Trajectories of brain development reveal times of risk and factors promoting resilience to alcohol abuse during adolescence

Abstract

Alcohol Use Disorder (AUD) is recognized as harmful for the developing brain. Numerous studies have sought environmental and genetic risk factors that predict the development of AUD, but recently identified resilience factors have emerged as protective. This chapter reviews normal processes of brain development in adolescence and emerging adulthood, delineates disturbed growth neurotrajectories related to heavy drinking, and identifies potential endogenous, experiential, and time-linked brain markers of resilience. For example, concurrent high dorsolateral prefrontal activation serving inhibitory control and low nucleus accumbens activation serving reward functions engender positive adaptation and low alcohol use. Also discussed is the role that moderating factors have in promoting risk for or resilience to AUD. Longitudinal research on the effects of all levels of alcohol drinking on the developing brain remains crucial and should be pursued in the context of resilience, which is a promising direction for identifying protective biomarkers against developing AUDs.

1. Epidemiology of Alcohol Use in Adolescence

Alcohol Use Disorder (AUD) is a major public health concern being an identified cause in almost 6% of deaths and more than 200 fatal diseases (World Health Organization, 2018). Numerous studies have considered the burden of AUD and have noted behavioral consequences and brain disturbances (Bernardin, Maheut-Bosser, & Paille, 2014; Bühler & Mann, 2011; Le Berre, Fama, & Sullivan, 2017; Oscar-Berman & Marinković, 2007; Sullivan & Pfefferbaum, 2019), which in turn increase the risk of perpetuating heavy drinking (Zahr, Pfefferbaum, & Sullivan, 2017). In the last decade, efforts using prospective longitudinal designs have accelerated investigations of the potential vulnerabilities and consequences of alcohol use in adolescents and young adults. Indeed, excessive alcohol use is prevalent in youth, and epidemiological surveys have found that the younger adolescents initiate drinking, the more likely they are to develop AUD (Enoch, 2006; Grant et al., 2015).

Adolescence is a critical period of development when biological, social, cognitive, and affective changes shape adult behavior. It is during this developmental stage that experimenting with risky behaviors, such as alcohol use, is initiated. The National Institute on Alcohol Abuse and Alcoholism (NIAAA, 2018) reports that by age 18, 58% of youth have already drunk alcohol, while about 60% of college students have drunk in the last month, and 20% meet criteria for an AUD diagnosis. Unlike adult drinking patterns, which usually entail multiple, moderate drinking episodes during the week, underage drinking is typically done on weekends and often at binge levels. Binge drinking denotes consumption of high quantities of alcohol in a single occasion, corresponding to a Blood Alcohol Concentration level of 0.08g/dl or greater. Other patterns of alcohol misuse are also reported in youth studies (details are described in Table 1).

Table 1.

Alcohol misuse patterns described in youth and definitions of binge drinking according to countries.

	Quantity	Time frame	Frequency	Country and alcohol definition
HAZARDOUS DRINKING
AUDIT	AUDIT ≥ 8	/	/	International 8 (UK), 10 (Europe), or 14 (USA) grams of pure ethanol in one alcohol drink
AUDIT-C	AUDIT-C ≥ 3 (female) AUDIT-C ≥ 4 (males)
HEAVY DRINKING
Cahalan inventory	3–4 drinks on average	On one occasion	At least twice a month	USA
BINGE DRINKING
NIAAA	4+ (female) 5+ (male) BAC level: .08g/dL	Within less than 2 hours	/	USA
SAMSHA	4+ (female) 5+ (male)	On one occasion	At least one day in the past month	USA
WHO	6+	On one occasion	At least once	International

Notes. NIAAA is the National Institute on Alcohol Abuse and Alcoholism. SAMHSA is the Substance Abuse and Mental Health Services Administration. WHO is the World Health Organization.

2. Overview of Factors Associated with Alcohol Misuse in Adolescence

Several factors have been identified that contribute to persistent alcohol misuse during adolescence, including low socioeconomic status (SES), early age of alcohol use onset, binge drinking, and family history of AUDs (Brown & Tapert, 2004). Binge drinking has been identified as a predictor of a future AUD (Bonomo, Bowes, Coffey, Carlin, & Patton, 2004; Gowin, Sloan, Stangl, Vatsalya, & Ramchandani, 2017; Jones, Steele, & Nagel, 2017). In particular, repeated alternations between drunkenness and abstinence characterizing binge drinking is purported to induce cognitive impairments and brain disturbances (Stephens & Duka, 2008) similar to those observed following repeated withdrawal episodes in adulthood (Becker, 1998; Duka, Townshend, Collier, & Stephens, 2003). Binge drinking could thus impair neuroplasticity and favor the development of a future AUD (Loheswaran et al., 2016). Cross-sectional studies have now established that binge drinking in youth is related to central nervous system (CNS) disturbances (Cservenka & Brumback, 2017), cognitive dysfunction (Carbia, López-Caneda, Corral, & Cadaveira, 2018), and emotional difficulties (Lannoy, Duka, et al., 2021). Such episodes are thus thought to increase the risk of uncontrolled and problematic drinking in youth (e.g., Bell et al., 2013). Nevertheless, longitudinal studies focusing on adolescents before drinking onset are necessary to confirm these results and establish causality.

Genetic predisposition is another mechanism contributing to hazardous drinking, given that binge drinking and drunkenness rates in the transition from adolescence to adulthood have been linked to a GABRA2 polymorphism (Dick et al., 2014). The hypothesis is that excessive drinking in adolescence favors the expression of gene variants associated with AUD vulnerability (Agrawal et al., 2009). Thus, youth with a genetic vulnerability, often assumed from a positive family history (FHP) of AUD, are at a higher risk for binge drinking and alcohol misuse relative to family history negative (FHN) youth (Gowin et al., 2017).

Family history of AUD is established when an individual has biological relatives (most often biological parents) diagnosed with AUD, generally across multiple generations (Alterman, 1988; Schuckit, 1985;1995). Family history of AUD has been associated with heightened risk of alcohol-related problems (e.g., social and academic difficulties, higher rates of hangover; Elliott, Carey, & Bonafide, 2012). This linkage has been supported by a population-based cohort study showing that offspring from high-density families (i.e., multiple relatives with AUD) have a heightened risk for alcohol misuse and AUD. High-density AUD families are also associated with low socio-economic status and high risk of psychiatric disorders (Kendler, Ohlsson, Bacanu, Sundquist, & Sundquist, 2020). Critically, these social and psychiatric risks are attenuated by achieving high levels of school attainment, promoting resilience, which has been measured as the ability to cope with stressful life events (Kendler et al., 2020). Compared with their peers, FHP youth also exhibited impairments in executive and affective processing and disrupted functional brain activations even before the onset of an AUD (Cservenka, 2016).

Youth from AUD parents were found to be less reactive to alcohol effects, as observed with electrophysiological measures (e.g., Schuckit et al., 1988). Electrophysiological studies also indicated a developmental delay in FHP youth, revealed at the P300 level (Hill et al., 1999; Porjesz et al., 1998; Rangaswamy et al., 2007), a component related to cognitive and decisional processes (Polich, 2007). The proposal is that genetic variations modulate event-related oscillations and contribute to AUD vulnerability (Porjesz et al., 2002; Chen et al., 2009). In addition, brain volume differences were observed between FHP and FHN youth during adolescence in the amygdala, orbitofrontal cortex, and cerebellum (Hill et al., 2001; Hill et al., 2007; O’Brien and Hill, 2017). These differential brain volumes have been associated with P300 developmental delay (Hill et al., 2001) and predicted substance use in young adulthood (O’Brien and Hill, 2017). Longitudinal studies further indicated brain changes related to substance use in FHP youth compared to controls, notably, greater activation of the nucleus accumbens during reward anticipation (Cope et al., 2019). Nevertheless, potentially consistent with previous observations (Kendler et al., 2020), some studies targeting FHP adolescents showed that they did not differ from their peers (Bjork, Knutson, & Hommer, 2008; Müller et al., 2015). In these studies, striatal activation (nucleus accumbens) during reward elicitation was similar between matched FHP and FHN adolescents. Clearly then, genetic makeup can heighten an individual’s risk but does not determine the ultimate engagement in excessive drinking or the development of an AUD (Jones et al., 2017). Therefore, protective factors of resilience that account for avoiding alcohol misuse despite all odds need to be identified.

Next, we describe the presumed effects of alcohol use on the developing brain and recognize that some effects might be pre-existing. Then, we approach the role of resilience and focus on brain measures to identify which markers may protect from alcohol misuse development. Finally, we review moderators found to be involved in the pathway toward AUD.

3. Alcohol Use during Adolescence: Its Effect on Brain Structure and Function

The normal developmental trajectory of brain neurobiology is characterized by a decrease in cortical (Giedd, 2008; Gur et al., 2019; Pfefferbaum et al., 2018; Raznahan et al., 2014) and cerebellar (Giedd, 2008; Sullivan et al., 2020) gray matter volume, often interpreted as evidence for neuronal pruning and restructuring in response to environmental interactions occurring postnatally through adolescence (Campbell & Feinberg, 2009; Feinberg, Higgins, Khaw, & Campbell, 2006). An increase in white matter volume throughout adolescence and emerging adulthood likely represents increased structural and functional connectivity, again in response to an individual’s experience and typical developmental processes (Giedd, 2008; Pfefferbaum et al., 2018; Raznahan et al., 2014; Sullivan et al., 2020). Brain development is influenced by myriad factors, notably biological (physiological, genetic) and psychosocial (external and internal milieu), which can alter brain structure, function, and behavior (Walker et al., 2011). These changes affect functioning and can have long-lasting consequences even into later adulthood, thus implicating adolescence as a time of vulnerability for developing psychiatric disorders (Meyer & Lee, 2019), including addiction (e.g., Bell et al., 2013; Silveri & Spear, 2002; Spear, 2010; Stephens & Duka, 2008; Witt, 1994). While adolescence is a vulnerable period for the development of neuropsychiatric pathology, it also constitutes a critical period when protective factors can exert a mitigating effect (Figure 1). For example, positive rearing environment and care-giver interactions promote problem-solving and behavioral regulation, which can concomitantly favor brain and behavioral development (Walker et al., 2011). Nevertheless, even apparently ideal rearing conditions are not a guarantee to thwart risk-taking or thrill-seeking behavior.

Figure 1.

LEFT: Differing trajectories of brain and behavioural development as a function of exposure to risk and protective factors. The cumulative effect is illustrated by the progressive strengthening (darker lines) of the trajectories over time. Figure and caption reprinted with permission from Walker et al. (2011).

RIGHT, TOP: Comparison of different developmental courses of selective frontal and subcortical structures relevant to addiction.

RIGHT, BOTTOM: Distributions of median ages of the incidence of onest of psychiatric disorders occurring from pre-adolescence to emerging adulthood. Figures reprinted with permission from Meyer et al. (2019).

The development of brain structures and circuitry during adolescence are characterized by heterochronicity in maturation and a possible source of vulnerability. For instance, the limbic brain system matures quickly during early adolescence (Meruelo, Castro, Cota, & Tapert, 2017; Morgane, Galler, & Mokler, 2005; Steinberg, 2008), while the prefrontal cortex matures gradually over the course of adolescence (Bava & Tapert, 2010; Brown & Tapert, 2004; Giedd, 2008; Pfefferbaum et al., 2018; Raznahan et al., 2014; Steinberg, 2008). As a hypothetical example, asynchronous brain systems development may result in enhanced reward sensitivity linked to early striatal development and immature executive skills linked to later prefrontal development, rendering adolescents susceptible to risk-taking behavior (Jadhav & Boutrel, 2019), such as engagement in alcohol drinking. In addition, excessive alcohol use is a particular concern in adolescence because the young brain is highly susceptible to alcohol-related consequences. Ample evidence is available from animal studies, showing that adolescents are more sensitive to various alcohol effects (Nixon & McClain, 2010; Obernier, White, Swartzwelder, & Crews, 2002; Spear, 2014). Notably, intermittent ethanol exposure (which can model binge drinking in human) impacts adult behavior with long-term effects on alcohol drinking, emotional symptoms, memory, and executive functions (Crews, 2008; Crews et al., 2019; Hauser et al., 2019; Varlinskaya, Hosová, Towner, Werner, & Spear, 2020). Prospective human studies support the deleterious effects of adolescent heavy drinking on brain volume and integrity (Jernigan et al., 2016; Pfefferbaum et al., 2018; Sullivan et al., 2020; Zhao et al., 2019). Adolescence is thus defined as a developmental window of vulnerability for the onset of excessive and potentially problematic alcohol use (e.g., Bell et al., 2013; Spear, 2018; Varlinskaya & Spear, 2015). Studies showing an association between alcohol use and neuroradiologically-determined neuropathology in adolescent human samples are described next.

3.1. Brain structural disturbance

Neurobiological measures from young drinkers provide objective and prospective evidence that alcohol disturbs neuromaturation. Both alcohol use initiation and binge drinking led to cortical thinning in frontal, anterior cingulate, and posterior cingulate cortices in adolescents and young adults (Luciana, Collins, Muetzel, & Lim, 2013; Mashhoon et al., 2014). Regionally thin cortices were also related to greater drinking quantity and frequency and to recent alcohol use (Mashhoon et al., 2014). Research focusing on substance-naive adolescents at baseline corroborates that thinner dorsolateral prefrontal cortex and inferior frontal cortex predicted binge drinking and externalizing symptoms in late adolescence, especially in youth who were FHP (Brumback et al., 2016).

In addition, smaller gray matter volumes in frontal, striatal, and limbic regions were found in adolescents (Heikkinen et al., 2017) considered hazardous drinkers (Table 1). This result observed in young adults, was supported, with evidence for accelerated decline of gray matter volumes in the hippocampus and para-hippocampus over 2 years of binge drinking; functionally, poorer memory and more memory blackouts were associated with greater volumetric decline in the hippocampus (Meda et al., 2018). Longitudinally, heavy drinking (Table 1) was found to disrupt normal brain development. In particular, adolescents who became heavy drinkers showed accelerated volume loss in frontal (Pfefferbaum et al., 2018) and cerebellar (Sullivan et al., 2020) gray matter; decreased frontal volumes were also related to the maximum number of drinks consumed on one occasion in the past year (Pfefferbaum et al., 2018).

Regarding white matter, a longitudinal study provided evidence for attenuated white matter volume growth in drinking adolescents and young adults (Squeglia et al., 2015). Results from diffusion tensor imaging, which assesses the microstructure of fiber tracks, showed poorer white matter integrity in hazardous and binge drinkers compared with controls (Pohl et al., 2016). Lower fractional anisotropy (an index of fiber organization), greater radial diffusivity (an index of edematous tissue or myelin disturbance), and greater axial diffusivity (an index of abnormally thin axonal caliber) occurred in several brain regions of heavy-drinking adolescents and young adults relative to no-to-low drinking youth (Pohl et al., 2016; Shen et al., 2019; Smith et al., 2017; Zhao et al., 2020). In young adults, low fractional anisotropy was related to binge-drinking intensity and poorer working memory (Smith et al., 2017). Of particular relevance is the longitudinal analysis of Zhao et al. (2020) conducted prospectively on 451 adolescents over a 5-year interval that presented seminal findings supporting excessive drinking in young adolescents as likely causative of accelerated decline in and deviation from the no-to-low drinking group in the trajectories of white matter microstructural development in selective frontal system.

3.2. Brain functional impairments

Numerous studies using functional MRI brain measures indicate modified brain activity among young drinkers relative to their no-to-low drinking counterparts or even to themselves before initiating drinking. Young adult binge drinkers exhibited greater frontoparietal activation during working memory than no-to-low drinkers with no detectable hippocampal activity (Schweinsburg, McQueeny, Nagel, Eyler, & Tapert, 2010). When attempting to inhibit negative emotions elicited by viewing unpleasant images, binge drinkers showed lower frontal and striatal activations than non-bingers, and greater binge drinking frequency was predictive of its extent (Cohen-Gilbert et al., 2017).

Longitudinal studies indicated that youth who reported a history of heavy or binge drinking (Table 1) within the previous three years exhibited reduced activation in frontal and striatal networks when correctly inhibiting an automatic response at baseline compared to youth who remained no-to-low drinkers (Norman et al., 2011; Wetherill, Squeglia, Yang, & Tapert, 2013). A three-year follow-up revealed that binge drinkers had greater frontoparietal and cerebellar brain activation, whereas controls who remained no-to-low drinkers showed a decrease in these brain regions (Wetherill and colleagues, 2013), suggesting that alcohol misuse disrupts the developmental trajectory of executive control. Furthermore, longitudinal studies found that before engaging in binge drinking, adolescents who later went on to binge drink had abnormally low frontoparietal activation during working memory (Squeglia et al., 2012) and decision-making (Jones, Cservenka, & Nagel, 2016) processing, which are two different components of frontoparietal functioning. After initiation of binge drinking, attenuation of dorsal striatal activation was observed during decision-making and was negatively correlated with the total number of binge-drinking episodes in the past 90 days (Jones et al., 2016). Increased activation in the reward circuitry (nucleus accumbens, precuneus) and in the insula during decision-making, under risk, also predicted the onset of binge drinking in adolescents (Morales, Jones, Ehlers, Lavine, & Nagel, 2018; Xiao et al., 2013), suggesting that increased striatal and limbic activation in a risk/reward context is related to higher risk-taking behavior during adolescence.

Neuroimaging studies inform us about brain responsivity to affective stimuli and alcohol-related cues. Young adults who binge drink exhibit different activation patterns from non-bingers when processing emotional stimuli (Lannoy, Dricot, et al., 2021; Maurage, Bestelmeyer, Rouger, Charest, & Belin, 2013; Rae et al., 2020). For example, high frontal activation in the face of low temporal activation noted in young bingers was hypothesized to reflect functional compensation of possible alcohol-related disruption (Maurage et al., 2013). A focus on implicit brain responses to various emotional contents showed that binge drinkers, relative to controls, had low insular activation when processing happiness, low frontal and cingulate activation when viewing sadness, high cingulate activation when viewing fear, and high cerebellar activation when viewing anger (Lannoy, Dricot, et al., 2021). In response to pain-related pictures, binge drinkers also had higher activation in the fusiform and inferior temporal gyrus than non-binge drinkers (Rae et al., 2020). Distinct CNS responses are elicited during emotional and social processing, and these are assumed to be involved in alcohol misuse and its perpetuation; however, longitudinal studies are needed to delineate and confirm these results. In an alcohol-related context, longitudinal studies revealed that moderately drinking adolescents (Nguyen-Louie, Courtney, et al., 2018) and young adults (Courtney, Infante, Bordyug, Simmons, & Tapert, 2020; Dager et al., 2014), who later transitioned to heavy or binge drinking, showed elevated frontolimbic and frontostriatal activity in response to alcohol cues at baseline testing.

Finally, functional connectivity analysis has revealed the relevance of specific neurocircuitry that contributes to, or is altered by, binge drinking. A longitudinal study indicated increased functional connectivity in the Default Mode Network (DMN) in binge drinkers over two years, whereas controls showed decreased functional DMN connectivity over the same duration (Correas et al., 2016). A six-year study further showed that an earlier age-of-onset of weekly drinking was associated with higher context-dependent functional connectivity between the bilateral posterior cingulate and cortical (right frontal, parietal, and left lingual areas) and subcortical (right cingulate precuneus) areas, this being related to poor attentional abilities (Nguyen-Louie, Simmons, et al., 2018). In young adults, stronger correlations between reduced frontal activation during inhibition and increased striatal activation during monetary reward distinguished binge drinkers from controls (Weafer, Crane, Gorka, Phan, & de Wit, 2019).

Short conclusion:

This subsection delineated cross-sectional and longitudinal evidence that structural and functional brain regional and neurocircuit disturbances appear to be attributable to the initiation of alcohol misuse in adolescents and young adults. First, studies highlighted predisposing brain vulnerabilities by indicating that pre-existing differences in brain functional responses in frontal and striatal networks can be observed in future drinkers (Figure 2), supporting the role of higher reward-seeking and lower executive control in the engagement of excessive drinking. Some studies also reported functional changes in the cerebellum among drinkers, but additional longitudinal research is needed to determine its role in alcohol use onset and maintenance. Second, evidence supports functional consequences related to alcohol misuse by showing that youth who initiated hazardous or binge drinking exhibited anomalous brain changes in frontolimbic and frontostriatal areas and functional aberrations in DMN connectivity. Alcohol-related consequences on neurostructural integrity were evident from disturbed developmental trajectories of gray matter and white matter volume and cortical thickness (Figure 3). Overall, these observations are consistent with addiction models (Sullivan & Pfefferbaum, 2019; Wise & Koob, 2014) emphasizing the conjoint role of executive control and reward-processing, and their associated neural circuits, as a vulnerability factor predisposing youth to engage in alcohol abuse and as a consequence of having engaged in drinking alcohol to excess.

Figure 2.

(A) Brain activation during receipt of monetary reward (Doors task: Win>Loss) in left VS. Significant peak activation was observed at the [−6 18 −2] MNI coordinate after correcting for family wise error (pFWE < 0.05) within the reward mask. (B) Brain activation during successful response inhibition (StopInh>Go) within the frontal-insular-subcortical (FIS) mask. Significant (pFWE < 0.05) peak activation was observed in a large right prefrontal cluster encompassing precentral, middle, and inferior gyri and anterior insula. The top scatter plot presents the negative relationship between extracted BOLD signal from a 10 mm radius spherical region centered at the [−6 18 −2] peak in the left VS (A) during receipt of monetary reward and extracted BOLD signal from a 10 mm radius spherical region centered at the [42 4 32] peak in the right IFG during response inhibition (B). The bottom scatter plot presents the negative relationship between extracted BOLD signal from the left VS (A) during receipt of monetary reward and extracted BOLD signal from a 10 mm radius spherical region centered at the [28 2 50] peak in the right MFG during response inhibition (C) The negative relationships indicate that less right frontal activation during response inhibition was associated with greater left VS activation during reward. Figure and caption taken with permission from Weafer et al. (2019).

Figure 3.

Top left: The plots connect values at baseline and 1- and 2-year follow-ups for 356 adolescents who continued to meet no/low criteria (gray) and heavy drinkers (green) as a function of baseline age; the linear mixed-effects model (lmer) fits with one and two standard deviations, computed separately for no/low and heavy drinkers, are also plotted.

Top right: The figure shows lateral and medial views of the left hemisphere. Regions showing faster declines in cortical volumes in the heavy drinkers relative to the no/low drinkers, displayed in bright orange, are false discovery rate corrected (p<,0.025).

Top figures and captions taken and modified with permission from Pfefferbaum et al. (2018).

Bottom left: Trajectories (i.e., regression lines) of individual heavy drinkers (green). The lmer fits with 61 and 2 SD separately computed for no/low drinkers (gray) and heavy drinkers (green) are also plotted.

Bottom right: Jitterplots of slopes (expressed as % change per year) by tissue type for the 328 no/low drinkers (gray) and all 220 youth who initiated heavy drinking (black). Red asterisks mark differences from the no/low drinking group in slopes meeting correction for multiple comparisons. Black asterisks note differences with p ≤ .05. The cerebellar images to the right of the jitterplots display in gray the lobules showing group differences (p ≤ .05) in slopes.

Bottom figures and captions taken and modified with permission from Sullivan et al. (2020).

4. Resilience and Brain Mechanisms

Parsing resilience into its component processes is central for recognizing factors that can be independently affected positively, negatively, or indifferently during adaptations to stressful life events or unfavorable environments (Alim et al., 2012). The many studies using subjective measures have shown that self-reported resilience mitigated risks for developing an AUD (Kennedy et al., 2019; Long et al., 2017). Beyond this psychological evidence, neuroscience studies indicated that compensation and reorganization of brain functions may offset vulnerability and thus constitute resilience markers (Cicchetti & Blender, 2006; Cousijn, Luijten, & Feldstein Ewing, 2018). This proposal of brain resilience received support in findings from a large European consortium (Burt et al., 2016), reporting that larger gray matter volumes in the right middle and superior frontal gyri were shown in adolescents (12–17 years old) who had experienced adversity (stressful live events assessed through the following domains: family, accident/illness, sexuality, autonomy, deviance, relocation, and distress) and exhibited positive adaptation (good rule-abiding conduct, social skills, relationships, academics, and absence of internalizing problems). Whether such evidence for resilience to trauma generalizes to resistance to drinking remains to be established. Some indications suggest possible resilience brain markers in youth who, despite a genetic vulnerability (family history of AUD), did not develop alcohol misuse.

4.1. Resilience Brain Markers

Neuroimaging research supports the existence of specific brain correlates associated with resilience in youth with a family history of AUD. Comparing resilient (non-drinkers) to vulnerable (hazardous drinkers, Table 1) emerging adults (18–25 year-olds) of AUD parents, cross-sectional results showed that the vulnerable group reported emotional distress, risky substance use, and greater activity in the middle frontal gyrus when rating negative emotional pictures compared with their resilient counterparts and showed higher activation of the posterior cingulate cortex during a working memory challenge (Brown-Rice et al., 2018). In another cross-sectional study, blunted nucleus accumbens responsivity was observed during decision-making in FHP participants who did not present alcohol use problems, suggesting that this attenuated activity is protective (Yau et al., 2012).

A longitudinal study assessed self-reported psychological resilience in early adolescence (12–15 years old) and related it to measures taken in late adolescence (Weiland et al., 2012). Greater resilience was predictive of lower substance use, fewer alcohol-related problems, and better executive performance. Associations also emerged between resilience and subthalamic nucleus and pallidum activation detected with fMRI during a working memory task. Moreover, compared with a low resilience group, youth with high resilience had stronger functional connectivity between the subthalamic nucleus and median cingulate cortex (Weiland et al., 2012). In the same cohort, psychological resilience in adolescents with a family history of AUD and carriers of the G allele (GABRA2) was associated with lower activation in the inferior parietal cortex during emotional processing and less severe externalizing behaviors (Trucco, Cope, Burmeister, Zucker, & Heitzeg, 2018); however, which resilience factor exerted the greatest influence remains unknown.

Several studies also directly compared resilient and vulnerable youth with prospective evaluations, for which follow-up was conducted after the onset of hazardous drinking or substance use in vulnerable youth. The relevance of functional connectivity between frontal and striatal brain networks (dorsolateral prefrontal cortex and posterior cingulate) was found, such that resilient youth (11–16 years old) had greater functional connectivity between the left dorsolateral prefrontal cortex and the left posterior cingulate cortex than vulnerable youth before the onset of excessive drinking (Martz et al., 2019). These two networks were also explored separately by evaluating brain responses during executive control and reward processing (adolescents from 12 to 14 years old at baseline). Greater activation in the right dorsolateral prefrontal cortex during correct inhibition was a significant predictor of resilience, whereas neural activation in the ventral striatum (nucleus accumbens) was not a significant predictor of resilience (Martz, Zucker, Schulenberg, & Heitzeg, 2018). Functional brain responses were also evaluated during passive viewing of emotional words (positive, negative, or neutral). Resilient adolescents (11–17 years old) had greater activation in bilateral orbital frontal gyrus and left insula in response to emotional stimuli, whereas vulnerable ones showed higher activation of the dorsomedial prefrontal cortex and lower activation of the ventral striatum and the amygdala bilaterally. The vulnerable group also had more externalizing symptoms, such as aggressive and delinquent behaviors, than the resilient group (Heitzeg, Nigg, Yau, Zubieta, & Zucker, 2008).

Short conclusion:

Specific patterns of brain responsivity may serve as potential neurophysiological resilience markers and are observed principally in prefrontal and striatal systems (Figure 4). To wit, greater activation of the dorsolateral prefrontal cortex occurs during executive control in resilient versus vulnerable youth. The dorsolateral prefrontal cortex is considered a primary hub of the executive network, involved in goal-directed behaviors, and has a role in excessive drinking regulation (Abernathy, Chandler, & Woodward, 2010). It has also been related to impaired adaptation in adult AUD patients (Beylergil et al., 2017), supporting its role as a neurophysiological resilience marker. Functional connectivity between the dorsolateral prefrontal and posterior cingulate cortices and the overall conjoint role of frontal and limbic brain regions has also been supported as related to resilience. In striatal regions, lower activation of the nucleus accumbens during reward processing is associated with resilience. Indeed, the nucleus accumbens is considered a fundamental driver of reinforcement components of alcohol-related behaviors (Casey & Jones, 2010; Koob et al., 2014), and its lower activation may act as a resilience marker by reducing incentive salience, alcohol cue-reactivity, and further consumption.

Figure 4.

Extracted task-related functional connectivity change with NAcc during Incentive>Neutral anticipatory processing by family history group. L, left; R, right; NAcc, nucleus accumbens; SSMA, supplementary sensoriomotor area; PoG, postcentral gyrus; SMA, supplementary motor area; OCC, occipital. Figure and caption taken with permission from Weiland et al. (2013).

5. Moderators of the pathway to AUD

The study of both risk and protective resilience factors should be pursued together to identify patterns of brain structural, functional, and behavioral factors that distinguish vulnerable from resilient youth as it pertains to drinking. As noted by Brown and Tapert (2004), other factors may be involved in gateways towards alcohol misuse and AUD such as earlier age of alcohol use onset and lower SES. Given the complexity and heterogeneity of factors forming the constellation of different outcomes, data-driven machine-learning analysis of large and complex datasets is providing new leads beyond detection with traditional univariate analysis toward identifying specific and robust risk and resilience factors. Findings support the role of early drinking onset, stressful life events, and neural activation related to language and social processing (Gowin, Manza, Ramchandani, & Volkow, 2020; Ruan et al., 2019; Squeglia et al., 2017) and will require replication with prospective, hypothesis-directed studies.

Most of the studies described in this review compared groups of similar age, SES, or childhood history, which can statistically parse the contributions of each variable to the overall variance of a target outcome but preclude understanding the role of specific factors in isolation. What was determined, however, was that early drinking onset (~14 years old) predicted dependence symptoms and the development of AUD (Conde, Peltzer, Gimenez, & Cremonte, 2020; Dawson, Goldstein, Patricia Chou, June Ruan, & Grant, 2008), such that differences in brain functioning were salient between youth who started to drink early versus their peers who refrained from drinking (Baker et al., 2019; Zhao et al., 2020). Numerous studies have also identified stressful life events such as childhood trauma history or poor SES as predictors of alcohol misuse. Longitudinal studies showed that childhood trauma predicted peri-adolescent binge drinking (De Bellis et al., 2020) and alcohol-related problems (Hagborg, Thorvaldsson, & Fahlke, 2020). Disturbed brain functional connectivity (Silveira et al., 2020) and neural oscillations (Meyers et al., 2019) have also been related to trauma history and have predicted AUD in adolescents. Similarly, longitudinal studies showed that low SES and financial difficulties influenced drinking in adolescents (Nooner, De Bellis, Clark, Thompson, & Brumback, 2020; Poonawalla, Kendzor, Owen, & Caughy, 2014) and interacted with genetic factors to predict alcohol-related problems (Barr, Silberg, Dick, & Maes, 2018; Davis & Slutske, 2018). Taken together, these negative factors increased the likelihood of developing AUD. Conversely, some support has been given for the role of resilience in explaining low substance use in youth with trauma history (Wingo, Ressler, & Bradley, 2014). Critically, specific neuroanatomical and neurophysiological markers of reliably-detected resilience factors have yet to be identified. Further, under what conditions do aversive factors such as trauma become sources of resilience need delineation.

6. Trajectories of risk and resilience: Integration of findings

Research to date may guide our understanding of pathways to AUD and alternative routes to avert AUD, as summarized in Figure 5.

Figure 5.

This model depicts the role of different factors involved in the pathway toward AUD: (1) Risk factors for drinking in pink: poor control abilities (frontal network), high reward sensitivity (striatal network), as hypothesized as predisposing factors; emotional processes (limbic network) and executive and emotional processes (cerebellar network) to be investigated; and gray and white matter compromise as consequence of binge/hazardous drinking, increasing the risk of AUD. (2) Family and environmental factors as influencing binge/hazardous drinking. (3) Resilience-related brain activations, potentially moderating the risk associated with family and environmental factors and the pathway toward or away from AUD.

Pathways to AUD can be represented by the complex interaction of multiple factors, each contributing a specific weight to the overall risk. A comprehensive evaluation of family and environmental factors (e.g., family history of AUD, stressful life events) may help to determine alcohol behaviors in adolescence but also explain later risk of developing AUD. Alcohol use itself can modify specific brain physiology that engender AUD and perpetuate alcohol use. This review noted predisposing differences in frontal and striatal functional responses between future drinkers and non-drinkers. As highlighted in addiction models (Wise & Koob, 2014), these brain regions may be related to behavioral outcomes favoring excessive drinking such as high incentive salience (alcohol desire) and low executive function (inability to control alcohol desire). Differences in limbic functional activation have also been explored in drinkers, although further controlled longitudinal studies are needed. Current findings from prospective research have identified specific drinking consequences, showing that the onset of alcohol misuse affects gray and white matter brain volumes and integrity, especially at frontal sites and frontally-based circuitry, which in turn, may reinforce the risk of uncontrolled drinking.

A model that emerges from the constellation of findings reviewed on risk and resilience brain markers entails a see-saw pattern of functional brain activation that is dependent on and modified with environmental, genetic, and developmental influences. Risk and resilience factors may be ubiquitously present but depending on the strength vs. weakness of relevant factors and developmental timing drinking or non-drinking will be manifest. Thus, regarding resilience, protective factors would enable positive adaption and low alcohol use despite unavoidable high vulnerabilities such as family history and genetic factors. Recognition of the continuing development of frontally-based functional and structural systems throughout adolescence provides temporal markers of heightened vulnerability. Indeed, resilience—as described by high dorsolateral prefrontal and low striatal functional activations—may reduce the chances of engaging in excessive and binge drinking during adolescence and have the consequence of averting the onset of later AUD.

Future Directions

Additional research is needed to delineate the individual and interactive effects of risk and resilience factors affecting the initiation of youthful drinking and AUD. To improve our knowledge, we have identified three points that deserve serious research focus. A hope is that identification of behavioral and brain biological markers of risk and resilience will enable recognition of at-risk youth and offer avenues for diverting attraction to drinking.

The role of the cerebellum as a site of disruption and compensation.

Functions of the cerebellum are traditionally linked to the long-term adverse consequences of AUD on gait and balance (Fitzpatrick, Jackson, & Crowe, 2008; Sullivan, Rose, & Pfefferbaum, 2006). It is now clear, however, that the cerebellum is also involved in executive, language, and affective processes, modeled as the “Cognitive and Affective Syndrome” (Schmahmann, 2019; Schmahmann & Sherman, 1998). Selective relations between cognitive and emotional difficulties and regional cerebellar structure or function in adults recovering from AUD (Pitel, Chanraud, Müller-Oehring, Pfefferbaum, & Sullivan, 2013; Sawyer et al., 2016; Sullivan, Zahr, Saranathan, Pohl, & Pfefferbaum, 2019) provide a foundation for pursuing such relations in heavily drinking youth. Indeed, exploration of such vulnerabilities and drinking-related changes in the cerebellum of young drinkers remain a critical yet under-appreciated research focus. The few such extant studies reported differences in functional cerebellar responses during executive (Wetherill et al., 2013) and emotional (Lannoy, Dricot, et al., 2021) processing that distinguished young drinkers from low or no drinkers. A recent structural MRI study from the NCANDA multi-site consortium found accelerated cerebellar volume decline in youth who initiated heavy drinking compared with youth who remained no-to-low drinkers over 4-years of longitudinal study (Sullivan et al., 2020). These findings offer ample justification for expansion of the cerebellar research avenue in youth, especially given functional relevance of frontocerebellar circuitry in adult AUD with respect to its disruption (Sawyer et al., 2016; Sullivan & Pfefferbaum, 2005; Zahr et al. 2017) and as a coordinated system of compensatory functional repair with sobriety (Chanraud, Pitel, Müller-Oehring, Pfefferbaum, & Sullivan, 2013; Chanraud & Sullivan, 2014).

The role of drinking characteristics in AUD risk.

Animal models strongly support the role of intermittent alcohol exposure in inducing and enhancing voluntary drinking in adolescent mice and rats (e.g., Bell et al., 2013; Crews et al., 2019). Whether intermittent drinking style accelerates or is a causative antecedent to human AUD, however, remains unknown. Consistent with the animal models is the proposal that adolescents who present binge-drinking features (e.g., hangover, blackout, unconsciousness) early in life are particularly vulnerable to exhibit cognitive impairments and disturbed brain functional responses (Min et al., 2019; Nguyen-Louie et al., 2017; Nguyen-Louie, Simmons, et al., 2018). Age at first alcohol-related unconsciousness episode as well as repetition and duration of these episodes are also identified as salient predictors of cognitive and brain alterations (Min et al., 2019); however, prospective evaluations are needed to support this proposal. Similarly, cross-sectional results suggest that altered cortical gyrification was specifically related to hangover and binge-related symptoms (Hua et al., 2020). Longitudinal evaluation of overall alcohol use in adult drinkers has offered promising avenues by showing the associations between unconsciousness episodes and risk for dementia (Kivimäki et al., 2020). Future studies should pursue this question longitudinally by quantitating youthful drinking characteristics. Although human studies do not have the same control level as animal model, the implementation of prospective evaluation offers a well-controlled human natural experiment.

The role of resilience brain markers in adverting AUD.

We identified several neurophysiological resilience markers that enhance resistance to initiating heavy drinking, but much work remains to be done to delineate the contexts that makes them viable in contrast to contexts where the same markers may be detrimental to resilience or inoperative as a resilience resource. A further question involves identifying biological and psychological mechanisms of resilience that are protective against unfavorable environments, experiences of trauma, or genetic risk. Additional considerations entail discovering factors that moderate resilience abilities, for example, sleep habits (Hairston et al., 2016). Prospective longitudinal studies will provide the most compelling answers to these questions and bring further insights into the risk and resilience factors involved in the dynamic pathways toward AUD and those resistant to AUD.

Abbreviations

AUD

Alcohol Use Disorder

CNS

Central Nervous System

DMN

Default Mode Network

FHP

Family History (of alcohol use disorder) Positive

FHN

Family History (of alcohol use disorder) Negative

NIAAA

National Institute on Alcohol Abuse and Alcoholism

SES

Socioeconomic status