Coffee pops up in the news every now and then, especially when large population-based studies like the one we are about to discuss are published. Usually people have no reason to fear, but skeptics exist.
There has been research showing increased mortality in folks consuming upwards of 6 cups of coffee a day. The thing is, those people smoke a lot more. Once the effects of smoking are removed from the equation (in a type of statistical analysis called a regression analysis) these ill effects disappear.
The study, titled “Association of Coffee Consumption with Total and Cause-Specific Mortality in Three Large Prospective Cohorts,” investigated the relationship between smoking and coffee consumption on mortality (death). They also looked at cause-specific mortality: deaths from specific diseases.
So here’s what happened. They took data from three different prospective cohort studies. These types of studies take a group (a cohort) of ideally randomly selected people, take some kind of measurement or collect data from a survey, then check up with them way later in the future and see what happened to them. In this case they looked just at death (these people were followed up for a long time).
The researchers collected data from three different studies: the Nurses’ Health Study (NHS), the Nurses’ Health Study II (NHSII), and the Health Professionals Follow-up Study (HPFS). These large studies began collecting data many years ago. They consist of medical professionals like nurses and doctors. The details if you are interested are listed under “study population” in the paper. It’s very important to know what population was studied when you read a paper and then ask yourself if the results of the study are generalizable to the rest of the population.
The NHS collected data from women only, and the HPFS, men only. Together, these three studies followed over 200,000 people for 28 years and 4,690,072 person-years (the amount of people times the number of years, something like that) of follow-up.
Here’s what they found:
Proportion of those who never drank coffee and never smoked:
- NHS: 63%
- NHS II: 80%
- HPFS: 71
Proportion of those who drank >5 cups/day of coffee and never smoked:
- NHS: 24%
- NHS II: 35%
- HPFS: 25%
All-cause mortality (death from any cause):
There were five total groups based on amount of coffee consumption: 0 cups per day, less than one cup a day, between 1-3 cups, between 3-5 cups, and over 5 cups a day.
A lot of associations tend to follow a non-linear curve, meaning the effect of whatever is being studied doesn’t simply go up with higher doses of it (or down with higher doses or lower doses). In fact for exercise and all-cause mortality the curve is J-shaped; those who don’t exercise at all have increased mortality, then those that exercise moderately have decreased mortality, and the heaviest exercisers see a small increase in their mortality but not as much as the non-exercisers (it’s a sideways J..I’ll add it in here later perhaps as an update).
The shape of the curve for coffee consumption and all-cause mortality is similar; those who consume a lot of coffee see a slight increase in their mortality, and those that drink moderately see a decrease in their mortality relative to those that do not drink coffee at all.
The thing is, those people who drink a lot of coffee smoke more too.
As a result, increased lung cancer and respiratory disease cases were observed in the groups that drank more than even 1 cup of coffee. Once this relationship was studied in never smokers, there was no more trend for lung cancer with increased coffee consumption. Smoking in this case is a huge confounding variable; it interacts with what we’re looking at making it harder to see a true effect from what we are interested in studying.
All right, well there you have it! I will update this post later with some more statistical information and pictures, but for now, all you need to know is that these three large prospective cohort studies found that drinking between 1-5 cups a day reduced mortality rates from all causes as well as cardiovascular disease (death rates among coffee drinkers was cut in half for those who suffered from diabetes and neurological diseases as well as suicide).
This means that without respect to individuality, or personalized medicine, or constitutions, overall, a cup or five of java is likely to help you live longer. Your chances of death are around 10-20% less from any cause if you are a regular coffee-drinker (For the exact numbers you need to look at the hazard ratios, provided in the tables at the bottom of the paper around page 50 onwards. You will see that the HR’s are between 0.8-0.9. If the HR is 0.9 for 1-3 cups a day of coffee then it means the risk of death is reduced by 10%).
Now I find those people who never drank coffee to be pretty interesting people. Would they live slightly longer if they drank coffee? See that’s not a question that this paper can really answer. You can say that those people who did drink coffee died less often than those who did not drink any coffee whatsoever. But to say that those people who did not drink any coffee at all would live longer is different because as usual the chicken and egg question comes up.
Are the people who don’t drink coffee genetically different than those who prefer coffee? What about those people that drink more than five cups of coffee a day and smoke a lot? Are those people constitutionally different? This is interesting to me because I like the personalized aspect of medicine; it allows us to fine tune information from large studies like this and think about the greater context of a god damn person (excuse my language).
In general, since there were a large number of people in the study, it is assumed that they are mostly similar and that they would actually live longer with coffee. But since no genotype analyses were done I think that question is unanswerable. My 23andMe results did tell me that I have some gene that predicts that I would consume less caffeine than average, so there is a genetic basis to this. Also those people that smoke and drink a lot of coffee perhaps need more stimulation. There are theories that differences in the dopamine receptor subtype 2 are responsible for explaining these differences.
Furthermore, the population studied was probably more stressed out than the average population I would imagine as they are mostly nurses and doctors. How does that add to the equation? Maybe those that drank coffee were better able to withstand the stresses of their job better than those that did not drink any coffee at all. Stress can cause free radical damage and if you’re not taking anything to alleviate that then those free radicals could accelerate the aging process. Also, those that drank the most coffee drank fewer sugar-sweetened beverages. They did do some separate analyses looking at that I believe but didn’t find anything noteworthy to report. But if those people that didn’t drink any coffee drank more sugar, we have another confounding variable to add to the equation, because this data almost makes it sound like if you DON’T drink coffee you will die sooner than those people that do drink coffee.
Out of all the mechanistic explanations for why coffee may prevent mortality, I can tell you there is a substitute. Antioxidants? Well a gazillion things have antioxidants. Big whoop. Caffeine is a stimulant and improves circulatory function? Big whoop a bunch of things did that.
And that is one problem with interpreting large studies like this: you study what most people do and you don’t study what is possible because that is something fewer people do and there aren’t enough of them to inform larger studies. What is nice thogh is that if there is a one-size-fits-all effect from anything and you see it in a very large sample, something is very likely to be actually going on.
So with that being said, coffee is definitely good for you especially considering that risk of cardiovascular disease-related deaths was less in those that drank a lot of coffee because those people smoked more too (if I find the exact statistics on that I will share but I don’t see it anywhere in the paper).
If you don’t drink coffee are you at risk of dying sooner? No, unless you consume more soda or take part in unhealthy activities. People who drink coffee may need to move around more and be more active in general so the anti-sedentary factor likely plays a large role as well.
Anyway, I hope you learned something and are interested in the big picture like I am!
I will soon discuss different types of coffee preparation methods as well as the results of my new pour-over filter…it should arrive tomorrow. Stay tuned, check out the gram, and sleep well because that is really important. So don’t check out the gram with or without nightmode on your phone right before bed in the dark unless you want to suppress your melatonin and GH production!