethanol exposure alone. Consistent with this, chronic alcohol appears to exacerbate the effects of TD across the lifespan on behavioral deficits and suppression of frontal cortical activity. Overall, chronic ethanol consumption, TD or both conditions together, damage critical neural circuits, but to different degrees, and these differences can be modulated by the age of the animals, severity of treatment, ethanol concentration, timing, recovery, and sex.

4.4. Conclusions: Similarity between WKS and Dementia

AUDs are associated with an elevated risk of all types of dementia (Rehm et al. 2019). Some studies have used the term alcohol-related brain damage to include neurocognitive disorders related to chronic alcohol use, including WKS and ARD. Similar to WKS, ARD seems to be a direct result of TD and possibly ethanol neurotoxicity, suggesting that ARD has a multifactorial etiopathology (Arts et al. 2017; Ridley et al. 2013). Although evidence suggesting that the two disorders have overlapping clinical symptoms, such as peripheral neuropathology and ataxia, the cognitive profile of ARD entails impairments in visuospatial function, memory, and executive tasks, whereas WKS patients show deficits on executive tasks in conjunction with memory deficits (Ridley et al. 2013; Smith and Atkinson 1995). A recent study showed that both disorders are more prevalent in men (Palm et al. 2022). In addition, WKS most commonly is first diagnosed in people aged 50–59 years, in contrast to ARD that occurs commonly in people aged 70–79 years. Both diseases tend to be diagnosed at a younger age (middle age to middle old) relative to other progressive neurocognitive disorders (Palm et al. 2022).

In summary, WKS and AD are two major neurocognitive disorders that lead to memory impairment. Diencephalic amnesia is most described in WKS patients (Segobin and Pitel 2021). In contrast, AD shows a medial temporal lobe amnesia, with hippocampal atrophy at the early stage and progressively extending to neocortical areas (Braak and Braak 1991). However, both disorders damaged in the same degree the anterior thalamic nuclei, cingulate cortex, and hippocampus (only in moderate AD), with the mediodorsal thalamic nuclei and mammillary bodies more severely damaged in WKS than AD (Segobin et al. 2023). These findings reinforce the importance of examining brain networks involved in memory function and how they are disrupted in these neurocognitive disorders.

According to Gibson and colleagues (2022), WKS and AD have similarities in clinical manifestations and molecular mechanisms, and both diseases have TD as a common factor. Neuronal loss, increased neuroinflammation, cholinergic dysfunction, alteration of neurofilaments, exacerbation of amyloid plaques are common features of TD in rodents, (review: (Gibson et al. 2022; Hazell et al. 2001; Ke and Gibson 2004; Kopelman 1991) with aggravation of these effects in transgenic models of AD (Calingasan et al. 1996; Karuppagounder et al. 2009). In conclusion, thiamine deficiency, chronic ethanol consumption and aging are key factors that increase susceptibility to ARBD and other dementias. Alone or combined, thiamine deficiency and ethanol become increasingly relevant as individuals age, causing serious health issues, including cognitive impairments and dementia, which appear to be more pronounced with aging. However, future research is needed to better understand the genetic, molecular, and neurobehavioral changes that emerge across the lifespan due to natural aging, TD and chronic ethanol consumption.

5. Alcohol use disorder as a risk for AD and related dementias.

AD is the most common form of dementia, affecting ~30 million people worldwide (Holtzman et al. 2011). It is characterized by the aggregation of extracellular amyloid plaques, the accumulation of intracellular neurofibrillary tau tangles, and neurodegeneration, which begins to accumulate 15–20 years before the onset of clinical symptoms (Jack et al. 2010). Thus, it is important to identify factors that can reduce or prolong this presymptomatic period. While a few studies suggest that low-to-moderate ethanol consumption may reduce the risk of AD (Koch et al. 2019; Luchsinger et al. 2004), several epidemiological studies have identified alcohol use disorder (AUD) as a risk factor for AD and AD-related pathology (Harwood et al. 2010; Rehm et al. 2019; Schwarzinger et al. 2018). Preclinical studies provide additional evidence that chronic ethanol exposure drives AD-related pathology, however the biological mechanisms linking the two conditions are poorly understood. In the following sections, we will review the current literature describing the relationship between AUD and AD. We will begin by describing the pathological biomarkers associated with AD, namely amyloid-β (Aβ) and tau. We will then review clinical and preclinical studies characterizing how alcohol impacts AD pathology in humans and transgenic animal models. Lastly, we will review potential mechanisms by which chronic alcohol misuse promotes AD pathology.

5.1. AD-related pathological markers.

Amyloid plaques are primarily comprised of Aβ, a post-translational cleavage product of amyloid precursor protein (APP). APP is a transmembrane cell surface protein that is broken down through competing pathways that produce pathologically inert substrates (non-amyloidogenic) or Aβ (amyloidogenic) (Haass et al. 2012). In the nonamyloidogenic pathway, APP is cleaved midway through the Aβ domain by an α-secretase enzyme, which produces a truncated APP C-terminal fragment-α (CTF-α). CTF-α is subsequently broken down by γ-secretase (Haass et al. 2012). In the amyloidogenic pathway, APP is cleaved by a β-secretase enzyme at the N-terminus end of the Aβ domain, producing a CTF-β peptide. CTF-β is then processed by γ-secretase to produce Aβ peptide. Aβ40 and Aβ42 are the most abundant Aβ products, with this pathway favoring Aβ40 production over Aβ42 (Citron et al. 1992; Haass et al. 2012). Of the two, Aβ42 is more aggregate-prone than Aβ40 and CSF Aβ42 levels decrease with AD progression, indicating increased deposition in the brain; CSF Aβ40 levels are unchanged with AD progression (Blennow et al. 2015). In fact, a reduced CSF Aβ42/40 ratio is a biomarker of AD progression and is a predictor of elevated phosphorylated tau levels (Wiltfang et al. 2007; Hansson et al. 2007). In neurons, Aβ is produced in endosomes and released at the synapses in an activity-dependent manner (Bero et al. 2011; Cirrito et al. 2008; Cirrito et al. 2005; Hettinger et al. 2018; Verges et al. 2011). Once in the extracellular space Aβ forms oligomers and aggregates into extracellular plaques in a concentration-dependent manner (Yan et al. 2009). While the severity of Aβ pathology does not correspond to cognitive decline, plaque pathology precedes tau pathology and neurodegeneration by several years (Jack et al. 2010; Musiek and Holtzman 2015). Tau protein is primarily expressed in neurons, where it binds to tubulin and functions to stabilize microtubules and helps regulate axonal transport (Wang and Mandelkow 2016). Tau contains 85 phosphorylation sites and is phosphorylated by several kinases (e.g., GSK3, Cdk5, MAPK, PKA, CaMKII) (Guo et al. 2017). Normal phosphorylation of tau regulates its distribution and function along the axon. However, tau hyperphosphorylation decreases its binding affinity to microtubules, subsequently promoting aggregation into paired helical filaments and neurofibrillary tangles (Castellani and Perry 2019). Tau pathology is a better predictor of cognitive impairment than Aβ and correlates with cognitive decline and neurodegeneration (Giannakopoulos et al. 2003; Nelson et al. 2012).

In AD, tau pathology begins to accumulate in the entorhinal cortex, locus coeruleus, and medial temporal lobes (Beardmore et al. 2021; Crary et al. 2014). As AD progresses, pathological tau propagates trans-synaptically in a prion-like manner eventually spreading throughout the neocortex (Guo and Lee 2011). Recent studies have provided evidence that basal forebrain degradation precedes neurodegeneration in the entorhinal cortex. Longitudinal neuroimaging data showed that degradation in the nucleus basalis precedes neurodegeneration in the entorhinal cortex, and this degradation is exacerbated by the presence of Aβ and tau (Fernández-Cabello et al. 2020; Schmitz and Nathan Spreng 2016). According to the A/T/N (amyloid/tau/neurodegeneration) framework model of AD, Aβ, tau, and neurodegeneration are all required for a diagnosis of AD. Therefore, we will discuss how alcohol drives these pathological biomarkers in humans and rodent models of pathology in the following sections.

5.2. Human Studies: The role of AUD as a risk factor for AD.

While epidemiological studies have begun to identify AUD as a risk factor for AD, there are conflicting studies on the degree to which alcohol use and misuse impact dementia, and the mechanisms by which it promotes cognitive decline. A 2004 study conducted on the Washington Heights-Inwood Columbia Aging Project cohort, which included Medicare beneficiaries aged 65 and over, reported that light and moderate alcohol intake (1 serving/month to 3 servings/day) was associated with a lower risk of dementia in individuals without the ApoE4 allele (Luchsinger et al. 2004). Similarly, a meta-analysis evaluating alcohol intake and dementia risk from 15 epidemiological studies conducted across multiple countries, found that moderate drinkers (<40 g/day) had a lower dementia risk than individuals who abstained from alcohol (Mewton et al. 2023). One study using Pittsburgh Compound-B ([¹¹C]PiB) PET imaging compared amyloid deposition and cortical thickness in healthy middle-aged adults (40–65 years-old, n=20), and those with a history of AUD (40–65 years-old, n=19). The researchers hypothesized that individuals with a history of AUD would show greater Aβ deposition; however, Aβ was not detected in either group (Flanigan et al. 2021). Despite this, the study reported that individuals with AUD showed reduced cortical thickness in the inferior temporal gyrus, middle temporal gyrus, and fusiform gyrus, as well as reduced grey matter volumes in the hippocampus (Flanigan et al. 2021). An MRI study conducted in 2021 reported that individuals with AUD showed smaller grey matter volume in the sensorimotor complex and hippocampal formation (Zhornitsky et al. 2021). While reduced cortical thickness and grey matter in these areas is associated with AD progression, it is typically preceded by the presence of amyloid and tau pathology (Dickerson et al. 2009; Mattsson et al. 2014). Thus, alcohol misuse may promote AD-related neurodegeneration before the onset of amyloid and tau pathology.

Conversely, several epidemiological studies have reported a stronger relationship between heavy alcohol misuse and dementia or AD. A nationwide cohort study conducted in France from 2008–2013 reported men and women with a history of AUD had a 3-fold increased risk of dementia, while 16.5% of men and 4% of women with dementia had a history of AUD (Schwarzinger et al. 2018). Another nationwide cohort study from Finland identified alcohol misuse as the strongest modifiable risk factor for dementia and found that it was associated with a 2.2-fold increased risk for vascular dementia and a 1.6-fold increased risk for AD (Kauko et al. 2024). In the DELIVER cohort (Decoding the Epidemiology of LIVER disease in Sweden; 1987–2020), individuals with AUD had a 4.6-fold increased risk of dementia (Zhao et al. 2023). Finally, in the Whitehall II study, 35–55-year-old participants were recruited in London between 1985 and 1988. Alcohol consumption was frequently assessed (8 times between 1985 and 2016) and dementia outcomes were reported through national databases until 2017. This study reported an interesting relationship between dementia outcomes and alcohol abstinence, moderate alcohol use (1–14 units/week), and heavy alcohol use (>14 units/week). The association between alcohol use and dementia appeared to follow a U-shaped curve, with moderate alcohol use conferring a lower risk for dementia than abstinence or heavy alcohol use (Sabia et al. 2018). This study links conflicting studies discussed above, demonstrating that moderate alcohol consumption may be protective against AD and dementia, whereas heavy alcohol misuse may exacerbate dementia. Collectively, these epidemiological, population cohort, and clinical studies have identified alcohol misuse as a novel risk factor for AD and AD-related dementias and have laid the foundation for a new field of research.

It is important to note, however, that these approaches have a few important limitations. First, many of these studies rely on self-reporting for alcohol intake assessment, which could be unreliable in some cases. Furthermore, individual differences in biological variables that impact ethanol metabolism (e.g., sex/gender, height, age, body composition, genetics) could influence AD risk and progression. Next, many of these studies did not report the diagnostic criteria with which an AD diagnosis was made. Thus, it is unknown whether individuals in these studies identified as AD patients truly met the appropriate diagnostic standard (i.e. A/T/N framework). Lastly, studies using human subjects are limited in their ability to manipulate environmental or genetic factors and their ability to characterize clinical and pathological outcomes in individuals. Thus, preclinical studies using animal models can provide further details on the genetic and environmental factors that drive AD-related pathology under the influence of alcohol. In turn, these studies can identify treatments or interventions that will limit the risk of AD in AUD patients.

5.3. Preclinical evidence for AUD as a driver of AD-related pathology.

As discussed in the chapter by Anton and colleagues, preclinical studies have begun to characterize and explore the mechanistic links between alcohol exposure and AD-related pathology. In one study, adult 3-month-old 3xTg-AD mice (Table 1) were exposed to voluntary ethanol intake for four months, then euthanized one month after ethanol cessation. Ethanol exposure was associated with increased Aβ42 levels in the lateral entorhinal cortex and prefrontal cortex (Hoffman et al. 2019). Tau levels were also increased in the lateral entorhinal cortex, medial prefrontal cortex, and amygdala, and there was increased p-tau in the CA1 region of the hippocampus (Hoffman et al. 2019). In another study, 10 weeks of ethanol exposure via a moderate two bottle choice drinking paradigm, 8-month-old APP/PS1 mice (Table 1) showed an increased number of plaques in the hippocampus. Interestingly, there were a greater number of smaller plaques in the hippocampus and cortex in ethanol-treated mice, suggesting that ethanol may be either limiting plaque growth, or promoting greater plaque proliferation. Despite this increased Aβ deposition and proliferation, ethanol exposure did not increase APP, CTF-β, or BACE-1 protein levels (Day et al. 2023). Ethanol-treated APP/PS1 mice also had reduced brain mass compared to APP/PS1 controls, indicating that long-term ethanol exposure may promote neurodegeneration in the presence of Aβ overexpression (Day et al. 2023). In another study, five-month-old 3xTg-AD mice were treated with 5 g/kg of ethanol 5 days/week for 3 months, then aged to 14 months without any additional ethanol. Western blot experiments showed that hippocampal Aβ42 levels were increased in ethanol-exposed female mice, but not males. Ethanol exposure also increased cortical and hippocampal t-tau as well as cortical p-tau levels, but only in female mice (Tucker et al. 2022). Collectively, these studies demonstrate that ethanol exposure during adulthood can exacerbate amyloid burden, tau pathology, and neurodegeneration.

Table 1:

Selective list of transgenic rodent models of AD-related pathology used in alcohol studies described in this chapter. Included are the model names, strain background, species, and types of AD-related pathology expressed.

Model	Strain, Species	Amyloid Pathology	Tau Pathology	References
3xTg-AD	B6;129, mouse	Yes	Yes	(Oddo et al. 2003)
APP23/PS45	C57BL/6, mouse	Yes	No	(Huang et al. 2018)
APP/PS1	C57BL/6J; mouse B6C3, mouse	Yes	No	(Jankowsky et al. 2004)
Tg2576	B6SJL; mouse	Yes	No	(Hsiao et al. 1996)
P301S	C57BL/6J; mouse	No	Yes	(Yoshiyama et al. 2007)
Tg-F344-AD	Fischer 344, rat	Yes	Yes	(Cohen et al. 2013)

Adolescence is an especially vulnerable period of neurodevelopment and binge-like ethanol exposure during this period can have long-lasting neurobehavioral consequences (Crews et al. 2019; Spear 2018), which may extend to an increased risk for AD and AD-related pathology. Six-week-old APP23/PS45 mice (Table 1) were exposed to ethanol via a drinking-in-the dark paradigm for 4 weeks (4 hours of ethanol exposure during their dark cycle). Mice were euthanized after two days of withdrawal and brains were homogenized for Western blot experiments. In these mice, ethanol exposure increased APP, BACE-1, and Aβ40 and Aβ42 protein levels, which translated to an increased number of amyloid plaques in the hippocampus (Huang et al. 2018). In another study, adolescent 3xTg-AD mice were exposed to an adolescent intermittent ethanol (AIE) exposure paradigm for 30 days (5 g/kg; 2 days on, 2 days off; P25-P55), and pathology was assessed when mice were 6 months-old. AIE-exposed mice had greater intracellular Aβ42 staining in the subiculum, entorhinal cortex, and amygdala (Barnett et al. 2022). Western blot experiments also showed that AIE-treated mice had increased p-tau-181 in the hippocampus, but not in the cortex (Barnett et al. 2022). Lastly, adolescent ethanol exposure also led to increased hippocampal Aβ42 protein levels at 6- and 12-months of age (Ledesma et al. 2021).

There is also evidence that ethanol exposure during the adolescent period accelerated the appearance of cognitive decline in AD models. In female 3xTg-AD mice, AIE did not induce impairments in novel object memory, but did reduce recall of spatial memory in the Morris water maze and decreased exploration (Barnett et al. 2022). In the APP/PSEN model, AIE dramatically increased time to reach the goal box and errors made in the Hebb maze selectively at middle age (Ledesma et al. 2021). Prenatal alcohol exposure (PAE) may also accelerate AD-related cognitive decline. 3xTg-AD mice exposed to a PAE paradigm showed spatial memory impairments and behavioral inflexibility at 4 months of age, compared to unexposed 3xTg-AD mice. Interestingly, PAE and unexposed 3xTg-AD mice showed similar behavioral deficits by 6 months of age (Tousley et al. 2023). These cognitive deficits also corresponded with reduced GABAergic interneuron function. PAE-treated 3xTg-AD mice showed deceased spontaneous inhibitory post-synaptic current (sIPSC) frequency which corresponded with reduced PV+ GABAergic interneurons in the mPFC at four months of age, indicating a potential hyperexcitability phenotype (Tousley et al. 2023). In the same study, 4-month-old PAE 3xTg-AD mice showed elevated Aβ immunofluorescence in the medial prefrontal cortex (mPFC), however these particular results should be viewed skeptically given a small n (n=3 per group), as the findings are likely underpowered (Tousley et al. 2023). Future studies examining how PAE drives AD-related pathology should seek to expand these findings. Collectively, these emerging studies demonstrate that ethanol has a clear impact on Aβ deposition, and identify alcohol misuse as an early, modifiable risk factor for AD. Importantly, these studies also provide evidence that chronic ethanol exposure during multiple phases of life (i.e., prenatal, adolescence, early adulthood, middle adulthood) leads to increased AD-related pathology during late adulthood. Despite this growing body of evidence clearly showing a relationship between chronic ethanol exposure and AD-related pathology, the mechanisms linking the two remain unknown. The data suggest that alcohol exposure significantly influences the trajectory of AD-like brain pathology, effects that appear to compound with age. Therefore, any potential mechanism must be one that is both persistent after the cessation of ethanol and one that drives AD-related pathology. Thus, the following section will discuss two potential mechanisms that preclinical studies are currently focused on: neuroinflammation and neuronal hyperexcitability.

5.4. Potential mechanisms driving AD-related pathology.

5.4.1. Neuroinflammation

A few studies have begun to investigate how ethanol exposure impacts microglia and neuroinflammation in the context of AD-related pathology. In one study, Aβ42 phagocytosis was reduced in rat microglia cultures after ethanol exposure in vitro (Kalinin et al. 2018). Conversely, another study showed that ethanol exposure may promote Aβ phagocytosis when ethanol consumption occurred between the ages of 10–14 months in F344 rats. Specifically, immunofluorescent labeling showed that iba1+ cells co-localized with Aβ42 in female, but not male, F344 rats who consumed 10% ethanol following a 2-day on/2 day off, single bottle procedure for 6 months (Marsland et al. 2023). These conflicting results may be due to differences in models (cell culture vs animal model) or ethanol exposure length (acute vs chronic) but are certainly in accord with human studies showing that patterns of drinking (moderate vs heavy consumption) may produce opposite outcomes (Livingston et al., 2020). Thus, further studies are needed to better understand how ethanol influences Aβ phagocytosis. Despite these differences there appears to be consensus of the long-term impact of chronic ethanol exposure on neuroinflammation. In one study, 3xTg-AD mice were exposed to a binge-like ethanol exposure paradigm (5 g/kg; 2 days on/2 days off) for 30 days during adolescence (P25-P55), then aged to 6.5 months without any additional ethanol. Ethanol-exposed mice showed increased levels of the proinflammatory cytokines INFα, IL1β, MCP1, TNFα, and TLR4. Interestingly, cytokine levels positively correlated with Aβ42 and p-tau-181 protein levels (Barnett et al. 2022). In another study, female 3xTg-AD mice were exposed to a binge-like ethanol exposure (5.0 g/kg; 5 days/2 days off) for 3 months (at age 5–8 months), then aged to 14 months without any additional ethanol; t-tau and p-tau-214 were increased in the cortex while Aβ42 was increased in the hippocampus. Importantly tau and Aβ levels were positively correlated with neuroimmune gene expression (Tucker et al. 2022).

Our laboratories have recently used the TgF344-AD model to examine whether adolescent ethanol, both chronic drinking and AIE, alter the progression of AD-related cognitive decline and neuropathological markers. In the chronic drinking models, with moderate BEC’s it was found relatively minimal effects of lifelong alcohol consumption on cognitive function, despite signs of increased anxiety in multiple neurobehavioral tasks. When plaque pathology was examined, we observed sex-specific effects in which females showed increased colocalization of Aβ42 in iba1+ cells (Marsland et al., in prep). In contrast, heavy binge-type ethanol exposure using the AIE model found sex-specific effects of AD transgenes, AIE, and their interaction manifested in distinct behavioral and neurobiological outcomes. In male rats carrying AD transgenes, spatial navigation deficits were evident by 3 months of age that were unaffected by adolescent ethanol exposure. Furthermore, AD transgenes combined with AIE led to pathological changes characterized by increased pTau levels and a shift in the balance of proNGF/NGF receptors favoring cell death mechanisms. Conversely, in female TgF344-AD rats, AIE accelerated AD-induced cognitive decline. This was accompanied by a reduction in Tropomyosin receptor kinase A (TrkA) receptors due to AIE and a decrease in Vesicular acetylcholine transporter (VAChT) resulting from AD transgenes. The observed dysregulation in the cholinergic system suggests its potential contribution to the exacerbated behavioral impairments in female rats carrying AD transgenes exposed to adolescent ethanol.

Together, these studies provide evidence that chronic ethanol exposure has lasting effects on AD biomarkers and a proinflammatory phenotype (see also Section A in this chapter), especially in models of heavy ethanol exposure. However, another central problem facing the field is the extensive reliance on transgenic models of AD in determining the relationship between alcohol exposure and AD risk, which may overstate this association due to the relatively low incidence of familial AD. Therefore, future studies should investigate whether ethanol drives neuroinflammation through increased pathology or drives pathology through increased neuroinflammation in both genetically vulnerable (familial AD) and genetically typical (sporadic AD) rodent models.

5.4.2. Neuronal Excitability

AUD is characterized by neuronal hyperactivity during periods of withdrawal (Ariwodola and Weiner 2004; Cheaha et al. 2014; Slawecki et al. 2006; Wang and Mandelkow 2016). In one study, chronic intermittent ethanol (CIE) exposure increased the frequency of spontaneous interictal spikes and lowered seizure thresholds throughout the brain (Alberto et al. 2023). Ethanol exposure bidirectionally altered N-methyl-D-aspartate receptor (NMDAR)-mediated synaptic activity in multiple regions of the brain during periods of intoxication and withdrawal. In the agranular insular cortex, acute ethanol exposure inhibited NMDAR currents and disrupted long-term potentiation (LTP) (Shillinglaw et al. 2018). In CIE-treated mice, acute ethanol exposure reduced NMDAR-mediated excitatory synaptic transmission in the central amygdala (Roberto et al. 2004). During withdrawal from a 10-day CIE treatment, NMDAR function was increased in the basolateral amygdala during withdrawal (McGinnis et al. 2020). Withdrawal from CIE treatment also enhanced Glun2B-dependent LTP in the bed nucleus of the stria terminalis (Wills et al. 2012). These findings are relevant in the present context because neuronal activity also regulates Aβ and tau pathology in rodent models of Aβ and tau overexpression. In vivo microdialysis experiments demonstrated that pharmacological stimulation of neuronal activity increased tau levels in vivo, whereas pharmacological inhibition had no effect on tau (Yamada et al. 2014). Electrical, pharmacological, and optogenetic stimulation of neuronal activity also increased Aβ levels in vivo, whereas pharmacological inhibition decreased Aβ levels in Tg2576 mice (Bero et al. 2011; Cirrito et al. 2008; Cirrito et al. 2005; Yamamoto et al. 2015). Consequently, targeting neuronal activity may reduce Aβ deposition over time. Consistent with this, treatment with the NMDAR antagonist memantine decreased Aβ levels, amyloid plaque formation, and behavioral deficits in mouse models of Aβ overexpression (Dong et al. 2008; Stazi and Wirths 2021). In 7-month-old APP/PS1 mice, 28 days of unilateral vibrissal deprivation reduced amyloid plaque formation and growth in the barrel cortex of the corresponding hemisphere (Bero et al. 2011). Given the innervation between the vibrissae and the barrel cortex, this decrease was likely due to reduced neuronal activity as there were no differences between microglial or astrocytic activation (Bero et al. 2011). Thus, ethanol-induced disruptions in the brain’s excitatory/inhibitory balance may ultimately exacerbate the activity-dependent production and propagation of Aβ and tau.

A few studies have begun to investigate how chronic ethanol alters brain excitability in rodent models of AD-like pathology. In one study, 5.5-month-old APP/PS1 mice were exposed to a moderate 2-bottle choice drinking paradigm (20% ethanol, 12 hrs/day, 4 consecutive days/week) for 10 weeks. Ethanol-treated APP/PS1 mice had increased cortical GluN2B mRNA levels, compared to ethanol-treated wildtype mice. In the same study GABA_AR, α5 subunit mRNA levels were elevated in water-treated APP/PS1 mice, which was decreased in ethanol-exposed APP/PS1 mice (Day et al. 2023). This data suggests that even moderate levels of ethanol exposure may disrupt the brain’s excitatory/inhibitor balance in APP/PS1 mice. In another study, 3-month-old P301S mice were exposed to an intermittent 2-bottle choice drinking paradigm (20% ethanol, 24 hrs/day, Monday/Wednesday/Friday) for 16 weeks. In locus coeruleus (LC) neurons, the action potential threshold was decreased in ethanol-exposed males and females, in both wildtype and P301S mice. This translated to increased neuronal excitability in the LC of ethanol-exposed male and female, wildtype and P301S mice; ethanol-exposed P301S female mice showed the greatest increase in excitability (Downs et al. 2023). Thus, chronic ethanol exposure throughout early- and middle-adulthood rendered LC neurons to a hyperexcitable state, potentially driving AD-related pathology. Collectively, these studies demonstrate that chronic ethanol exposure during early- and middle-adulthood exacerbated brain excitability in the context of AD-related pathology. In these studies, measures of excitability were taken during an early withdrawal period (~72 hrs post-ethanol). Furthermore, these studies were also conducted in animals at an age when amyloid or tau pathology first begin to appear. It is unclear how ethanol-induced brain excitability changes over time and with age. Thus, future studies should consider the long-term consequences of ethanol exposure during different phases of development on brain excitability and AD-related pathology. Furthermore, as discussed with the ethanol-associated neuroinflammatory phenotype, it is unclear whether ethanol-induced hyperexcitability is the cause or consequence of increased pathology. Thus, future studies investigating this mechanism should seek to characterize the directionality of this relationship.

5.5. Concluding remarks

Ongoing research continues to provide evidence that alcohol misuse or chronic ethanol exposure increases the risk of AD and drives AD-related pathology; however, there are a few key issues that need to be addressed in this growing field. While epidemiological and clinical studies provide evidence that heavy alcohol misuse increases the risk for dementia and AD, mild-to-moderate alcohol usage may offer a protective effect relative to complete abstinence. Here, we identified neuroinflammation and neuronal hyperexcitability as potential mechanisms by which alcohol misuse exacerbates AD-related pathology. It is important to note, however, that alcohol has widespread effects on the nervous system and peripheral organs, and that AD is a multifaceted disease with many genetic and environmental risk factors. Thus, the relationship between AUD and AD may not be limited to a single mechanism. Moreover, it is also important to account for bidirectionality between the systems disrupted by alcohol misuse and those exacerbating AD-related pathology. Thus, future research in this field should be done from a multidisciplinary approach.

6. General Conclusions

Natural aging is associated with a wide range of cellular, structural, and circuit-level disruptions that collectively contribute to aging-related cognitive decline, mild cognitive impairments, and for some individuals, ADRD. The contribution of lifestyle factors such as patterns, frequency, and intensity of alcohol consumption appear to modify the trajectory of healthy brain aging, and the mechanisms contributing to these changes are just now beginning to emerge. The aged brain is associated with both heightened basal inflammation and delayed recovery of inflammatory processes, at least in part due to deficits in anti-inflammatory shutoff mechanisms that govern inflammation in the young brain. Emerging evidence suggests that natural aging-related changes in inflammation may contribute to earlier emergence of dementia, development of alcohol-related brain damage due to excessive inflammation (directly), or through thiamine deficiency (indirectly). Binge-like alcohol exposure during adolescence, a developmental period during which ethanol consumption typically peaks for most individuals, is also associated with reduced cholinergic output from basal forebrain cholinergic neurons. In this sense, it is perhaps noteworthy to mention that cholinergic signaling has been shown to have moderate anti-inflammatory effects itself. If caught early, loss of the cholinergic phenotype after chronic ethanol (with or without thiamine depletion) may be rescued by thiamine replacement, regular exercise, and other corrective measures that provoke a pro-neurotrophic response. However, once diencephalic damage associated with severe thiamine depletion is instantiated, cognitive deficits become irreversible. The same also appears to be true in the case of ADRD; as alcohol consumption and aging progress, the accumulation of cardinal neuropathological features of ADRD (amyloidopathy and tauopathy) appears to set the aging brain onto an irreversible course of cognitive dysfunction for which no effective treatments currently exist.

Some caveats to this framework should also be considered. For instance, substantial evidence from human epidemiological studies supports the notion that low to moderate levels of alcohol consumption may confer a slight protective benefit against dementia relative to non-drinkers. Furthermore, most preclinical studies examining alcohol-induced modulation of ADRD-like pathology have shown effects predominantly in transgenic rodent models of AD, which may bear relevance more to familial AD (~5% of AD) than sporadic AD (~95% of AD cases). In contrast, there is a relative paucity of studies examining the myriad of genetic and environmental factors that might confer a protective benefit that mitigates the untoward effects of alcohol consumption on the progression of cognitive dysfunction into dementia and its associated neurological diagnoses (MCI, WKS, ARBD, ADRD). To be sure, much work remains to be done.