Do you ever wake up in the morning and think, hmmm, this would be a great week to try cutting out coffee? No? Well, I have some good news for you. A recent study done by the Harvard School of Public Health analyzed the coffee-drinking habits of over 208,000 people over decades. In the time frame of the study, 277 of those people died by suicide. Interestingly, those people who drank 2-4 cups of coffee a day had a 50% lower risk of death by suicide. The researchers thought caffeine might be the player here, as caffeine has chemical activity that not only increases alertness (as we well know) but also has some antidepressant effects.
The data is consistent with a bunch of previous studies over the years. I reviewed many of them in my previous article, Coffee Happiness. There’s been a lot of news coverage about the most recent study, and I’ve seen several writers touting the benefits of caffeinated beverages in general, but the research has been fairly consistent in showing only coffee has enough caffeine to be have the antidepressant effect. Tea and cola consumption, in these large observational studies, have not been correlated with reduced suicides or reduced depression. (In full disclosure, I am a tea drinker and rarely have coffee.)
There is also a “sweet spot” of coffee consumption, somewhere right around 4 cups a day (though the Harvard scientists say 2-4 cups, other research has pointed to slightly higher amounts being protective, so I’m hedging my bets). Certainly there is such a thing as too much coffee. In a Finnish study from 2000, those who drink more than 7 cups of coffee have an increased suicide risk compared to moderate Finnish coffee drinkers. The Finns have some of the highest coffee consumption in the world and one of the higher suicide rates.
The strength of these sorts of studies are the sheer numbers. It is awfully hard to get good data on risk factors for suicide simply because completed suicide is relatively rare. But if you follow 200,000 people for several decades, you have enough people committing suicide in that cohort to gather some statistically significant data. The limitations of these studies are profound. They are observational and tell us nothing about cause and effect (we can only guess at that). For those reasons, when large observational studies tell me that moderate exercise or fish or oatmeal is good for me, I yawn.
We’ve all been told that these things are good for us by the health authorities for the past gazillion years. Healthy people with adequate impulse control and executive planning capabilities will therefore adjust their lives to eat more fish and get some exercise. They are also far more likely to not drink heavily, not smoke, and get the appropriate amount of sleep etc. etc. etc. And while the epidemiologists say they can pick apart all these variables if they ask the right questions and use the right computer program to take out one variable, I’m skeptical.
But the same limitations that plague the results of these large observational studies that seem to form the backbone of public health advice these days is actually a strength of the coffee study. Coffee consumption is not one of those healthy things we’ve been told to do. In fact, we are typically told to limit caffeine, because it can make you anxious and irritable and wreck your sleep. So the coffee drinkers are not typically the magical healthy cohort of people who all drink green smoothies and eat lentil soup. In fact, coffee drinkers are more likely to smoke, drink, get poor sleep, and don’t exercise as much as those who don’t drink coffee. Therefore finding that coffee drinking is associated with less depression and less completed suicide even though coffee drinkers tend to have less healthy habits in general is very interesting.
The other limitation of these large epidemiological studies is that they are based on rather infrequent food frequency questionnaires (FFQs). Yes, all these hundreds of thousands of people were given FFQs every four years. So imagine being asked how many hamburgers you ate, say, on a weekly basis on average the last four years and how accurate you could possibly be on these surveys. But coffee consumption tends to be a daily activity, perhaps with more on workdays and less on weekends and holidays, but I would be more apt to believe the accuracy of someone who told me, yes, I drank a pot of coffee a day every day for the past four years than someone who said he ate 3.2 cheeseburgers a week.
Can we think of any confounders that weren’t accounted for in the study? One group we do automatically remove in a large cohort of coffee drinkers are the people who are too anxious to tolerate coffee, and anxiety is a known risk factor for suicide. I’m sure there are other confounders that could explain part of coffee’s seeming protection from suicide. But it certainly is plausible that coffee could have anti-suicide effects, and in the right doses, it's associated with greater longevity, health, and happiness, so drink up.
So far as my clinical practice, I don’t generally advice people to stop drinking coffee unless they have awful, intractable insomnia. I just tell people to try to keep it less than 8 cups a day.
Copyright Emily Deans, MD