Friday, June 08, 2012

Questions for journal club on neuroeconomics and psychiatry

What is neuroeconomics?

See Loewenstein et al. (2008) for a detailed introduction into how economics, psychology, and neuroscience  are beginning to converge.

What does neuroeconomics have to offer to psychiatry (and vice versa)?

Psychiatric diagnoses are typically based on clusters of symptoms and are inherently atheoretical. Economic concepts may offer a potential integrated computationally based framework for understanding psychiatric problems and even for improving how psychiatric conditions are defined.

What are the economic concepts of key relevance to psychiatry?

Single-dimension utility, marginal rate of substitution, expectation, uncertainty, gain-loss asymmetry, social preferences, time preferences, revealed & real preferences, others?. Each of these is described and their relation to psychiatry and disorder discussed in detail as summarized below:

Single-dimension utility: A global account of all choice behaviour. Studies examining neural valuation networks provide some support for the idea that a global valuation system overlaps with brain areas & networks implicated in psychiatric dysfunction. Monamine neurotransmitters are strongly implicated in both psychiatric disorder and the evaluation of rewards/punishments.

Example of integration of neuroeconomics & psychiatry in this area: Forming a quantitative measure of the severity of depression by examining response to financial incentives.

What does this tell us about the cause or mechanisms involved in depression?

How would the utility function of a depressed person differ from others (higher valuation of costs- including enduring risks)?

How could this research design be integrated with functional imaging/other tools of neuroscience?

Marginal rate of substitution: The basic idea here is the amount of one good a person will exchange/give up for a certain amount of another good. However, addicts often resist substitution from heroin to methadone, from cigarettes to nicotine replacement devices and so forth. Addictive drugs may be readily available (as opposed to more complex rewards), may be valued far more greatly than other goods and may attenuate the value of other rewards.

Is it tautological to say that a low marginal rate of substitution for addictive goods is a risk factor for addiction?

Hasler proposes that people may identify with their disorder and may feel unwilling to give up this part of their identity and compares this to the endowment effect. Alternatively, the symptoms of disorder could be valued (e.g. rituals in OCD, delusional ideas in schizophrenia, body form in anorexia) and people may not be willing to give these up or substitute them for alternative behaviours/thoughts/bodily states and so on.

Is there evidence for the role of synaptic plasticity in the substitution of rewards?

Expectation: Habit learning is affected in many psychiatric conditions leading to dysfunctional symptoms like the avoidance of perceived but non-existent threats (as in PTSD). Some emotional states reflect the expected utility of an anticipated outcome (potentially anxiety, fear, depression, hope-optimism).

"Hot stove" effect - avoidance brings success and is repeated leading to what Hasler calls biased sampling which will include the neglect of positive alternatives. This failure to explore is key to attachment theories and is one interpretation of why the profound lifetime effects of early childhood maltreatment/abuse occur.

Disengagement is mentioned but not elaborated. It is likely to be an expectation driven effect that is prominent in depression (excessive disengagement) and symptoms linked to perfectionism (failure to disengage).

Uncertainty: Decision to take risks depends on the concavity of the utility function (standard model) and the probability weighting function (prospect theory).

Can neuroticism be construed as a negative attitude towards risk? (certainly linked to poor decision-making in contexts of evaluation of risk as evident for example in a paper by Denburg et al. (2009) on Iowa gambling task performance).

An impaired evaluation of risk may contribute to stress-induced psychopathology, and an aversion to or adverse reaction to the ambiguity inherent in social relationships may be a key component of borderline personality disorder, social anxiety, conduct disorder (what are you looking at?!).

Gain-loss asymmetry: Loss aversion implies that losses loom larger than gains (potentially 2:1) and there is also diminished sensitivity to larger losses (see Fig 2.).

Might anxiety and depression be linked to excessive loss aversion?

Bipolar patients do not appear to show the normal pattern of diminished sensitivity to losses or gains, which is one explanation for their excessive pursuit of rewards and feelings of despair.

Social preferences: Although people are self-interested they also value the well-being of others, often behave altruistically, are averse to inequity, betrayal, and "free-loading". Economic games can be used to produce a behavioural and computational account of social preferences and when combined with neuroimaging a neural account which can be used to estimate the extent of dysfunction.

Examples: The ultimatum game- person A offers B a certain allocation of a sum X, B can choose to accept this allocation or reject it in which case neither receive money.

An exaggerated response to unfairness could place the individual at risk for affective disorders like depression. A diminished response to unfairness could indicate psychopathy (characterized by a lack of moral disgust).

Example: Trust games - person A gives B a certain allocation of sum X, B can choose to send some of the allocation back to A (this is typically doubled or tripled). Deficient in trust shown amongst those with borderline personality disorder.

Depression has been characterized as an involuntary defeat syndrome which is an adaptive response to hierarchical conflict indicating one has accepted a low rank. This can be studied through economic games (e.g hawk-dove).

Time preferences: Intertermporal choices - trade-off gains and losses over varying time horizons (e.g. any credit decision).

Internalizing and externalizing problems are linked to self-control and future orientation (potential overinhibition, overcontrol, excessive future orientation vs. disinhibition, lack of control, present orientation). Present-orientation could be a risk for substance abuse, ADHD, mania, conduct disorder, impulsive aspects of psychopathology (potentially suicide risk, psychopathy).

Anhedonia is associated with an inability to experience contemporaneous rewards. Abraham Lincoln example - anhedonia may facilitate future orientation for which Lincoln was renound. Acute depression, on the other hand, is linked to hypersensitivity to future negative outcomes.

Becker & Murphy rational addiction model predicts that present-orientated people will be at risk of addiction and that their consumption may be utility maximizing. Hasler proposes that the model enables a precise understanding of how stress affects consumption, the need for quick withdrawal of addictive substances, and the need for a substance even if the positive affective impact has diminished.

Though there is evidence that time and risk preferences are similar to the general population in those with psychiatric disorders such as addiction and bulimia nervosa, it is also clear that visceral factors (Loewenstein, 1996) can induce a present-orientation and that such factors are typically not fully considered when people plan goal directed actions.

Revealed and real preferences

Is there a disconnect between revealed and real preferences in psychiatric disorders?

Important problem, as up to 50% of the population experience a psychiatric disorder during their lifetime (Kessler et al., 2005) and it is likely that mild psychiatric symptoms were selected for throughout our evolutionary history.

Example of alcohol addiction in healthy present-orientated people as potentially representing a rational utility maximizing strategy.

Early experiences may condition exploratory behaviour and the susceptibility to learned helplessness where people become underresponsive to the pay-off probabilities of the current environment.

Can CBT be considered a corrective force ameliorating the disconnect between desired and revealed outcomes and thus eliminating subobtimal utility maximization?

How is normality and optimal functioning currently defined in psychiatry at what might neuroeconomics add to this?

What would psychiatry look like if it were computationally based or in a more extreme scenario if it were based on economic concepts as described above?

What does neuroeconomics bring to psychiatry that it has been lacking? A systematic understanding of motivation, effort and reward may counterbalance the recent focus on executive functioning deficits (e.g. attentional impairment & memory problems).

No comments: