Real vs Placebo Coffee

By Neuroskeptic | November 2, 2009 1:41 pm

Coffee contains caffeine, and as everyone knows, caffeine is a stimulant. We all know how a good cup of coffee wakes you up, makes you more alert, and helps you concentrate – thanks to caffeine.

Or does it? Are the benefits of coffee really due to the caffeine, or are there placebo effects at work? Numerous experiments have tried to answer this question, but a paper published today goes into more detail than most. (It caught my eye just as I was taking my first sip this morning, so I had to blog about it.)

The authors took 60 coffee-loving volunteers and gave them either placebo decaffeinated coffee, or coffee containing 280 mg caffeine. That’s quite a lot, roughly equivalent to three normal cups. 30 minutes later, they attempted a difficult button-pressing task requiring concentration and sustained effort, plus a task involving mashing buttons as fast as possible for a minute.

The catch was that the experimenters lied to the volunteers. Everyone was told that they were getting real coffee. Half of them were told that the coffee would enhance their performance on the tasks, while the other half were told it would impair it. If the placebo effect was at work, these misleading instructions should have affected how the volunteers felt and acted.

Several interesting things happened. First, the caffeine enhanced performance on the cognitive tasks – it wasn’t just a placebo effect. Bear in mind, though, that these people were all regular coffee drinkers who hadn’t drunk any caffeine that day. The benefit could have been a reversal of caffeine withdrawl symptoms.

Second, there was a small effect of expectancy on task performance in the placebo group – but it worked in reverse. People who were told that the coffee would make them do worse actually did better than those who expected the coffee to help them. Presumably, this is because they put in extra effort to try to overcome the supposedly negative effects. This paradoxical placebo response reminds us that there’s more to “the placebo effect” than meets the eye.

Finally, no-one who got the decaf noticed that it didn’t actually contain caffeine, and the volunteer’s ratings of their alertness and mood didn’t differ between the caffeine and placebo groups. So, this suggests that if you were to secretly replace someone’s favorite blend with decaf, they wouldn’t notice – although their performance would nevertheless decline. Bear that in mind when considering pranks to play on colleagues or flatmates.

It looks like science has just confirmed another piece of The Wisdom of Seinfeld:

Elaine: Jerry likes Morning Thunder.
George: Jerry drinks Morning Thunder? Morning Thunder has caffeine in it. Jerry doesn’t drink caffeine.
Elaine: Jerry doesn’t know Morning Thunder has caffeine in it.
George: You don’t tell him?
Elaine: No. And you should see him. Man, he gets all hyper, he doesn’t even know why! He loves it. He walks around going, “God, I feel great!”
Seinfeld, “The Dog”


ResearchBlogging.orgHarrell PT, & Juliano LM (2009). Caffeine expectancies influence the subjective and behavioral effects of caffeine. Psychopharmacology PMID: 19760283

CATEGORIZED UNDER: drugs, papers, placebo
  • Anonymous

    “He loves it. He walks around going, 'God, I feel great!'”

    Not that there's anything wrong with that…

  • Dubito

    Hah! I KNEW it!

    Coffee as a placebo?

  • monicajane

    not quite the topic of the post but related:

    I once was given caffeine when I asked for decaf because I'm sensitive to caffeine and never drink regular coffe. I thought I had gotten decaf…my body knew different…

    felt like I was on speed.

    I went to the counter and they acknowledged they had accidentally given me caffeinated coffee.

    I don't order decaf when I'm out anymore.

  • Neuroskeptic

    Apparently, I once bought a pot of decaf coffee by mistake, instead of my usual, and didn't notice. I only found out about this when someone who saw it later said –

    “You used to drink decaf didn't you?”
    “But there was that time you had a whole pot of it”

  • Mal

    The other day a friend bought me a 20 oz mocha from a coffee place I've never been to. I felt just great drinking it, but then I fell asleep in class anyway. Given that a drink that size usually has 300+ mg caffeine and leaves me going like crazy for many hours, I'm pretty sure the initial alertness was placebo and the drink was decaf.

  • Anonymous

    So, what did they do about the fact that decaf is reduced caffeine, not caffeine-free? (The treatment applied to the beans removes more than half, but certainly far from all.)

  • Captain Skellett

    The placebo effect is certainly more complex than it often appears – that's why it's so good to do double blind trials every time. Your post makes me REALLY want a coffee!

  • Andrew

    Let's call the difference in the enhancement/impairment strata of the placebo group the expectation effect. Why is it only present in the placebo group but not the caffeine group? Not only is there no significant difference between the strata for the caffeine group, but to the extent that there is an expectation effect, it has the opposite sign of the placebo group makes your explanation for it implausible.

  • Ronald Pottol

    But it was not their usual coffee, so they may have attributed the diff to that.

    Consumer Reports did it by adding caffeine to decaf, they found that while no women could tell the difference, 25% of the men could.

  • Neuroskeptic

    Anonymous: They say that:

    “Folgers decaffeinated coffee (22 g) was prepared with
    500 ml of water using a drip coffee maker. Independent
    laboratory analysis indicated that this coffee typically
    contains approximately 11 mg of caffeine per 12 oz serving
    (University of Florida College of Medicine Forensic
    Toxicology Laboratory, Gainesville, Florida, USA).”

    11mg is a very small amount. By comparison a normal 8oz coffee is about 100 mg, a normal tea is about 50mg, and a 12 oz coke is about 40mg.

    I think it's very unlikely it would have more than a trivial effect, if any.

  • Anonymous

    Whats with the argument from Andrew?

  • Neuroskeptic

    He's referring to:

    “there was a small effect of expectancy on task performance in the placebo group – but it worked in reverse. People who were told that the coffee would make them do worse actually did better than those who expected the coffee to help them. Presumably, this is because they put in extra effort to try to overcome the supposedly negative effects.”

    Which is what the authors of the paper say as well. It's true that this doesn't hold for the caffeine group – I'm not sure why this is.

  • Maht

    You'd be surprised (or maybe not) how many people can sense the state of their body.

    For instance see :

    It is a factor one should legislate for in control groups when doing experiments such as these.

  • Graham

    I think the key here is that the participants where regular coffee drinkers. I believe the effects of caffeine diminish after regular daily use. I know, for myself, if I take a week off I am practically vibrating with one cup. I might also guess that regular drinkers are a self selecting group who are less sensitive to caffeine (since they experience less undesirable effects.) Would be interested to see it repeated with a mixed group.

  • jim

    I've noticed that on the (thankfully) few times that I've been given decaf without knowing it I feel a mentally sharpening boost but it doesn't sustain. Presumably the smell, taste, cup-holding, location (etc) all combine for an association effect but without the real stuff it doesn't last. It would be interesting to be hooked up to a pile of body sensors at the time to see how physiologically real the effect is. I'd guess that it's weaker but present.

  • Laurent

    I've always known it! Thanks for confirming my tought!

  • Anonymous

    That instant blast of energy you feel the instant you down your coffee is placebo, as it takes a while to kick in. Thats not to say it isn't real though. Your brain has natural stimulants which it can use to emulate the effects of caffeine.

    There are two placebo effects at work in this study. The decaf groups probably reached a negative conclusion while taking the test (“caffiene has no effect on my performance”) so we see an inverse placebo effect. The caffeine group probably reached a positive conclusion ( “caffiene has an effect on my perfornance”) because they were noticing it's effects

    Its important to have an exit poll (based on your experence today, what do you think this study will conclude?) in order to determine the direction of the placebo effect

  • Anonymous

    A better way of putting it: “do you your results make it more, or less likley that we will conclude blah”

  • Jayarava

    Don’t know about placebo effect but what’s going on with the apostrophes in the comments? They’re all coming out “&#039”

    • Neuroskeptic

      They must have got corrupted when the blog was imported into Discover :(



No brain. No gain.

About Neuroskeptic

Neuroskeptic is a British neuroscientist who takes a skeptical look at his own field, and beyond. His blog offers a look at the latest developments in neuroscience, psychiatry and psychology through a critical lens.


See More

@Neuro_Skeptic on Twitter


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Collapse bottom bar