In the United States, 18 million individuals (7.4% of the 15 and older population, according to estimates from 2010) report having an alcohol use disorder (AUD), with nearly 12 million of these individuals reporting alcohol dependence (1). Recent changes to the diagnostic definition of AUDs in the updated DSM-V eliminate the clinical distinction between AUDs and alcohol dependence, opting to categorize them together under the umbrella category of AUDs and describe the broad disorder as a “… problematic pattern of alcohol use leading to clinically significant impairment or distress …” as well as requiring concurrent escalation of alcohol intake, craving for alcohol, and significant disruptions to personal and professional conduct (2). In 2011, AUDs cost the United States $223.5 billion, an estimation which includes the cost of medical treatment, judiciary involvement, and loss of productivity (3).

However, these statistics, while useful in conveying the gravity of the alcohol abuse problem in the United States, do not provide insight into the recovery process nor the continuing health and cognitive disparities these individuals face into periods of abstinence from alcohol consumption. Additionally, long-term alcohol abuse results in significant, non-economic personal costs, including devastating bodily harm, with some of the most striking effects apparent in the brain. Evidence from human and animal studies suggest that select regions of the cortex, particularly the prefrontal cortex (PFC) and hippocampus, may be more sensitive to the deleterious and damaging effects of long-term alcohol use than others, and recovery of cognitive function sensitive to these regions may occur at different times into periods of prolonged abstinence (4–7).

Alcohol is widely known to acutely alter cortical function by modulating inhibitory and excitatory receptor function on neuronal processes (8–10). By repressing excitatory transmission (8, 11–15) and concurrently enhancing inhibitory transmission (16–21), alcohol acutely acts as a systemic depressant. Over repeated, chronic exposures, neuronal transmission achieves a homeostatic state in the presence of alcohol (22), and cognition can resemble that of non-dependent function. However, during periods of abstinence when alcohol is absent from the system for extended phases, effectively disrupting the previously described modified homeostasis, cognitive function is significantly impaired (due to the absence of alcohol as critical modulating factor), and these cognitive impairments persist for some time. Interestingly, these cognitive perturbations, in some instances, do recover to or near pre-dependency levels. What follows is a description and synthesis of how alcohol modulates PFC and hippocampal function, what changes occur as occasional alcohol consumption becomes chronic consumption, and what cognitive impairments are present during acute withdrawal.

It is worth noting, while outside the general scope of this review, that chronic alcohol use does result in structural and/or functional atrophy in regions outside of the PFC and hippocampus and that these additional changes cannot be eliminated as potential modulators of the deleterious effects observed in the PFC and hippocampus (23). Further, research into the cognitive capacities of alcoholic individuals has identified cognitive disorders, such as Wernicke–Korsakoff syndrome, alcohol dementia, and Marchiafava–Bignami disease, which are directly related to long-term alcohol abuse and cloud our understanding of alcohol’s solitary effects on cognitive functioning (24, 25). Similarly, age and concurrent drug use can additionally complicate our understanding of alcohol’s impact; therefore, for the purpose of this review, studies including subjects with chronic alcohol use without poly drug use were evaluated.

A strong body of evidence in alcohol-dependent individuals has demonstrated that various cognitive capacities do return to (or nearly to) non-dependence levels of performance. However, the details of this recovery vary widely in terms of temporal resolution based primarily on the cortical structure of interest, and it is difficult to disseminate apparent recovery of damaged regions from compensation by other cortical regions with regards to behavioral function and performance alone. For example, studies appear to suggest that cognitive deficits due to PFC damage from alcohol abuse recover on a shorter timescale compared with those dependent on the hippocampus. However, as the functionality of the PFC and hippocampus is intricately related, there is a clear challenge to designing studies to directly address the explicit temporal recovery of specific structures in humans. Therefore, the findings presented here are from studies addressing broader questions of functionality in alcoholics.

With respect to the PFC damage, recovery of cognitive function in this region is critical to the persistence of abstinence from alcoholism and avoidance of relapse in dependent individuals (97). A recent met-analysis of human literature (62 sources in all) demonstrated that cognitive impairments sensitive to the PFC in individuals with AUDs identified in recent abstainers (98–101) are primarily alleviated or “normalized” (meaning performance is comparable to non-dependent individuals) by 1-year abstinence of alcohol use (102). Similarly, improvements in executive functioning occurring as soon as 6 months into abstinence has been reported (95, 103). However, as proposed and reviewed by Oscar-Berman et al. (104), it is plausible that the recovery of PFC function is more the result of compensatory activity in associated regions of the cortex rather than distinct recovery or repair of the PFC itself.

With regard to hippocampal functionality, human studies evaluating episodic memory in dependent, long-term abstinent individuals have reported similar findings to those relating to the PFC, but the outcomes of the studies have not been entirely equivocal. For example, multiple studies have reported impaired performance on tasks of episodic memory (105–107), and that “normalization” of episodic memory performance in alcohol-dependent subjects has taken place by 1 year of abstinence (95). However, there is evidence that hippocampal dysfunction remains impaired years after abstinence (5, 108). The potential distinction of these two seemingly disparate findings may be the result of (A) many of the studies not evaluating function beyond 1-year abstinence and (B), as described previously, episodic memory is not entirely exclusive of hippocampal function. Therefore, it is possible that, while episodic memory function returns, other facets of hippocampal function remain perturbed long into abstinence from alcohol. Taken together, the current evidence suggests that the recovery of cognitive functionality in abstinent alcohol-dependent individuals is sensitive to the duration of the abstinence period, with the PFC returning to “normative” levels prior to the hippocampal formation.

A wealth of evidence from clinical findings demonstrates that acute alcohol exposures can inhibit cognitive capacities. Interestingly, it is primarily following withdrawal from chronic alcohol exposure that individuals experience persisting, severe cognitive impairments. As eloquently described in Oscar-Berman et al. (104), studies involving human subjects and drugs of abuse are often rife with complicating and confounding factors, including family history, genetic predisposition, and past life events and experience, much of which cannot be controlled for. While clinical studies are limited to observational investigations into the deleterious cortical adaptations subsequent to chronic alcohol exposure, preclinical models have been successful at informing and elaborating our understanding of the cellular and molecular changes, which may explain the mechanisms underlying cognitive disparities in abstinent alcohol-dependent subjects. Further, preclinical models of alcohol dependence have generated evidence suggesting that the distinct cellular compositions of the PFC and the hippocampus may be the basis for the differential cognitive recovery in these regions in abstinent individuals. Therefore, the following sections will discuss preclinical models of alcohol addiction and dependence with specific focus on cognitive impairments dependent on the PFC and hippocampus and will elucidate the associated cellular and molecular changes in these regions.

Rodent models of alcohol dependence have been instrumental in furthering our understanding of both the cognitive and neurobiological impact of withdrawal from alcohol dependence, as well as providing critical insight into the potential mechanisms of the pathological state associated with and resulting from alcohol withdrawal in dependent animals. While studies targeting examination of one explicit region or feature are impossible in human populations, particularly with regards to the effects of drugs of abuse, animal models have been instrumental tools in allowing for the fine manipulation of explicit cortical regions and functions.

Multiple studies employing rodent models have investigated the impact of alcohol dependence on prefrontal cognitive capacity. Growing evidence suggests that the rodent medial prefrontal cortex (mPFC) likely represents a functional homolog of the human medial and dorsolateral PFC (109). Reports using various rodent models of alcohol dependence [including chronic intermittent ethanol vapor exposure (CIE), liquid diet, two bottle choice; for paradigm overviews, see Ref. (110)] have found behavioral inflexibility (111), impaired extinction (112), impaired set-shifting (113), and impaired working memory (114, 115), all tasks which require a fully functioning PFC. Further, two of these studies (112, 113) linked the disruption in frontal cortical function to alcohol-induced dysregulation of the N-methyl-d-aspartate glutamatergic receptor (GluN) system. Two studies have investigated PFC functions into periods of abstinence following chronic ethanol exposure via CIE (10 days abstinence; (116)), or liquid diet (6 weeks abstinence; (114)). Interestingly, at 10 days into abstinence there is a lack of impairment in cognitive flexibility while at 6 weeks into abstinence there were severe impairments in working memory. Furthermore, investigation of anxiety-like behavior, 6 weeks into abstinence, demonstrated a lack of emotional behavioral deficit in abstinent animals (114). Taken together, it is evident that the paradigm of ethanol experience and the type of behavioral investigation are critical when determining alterations in PFC-dependent functions during abstinence, and that some PFC-dependent behaviors are less sensitive to the neurobiological alterations in the PFC in abstinent animals compared with others.

Animal models have also been critical in resolving the explicit impact of chronic alcohol on the functionality of the hippocampus. Similar to the studies in animal models of alcohol dependence, which replicated the PFC impairments observed in humans, studies in animals exposed to translationally relevant models of chronic alcohol exposure have reproduced and expanded on the findings from human subjects. These studies have resulted in numerous structural and functional abnormalities of the rodent hippocampus similar to those seen in human studies. For example, studies in rodents employing forced chronic consumption demonstrate long-term exposures to alcohol resulted in extensive impairment in spatial memory (117–122). Unfortunately, behavioral disparities in these preclinical models have been limited to the spatial and contextual processing functions of the hippocampus with no reference to the temporal discrimination role of this structure. Nevertheless, it is clear that chronic alcohol exposure critically impairs hippocampal function in preclinical models similar to those previously discussed in clinical settings, although there remain unanswered questions in this field with regard to the complete profile of hippocampal cognitive impairments. The remainder of the review will focus on the hippocampus and provide a brief overview of the cellular and molecular mechanisms in the hippocampus that could contribute to the long-term impairments in the behaviors dependent on the hippocampus in preclinical models of AUDs.

The regionally differential rates of cognitive recovery following abstinence from alcohol use are potentially consequent to the neurogenic properties (or lack thereof) of each region. To be more specific, cognitive function relying on the frontal cortical region in humans has been described as being recovered at an earlier time in abstinence than cognitive functions specific to the hippocampal formation of the limbic system as previously discussed. It is possible that this disparity is due to, at least in part, the ongoing adult neurogenesis in the hippocampus which occurs at a much lesser rate in the PFC of mammals (146); neurons which would be generated during critical periods of withdrawal would be developing into mature neurons during a time of negative affect (147, 148), potentially resulting in a pathologic phenotype and dysfunctional characteristics (149). This problematic phenomenon would be far more impactful in a region with high neurogenesis (such as the hippocampus) as compared with a region of low or absent neurogenesis, where the typical functioning of the existing circuitry may return upon complete washout of the drug.

Adult mammalian neurogenesis is a widely accepted phenomenon, as evidence demonstrates the existence of mitotically active cells in distinct regions of the brain, one which is the granule cell layer of the DG of the hippocampus. Neurogenesis, or the process of proliferation, differentiation, and maturation of neural progenitor cells to fully functional and integrated neuronal components of the surrounding network (150, 151), has been confirmed in numerous mammalian species, including humans (152). Assessment of cell number and structure at various time points following cell birth can provide insight into the impact of exogenous factors on the neurogenic process in the hippocampus [for comprehensive review of granule cell development see Ref. (153)].

The explicit functionality of these adult-born cells is still a topic of contention. Hippocampal-sensitive learning has been shown to positively influence proliferation and survival of new neurons [reviewed in Ref. (154)]; inversely, increases in proliferation or survival of newly born neurons can increase performance on hippocampal-sensitive tasks, while reductions or ablations of neuronal proliferation results in problematic cognitive performance [reviewed in Ref. (155)]. Acquisition, retention, and extinction of trace fear conditioning (TFC; a hippocampus sensitive task) has been shown to be sensitive to changes in neurogenesis (156) due to or as a result of its hippocampal-dependence (157), but as yet, investigations into the impact of clinically relevant models of chronic alcohol on TFC performance have not been reported.

The goal of this review was to provide initial evidence in support of the proposal that the cognitive recovery of the hippocampus and the PFC following abstinence from long-term alcohol abuse occur at different rates, potentially due to their difference in cellular composition and neurogenic functionality. For example, clinical evidence supports recovery of certain PFC-dependent tasks in times of abstinence from alcohol at different rates compared with hippocampal-dependent tasks. Preclinical findings in animal models of alcohol exposure support the clinical observation; mechanistic studies support that this temporally differential rescue of PFC-dependent tasks is potentially due to the neurogenic deficits in the hippocampus during abstinence, such that the birth of new neurons during periods of negative affect result in the persistence of the hippocampal-specific cognitive disparities.

Many questions remain unanswered with regard to human hippocampal function during periods of alcohol abstinence. For example, it is clear that employing cognitive therapy can support individuals in successful attempts at abstinence. Given that extinction training is being adopted in clinical behavioral therapy to promote recovery from relapse (175), it is critical to investigate similar potential therapeutic strategies (be it behavioral or pharmacological), which will serve this purpose not only to ameliorate the cognitive disparities in these individuals but to facilitate dependent individuals in avoiding relapse to alcohol abuse.

MS was responsible for the article concept and drafted the manuscript. CM and MS provided critical revision of the manuscript for important intellectual content. Both authors critically reviewed content and approved final version for publication.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.