Decision Making In Adolescents: Effects of Sleep and Stress

By MING ZHANG, Guangzhou, China
Pioneer Research Program, 2015
Professor Jagmeet Kanwal

CONTACT AUTHOR

​"Adolescents are often associated with the negative image of staying wide awake at midnight, unable to get up in the morning, and sleepy in class. While adults tend to criticize such “unhealthy” sleep behavior, studies have shown that the adolescent brain is programmed to have unique sleep patterns."

ABSTRACT
​Adolescence is characterized by impulsive and risky decision making. Considering that the brain continues to develop throughout adolescence, the author of this paper hypothesized before researching on the topic that different developmental paces of different brain regions cause risky decision making in adolescence, which are exacerbated by stress and sleep patterns in this period. During research, it is found that the prefrontal cortex, a brain region associated with cognitive functions, is less developed in adolescence than the striatum and amygdala, which play important roles in reward and emotion processing...(continued)

Therefore, the striatum and amygdala are able to override the prefrontal cortex and prevents an adolescent from effectively making optimal decisions. Moreover, through experiments and brain scanning techniques such as fMRI, stress and sleep deprivation experienced in adolescence are found to further reduce activation in the prefrontal cortex and increase activation in the other two brain regions, making the existing imbalance of control in the three brain regions even more significant. Such findings suggest that adolescents need to learn to cope with their stress and get enough sleep for better performance inside and outside school. To gain a better understanding in this topic, further research needs to be done to investigate brain regions other than the three discussed extensively in this paper and determine their impact in adolescent decision making as well as their relationships with one another.

​1. INTRODUCTION
Adolescence is broadly recognized as the transition period between childhood and adulthood during which an individual physically and mentally matures. It starts with the onset of puberty and ends when social independence is achieved.[i] Patterns of behavior not present in children or adults are found in adolescents, who are characterized by risky decision making and impulsive sensation seeking.[ii] [iii] Moreover, adolescents experience stress and sleep patterns that are unique to their developmental period. In order to learn about the neural mechanisms behind adolescent behaviors as well as the relationship between decision making and different physical and mental states, researchers have looked into the brain for answers.

The human brain is a complicated structure divided into different regions that function in cooperation with one another while serving various purposes individually. Figure 1.1 shows some major brain structures along with their functions.[i] Older brain regions, including the brainstem, the thalamus, and the cerebellum, support basic functions of life such as heartbeat and movement.[ii] The limbic system, located under the cerebrum, is involved with functions such as memory and emotion. The cerebral cortex is the control and information processing center for the whole brain and is referred to as the newer part of the brain because its size increases along the hierarchy of animals.

Different parts of the brain are connected by networks of neurons, which transmit messages in the forms of electrical and chemical signals. As shown in figure 1.2, a neuron consists of three parts: the axon, the cell body, and the dendrite.[i]

Figure 1.3

Electric signals enter a neuron from its dendrites, pass through the cell body, and leaves at its axons, proceeding to the next neuron. At a inter-neuronal level, the axon of one neuron is linked to the dendrite of the next neuron and the space in between is called a synapse (Figure 1.3).[i] Neurons communicate with each other by releasing chemicals called neurotransmitters, which travel through the synapse to bind with the next neuron and, consequently, create a neural impulse when there are enough signals received by the dendrite of the next neuron.[ii] 
          Neurotransmitters play a crucial role in ensuring that the brain functions normally. Malfunction of neurotransmitters is linked to problems such as depression, seizures and even schizophrenia.[iii] Just as the brain is divided into different regions, each serving its own purpose, different types of neurotransmitters (e.g. serotonin, glutamate) specialize in different functions. One type of neurotransmitter called dopamine is closely associated with pleasure and reward seeking.
          Although 1) infants are born with disproportionately large heads that indicate relatively developed brains compared to the rest of the body and 2) the brain reaches approximately 90% of its adult size at the age of six, the human brain continues to develop long after birth and experience further structural changes, becoming fully mature when adulthood is reached.[iv] Moreover, different brain parts follow different tracks of development, resulting in some brain regions having dominant control over others during certain periods.[v]
 
Hypothesis: Risky decision making in adolescence is caused by the uneven pace of development of different brain regions, which are exacerbated by sleep deprivation and increased levels of stress.
 
          This paper will discuss decision making in the context of adolescence. By analyzing the developing adolescent brain and the neural mechanisms behind sleep and stress in this period, it will attempt to explain the adolescent tendency towards suboptimal decision making. This paper will also seek connections between these elements and consider their implications to adolescents in real life situations.

---

2. THE ADOLESCENT BRAIN

2.1     The Triadic Model
​The Triadic Model (Figure 2.1) is proposed by Monique Ernst to provide a framework for studying the adolescent brain.[i] According to the model, three neural systems interact to decide adolescent behaviors in different situations: approach (motivation), avoidance (emotion), and regulation.[ii]
​
          Motivation, which is a force or influence that leads to behavior, can be measured by how much a subject is willing to work toward a certain goal. Confusion between motivation and emotion may arise from the expression of “motivation to avoid,” but in this system motivation mainly refers to positive motivations, leaving the avoidance aspect to the emotional system. The brain region that has the most impact on this system is the striatum, which is part of the basal ganglia. Nucleus accumbens, located inside the ventral striatum, has a major influence on this system as it involved in predicting reward outcome.
          Emotion refers to feelings that subjects experience (mostly subjective) that influence the direction of their behaviors. The amygdala, which is part of the limbic system, plays a predominant role in emotions. Positive and negative emotions point to different directions, with positive associated with motivation and approach, and negative associated with avoidance. In the Triadic Model, emotion is connected to avoidance because based on available research results, this neural system has a stronger association with negative emotions. However, since most research studies have focused on negative instead of positive emotions, further research is needed to determine whether such a disproportionate association actually exists.
          Regulation is the balance between two neural systems and requires higher-level and more complex operations. This system is associated with the prefrontal cortex (PFC) because of its role in affective and cognitive processes. The PFC regulates the whole system by interacting with the striatum and amygdala, therefore controlling approach and avoidance behaviors.
          Based on the description above, it is important to note that the three systems are not entirely separate from each other; brain regions responsible for the systems sometimes overlap from neural, psychology, and functional perspectives, creating a much more complicated relationship between the three. The amygdala and the striatum both have some level of influence on avoidant and reward processing. The amygdala overall plays a dominant role in avoidant behavior, while the striatum plays a more important role in reward-driven behavior. The three systems are distinctly defined in order to simplify the model. However, these overlaps are important in addressing potential functional biases within systems and may change throughout the development of the brain, forming different patterns that characterize different stages such as adolescence.
 
2.2      Adolescent brain development
          Based on the Triadic Model, the study of the adolescent brain can be focused on three brain regions: the prefrontal cortex, the amygdala, and the nucleus accumbens. The development of these regions is reflected through a process called synaptic pruning. An overproduction of axons and synapses during puberty and rapid pruning in later adolescence are observed in multiple brain regions, especially the three mentioned above.[iii] The completion of pruning indicates that the involved brain region is developed. By studying the differences in pruning in different brain regions, researchers have found distinct developmental trajectories of different neural systems.
          During adolescence, prefrontal regions are found to be more active compared to other regions. Because the prefrontal regions are the higher level processing center of the brain, the change in activity level suggests linear increase in cognitive control and regulation.[iv] Based on this pattern of development of prefrontal regions, the fact that adolescents make less rational decisions compared to adults can be explained by their developing prefrontal cortex. Nonetheless, protracted development of prefrontal regions itself cannot explain the differences in behavior between adolescents and children because the pattern would indicate that children are worse in behaving appropriately due to their underdeveloped prefrontal cortex. However, adolescents are observed to be more impulsive than both adults and young children, contradicting the linear increase in cognitive control. Therefore, the development of other brain regions must contribute to adolescent behavior patterns as well.
          Synaptic pruning studies suggest that subcortical limbic regions (e.g. amygdala, nucleus accumbens) mature earlier than cortical (prefrontal) regions based on their less prolonged pruning, meaning that both the amygdala and the nucleus accumbens are relatively developed during adolescence compared to the PFC.[v] Dopamine receptors, which play an important role in the communication between cortical and subcortical regions, is overproduced in the amygdala and accumbens during early adolescence.[vi] Given the influence of dopamine on motivation, such a peak can explain the enhanced accumbens activity to rewards and elevated amygdala activity to emotional stimuli in the adolescent brain. Consequently, the limbic regions of the brain have more control on an individual’s behavior during adolescence than the cortical regions.
Neuroimaging studies of the brain also support the notion of different developmental trajectories influencing adolescent behavior. The two main methods used are magnetic resonance imaging (MRI) and functional magnetic resonance imaging (fMRI). As shown in Figure 2.2.1, MRI produces structural images of the brain, while fMRI is used to show brain activity of different regions.[i] MRI studies show that the PFC is one of the last regions to mature.[ii] fMRI studies reveal the bottom-up limbic and top-down prefrontal control regions relationship in the human brain and track the functional development of cortical and subcortical regions. Figure 2.2.2 shows different developmental trajectories for limbic and prefrontal regions of the brain.[iii] During adolescence, limbic regions are already close to maturation, while prefrontal regions are still developing in a linear pattern, making this period most seriously affected by the difference in developmental stages of the two regions. fMRI studies have also shown positive correlation between limbic subcortical activity and suboptimal choice behaviors. Since the limbic system is involved in motivation and emotional reactivity and the PFC serves to control it, it is reasonable for adolescents to display behaviors that appear to be less thoughtful.
          Figure 2.2.2 also explains why adolescents make better decisions in hypothetical situations than in real life ones. When emotion is involved, an adolescent will make decisions based more on his or her emotions than rational judgement because the amygdala and the striatum override the PFC. In hypothetical situations, emotional input is much less and therefore adolescents are able to analyze and consider different factors equally.

---

Figure 3.2.3

3. SLEEP AND DECISION MAKING
​
3.1      Adolescent sleep pattern
          Adolescents are often associated with the negative image of staying wide awake at midnight, unable to get up in the morning, and sleepy in class. While adults tend to criticize such “unhealthy” sleep behavior, studies have shown that the adolescent brain is programmed to have unique sleep patterns.
          After puberty, a teenager’s circadian rhythm moves back gradually.[i] In fact, melatonin is produced about three hours later in teens than in children or adults. If a teenager used to go to bed at 10pm and wake up at 7am, he or she would now go to bed at 12 to 1 a.m. and wake up at 9 to 10 a.m.. However, a regular high school typically starts at no later 8 a.m., not giving adolescents the chance to make up for the two hours they lose at night.
          Delayed sleep in adolescence is found to be connected to secretion of melatonin, a hormone that has to do with darkness and plays an important role in sleep. The body starts to secrete melatonin when the environment gets dark and its amount normally peaks at midnight, making people feel drowsy. In the morning, when the surroundings get brighter again, the amount of melatonin in the body decreases and tell people to wake up (Figure 3.1.1).[ii] Therefore, melatonin, which is influenced by exposure of light, serves as a signal telling the body when to sleep and when to wake up.
          Adolescence, as a critical developmental period, requires plenty of sleep. While nine hours of sleep is ideal for adolescents each day, they only get about seven hours of sleep on average. Moreover, adolescents sometimes procrastinate and leave most of their work to the end and consequently have to stay up late before deadlines. Therefore, most adolescents are more or less sleep deprived. Sleep deprivation can cause brain regions to function abnormally and consequently lead to impaired decision making in adolescents.
 
3.2     Sleep deprivation and the brain
          The PFC, a brain region crucial in the process of decision making, is especially vulnerable to sleep loss.[iii] An experiment done by Killgore et.al, using the Iowa Gambling Task (IGT), shows that people display impaired ability to make decisions based on experience of reward and punishment after 49 hours of sleep deprivation.[iv] The IGT is a paradigm that mimics real-world decision making under risk in a lab setting.[v] In the IGT, subjects are provided four decks that represent high reward, low reward, high punishment and low punishment. The IGT is set in a way that choosing the high reward and punishment side results in overall loss, while choosing the low reward and punishment side results in overall gain. During the study subjects gradually learn to take low risk so that they can gain, but it is found that those with lesions in the ventromedial prefrontal cortex (vmPFC) learn poorly compared to others. Based on the findings of another experiment that sleep deprivation reduces glucose metabolism within the PFC,[vi] researchers hypothesized before the experiment that sleep deprived subjects would show a deficit in decision making similar to that of patients with damaged vmPFC.
          The results, shown in figure 3.2.1, correspond to the hypothesis made by researchers before the experiment. Before sleep deprivation (baseline), subjects are able to identify the strategy to gain as much as possible as they go through more and more trials. In contrast, sleep deprived subjects show no overall improvement in the task over time, although they still perform slightly better than subjects with lesions in the vmPFC.
          Because the PFC is particularly involved in decision making with uncertainty, impairments in the PFC make people fall back to routines even when they need to modify their old solutions to suit variables in a new situation.[vii] As the experiment by Killgore et al. demonstrates, sleep deprived people are unable to adjust their strategy in the IGT according to patterns of loss and gain associated with different decks. When the PFC is disrupted, other cortical regions that seem less vulnerable to sleep loss take over and lead people to handle problems in a way that they are most familiar with instead of figuring out the best method. Therefore, not getting enough sleep can cause adolescents to lose creativity when dealing with tasks and prevents learning based on experience.
          One study done with mice shows that sleep deprivation (SD) leads to changes in intrinsic neural properties in the PFC. It causes opposite changes at synaptic and membrane levels, with a decrease in synaptic output, as seen in changes in miniature excitatory postsynaptic currents (mEPSCs), and increase in membrane excitability, which is reflected through action potential firing frequency (Figure 3.2.2).[viii] Although the changes are “intrinsic,” meaning that overall the effects appear to cancel out, such changes are likely to have some negative influences on PFC neurons as there is a shift in the input/output of PFC neurons. Further studies need to be done for researchers to gain a better understanding of how changes at a neuronal level makes a larger impact on the PFC overall.
          Sleep deprivation is also found to have an influence on subcortical regions such as the nucleus accumbens. By adopting a modified version of the IGT and
scanning the brains of participants in the experiment, Venkatraman et al. find that sleep deprivation leads to increased activation in the nucleus accumbens (Figure 3.2.3).[ix] The exact meaning of each column is not specified in this paper, but it is obvious from the diagram that in every case the nucleus accumbens is found to be more active during sleep deprivation. In the study, the nucleus accumbens is found to be more active with riskier choices in both states (with and without sleep deprivation). Therefore, elevated nucleus accumbens activity with sleep deprivation implies for riskier behavior.
          Figure 3.2.4 illustrates how sleep deprivation influences the balance between the three neural systems within the Triadic Model. Most of the studies done with sleep deprivation and decision making involve sleep deprivation for over 24 hours. The severity of sleep loss applied in labs are not so common in real life for adolescents, since it is rare for them to stay up all night without any sleep and not making up for the loss later. However, these studies do indicate a general trend of disruption in brain function if one does not get enough sleep.

---

4. STRESS AND DECISION-MAKING
​
4.1      Stress in adolescence
          Adolescents bear stressors from various areas such as academic, peer, and family. In order to get into good colleges, they have to take challenging courses at school that come with demanding coursework while excelling in extracurricular activities at the same time. In their social context, adolescents may experience conflicts with their peers while striving to have a sense of belonging. They also begin to get involved in romantic relationships and have to cope with relevant issues. At home, adolescents are often described by their parents as rebellious and have tense relationships with them. Based on the scope and scale of stress, adolescence can be considered a more challenging developmental period than childhood and even adulthood.

4.2      Stress and dopamine
          Stress and other aversive stimuli have been found to have an influence on dopamine and dopaminergic neurons. Anstrom and Woodward find that restraint, a model of acute stress, increases dopamine release by increasing firing rates and burst firing of dopamine neurons (Figure 4.2.1).[i] Figure 4.2.1 shows that compared to the baseline state, the brain experiencing restraint have neurons with significantly higher firing rates and percentage of spikes found in bursts, a neuronal property that reflects the activation of a neuron. Elevated dopamine activity means that the brain would be more responsive to reward, consequently leading to more reward-biased decisions.
 
4.3      Stress and brain regions
          Researchers have found chronic stress induced atrophy in the medial PFC (mPFC) and hypertrophy of the dorsolateral striatum (DLS) and dorsomedial striatum (DMS).[ii]
           As shown in Figure 4.3.1, stress causes the volume of the mPFC to decrease, meaning that this region of the brain is less active after stress exposure. In contrast, the striatum (both DMS and DLS) is found to be more active under stress. In this case, the striatum has more control in the system and the PFC has less control, similar to the sleep deprived state discussed earlier (Figure 4.3.2). Therefore, stress can also make people fall back to old solutions to problems and fail to deal with new situations effectively.
          Fortunately, stress-induced changes in the brain is reversible.[iii] In an experiment performed by Soares et al., subjects are tested after exposure to prolonged stress and then retested after a 6-week stress-free period. Through volumetric data analysis, it is found that brain regions are able to recover from structural changes caused by stress. In figure 4.3.1, upper panels represent volumetric variations in subcortical regions and the lower panels represent changes in cortical regions. Since the colors before and after stress recovery indicate opposite directions of volumetric change, it can be concluded that the brain is able to return to its original state if given time to relax after stress exposure.
 
5. CONCLUSION AND DISCUSSION

          This paper explained development of three brain regions during adolescence and the neural mechanisms of stress and sleep on decision making. The hypothesis stated that risky decision making in adolescence is caused by the uneven pace of development of different brain regions, which are exacerbated by sleep deprivation and increased levels of stress. Research on the human brain shows that different brain regions have distinct trajectories. Impulsive and risky decision making during adolescence is caused by imbalance of development between subcortical regions and cortical regions. The PFC, the amygdala, and the striatum together lead to the adolescent patterns of decision making. Because the PFC is relatively underdeveloped in adolescence than the amygdala and the striatum, these two regions are able to override the PFC in situations where reward is presented or emotion is involved. However, linear development of the PFC still indicates that adolescents, compared to children, have better analytical ability and can make judgements that reflect the corresponding level of their PFC development in hypothetical scenarios. In addition, the dopamine system in the brain plays an important role in reward seeking. During adolescence there is a peak in dopamine release and remodeling of the dopamine system, meaning that decisions are more reward-based.
          Adolescents are often sleep deprived due to a mismatch between the shift in their internal rhythm and external rules that make them wake up earlier than they should. Sleep deprivation can suppress activity in the PFC while increasing activation of the nucleus accumbens in the striatum (as illustrated in figure 3.2.3). Moreover, adolescents experience stress from various aspects of life, which can cause volumetric changes in multiple brain regions in a similar way as when they are sleep deprived. Considering the roles of the PFC and striatum in decision making, sleep deprivation and stress can lead to less thoughtful and more risky actions. Stress also has an effect on the dopamine system, which leads to more reward-biased decisions in stressed individuals.
          Therefore, sleep deprivation and stress in adolescence tend to aggravate the imbalance of control within the three key brain regions appointed by the Triadic Model that is originally caused by the uneven paces of development of those brain regions. Reducing the level of stress experienced by adolescents and making sure that they get enough sleep can minimize external influence on the balance of the neural systems. It is important to note that effects of impaired decision making extends beyond the feeling of sleepiness. While caffeine and short naps can make people feel refreshed, they cannot replace sleep at night as they fail to correct fundamental physiological disturbances that prevent the brain from recovering from sleep loss and functioning optimally.[iv] Moreover, adolescents need to find ways to relax after stressful events so that the brain can recover and functional normally.
          In this paper, the Triadic Model is adopted to provide a framework to focus the author's research on three brain regions. Although the PFC, the striatum, and the amygdala together have a major influence on adolescent decision making, it is undeniable that other cortical and subcortical brain regions also have some degree of control over decision making and are influenced by both stress and sleep deprivation. To get a more accurate account of the relationships between factors involved in the process of decision making, these brain regions will need to be considered in addition to the three main ones discussed in this paper.

​References