Many people take caffeine always, or never. But the evidence is clear: for some tasks, drink coffee -- for others, don't.
Use  caffeine for short-term performance on a focused task (such as an exam).
Avoid  caffeine for tasks that require broad creativity and long-term learning.
(Disclaimer: The greater altertness, larger short-term memory capacity, and eased recall might make the memories you do make of higher quality.)
At least, this is my take. But the issue is convoluted enough that I'm unsure. What do you think?
New Comment
109 comments, sorted by Click to highlight new comments since:
Some comments are truncated due to high volume. (⌘F to expand all)Change truncation settings

I like my women the way I like my coffee: detrimental to hippocampal neurogenesis, but conducive to short term memory and attentional control.

You, steven0461, are my hero of the day. Thanks for your brililant wit, may I quote you?
Hey steven, you have been quoted on TV tropes now, the single most prestigious recognition anyone can receive. can I have your autograph?

Memory formation and memory retrieval are very different tasks, so one should be specific when making claims like "Caffeine helps long term memory." For example, if caffeine only hinders long term memory formation, but not retrieval, then this would suggest using it during an exam, but not while studying. If vice versa, then vice versa.

Unfortunately for our purposes, the authors of your first article have blurred this distinction in their abstract, no doubt because it was not the subject of their study: their method was to add caffeine to rats' water supplies, without controlling the timing of the doses in relation to the events of formation and retrieval.

I was happy to find your last article addresses precisely this question:

Groups of 12 adult male Wistar rats receiving caffeine (0.3-30 mg/kg, ip, in 0.1 ml/100 g body weight) administered 30 min before training, immediately after training, or 30 min before the test session were tested ... Post-training administration of caffeine improved memory retention at the doses of 0.3-10 mg/kg ... but not at the dose of 30 mg/kg. Pre-test caffeine administration also caused a small increase in memory retrieval .... In contrast

... (read more)
I would hesitate under any circumstances to take a dose of 30 mg/kg.

What do you think?

I'm happy to free-ride off of your opinion. Relative to the other things I'm working on, deciding whether to drink coffee or not seems like a fairly minor optimization. And doing my own research when I've already got this (seemingly solid) analysis of yours is an even smaller optimization.


If I had a name like yours i'd watch out for caffeine bias ;)

It's sad to see such denial, and yet so humbling. ;P (PROTIP: Maxwell House)
xamdam nice!

Re: "Use caffeine for short-term performance on a focused task (such as an exam)."

Really? What about state-specific memory? If you are intoxicated by caffeine during an exam, don't you need to be taking it during the revision process as well?

Sure, I can imagine caffeine impeding long-term learning from exam revision. But I find the increased focus to be much more important, for an exam that I've already studied for, and for material I will very likely never need to know in quite as much detail ever again. There's 2 different kinds of studying I do. Studying conceptually for the long term, and cram time for a specific exam fitting in all the fine details, and then quickly regurgitating them. If it takes exam revision to significantly enhance the former, then I already learned too little, too late. That said, I commonly use caffeine for the latter with no regret of the side effects.
Tim's point is that things that you learn while on a specific substance (and so in a specific brain state) are actually more accessible while in that same state. So if you study without caffeine then perform with caffeine you don't access the memories as well as if you had studied with caffeine too. The reverse applies as well. As you say, the focus benefit you get probably gives you a net benefit from the caffeine regardless but state specific learning is worth considering.

It's not clear to me exactly what types of learning or memory cause the hippocampus to grow, even after reading this article. I don't even think they have a clue whether there's a training effect such that the added neurons make it generally easier to process new inputs for learning/memory - that is, they're seeing a correlation and speculating that there's causation in both directions.

However, I'm reminded that severe stress (see Sapolsky's "Zebras") causes the hippocampus to shrink (and the amygdala, which is apparently involved in fear, to gr... (read more)

I wonder how much of the beneficial effects of coffee are exactly the effects you would get from stress. Stress here being the fight-or-flight response. The theory is that the body diverts resources to make sure you survive (presumably, to your muscles and your short-term memory and executive function) and away from long-term maintenance (reproductive function, immune system, long-term memory formation). I've read somewhere that a component of sleepiness is modulated by amount of cortisol in the body. According to wikipedia, coffee stimulates production of cortisone. Maybe this is evolution adapting the fight or flight response to the circadian rhythm, which explains why they are related. Perhaps in the case of diminished hippocampal neurogenesis it's just from stress. And maybe the beneficial effects from coffee are too.

Caffeine, of course, is rather addictive.

So one might (and I do) find it difficult to optimize finely according to what tasks one is attempting. The addictive nature of the drug probably explains the "always or never" consumption pattern.

Personally I drink 2-3 espressos every other day or so and find it easy to maintain that level. In fact, I wish I were more addicted to caffeine, because I recently realized that I can write better with it, hands down. I would sometimes go for a week or two without drinking much caffeine because I thought it was unnecessary and slightly hurt my stomach, but now I regret neglecting it.

I cycle caffeine and ephedrine to avoid the withdrawal and dependence effects. Of course, I always combine Ephedrine and caffeine on important test days. There seems to be a multiplier effect. The improved focus is especially important on long tests, which generally become a battle of attrition.

I also used the EC stack as a weight-loss tool with great success. It powerfully wards off hunger, and I simply forget to eat meals when I'm using EC.

I asked my wife, an ICU nurse, about this combination of drugs as a concentration enhancer/weight loss tool. According to her, use of ephedrine in anything but a medical emergency is a really bad idea. Wrong dosing/accelerated uptake due to other factors could mean a heart attack really really easily. Probably you already know this, but I can't assume. She also told me that, if you are going to choose a strong stimulant, there are a lot better choices than ephedrine (sorry, forgot specifics).
Do we have numbers here? Any statistical evidence that it's worse than related but patentable Sudafed?
None of the drugs mentioned in this thread are still under patent. The only relevant current patent is on "minus pseudoephedrine," but that never reached market. Also, I think you're confusing (ephedrine, pseudoephedrine) with (pseudoephedrine, phenylephrine). (ETA: and phenylephrine isn't under patent, either)
Nope. I personally don't care at all, strong black tea is as far as I go with self-medication, but I thought knb & others here ought to know that there are risks. According to Wikipedia, Health Canada recalled all ephedrine-related weight loss/bodybuilding meds in 2002. That's enough to give me pause. Minimax, people!
Minimax? I don't think a decision, however wise, to avoid self-medication can be described as minimaxing.
I guess it doesn't strictly fit the criteria. I mean that it involves taking what seems to me a substantial risk for a small-to-medium, temporary increase in productivity.
There are many things stronger than black tea that can be used without substantial risk. Some of them will serve to better optimise the goals that are being subject to minimaxing. Avoiding all such options out of conservatism isn't necessarily a bad idea and it is certainly what can be described as 'appropriate' or 'sensible' but it is not minimaxing.
Yep, good point. I am aware that ephedrine causes heart problems in some people. I really should have mentioned that. However those outcomes are very rare, ephedrine has been consumed for centuries (as ephedra tea) with a good safety record. I wouldn't ever take more than 25 mg of ephedrine, especially if you take it with caffeine. However, ephedrine is very commonly used by bodybuilders (I was actually turned onto ephedrine by bodybuilders at my gym) who want to cut weight, and other athletes. Dangerous outcomes usually happen in people who take ephedrine before vigorous exercise. Ephedrine before exercise is extremely common, but I wouldn't recommend it for most people.
Do you have reason for ignoring the general medical wisdom that methylphendidate and amphetamine are safer and more effective strong stimulants than ephedrine? Based on your comparison of test taking to a battle of attrition, I don't think you'd have any problem getting a diagnosis of ADHD-PI. Have you already tried methylphenidate and amphetamine and concluded they are less effective for you than caffeine+ephedra?
Well, we do have good data on the value of general medical wisdom in this community, no?
I presume knb has done enough research to know how not to give himself ten times the safe dose.
Yeah... I suspect that knb would be much better off with a prescription for methylphenidate. Based on knb comparing test taking to a battle of attrition, which is not a normal comparison, I don't think knb would have any trouble getting a diagnosis for ADHD-PI. I suspect that small doses of methylphenidate are more effective than caffeine in most scenarios, but most people here are unable to do that comparison without going out and getting a diagnosis of ADHD.
NB This works best with caffeine tablets not coffee or pop, which, when combined with Ephedrine can cause an annoying diuretic effect. A 200 lbs man only needs 50mg caffeine and 25 Ephedrine to keep him going, btw. NB: If people can't find Ephedrine, then the dirty sister substitute SudaFed (make sure it has psuedoephedrine in it) will serve in a pinch when you are cramming/testing. Funny how their slogan is "UNBLOCK YOUR HEAD". Their caps not mine.

What do you think?

Your summary roughly matches my own research and is confirmed, for what it is worth, by my own anecdotal experience.

I'll note that similar (but stronger) effects can be expected from the more direct stimulants (amphatamine, methamphetamine, etc.).

ETA: Regarding attention control, be aware that sometimes increased 'attention control' comes in the form of increased focus on the immediate task which can actually reduce the ability to switch tasks smoothly. This can affect the balance of attention you place on social details relative to task details when the situation at hand requires both.

There are conflicting issues though. There are studies (that I read years ago, and have no link to) that show that consistency is better... that consistent low-level caffeine drinkers are more alert than their non-caffeine colleagues, but less jittery than high-caffeine people (optimum seemed to be 2-3 cups per day).
Associated with that would be method of consumption: concentrated does (espresso) v. sipping american coffee over an afternoon. Using is in a "targeted" manner might fail you: If you are not particularly used to the effects and sudde... (read more)

The sensitivity to irregular caffeine users is just due to lack of tolerance. It can still be avoided by just reducing the dosage compared to regular caffeine users.

Is there a good source of documentation for the expected side-effects of coffee - the "down" period of reduced mental capacity that often occurs a few hours after drinking, the effect of disrupted sleep cycles for those who do not normally drink the stuff, etc...?

Here's a recent Lifehacker article covering some of the same ground as well as some other stuff, including withdrawal: What Caffeine Actually Does to Your Brain

Interesting article, though I notice the top featured comments claims it's "missing essentially every piece of work done in the last FIFTEEN years." Nonetheless it doesn't give any reason to think that the downsides of caffeine that the article points out aren't real.

Although the evidence is far from conclusive, regular caffeine consumption may have neuroprotective effects, perhaps more likely among women and older people.

  1. Ritchie K, Carrière I, de Mendonca A, Portet F, Dartigues JF, Rouaud O, Barberger-Gateau P, Ancelin ML. The neuroprotective effects of caffeine: a prospective population study (the Three City Study). Neurology. 2007 Aug 7;69(6):536-45.

  2. Rosso A, Mossey J, Lippa CF. Caffeine: neuroprotective functions in cognition and Alzheimer's disease. Am J Alzheimers Dis Other Demen. 2008 Oct-Nov;23(5):417-22.

... (read more)

I have sort of been doing what you recommend anyway, because I found needing coffee to be normally awake (like my father does) absolutely unacceptable. I like coffee, but I never drink it unless there is a specific reason to. The same for tea, and I almost never drink coke anyway because I prefer the taste of most other non-alcoholic beverages, including cold tape-water.

But I never noticed any effect on me, so recently I haven't bothered drinking coffee for anything but social reasons ( a cup every two weeks or so). Maybe I'll try it again next time. Or ... (read more)

It's funny, but I've never noticed a caffeine effect from tea. No matter how much I drink. Coffee, on the other hand, can have me feeling bouncy and downright high after a single cup. What could explain this? Do I just suck at steeping my tea long enough, or something? Anyway, I agree with your approach. Caffeine is physiologically addictive, and if you use it often enough, it stops being a fun bonus and becomes something you need just to feel normal, and that's a ridiculous state of affairs.
Tea has other alkaloids beyond caffeine. Probably the most notable is theanine, which is relaxing.
Kevin, You got it. Tea has L-Theanine in it, which keeps you from getting the coffee jitters. I just got myself some L-Theanine pills to test out how it works on my concentration.
I take some Theanine pills every now and then (Theanine Serene). I tried them originally to help with occasional bouts of insomnia. They didn't help with that at all, but they definitely do relax me in the rare moments I get a bit stressed from overworking.
Wonderful stuff, theanine. I converted to green tea extract as my preferred caffeine source once I learned green tea is a good source. (Then I bought bulk theanine powder and explicitly chose my dose!)
May I propose an experiment (and report back) involving two big strong mugs of coffee, one decaff, one not, and both unfamiliar brands so you can't reliably taste the difference?
I actually wasn't expecting any effect from the coffee. I hardly ever drink coffee, so I haven't formed a mental association between the taste and smell and increased alertness. It's hard to say without a blind test, of course, but it's still puzzling to me.
I like the experiment but note that the findings would be confounded somewhat. Heavy coffee drinkers have been found to get an increased performance on tasks requiring alertness just from the smell! They get another boost once the caffeine actually gets digested and flows to the brain but it may be difficult to establish a negative finding if both the decaff and caffeinated seem to give an effect.
Probably both with lots of cream and sugar is a more reliable way of masking any taste difference.
Agreed... but we want to test how much his reaction to his coffee is dependent on the experience of drinking coffee versus it's actual caffeine content, so I'd suggest he takes it reasonably close to how he normally takes it, to get the same trigger. We already know he doesn't react to a caffeinated drink (his tea) that doesn't taste like his usual coffee. In other words, if he normally takes it black with no sugar, lots of cream and sugar might break the experiment.
For tea to be awakening it's important to remove the leaves from the cup after a few minutes. If you leave the leaves in the tea other substances go into the tea that counteract the caffeine effect.
Do you have a citation for this claim?
At the moment I don't find a reputable english source however, I can however elaborate the argument a bit more: Tea contains both caffeine and theanine. Theanine relaxes the body and reduces stress while coffeine does the opposite. If you brew the tea longer than the ratio of caffeine to theanine that gets released changes.
My understanding was that caffeine was dissolved almost immediately, on the first steep or two, and this was a standard recommendation for those who wish to cut down their caffeine consumption - throw out the first steep and drink the others. So the ratio could only be changing in theanine's favor, which doesn't sound like a bad thing.
I agree, with the exception that tea relieves extremely severe fatigue for a while.

Avoid caffeine before tasks requiring manual dexterity.

Tell me more? It doesn't particularly surprise me but I'm not familiar with that specific effect...
It makes my hands shake, especially when I'm nervous. So that adds: Avoid caffeine before playing poker.
Or you could just take more, so that the nervousness is swamped by the general handshakery...
What I've heard pointing in the same direction: a martial artist who gives up caffeine (very difficult for him) before tests, a juggling website which recommended against caffeine.

In the study which found impaired neurogenesis the rats were consuming caffeine chronically. Were the impairments in neurogenesis due to sleep deprivation which would not occur in morning-only human coffee drinkers?

I like your meta-analysis on to which kinds of tasks coffee works better.

I add something on the how much. Frequent small doses gives you better results than few large doses.

Actually, whenever in the absence of further specific evidences, I have found that small-doses-many-times is a good rule of thumb for a vast array of substances (eg: food in general, sugar) if the goal is to maintain a stable, productive mental state.


I don't trust a single one of these findings, as I don't trust their proxies to be of any real use, I don't trust their methodology to be representative of the real world, and frankly I'm extremely skeptical that any result I see is not classical statistics shenanigans before I at least see it confirmed by a meta-analysis based on large number of trials.

Do you also distrust neurochemistry? Caffeine has a known mechanism of action. It's not poorly understood like modafinil. Caffeine binds to certain types of adenosine receptors, but it doesn't activate them. This prevents adenosine from binding to and activating the receptors. Adenosine does a lot of things, but one thing it causes is sleepiness. If you increase the amount of adenosine in someone's brain, they get sleepy. The amount of adenosine in the brain naturally increases over the day until you fall asleep. So if adenosine makes you slow and sleepy, blocking it should make you sharper and more alert. It's no surprise that most people report caffeine having that effect on them.

So, there are a bunch of studies. There's a known mechanism of action. There are lots of caffeine users who can confirm the effects predicted by the studies and the neurochemistry. You can even get firsthand evidence if you try the drug yourself. What more do you want?

Basically yes. Brain are evolved complex homeostatic systems, not engineered systems with neat subsystems in which chemicals are control knobs; randomly changing various things in brains will not get you clean results. Findings that something interacts with some enzyme has almost no relevance to question of influence over real world effects. By the way caffeine doesn't even make you not sleepy. Modafinil and amphetamines do, in different ways.
In the paragraph before that, you said you didn't trust neurochemistry to explain how drugs work. Why do you then accept amphetamine's mechanism of action (or at least accept that it differs from modafinil's), but not caffeine's? And I'm pretty sure caffeine makes people less sleepy. It's not as powerful as modafinil or amphetamines, but it's effective.
Which people, though? Not all of them, every time, because as a heavy user, I have noticed that coffee sometimes seems to bring on drowsiness instead of dispelling it. I say "seems to" because I've not done any experiments, but I've heard it happens to other coffee drinkers as well.
By messing with adenosine receptors caffeine reduces one of the mechanisms behind 'sleepiness', that is, the sleepiness you get from having been awake too much. Because caffeine also has general stimulant effects, particularly at large doses it can also produce the seemingly paradoxical effects that stimulants have on some people. Sure, take enough caffeine and amphetamine and you can stay up and remain approximately functional. But sometimes taking stimulants makes people tired. For example, if I drink a red bull and take a couple of Ritalin tablets I can either fall straight to sleep or stay up for another 36 hours, depending on whether there is task that latches on to the ensuing over-focused attention control systems and overrides the sleepiness impulse. (Thus goes the typical explanation that I hear from various sources. I take it with a grain of salt. I trust the actual results of experiments somewhat more than the stories that get told in the accompanying discussion.)
My parents drink coffee before going to bed. I've always assumed that it was because withdrawal during the night would be worse than the "stay awake" effect, but maybe there's another explanation.
That sounded more interesting out of context than it should have.
Ha! And totally unintentional, I assure you :)
I've found that caffeinated sleep tends to be fairly satisfying and invigorating. The trick (for me, anyway) is to fall asleep before full onset. 100mg caffeine pills do the job nicely, assuming I'm sleepy enough to nod off in time. The penalty for failure is pretty stiff though - I've lost half a night's sleep that way.
Curiously, I've never experienced any sort of withdrawal symptoms from caffeine. When I've had a reason to stop drinking coffee and tea for a week, I've simply done so with no discomfort.
true, I usually feel sleepy after drinking coffee-but I rarely if ever drink it anyway.
Caffeine might increase energy levels, and do other things, but it is ridiculously inaccurate to say "caffeine makes you less sleepy" as its main effect. Try modafinil/caffeine double blind test - modafinil is not "stronger caffeine", it's completely different, by actually reducing sleepiness without any energy boost. Amphetamines are even more complicates.

What better sources of information should we be updating on?

Anyone have any experience or advice about optimal dosing? What is the maximum amount of caffeine I can consume while still avoiding dependency? What pattern of consumption is best?

Personally I find caffeine to be amazing for basically everything, but it becomes too weak after you build some tolerance, and withdrawal seems to affect me particularly badly.

5Swimmer963 (Miranda Dixon-Luinenburg)
I've dealt with the tolerance effect by using caffeine only intermittently. After an extremely hectic semester, when my policy was "drink coffee on mornings when I'm up before 6 am, and not on mornings when I get to sleep in to a normal time." The result: I was more tired when I slept in, because of caffeine withdrawal, and when I did drink coffee on early-wakeup days, it no longer made me feel especially alert or cheerful, but rather brought my energy and mood up to average. My current policy is to drink coffee only when I get up early and I find myself feeling extremely tired and groggy, and I have good reason to need my alertness (i.e. for an exam). If I have an exam and I got to sleep in that day, no coffee, whether or not I'm groggy. If I'm sleep deprived, I wait on drinking coffee until I actually feel tired and groggy, because often if I'm busy enough, I'll stay alert for the whole day anyway. If I didn't get much sleep and I'm groggy, but it's a weekend and/or I don't have anything important to do, no coffee...I just take a nap. I have only one coffee per day, max, and two days per week max, preferably not consecutive days. Probably the only reasons I was able to make this change are that a) instead of waking up at 5 am on four or five days out of the week and getting around five hours of sleep most nights, I get up at 7 or later most days and can get 8 hours of sleep per night, and b) I was able to wean myself off coffee during a two-week family vacation, when I could sleep 10 hours a night and laze around all day. I don't think I would have made it through the caffeine withdrawal if I was extremely busy at the time and had important things to do. That being said, the transition was very worth it. I've acclimatized so that I don't feel especially groggy without caffeine, and if I do, I have non-caffeine methods that work, i.e. taking a 5-minute nap while on break at work, or getting up and walking around when I start to get sleepy. And when I do drink co

In vitro, caffeine actually encourages hippocampal dendrite growth - unless this finding (from 1999?) was wrong. That said, the studies claiming that caffeine impairs some types of learning and hippocampal neurogenesis in live rats is probably more applicable in predicting the effects of human consumption.

To reason about Caffeine consumption, or the consumption of any mental enhancer, one should not consider only the damages and benefits of it. But the difference between those and the ones you'd have from consuming other enhancer.

Compare it with other stuff, not just being sober, otherwise, you are falling into the fallacy of the bicameral mind. The mind that evolved to think that everything is 0 or 1.

I agree, and would go further: you should consider all possible bags of drugs, because drugs have interactions. I don't see what "everything is 0 or 1" has to do with "bicameral mind)", a phrase I've long hated, but just now found is an interesting hypothesis and not just a bit of jargon.

What about Spiking that coffee with Provigil or Ritalin?

What if you have neither Ritalin nor Provigil?
I haven't noticed that Provigil is too terribly hard to get, once Doctors know that it isn't an Amphetamine, they will usually prescribe it... Remember though, it is usually prescribed in dosages for narcoleptics, so only half a tablet is needed... Ritalin, well, I guess that you really need to either be ADD to get that, or be willing to enter into a black market arrangement.
Selegiline goes well with caffeine. Unfortunate association with cigarretes aside, nicotine lozenges or patches are easily accessible and provide some benefitial effects too, depending on what kind of performance you desire from your brain at the time. I don't recommend actually spiking the coffee with a nicotine lozenge. But, come to think of it I've never tried. Given how habit forming coffee is already, adding nicotine to the mix would be... interesting to say the least.