ADHD children make poor decisions due to less differentiated learning processes

Which shirt do we put on in the morning? Do we drive to work or take the train? From which takeaway joint do we want to buy lunch? We make hundreds of different decisions every day. Even if these often only have a minimal impact, it is extremely important for our long-term personal development to make decisions that are as optimal as possible. People with ADHD often find this difficult, however. They are known to make impulsive decisions, often choosing options which bring a prompt but smaller reward instead of making a choice that yields a greater reward later on down the line. Researchers from the University Clinics for Child and Adolescent Psychiatry, University of Zurich, now reveal that different decision-making processes are responsible for such suboptimal choices and that these take place in the middle of the frontal lobe.

Mathematical models help to understand the decision-making processes

In the study, the decision-making processes in 40 young people with and without ADHD were examined. Lying in a functional magnetic resonance imaging scanner to record the brain activity, the participants played a game where they had to learn which of two images carried more frequent rewards. In order to understand the impaired mechanisms of participants with ADHD better, learning algorithms which originally stemmed from the field of artificial intelligence were used to evaluate the data. These mathematical models help to understand the precise learning and decision-making mechanisms better. “We were able to demonstrate that young people with ADHD do not inherently have difficulties in learning new information; instead, they evidently use less differentiated learning patterns, which is presumably why sub-optimal decisions are often made”, says first author Tobias Hauser.

Multimodal imaging affords glimpses inside the brain

In order to study the brain processes that triggered these impairments, the authors used multimodal imaging methods, where the participants were examined using a combined measurement of functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) to record the electrical activity and the blood flow in the brain. It became apparent that participants with ADHD exhibit an altered functioning in the medial prefrontal cortex – a region in the middle of the frontal lobe. This part of the brain is heavily involved in decision-making processes, especially if you have to choose between several options, and in learning from errors. Although a change in activity in this region was already discovered in other contexts for ADHD, the Zurich researchers were now also able to pinpoint the precise moment of this impairment, which already occurred less than half a second after a feedback, i.e. at a very early stage.

Psychologist Tobias Hauser, who is now researching at the Wellcome Trust Centre for Neuroimaging, University College London, is convinced that the results fundamentally improve our understanding of the mechanisms of impaired decision-making behavior in people with ADHD. The next step will be to study the brain messenger substances. “If our findings are confirmed, they will provide key clues as to how we might be able to design therapeutic interventions in future,” explains Hauser.

Literature:

Tobias U. Hauser, Reto Iannaccone, Juliane Ball, Christoph Mathys, Daniel Brandeis, Susanne Walitza & Silvia Brem: Role of Medial Prefrontal Cortex in Impaired Decision Making in Juvenile Attention-Deficit/Hyperactivity Disorder, in: JAMA Psychiatry

(Image caption: This is the happiness equation, where t is the trial number, w0 is a constant term, other weights w capture the influence of different event types, 0 ≤ γ ≤ 1 is a forgetting factor that makes events in more recent trials more influential than those in earlier trials, CRj is the CR if chosen instead of a gamble on trial j, EVj is the EV of a gamble (average reward for the gamble) if chosen on trial j, and RPEj is the RPE on trial j contingent on choice of the gamble. The RPE is equal to the reward received minus the expectation in that trial EVj. If the CR was chosen, then EVj = 0 and RPEj = 0; if the gamble was chosen, then CRj = 0. The variables in the equation are quantities that the neuromodulator dopamine has been associated with in previous neuroscience studies. Credit: Robb Rutledge, UCL)

Equation to predict happiness

The happiness of over 18,000 people worldwide has been predicted by a mathematical equation developed by researchers at UCL, with results showing that moment-to-moment happiness reflects not just how well things are going, but whether things are going better than expected.

The new equation accurately predicts exactly how happy people will say they are from moment to moment based on recent events, such as the rewards they receive and the expectations they have during a decision-making task. Scientists found that overall wealth accumulated during the experiment was not a good predictor of happiness. Instead, moment-to-moment happiness depended on the recent history of rewards and expectations. These expectations depended, for example, on whether the available options could lead to good or bad outcomes.

The study, published in the Proceedings of the National Academy of Sciences, investigated the relationship between happiness and reward, and the neural processes that lead to feelings that are central to our conscious experience, such as happiness. Before now, it was known that life events affect an individual’s happiness but not exactly how happy people will be from moment to moment as they make decisions and receive outcomes resulting from those decisions, something the new equation can predict.

Scientists believe that quantifying subjective states mathematically could help doctors better understand mood disorders, by seeing how self-reported feelings fluctuate in response to events like small wins and losses in a smartphone game. A better understanding of how mood is determined by life events and circumstances, and how that differs in people suffering from mood disorders, will hopefully lead to more effective treatments.

Research examining how and why happiness changes from moment to moment in individuals could also assist governments who deploy population measures of wellbeing to inform policy, by providing quantitative insight into what the collected information means. This is especially relevant to the UK following the launch of the National Wellbeing Programme in 2010 and subsequent annual reports by the Office for National Statistics on ‘Measuring National Wellbeing’.

For the study, 26 subjects completed a decision-making task in which their choices led to monetary gains and losses, and they were repeatedly asked to answer the question ‘how happy are you right now?’. The participant’s neural activity was also measured during the task using functional MRI and from these data, scientists built a computational model in which self-reported happiness was related to recent rewards and expectations. The model was then tested on 18,420 participants in the game ‘What makes me happy?’ in a smartphone app developed at UCL called 'The Great Brain Experiment'. Scientists were surprised to find that the same equation could be used to predict how happy subjects would be while they played the smartphone game, even though subjects could win only points and not money.

Lead author of the study, Dr Robb Rutledge (UCL Wellcome Trust Centre for Neuroimaging and the new Max Planck UCL Centre for Computational Psychiatry and Ageing), said: “We expected to see that recent rewards would affect moment-to-moment happiness but were surprised to find just how important expectations are in determining happiness. In real-world situations, the rewards associated with life decisions such as starting a new job or getting married are often not realised for a long time, and our results suggest expectations related to these decisions, good and bad, have a big effect on happiness.

"Life is full of expectations - it would be difficult to make good decisions without knowing, for example, which restaurant you like better. It is often said that you will be happier if your expectations are lower. We find that there is some truth to this: lower expectations make it more likely that an outcome will exceed those expectations and have a positive impact on happiness. However, expectations also affect happiness even before we learn the outcome of a decision. If you have plans to meet a friend at your favourite restaurant, those positive expectations may increase your happiness as soon as you make the plan. The new equation captures these different effects of expectations and allows happiness to be predicted based on the combined effects of many past events.

"It’s great that the data from the large and varied population using The Great Brain Experiment smartphone app shows that the same happiness equation applies to thousands people worldwide playing our game, as with our much smaller laboratory-based experiments which demonstrate the tremendous value of this approach for studying human well-being on a large scale."

The team used functional MRI to demonstrate that neural signals during decisions and outcomes in the task in an area of the brain called the striatum can be used to predict changes in moment-to-moment happiness. The striatum has a lot of connections with dopamine neurons, and signals in this brain area are thought to depend at least partially on dopamine. These results raise the possibility that dopamine may play a role in determining happiness.

Brain’s dynamic duel underlies win-win choices

People choosing between two or more equally positive outcomes experience paradoxical feelings of pleasure and anxiety, feelings associated with activity in different regions of the brain, according to research led by Amitai Shenhav, an associate research scholar at the Princeton Neuroscience Institute at Princeton University.

In one experiment, 42 people rated the desirability of more than 300 products using an auction-like procedure. Then they looked at images of paired products with different or similar values and were asked to choose between them. Their brain activity was scanned using functional magnetic resonance imaging (fMRI). After the scan, participants reported their feelings before and during each choice. They received one of their choices at the end of the study.

Choices between two highly valued items (high-high), such as a digital camera and a camcorder, were associated with the most positive feelings and the greatest anxiety, compared with choices between items of low value (low-low), like a desk lamp and a water bottle, or between items of different values (low-high). Functional MRI scans showed activity in two regions of the brain, the striatum and the prefrontal cortex, both known to be involved in decision-making. Interestingly, lower parts of both regions were more active when subjects felt excited about being offered the choice, while activity in upper parts was strongly tied to feelings of anxiety.

This evidence that parallel brain circuits are associated with opposing emotional reactions helps to answer a puzzling question, according to Shenhav: “Why isn’t our positivity quelled by our anxiety, or our anxiety quelled by the fact that we’re getting this really good thing at the end? This suggests that it’s because these circuits evolved for two different reasons,” he said. “One of them is about evaluating the thing we’re going to get, and the other is about guiding our actions and working out how difficult the choice will be.”

The study, “Neural correlates of dueling affective reactions to win-win choices,” was published July 14 in the Proceedings of the National Academy of Sciences. Shenhav conducted the research as a graduate student at Harvard University, along with Professor of Psychology and Neuroscience Randy Buckner, the study’s senior author.

A second fMRI experiment showed that the same patterns of emotional reactions and brain activity persisted even when the participants were told before each choice how similarly they had valued the items. Their anxiety didn’t abate, despite knowing how little they stood to lose by making a “wrong” choice. In a third experiment, Shenhav and Buckner tested whether giving people more than two choices increased their levels of anxiety. Indeed, they found that providing six options led to higher levels of anxiety than two options, particularly when all six of the options were highly valued items. But positive feelings about being presented with the choice were similar for two or six options.

This suggests that the anxiety stems from the conflict of making the decision, rather than the opportunity cost of the choice — an economic concept that refers to the lost value of the second-best option. The opportunity cost should be the same, regardless of the number of choices. In addition, subjects in this final study were given an unlimited amount of time to make a decision, compared with 1.5 seconds in the first two studies. The results showed that time pressure was not the main source of anxiety during the choices.

At the end of each study, participants had a surprise opportunity to reverse their earlier choices. Higher activity in a part of the brain called the anterior cingulate cortex around the time of an initial choice predicted whether that decision would later be reversed. Previous work has shown that this brain region is involved in assessing how conflicted an individual feels over a particular choice; this result suggests that some choices may have continued to elicit conflict after the participant made a decision, Shenhav said. The researchers also found that people who reported more anxiety in their daily lives were more likely to change their minds. 

This work could explain why ostensibly positive options can evoke a mixture of positive and negative responses, which are not explained by purely economic analyses of choice. “Rationally, there’s no reason why when you put one good thing with another good thing, you should feel worse about the situation,” said Brian Knutson, an associate professor of psychology and neuroscience at Stanford University, who is familiar with the work but was not involved in it. “The neuroimaging tells us that these different mechanisms are fighting with each other,” he said. “Understanding that dynamic can help us understand why decisions that we think should make us feel better can actually make us feel worse.”

According to Shenhav, this research could shed light on the neural processes that can make more momentous choices so paralyzing for some people — for instance, deciding where to go to college or which job offer to take. But he admits that even more trivial decisions can be tough for him. “I probably experience more win-win choice anxiety than the average person,” he said. “I’m even terrible at choosing where to eat dinner.”

Monkeys also believe in winning streaks

Humans have a well-documented tendency to see winning and losing streaks in situations that, in fact, are random. But scientists disagree about whether the “hot-hand bias” is a cultural artifact picked up in childhood or a predisposition deeply ingrained in the structure of our cognitive architecture.

Now in the first study in non-human primates of this systematic error in decision making, researchers find that monkeys also share our unfounded belief in winning and losing streaks. The results suggests that the penchant to see patterns that actually don’t exist may be inherited—an evolutionary adaptation that may have provided our ancestors a selective advantage when foraging for food in the wild, according to lead author Tommy Blanchard, a doctoral candidate in brain and cognitive sciences at the University of Rochester.

The cognitive bias may be difficult to override even in situations that are truly random. This inborn tendency to feel that we are on a roll or in a slump may help explain why gambling can be so alluring and why the stock market is so prone to wild swings, said coauthor Benjamin Hayden, assistant professor brain and cognitive sciences at the University of Rochester.

Hayden, Blanchard, and Andreas Wilke, an assistant professor of psychology at Clarkson University, reported their findings in the July issue of the Journal of Experimental Psychology: Animal Learning and Cognition.

To measure whether monkeys actually believe in winning streaks, the researchers had to create a computerized game that was so captivating monkeys would want to play for hours. “Luckily, monkeys love to gamble,” said Blanchard. So the team devised a fast-paced task in which each monkey could choose right or left and receive a reward when they guessed correctly.

The researchers created three types of play, two with clear patterns (the correct answer tended to repeat on one side or to alternate from side to side) and a third in which the lucky pick was completely random. Where clear patterns existed, the three rhesus monkeys in the study quickly guessed the correct sequence. But in the random scenarios, the monkeys continued to make choices as if they expected a “streak”. In other words, even when rewards were random, the monkeys favored one side.

The monkeys showed the hot-hand bias consistently over weeks of play and an average of 1,244 trials per condition. “They had lots and lots of opportunities to get over this bias, to learn and change, and yet they continued to show the same tendency,” said Blanchard.

So why do monkeys and humans share this false belief in a run of luck even when faced over and over with evidence that the results are random? The authors speculate that the distribution of food in the wild, which is not random, may be the culprit. “If you find a nice juicy beetle on the underside of a log, this is pretty good evidence that there might be a beetle in a similar location nearby, because beetles, like most food sources, tend to live near each other,” explained Hayden.

Evolution has also primed our brains to look for patterns, added Hayden. “We have this incredible drive to see patterns in the world, and we also have this incredible drive to learn. I think it’s very related to why we like music, and why we like to do crossword puzzles, Sudoku, and things like that. If there’s a pattern there, we’re on top of it. And if there may or may not be a pattern there, that’s even more interesting.”

Understanding the hot-hand bias could inform treatment for gambling addiction and provide insights for investors, said Hayden. “If a belief in winning streaks is hardwired, then we may want to look for more rigorous retaining for individuals who cannot control their gambling. And investors should keep in mind that humans have an inherited bias to believe that if a stock goes up one day, it will continue to go up.”

The results also could provide nuance to our understanding of free will, said Blanchard, who was drawn to the study of decision making during prior graduate training in philosophy. “Biases in our decision-making mechanisms, like this bias towards belief in winning and losing streaks, say something really deep about what sorts of creatures we are. We often like to think we make decisions based only on the information we’re conscious of. But we’re not always aware of why we make certain decisions or believe certain things.

“We’re a complex mix of biases and heuristics and statistical reasoning. When you put it all together, that’s how you get sophisticated behavior. We don’t know where a lot of these biases come from, but this study—and others like it—suggest many of them are due to cognitive mechanisms we share with our primate relatives,” said Blanchard.

Upfront and personal: Scientists model human reasoning in the brain’s prefrontal cortex

Located at the forward end of the brain’s frontal lobe, the mammalian prefrontal cortex (PFC) is the seat of many of our most unique cognitive abilities – collectively referred to as executive function – including planning, decision-making, and coordinating thoughts and actions with internal goals. That said, perhaps its most important attribute – one that is apparently unique to H. sapiens – is reasoning which, based on Bayesian, or probabilistic, inference, mitigates uncertainty by informing adaptive behavior. While the structural details of this remarkable process have historically remained elusive, scientists at Institut National de la Santé et de la Recherche Médicale, Paris, and Ecole Normale Supérieure, Paris and Université Pierre et Marie Curie, Paris have recently employed computational modeling and neuroimaging to show that the human prefrontal cortex involves two interactive reasoning pathways that embody hypothesis testing for evaluating, accepting and rejecting behavioral strategies. More specifically, their model describes behavior guided by reason in the form of an online algorithm combining Bayesian inference applied to multiple stored strategies with hypothesis testing that can update these strategies. In addition – as proposed in a previous work – the scientists conclude that since the frontopolar cortex (FPC), located in the anterior-most portion of the frontal lobes, is human-specific and is a key component in executive function decision-making, the ability to make inferences on concurrent strategies and decide to switch directly to one of these alternative strategies is unique to humans as well.

Prof. Etienne Koechlin discussed the paper that he, Dr. Maël Donoso and Dr. Anne G. E. Collins published in Science.

Read more

Does ‘free will’ stem from brain noise?

Our ability to make choices — and sometimes mistakes — might arise from random fluctuations in the brain’s background electrical noise, according to a recent study from the Center for Mind and Brain at the University of California, Davis.

"How do we behave independently of cause and effect?" said Jesse Bengson, a postdoctoral researcher at the center and first author on the paper. "This shows how arbitrary states in the brain can influence apparently voluntary decisions."

The brain has a normal level of “background noise,” Bengson said, as electrical activity patterns fluctuate across the brain. In the new study, decisions could be predicted based on the pattern of brain activity immediately before a decision was made.

Bengson sat volunteers in front of a screen and told them to fix their attention on the center, while using electroencephalography, or EEG, to record their brains’ electrical activity. The volunteers were instructed to make a decision to look either to the left or to the right when a cue symbol appeared on screen, and then to report their decision.

The cue to look left or right appeared at random intervals, so the volunteers could not consciously or unconsciously prepare for it.

The brain has a normal level of “background noise,” Bengson said, as electrical activity patterns fluctuate across the brain. The researchers found that the pattern of activity in the second or so before the cue symbol appeared — before the volunteers could know they were going to make a decision — could predict the likely outcome of the decision.

"The state of the brain right before presentation of the cue determines whether you will attend to the left or to the right," Bengson said.

The experiment builds on a famous 1970s experiment by Benjamin Libet, a psychologist at UCSF who was later affiliated with the UC Davis Center for Neuroscience.

Libet also measured brain electrical activity immediately before a volunteer made a decision to press a switch in response to a visual signal. He found brain activity immediately before the volunteer reported deciding to press the switch.

The new results build on Libet’s finding, because they provide a model for how brain activity could precede decision, Bengson said. Additionally, Libet had to rely on when volunteers said they made their decision. In the new experiment, the random timing means that “we know people aren’t making the decision in advance,” Bengson said.

Libet’s experiment raised questions of free will — if our brain is preparing to act before we know we are going to act, how do we make a conscious decision to act? The new work, though, shows how “brain noise” might actually create the opening for free will, Bengson said.

"It inserts a random effect that allows us to be freed from simple cause and effect," he said.

The work, which was funded by the National Institutes of Health, was published online in the Journal of Cognitive Neuroscience.

Rats show regret, a cognitive behavior once thought to be uniquely human

New research from the Department of Neuroscience at the University of Minnesota reveals that rats show regret, a cognitive behavior once thought to be uniquely and fundamentally human.

Research findings were recently published in Nature Neuroscience.

To measure the cognitive behavior of regret, A. David Redish, Ph.D., a professor of neuroscience in the University of Minnesota Department of Neuroscience, and Adam Steiner, a graduate student in the Graduate Program in Neuroscience, who led the study, started from the definitions of regret that economists and psychologists have identified in the past.

"Regret is the recognition that you made a mistake, that if you had done something else, you would have been better off," said Redish. "The difficult part of this study was separating regret from disappointment, which is when things aren’t as good as you would have hoped. The key to distinguishing between the two was letting the rats choose what to do."

Redish and Steiner developed a new task that asked rats how long they were willing to wait for certain foods. “It’s like waiting in line at a restaurant,” said Redish. “If the line is too long at the Chinese food restaurant, then you give up and go to the Indian food restaurant across the street.”

In this task, which they named “Restaurant Row,” the rat is presented with a series of food options but has limited time at each “restaurant.”

Research findings show rats were willing to wait longer for certain flavors, implying they had individual preferences. Because they could measure the rats’ individual preferences, Steiner and Redish could measure good deals and bad deals. Sometimes, the rats skipped a good deal and found themselves facing a bad deal.

"In humans, a part of the brain called the orbitofrontal cortex is active during regret. We found in rats that recognized they had made a mistake, indicators in the orbitofrontal cortex represented the missed opportunity. Interestingly, the rat’s orbitofrontal cortex represented what the rat should have done, not the missed reward. This makes sense because you don’t regret the thing you didn’t get, you regret the thing you didn’t do," said Redish.

Redish adds that results from Restaurant Row allow neuroscientists to ask additional questions to better understand why humans do things the way they do. By building upon this animal model of regret, Redish believes future research could help us understand how regret affects the decisions we make.