News for hackers. Stuff that matters.
Re:Beware the Nocebo Effect by mistercow on Friday August 10, 2012 @07:31PM (#4368824)
One really important point about the placebo effect that a lot of people miss is that a lot of it is not a real physiological change caused by the placebo, but is a result of bias in both the subject and the experimenter. Over any period of time, a person will have a great many variations in mood, pain, etc. (which a subject is able to self-examine), and a great many variations of biological function (which a researcher is able to measure). When those variations correspond to effects that a subject or researcher has been primed to expect, the subject or researcher is more likely to notice that effect. This is simple confirmation bias. If you think trained scientists couldn't possibly stumble on that kind of pitfall, you should read about N-rays[1].

This is just as likely to be true for the "nocebo" effect as for the placebo effect. Dizziness, for example, is a very common symptom for people to experience for myriad benign reasons. Orthostatic hypotension (head rush) is a common cause of dizziness, and it is affected by a lot of minor factors like hydration, vasodilation, how long you've been sedentary, how fast you stand up, etc. This may happen to you three times in a week, and you'd never pay it any mind. But if a doctor told you that a medicine you were taking might cause dizziness, you would notice it.

So to this extent, a placebo (or nocebo) effect can be accounted for with absolutely no "mind over matter" style physiological change whatsoever. It is plausible that some specific placebo effects like reduction/increase of pain response, anxiety, and mood could actually be real psychological effects involving suggestion. But if you hear of a placebo doing something and you think "how could a belief in a placebo possibly have that kind of effect", it's very likely that you are at least partly seeing mere measurement error.

[1] http://en.wikipedia.org/wiki/N-rays

Re:Beware the Nocebo Effect by sp332 on Friday August 10, 2012 @08:40PM (#4368983)
That's interesting but I'm sure it doesn't explain everything. For example pain is actually experienced differently (confirmed by watching nerve pathways light up on fMRI scans) depending on people's expectations. http://www.ninds.nih.gov/news_and_events/news_articles/news_...
Re:Beware the Nocebo Effect by mistercow on Friday August 10, 2012 @09:01PM (#4369026)
I did actually cover that in my last paragraph. There are certain effects that are known to be highly subjective and therefore it is not implausible that a real effect from placebos could be seen, even accounting for biases.
Re:Beware the Nocebo Effect by gnaritas on Friday August 10, 2012 @09:40PM (#4369105)
Don't you just hate it when you head off obvious counter arguments and someone comes along and makes it anyway!
Re:Beware the Nocebo Effect by TheComedian on Friday August 10, 2012 @11:55PM (#4369374)
Though I understand that you aren't denying the placebo effect, I'm confused about why you feel that a large amount of the placebo effect is due to bias. After all, most drug trials are double-blinded and placebo controlled, so surely in these trials the placebo effect is legitimate (assuming no scientific fraud).
Re:Beware the Nocebo Effect by Haplo on Saturday August 11, 2012 @03:27AM (#4369665)
The placebo is actually there in double-blinded tests to remove any bias and other factors related to the experiment. And no matter how good/legitimate the experiment, any measurement pretty much always causes changes.

The actual effect of the placebo is usually at most a small portion of all the factors that are measured by it. Of course this depends heavily on the experiment in question and some drugs (like psychological ones) have a higher placebo effect than others.

A lot of people seem to have the idea that the placebo effect is very big for medication outside of experiments - but most of the time it's very small to non-existing.

Re:Beware the Nocebo Effect by Semaphor on Saturday August 11, 2012 @07:38AM (#4369903)
Considering the placebo effect (unless you think bioresonance therapy is true) cured me of several allergies, I'm pretty happy with how strong it is:)
Re:Beware the Nocebo Effect by mistercow on Saturday August 11, 2012 @09:43AM (#4370070)
Or your allergies just went away as you got older, and you attribute that to the placebo effect.
Re:Beware the Nocebo Effect by pygy_ on Saturday August 11, 2012 @11:08AM (#4370228)
There are complex interactions between the nervous and immune systems.

I don't think that anything has been demonstrated regarding allergies, but I wouldn't rule out a therapeutic effect in this case. The absence of evidence is not an evidence of absence.

Re:Beware the Nocebo Effect by Semaphor on Tuesday August 14, 2012 @08:56PM (#4383882)
I must say I honestly don't know if they were allergies in the medical way. For quite some time I couldn't drink large (>1L/week) of milk while milk products were fine. I always called that an allergy and only learned much later that it's not;)
Re:Beware the Nocebo Effect by Semaphor on Tuesday August 14, 2012 @08:25AM (#4380513)
I think the therapy took around 1 month, would have been a pretty sick coincidence:)

I went from not being able to eat anything with wheat or soy lecithin to having no problem with it:)

Re:Beware the Nocebo Effect by mistercow on Saturday August 11, 2012 @09:41AM (#4370067)
> I'm confused about why you feel that a large amount of the placebo effect is due to bias

Two reasons: 1) bias is a sufficient explanation, 2) the prior probability for biases effecting experiments is huge, while the prior for beliefs having a strong effect on physiology (with the exception of highly subjective phenomena like pain and mood) is comparatively quite low.

>so surely in these trials the placebo effect is legitimate

It would superficially seem so, but the key point to remember is that cognitive biases are not something we consciously apply. They are instinctive heuristics that worked really well at helping us survive in the ancestral environment (long before the concept of empirical tests), and even when we know about them, we can't turn them off. Cognitive biases work a lot like optical illusions. You've probably seen the checker shadow illusion[1]. You can understand how the illusion works, and know full well that the two squares are the same color, and you can even watch an animation that proves it to you[2], but when you look at the final image, square A will always look darker than square B. Knowing about the illusion doesn't fix it.

So even placebo controlled studies do not allow us to be unbiased in our perception. What they do is allow us to measure the effects of our biases, so that we can compensate for those effects in our calculations.

[1] http://upload.wikimedia.org/wikipedia/commons/6/60/Grey_squa...

[2] http://upload.wikimedia.org/wikipedia/commons/9/93/Optical_i...

Re:Beware the Nocebo Effect by SeanDav on Friday August 10, 2012 @09:58PM (#4369143)
Not sure I get what you are trying to say here. Are you saying placebo/nocebo effects are real or not?

Using "measurement error" as an argument is kind of insulting to the testers as most of them have got to have a grasp of statistics and other factors in order to publish results that aren't going to get them laughed at by the scientific community.

How do you explain placebo effects in double blind tests?

Re:Beware the Nocebo Effect by mistercow on Friday August 10, 2012 @10:18PM (#4369183)
>Are you saying placebo/nocebo effects are real or not?

It's not a yes or no answer. Part of what we call "placebo effect" is certainly the result of priming and confirmation bias, but for some specific subjective phenomena (like mood), it is plausible that the effect is real.

>Using "measurement error" as an argument is kind of insulting to the testers as most of them have got to have a grasp of statistics and other factors in order to publish results that aren't going to get them laughed at by the scientific community.

>How do you explain placebo effects in double blind tests?

Being aware of cognitive biases doesn't immunize you to them. There isn't even very good evidence that being aware of cognitive biases reduces your susceptibility to them. There is no shame in that, and any scientist who is insulted by the suggestion that s/he is susceptible to biases should go back to Science School. What being aware of biases does allow you to do is design experiments that compensate for those biases.

And that's the whole reason that we have double blind tests. We know that since humans have to administer and participate in experiments, a lot of biasing factors are going to come into play. So we have come up with designs like that which allow us to measure the extent to which the effects we appear to see are due to biases (or the power of suggestion), and how much the effects can be attributed to an actual effect of the drug.

Re:Beware the Nocebo Effect by pessimizer on Friday August 10, 2012 @10:09PM (#4369165)
If those are the only two possible answers that you'll accept, he's not saying anything you will hear.

>Over any period of time, a person will have a great many variations in mood, pain, etc.

>When [...] a subject or researcher has been primed to expect [an effect], the subject or researcher is more likely to notice that effect.

>to this extent, a placebo (or nocebo) effect can be accounted for [by these measurement errors.]

>It is plausible that some specific placebo effects like reduction/increase of pain response, anxiety, and mood could actually be real psychological effects[.]