Alcohol, adolescence and the addicted brain
By: Tjitske de Graaf Student number: 0513725 Supervisor: Ozlem Korucuoglu
University of Amsterdam
1. INTRODUCTION
“Youth are heated by Nature as drunken men by wine” - Aristotle
1.1 Background
Adolescence is a period of powerful transition from childhood to adulthood and is characterized by dramatically physical, psychological and social changes. In this transitional stage in their life young people learn to understand abstract ideas, gain cognitive flexibility (Rutter & Rutter, 1993), establish and maintain satisfying relationships (Arnett, 2000) and acquire a mature sense of themselves and their purpose. Adolescence is also accompanied by a variety of physical modifications such as the appearance of the secondary sexual characteristics, as a result of hormonal release. Other biological alterations include physical growth due to an increase in muscle mass, height and weight (Coleman, 1990)The sudden and rapid physical changes that adolescents experience often lead to enhanced self-consciousness and self-awareness. Although the beginning of adolescence is characterized by distinct and dramatic physiological changes, the end of adolescence has less clear biological boundaries. Meanwhile established relationships are altered as a result of increased freedoms and independence gained by the aging process. In the course of adolescence, it is normal for young adults to separate from their parents and establish their own identity (Arnett, 2000) Simultaneously, teenagers draw closer to their peers. This time is life is often accompanied with the increase of risk behaviour. Previous studies suggest that teenagers tend to be more sensitive to rewards than either children or adults and less sensitive to risks (Steinberg, 2004). As a result adolescents have a increased tendency to act upon behaviours that adults and children would classify as risky.
Traffic injuries, drug and alcohol use, sexual risk taking and crime are all statistical indicators that adolescents engage in more risky like behaviour than children or adults (Steinberg 2008; Blum & Nelson-Mmari 2004; Williams et al. 2002). Adolescents take risks to test and define themselves and more often this is related to the presence of peers (Millstein & Halpern-Felsher, 2003). Risky behaviours are not only harmful but also beneficial when teenagers learn new skills and prepare themselves for future challenges (Steinberg, 2004). By taking risks the adolescent is able to discover oneself and his or her place relating to others and the larger world; it may allow the adolescent to explore adult behaviour and privileges (Silbereisen & Reitzle, 1992) and to accomplish normal developmental tasks (Muuss & Porton, 1998). A lot of research has focused on the question why teenagers are more inclined to engage in risky behaviours compared to children or adults (Steinberg, 2008). Assumptions
that young adults are worse at perceiving risks or are less risk averse were found to be incorrect. Instead research suggests that 15-year olds have similar logical reasoning and basic information-processing abilities as adults (Reyna & Farley, 2006). Rather the finding that the brain keeps developing well into early adulthood has been implemented as underlying the behavioural changes observed in adolescents (Steinberg, 2005, 2008).
Many of the behaviours that are associated with health risks, like smoking and drinking are initiated during adolescence and are continued into adulthood. Therefore risk taking behaviour during this time in life affects health and wellbeing later in life (Jackson, Henderson, Frank, & Haw, 2012). Among European adolescents, alcohol is the drug of choice and binge drinking occurs frequently (Hibell et al., 2007). Over 60 per cent of the Dutch teenagers start drinking between the ages of 12 and 16 years (Hagemann, 2010), although Dutch law prohibits selling alcohol to adolescents under the age of 16 years. Just a little under half, 45 per cent, of all young adults drink five or more glasses of alcohol on respectively a Friday or Saturday night (Verdurmen et al., 2011). One of the key indicators in determining whether someone will develop a dangerous relationship with alcohol is the age onset of alcohol use (DeWit, Adlaf, Offord, & Ogborne, 2000; Grant & Dawson, 1998). This has proven to be relatively young for the Dutch adolescents. Repeated use can ultimately lead to dependent behaviour and result in alcohol use disorder (AUD). Usually, the motivation to regulate addictive behaviours is low in teenagers, because adolescents often do not recognize their alcohol and drug use as problematic (Wiers et al., 2007).
Of course not all binge drinking kids will grow up to be addicts. Almost all adolescents in the Western world drink, while only a few of them will become addicted. The larger sum will consume large amounts of alcohol through their teenage years and student lives, but will eventually return to moderate drinking. For some, however, the alcohol consumption escalates and it becomes a chronic condition (Rehm et al., 2003). Other risk factors that contribute to the development of AUDs besides the first drinking age are repeated alcohol exposure (DeWit et al., 2000; Grant & Dawson, 1998; Hawkins et al., 1997; Labouvie, Bates, & Pandina, 1995) and an existing family history of alcohol abuse (Dawson, 2000; Goodwin, 1979; M. A. Schuckit, 2009). Although these indicators statistically suggest who is more likely to have alcohol abuse problems, these do not predict which individuals will eventually become addicts. A major challenge in understanding AUDs lies in uncovering why some individuals become addicted when exposed to (repeated use of) alcohol, whereas others do not. Therefore it is important to investigate the difference in these individuals.
1.2 Aims of the study
The difference between social users and alcohol abusers is likely the result of the interplay between numerous factors such as social, psychological, environmental, and behavioural and (neuro-) biological. Because of the suggestion that brain plasticity plays an important role in developing of typical adolescent behaviour, I would like to investigate whether underlying neural differences also accounts for an individual likelihood of developing an addiction. Because not only brain plasticity may lead to an increase in risk seeking behaviours such as consuming alcohol, drinking itself may influence the developmental trajectory of the brain. Therefore I would like to investigate the current status of research on the functional distinction which might underlie the differences between moderate drinking adolescents and high risk drinkers. The paper unfolds as follows, I will first current theories regarding brain maturation and adolescence and the behavioural changes that are the likely result of the changing brain. Second, I describe the current literature regarding the effects of alcohol on brain structure and function. Third, I will look at genetic factors that influence the risks of becoming an addict. Fourth, I describe the cognitive models that are thought to underlie behavioural characteristics of addicts. Finally, based upon this review I suggest possibilities for future research directions.
2. RE-WIRING THE BRAIN
2.1 Brain maturation during adolescence
For many years it was thought that brain maturation was set at a fairly early age and completed by the time the child reached puberty. In fact brain size changes only little beyond the age of six (Pfefferbaum et al., 1994). However it now appears that brain development is a complex process which continues well into adolescence and early adulthood (Baird et al., 1999; McQueeny et al., 2009). It seems that the brain is rewiring itself during this transitional time in life. Specifically modifications in white matter (nerve fibres) and grey matter (neuron cell bodies, dendrites and glial cells) occur as the brain attenuates unused connections and strengthens other connections (Giedd et al., 1999). The neuroanatomical alterations of the brain have been made visible by improved brain imaging techniques, such as magnetic resonance imaging (MRI), Diffusion Tension Imaging (DTI) and functional MRI (fMRI). For example, these brain imaging techniques make it possible to study both structural and functional characteristics of the living brain (Steinberg, 2005). In accordance with findings from previous animal studies it was found that the brain undergoes a series of anatomical changes through adolescence (Crews, Braun, Hoplight, Switzer, & Knapp, 2000; Giedd et al., 1999). These include growth of the prefrontal cortex (PFC), of the limbic system structures and of white matter association fibres (McQueeny et al., 2009). In addition, the brain developments are linked to advancements in cognitive functioning and emotional processing (Bava & Tapert, 2010)], thereby suggesting that improved cognitive performance is reflected within the structural brain alterations.
As previously been mentioned white matter and grey matter densities of the brain undergo a series of alterations when growing up. Grey matter volume seems to follow an inverted U-shape with a peak around the ages of 12 till 14 (Giedd et al., 1999; Gogtay et al., 2004). At first during pre-puberty the amount of grey matter increases with the rapid growth of neuronal cells. Until recently this process of proliferation was only thought to happen in the first three years of childhood, for a review, see (Paus, Keshavan, & Giedd, 2008). Second, the grey matter volume declines again when the brain rewires itself through a process of synaptic pruning. This process eliminates unused connections, the so called use it or lose it principle (Giedd et al., 1999). The first areas of the brain to be pruned are those involving primary functions, such as motor and sensory areas. While areas involved in more complex processes
such as the lateral and PFC are pruned at a later stage in adolescence (Crews et al., 2000). The nonlinear pattern and time course of grey matter maturation is likely to coincide with measures of cognitive functioning (McQueeny et al., 2009; Sowell, Trauner, Gamst, & Jernigan, 2002).
The decline in grey matter is also facilitated by the increase in white matter, resulting in a net loss for grey matter (Paus et al., 2008). In contrast to grey matter, white matter increases linearly in the course of adolescence in certain brain regions, in particular in the frontal and parietal cortices (Giedd et al., 1999). Also modifications in the sub cortical regions occur, particularly in the basal ganglia. These changes are said to involve greater cognitive control by fine tuning and strengthening the connections between these sub cortical structures and the PFC (Eluvathingal, Hasan, Kramer, & Fletcher, 2007). The increase in white matter is the result of myelation, a process in which nerve fibres become sheathed in myelin. Myelin insulates and hence increases the speed of neural transmission and improving the efficacy of information processing. In addition, increases in white matter volume are accompanied by progressive changes in white matter integrity, such as the magnetization-transfer ratio (MTR) in MRI, and mean diffusivity (MD) and fractional anisotropy (FA) in DTI (Barnea-Goraly et al., 2005; Paus, 2010). A number of recent cross-sectional studies discovered an overlap between the increase in white matter and cognitive performance , for a review see (Bava & Tapert, 2010), although this is not surprising while the increase in white matter is also associated with maturation, and maturation in its turn with enhanced cognitive abilities.
In addition to the structural remodelling described above, neurotransmitter systems also undergo substantial changes through adolescence. Since most of the neural connections that are pruned contain neurotransmitter receptors, alterations to the neurotransmitter systems are in synchrony with brain development and maturation of cognition. For example, changes in the inhibitory GABAergic system during adolescence coincide with executive functioning and working memory performance (McQueeny et al., 2009). Pruning does not occur equivalently across all types of synapses. Most pruned synapses are excitatory, involving glutamate receptors. In contrast to the decrease in glutaminergic input, the dopaminergic input is elevated (Andersen, Rutstein, Benzo, Hostetter, & Teicher, 1997; Spear, 2009). Rodent studies showed that dopamine (DA) concentrations at the time of adolescence are highest in the PFC. In contrast, concentrations in posterior regions including the partial lobe peak earlier in childhood. This imbalance of DA concentrations has been proposed to underlie characteristic adolescent behaviour (Andersen, Thompson, Rutstein, Hostetter, & Teicher, 2000; Casey & Jones, 2010; Steinberg, 2005; Wahlstrom, Collins, White, & Luciana, 2010).
Because this neurotransmitter is thought to play a critical role in the brain’s reward system and therefore remodelling of the dopaminergic pathways in this transitional time is likely to have important aspects in both socio-emotional and cognitive control networks (Spear, 2000, 2009).
In sum, brain maturation results in a number of distinct changes and is probably involved in specific behavioural and cognitive aspects of teenagers. These findings suggest that cognitive changes could be reflecting within structural brain changes. Nonetheless, knowing more about the structure of the brain does not necessarily tell us more about the function of the brain. Ascribing imaging measures to underlying cellular and molecular events is challenging. A future challenge is to determine the cellular changes that underlie these macrostructural observations by the use of multimodal imaging techniques like fMRI, MRI-based morphometry and studies with non-human animals in which imaging and histological studies can be performed parallel (Casey, Giedd, & Thomas, 2000).
BOX 1 THE REWARD SYSTEM
The reward system was first discovered by Olds and Milner, who identified that certain regions of the brain acted as a reward in teaching rodents to run through mazes or solve other problems (Koob & Weiss, 1992; Pierce & Kumaresan, 2006). Stimulations of those particular brain regions seemed to provide pleasure. Dopamine (DA) was found to be one of the main neurotransmitters involved in this reaction and was subsequently dubbed ‘the pleasure chemical’. The reward circuitry consist of the mesolimbic and mesocortical pathways (see figure 1), with the first being more prominent. The mesolimbic pathway starts at the ventral tegmental area (VTA) and goes via the medial forebrain bundle towards the nucleus accumbens (NAc) of the limbic system (Jones & Bonci, 2005; Robbins, 2002). Similar to the mesolimbic pathway, the mesocortical pathway is initiated at the VTA but it terminates in the PFC. The VTA is the primary release site for DA but also sends information to the tegmental pendulopontine nucleus (TPP), a brain region that is involved in non-DA-mediated reward signalling (Koob & Nestler, 1997; Pierce & Kumaresan, 2006; Robbins, 2002). The reward system is often been implicated in addiction and addiction behaviours. All addictive drugs have reinforcing properties by enhancing ventral striatum response (Compas, Orosan, & Grant, 1993; Ernst, Pine, & Hardin, 2006). A wildly held view why people become addicted to drugs is that they increase the activity of a common dopaminergic reward system in the brain: this view is referred to as a "positive reinforcement" view of addiction to drugs (Casey & Jones, 2010; Zuckerman & Neeb, 1979). Addictive drugs (such as cocaine, heroin and alcohol) create a shortcut to the brain's reward mechanism by flooding the NAc with DA (Berridge
& Robinson, 1998). It has been said that these drugs 'hijack' the brain. But drugs do more than provide the DA jolt that induces ecstasy and mediates the initial reward and reinforcement. Repeated exposure to drugs leads to gradual adaptations in the reward system that give rise to addiction. Because DA is also important in learning and memory and interacts with another neurotransmitter, glutamate, to take over the brain’s system of reward-related learning (Berridge & Robinson, 1998; Bjork, Smith, Chen, & Hommer, 2010; Schultz, Dayan, & Montague, 1997). However, recent studies have shown that glutamate may play as large, or even possibly a larger role than DA does. Recent studies have shown the brain structures rich in glutamate are involved in learning processes which play a part in developing drug cravings, one of the central aspects of compulsive drug seeking behaviour (for a review see (Tzschentke & Schmidt, 2003). Glutamate is both directly and indirectly involved in the modification of the activity of the dopaminergic system, they interact in complex way in the NAc. It has also been suggested that glutamatergic projections to the Nac induce drug seeking behaviour (Kalivas & Volkow, 2005). A crucial challenge is to determine whether the brains of people who are addicted to drugs respond to a reward in the same way as controlled brains do.
2.2 Functional remodelling of the adolescent brain
The developmental changes in the adolescent brain coincide with numerous behavioural alterations. Advanced cognitive skills such as executive functions, planning, problem solving, and inhibitory control are all developing at this time (Anderson et al., 2011). In addition to alterations in neuroarchitecture there are also functional changes visible in the brain of teenagers. The functional changes are predominantly observed in the cognitive control network and the socio-emotional network. The cognitive control network is associated with the dorsal lateral prefrontal cortex and parietal cortices (Steinberg, 2004). Whereas the ventral striatum (VS), amygdala, orbitofrontal cortex (OFC), medial prefrontal cortex (mPFC), and superior temporal sulcus are regions implicated in the socio-emotional network (Steinberg, 2004).
Cognitive control, also referred to as executive functioning, emerges in early childhood and is enhanced throughout the process of aging. This mechanism ensures that people focus their attention on relevant information and shield it against distraction from irrelevant information (Durston & Casey, 2006). Also higher-level cognitive functions such as
attention and working memory are thought to rely critically on control processes. The dlPFC and parietal cortex are found to be active during basic cognitive control functions such as working memory, task switching and inhibition tasks. These areas show enhanced activity during these tasks (Casey et al., 2004; Casey, Thomas, Davidson, Kunz, & Franzen, 2002; Schweinsburg, Nagel, & Tapert, 2005). Imaging studies examining performance on tasks requiring cognitive control (e.g., Stroop, flanker tasks, Go/No Go, anti-saccade) have shown that adolescents tend to recruit the cognitive control network less efficiently than adults (Durston & Casey, 2006; Hare et al., 2008). Furthermore regions which correlate with task performance such as cognitive control areas become more activated with age. These findings suggest that there is a role for improved performance of executive functioning and maturation. The socio-emotional network processes social and emotional information and is very active during puberty. This network is implicated in the characteristics that adolescence are easily aroused and experience more intense emotions and are more sensitive to social influence. The socio-emotional network is also implicated in reward and most functional imaging studies have focused their attention on activity in the ventral striatum during reward delivery (Geier, Terwilliger, Teslovich, Velanova, & Luna, 2010) and during reward anticipation (Van Leijenhorst et al., 2010). One of the first studies to examine reward related processes across the different transition stages from childhood to adulthood was completed by Galvan and colleagues, they showed that VS activations was sensitive to varying magnitudes of monetary reward (Galvan, 2010). More recent studies show that adolescent had heightened VS activation in anticipation of a reward when compared to adults and children (Casey, Jones, & Hare, 2008; Drevets & Raichle, 1998; Galvan, 2010; Geier et al., 2010). Similar findings were observed during reward delivery (Van Leijenhorst et al., 2010), suggesting that risk taking tendencies of adolescents' may be related to unique neural characteristics that increase their anticipation to reward and sensitivity to rewarding outcomes. For example, in a study from Geier et al it was found that BOLD signal showed attenuation in the VS when adolescents were required to assess an incentive cue, but showed elevations during reward anticipation, thus suggesting that adolescents have more difficulty in assessing potential reward outcomes (Geier et al., 2010). The same study indicates that during anticipation of reward this brain regions showed exaggerated reactivity compared to adults. Also increased activity in the socio-emotional network brain regions was found to predict the selection of risky choices compared to more conservative choices (Ernst et al., 2006; Spear, 2009). These findings suggest that hyper-activation of the striatum leads to reward seeking behaviour. DA release during rewarding events is thought to underlie ventral activation (Aarts et al., 2010;
Koepp et al., 1998) and may therefore lead to the incentive to seek additional rewards and therefore in reward-seeking behaviour (Bava & Tapert, 2010). In contrast, increase in activation of the dlPFC has been linked to account for the improved performance of cognitive task during development (Casey & Jones, 2010; Durston & Casey, 2006) and similar activation during reward-risk conflict appears less active in adolescents compared to that in adults (Bjork et al., 2010; Gilman, Ramchandani, Davis, Bjork, & Hommer, 2008). Suggesting the involvement of the socio-emotional network during anticipation of rewards, reward receipt and high risk choice making.
Neuroimaging studies that investigated the underlying neural correlates of low and high choices found high risk choices activate reward related VS and mPFC, while low risk choices activate control related dlPFC. Interestingly, activation of the mPFC was positively associated with risk-taking propensity, whereas activation of the dlPFC was negatively associated with risk-taking propensity (Van Leijenhorst et al., 2010). Consequently it can be said that the cognitive control network is linked to low risk tendencies and the socio-emotional network to high risk behaviour.
2.3 The neurobiological model of adolescent risk taking
As a result of combining data from the imaging studies with the structural data from both human and rodent models, researchers have proposed a neurobiological model of adolescent development (Spear, 2000). This model proposes that the maturational gap between cognitive and affective processes might result in the tendency of teenagers to involve in risk behaviours (see figure 2). It is postulated that there is an asynchronous development of reward and control systems that enhance adolescents’ response to incentives and risky behaviours (Casey et al., 2008). The hypothesis states that bottom up limbic systems which are involved in emotional and incentive processing are earlier developed than top down prefrontal systems which are involved in behavioural control. Particularly, in situations with high emotional salience, the more mature limbic regions will override prefrontal regions, resulting in poor decisions. The relatively late development of the cognitive control network may influence risk behaviour in adolescents (Steinberg, 2008). For example, an adolescent may choose a potential reward although this might be the best rational choice (Gladwin, Figner, Crone, & Wiers, 2011).
FIGURE 2 NEUROB IOL OGICAL MOD EL OF AD OLESCENT DEV ELOPMENT. THIS F IGURE SHOWS THE D IF F ERENT DEV ELOPMENTAL TRAJECT ORIES F OR SIGNALLINGOFSUB CORTICAL ANDPR EF RONTAL TOP-D OWN CONT ROLREGIONS DURING TRA NSITIO NINTO AD ULTHOOD. THELIMB IC PRO JECTION S DEV ELOPSO ONE RTHANTHEPR EF RONTALCONT ROLREGIO NS. DURINGAD ULTHOODTHEF RONT ALLIMB IC CIRCUITRYISF ULLY MATURE. REPRINTED F ROM CURRENT OPINI ON IN NEUROB IOLOGY, VOL. 20, SOMERV ILLE LH, AND CASEY BJ. DEV ELOPMENTAL NEUR OB IOLOGY OF COGNITIV E CONTR OLANDMOTIV ATIONALSY STEMS, 179- 159. COP YRIGHT 2010 (CASEY & JONES, 2010)
The model suggests that decision making is often the result of competition between the cognitive control network and the socio-emotional network (Drevets & Raichle, 1998; Steinberg, 2008). This competition is not limited to adolescents as it has been found that also adult decision making cannot always be explained by the rational principles of economic theory. In this model risky choices are thought to be the result when the socio-emotional network dominates the cognitive control network. Other dual-process-models also assume that adolescents take risks because their decisions are based upon reacting rather than deciding (for a review see (Reyna & Rivers, 2009)). The time gap between developments of the two networks during transition to adolescence might therefore be one of the reason that young adults tend to make more emotional and less rational decisions Evidence that support this model shows that adolescent risk-taking follows an inverted U-shaped pattern for reward related regions with a peak in adolescence, and a linear pattern for regions associated with cognitive control (Casey & Jones, 2010). Although the structural alteration in the brain do not automatically lead to functional changes it seems that combining evidence from both types of studies does propose a hypothesis that explains why adolescents behave the way they do (Casey et al., 2008; Guerri & Pascual, 2010; Spear, 2000).
However, a possible correlation between structural and functional (BOLD signal) changes is still under investigation. Only a few studies have examined the link between grey and white matter alterations and functional observation (Blakemore, 2012). These studies have focused their attention at executive control, relation reasoning, stimulus independent thought, reading and, working memory (Dumontheil, Hassan, Gilbert, & Blakemore, 2010; Lu et al., 2009; Olesen, Nagy, Westerberg, & Klingberg, 2003). Overall they show that the neuroanatomical remodelling is only partly involved in the developmental pattern of the BOLD signal. Although some regions were found to be positive correlated for both structural
and functional changes, other regions were not. Future research should focus on disentangling possible causal contributions between these two.
I should point out that recently critics state that this dual-system model is both simplistic and speculative and does not account for individual and contextual differences (Crone & Dahl, 2012). For example, this model does not include the low LR and high LR distinction, does not differentiates between type 1 and type 2 alcoholics and forgoes gender effects (see below). A recent meta-analysis study that researched areas involved in decision making in adolescent disputes the claims that this model makes (Crone & Dahl, 2012). Studies showed both an task-dependent increase or decrease in lPFC and mPFC during maturation (Crone & Dahl, 2012). They argue that this great diversity of findings cannot be explained by a frontal cortical immaturity concept. For example, performance monitoring studies did not confirm to the dual process model (Cohen et al., 2010). To the contrary it was found that the frontal cortical network was engaged to the same extent in participants of different age groups but under different experimental conditions. Both PFC and parietal cortex were activated in early adolescents and adults; however this activation followed positive performance feedback in the youngsters whereas this same activity in adults was the result of negative feedback (Van Den Bos, Güroğlu, Van Den Bulk, Rombouts, & Crone, 2009). Instead the researchers propose a new theory of flexibility of the frontal cortical network. This new theory suggest that cognitive control capacities are not necessarily immature during adolescence, but, rather, there is a great deal of flexibility. The degree to which adolescents engage in successful self-regulation is influenced by the motivational salience of the social context. This new model of flexibility that the authors propose contains two key aspects; the first focuses on social–affective engagement and goal flexibility, and the second focuses on the role of pubertal hormones in social–affective engagement. Evidence for this theory is derived from studies which show that PFC activity in adolescence is sensitive to context and that activity can be enhanced by training (Jolles, Van Buchem, Rombouts, & Crone, 2012).
3.1 The effects of alcohol on the brain
Recent research suggests that alcohol use affects adolescents and adults differently, which is intuitive given what we now know about the changes in the teenage brain. The significant structural and functional remodelling of the adolescent brain and at the same time exposure to heavy drinking could interfere with the developmental trajectory of cognitive and behavioural functioning or even permanently 'fix' addictive behaviour in the brain (Crews et al., 2005; Silveri & Spear, 1998). Heavy drinking is associated with a wide range of neural consequences in adults (Pfefferbaum et al., 1994). For example, chronic alcohol exposure is associated with cortical and white matter volume loss besides decreases in choline and N-acetylaspartate, reduced GABA receptor efficacy, and impaired neurogenesis (Crews et al., 2005). This leaves the question that what effects alcohol has on adolescent brain development. Most information of the effects of acute and chronic ethanol exposure comes from animal experimentation studies because of the ethical constraints in performing such studies in human adolescents. Multiple animal studies have shown that alcohol has different effect upon the brain of adolescents’ rodents than of adults (Acheson, Stein, & Swartzwelder, 1998; Hiller-Sturmhöfel & Swartzwelder, 2004; Nagel, Schweinsburg, Phan, & Tapert, 2005). Similar to humans, animals undergo a period which can be marked as adolescence and in which a reorganization of the brain takes place (Hiller-Sturmhöfel & Swartzwelder, 2004; Spear, 2000). Structural imaging further exhibit that the teenage brain is extremely vulnerable to the deleterious effects of drinking; particularly the cortex and hippocampus are affected (De Bellis et al., 2000; Medina et al., 2008; Nagel et al., 2005; Tapert, Schweinsburg, et al., 2004; White & Swartzwelder, 2004). Binge drinking can result in brain structure abnormalities, deficiencies in memory and poor academic performance (López-Frías et al., 2001; Medina et al., 2008; Squeglia, Jacobus, & Tapert, 2009; Squeglia, Schweinsburg, Pulido, & Tapert, 2011; Tapert, Schweinsburg, et al., 2004). De Bellis and colleagues reveal that hippocampal and prefrontal white matter volume appears smaller in binge drinking teenagers ( De Bellis et al., 2000). In accordance other studies show that there are alterations in anisotropy in the corpus callosum in alcohol using teenagers (De Bellis et al., 2008). Altered FA has also been found in frontal, temporal and parietal regions of binge drinking adolescence (McQueeny et al., 2009). These studies suggest that drinking leads to atypical white matter developmental trajectories. Although these studies indicate that heavy drinking during this period in life is detrimental and leads to impaired cognitive performance and brain health, it is left to determine whether these are the cause of substance use or whether these features predate substance use.
Most studies which examine the effect of substance abuse on neurocognitive development have been performed with young adults that already have a problematic relationship with drinking. These cross-sectional studies show that among teenagers who already suffered from AUDs, various cognitive impairments are observed when compared to healthy controls. For example, they exhibited poorer retrieval, weakened attention and troubled information processing (Tapert & Brown, 1999). Also deteriorated visuospatial functioning could be indicated in young adults who were being treated for abuse of this substance (Tapert, Schweinsburg, et al., 2004). Post drinking symptoms such as hangovers and withdrawal do not only suggest intense intake of the intoxicant but has also been linked to brain alterations. Teenagers who showed more signs of withdrawal symptoms were also shown to have poorer memory and abnormalities in brain response (Tapert, Pulido, Paulus, Schuckit, & Burke, 2004). A longitudinal study that examined the influence of binge drinking and incurring hangover on neurocognitive functioning reported that deficits in functioning are related with heavy drinking (Squeglia et al., 2009). Moreover they argued that direct and indirect changes of neuromaturation due to substance use would likely extent into adulthood. For instance, a decrease in volume of the hippoampus and comprised quality of white matter.
Although there have been some studies that investigate the effect of alcohol intake during the developmental stage of adolescence, none of the studies with this subject group have examined what structural and functional differences might underlie the difference in becoming an addict. The immense neural changes that occur in the adolescent brain maturation stages and the vulnerability of the developing brain to the damaging effects of ethanol, suggest that alcohol exposure at this time would most likely have a significant impact on the adult brain functions (Hiller-Sturmhöfel & Swartzwelder, 2004)
3.2 Acute alcohol and the functional brain
.The negative effects of acute alcohol administration on cognitive performance are well documented. Functional imaging has shown that adult alcoholics have difficulties with controlling themselves as the result of impaired executive functioning. Previous research examining the neural correlates of acute ethanol effects suggests that moderate-to-high alcohol doses acutely disrupt brain activity (Volkow et al., 1988). For example, intoxication leads to a shift in activity from the cortical area's to the limbic regions. Others suggest that intoxication lead to increased PFC activation but only at low doses (Sano et al., 1993).
Conversely, high doses of alcohol lead to a decrease in PFC and parietal cortex activity (Levin et al., 1998; Söderlund, Grady, Easdon, & Tulving, 2007; Van Horn, Yanos, Schmitt, & Grafton, 2006). Although this could also be the result of compensatory processes under low intoxication. The negative effect of drinking on higher level functioning (Van Horn et al., 2006) is visible in regions necessary for goal directed motor action, frontal and parietal circuitry as well as a general suppression of cerebellar activity. In alcoholic conditions a decrease in activation was visible. Further studies observed impaired episodic memory in drunks, although stimuli-dependent. This suggest that impaired performance might be associated with a decreases in PFC activity and an increase in activity of the parahippocampal gyrus (Söderlund et al., 2007). Suggesting that alcohol impairs memory by interfering with encoding.
Multiple functional imaging studies have examined various aspects of cognitive functioning and inebriated subjects. Marinkovic et al. studied the effects of alcohol on cognitive control by examining the ability to inhibit prepotent responses in favour of rewarded ones (Marinkovic, Rickenbacher, Azma, & Artsy, 2012). It has been suggested that impulsivity is the result of impaired ability to inhibit prepotent responses (Fillmore, Vogel-Sprott, & Gavrilescu, 1999). BOLD measurements were derived from social drinkers in a placebo/acute alcohol study. With a modified four-color Stroop task, they showed increased reaction time and more errors on incongruent for the intoxicant condition. Intoxication resulted in attenuated anterior cingulated cortex (ACC) activation in high conflict trials and erroneous responses. The ACC has been linked to cognitive control and fMRI studies showed that the ACC is important in error monitoring and behavioural adjustments (C. S. Carter et al., 1998). Thus, alcohol may interfere with the inhibitory control of the ACC (Ridderinkhof et al., 2002), by rendering a person less able to focus attention. Or by suppressing certain responses or initiate purposeful behaviour (Marinkovic et al., 2012). The inability to maintain inhibitory control over drinking is both accounted as a dispositional risk factor and as a consequence of excessive drinking (Field, Wiers, Christiansen, Fillmore, & Verster, 2010). Other researchers have used an fMRI "Go/No-go" paradigm to examine the influence of ethanol dosage on brain activity and behavioural change (Anderson et al., 2011). A high dose of this substance leads both to an increase in reaction time but also increases "No-go" false alarms. A moderate dose did not result in significant changes in reaction time compared to sober controls, however errors did increase. Also a dose dependent decrease in cortical activation was found in the dorsal anterior cingulate, bilateral anterior insula, supplementary motor area, dlPFC, left sensorimotor cortex, and some temporal lobe regions.
Gundersen and colleagues researched the effect of alcohol on neuronal activation during increasing levels of cognitive load in a working memory n-back task (Gundersen, Grüner, Specht, & Hugdahl, 2008). They observed decreased activation in the anterior cingulate cortex (ACC) and cerebellum during a high load trials, under a moderate intoxicant dose. In a subsequent study the authors also show that task context plays an important role in determining the effect of ethanol on brain activation during WM processes (Gundersen, Specht, Grüner, Ersland, & Hugdahl, 2008), because there is an increase in prefrontal and ACC activation when alcohol is expected but not received, while the activity in the same area's decreases when the substance is received (Gundersen, Grüner, et al., 2008; Gundersen, Specht, et al., 2008). Thus intoxication and expectancy have complementary effects on neuronal activation.
The subjective response between different types of drinkers has also been investigated. Prior studies show that individual differences in subjective responses to alcohol may affect the individual’s vulnerability to developing AUDs and is influenced by genetics ( Schuckit, Smith, Trim, Fukukura, & Allen, 2009). Other subjective response experiments report a difference in response between high drinkers and social drinkers where measured in the reward system while presenting emotional cues (Gilman et al., 2008). The high risk drinkers demonstrated both a reduced subjective response and a blunted response in the brain’s reward system. Activation of the NAc was visible in both groups and may contribute to subjective experience of pleasure and reward during intoxication. However only in the social drinking group did ethanol elicit a significant activation of this brain region.
Low level response studies have also been used to identify differences between healthy groups with different risk for developing AUDs (Schuckit et al., 2012). People who have a low level response (low LR) to alcohol need more drinks in order to experience alcohol related effects compared to people with a high level response (high LR). This difference in response is genetically influenced and accounts for an enhanced risk for AUD (Chung & Martin, 2009; Schuckit et al., 2007). Contrary to high LR subject, low LR subject might relate to different brain processing during placebo or intoxicant conditions while performing a stop signal task. Low LR individuals typically show more activity to erroneous or difficult trials under placebo conditions than matched high level response subjects, but less activity after a moderate dose of alcohol. Similar to prior studies high and low LR groups differed on measures of brain activity, but not on task performance (Padula et al., 2011; Tapert, Pulido, et al., 2004; Tolentino et al., 2011; Trim et al., 2010). These results may
suggest a brain mechanism that contributes to how low LR individuals might enhance the risk of future heavy drinking and substance dependence (Schuckit et al., 2012).
In a similar study with low LR and high LR to alcohol, findings were that subjects with a low LR had a load-dependent higher functional activation in dlPFC, PPC and visual cortex (Padula et al., 2011). Alcohol did not affect performance (errors or response latency), but attenuated the working memory load–dependent activation in the dorsolateral prefrontal cortex. The pattern of neural effects was consistent with the cognitive demands of the working memory task. Subject exhibited longer response latencies and more errors. In the placebo condition subjects with a low LR showed enhanced activation in dlPFC and PPC compared to the high level response subjects. This suggests that alcohol either greatly diminishes or reversed the low LR group difference seen after placebo.
Acute effects on the emotional processing might also be another important factor in becoming addicted. A recent alcohol challenge study also reported reduced amygdale activity to fearful and angry faces (Sripada, Angstadt, McNamara, King, & Phan, 2011). Acute administration of ethanol attenuates amygdale reactivity to social signals of threat in social drinkers and reduces the activation of regions associated with threat or fear. The intake of alcohol and processing of fearful faces stimuli resulted in the activation of the amygdale and limbic regions but solely in the social drinkers. While severe drinkers did not exhibit any difference in response to fearful or neutral faces nor was there a difference while intoxicated (Gilman, Ramchandani, Crouss, & Hommer, 2012; Gilman et al., 2008).
In sum, the studies described above have shown that the intake of an intoxicant substance has numerous effects on various types of cognitive functions. However it remains unclear which brain regions may be affected early in the clinical course of becoming an addict and which areas of impairment might be associated with identified neuropsychological differences (Clark, Thatcher, & Tapert, 2008). Researchers have been unable to identify what predicts the course of alcohol use for all or even most young people. Instead, findings provide strong evidence for wide developmental variation in drinking patterns within this special age group (Schulenberg et al., 1996). Complex behaviours, such as the decision to begin drinking or to continue using alcohol, are the result of a dynamic interplay between individual, genes and environment. Vulnerability to develop a drug addiction is influenced by a combination of genetic and environmental factors. There are large individual differences in adolescent social and behavioural development. Understanding the mechanisms that lead to these differences may shed light on why some adolescents become heavy drinkers, while others do not. Genetic factors are likely to play an important role in individual differences (Green et al., 2008). It is
well documented that genetic polymorphisms (inter-individual variation) affects cognitive ability, and is specifically associated with structural differences in specific cortical structures (Toga & Thompson, 2005).
3.3 Arterial spin labelling, an alternative imaging technique
Functional MRI (fMRI) using the blood oxygenation-level dependent (BOLD) response is by far the most commonly employed method for mapping functional activation. The BOLD response is derived from the magnetic susceptibility differences between oxygenated and deoxygenated haemoglobin (Ogawa et al., 1992) and only indirectly measures neuronal activity. The BOLD response reflects a complex interaction of blood flow, blood volume, and oxygen extraction fraction (Buxton, Uludağ, Dubowitz, & Liu, 2004). Arterial spin labelling (ASL) is an emerging non-invasive technique which operates on a similar basis as 15O PET scanning (Detre & Alsop, 1999). While PET studies need injection of tracer, instead ASL uses water in blood as an endogenous perfusion contrast agent (Liu & Brown, 2007). Water protons in arterial blood are magnetically labelled and fMRI images are acquired both with and without tagged blood, cerebral blood flow (CBF) is measured by subtracting the two images (figure 3). In contrast to BOLD, ASL measures directly and quantitatively localized perfusion changes. There are 3 main techniques available, continuous arterial spin labelling (CASL), pulsed arterial spin labelling (PASL) and a combination of both pseudo-continuous arterial spin labelling (psCASL) (Liu, Wong, Frank, & Buxton, 2002; Sutton, Ouyang, Karampinos, & Miller, 2009).
Figure 3: Arterial spin labelling. The left panel shows the proximal tagging pulse being applied. The middle panel is included to depict the transit of the bolus of tagged blood from the tagging plane to the imaging plane. The right panel shows the arrival of the tag at the imaging slice after the transit time. Reprinted from (Sutton et al., 2009)
Similar to BOLD, ASL also indirectly reflects to neuronal activity, although the quantitative measure of CBF may be more closely tied to neuronal function (Fernández-Seara et al., 2005; Liu & Brown, 2007). One study showed a more linear relation between CBF changes and
neural activity then with BOLD (Miller et al., 2001). Other studies shown a decreased inter-subject and inter-session variability in comparison with BOLD, suggesting a more direct link between CBF and neuronal activity (Aguirre, Detre, Zarahn, & Alsop, 2002; Tjandra et al., 2005; Wang et al., 2003). ASL is most likely to be used in cognitive neuroscience to study slowly evolving changes in psychological state, such as mood changes, and to help with the interpretation of BOLD findings in pharmacological studies.
Many researchers still prefer BOLD over ASL because of some disadvantages. The temporal resolution of perfusion fMRI is inherently poorer than BOLD fMRI because of the necessity to acquire two sets of images (tag and control) and the need to wait for blood to flow into the imaging region (Liu & Brown, 2007). Typically the image acquired after magnetic tagging is subtracted from a control image to remove the effects of static magnetization and to control for imaging effects of the tagging experiment unrelated to the rate of blood flow. The signal to noise ratio is also low, due to the inflowing blood magnetization this will typically be only about 1 per cent of the tissue. In ASL in contrast to BOLD there is often no whole-brain analysis, due to the need to acquire data before the tagged blood signal has fully relaxed which results in a maximum of slices than can be acquired.
However with pharmacological studies ASL has an advantage over BOLD. Due to the dependence on hemodynamic regulation, the BOLD signal is sensitive to anything that can alter hemodynamics or the neurovascular coupling, including pharmacological agents, disease, and so on (Borogovac & Asllani, 2012). Alcohol (ethanol) is a vasodilator, and has been shown to decrease brain BOLD response (Levin et al., 1998). Therefore the BOLD response can partly be caused by other effects than neural activation, which in turn would affect task-induced BOLD changes (Aguirre et al., 2002). Another important aspect which one needs to be aware of is the effect of alcohol on the hemodynamic response function (HRF). Luchtmann showed that alcohol prolonged the time course of the HRF in some regions (visual cortex, primary motor cortex) but also affected underlying patterns of neuronal activation in areas such as the supplementary motor area (Luchtmann, Jachau, Tempelmann, & Bernarding, 2010). This could affect group comparisons in fMRI studies. By combining both methods, ALS and BOLD, one can derive the vascular from neurally based changes underlying BOLD. This ‘calibrated’ BOLD method uses mild hypercapnia to increases CBF but not the cerebral metabolic rate of oxygen (which is changed in BOLD) to derive both signals (Davis, Kwong, Weisskoff, & Rosen, 1998).
A limited understanding of physiology underlying BOLD (Iannetti & Wise, 2007; Liu & Brown, 2007) and because of the vasoactive properties of alcohol it may not be possible to interpret the BOLD studies unambiguously, due to the hemodynamic changes underlying the BOLD signal effects in acute intoxication studies. Multiple studies using positron emission tomography (PET) or single photon emission tomography (SPECT) reported specific alcohol dose dependent CBF changes (Mathew & Wilson, 1986; Sano et al., 1993). Because ASL is still an upcoming method there are just a few studies that combine this technique with the effects of acute alcohol (Durazzo, Gazdzinski, Mon, & Meyerhoff, 2010; Khalili-Mahani et al., 2011; Rickenbacher, Greve, Azma, Pfeuffer, & Marinkovic, 2011; Tolentino et al., 2011). ALS has been used to examine brain blood perfusion of alcoholic dependent individuals in a non-intoxicated state (Clark et al., 2007). Most studies compare ASL with the 15O-PET technique (Khalili-Mahani et al., 2011) with comparable findings, although a gender effect could be observed (Rickenbacher et al., 2011). An ASL study with into treatment seeking alcoholics revealed that individuals that resumed drinking had lower frontal CBF both at baseline and after 34 days of abstinence in comparison with those that remained abstinent over the year (Durazzo et al., 2010).
Tolentino and colleagues was the first to examine the effect of alcohol on CBF in subgroups with different future risk for AUD (Tolentino et al., 2011). Individuals with a low level of response (low LR) to alcohol have a lower sensitivity to this drug and are more prone to future alcoholic problems. After placebo both groups, high and low LR to alcohol subject had similar CBF levels. However after alcohol administration (similar blood alcohol content) the low LR group had significantly less increase in CBF particularly in frontal regions. These findings further support the data previously observed with PET studies using receptor-specific, CBF, or metabolic ligands (Volkow et al., 1988) and earlier reported low level response studies (Chung & Martin, 2009; Schuckit et al., 2007).
A separate resting state study that investigated the gender effect of acute intoxication on CBF showed perfusion increased bilaterally in frontal regions in men but not in woman (Rickenbacher et al., 2011). While woman exhibited stronger cortical perfusion during placebo conditions in the left hemisphere frontal, parietal and temporal areas. The gender differences observed are likely the result of hormonal, metabolic and hemodynamic regulation systems. Furthermore this study suggested that alcohol induced perfusion correlated positively with dopaminergic reward system. Cortical perfusion data have also shown that individuals that relapse to alcohol use have lower frontal grey matter perfusion when compared to controls and abstainers (Durazzo et al., 2010). While controls and participants that remained
abstinent for at least 12 months were equivalent on frontal grey matter perfusion. Suggesting a premorbid and/or acquired neurobiological abnormalities resulting in an increased likelihood of relapse in these individuals. These individuals’ response to alcohol indicates that CBF needs to be separated from the BOLD response in order to interpret BOLD signal changes. Although ASL is theoretically a better technique to measure the effects of acute alcohol, because of the vasodilating properties of alcohol and that BOLD measures several underlying processes, it has been scarcely used. The few that did (Durazzo et al., 2010; Khalili-Mahani et al., 2011; Rickenbacher et al., 2011; Tolentino et al., 2011) all focused on the effect of alcohol on the resting state network.
4. RISKY GENES
4.1 Genetic polymorphisms and addictive behaviours
In the past the relationship between brain function and genes has focused on single gene polymorphism. However now it is known that very little diseases could be caused by a single mutation. The complexities of of post-genomics, with its focus on gene/gene, gene/epigeneticand gene/environment interactions challenge this simplicity (Davies, 2012). Gene-environment interactions can account for much more of the aetiology of psychiatric disorders. For example, schizophrenia is only 50% concordant in genetically identical (monozygotic) twins. Furthermore environmental events are thought to result in long=term developmental changes in brain protein chromatin (Nagy & Turecki, 2012). Epigenetic approaches investigate how a set of reversible heritable changes in the functioning of a gene can occur without any alterations to the DNA sequence. These mechanisms include DNA methylation, chromatin conformational changes through histone modifications, ncRNAs and, most recently, 5-hydroxymethylcytosine. Emerging neuroimaging techniques could contribute to establishing the link between genetic susceptibility and cognition and behaviour as well as the link between genetic susceptibility and neuropsychiatric disorders (Mattay & Goldberg, 2004).
Changes in the gene expression in the brain reward regions are thought to contribute to the pathogenesis and persistence of drug addiction (Guerri & Pascual, 2010). Recently there is an increase in genetic phenotyping studies that also examine the genetic underpinnings of the risk for alcoholism. Familial and twin epidemiological studies have shown that genes contribute to the vulnerability of becoming an addict, with estimates of heritability of 30–60%
(Goldman, Oroszi, & Ducci, 2005). This heritability was first demonstrated in alcoholics and dependence of this subject was linked to regions on chromosomes 1, 2, 3, 4, 7 and 8 (review (Edenberg et al., 2006)). While low LR alcoholism is shown to be linked to chromosomes 1, 2, 9, 21 (Schuckit & Smith, 2001). After the chromosomes are revealed, through linkage studies, candidate genes can be identified. In contrast to the linkage approaches, association studies are designed to analyse whether a single gene and its polymorphic structure effects the examined disease. Multiple genes have been appointed as candidate genes, genes that influence the risk of AUDs and related phenotypes. For example, the GABA receptor subunit A2 (GABRA2) and the muscarinic acetylcholine receptor M2 (CHRM2) play both a crucial role in alcohol dependence (Edenberg et al., 2004; J. C. Wang et al., 2004). Subsequently for a detailed analysis of these genes, single nucleotide polymorphisms (SNPs) were genotyped in positional candidate genes located within the linked chromosomal regions, and analysed for association with these phenotypes (Edenberg et al., 2006). Alterations in the coding regions of genes may alter the final protein product, this is the case in the A118G variant of the mu-opioid receptor gene (OPRM1) (Kreek, Nielsen, Butelman, & LaForge, 2005; Kreek, 1996).
Because alcoholism is a complex disease with the interplay of various personality traits, differences in neurobiochemistry, dependent on (social) environment and influenced by genetic traits such as heritability, numerous candidate genes have been postulated. These include dopaminergic, GABAergic, glutamatergic, opioid, cholinergic and serotonergic neurotransmitter systems as well as the neuropeptide Y (for a review see (Köhnke, 2008)). For example genes that were characterized as impulsive personality traits were located in the serotonergic system, the dopaminergic system, the monoamine metabolism pathway and the noradrenergic system and are all associated with alcoholism or some other addiction (M. A. Schuckit & Smith, 2001).
The opiodergic system is thought to mediate some of the rewarding pharmacological effects of alcohol, such as feelings of euphoria (Herz, 1997; Kreek, 1996). This has raised an interest in research to genes encoding for endogenous opioid receptors. With a particular focus on the mu-opioid receptor gene 1 (OPRM1). One of the most widely studied polymorphisms of the OPRM1 gene is the A118G single nucleotide polymorphism (SNP) located in the 118th position in axon 1, which codes for the AsnAsp40 substitution (rs1799971) (L. A. Ray, Mackillop, & Monti, 2010). Molecular studies of this polymorphism found that functional significance of this SNP suggested that the G allele has deleterious effects on both mRNA and protein production (H. Zhang et al., 2006; Y. Zhang, Wang, Johnson, Papp, & Sadée, 2005). Animal testing also shows that this SNP is associated with
increased alcohol response, consumption, and preference (Barr et al., 2007). Multiple studies have searched for the relationship between the A118G SNP of the OPRM1 gene and substance use disorders such as alcoholism (Bergen et al., 1997; Hernandez-Avila, Wand, Luo, Gelernter, & Kranzler, 2003; Kranzler, Gelernter, O’Malley, Hernandez-Avila, & Kaufman, 1998) while other did not find any association (Crowley et al., 2003; Franke et al., 2001; Gelernter, Kranzler, & Cubells, 1999; Loh, Fann, Chang, Chang, & Cheng, 2004). However recent research that focused on subjective response instead alcohol dependence suggest an enhanced response for individuals with at least one copy of the G allele (Ray & Hutchison, 2004) and was corroborated with a subsequent study (Ray & Hutchison, 2007). These individuals had higher scores of subjective intoxication, sedation and stimulation.
It was also argued that male carriers of the G allele reported higher level of alcohol craving following cue-reactivity (Van Den Wildenberg, Janssen, Hutchison, Van Breukelen, & Wiers, 2007). OPRM1 polymorphism is also associated with functional changes in mesocorticolimbic structures. Exposure to alcohol taste cues resulted in activation of these reward structures and that this activation was enhanced by a priming dose of alcohol in some of the structures in G118 allele-carrier individuals (Filbey et al., 2008). Likwise, activity in the striatum correlated with drinking behaviour in the same individuals. Contrary participants without this polymorphism showed activity in the mPFC. These findings indicate that alcohol cues elicit a higher biological response and may influence the development of incentive salience in A118G SNP individuals. Overall these results suggest that endogenous opioids might be involved in the reinforcing effect of rewarding subjective responses of alcohol. Recently, it has been indicated that both heavy drinkers and alcoholic patients have an approach bias for alcohol by using a novel alcohol approach avoidance task (alcohol -AAT) (Wiers, Eberl, Rinck, Becker, & Lindenmeyer, 2011; Wiers, Rinck, Dictus, & Van Den Wildenberg, 2009). In this task participant either pull or push a joystick in response to specific beverage related stimuli (Wiers et al., 2009). The rationale behind using joystick tasks to measure action tendencies is that arm flexion (pulling) is associated with more positive evaluations and approach motivational tendencies, whereas arm extension (pushing) is associated with more negative evaluations and avoidance motivational tendencies (Cacioppo, Priester, & Berntson, 1993). A training version of this task has also been used as a treatment strategy of alcohol abusers (Gladwin et al., 2011; Wiers et al., 2011). This bias was moderated by the G allele of the OPRM1 gene: carriers of a G allele demonstrated a particularly strong approach bias for alcohol, as well as for other appetitive stimuli.
The dopaminergic pathway also interacts with mu-opioid receptors. Alcohol induced dopamine release is said to be the consequence of increased opiodergic activity, which in turn inhibits GABAergic neurons and subsequently disinhibits dopaminergic neurons (Koob & Nestler, 1997; Niesink, Japsers, Kornet, & van Ree, 1999). Consequently, genes associated with brain dopaminergic activity have been commonly studied candidates. Abnormalities in the dopaminergic neurotransmission have been proposed to underlie the pathogenic mechanism of alcoholism (Hietala et al., 1994; Repo et al., 1999; Staley & Mash, 1996). Alcohol-induced drug reward is largely mediated via DRD2 (Blum et al., 1990; Comings et al., 1991) and neuroimaging evidence indicates that patients with alcohol dependence have decreased striatal DRD2 density (Repo et al., 1999). Besides, it is well documented that striatal dopamine activity is involved in reward processing.
The D2 dopamine receptor gene (DRD2) region is one of the most extensively investigated gene regions associated with dopamine receptor function, particularly the DRD2/ANKK1-TaqIa polymorphism (Bowirrat & Oscar-Berman, 2005). In humans, presence of an A1 allele of this polymorphism is associated with reduced expression of dopamine D2 receptors in the striatum. Another extensively studies polymorphism is found in the C957T SNP and is associated with higher striatal DRD2 binding potential (Hirvonen et al., 2009). Also this C-allele mutation is associated with an increase in mRNA stability and protein translation (Duan et al., 2003). Human imaging studies have revealed that decreased striatal DRD2 density is associated with executive function impairments, such as response inhibition, working memory and planning (White, Lawford, Morris, & Young, 2009). There is some preliminary evidence that C957T SNP is involved in impulsivity (M. J. White et al., 2009). Recent studies indicated that individuals with a homozygote polymorphism show impaired executive functions (Rodriguez-Jimenez et al., 2006; Xu et al., 2007). Moreover, this effect was associated with the COMT Val158Met polymorphism (Xu et al., 2007). Although no functional studies have examined the relationship with alcohol and DRD2 a nicotine administration study indicates that the homozygote polymorphism impaired working memory and decreased activity in the ACC, lateral prefrontal, and parietal cortices.
The catechol-O-methyltransferase (COMT) Val158Met polymorphism contains a common functional polymorphism, which lies in a region strongly implicated in schizophrenia (Levinson, 2003). This polymorphism is characterized by a substitution of methionine (Met) instead of valine (Val) at codon 158 which results in a 3- to 4-fold decrease in the activity of the COMT enzyme (Lachman et al., 1996). The COMT gene has been one of the most widely studied genes for increased risk of psychiatric symptoms and behavioural
problems, as a result of the critical role of COMT in the metabolism of central nervous system dopamine and norepinephrine. Carriers of the Val allele have been associated with impaired prefrontal activation in executive functioning, while carriers of the Met allele showed greater activation in the limbic regions in emotions processing (Mier, Kirsch, & Meyer-Lindenberg, 2010). However, research examining the association between the COMT Val158Met polymorphism and a broad range of psychiatric disorders has produced mixed findings (Laucht et al., 2012). The relationship between alcohol use and the Val148Met polymorphism has also resulted in inconclusive findings. Some found a correlation between Val/Val genotype and substance abuse (Vandenbergh, Rodriguez, Miller, Uhl, & Lachman, 1997); others found that late-onset alcoholics are often of a homozygote Met genotype (Tiihonen et al., 1999). It has been proposed that we can discriminate between two types of alcoholics, type I consist of late-onset, milieu-limited AUDs and type II early-onset, male-dominated, genetically influenced and impulse alcoholism (Cloninger, Bohman, & Sigvardsson, 1981). It has been suggested that different polymorphism underlie the differences in behavioural type 1 and type II alcoholic.
5. APPROACHING THE CUE
5.1 Cues, craving and the reward system
The social environment and context may also play a vital role in determining whether someone becomes an addict. Appetitive stimuli are likely to influence the behaviour of substance dependent subjects. Alcohol expectancy research contributed significantly to our understanding of problem drinking in young adults. Alcohol expectancies are implicit processes regarding the outcomes associated with the individuals' alcohol use and influence the initiation and duration of excessive drinking (Brown, Goldman, & Christiansen, 1985). For many years it was thought that addicts exhibited Pavlovian-like response to substance cues, however a meta-analysis in 1999 revealed that this is almost never the case (Carter & Tiffany, 2001). The substance-related cues did produce subjective craving and physiological arousal but did not act as a conditioned response. More recent theories state that substance-related cues acquire incentive-motivational properties, hereby altering the way in which the cue is perceived. The best known model is incentive-sensitization theory which states that repeated administrates of a substance produces a dopaminergic response that becomes sensitized with each following administration (Robinson & Berridge, 1993). Because of this sensitization the substance would acquire strong motivational properties, so that obtaining and self-administering the substance becomes an important goal, and strong subjective cravings for the substance develop (Field, Schoenmakers, & Wiers, 2008).
Functional imaging studies have revealed that substance cues have influence on the brain regions associated with the reward system. Upon exposure to alcohol related cues a positive correlation between craving and brain activity of PFC and limbic regions are to be seen in alcoholics (Myrick & Anton, 2004). The study further showed that exposure to these cues leads to higher craving ratings in drink dependent subjects compared to social drinkers. A cue reactivity study showed that AUDs adolescent showed substantially greater brain activation to pictures of alcoholic beverages that healthy teens (Tapert et al., 2003). The activity observed was predominantly in the left anterior, limbic and visual system areas. These results suggest that there is an association between BOLD response in reward system regions. Similar findings were reported by a combined PET/fMRI study that assessed alcohol cue reactivity in heavy drinkers (Heinz et al., 2004). They shown that VS dopamine D2 receptor availably was lower in recovering alcoholics and was inversely related with self-reported
craving. Striatal D2 availability in alcoholics was negatively correlated with alcohol induced activation in ACC and the mPFC. Thus alcohol facilitates dopamine release in the core region of the brain reward system. Although it is impossible to determine whether the availability of the D2 receptors or greater endogenous dopamine release results in the craving or cue-induced activation.
Ray et al. examined the effect of repetition priming and craving in high risk college students by exposure to alcohol and drug related cues (Ray, Hanson, Hanson, & Bates, 2010a; Ray, Hanson, Hanson, Rahman, & Bates, 2010b). Findings suggest that the brains of these individuals are similar to AUD subjects, suggesting that neural reactivity to alcohol cues may exist prior, or very early in the course of substance use. Enhanced activation in left caudate, left PFC, anterior cingulated and right insula could be observed. However, after repeated exposure to alcohol stimuli a suppression of neural activity could be observed. This is the result of repetition priming effect and suggests a potential role for the maintenance of substance use behaviour. Another fMRI study that investigated craving showed that in AUD subjects, compared to controls, there was significant increase in activation in the PFC and limbic regions when exposed to alcohol cues relative to non-alcoholic cues (Park et al., 2007). Limbic activation is one component of cue-induced craving and is likely involved in appetitive craving (Childress et al., 1999). These results support the idea that environmental stimuli associated with drinking influences the desire to consume alcohol in alcohol-dependent individuals (Myrick et al., 2004).
5.2 Approach tendencies and addictive behaviours
Recently researchers have advocated that specific biases in the cognitive processing of addictive people are a critical aspect of cue reactivity. It has been well documented that substance users gaze longer at substance-use related cues and demonstrate a stronger behavioural approach bias toward these cues (Field et al., 2008, 2010). Furthermore drug-related cues capture early selective attentions in drug abusers (Wiers et al., 2007). These response biases (attentional bias and approach bias) are mediated by underlying mechanism of incentive motivation as would be predicted by the incentive sensitization model.
Attentional bias and craving have been shown to have a mutually excitatory relationship, such that attentional bias towards alcohol elicits craving and in the same way, craving elicits attentional bias, forming a positive feedback loop (Field et al., 2008). Responses to stimuli with a positive valence are compatible with a behavioural approach tendency, whereas
