“What work of genius has ever been composed on chamomile?” asks Michael Pollan in his audiobook Caffeine, as he tries to quit his habit. America runs on caffeine. Roughly 85 percent of us drink at least one caffeinated beverage every day. Here’s the latest on how caffeine affects our health.

Click here to see our chart of caffeine content in foods and beverages.

Caffeine 101

How does caffeine work?

“The stimulant effects that most people associate with caffeine are due to caffeine’s ability to block adenosine receptors in the brain,” explains Marilyn Cornelis, associate professor of preventive medicine at the Feinberg School of Medicine at Northwestern University.1

Adenosine, a natural sedative, builds up during waking hours and dissipates during sleep. But when caffeine blocks the adenosine receptors, adenosine can’t do its job...so you don’t feel drowsy.

Like with most dependencies, you can build up a tolerance to caffeine.

“The more caffeine you consume, the more adenosine receptors your brain makes,” notes Cornelis. So you need even more caffeine to block those extra receptors and keep you alert.

How much caffeine is too much?

The Food and Drug Administration says that most adults can safely consume up to 400 milligrams a day.

But caffeine’s impact on people varies.

caffeine diagram
Why caffeine keeps you up
Caffeine (C) binds to adenosine receptors in the brain. That prevents adenosine (A), which is a natural sedative, from attaching itself and making you feel drowsy.
Jorge Bach/CSPI.

“On average, the half-life of caffeine—which is the amount of time that it takes to break down half the caffeine in your body—is about three to five hours,” says Cornelis. So if you drink a cup of coffee at, say, 2 p.m., a quarter of its caffeine may still be coursing through you between 8 p.m. and midnight.1

“But there’s large genetic variability in the activity of the key enzyme that metabolizes caffeine,” says Cornelis.

What’s more, caffeine metabolism can change. For example, it speeds up if you smoke cigarettes and slows if you take birth control pills or are pregnant.

Of course, you don’t need a genetic test to find out how caffeine affects you.

“Those tests only look at the gene for metabolizing caffeine,” Cornelis points out. “But maybe you have a genetic variant in the adenosine receptor, which could make you more or less sensitive to caffeine’s effects on sleep. The test won’t tell you that.”

What’s more, most people know how much caffeine they can handle. “People with the genetic variant that is linked to slower caffeine metabolism generally consume less caffeine,” says Cornelis.2

“If you drink one cup of coffee and you’re off the wall for the rest of the day, that’s more informative than a genetic test.”


“I’m sleeping like a teenager again and wake feeling actually refreshed,” says Michael Pollan about his breakup with caffeine.

But Pollan’s experience may not apply to everyone.

In one of the few studies to look, researchers had 66 young regular caffeine users who had trouble sleeping go cold turkey. Over the next week, the volunteers spent no more time asleep and took no less time to fall asleep than before.3

At Starbucks, some varieties of coffee have more caffeine than others. Check our chart or Starbucks’s website.

Even so, it may be wise to avoid high doses of caffeine later in the day.

In an industry-funded study, researchers gave 12 young regular caffeine users either a placebo or 400 mg of caffeine six hours, three hours, or immediately before bedtime on separate days. Compared to when they got the placebo, the volunteers slept about an hour less no matter when they got the caffeine.4

Studies show that people take longer to fall asleep, get less deep sleep, and wake up more often throughout the night when they’re given caffeine right before bed.5 (There’s little research on whether having caffeine in the morning or early afternoon also impairs sleep.)

If you think caffeine is keeping you from getting a good night’s rest, try cutting back or cutting it out after midday.

Type 2 diabetes

“Coffee consumption has been consistently linked with a lower risk of type 2 diabetes in cohort studies from across the world,” says epidemiologist Rob van Dam of the National University of Singapore and the Harvard T.H. Chan School of Public Health.

In one analysis of 28 studies that tracked roughly 1.1 million people for an average of 11 years, the risk of type 2 diabetes was 6 to 8 percent lower with each daily cup of coffee, up to six cups a day.6

(Keep in mind that a cup is 8 ounces, but many mugs hold 10 to 12 oz. At Starbucks, the only size on some menu boards—a “grande"—is 2 cups, or 16 oz. A “venti” is 2½. Want just 1 cup? Ask for a “short.”)

To prove that coffee prevents type 2 diabetes, you’d need a study that randomly assigned people to drink coffee or a placebo for years.

hot tea
In one study, steeping black tea bags for 3 minutes instead of 1 increased the caffeine by about 65 percent.
MK studio/stock.adobe.com.

Surprisingly, “in short-term trials, caffeine reduces insulin sensitivity,” says van Dam.7 In other words, caffeine makes insulin less effective at moving sugar from your blood into your cells, which could eventually lead to type 2 diabetes.

However, “the body adapts to the effects of caffeine, usually within a week or so,” notes van Dam. “In trials lasting several weeks, caffeinated coffee has no detrimental effects on insulin sensitivity.”8

But coffee is more than a caffeine delivery system.

“Coffee is complex, and it contains compounds that, in animal studies, have beneficial effects on insulin sensitivity and blood sugar,” says van Dam.

That would explain why studies that track people for years show a lower risk of type 2 diabetes with both regular and decaf coffee.

“Components of coffee other than caffeine may be responsible for the beneficial effects of coffee consumption on diabetes risk,” says van Dam.

Exercise performance

“If we put you on a treadmill and told you to go as long as you can, you’ll run farther if you’ve had caffeine than if you haven’t,” says Matthew Ganio, who heads the department of health, human performance, and recreation at the University of Arkansas.9

(Researchers typically give people between 1.4 and 2.7 milligrams of caffeine per pound of body weight an hour before exercise. For a 150-pound person, that means roughly 200 to 400 mg of caffeine. And more isn’t better. Higher doses don’t give more of a boost and can cause side effects.)

On average, people who have had caffeine run about 12 percent farther when they push themselves to exhaustion. “But there’s almost never a setting where you’re pushing yourself until you can’t go anymore,” Ganio points out.

So he looked at the research that gauged whether caffeine could slash the time it takes to, say, run a 5-kilometer race or row 2,000 meters. After analyzing the results from 21 studies (which enrolled mostly young, fit men), he calculated that caffeine could boost performance by about 3 percent.10

“That may not seem like a lot to the recreational athlete,” says Ganio, “but it would be quite the difference in the Olympics.”

Not an Olympian?

“The average person can still reap the benefits,” says Ganio.

In one study, researchers enrolled 12 Australian men who typically exercised less than an hour a week. On two separate days, the men took a placebo or 2.7 mg of caffeine per pound of body weight an hour before riding stationary bikes for 30 minutes. On the day they got caffeine, the men burned 5 percent more calories, pedaled 5 percent harder, and pushed their heart rate 5 percent higher, all without feeling like they were working harder.11

Few studies have looked at the exercise-boosting effects of caffeine in older adults, Ganio points out.

“But I can’t think of a reason why it wouldn’t help them.”

In one study, 19 British adults aged 61 to 79 performed a battery of physical tests an hour after taking a placebo or 1.4 mg of caffeine per pound of body weight. On the day they got the caffeine, the volunteers were able to do one more bicep curl in a 30-second test and walk about 100 feet farther in a 6-minute test than when they took the placebo.12

Why might the caffeine in two-or-so cups of coffee give your workout a jolt?

For one, “it reduces feelings of fatigue,” says Ganio. Caffeine can also blunt the perception of muscle pain. “And, at a given pace, people will rate exercise as less difficult when they’ve had caffeine.”

Taken together, that means that caffeine may make exercise feel a bit easier.

“In turn, maybe you’re able to push yourself a little harder,” says Ganio.

LDL cholesterol

At typical intakes, caffeine doesn’t raise your risk of heart disease.13

But some methods of brewing coffee—regular or decaf—could harm your heart by raising your LDL (“bad”) cholesterol.

Fortunately, that’s not the way Americans typically drink theirs.

“When coffee is prepared using a drip-filter method, a compound called cafestol remains in the filter and doesn’t make its way into your cup,” van Dam explains.

And it’s cafestol that raises LDL. (No one has tested whether wire mesh filters can trap cafestol as well as paper does.)

In coffee made with a French press or boiled (like Scandinavian or Turkish coffee), the cafestol isn’t filtered out. Love a latte? Brewing espresso filters out about half the cafestol.14

coffee ice creams
A serving of Ben & Jerry’s coffee ice cream has 65 mg of caffeine. A serving of Talenti has none. But you can’t tell that from the labels.
Kaamilah Mitchell/CSPI.

The rise in LDL from unfiltered coffee isn’t trivial.

In one study, 64 Dutch adults were randomly assigned to a no-coffee control group or to drink 30 oz. (nearly 4 cups) of either filtered coffee or unfiltered coffee every day. After 11 weeks, the average LDL level of the unfiltered-coffee drinkers was 16 points higher than the level of the filtered-coffee drinkers.15

How much unfiltered coffee is too much? “It’s difficult to establish a clear cutoff,” says van Dam. “But each additional cup seems to increase LDL more.”

His advice: “For people with elevated LDL levels, it seems prudent to avoid drinking unfiltered coffee daily and to try switching to another preparation method.”

That French press? Save it for the occasional treat.


Caffeine improves dopamine signaling in the brain. And loss of dopamine causes the movement problems that occur in Parkinson’s disease.

Could caffeine help protect against Parkinson’s or slow its progression?

“It’s quite clear that people who consume more caffeine have a lower risk of Parkinson’s disease,” says Ron Postuma, a neurologist at McGill University in Montreal.

In one study, researchers tracked roughly 138,000 men and women for up to 32 years. In women who had never used hormone replacement therapy and in men, those who consumed the most caffeine (about 600 to 800 mg a day) had a 38 percent lower risk of Parkinson’s than those who consumed the least. (It’s not clear whether caffeine lowers the risk of Parkinson’s in women who take hormones after menopause.)16

“You would assume that would mean that caffeine must be protective against Parkinson’s disease,” says Postuma. But something else could be at play.

For example, “Parkinson’s disease has a period of about 15 to 20 years where it’s already in your brain, but you don’t show symptoms,” Postuma explains.

“During that time, people might start losing sensitivity to caffeine, so they use it less.”

Could caffeine help people who already have Parkinson’s?

Postuma randomly assigned 118 patients to take either a placebo or 200 mg of caffeine twice a day for six months.17

“We didn’t see any obvious motor benefit in the patients who got caffeine,” he says. “I don’t think caffeine can be recommended for long-term management of Parkinson’s symptoms.”

Caffeine caution

Caffeine is safe for most adults if they don’t overdo it. But it has some downsides:

  • Too much caffeine can make you jittery and disturb your sleep.
  • Caffeine doesn’t cause hypertension, but it can raise your blood pressure for several hours.
  • Energy drinks have been linked to irregular heart rhythms and increases in heart rate and blood pressure. Don’t get carried away.

1Neurosci. Biobehav. Rev. 71: 294, 2016.
2Mol. Psychiatry 20: 647, 2015.
3J. Sleep Res. 2020. doi:10.1111/ jsr.13048.
4J. Clin. Sleep Med. 9: 1195, 2013.
5Sleep Med. Rev. 31: 70, 2017
6Diabetes Care 37: 569, 2014.
7Nutr. J. 15: 103, 2016.
8Nutr. J. 10: 93, 2011.
9Int. J. Sport Nutr. Exerc. Metab. 14: 626, 2004.
10J. Strength Cond. Res. 23: 315, 2009.
11J. Sports Sci. 30: 1235, 2012.
12J. Nutr. Health Aging 18: 883, 2014.
13Circulation 113: 2045, 2006.
14Curr. Cardiol. Rep. 15: 403, 2013.
15Arterioscler. Thromb. 11: 586, 1991.
16J. Parkinson’s Dis. 7: 677, 2017
17Neurology 89: 1795, 2017.



Subscribe to Nutrition Action

Nutrition Action is completely independent. We accept no advertising and take no donations from corporations or the government. So we’re free to blow the whistle on dishonest products and to applaud the good ones.

Sign up today