Undermining the Placebo Effect

[scepticat]

Apologies if this has been brought up before.

I onetime complained to my doctor that I was experiencing energy crashes after eating simple sugars. I told him after an initial high, I felt somewhat exhausted and ineffective, and I wanted to know if it was normal and if something could be done about this.

He told me that it was my beliefs concerning blood sugar regulation caused me to believe I was experiencing this. He said it was a placebo effect. He said that the belief (that eating simple sugars caused an abrupt rise in blood sugar levels and a corrosponding release of insulin which caused blood sugar to fall to sub fasting levels) could be ruled out because when you give people experiencing such symptoms a drink containing no sugar, they report that they experience the same symptoms.

However, as I recently learned, this is not the entire story. What follows is very interesting:

After you drink a coke, insulin release is triggered. If you do this enough times, it is now known that your body will begin to release insulin before you begin to consume the drink. And if you are given a diet coke in a normal coke can and cannot taste the difference, the same insulin release will occur.

In cases where this phenomenon is possible, extra precautions need to be taken before claims concerning the placebo effect can be made. Many studies in the past should be reexamined accordingly.

 

[Rolfe]

References?

Rolfe.

 

[Acleron]

Interesting if true but I cannot quite see how it undermines the placebo effect.

 

[scepticat]

Sorry guys, I have no written references. Expect me to provide one in the coming days. I learned of this phenomenon in a neuroscience class at the University of Pennsylvania from a respected proffessor. I will ask him where I can find out more.

It undermines the placebo effect in the following circumstances:

If a person reports feeling a certain way after being administered a placebo, it may because they believe the treatment causes a specific reaction and cannot accurately make a self evaluation and thus they imagine things.

On the other hand, they may have been conditioned by prior experiences, and in the insulin example I described, it's not imagined, it's the result of the brain making a prediction and producing a reaction based on previous experiences.

 

[tkingdoll]

On the other hand, they may have been conditioned by prior experiences, and in the insulin example I described, it's not imagined, it's the result of the brain making a prediction and producing a reaction based on previous experiences.

Isn't that what a placebo effect is? Experiences = expectations. It's our expectations that cause the placebo effect, surely? Plus, in order for that experience to exist, then the 'thing' must have at some point had a genuine effect so I'm not sure how a real experience is relevant for most testing. Imagine you're testing homeopathy. In order for someone to get an effect as you describe, they must have experienced real effects from homeopathy in the past.

 

[sthomson]

Isn't that what a placebo effect is? Experiences = expectations. It's our expectations that cause the placebo effect, surely? Plus, in order for that experience to exist, then the 'thing' must have at some point had a genuine effect so I'm not sure how a real experience is relevant for most testing. Imagine you're testing homeopathy. In order for someone to get an effect as you describe, they must have experienced real effects from homeopathy in the past.

The Placebo Effect is many different things. One of those things is certainly conditioning like skepticat describes.

 

[fls]

There is a training effect that has been demonstrated in several conditions. For example, if you relieve pain with morphine and at some point in the course of treatment substitute saline, the pain will be relieved by the saline, but not if you administer an opioid antagonist (which would block the action of morphine and endogenous (produced by the body) opioids). However, if you initially use non-opioid analgesics, then the pain-relieving effect of the saline is not blocked by opioid antagonists. These conditioning effects have also been studied in Parkinson's disease and the immunosuppression from cyclosporin A. They do wear off, and they are affected by expectation.

I'm not sure what is meant by 'undermine' though.

Linda

 

[fls]

Scepticat,

Do you mean it undermines our ability to distinguish placebo effects from treatment effects?

Linda

 

[Ivor the Engineer]

I knew I'd read about this somewhere before.

From: Neuroscience: Exploring the Brain 2nd ed. p.536/7

Insulin release by the pancreas is controlled in a number of ways. Consider your pancake breakfast. During the cephalic phase, when you are anticipating food, the parasympathetic innervation of the pancreas (delivered by the vagus nerve) stimulates the beta cells to release insulin. In response, blood glucose levels fall slightly, and this change, detected by the neurons of the brain, increases your drive to eat, in part by activation of the NPY neurons of the arcuate nucleus.

 

[scepticat]

Linda,

I mean that it undermines our ability to claim that a treatment works only because of the placebo effect rather than by any valid physiological mechanism.

In other words saying "I could give you a drink containing no sugar and you would give the same report" doesn't mean that I imagined the effect in the first place.

 

[Dancing David]

There are lots of ways the placebo effect works, they are not limited to what you doctor told you about drinking soda.

Confirmation bias and return to the mean are another two ways that it works.

 

[Jeff Corey]

Apologies if this has been brought up before.

I onetime complained to my doctor that I was experiencing energy crashes after eating simple sugars. I told him after an initial high, I felt somewhat exhausted and ineffective, and I wanted to know if it was normal and if something could be done about this.

He told me that it was my beliefs concerning blood sugar regulation caused me to believe I was experiencing this. He said it was a placebo effect. He said that the belief (that eating simple sugars caused an abrupt rise in blood sugar levels and a corrosponding release of insulin which caused blood sugar to fall to sub fasting levels) could be ruled out because when you give people experiencing such symptoms a drink containing no sugar, they report that they experience the same symptoms.

However, as I recently learned, this is not the entire story. What follows is very interesting:

After you drink a coke, insulin release is triggered. If you do this enough times, it is now known that your body will begin to release insulin before you begin to consume the drink. And if you are given a diet coke in a normal coke can and cannot taste the difference, the same insulin release will occur.

In cases where this phenomenon is possible, extra precautions need to be taken before claims concerning the placebo effect can be made. Many studies in the past should be reexamined accordingly.

Yes, it appears to be a classical conditioning phenomenon. One of my profs in grad school had shown that saccharine triggered insulin release (or at least the change in blood sugar) for his dissertation. Classical conditioning is a more parsimonious explanation than "belief".

 

[fls]

Linda,

I mean that it undermines our ability to claim that a treatment works only because of the placebo effect rather than by any valid physiological mechanism.

How so? The mechanism is not at all specific to the placebo, and is not relevant to the vast majority of cases where a placebo group is present. Whether or not there is a physiologic effect depends upon training, not the placebo. And most studies that use a placebo group do not involve an opportunity for training to take place.

In other words saying "I could give you a drink containing no sugar and you would give the same report" doesn't mean that I imagined the effect in the first place.

The problem with your original example is that falls in blood sugar do not present as "exhausted and ineffective". Regardless of whether or not you were able to document that you trained yourself to release insulin, it can't be assumed to account for the way you feel. I realize that the "sugar crash" has entered popular lore, but it hasn't been found to correspond to a physiologic effect, let alone a physiologic effect that can be subject to training.

Linda

 

[Jeff Corey]

There are lots of ways the placebo effect works, they are not limited to what you doctor told you about drinking soda.

Confirmation bias and return to the mean are another two ways that it works.

I find it useful to distinguish between those terms. For example, if you wished to demonstrate the placebo effect, you would have to control for statistical regression to the mean and other threats to internal validity.

 

[Mojo]

After you drink a coke, insulin release is triggered. If you do this enough times, it is now known that your body will begin to release insulin before you begin to consume the drink. And if you are given a diet coke in a normal coke can and cannot taste the difference, the same insulin release will occur.

This sounds like the cephalic phase effect thingy that Kumar tried to propose as a mechanism for homoeopathy.

 

[fls]

I find it useful to distinguish between those terms. For example, if you wished to demonstrate the placebo effect, you would have to control for statistical regression to the mean and other threats to internal validity.

Especially if what you wish to study are effects specific to the placebo, such as those brought up in this thread. I also think it has been confusing to lump together what have turned out to be quite disparate effects.

Linda

 

[sol invictus]

Bear in mind that if you experience a reaction of any kind there must be a low-level physiological explanation for it. The fact that you know the explanation in this case does not mean it isn't an example of the placebo effect.

Here are two relevant features of this example:

1) It's your belief that the drink is normal coke that stimulates the reaction. If you were given a glass of water, or a glass of clear sugar syrup with the same sugar content as coke, the reaction would not occur (at least not the part of it that comes before ingesting the drink).

2) There is a conditioning element here - you've had coke before, and it's reasonable to suppose your body has a specific conditioned response to the sugar in it. The classic example of a placebo is when you react to a neutral pill that you're told contains a drug - but not one you've ever had before. In that case your body can't have been conditioned, at least not in any specific way, so it's a cleaner example.

Personally, given 1), I would say that despite 2) this is an example of the placebo effect, because it would not occur were it not for your beliefs about the drink. The conditioning just muddies the waters a little. However I don't know what the accepted definition is in psychology/medicine.

 

[Cuddles]

Surely this is just a plain old Pavlov response? Ring bell and then feed dog enough times and it starts salivating whenever it hears the bell. See coke and get more sugar in blood and you produce insulin whenever you open a can. Nothing new here at all.

 

[Jeff Corey]

Surely this is just a plain old Pavlov response? Ring bell and then feed dog enough times and it starts salivating whenever it hears the bell. See coke and get more sugar in blood and you produce insulin whenever you open a can. Nothing new here at all.

That's what we were talking about when we referred to "classical conditioning". It's also called respondent or Pavlovian conditioning.

 

[sol invictus]

Right - but it's not so clear how to distinguish that from the placebo effect.

Suppose you take aspirin regularly to kill headaches. Then one day someone secretly swaps your aspirin for some inactive pills, but when you take one your headache still goes away. Placebo effect, right?

But why did your headache go away? What was the specific physiological mechanism responsible? Maybe it's something at the level of brain chemistry, maybe it's not understood yet, but it's there. And it may well be a type of Pavlovian conditioned response.