 So, Shep, Jason and I talk a lot about how our expectations can shape our experiences. Do you know if there are ways that expectations can even shape our physiological experiences? Sure, expectations is effective physiological experiences all the time. If you expect to eat a meal because it's the time of day when you normally eat a meal or you're in an environment in which you normally eat a meal or preparations for cooking are being done, you're secreting insulin, your gallbladder is contracting to eject bile into your gastrointestinal systems. Your physiology is changing in anticipation of a physiologically significant event all the time. Okay, so that's an example from food. Are there any other examples from everyday life where that happens? If you expect to take a drug, you make responses that prepare you for a drug. Just like you make responses in anticipation for food that prepares you for food. Okay, so how does that mechanism work? What do you mean you're expecting and how do you know what to expect from a drug? Well, if you're in an environment in which... I take a common example of somebody who has a cocktail in the evening when they come home from work. Many people report that if on some peculiar occasion they drink the same amount of alcohol at lunchtime, for example, they get more intoxicated than if they drink it in the evening. And that's because the usual cues for the alcohol that cause them to prepare for the alcohol are not present in the afternoon because they haven't normally drunk at that time, whereas in the evening they are present. So the effect of the drug is more pronounced if you don't prepare for it than if you do prepare for it. And that's been seen with many effects of many drugs. But I still don't understand. It's exactly the same. Say we're in a highly controlled situation. I'm taking exactly the same amount of alcohol. How can I get more drunk? It's the same amount of the drug, but it has a completely different effect. If you take a drug, any drug, what it does is engender responses that attenuate the effect of the drug in order to keep you in a normal homeostatic state. So if the drug causes your blood vessel diameter to increase, for example, there'll be parasympathetic activity that causes the blood pressure diameter to decrease. If the drug causes your heart rate to increase, the increased heart rate will initiate homeostatic responses that will decrease. So are you saying homeostatic response? What does that mean? Are you saying your body is easier doing this? We can only survive in an extraordinarily narrow range of physiological parameters. If we get too hot, if we get too cold, if the potassium concentration in our blood gets beyond certain limits, we get sick and then we die. And we have this elaborate homeostatic machinery, which was first elaborated by Claude Bernard and later by Walter Cannon, that acts to keep us that when something happens it disturbs our internal environment, such that in a way that's contrary to what we need to survive, that initiates responses that counter that effect and these are the homeostatic responses. And a drug certainly can be a major physiological insult and the reason why you don't die when you get the drug is because the petrobations that are produced by the drug induce responses that counter the effects of these internal disturbances. Okay, so I can imagine that this same kind of phenomenon would affect things like overdoses. So I think my intuitive idea of what an overdose is, is if I take too much of a drug, if I inject too much heroin than I normally do, then I will die. Is that the way overdoses work? So if you do inject too much of a drug, you'll die. But the interesting fact about overdoses of many drugs, and especially heroin, is that these very drug-experienced and presumably very drug-tolerant individuals die after receiving a dose of the drug that would not be expected to kill them. I'm going to document, for example, that many people buy a drug from a common drug supplier and they all inject about the same amount of the drug and only one individual will suffer the overdose and the others not only don't suffer an overdose but don't see anything particularly unusual about the drug on that occasion. You can find instances in which an individual died after self-administering a dose of the drug that you can demonstrate was a smaller dose than they survived the previous day. So it's not an overdose and this has been known for a long time as the term is usually used. It's some peculiar idiosyncratic reaction to the drug that suffered by the victim. And the reason why it occurs is because they don't make the compensatory response in anticipation of the drug. Why wouldn't they make the compensatory response in anticipation of the drug? It's because on the occasion of the overdose they get the drug in unusual circumstances. This has certainly been demonstrated in several experiments with animals concerning overdose to heroin, pentabarbital, and alcohol in that the drug-experienced rat or mouse that gets the drug in an environment other than that previously paired with the drug is likely to suffer an overdose. It's also been demonstrated with people. Of course you can't do the experiments with people that you can do with animals but there are many, many case reports of individuals who suffer an overdose when they take the drug in circumstances for them which are unusual and that would be expected if they don't make the drug preparatory response. And of course there's one well-known study done in Barcelona where they looked at admissions to a hospital for heroin overdose. If you're admitted to a hospital you're going to survive the overdose because you're given a drug called an opiate antagonist that displaces the opiate molecule from its site of action in the brain and so you survive. So you can question the individual about their drug use on the occasion of the overdose. And so what they looked at in this Spanish hospital were people who were admitted for heroin overdose and survived and they could question them about what the circumstances of drug administration were on the occasion of the overdose but they also had people admitted to the hospital for any of a variety of reasons that people get admitted to the hospital, fights on the street, automobile accidents, other traumatic events having nothing to do with drug use but you can evaluate blood levels of morphine which is what heroin is converted to when it's administered and you have another group of people that have recently taken heroin but were admitted to the hospital at about the same time as the overdose victims but not for heroin overdose for some other reason and the people that were admitted for heroin overdose tended to take the drug in unusual circumstances remember they could be revived by giving the opiate antagonist a question about it and the people with about the same blood levels of opiate who were admitted to the hospital but didn't suffer an opiate overdose and so were tolerant to the drug they took it in their usual place on that occasion. So it seems you need to be careful if you're engaging in this sort of behaviour you might be about to take the same physical amount of a drug but if you're in a new environment or a new situation then it's going to affect you differently so my question would be making no value judgments about whether they should or should be engaging in this behaviour what advice would you have young college students who need to get drunk quickly with the lowest amount of money based on that research? If you want to get drunk quickly then you should drink your beverage in circumstances that for you are unusual and so there was an experiment that was done some years ago that showed that college students get more intoxicated drinking the same amount of alcohol in a novel peppermint flavoured blue beverage than they do if they drink it in a beer flavoured and coloured beverage so just by changing the colour and the flavour of the beverage it causes more intoxication indeed we recently wrote a paper showing that there is one beverage that's widely sold at least in North America called Forloko, F-O-U-R-L-O-K-O that has been implicated in many cases of alcohol poisoning especially among college students and what Forloko is is a beverage that doesn't have very high alcohol content about the same as a bottle of wine about 12% alcohol but it's presented in very novel flavours that in the past have not been paired with alcohol so sort of confectionary flavours, a watermelon flavour or other flavours that apparently defy description but they're novel flavours for alcohol and also perhaps the manufacturers of this beverage realise they know about the effect the effect because they introduced a product some years ago called Forloko XXX Limited Edition and what Forloko XXX does is change the flavour every 4 months so the idea is that if you're used to drinking watermelon Forloko then perhaps you would subsequently get lemon lime Forloko and so even though you would eventually form an association between watermelon flavour and alcohol when you switch to this other flavour it would again have an intoxicated effect So am I right in saying that your body doesn't have a chance to learn about the new flavour and therefore compensate for the effect of alcohol That's right, until you have it on a number of occasions Is there any lab based evidence for the difference you were talking about between experiencing a drug in one situation versus another? Yeah, it's been demonstrated with heroin with a barbiturate pentabarbital and with alcohol described a heroin experiment to you There are a large number of rats that were prepared with intravenous cannula that is they could be administered a drug through a vein without actually piercing the skin with a hypodermic needle all the time and these rats were injected with heroin once every other day on a gradually increasing scale that is a dose would be gradually increasing each day so they built up the tolerance to it So they'd get a small dose on day 1 a somewhat larger dose on day 3 a somewhat larger dose on day 5 On those days when they weren't administered heroin through the vein they were administered an inert substance through the vein A saline It was a sugar solution which was the vehicle for the heroin and on those alternate days when they got the sugar solution it was done in a different environment so there were two different rooms, imagine that differed along several dimensions and imagine on odd numbered days they got heroin in a heroin room and on even numbered days they got saline in a saline room The exact rooms of course account the balance and then on a final test session everybody gets a large dose of heroin but half the rats get the large dose of heroin intravenously in the environment where they previously got heroin in the heroin room the other group gets the large dose of heroin in the saline room in the sugar room, the room that was previously paired with the vehicle Is there any physical difference between those two rooms It's just that they received it in those rooms There's a physical difference in the terms of different sizes, different odors but it's counterbalanced, what was the sugar room for half the rats was the heroin room for the other half the rats So it doesn't matter, the environment in the rooms doesn't matter So prior to that final test session you have two groups of rats that got the same doses of heroin equally often and at the same intervals but the rats that got heroin in the sugar room died as a result of this final dose of heroin and the rats that got heroin in the usual heroin environment tended to survive So it's the same thing as the Barcelona study People that self-administer heroin where they expect heroin tend to survive People that administer heroin where they don't expect to receive heroin tend to die So what's the explanation for that? If you are in an environment that you previously have not gotten heroin you don't make the conditional homeostatic response that's going to attenuate the effect of the drug So your body doesn't prepare for the assault of the drug when you've learned it in one situation versus another It's not just with heroin, it's with anything, with coffee There's a more recent study in which experienced coffee drinkers drank espresso, strong coffee and you measured blood pressure and these experienced coffee drinkers had almost no effect on blood pressure You take these same people and give them the caffeine intravenously and you do it in a way that mimics the blood levels of caffeine that they got when they ingested the espresso which you were measuring all the time So it's a slow intravenous drip designed to mimic the same blood levels of caffeine that they had when they injected the caffeine normally and what you find is that it has a dramatic effect on blood pressure So are these people tolerant to caffeine, these experienced coffee drinkers? It depends, it depends on whether they get it where they expect it that is drinking coffee instead of all the cues that in the past have impaired with caffeine the sight of the coffee, the smell of the coffee, the distinctive temperature of the coffee or do they get it in a way that eliminates these cues that have impaired with caffeine that is presenting it intravenously Interesting, so what would happen if you accidentally took decaffeinated coffee or you took decaffeinated coffee and you weren't expecting it? That's a fascinating question What I would expect, and this is an experiment that I certainly would encourage people to do if you are an experienced coffee drinker, and by the way, what I assert will happen is what people tell me anecdotally do happen and you on some particular occasion get decaffeinated coffee when you normally get caffeinated coffee it should have a soporific effect as you would prepare for the activating effects of the caffeine and there would be no caffeine there and you would have a conditioned response which would be sleepiness So sorry, what do you mean by soporific? You get very sleepy just like caffeine causes you to be alert the anticipation, the reason why the experienced coffee drinker is not hyperactive when I have coffee is because all these cues elicit a response that counter the effects of the coffee If I don't get that caffeine, I just have decaffein Shouldn't I just feel a little bit down? Why would I go to the other end and get sleepy? Well yeah, you'd feel a little bit down in a dramatic way You get sleepy, whatever caffeine produces in the body that counters the effect of caffeine that's what decaffeinated coffee should do So my body is pushing me down expecting a caffeine hit but it doesn't get it and so I don't stay at a normal level I've been pushed down but I don't get the actual drug to pull me back up to the middle That's right, exactly Awesome So your whole line of research reminds me of another sort of idea that's similar and probably related to the placebo effect Can you tell me about the placebo effect? What is it? The placebo effect is interesting both historically and its mechanisms The placebo effect really was publicized as a result of a paper I think in 1955 by somebody named Beecher called the Powerful Placibo and he didn't do any new research but he evaluated existing literature and he showed that for a variety of treatment modalities about a third of the, for a variety of different disease states about a third of the people that don't get treated get better and that's the widely cited statistic where about 30 or 33% of people who get a placebo get better So what do you mean by a placebo? What's a placebo? A placebo, well a placebo can be anything it could be if the treatment is a drug a placebo can be something that looks like the drug that doesn't contain the active ingredient if the treatment consists of the application of a machine like a hearing aid a placebo would be something that looks like that hearing aid Not turned on So how does the placebo effect work? What's the mechanism that allows you to somehow have a physiological reaction like an active drug or substance? Well, when you say have a physiological reaction the placebo effect is based on the patient's report of the feeling better and that's to be distinguished from a physiological reaction although there are occasional isolated reports there are really not good studies that are stood up to replication that say that a placebo can have a physiological effect reducing tumor regression, for example or rather placebo effects are seen primarily in two areas analgesia, that is pain relief the patient reports that they're in less pain and depression, the patient reports it to less depression and these are all kinds of private symptoms they're only available to the patient they're not available to anybody else outside the patient and so the patient has to report of them in terms of public symptoms blood pressure, asthma symptoms and things like that placebo effects are much less pronounced so let me tell you about an interesting study that was done recently where patients who had asthma had a, there's a drug that's effective called albuterol and if patients get albuterol you can demonstrate that the symptoms are released by changes for example in respiratory volume so if you give half your patient well they gave a third of the patients nothing a third of the patients an albuterol inhaler and a third of the patients a placebo inhaler that didn't have the active drug in it and the patients that got the placebo inhaler reported they felt as good as the patients that got the actual active ingredient inhaler and both of them did better than the patients who got nothing but if you actually measure respiratory volume it was only the patients who got the active albuterol inhaler that performed better so if the patient says they feel better that's usually taken as evidence of the placebo effect but often if you measure whether they get better and if you have some physiological index respiratory parameters then you will find that there isn't a placebo effect what does it mean for something to be a real effect though so how would you react to someone who said well even though you feel or you're reporting you have less depressive symptoms you don't really have less depressive symptoms because I gave you a placebo but that seems strange because I do have less I am actually reporting feeling people make mistakes and the mistakes are understandable let's take the case of a physician that has to make a medical diagnosis suppose you come to the doctor and say I have pains in my pelvic area well one possibility is appendicitis and that would be terrible because if the appendix bursts then your life is at risk so you want to remove an inflamed, infected appendix so you go through a lot of diagnostic tests and so on you might say well I never want to have unnecessary surgery so I want to go to a physician that never removes a healthy appendix that might be what you think but that's not what you want because the physician that never removes a healthy appendix simply doesn't have to would always decide that you don't have appendicitis so you might say well I want to go to a physician that only removes diseased appendix well you wouldn't want to do that because that physician is going to make some mistakes in terms of he's going to have to be so sure that the appendix is diseased that there will be some diseased appendices that the physician is going to mix so in fact it turns out that even with the best diagnostic procedures physicians that do appendix surgery are going to do unnecessary surgery 10-20% of the time because they don't want to make the type of error which is to miss a burst exactly there are two types of mistakes and one type of mistake is horrendous that is missing an affected appendix the other type of mistake a false negative the other type of mistake a false positive where the patient doesn't really have appendicitis but you do the surgery well that's pretty bad you're doing unnecessary surgery but it's not as bad as missing an inflamed appendix so we understand the physician can make a false positive errors when considering the costs and benefits of the different types of errors that they make well now the patient is in pain and gets a drug or gets a substance which the physician assures the patient will alleviate the pain well it's just like the physician had a hard time diagnosing the appendicitis the patient has a hard time deciding whether the pain or the depression or whatever is better or not the pain is fluctuating in intensity they have to somehow compare the current level of pain presumably arithmetic mean of the pain in the recent past so it's a very difficult decision to make the patient if the patient in fact gets an inert substance but doesn't know that doesn't know whether they get an inert substance or a placebo then if they make one type of mistake they'd be suppose they actually got an active ingredient and if they said I don't feel better they're disputing the expertise of the physician they might be labelled a complainer they might be prolonging their stay in the hospital so there are pressures on the patient to respond positively so placebo effect for a patient is a false positive response and that is mistakenly saying I feel better when I don't feel better and it is understandable as a false positive response that the clinician makes in diagnosing appendicitis in the patient it's interesting you said before that a placebo can be anything are there any factors that affect the strength of a placebo effect? I guess if the patient is more and more convinced that the substance that they get might be one with active ingredients because for example it's very expensive or it's it's it's it comes from a bottle it looks like an orthodox medicine bottle or a big syringe that's right but I don't want you to think that all placebo effects are simply false positive response in fact most of them and all of them in Beecher's case in 1955 are explained by a much simpler mechanism which is simply regression to the mean that is most people that are sick repoint approach of clinician seek treatment when they're sickest and most people who are sick get better and now we're back to homeostasis because they're homeostatic machinery you're right we'll return them to health and so if you so the people that get the placebo get better just because everybody just because the majority of people who are sick get better and they just happen to get a placebo and it's been pointed out since at least in 1990s in order to really demonstrate a placebo effect you need at least a three arm study where one group gets a drug one group gets a placebo and a third group gets no treatment and you have to find a difference between a placebo group and a no treatment group in order to assert a placebo effect you can't just say that the placebo group got better can you unpack that for me a bit so that sounds like a very complicated experiment or clinical trial a placebo effect of the drug and then of a placebo and then of a no treatment so can you tell me how adding each of those conditions allows you to say make different conclusions about the experiment imagine a group that is sick imagine a large group of people that are somehow equally in pain equally distressed by pain a third of them get a pain medication and most of them get better a third of them get an inert substance and some of them no active ingredient a lot of them are going to get better just because whatever caused the pain healed as a result of the passage of time so is it the group that got the placebo did they get better because everybody because most people who get sick get better or did they get better simply as a result of the passage of time so you have a third group that was in pain and they're told well we're really busy now we're going to have to schedule your appointment for two weeks from now and if during that two week period there's about the same level of symptom relief in the waiting for the next group as in the placebo group then the placebo group probably isn't really a placebo effect it just reflects the fact that time heals so I thought that's surprising to me because my intuition or my common understanding of the placebo effect was that if there is nothing active and you get better then it was a placebo effect but you're saying that that's not necessarily better most of the things that afflict us except the last thing we survived my name is Shep I think about anticipation