A distinct adolescent phase as the transition between childhood and adulthood is characteristic for several mammalian species, defined by sexual maturation and a period of brain remodeling. Depending on species this process happens over a few weeks or several years, as in humans. Next to distinct maturation in the brain, adolescence is accompanied by notable behavioral changes. For instance, adolescents show more grooming behavior, social interactions, and playfulness, as well as increased novelty seeking, risk-taking and altered reward sensitivity (Spear, 2000). Those behavioral modifications are believed to have adaptive value. For example, becoming independent, by exploring the environment and leaving the protected childhood nest, as well as developing social skills that are important to become an adult member of their social environment (Crews, He, & Hodge, 2007). But those rapid changes observed in brain and behavior seem to make adolescents as well more vulnerable to adverse developments, such as depression, eating disorders and substance abuse (Kelly, Schochet, & Landry, 2004). Especially novelty-seeking and risk-taking have been suggested to be involved in increased susceptibility to adolescent drug-using behaviors. Likewise, substance abuse in adolescence seems to be connected to an increased risk for continued or later reemerging drug consumption.
Adolescence is related to widespread changes in hormonal production, such as the activation of the hypothalamus-pituitary-gonadal axis (HPG), gonadotropin-releasing hormone (GnRH), and the consequent production of luteinizing hormone (LH) and follicle stimulating hormone (FSH). While those hormones are involved in maturational aspects of puberty, risk-taking behavior in adolescence can be connected to the steroid hormone testosterone. During adolescence, testosterone increases in both genders, though this increase is more significant in boys (Grumbach, 2002). This gender difference is in line with the finding that risk-taking behavior is more prominent in adolescent boys than girls (Kelly et al., 2004). Further, individual differences in testosterone have been attributed to behavioral problems and other psychopathology in adolescent boys and girls (Granger, Shirtcliff, Zahn-Waxler, Usher, Klimes-Dougan, & Hastings, 2003). One study found, that in adolescent boys the level of testosterone was associated with non-aggressive risk-taking (Vermeersch, T’sjoen, Kaufman, & Vincke, 2008). However, those results were related to social factors, meaning that boys with higher levels of testosterone tended to have friends who engage in increased risk-taking. Similar results have been shown in adult males who showed increased risk-taking behavior after drug-induced elevation of testosterone (Goudriaan, Lapauw, Ruige, Feyen, Kaufman, Brand, , 2010). Moreover, Peper, Koolschijn, and Crone (2013) reported that testosterone levels in male and female adolescents were not just related to risk-taking behavior, but characteristics of the orbitofrontal cortex (OFC). The OFC is involved in decision-making processes, and OFC morphology was identified as a significant mediator between testosterone and risk-taking behavior. Interestingly, there was a gender difference in this association. For males, smaller grey matter volume in the OFC was linked to higher testosterone, whereas for females it was a smaller OFC surface area. But not just risk-taking behavior can be linked to testosterone levels, novelty-seeking and sensation-seeking have been found to be correlated to salivary testosterone levels in females (Kerschbaum, Ruemer, Weisshuhn, & Klimesch, 2006). Further, a study with all male subjects suggests that testosterone and a variant of the dopamine receptor gene (7R+ allele) may both predict novelty-seeking behaviors in adolescence (Campbell et al., 2010). But while testosterone is linked to risk-taking as well as novelty seeking behaviors in several studies, it has been implied that this characteristic adolescent behavior may be even more initiated by significant changes in brain structure and the dopaminergic system (Spear, 2000; Steinberg, 2008).
Adolescent Brain Development
The adolescent brain undergoes substantial changes, such as extensive pruning, while on the other hand myelination, especially in the human PFC increases (Giedd et al., 2006). The Corticolimbic circuit is among the brain structures that undergo extensive remodeling and are likely to be involved in characteristic behaviors such as risk taking, novelty and sensation seeking, as well as associated reward sensitivity (Crews et al., 2007). One of the related structures is the prefrontal cortex (PFC), which is involved in decision-making processes, including risk-reward tradeoff, behavioral inhibition, and planning. Further, during adolescence, the PFC connectivity is still undergoing myelination. Of particular interest is the amygdala-PFC connection, involved in emotion regulation and threat learning (Monk et al., 2008). In a study with participants between four and 22 years old, the amygdala-prefrontal circuit has been shown to switch from positive connectivity until the age of ten, to negative functional connectivity in individuals older than ten years of age (Gee et al., 2013). Further, it was shown that this transition was accompanied by a decrease in amygdala reactivity and a decline in reported childhood anxiety. The PFC has been implied as a regulator of amygdala reactivity and the switch to negative connectivity could reflect one essential mechanism for the increase in novelty-seeking and risk-taking. Also, the ventral striatum or nucleus accumbens is considered part of the corticolimbic circuit and a primary target in the dopaminergic system. Changes during the transition to adolescence could play a role in increased reward-seeking behaviors, such as in drugs, and possibly even in impulse control (Dalley et al., 2007; Kelly 2004). One cross-sectional study investigating responses to a reward-based paradigm in adolescents showed amplified activity in the nucleus accumbens compared to children and adults, while their activation in the orbitofrontal cortex was similar to children (Galvan et al., 2006). Thus, maturational differences may be involved in disproportional activation. The transition within the corticolimbic circuit and the associated dopaminergic projections may play a role in increased risk-taking, novelty or reward-seeking behaviors, contributing and predispose to vulnerability to substance abuse.
But it is not just brain development that enables characteristic behaviors and higher vulnerability. In turn, those behaviors may alter brain development and have a lasting effect on future development. The increased plasticity through adolescent brain remodeling may enable stronger and lasting learning mechanisms. For instance, adolescents who form drinking habits before the age of 15 have an increased risk to develop substance abuse disorders later in life (Grant and Dawson, 1998). Subjects with adolescent alcohol use disorder show smaller hippocampal volumes compared to healthy controls (Nagel, Schweinsburg, Phan, & Tapert, 2005), the extent of reduction in hippocampal volume was associated with age of onset and duration of the adolescent alcohol use disorder (DeBellis et al., 2000). Further, adolescent alcohol abuse was found to be related to lower volumes in the prefrontal cortex and decreased white matter connectivity in the PFC (Bellis, Narasimhan, Thatcher, Keshavan, Soloff, & Clark, 2005).
The Neurotransmitter System
Increased risk-taking and altered reward sensitivity are considered critical components of adolescent drug vulnerability. While these behaviors can be connected to changes in brain connectivity, it is also hypothesized that increased reactivity in the adolescent dopaminergic system may contribute (Wahlstrom, White, & Luciana, 2010). The dopamine system is assumed to play a role in behavioral activation, whereby especially the connections to the corticolimbic circuit, including the ventral striatum and the associated dopaminergic system can be connected to motivation and reward processes (Robbins & Everitt, 2002). The increase in activity of dopaminergic neurotransmission is assumed to help the adolescent turning motivation into rewarding actions, encouraging independent behavior in this critical transitional phase into adulthood. This increased activation in the dopaminergic system may lead to increased reward-motivated approach behaviors, while behavioral control or executive function is still immature (Wahlstrom et al., 2010). The reorganization of the dopaminergic system during adolescence is characterized by significant overproduction of dopamine receptors and transporters in relevant brain regions such as the hippocampus, cortex, striatum and nucleus accumbens. The rapid increase is followed by subsequent pruning and finally stabilization in early adulthood, suggesting a complex remodeling of the reward system (Andersen, Thompson, Rutstein, Hostetter, & Teicher, 2000; Crews et al., 2007). This comprehensive process of maturation and reorganization may contribute to the increased preservation of adolescent behaviors such as substance abuse into later adulthood. For instance, it is assumed that the transition from voluntary drug use to a habituated substance abuse is paralleled by a shift of control from the prefrontal network to increased control in the striatum with its dopaminergic innervations (Everitt & Robbins, 2005). This transition could be amplified during a sensitive period with increased neuroplasticity and restructuring of involved circuits.
Consequently, this suggests that substance abuse during adolescence impacts the remodeling of the dopaminergic system, leading to long-term effects on related behaviors. This effect has been reported using an animal model. Adolescent rats that were exposed to ethanol showed an increased and prolonged activation of the dopaminergic system to a subsequent treatment during adulthood compared to rats that had no juvenile ethanol experience (Pascual, Boix, Felipo, & Guerri, 2009). Further, animals with adolescent exposure showed significant down-regulation of the D2 dopamine receptor expression in the prefrontal cortex, along with epigenetic modifications in striatum and nucleus accumbens. This indicates an effect of alcohol exposure on adolescent brain maturation with long-term changes in the dopaminergic system, and consequently implied alterations in reward processing.
Similar to dopaminergic system, nicotine receptors are as well overproduced during adolescence (Naeff, Schlumpf, & Lichtensteiger, 1992; Trauth, Seidler, McCook, & Slotkin, 1999). And analogous, Trauth et al. (1999) found that adolescent exposure to nicotine results in a significant up-regulation of receptors in the midbrain, hippocampus, and cerebral cortex. Further, the adolescent nicotine exposure leads to long-term changes that remained stable even after exposure termination. This may contribute to the increased and lasting vulnerability to adolescent-onset substance abuse. Similar results were found in a study looking at self-administration of nicotine in rats (Adriani et al., 2003). Animals that received nicotine during adolescence showed increased self-administration behavior as adults, compared to animals pretreated with saline. These long-term changes in sensitivity to nicotine were related to an up-regulation in the nicotinic acetylcholine receptor. Additionally, it has been shown in mice that increased novelty seeking occurs during adolescence compared to juvenile mice and adults. Using an elevated plus-maze paradigm adult and juvenile mice showed increased avoidance behavior towards the open arm of the plus-maze, while adolescent mice entered the open arm more frequent and spent approximately equal time between open and closed areas (Macri, Adriani, Chiarotti, & Laviola, 2002). Interestingly, the authors reported no difference in risk-assessment behaviors between the age groups, indicating increased willingness to take risks, as well as novelty and exploration drive in adolescent mice. In a follow-up study, those results were replicated and shown that nicotine consumption during adolescence can change this behavioral pattern (Adriani et al., 2004). That is, adolescent mice that were pre-exposed to nicotine exhibited less exploration on the open arm. Further, the adolescent nicotine exposure down-regulated the GluR2 and GluR3 subunits of the AMPA glutamate receptor. Also, adolescent exposure affected adult behavior in such that animals showed increased locomotor activity in the context of novelty, which is an established indicator of increased vulnerability of addictive properties of psychostimulants (Piazza, Deminière, Le Moal, & Simon, 1989).
Models of Vulnerability
Several animal models try to include individual personality traits to study vulnerability to substance abuse. For example, Dalley and colleagues (2007) measured trait impulsivity in rats and reported that high impulsivity scores related to higher susceptibility to cocaine. Also, they were able to show that impulsivity in rats was associated with dopamine receptors in the nucleus accumbens and that those deviations in dopamine receptors were observable before first exposure to the drug. The dissociation of individual phenotypes is one useful strategy to understand the influence of vulnerabilities to substance abuse. Next to predisposing traits, individual differences in substance use may play a role in susceptibility. In rodent models, locomotor activity in response to novelty has been used as a model for novelty-seeking behaviors (Piazza et al., 1989). They showed that higher locomotor response to this paradigm predicts increased self-administration of substances when available, indicating a higher vulnerability in animals with increased novelty seeking behavior. Combining both of those traits into one model studies were able to show that novelty-seeking may be a good predictor of drug self-administration. While impulsive behavior traits are more related to actual addictive behavior (Belin, Berson, Balado, Piazza, & Deroche-Gamonet, 2011; Belin, Mar, Dalley, Robbins, & Everitt, 2008). Accordingly, showing that the combination of both traits may have the highest predictability for the development of addictive behavior.
Increased risk-taking and novelty-seeking behaviors seem to occur during adolescence across several mammalian species and are considered factors of higher susceptibility to substance abuse. This increased vulnerability has been attributed to extensive remodeling in the brain, with an emphasis on the corticolimbic circuit, as well as the neurotransmitter system. Moreover, the different rates of maturation may explain part of the phenomenon such as an early increase in novelty and reward-seeking behaviors, combined with late maturing self-control systems, such as the orbitofrontal cortex. The increased neuroplasticity due to adolescent brain remodeling may contribute to the stabilization of adverse behaviors and increased long-term vulnerability to drugs after adolescent exposure. However, not all adolescents show increased vulnerability, differences in individual personality traits seem to be one moderating factor. Further research identifying the early marker of susceptibility and underlying neurological factors could inform future prevention of substance abuse.