What's Wrong With the Coffee Mortality Study? You Tell Us.

By Sarah Zhang | May 18, 2012 11:47 am

coffee stain

A recent study suggesting a link between coffee drinking and longer lives has prompted a flurry of coverage—some snarky, some cautious, but mostly celebratory. (We see you there, reaching for another cup of coffee.)

The study published at the prestigious New England Journal of Medicine is about as good as observational epidemiology studies go, but it’s limited by virtue of being observational. Last month on our Crux blog, Gary Taubes wrote a hard-hitting piece about the problems with observational studies. A major limitation of surveying people about their lifestyle habits is that correlation does not imply causation. It can’t prove coffee drinking actually led to living longer. There are always confounding variables.

In this coffee study, for example, they initially found that coffee drinkers died younger, but coffee drinkers are also more likely to be smokers. When they controlled for smoking as a confounding variable though, the result flipped: coffee drinkers lived longer. The researchers recognized there are other confounding variables too, and this is the entire list the researchers controlled for, taken directly from the paper:

The multivariate model was adjusted for the following factors at baseline: age; body-mass index (BMI; the weight in kilograms divided by the square of the height in meters); race or ethnic group; level of education; alcohol consumption; the number of cigarettes smoked per day, use or nonuse of pipes or cigars, and time of smoking cessation (<1 year, 1 to <5 years, 5 to <10 years, or ≥10 years before baseline); health status; diabetes (yes vs. no); marital status; physical activity; total energy intake; consumption of fruits, vegetables, red meat, white meat, and saturated fat; use or nonuse of vitamin supplements; and, in women, use or nonuse of postmenopausal hormone therapy.

But are there more confounding factors lurking unaccounted in the numbers? The Boston Globe interviewed Dr. Jeffrey Drazen, the editor-in-chief of NEJM, who admitted the studies like this are problematic and the decision to publish was controversial within the journal. He points out that the study did not control for health insurance, high blood pressure, or high cholesterol levels.

At first the list of confounding factors from the paper looked pretty exhaustive (and long) to us, but there’s always more factors to consider. So let’s play a game of find the confounding factor. What else do you think should the researchers have controlled for?


CATEGORIZED UNDER: Health & Medicine
  • Kieron

    How much people sleep, & alternate caffeine intake.

  • IMEarlGrey

    It correlates highly with education, but I notice income isn’t on there. My hunch is that those who have higher income have better access to health care (also not listed in the controls). Perhaps those who have more income are more likely to frequent coffee shops, or work the long hours at white-collar jobs, where coffee can be consumed throughout the day (as opposed to those who work more in physical labor, where sipping on a coffee all afternoon is presumably less feasible).

    Either way, income and/or access to health care seem to be the biggest factors not in the list. I’d love to believe my caffeine addiction is providing long-term health benefits, but I’d need to see these correlations controlled for before I’m willing to consider my habit “healthy.”

  • Jay Fox

    Did they sort out exactly how the coffee was consumed? Some might claim that highly sweetened coffee drinks would be “less healthy” than coffee consumed black or plain.

  • Ben

    How about income? If coffee is a luxury good, one would expect higher income to correlate with coffee consumption.


    And higher income might be more closely related to mortality (directly via health care/quality of living, secondarily via stress?).

  • Audrey

    “1 cup” does not tell the amount of caffeine people consume on a daily basis.
    However the study claims to show that coffee drinking makes people live longer. I’m just insolently assuming we are talking about caffeine, not the movement required to actually put the coffee in your mouth.

  • http://leefalin.com Lee Falin

    Since the paper is behind a paywall I haven’t read it in its entirety yet, but from the excepts I’ve seen it doesn’t appear that they take into account family history of cardiovascular disease, which is a well known risk factor.

  • BJM

    @Audrey: “…assuming we are talking about caffeine…”
    According to the account at the link below, the study indicates: “…researchers found similar risk outcomes for those who drank regular coffee and those who downed decaf.”


  • Jose

    coffee is consumed mostly in the morning,afternoon or evening.
    Location/place of residence -somebody living near a factory or a nuclear plan?…
    Consumption of tea/cocaine/other legal or ilegal drugs with similar effects to caffeine

  • RJD

    Ad infinitum. They would do better randomizing.

  • Diane

    >> “So let’s play a game of find the confounding factor. What else do you think should the researchers have controlled for?”


  • Audrey

    @BJM: Thank you very much! So I was indeed insolent! But now I want a control group of people pretending to drink coffee and see what happens…

  • Hudasx

    Given the sample size of that study, can’t we just accept that coffee can be good for us?

  • Superchkn

    Did they account for a social or behavioral aspect? My grandparents would drink a lot of coffee, but they would primarily do so when meeting up with others in their community. Perhaps that offers a similar effect as marriage? They would usually have a few different times throughout the day when they would meet at a coffee shop and meet with friends. This is observing an older population and I wonder how many of the members of the group still had jobs at this time, so the typical working-age view of coffee consumption may not apply.

  • jan

    There is such a vast difference in coffee produced ,processed and consumed that you might as well compare apples and oranges……set a standard first.

  • Greg

    To respond to one of the above comments: Decaf is not caffeine free, it just has reduced caffeine.
    One important variable not included is sleep apnea, especially untreated sleep apnea. Although BMI correlates with sleep apnea, it is not predictive. Caffeine improves alertness and concentration making it less likely for a person with sleep apnea to die in an auto-accident, especially those notorious single-vehicle accidents. An argument can also be made for anyone else likely to be sleepy even if not on a regular basis, such as shift workers and long haul truck drivers.

  • http://discovermagazine.com Iain


  • Audrey

    Well, even though there’s caffeine in decaf, is it reasonable in a study focused on coffee drinking effects to compare people drinking strong espressos with those having light coffee or even decaf, and make conclusions? If they had considered education, income, health insurance, family history, etc… what would we be looking at anyway? It’s also pretty unlikely that we all “metabolize” caffeine the exact same way, even with a similar intake.

    Considering that “coffee drinking or not” can be defined as a single variable is confounding already. So why bother with subtler factors?

    Anyway, it’s more challenging to answer the question “What’s not confounding in this study?”.

  • blindboy

    I really don’t see much value in the study. There is no existing evidence to suggest that coffee might have a significant impact on longevity. In those circumstances trying to extract whatever minimal effect it might have from the similarly minimal influence of the numerous other factors already mentioned would seem to be both futile and pointless.

  • Tony D

    Agree on the difficulty with all the variables to conclude on positive effects of coffee. However, the study set out to determine if there were any negative effects related to coffee, which I believe the results can lead us to conclude that there are no significant negative effects on mortality.
    Drink up, but still take care of your self.

  • Jenk

    It seems an almost insurmountable task to use statistical data and correlations to support a claim for something like longevity. There are simply too many contributing factors in the human experience. Clone mice, force an identical lifestyle and try again, I drink lots of coffee and would love outside approbation!

  • granny’s kid

    well i dont know much about confounding factors, but as an observation i remember all my “granny aunts & uncles” (all my grandmas siblings & cozins) drinking good strong black stovetop coffee- the rot gut that sits there & percs all day- like it was water…. they all grew up dirt poor dust bowl farm kids before the Great Depression and worked hard and ate light–
    and every single one of them survived into their 80’s or 90’s…

    my grandpa who married into her family grew up & lived the exact same way, swigged the same amount of coffee, and died in his mid 60’s of Alzheimers.

    im no “expert”, but really it seems to me it could have more to do with family genetics than coffee in the blood…
    but i am my grandma’s AND my grandpa’s granddaughter, either way! and here’s to Coffee regardless :)

  • Alan Andrew

    It also may not be linear: caffeine may prolong life to a point, but too much caffeine may shorten life.

  • Brian Too

    Cat person versus dog person.

  • Bram Floria

    Over-active imagination.

  • floodmouse

    They should have controlled for:

    Type of coffee: Freeze-dried coffee, especially in aluminum packets, can’t have as much potential flavenoid impact as fresh-ground coffee beans. Also I believe exposure to aluminum has been correlated with some negative health impacts.

    Laxative & diuretic effect: One of the major effects of coffee (or tea) is to purge a lot of garbage out of the digestive tract that otherwise might sit there and fester. In other words, if a lot of chemical-laced processed food is sitting around in your digestive tract for a longer period of time, more toxic chemicals will get absorbed into your tissue instead of getting excreted. To do a proper study, somehow you would have to control for the non-coffee drinkers who achieved a similar fast turnaround using just laxative chemicals, alternative hot beverages, or exercise. Good luck getting reliable reporting on bowel habits from a questionnaire.

  • http://None Tom Russel

    Not controlled for gender of the subjects, except for a very narrow criterion only applicable to women.

  • Siobhan

    To be fair this was a tough study to undertake and I don’t expect they would expect everyone to take it completely as fact – as it is obvious that it would be near to impossible to account for every other factor effecting longevity.

    However, if you control for something how do you know that coffee consumption levels don’t affect the very thing you have controlled for. ie coffee could some how in a yet undiscovered way decrease your BMI – making you more healthy and increasing your life expectancy. I guess what I’m saying is that if you control for something that may effect life expectancy how do you know that this isn’t the mechanism by which coffee works its life increasing magic?

    I’d like to think everything is good in moderation until someone can pinpoint exact mechanisms by which something has a positive effect.

  • Wil

    I am baffled by the continuing lure of doing garbage studies.

    This group undoubtedly knew before they even began, that they they did not have enough money, mathematical skills, or skeptical attitude to prove, or to disprove, anything what-so-ever. And yet they still took the time and spent the money to do it. Why?

    We all hear of several studies per week, week in and week out, that do not really prove or disprove anything at all. Why do people waste a portion of their lives doing that? It’s as if they hunger to dig ditches, and then fill them up again.

  • TheRant

    *Rant begin

    Without being specific, I must just go ahead and say that when it comes to genetics, that is not, and should never be, an “answer” to why people live longer or shorter lives than others. I know people who have used “it’s all genetics, anyway,” as a reason to do all sorts of unhealthy things to their body from smoking to drugs to drinking to the foods they eat. Some people will live longer than others, of course. However, people that live into their 90s living that extremely unhealthy diet and lifestyle while health gurus die in their 30s from heart attacks are the exceptions to these rules. They’re not evidence to the contrary. To anyone who may already do this, please don’t go around telling people that the most important factor is genetics. It’s a huge factor, yes, but people don’t need yet another reason to justify unhealthy lifestyles (here’s looking at us, America). – American concerned for just-informed-enough-to-make-bad-choices Americans

    *Rant over

  • Karen

    I liked the prior idea of pretending to drink coffee. Substitute warm water or some other tasty warm substance. Does the practice of having a nice hot cup of coffee lead to 10-15 minutes of meditative downtime? Two or three times a day? It’s often associated with breaks – “coffee break” time at work etc. I know I will rarely have a decaf, but when I do, it’s because I visualize holding that warm cozy mug in my hands, sitting quietly outdoors listening to the birds etc. Perhaps it’s not the substance but the related relaxation time. Or, as another commented, the relaxed socializing that often occurs with it.

  • http://tompetrie.net Thomas Petrie

    Like others, I question the motive behind this study. No one noted that coffee is a crop that is heavily sprayed with pesticides, unless organically grown. At least two common sprays used in coffee growing contain fluoride (e.g., Sufuryl Fluoride and Sodium Aluminum Fluoride/Cryolite). Doesn’t anyone ever wonder why the discussion of fluorides/pesticides in one’s coffee is almost never discussed in the media or investigated by researchers? If we really want to study coffee, we should study people who drink organic coffee versus those who drink commercially raised coffee. We should study various health parameters (as has Steven Cherniske in Caffeine Blues, 1998). As has been noted by other commenters here, this study hasn’t taught us anything.

    As an individual sensitive to the toxic effects of fluoride, I can attest to the fact that my joints do NOT hurt after drinking organically raised coffee, but do hurt after commercially raised varieties–especially if the coffee is instant. It’s not universal, and that’s likely because the fluoride content of coffee varies between varieties. I say let’s TEST the coffee for residues.

    While this writer likes his coffee, it is NOT a healthy beverage and I’m tired of seeing studies–such as this NEJM one–sugar coating this issue. Like many observational studies, it has just obfuscated this issue, not spread light on it.

  • Julian

    Drink Colombian coffee. You will never see a crop-dusting plane negotiating the severe curvatures of the coffee growing Andes mountains.


Discover's Newsletter

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


80beats is DISCOVER's news aggregator, weaving together the choicest tidbits from the best articles covering the day's most compelling topics.

See More

Collapse bottom bar