What is the neuroscience explanation for refusing mental treatment?

Neuroscience may be starting to approach behavioral questions.

This post resulted from the Quora question that is the title of this post.

I answered, “I could make up a neurosciency explanation, but I won’t.”

He responded, “Please, i would like to know you line of reasoning on a possible neuroscience explanation on why someone refuse mental treatment, i need to know.”

While I don’t know of a research paper that covers this topic, I can see odd bits of the most modern neuroscience that may provide insight. I will provide a neurosciency answer, since there are so few other applications of modern neuroscience to behavioral issues, especially behavioral issues which have proven resistant to traditional therapies.

The question is important because we often see and experience people doing things which are clearly injurious, or in this case, not doing things which they probably know would help them.

Which end is up?

This post is based on the work of two pioneering 21st century neuroscientists, Karl Friston and Lisa Feldman Barrett. Their work has upended our understanding of what it means to be human, and their work has been largely accepted by neuroscience and related fields as the basis of how the brain works.

By upended I mean reversing our understanding of how living beings work from externally driven to internally driven. Over human existence this understanding has reversed several times, and in truth it’s both. We seem to be converging on a useful understanding.

The best way to understand which way is up is to ask “Where does my life come from?”

In the neuroscience classes I’ve taken, life starts with a sensation, which travels toward the brain, which magically provokes a motor action that might affect the world. While this view is useful for understanding how nerves work, it is useless for understanding how life works.

Now we know that the brain has a model of the world that predicts what will happen outside the brain and uses that prediction to act in the world to assure our survival, while using the senses to update the model. Some neuroscientists refer to this model as an hallucination, and that is not far from the truth.

Karl Friston

Neuroscience in the 20th century was about microscopes. Santiago Ramón y Cajal used the microscope to provide the first rich view of the brain’s neural networks by drawing the extreme richness of a neuron’s connections. This provided our first real understanding of the the detail of the brain, but such a view can only tell us a little bit about the 86 billion neurons we now know make up a human brain.

In the last decades of the 20th century we adapted x-rays and positrons to provide images of brains that show neural structure, which only made Cajal’s neurons more tantalizing.

Neuroscience in the 21st century is about Magnetic Resonance Imaging (MRI), which shows neural structure in greater resolution, and Functional Magnetic Resonance Imaging (fMRI) that gave us our first understanding of how those neural structures work together as we think.

Karl Friston is the leading researcher for the understanding and use of fMRI. Out of that research he proposed the free energy principle of the brain (FEP). As far as we can tell FEP is how all biological brains work.

According to the FEP, the isolated brain predicts what will occur in the senses and adjusts its predictions based on errors detected in the predictions. The brain’s model of reality is close to reality but not the same. All our actions and memories are based on the imperfect model, not the senses.

This means the model is on top of our lives, with senses and muscles descending from it. This approach means we are sometimes fooled by the senses, but it also means we can live in the present, rather than 100–500 milliseconds later.

Lisa Feldman Barrett

Lisa Feldman Barrett is the leading neuroscience researcher on emotions. She discovered that our 20th century view of emotions is not how they work, and showed a new view that is fully accepted by the neuroscience community.

She demonstrated that what we call emotions really consist of two processes: affect and emotional behavior, and she showed that this separation is much more effective at explaining human behavior.

Affect is the collection of internal bodily processes that monitor and modulate all the activities inside our bodies, primarily through the autonomic nervous system (ANS). The best know aspects of the ANS are heart rate, breathing rate, blood vessel constriction, bladder and rectum control, intestinal control, and dozens of others.

The brain models the autonomic nervous system similarly to how it models the senses of the peripheral nervous system. Whenever the brain’s model of the ANS doesn’t match what is expected, we have an affective response which is analogous to when the brain fails to predict the peripheral senses.

Affective responses from the ANS can trigger actions just as prediction errors from the PNS can, and the results of such responses and errors can interact. All of this usually happens below the level of our conscious awareness.

Emotional behavior is the collection of memories we have about how we responded to difficult situations in the past. We draw on them when the brain’s model of reality encounters another difficult situation, and they become a portfolio of behavior that we can draw on in the future. We adapt the memories to new situations, and we remember each new emotional behavior.

We recognize difficult situations because they usually result from an affective trigger in the memory context of one or more previous situations. If we find that a particular behavior is effective at resolving our difficulty, we will repeat it. The behaviors we try and the ones we repeat are affected by the culture we live in.

We are the model of reality in our brain

What is the model?

The human’s model first appears midway through pregnancy as the brain is created, with only a small inherited knowledge of how the world works. Other animals are born with a more complete model of the world, but that makes adapting to a changing world harder.

Human babies are born with the animal superpower of being able to adapt to a world that may be nothing like the one its parents grew up in, while being able to adapt to the a world that their parents may be struggling in.

The model is really the source of our life. It is the basis for our actions, attention, and attitudes. Moreover, it is what we remember, not the senses themselves, though in the best case the model of reality is driven closer to reality by sensory input.

The downside of this superpower is that it is built on early memories, so it is subject to disruption and corruption by early trauma. Trauma modifies our model of reality without changing reality itself. The result can be a vicious cycle, where a traumatic experience, worse yet a repeated traumatic experience, becomes imprinted on the brain and therefore becomes the context for understanding future experiences.

All memory is recalled without conscious awareness as it is used to guide us through life. Only when a memory becomes part of the story we tell ourselves (often called consciousness), do we know consciously what is in our memory. Otherwise it is just the ground we walk on every day unnoticed.

Where is the model?

Our model of reality isn’t really stored anywhere. We create the model as needed from memories, based on the context of what is happening in this moment. We draw memories from various locations in the brain and construct what the brain thinks is the best model for the current circumstances.

The model of reality is created by and used by the limbic system. The limbic system consists primarily of the amygdala, prefrontal cortex, and hippocampus.

The thalamus reports sensory prediction failures as surprises to the amygdala. The amygdala is the first step in the processing of these surprises. In the 20th century the thalamus was identified as the source of fear, but 21st century imaging has shown that it really is where reports of surprise can start being turned into various responses, including fear.

The amygdala also contributes to our model of reality. It holds our most fundamental memories, not as memories per se, but as neural circuits formed by some of our earliest learning about the world. These are memories that we can never recall consciously but affect our model of reality and then process the prediction failures that result from the model.

The prefrontal cortex allows us to store, process, and manipulate survival strategies far beyond the capabilities of the amygdala. The prefrontal cortex is intimately connected to the thalamus, coordinating its higher level strategies with the more reflexive tactics of the amygdala. It also has memories of past survival events in its neural circuits and uses them to initiate survival strategies that can be applied rapidly to a wider variety of circumstances.

The hippocampus is where most of our memories are stored, often using areas of the nearby temporal lobe to store the longer term ones. Don’t confuse the activity level of the hippocampus to the lazy level of memories associated with daydreaming. The hippocampus is always running as fast as it can, recalling every memory that might provide context to the current moment.

Hippocampal memories are stored in associative neural circuits that scale to holding the memories of a lifetime. It can take longer to recall its memories because most memories are recalled as a result of recalling previous memories, and each subsequent recall depends on the varying context of reality provided by the model.

The result of the hippocampus recalling a memory is simultaneously a context for understanding what is happening now and a context for storing new memories about what his happening now. This storage context is the basis for our associative memory facility.

How is the model used?

The model predicts what will happen in the world about a half second into the future. A half second may not sound like a big deal, but it’s generally enough to keep us alive and allow us to thrive.

All animals with brains predict. There may have been a time when they didn’t but once any animal gave up sensory veracity for the reduced activation latency of prediction, it was incumbent on all animals to do the same.

Acting a half-second later than a competitor or predator would spell death. That’s why our brains work like this, but that’s not all we do with it.

In our normal human lives, we use it to sing and dance in groups, have sparkling conversations, walk down the street without collisions or tripping, play with others and participate in sports, and just about any activity that involves reactions.

The model is both about us and about the world we are experiencing at each moment. It predicts how the world will be in half a second and how we can interact with it. Specifically, it initiates the motor activity we need to fulfill the predictions and the PNS and ANS sensory responses that will result.

The peripheral nervous system predictions are presented downward to the respective sensory cortex regions of the cerebrum (visual, auditory, tactile), which turn the cognitive predictions into sensory predictions. Each sensory cortex region tells the thalamus what to expect from the respective sensory input, and each difference between prediction and sensory reality results in a surprise being reported to the amygdala.

Another result of a predictive failure is that an error code is projected upward through the sensory cortex to keep the model of reality closer to reality. Many error codes may have to be reported upward over milliseconds or even seconds before the model is up-to-date, and the updated model continues to make better predictions, which typically converge on what the model predicts.

Of course, we aren’t aware of this continual prediction, testing, correction, responding, etc., cycle. It happens entirely outside of our conscious awareness. Some aspects of the process can enter our attention, but it’s necessarily only a small sample, else we would be overwhelmed by the complexity of the process.

What are the implications of modeling reality?

This newer, upside-down, inside-out, view of how the brain works has many implications that will change how we think about the brain in coming decades. Here I will just list some without explanation before answering the question at hand.

• The first grand implication of Karl Friston’s work was the revolutionary work of Lisa Feldman Barrett. As Dr. Barrett’s works expands his discovery into her field of clinical psychology, it will change how we treat people with brain problems.
• Mental illness can be seen as structural and learning issues with the brain rather than personality defects.
• The modeling operation of the limbic system is not based on language, logic, and imagery, but on learned expectations of thousands of sensory channels in the PNS and ANS.
• Memory is not coded reality, but the state of all the expectations of our model of reality.
• What we remember and how we remember it depends on what is important to the survival of our body, not what is important to a hypothetical audience in our reality.

Models come from memory

Types of memory

For the purposes of this question there are three important memory types.

• memory in the amygdala
• memory in the prefrontal cortex
• memory in the hippocampus

These memory types are all mediated by the limbic system.

Amygdala memory

Memory in the amygdala is about how to deal with surprise reported by various parts of the brain, most notably the thalamus, the respective sensory cortex regions, and the parts of the prefrontal cortex that monitor the autonomic nervous system.

The surprise reports are all delivered to the amygdala because it is capable of arousing the body to respond to surprises. It is connected to the hypothalamus to initiate the release of what are referred to as stress hormones, but which are really arousal hormones. These hormones prepare the body to respond to changes perceived in the external world.

The most common arousal hormones are cortisol and adrenaline. They prepare the autonomic nervous system to support muscular activity by increasing heart rate and breathing, constricting blood vessels to ensure blood availability, and a dozen other local body adjustments.

In addition, the amygdala can trigger immediate behavior by muscle groups through activation of the motor cortex and bodily responses such as crying and bowel distress through the ANS. This only happens when trauma has trained the amygdala to provide immediate responses. Unfortunately, amygdala activation is both rare and ineffective.

Amygdala memory consists primarily of reactive circuits, and it’s memories and use are entirely outside of conscious awareness.

Prefrontal memory

Prefrontal memory can respond to the same surprises as the amygdala does, but its responses can be more varied, complex, and nuanced, though they ae a bit slower. Think of it as an extension of the amygdala.

Prefrontal memory learns adaptive emotional behaviors, including entire activity sequences that have been learned in traumatic environments, though these sequences typically don’t involve recognizable speech or highly coordinated motor activity.

As with amygdala memory, prefrontal memory consists of reactive circuits, and its learning and use is entirely outside of conscious awareness.

Hippocampal memory

This is what most of us think of as memory. Hippocampal memory has many subtypes, depending on what is being learned and the context under which it is learned. It includes both short and long term memory, with a consolidation process that turns some short term memories into long term memories.

Short term memories are managed in the hippocampus, and long term memories often use adjacent areas of the temporal lobe to form associative circuits. The different types of associations represented by these memories are not important to this discussion.

Hippocampal memories are not reactive. They are used to inform the limbic system’s model of reality; the model may initiate actions based on them, but these memories don’t make actions happen by themselves.

While hippocampal memories are not directly accessible to conscious awareness, they often reach conscious awareness as part of the stories about the world model they contribute to constructing.

Memory type interaction

These three memory types have some interaction. Specifically, they don’t inform each other, nor are they learned, unlearned, or relearned in a coordinated fashion. They all operate in parallel, and their actions vie for expression.

The primary interactions among the types is that each happens in its own time and typically suppresses responses that come later. If there is a amygdala-based response, it will happen first and may interfere with later responses. If there is a prefrontal cortex response, it will usually be a bit slower, but it will take place as long as it isn’t interfered with.

The hippocampal response is much slower and takes more processing, so it is often unable to happen because of prior interference.

An amygdala-mediated response can often be recognized because of its fast latency, simple execution, and repeatability. A prefrontal cortex response can often be recognized because it is relatively low latency, more complex execution, longer lasting, and with some variability. A hippocampal response is typically slower to start, with greater complexity and variability, and is sometimes in the person’s awareness and control.

While hippocampal memories can be suppressed by the other two memory types, they can also be suppressed by other hippocampal memories. Our memories are not stored or structured to provide a coherent view of the world; they are stored as events occur. They may complement or contradict other memories, and contradictions are only resolved when we recall and try to use them.

I’ll mention the worst-case response for completeness. Some severe trauma results in myelination of the circuits within the limbic system, resulting in complex responses from the prefrontal cortex in the time frame associated with amygdala responses. Such circuits often result in serious behavioral issues with little ability to change or heal. Fortunately this is rare.

Refusing mental treatment?

An apology

I am sorry to throw so much at the reader for what may seem like a simple question about behavior. The problem we all face is that neuroscience and psychology are changing radically as you read this.

The philosophers Patricia and Paul Churchland predicted late in the 20th century that folk psychology, which is the collection of commonplace understandings of human behavior, would be entirely replaced by neuropsychological explanations that will be much different and more fundamental and useful.

We are in the midst of that change, and very few people outside neuroscience understand the new paradigm well enough for me to use it as context.

Refusing treatment

The word ‘refusing’ is a reaction based on a model of the world. A person suggesting, offering, or recommending a treatment does so based on their model of the world, and the person refusing treatment does so on their model of the world. The past experiences and memories of the world are likely to be radically different between the two.

Both the offeror and refuser are being rational within their own understanding of the world. That is the position of modern neuroscience. In other words, the refusal is not irrational, it is a rational choice based on a different understanding of the world.

Survival

Often, neither the person recommending the treatment nor the person refusing it can let go of this conflict. The recommender often has personal or professional experience that tells them what will happen if treatment doesn’t happen, and often both see the conflict in terms of the survival of the person refusing treatment.

Closer to the answer

The underlying issue of the answer is that the human brain is incredibly complex in ways that we don’t fully understand.

We often see our “self” as a unified whole, as if there were a “soul” or other singular entity running the show. Modern neuroscience has looked hard for such a central mechanism and not found it.

Instead, neuroscience has found hundreds of functional areas that contribute to our survival and behavior. Moreover, these areas don’t contribute to a centralized behavior mechanism: they all just express themselves, and behavior becomes the summation of their activity, with little attempt to rationalize or integrate their activity.

In short, neuroscience has found that there is no singular “I” in us, just as the Buddha found millennia ago.

The answer

What you are reading is an attempt to answer the deep question “What is the neuroscience explanation for refusing mental treatment.”

The answer is that our brains are a competing collection of needs, expectations, memories, models, intentions, and understandings, and they don’t agree on everything.

A common internal disagreement is between our understandings (based on hippocampal memories) and our needs (based in the amygdala.) Our amygdala can carry primal memories and survival responses that worked in the past, and the amygdala holds a privileged position in the limbic system for affecting our behavior. For example, we can hold the following opposing views: “I trust my therapist as an authority, but I have been severely harmed by a person in authority.”

The lack of an answer

Neuroscience is now in the early stages of teasing apart the contributions of the various functional units in the brain. We are using brain imaging to find what functional units are activated for a specific individual for a specific issue, but that is both expensive and insufficient.

Imaging is expensive because brain imaging requires expensive equipment and expensive expertise to use it.

Imaging is insufficient because we are only now examining the causes and mitigations associated with various brain areas. For example, each of us has a left and right amygdala, and we don’t understand how they differentially contribute to our lives.

I am confidant that neuroscience will make great progress on these and other issues in the next decades, but that may not help those who are suffering now.