Perceptions of Threat and the Video Game From Hell

Their advice was usually employed when I reacted dramatically to a perceived threat, which in my teenage years was often. This is hardly unusual and may even have strong evolutionary roots. In humans (for whom planning of any kind is a species-defining cognitive salient), there is great selective pressure on the ability to anticipate threats and to plan for responses of any kind. Indeed, we devote a tremendous amount of neurological resources to evaluating threats and planning reactive strategies. We are able to do so even when no such threat exists, which is unusual in the animal kingdom.

If neural substrates exist for a complex, highly regulated behavior, any disruption of those substrates is likely to cause a behavioral change. Any deregulation of prognosticating neural substrates might lead to a broad spectrum of mental health problems, perhaps including panic attacks and anxiety disorders. Characterizing the interactions of such substrates is highly relevant to psychiatric professionals.

There has been recent progress in studies that focus on such planning activities. The researchers wondered: How does the brain respond to immi-nent versus remote threats? Which brain regions are involved and what light recent data may shed on issues such as anxiety is the subject of this month's column.

Two categories of background information need to be addressed to understand what happened--one behavioral, the other technological. Feel free to skip to "Video game from hell" if "executive function," "predatory imminence continuum," and "blood oxygen level-dependent (BOLD) signals" are working parts of your vocabulary.

As the luminary residents of the prefrontal cortex, executive functions control the ability to plan, set priorities, control impulse behaviors, shift attention, and weigh the consequences of actions. In humans these behaviors are developmentally regulated. The so-called teenaged brain lacks many of these functional elements, and this fact has been used to explain (or some would say excuse) a few of the more socially challenging habits of teenagers.

Another behavior involves responses to predatory threats. Cognitive neuroscience describes anxiety and fear responses to such threats in terms of the predatory imminence continuum. Put simply, this continuum is a function of the degree to which animals fear a given predator and the reassurance they have that they could escape if need be. It involves the lateral amygdala (the amygdala mediates the creation of emotions and their memories) and prefrontal cortex. These 2 regions are thought to coordinate reactions when the perceived threat is at a distance, resulting in classical "avoidance" or "escape" behaviors. A different system becomes activated as the threat becomes imminent, shifting more to the midbrain. The central amygdala becomes activated, and a region known as the periaqueductal gray (PAG) is recruited. Together, these regions coordinate "freezing" behavior.

Executive function and the predatory imminence continuum may interact with each other. There is evidence that the forebrain systems can actively inhibit midbrain defensive freezing maneuvers when the threat is distal, allowing the animal to move to safe ground. When the predator closes in and the aversive stimulus becomes unavoidable, the forebrain systems give up inhibitory signaling, which releases the animal into a more defensive posture.

Do these reactions occur in humans? Answering that question involves noninvasive detection during stressful experiences. The researchers in the experiments described below attempted to answer that question using high-resolution functional MRI (fMRI). fMRI measures a BOLD signal--a measure of changes in blood flow. Such changes have been shown to be closely associated with site-specific neural activation. fMRI takes "snapshots" of brain activation patterns at specific moments. Cleverly designed experiments executed while the subject is inside an fMRI device can reveal remarkable details about brain activation patterns. Brain activation responses to predatory threats were viewed with fMRI in the experiment described below.

As mentioned, the researchers were interested in the human perception of threat as an assessment of how proximal or distal the threat was so they devised a formal test of the predatory imminence continuum in humans. They were able to map behaviors by designing a most unusual video game, which was deployed as participants were given an fMRI. The fingers of the study subjects were also hooked to a device that could deliver an electric shock.

Subjects were allowed to manipulate a keyboard attached to a computer game similar to Pac-Man (Figure). The subjects' task was to evade an "electronic predator" in a 9 3 13-in gridded maze. The predator did not consist of ghosts, as in Pac-Man, but was a small red circle puttering around the maze. However, this was no ordinary red circle; if the subjects could not successfully evade it, they would receive a high- or low-intensity electric shock. It soon became clear to the subjects that this circle had the ability to follow, capture, and inflict real physical pain.

Before each experiment began, subjects were apprised of the likelihood of getting shocked during the trial. Contingencies could include a low probability, a high probability, or no probability of experiencing pain. This allowed researchers to view neural responses stimulated by the anticipation of an adverse stimulus before any chase commenced.

Once the chase commenced, brain activity was recorded at various times. The cleverness of this experiment came from the design of the game. The researchers could view how the brain anticipated pain during various intervals of the chase as the predator homed in on its victim. These included segments in which the predator was far from the subject, in which the predator began closing in on the subject, and in which the predator was very close to the subject.

Many controls were deployed over the investigation. These included "yoked trials," in which the subjects imitated the path of formerly executed chases. In some controls there were no predators; in other controls there were no threats of an electric shock. In all cases, fMRI images of the subjects' brains were obtained and compared.

What the investigators found was a brain cartographer's dream. When subjects were given advance notice that the chase was about to start, the researchers noticed a dramatic increase in activity in specific frontal cortical regions, including the orbitofrontal cortex, ventromedial prefrontal cortex, and anterior cingulate cortex (ACC). Why are these areas important? Some of these regions, such as the orbitofrontal region, are involved in planning and decision-making processes--the types of behaviors previously described as executive functions. The ACC assists in detecting when perceived unfavorable circumstances call for a change in behavioral strategy. It is specifically involved in threat detection. One way to interpret this is that the brain is detecting a potential threat and has begun to plan a series of actions to negotiate the upcoming task. The amygdala--the canonical mediator of human emotion--was also stimulated.

After the chase started, BOLD signals were detected in the cerebellum and PAG. Why is this important? The cerebellum is obviously involved in regulating certain types of motor behavior. PAG has been shown to be involved in organizing specific defensive responses in laboratory animals. This occurs whether the predator is natural or artificial. It is possible that similar functions occur in humans.

Researchers also uncovered something of a surprise in this phase of the response. Activity decreased in the ventromedial prefrontal cortex and amygdala. That last deactivation was not expected. Most cues that forecast threat, expected or not, involve this region of the limbic system. The brain was obviously shifting its priorities.

As the chase progressed, researchers began to notice increasing variability in the signals being detected. The variability was eventually explained as a series of patterns that depended on the proximity of the predator and the expected intensity of the shock. When the red circle was relatively remote, blood flow increased in the lateral amygdala and ventromedial prefrontal cortex. If subjects knew they were going to receive a mild shock, these remote BOLD signals were more intense. As the predator moved closer, however, the signals shifted to the central amygdala and PAG. If the subjects knew they were going to receive a strong shock, these signals were more intense.

Taken together, the researchers observed a predictable and, from a mammalian perspective, surprisingly familiar pattern. They noticed that as the predator approached, there was a general shift of BOLD signals away from the forebrain and into the midbrain. This notion was strengthened with the observation that the intensity of the signals in the PAG (and even its neighbor, the dorsal raphe nucleus) could be regulated, depending on whether the subjects feared the predator or felt they could escape from it. This is very similar to the predatory imminence continuum previously described. It is quite possible that by playing the video game from hell, researchers found a way to measure something that had previously been explained only in animals.

Do these data mean anything in terms of our understanding of anxiety or panic reactions in humans? The answer right now is "no," or more charitably, "not yet." But the data suggest many things, and certainly provide light for further and more penetrating research. As predatory threats move from distal to proximal perceptions, these data clearly show that differing neurological substrates are actively recruited. It is possible that these differences mediate the more subjective experience of fear. One way to look at anxiety responses is simply to say that a distal threat (or even a nonexistent threat) has somehow activated the proximal neural circuitry. This might cause a person to believe the threat is at the doorstep, even if no actual threat exists. There may even be a way to tease out anxiety from panic disorders. It is possible that anxiety is an aberrant reaction of the distal mechanisms, such as those found in the prefrontals. Panic may be described as an aberrant activation of the PAG and other substrates more associated with the midbrain.

These are just speculations, of course. The real contribution of these data is to outline how we perceive remote and imminent threats. The predatory imminence continuum, which is well characterized in animals, may be operating similarly in humans. My parents could not possibly know that their "90%" comment was an attempt to keep my developing forebrain and midbrain in proper regulatory shape. But in the end, that may be exactly what they were attempting to do. Perhaps video games serve a useful purpose after all.