Reward Prediction Error At The Gas Pump

A couple of years ago I was driving up to Minnesota resort country when I noticed something happened at gas stations. For decades, gas pump choices were arranged linearly with the lowest octane fuel on the left and the highest on the right.  The only difference was the occasional pump with diesel or racing fuel options and they were typically on the far right.  To illustrate I took this photo of a gas pump display on the way home form work yesterday.  Three years ago the display would have been (left to right) 87-88-91.  In this case, in addition to the octane shuffle there is a price dissociation.  Motorists have been trained for year to expect that 87 octane to be the least expensive and now it is not - 88 octane is the least expensive.  These changes at the pump reminded me of a paper I had read a few years ago by Wolfram Schultz one of the world experts in reward prediction error. 

Reward prediction error is basically the difference between a prediction about the nature of the reward and what really happens. See all of the definitions in the diagram below that are taken directly from Schultz.  Reward is clearly defined as well as predictions to both the positive and negative sides.

In the paper of interest Dr. Schultz introduces it by being confronted with a vending machine where he cannot read the language (Japanese).  There are 6 choices and his expectation is low that he will get his preferred choice so he pushes the second button.  He is surprised that it delivers exactly what he wants - black current juice. He points out that this experience will keep him pressing this second button until the machine is loaded differently and he does not get his expected juice.  He uses this as an example of positive reward prediction error (RPE) or the difference between his low expectation and the ideal outcome that resulted.  RPE systems are set up to optimize positive reward prediction error.  When the eventual negative outcome results in negative prediction error decision making will change in order to return to the positive RPE scenario and then no reward prediction error scenarios.

Getting back to the gas pumps. They appear to have been designed to defeat RPE at two levels.  First, the position of the buttons after everyone was conditioned to push the one at the far left.  The second, is the dissociation of gas price from button position.  There is no longer a linear correlation between octane and price.  I am not sure if both of these trends have been occurring over time or just recently. I do know from studying various gas pumps that the button positions do not necessarily reflect linear price or octane changes, but in some cases they still do. The good news is that they can both be overcome by carefully studying the octane button position and posted per gallon price rather than depending on the old learned patterns.  The other interesting aspect of the gas pump problem is that there don't have to be a lot of predictions.  The prediction error occurs only if the purchaser depends on old patterns without paying close attention to details. 

The neurobiology of RPE is more fascinating than the descriptive aspects. We know that the neurobiology of reward in the human brain is heavily dependent on dopaminergic systems in the ventral striatum.  Dopaminergic neurons code reward in the form of prediction error even  in complex tasks.  What is even more interesting  is that the coding is not in terms of quantifiable measures but subjective ones.  Utility functions incorporate subjective measures and can be used to determine the potential values of the reward.  Dopaminergic neurons code the utility of the received reward minus the utility of the predicted reward.  Looking at a computational model of the addictive process, Redish (7) discusses a value function V[s(t)] dependent on the state of the world s(t) and presents it as the calculation of expected future reward discounted by the expected time of reward:

As noted the value of delayed rewards are reduced and the actual discounting applied is based on empirical work on discounting in human and animal models.  The author in this case goes on to develop a computational model of addiction in this case that is based on reward prediction error and the fact that cocaine produces direct phasic increases in dopamine (DA).  The model is termed a temporal-difference reinforcement learning (TDRL) model.  According to RPE increases in DA occur after unexpected natural rewards. Over time DA release decreases, learning stops, and instead is paired with the cues for the reward.  That does not occur with a pharmacological reaction at the level of the dopaminergic neurons.  In that case, a drug like cocaine will release DA independent of the expected reward.  That produces a final state where a unexpected natural reward, a cue for a learned reward, and cocaine will all produce DA.

The Redish paper also looks in detail at a couple of associated issues. The first is rational addiction theory defined as the user maximizing value or utility over time. Long term rewards for quitting are discounted more than the short term penalties and therefore the user remains addicted.  In the author's model "the maximized function entails remaining addicted." (p 1946).  TDRL theory suggests that addiction is always irrational because the pharmacological effects of cocaine (in this case) always outweigh the associated DA surges from the universe of value functions available in the real world. Addictive drugs will produce an increase in DA, so that the user will not be able to encounter and learn a value function that is associated with an equal or greater DA surge than is produced by the drug.  Therefore the user remains addicted.  This has been taught in addiction seminars for years as the Hijacked Brain Hypothesis - meaning that the dopamine signal produced by addictive drugs overwhelms the dopamine signal produced by natural stimuli like eating, drinking water, sexual behavior, and social affiliation.  Both RPE and TDRL theory offer more explanatory power than the Hijacked Brain Hypothesis. 

As I wrap up this post, I pulled down the latest editions of the two major addiction texts to see what they had to say about reward prediction error, computational models of addiction, and some of the authors referenced in this post, especially Wolfram Schultz.  There were no references at all and the sections on the actual function of DA neurons in addiction was surprisingly thin. On the other hand a lot of concepts used in the field like salience are the direct product of these systems. In order to produce a more coherent picture of the neurobiology of addiction it is important to outline these DA systems and how they work normally and in addictive states.

I am hoping that addiction texts for clinicians will contain some of this information in the near future and ideally the chapters will be written by the scientists that have been studying these processes in some cases for decades.


George Dawson, MD, DFAPA

Supplementary 1:

Getting back to the gas pump example, considering the 3 octane ratings and the three prices that may or may not correspond to the octane ratings means that there are 6 possible combinations at any pump that need to be considered.  Any real world actor at the pump needs to consider this carefully when the gas pump has undergone a transition from the expected correlation between increasing octane ratings and price to one where this relationship does not exist. The advantage to the actor in this case is that all of this information is explicit and that behavior is more likely to be affected by negative prediction error when automatic selection behavior results in the wrong octane or fuel cost being selected.  That is unlike Dr. Schultz's example in the Japanese airport when he randomly chose a beverage and was unexpectedly rewarded.


Supplementary 2:

I can't say enough about the writings of Wolfram Schultz.  They are only peripherally mentioned in the addiction literature and yet his theories and experiments are some of the more important that I have read with regard to the neurobiological theories of addiction.

Papers of Wolfram Schultz - Journal of Neurophysiology Page Link

Home Page of Wolfram Schultz Link - contain some of the best PowerPoint slides that I have ever seen.

References:




1: Schultz W. Dopamine reward prediction error coding. Dialogues Clin Neurosci.2016 Mar;18(1):23-32. Review. PubMed PMID: 27069377 full text

2: Schultz W. Reward prediction error. Curr Biol. 2017 May 22;27(10):R369-R371.doi: 10.1016/j.cub.2017.02.064. PubMed PMID: 28535383 full text

3: Stauffer WR. The biological and behavioral computations that influence dopamine responses. Curr Opin Neurobiol. 2018 Apr;49:123-131. doi: 10.1016/j.conb.2018.02.005. Epub 2018 Mar 2. Review. PubMed PMID: 29505948. full text

4: Takahashi YK, Batchelor HM, Liu B, Khanna A, Morales M, Schoenbaum G. DopamineNeurons Respond to Errors in the Prediction of Sensory Features of Expected Rewards. Neuron. 2017 Sep 13;95(6):1395-1405.e3. doi: 10.1016/j.neuron.2017.08.025. PubMed PMID: 28910622. full text

5: Keiflin R, Pribut HJ, Shah NB, Janak PH. Ventral Tegmental Dopamine Neurons Participate in Reward Identity Predictions. Curr Biol. 2019 Jan 7;29(1):93-103.e3. doi: 10.1016/j.cub.2018.11.050. Epub 2018 Dec 20. PubMed PMID: 30581025.

6: Tobler PN, Fiorillo CD, Schultz W. Adaptive coding of reward value by dopamine neurons. Science. 2005 Mar 11;307(5715):1642-5. PubMed PMID: 15761155.

7: Redish AD. Addiction as a computational process gone awry. Science. 2004 Dec10;306(5703):1944-7. PubMed PMID: 15591205.

8: Sweis BM, Thomas MJ, Redish AD. Beyond simple tests of value: measuring addiction as a heterogeneous disease of computation-specific valuation processes. Learn Mem. 2018 Aug 16;25(9):501-512. doi: 10.1101/lm.047795.118. Print 2018 Sep. PubMed PMID: 30115772.  

Policy Makers Are Always The Weakest Link In Healthcare

When it comes to solutions to the opioid epidemic - talk is cheap. The last 20 years everybody has “the solution”. The AMA came up with a new version of theirs entitled “AMA Opioid Task Force Recommendations for Policymakers.”  Inspection of this page shows that it is basically a rehash of everything we have known over the past 10 years or longer. The only new message is that the AMA is now suggesting that policymakers should follow these recommendations. In this era of patient empowerment, direct advice to patients is completely missing.

Drug legalization advocates have apparently vilified the Reagan era “Just Say No” campaign to the point that attempts at primary prevention of substance use are now politically incorrect and forbidden. How can you possibly stop opioid and methamphetamine epidemics when there is a large and vocal advocacy for legalizing all drugs emboldened by the cannabis campaign? There are few reasonable voices out there saying “You know you can really live a better life without drugs or alcohol”.

It should probably come as no surprise that real action on the drug epidemic cannot be expected from a government that is unable to end a decades long streak of mass shootings. We hear the familiar refrain that people were “in the wrong place at the wrong time” or that they are “fallen heroes” and that “now is the time to move on”. A real adaptive response to mass shooting like even slightly tougher gun laws would make a difference. Despite hearing that “this is the last time that our children can be victimized” the mass shooting saga drags on - courtesy of local and federal governments.

The resolution of the opioid epidemic is another example of how our government doesn’t work on serious public health issues.  The epidemic has been in place for the past 20 years.  Using deaths by overdose as a proxy measure suggests that things may be improving the last couple of years, but the epidemic is far from resolved.  The more recent problem has been that people who started using prescription opioids have changed to heroin or fentanyl – the supplies of both are plentiful and less expensive than the street value of typically prescribed opioid pain tablets.

A few words about the points the AMA has in their graphic:

1. MAT - medication assisted treatment for opioid use disorder is considered a major advance in treatment. That applies both to methadone maintenance treatment and more recently buprenorphine maintenance with various preparations. Sustained-release naltrexone injections are also an option but they are more controversial due to the longer induction and wait time until the patient is safely covered by opioid receptor antagonism. The current AMA position is to remove prior authorization from these treatments so that they are more readily available. Some treatments are more cost-effective than others. It is not clear from the statement how the AMA hopes to remove these barriers particularly since they have not been effective in removing them for the past 30 years of utilization management or prior authorization. They may be counting on political leverage in this case but I don’t see it happening. Regulators and politicians could easily make this an exception to the current utilization management and prior authorization statutes that they have on the books but it should be apparent from that statement that they are the problem in the first place.

2. Mental health - the document cites the well-known correlation between mental illness and substance use. The document also cites the Mental Health Parity Addiction and Equity Act (MHPAEA) as meaningful but the only way this law gets enforced is if civil action is brought against healthcare companies. These healthcare companies are protected by legislation and they basically do whatever they want. The AMA Task Force suggests that healthcare company should be “held accountable” but that hasn’t happened in the 10 years since the MHPAEA has been passed.  The document suggests that a number of addiction specialists should be in the networks of these healthcare providers, but for 20 years politicians have been rationing mental health services to the point that county jails are currently our largest psychiatric institutions. The mental health suggestion in this document seems like another wish.

3. Comprehensive pain care and rehabilitation access - I would really like to see the numbers on this one. If anything there has been a tremendous proliferation of freestanding or chains of pain clinics over the past 20 years. That proliferation correlates directly with increasing opioid prescriptions. As far as I can tell there has been no movement at all in terms of determining what constitutes a quality pain clinic versus something else. This may have to do with the politics that wrung the word “quality” out of the healthcare system 30 years ago. There is also an access problem. In other words there has always been “non-opioid alternatives” like physical therapy but healthcare systems ration their utilization.  This might be another area where education is important and convincing people that a course of physical therapy even if their healthcare company makes them pay for it is potentially more beneficial than taking opioids and getting deconditioned for a period of time.

4.  Maternal and child health - there is no doubt that punishment-based paradigms can intrude on the parental relationships with children and result in destabilization of families. This usually occurs on a county by county basis and there are no statewide standards and no specific treatment facilities. The problem is compounded by the fact that most states consider social services to be as expendable as mental health services and it takes more than a suggestion to reverse that 20-year trend.  Recently, the child protection issue as a result of substance use has become so bad that additional tax legislation is needed just to cover this problem.

5. Civil and criminal justice reforms - the most significant reform suggested in this section is that MAT is continued when a person is incarcerated and after they are released. This is a tall order considering how difficult it is for anyone to access MAT in an outpatient setting. Jails and prisons have the absolute worst record. The evidence for that is people who are acutely taken off of methadone, buprenorphine, or other psychiatric medications at the time of incarceration. That can lead to weeks of opioid withdrawal symptoms and intense physical symptoms.  Despite many county jails considering themselves to be psychiatric hospitals very few of these places are equipped to assess and treat psychiatric disorders or do medication assisted treatment of substance use disorders.

That is the AMA WishList and all of its deficiencies. I have not seen a realistic assessment of the problem and how to reverse it in spite of the fact that there are two documented opioid epidemics in the medical literature and suggestions about how they were resolved. I never heard anyone referencing them. Medication assisted treatment was one component but there are other significant factors that no one seems to be talking about at this time.

Working in a residential treatment facility provides me with unique perspective on the problem. The continuum of care ranging from residential treatment to intensive outpatient treatment to date treatment to self-help groups like Alcoholics Anonymous and Narcotics Anonymous depends on a number of factors to make it work. First and foremost is a competent staff in the facility with reasonable boundaries and a supportive environment. Most medical facilities do not have this because of significant bias against people with substance use disorders. There are some treatment facilities that have similar biases and they should not be allowed to admit people until that problem is resolved. The measures recommended by the AMA Task Force are medically weighted and that means that treatment facilities need to have medical staff. If the facility needs histories and physicals done medical staff need to provide that function as well as comprehensive detoxification, treating associated medical problems, and providing psychiatric care and MAT. There is no point in having residential or outpatient treatment programs in a network if they cannot provide that level of care. People who need MAT should not be treated in facilities where they cannot get medical assessment and treatment.

That basic fact seems to be missing from the AMA Task Force guidelines, state regulations, and any discussion at the federal level about what kind of treatment is needed for people with active opioid use disorders.

The AMA could be of more service referring people to appropriately staffed treatment programs and advising the public on the source of all of these obstacles of care. As I have been writing here for years now those obstacles are a product of pro-business government policy at both the state and federal level and how those rationing businesses are able to operate. Until that basic flaw is corrected - I do not anticipate any increase in access to treatment (at least effective treatment), increased access to appropriate social services, or sudden revision of county jails to suddenly make them functional psychiatric units.

There are some changes that would make an immediate difference in the opioid epidemic instead of the continued evidence-based platitudes.  If there are any policy makers or politicians out here that are serious about making some changes - here they are:

 1:  Hold physicians harmless for providing MAT:

The suggestion that more physicians should be providing MAT for opioid use disorder has gone from a suggestion to more of a demand.  Just this weekend there have been debates about why Emergency Department Physicians aren't providing MAT for every person with OUD that they see.  My first thought when I saw that was: "Are they serious?" People are not presenting to EDs with casual use.  They are not people coming into clinic intentionally in withdrawal to start buprenorphine induction. They are generally people with very serious use problems who end up in EDs because of a different problem. Many of them are polysubstance users with multiple drugs on board and in many cases drugs that are typically flagged as having potentially serious interactions with buprenorphine.  Add to that the dearth of buprenorphine prescribers that will accept referrals from an ED and it makes perfect sense that Emergency Medicine physicians do not want to send people out with buprenorphine.

The physicians are not the problem, the practice environment is.  The solutions seem obvious to me.  The first is to indemnify the physicians for providing care that is harm reduction to patients with high risk. This already happens in state statutes that cover Good Samaritan provisions, mandatory reporting of child and adult protection concerns, and civil commitment and guardianship proceedings that hold the petitioners harmless for good faith activity.  MAT is a very similar endeavor. But I would not just stop at a vague statutory requirement. I would tie it in with abbreviated training for MAT.  When I took that training, at least half of the patient case examples were high risk with limited resources, psychiatric comorbidity, and they were using high levels of multiple substances.  The answer in each of these scenarios was to prescribe buprenorphine as a way to assist the patient with the OUD aspect of the problem. 

2:  Open up addiction clinics:

The idea that primary care physicians are all going to start seeing large volumes of these patients will not materialize as long as there is a problem with cross coverage.  I have seen it happen many times. A well intended physician starts prescribing buprenorphine and even in a mutli-specialty clinic has nobody else to assist and is on-call 24/7 for years until they burn out.  There has to be a structure in place where there are clinics that can handle large volumes of patients including the referrals from all of the local EDs and correctional facilities and provide adequate cross coverage for the physicians prescribing buprenorphine. 

3:  Decrease the training requirement:    

Unlike others - I don't think it can be eliminated for the reason I cited above.  The physicians and other prescribers need to know the high risk scenarios that they can treat.  I think it could probably be done in two hours with a case book of treatment scenarios.  The case can be made for collaborative care/mentoring arrangements with experienced physicians, but the funding of those scenarios should be seriously considered.   

4:  Provide temporary housing programs to take people directly from the ED and crisis appointments: 

As a former acute care psychiatrist - I know the uneasy feeling of providing brief opioid detox services and discharging patients with OUD to the street with medications that have street value.  There is no surer path to immediate relapse.  If we are really serious about helping people get established on MAT, they need a stable environment where it can happen. 

5:  States need to license substance use programs only if they provide medical services and MAT:   

If we are all serious about the effects of MAT in OUD it is time to start acting like it.  There is no longer an excuse or reason for not offering MAT to all patients in residential, extended care, or outpatient treatment programs.  There are no religious or ideological grounds that justify not offering these services and the license of all treatment facilities should depend on it.

These are my ideas about stopping the opioid epidemic that stop all of the platitudes in their tracks.  There is a rational way to proceed that does not depend on physicians sacrificing to keep the irrational system afloat. The rational way will cost money, but it will also save money but not in the way politicians usually talk about healthcare savings. It will save money and resources by saving lives, not investing in inadequate treatment, and finally putting a dent in the large circulating pool of opioid and polysubstance users that are circulating between emergency departments, inpatient units, drug treatment programs without MAT, detox units, shelters, and jails.    

George Dawson, MD, DFAPA

Medical Cannabis Does Not Prevent Opioid Overdoses

The political aspects of medical cannabis are undeniable. The legalization of cannabis for recreational purposes had no traction with American politicians or voters until it was promoted as a miracle drug.  With that widespread promotion medical cannabis is now legal in 33 states and recreational cannabis is legal in ten.  The legalization arguments also suggested that the US was behind other countries of the world when there are only two countries – Canada and Uruguay – where it is completely legal for medical or recreational sale and purchase.  In the world, 22 of 195 countries have legalized medical cannabis with widely varying restrictions on its use. The Netherlands is often cited as an example of recreational cannabis use, but most Americans don’t realize that it is illegal for recreational use and tolerated for use and sale only in specially licensed coffee shops.  The promotion of cannabis as a solution to the opioid overuse and chronic pain problems can be seen as an extension of the political arguments for legalization that outpace any science to back them up.

There was probably no greater hype about the purported benefits of medical cannabis than early data suggesting that it might decrease the rate of opioid overdoses (1). The sequence of events was supposed to be opioid users tapering off of opioids or using lower equivalent amounts because of medical cannabis use.  The original study covered the time period from 1999-2010 and suggested that states with medical cannabis laws had a lower mean opioid overdose mortality and that the annual rates of overdose progressively decreased over time.  The authors conclusion was:  “Medical cannabis laws are associated with significantly lower state-level opioid overdose mortality rates.”

Despite the usual caveats suggested by the authors in the original study the results of that study were heavily hyped by all cannabis promoters as was the discussion of many Internet forums.  The lay press, public, and politicians saw it as another reason to promote medical cannabis and recreational cannabis by association.

A study came out today in PNAS (2), that is an extension of the original data and it no longer comes to the same conclusion.  In this new study the authors replicated the opioid mortality estimates from the original study but when the data was extended from 2010 to 2017 – the improved opioid overdose mortality rates not only did not stay constant but they reversed themselves to that they were now on the average from -21% to +23%.  They provide an even more valuable analysis of this effect as spurious rather than a true positive or negative effect based on the low penetration of medical cannabis in the population at large (2.5%).  The authors focus on the problem of ecological fallacy – that is conclusions about individuals are drawn from aggregate data across the entire population.They point out that the states with the medical cannabis laws have a number of characteristics separating them from other states.  A recent good example of this fallacy was the New England Journal of Medicine (3,4) report that per capita chocolate consumption correlated with the number of Nobel Laureates in a particular country.  

This is a valuable lesson in scientific analysis. The political approach to the problem is all that most of the public sees. That approach is to grab any information that seems to agree with your viewpoint and run with it.  Big Cannabis and cannabis promoters have been doing this for almost 20 years now. The process of science on the other hand is slower and more deliberate.  It is not a question of a right answer but a dialogue that hopefully produces the right pathway. The authors of this study have added a lot to the dialogue about cannabis but also statistics and how statistical descriptions may not be what they seem to be. 

George Dawson, MD, DFAPA

References:

1: Bachhuber MA, Saloner B, Cunningham CO, Barry CL. Medical Cannabis Laws and Opioid Analgesic Overdose Mortality in the United States, 1999-2010. JAMA Intern Med. 2014;174(10):1668–1673. doi:10.1001/jamainternmed.2014.4005 (full text)

2:  Shover CL, Davis CS, Gordon SC, Humphreys K.    Association between medical cannabis laws and opioid overdose mortality has reversed over time.  First published June 10, 2019 https://doi.org/10.1073/pnas.1903434116  (full text)

3: Messerli FH. Chocolate consumption, cognitive function, and Nobel laureates. NEngl J Med. 2012 Oct 18;367(16):1562-4. doi: 10.1056/NEJMon1211064. Epub 2012 Oct 10. PubMed PMID: 23050509.

4:  Pierre Maurage, Alexandre Heeren, Mauro Pesenti, Does Chocolate Consumption Really Boost Nobel Award Chances? The Peril of Over-Interpreting Correlations in Health Studies, The Journal of Nutrition, Volume 143, Issue 6, June 2013, Pages 931–933, https://doi.org/10.3945/jn.113.174813


Attribution:

Above figure is from the original article (reference 2): "This open access article is distributed under Creative Commons Attribution-Non Commercial No Derivatives License 4.0 (CC BY-NC-ND).y"  See this link for full conditions of this license.

Spare The Venlafaxine.....

Venlafaxine is a commonly prescribed second-generation antidepressant. It is well-known to psychiatrists because it is a second line medication if SSRIs fail and for many psychiatrists it is another first-line antidepressant. In some head-to-head comparisons with SSRIs venlafaxine has a more favorable side effect profile. It does have the risk of discontinuation symptoms and typical antidepressant side effects. I have noticed that the dose escalation with venlafaxine seems to be out of proportion with SSRIs, bupropion, and third-generation antidepressants.

Consider the following venlafaxine related scenarios:

1. A colleague comes into my office late in the day and asks me: “Have you ever heard of venlafaxine causing sedation at higher doses?” The patient in question was just increased from 187.5 mg to 225 mg - the suggested max dose according to the FDA approved package insert. 

2. I am asked to consult on patient who had extensive pharmacogenomic testing in a different facility where she was told that she may need to take 350 to 450 mg of venlafaxine per day based on that genetic profile. She wants to make sure that she gets an adequate dose of the antidepressant and is currently on 225 mg.  I reviewed the limitations of that approach with the patient and potential side effects and I let her know that the commonest side effect I see in people taking high-dose venlafaxine is excessive sedation or low energy in the daytime. As we start to follow the recommended dose increase she discloses that she has had sedation even at the 225 mg level. We decreased the dose to 150 mg and that side effect is gone.  Her depression is also gone.

3. I see a significant number of patients taking more than 300 mg per day of venlafaxine from the same geographic location in the United States. They all tell me that the target dose in that location is 350 mg per day and they are all experiencing numerous side effects. Many had dose escalations into that range in a week or two - much faster than any increase I have done.

What is wrong with this picture? Why are there a significant number of people taking more than the recommended dose of venlafaxine in some cases much more and appearing to have side effects? The roots of this prescribing behavior can be traced back to old-school psychopharmacology. Proponents of that approach suggests that there may always be a group of outliers that need to take higher-than-expected doses of medications - typically antidepressants but there has also been a history of excessive dosing of antipsychotic medications. People were generally more cautious with more toxic medications like tricyclic antidepressants, monoamine oxidase inhibitors, lithium, and various addictive compounds. They also seem to be more cautious with SSRI type medications at least initially. It took over a decade for me to see a dose of sertraline in excess of the maximum recommended dose.  While it is true that there are always outliers in terms of dose-response what is the best way to approach that problem.

I have attended medical education courses where the lecturer suggested titrating the medication to the point of toxicity and then reducing it back down to the next lowest dose. That particular lecture was focused on treating anxiety disorders with SSRIs. I don’t think that is the best approach. The best approach to me is one where the patient recovers from anxiety or depression and the process does not experience a single side effect. I know that can be done because I have been doing it for decades.

That also brings me to what I think is a good research article that looks at optimal dosage ranges. It is a very large meta-analysis of fixed dose randomized clinical trials that utilize the specific antidepressants - citalopram, escitalopram, fluoxetine, paroxetine, sertraline, venlafaxine, and mirtazapine.  The trials were identified by searching the literature and looking for unpublished studies specifically by searching national drug licensing agencies and requests directly to pharmaceutical manufacturers. Outcomes were noted at eight weeks of treatment and defined as a 50% reduction on an observer rated scale for depression.  Dose equivalence among medications was determined from previous studies and the recommendations of the manufacturer. The article is written by researchers that I consider to be world experts in meta-analyses and the analysis of large data sets in psychiatry.

77 studies were identified from a total of 24,524 published references and 4030 unpublished records.  27 were published, 21 or unpublished, and 29 were both published and unpublished. The study showed too hard when treatment groups across all of the medications of interest between the years 1986 and 2013.

The authors calculated dose response, dropouts due to adverse effects, and dropouts for any reason. Relative risks (RRs) were calculated for specific doses. The dose outcome relationships for venlafaxine are included in the figures below from the original article.  The Response figure shows the significant increase of up to about 150 mg and then a much more modest increase beyond that. The Dropout figure shows a similar increase up to the 150 mg range. The Dropout for any reason was less remarkable. The authors calculated that the 75-150 mg dose of venlafaxine was equivalent to 20 to 40 mg of fluoxetine (click to enlarge graphic)

The authors conclude that optimal acceptability of SSRIs and venlafaxine and and mirtazapine occurs within the lower end of the licensed dose range. They reconcile this with serotonin transporter (SERT) studies that show that 80% SERT occupancy occurs at the minimum doses of SSRIs or venlafaxine with further dose increases showing small increase in SERT occupancy.

In the case of venlafaxine they suggest that noradrenalin reuptake transporter (NET) may require higher doses of venlafaxine in the 225 mg to 375 mg per day range. Given the lack of efficacy of atomoxetine, a logical question might be whether NET blockade adds much to the antidepressant effect.

The authors review other dose-efficacy studies of antidepressants and point out that they are variable. The variability ranges from optimal doses of fluoxetine in the 21-40 mg per day range to doses at the recommended lower end of the range being superior. Response to doses in the higher range were variable in some studies. One study found a significant greater response for high-dose antidepressants but the dose of 40-50 mg fluoxetine equivalents showed the greatest efficacy.

The authors considered strengths and limitations their study. They thought that their state-of-the-art meta-analysis was a strength as well as the size of the data set. They also examined dose dependency for both efficacy and tolerability and acceptability. The limitations they discussed included patient selection and dosing not reflecting clinical practice. They also discussed the calculation of dose equivalency among antidepressants and how that might be problematic.

Another obvious strength of this study is the calculation of relative risks for response across SSRIs, venlafaxine, and mirtazapine. The figures are modest but favor antidepressants across all dosage ranges with the exception of mirtazapine at the 60 mg dose.  The authors don’t seem to mention it but it would seem that the optimal dosage ranges could be predicted from the regulatory information since that is based on dose ranging studies and tolerability studies. In that regard, the conclusion about dose ranges don’t seem to be that surprising but they may be needed given what is happening clinically.

Getting back to the issue with venlafaxine I see people respond to dosing within the lower and of the range from 37.5 to 75 mg in many cases. That same response rate continues up to the 150 mg dose and then starts to diminish between two or 25 and 375 mg. Over that same range there is a significant increase in dropout rates due to adverse effects.

How clinicians approach this new information will be interesting. There will still be people like me and the conservative camp looking for the first signs of side effects and toxicity and deciding at that point whether to stop dose escalation. I explicitly tell people that the goal is not to experience any side effects and that I doubt that people “get used to” side effects. There are clearly clinicians out there who are doing exactly the opposite and that is increasing the dose of venlafaxine and advising people to either tolerate the side effects or expect that they will go away.

The balance between therapeutic effect and side effects is a central issue in all branches of medicine. In many cases, the severity of adverse effects like an allergic reaction determines the decision. In the case of the medication like venlafaxine making that decision can be complex. Some of the side effects like sedation and lethargy at high doses can mimic symptoms of depression. At this point in time neither pharmacogenomics or most plasma level determinations guarantees either tolerability or efficacy.  

Detailed analysis of the situation by an expert with a bias toward preventing side effects is required as the first step in any dose increase.

George Dawson, MD, DFAPA

References:

1:  Furukawa TA, Cipriani A, Cowen PJ, Leucht S, Egger M, Salanti G.  Optimal dose of selective serotonin reuptake inhibitors, venlafaxine, and mirtazapine in major depression: a systematic review and dose-response meta-analysis.  Published:June 06, 2019DOI:  https://doi.org/10.1016/S2215-0366(19)30217-2.

Attribution:

Above figure of the venlafaxine dose response and drop outs are directly from the paper in reference 1 and used per the Creative Commons Attribution 4.0 International Public License.

Attribution 4.0 International (CC BY 4.0)

Attribution 4.0 International (CC BY 4.0)

Chemical Imbalance As Advertising Meme

After a protracted discussion on the previous post, I thought I would go down to the University of Minnesota Biomed library today and look at the drug ads in psychiatric journals at about the time Prozac came out in 1987. I was interested in the trends before and after so I picked the years 1985 to 1995. I also picked the journals the American Journal of Psychiatry, Archives of General Psychiatry (currently JAMA Psychiatry), and the Journal of Clinical Psychiatry.  I was going to include JAMA and the New England Journal of Medicine.  They had about the same number of ads but none of them in that year contained ads for psychiatric medications.

This kind of search is labor intensive these days. There was a time on the early days of the Internet when entire journals with all of the ads were scanned in. As a subscriber I could have run that search from home.  These days, all of the ads are gone and the references are saved as text files only. In order to see historical ads - the hard copy of the original journal needs to be examined. Even then there are some problems.  I encountered some bound volumes where the ads were physically removed. There were two to three bound volumes per year and additional copies of the NEJM and JAMA - I may have looked at 75 bound volumes over 4 hours.

In many ways it was a walk down memory lane.  Clozaril and Haldol Decanoate ads were especially heavy in the early 1990s.  There were ads for medications that I prescribed all of the time like Navane and Pamelor and ads for drugs that I seldom prescribed like Stelazine, Serzone and Luvox. There were ads for new drugs that I would prescribe once like Paxil.  It was a reminder that despite all of the advertising - a  lot of drugs end up never being prescribed by physicians.  My reason for being there was to look for the origins of the term "chemical imbalance" in this advertising.

I decided to embark on this project because of all of the inaccuracy about the term, especially the tendency to blame psychiatrists for it. In my previous post, I attempted to point out that it is a fairly straightforward process to conclude that the human brain does not run on chemical imbalances - just based on the average scientific knowledge of physicians. On the advertising side, I was there for the first National Depression Screening Day in 1990 and that was the first time I heard the term. The event has been criticized as a venue for allowing a pharmaceutical company to showcase their product.  I participated in the event for 3 years and the advertising involved was much more subtle than is found today at NAMI walks for example. But the question is whether the advertising meme "chemical imbalance" was introduced at that time. Any event that happened 30 years ago is very hard to track. As the Public Affairs Rep for my District Branch of the APA, I had a lot of files about it that I subsequently trashed. I am guessing there were also some files on disk drives that would have been helpful.  This is a reconstruction without that data.

I successfully located the first Prozac ad in AJP from 1988.  The graphics are all iPhone photos so there is some distortion.  Chemistry is emphasized on page one as in the chemical structure, chemically unrelated to other antidepressants, distinctive chemistry, and the first highly specific and highly potent blocker of serotonin reuptake.

Why is this important? At the time most of the antidepressants being used were tricyclic antidepressants.  They could not claim any specificity and in subsequent ads manufacturers start to compare possible side effects based on transporter monoamine protein and receptor affinities. The Prozac molecule was being hyped as being chemically unique and with a better side effect profile. As Prozac started to sell more it became a blockbuster drug for Eli Lilly and at that point the manufacturers of other new antidepressants noted and the competition heated up.  There were some direct references to Prozac in the ads from competitors.

The best example is this Wellbutrin ad from AJP in 1991. Prozac is directly mentioned in the ad and reasons are given for choosing Wellbutrin over Prozac.  Being non-serotonergic is one of them and this is more of a counter to Prozac advertising as being a unique first highly selective serotonergic drug.  It gives little or no weight gain as a reason, but at the time I was seeing obese patients who were taking 80 mg of Prozac because their primary care physicians told them they could lose weight taking it. Of the other bullet points it seems that lack of sexual dysfunction would be  the most relevant. The marketing decision in this case was a conscious decision to go after the purported serotonergic effects of Prozac rather discuss the hypothetical mechanism of Wellbutrin.  The side effect of Wellbutrin that most physicians are concerned about - seizures - is in the smaller print below the bullet points.

Effexor came up with similar ads.  In the late 1980s and early 1990s, synaptosome technology was invented to look at binding affinities of central nervous system medications to specific receptor sites.  The quantitative aspects of these studies were generally globalized in the psychiatric literature to qualitative ballpark effects.  For example a plus or minus grading system could be used ranging from no effect at a receptor (-) to a robust effect (++++).  Effexor advertising used this to compare side effect profiles among the competitive antidepressants at the time.

This ad emphasizes that Effexor is "a structurally novel antidepressant and is chemically unrelated to any other available antidepressant."  It shows the table with comparisons to tricyclic antidepressants (TCAs) and selective serotonin reuptake inhibitors (SSRIs) and what might be predicted based on the in vitro synaptosome data with the qualifier that the clinical significance of that data is unknown.  Clinically most people are able to tolerate all three classes of medication but some will not.  The differences can't be predicted on the basis of the receptor binding studies because of receptor heterogeneity and differences in drug metabolism.  For example, I still prescribe TCAs. It is nortriptyline and it is the only one I have ever prescribed. At the doses I prescribe and per the table in the ad - it is as well tolerated as SSRIs and SNRIs (Effexor).  The ad appeared in the AJP in April of 1994.  On that basis the argument could be made that it is an appeal to the technical expertise of psychiatrists and it should contain this information.  That also points to a weakness in my informal advertising study and that is a lack of ads from the non-technical consumer literature from the same period. (see supplementary on a proposal).

I have 30 additional ads from the journals but the themes are roughly the same. An emphasis on medicinal chemistry and the suggestions that some chemistry is better than others. Interestingly, in my previous post the whole point was that this is the kind of argument that would not fly based on what the average physician knows about chemistry and molecular biology. Psychiatrists should know a lot more because the evidence for and against these theories had been reviewed in the psychiatric literature 20 years before these ads came out (1974-2002) (1).  And they are engaged in clinical practice and need to be skeptical of newly introduced products and claims.

What I did find so far is unequivocal evidence that the chemical imbalance meme was used to directly market antidepressants to the public.  The Zoloft ad embedded at the top of this page from 2001 is the first example.  The second example is this Paxil ad from the same year.

That is what I have so far.  See the Supplementary below to find out what you can do to complete the story. I don't have a problem with people telling me that their doctor told them that they have a chemical imbalance and their antidepressant is supposed to treat that. I don't have a problem with people saying that their psychiatrist told them that.  I do have a problem with people saying that all or even most psychiatrists say this and that psychiatrists are behind this meme.

There is an exaggerated focus on the mechanism of action of medications used for psychiatric indications. I have never heard anyone say their doctor told them about the mechanism of action of antibiotics or even their blood pressure medications. In the case of antibiotics it is clear that people demand them and they don't care what the risks or mechanisms are.  This advertising campaign may have something to do with the conversion of folk psychologists to folk psychopharmacologists.  A friend of mine also brought up an important aspect of this campaign that is also addressed by these manufacturers and that is legitimacy. For decades people with depression and anxiety were viewed as weak people with a questionable problem.  My friend told me that these ads confirmed that she had a serious problem that needed a serious solution and that it was finally acceptable to talk about it.  Say whatever you want about Big Pharma advertising but it apparently carried the message that current "Let's Talk About Mental Health" programs do - but over 15 years ago.

The attribution of an advertising meme to psychiatry and psychiatrists despite the fact it has never appeared in 30 years of psychopharmacology texts is not a trivial fact.  The advertising videos posted here  were viewed by tens of millions of people.  I hope to get more information and still have some people to contact. With any luck I will be able to fill in the additional data between the release date of Prozac in 1987 and the ads posted here from 2001.

Please send me anything you might have from those dates.



George Dawson, MD, DFAPA


References:

1.  Nathan KI, Schatzberg AF. Mood disorders. in Review of Psychiatry, vol 13. American Psychiatric Press, Washington DC(1994): p.171-184


Supplementary 1:

From the  information I posted above it is clear that chemical imbalance was an advertising meme introduced during the height of competition of blockbuster  antidepressant drugs.  The  common Wall Street definition of a blockbuster pharmaceutical is a product that generates sales of a billion dollars a year.  There are two important pieces of data that would be useful to complete the story.

The first is earlier ads with the term chemical imbalance. So far, I have two from 2001, but I am certain it appeared before that. I don't have time to search all of the popular literature.  If you subscribe to a magazine that has pharmaceutical advertising and keep all of the old volumes - take a look at the editions from about 1987 to 1995. If you see the term chemical imbalance please send me the image with the name and date of the periodical.  Let me know if you want credit for finding the image and I will give you full credit.

If you are a current or former pharmaceutical rep or marketing person and have access to any documents or videos with the chemical imbalance phrase please send it to me with the date it was being used. If you have recollections of how it was implemented and when I can also use that information but I am most interested in clear documentation like the videos I have posted. I have no interest in vilifying the pharmaceutical industry and understand the need for marketing and advertising.  I am just interested in the origins of this term and how it was implemented.

If you are an APA member and you were involved in the original National Depression Screening Day in 1990 - you may also have some information about this.  Please send it to me.

Thanks!


Supplementary 2:

All of the name brand drugs/medications mentioned in this post are currently generics or are no longer manufactured.  I have no affiliation with the original manufacturers or the generic drug industry.


Supplementary 3:

There are various Internet sites that attribute the term chemical imbalance to Pfizer or Lilly but they do not appear to be reliable - many appear to be antipsychiatry sites.  I would like to hear from people who were there at the time and can provide the necessary proof.  In those days (1986-1996) it would have been an internal memo or presentation.  Send me a copy if you have it.


Supplementary 4:

I had the opportunity to discuss this issue with a corporate attorney - especially the issue of available emails and memoranda dating back to 1987.  He told me that corporations hold this data only as long as the law states they needs to.  For example, if the law states the data must be held for 4 years it will be held exactly that long and then everything will be shredded.  If this information exists it will probably be in private hands.

Supplementary 5:

I got the expected low level feedback from a Twitter poster who thought he was making some point about this link on the Royal College of Psychiatrists web site suggesting that at least one of the causes of schizoaffective disorder was "a chemical imbalance".  I guess he really thought he had made me look foolish especially with the proclamation "You aren't psychiatry - they are."

https://www.rcpsych.ac.uk/mental-health/problems-disorders/schizoaffective-disorder

In fact, I can't tell who wrote this and whether or not it is a psychiatrist. I don't know what the RCP official position is.  I was happy to see that they are much more flexible than the anti-psychiatry Twitter posters I encounter.  There was a feedback form that I completed and advised them to lose the "chemical imbalance" and that replacing it with "unknown etiology" was preferable. What I would like to see is an exposition of the latest theories and a suggestion that the critics actually read psychiatric literature.  They would be less likely to perseverate the same criticism they have used for year after year. This poster also seemed to ignore the fact that the RCP public information was posted in 2015 - that's 14 years after the television ad posted at the top of this page.  Royal College of Psychiatrists - the ball is in your court.

Supplementary 6: (added on 1/11/2020):  I just learned today from an advertising expert in antidepressants that there was also a Zoloft ad from 2004 that used the term: 

"While the cause is unknown, Zoloft can help.  It works to correct a chemical imbalance in the brain that may be related to these symptoms."

Reference:

Cristina Hanganu-Bresch. Treat Her with Prozac: Four Decades of Direct-to-Physician Antidepressant Advertising in Drugs Media: New Perspectives on Communication, Consumption, and Consciousness (Hardback) (1st Edition) by Robert C. Macdougall (Editor), Drugs &. Media-Pasta Dura, 340 Pages, Published 2011 ISBN-10: 1-4411-1988-4 / 1441119884 ISBN-13: 978-1-4411-1988-9 / 9781441119889:

The Non-Existent Chemical Imbalance Theory

I keep looking for it and can never find it.  The above picture is my stack of psychopharmacology texts dating back to about 1980 and none of them mentions "chemical imbalance".  I could add another foot or two to that stack and there still would be no mention of this theory.

Why is that important? The main reason is that one of the favorite arguments by anti-psychiatrists is that real psychiatrists believe that psychiatric disorders are caused by a “chemical imbalance” in the brain. This criticism showed up on this blog several years ago in a post that I critiqued that was largely a screed against psychiatrists. Accusing psychiatrists of promoting a chemical imbalance theory is an almost perfect rhetorical strategy. It uses what essentially was a marketing device for antidepressants in the late 1980s to portray psychiatrists as excessively reductionist at the minimum and at the worst biologically naïve and dishonest.

My colleague Ron Pies, MD has written a recent piece on the historical, philosophical, and rhetorical aspects of this argument. What I hope to accomplish in this post is taking a look at the science behind why no psychiatrist would consider the brain to be a substrate run by “chemical imbalances”. Some might find this argument to be quite boring but I can attest to the fact that the premises used allowed me to state unequivocally to the first pharmaceutical rep to use the term that no such state exists in the brain.

The main factor has to do with how physicians are trained. There’s still a lot of confusion about whether a psychiatrist is a physician or not. I can assure anyone reading this that we all are. That means in order to get into medical school certain prerequisites at the undergraduate level have to be completed. That includes a year of general chemistry, a year of organic chemistry, and a year of general physics. A significant number of psychiatrists that I have encountered were chemistry majors. That training means that physicians in general have had exposure to physical science and how chemistry works in solutions and gases.  In these basic two or three component systems there are limited possibilities in terms of reaction outcomes. Even electrochemical reactions produce electron flow that decays predictably over time but that is not able to transmit any nuanced signal.  In other words the information content in these systems is low – too low to run biological organisms.

In the basic science years of medical school biochemistry, neuroanatomy, neurophysiology, pharmacology, and all of the associated molecular biology provided medical framework that all of the physical science can be mapped onto. The study of enzyme and receptor systems highlight the basic concept that the chemistry involved can only occur because it is in a specific microenvironment. That microenvironment includes the protein structure of the enzyme or receptor molecules as well as associated membrane components and cell signaling components. The intracellular and extracellular environments are exquisitely controlled as is the synaptic cleft. Many of the reactions involve additional acid-base and ionic gradients. The degrees of freedom in these many component and many phase systems are large. They are so large in fact that I have been unable to find an estimate of degrees of freedom for neurobiological systems.

A good example of the kind of microenvironments and complex interactions that I am taking about is the GABA_A receptor depicted diagrammatically below. The GABA_A receptor is a transmembrane cylindrical receptor that is a member of the pentameric ligand-gated ion channel superfamily.   The diagram is a top down view of the receptor complex cylinder highlighting that it is composed of 5 glycoprotein subunits.   Each subunit is composed of 4 domains with one domain that lines the chloride ion channel through the center of the receptor complex. Binding sites on these protein allow for allosteric modification of the cylindrical receptor to facilitate chloride ion influx and fast inhibition of neuronal signals.  Allosteric modulation of enzymes and receptors occurs when a molecule reversibly binds to the protein molecule resulting in inhibition or stimulation of the overall process.  For example, benzodiazepines bind to a specific site at the α-γ interface leading to increased affinity for GABA at the receptor sites and increased chloride ion influx. Benzodiazepines are the classic allosteric modulators of the  GABA_A receptor but there are others.  Barbiturates, anesthetic agents, neurosteroids and ethanol are also allosteric modulators at the GABA_A receptor.  The detailed structure of both the benzodiazepine and flumazenil binding sites on the human synaptic GABA_A receptor have only recently been detailed (1). 

The above paragraph is a glance into the types of systems that modern psychiatry is focused on.  In the case of the GABA_A receptor global inhibitory effects can be expected at some point, but there are not the product of chemicals floating about inside the body or brain. They are the effects of complex interactions between proteins, positive and negative modulators, neurotransmitter effects, ion fluxes, and additional signaling.  The effects result from where these receptors are located in the brain and central nervous system. The education of physicians assures that this level of complexity in the brain is appreciated as both the basis for normal physiology and also the basis for pharmacology and toxicology. It may be tempting to try to simplify things - but real brain function defies simplification.  The basic working of the GABA receptor was discovered when I was in medical school back in the 1980s. The lectures in those days showed a simple structure with an arrow showing increased chloride ion permeability but nowhere near the structure that we currently have. 

This is one set of receptors and modulators very simplified. To get more of the story read the 22 pages of reference 1.  To understand the brain and modern pharmacology much more needs to be understood. Forgetting about the term "chemical imbalance" is a good first step.

George Dawson, MD, DFAPA

References:

1: Zhu S, Noviello CM, Teng J, Walsh RM Jr, Kim JJ, Hibbs RE. Structure of a human synaptic GABA(A) receptor. Nature. 2018 Jul;559(7712):67-72. doi: 10.1038/s41586-018-0255-3. Epub 2018 Jun 27. PubMed PMID: 29950725; PubMed Central PMCID: PMC6220708.

2:  Human GABA-A receptor alpha1-beta2-gamma2 subtype in complex with GABA and flumazenil, conformation A.  Detailed structure from the above paper.