Margaret Voss
Margaret Voss is a professor of practice in ’s Department of Public Health, Food Studies and Nutrition. She teaches courses in nutritional biochemistry, metabolism and nutrigenomics. She shared her thoughts about recently released research that seems to indicate that drinking more coffee can increase longevity, and why this shouldn’t send everyone running for the coffee pot.
These studies are both large cohort studies in which data were analyzed post-hoc, or after the fact. The data are subsets of large health surveys, as opposed to data from a designed experiment meant to test a specific hypothesis about drinking coffee. The data for the first study were collected over a 16-year period; data for the second were collected between 1993 and 1996. Both studies used regression techniques.
In general, regression is a statistical method that can show a correlation between two variables. An example of this from the first paper is the relationship of increased coffee consumption with decreased incidence of cardiovascular disease in women. This is a valid observation and the results are presented in the form of a hazard ratio. To calculate hazard ratios, researchers compare a group (coffee drinkers) to a control group (non-coffee drinkers) and determine the likelihood of the treatment group dying from a disease compared to the control group. For example, if coffee drinkers are as likely to die from heart disease as the control group, the odds ratio would be one.
The authors report a statistically significant inverse association of coffee drinking with mortality related to circulatory disease among women. The hazard ratio for this relationship was 0.78, meaning female coffee drinkers were slightly less likely to die from circulatory disease. Although statistically significant, this is a very slight decrease from one. At the upper end of the interval reported (0.90), some female coffee drinkers would be only 10 percent less likely to die from circulatory disease than the control group.
Cardiovascular health can absolutely be improved through dietary change, but the most effective changes would involve decreased consumption of saturated fat and an increase in fruit and vegetable intake. This is why both studies, although valid, are not as exciting as the headlines might make it seem.
My explanation above breaks down the significance of the correlations reported, but nowhere did I mention what it was about coffee that could produce the results. In reality, we have no idea of why this relationship exists. These studies were not designed to test any particular hypothesis about coffee. They simply identified trends. So, we do not know if caffeine, an antioxidant or some other chemical compound in coffee was responsible.
The correlation might even be a result from some other lifestyle factor that most coffee drinkers have in common rather than anything in coffee itself. Maybe coffee drinkers sleep less and are slightly more active than non-coffee drinkers.
The only way to actually test the hypothesis that coffee increases longevity would be to do a randomized, controlled trial in which coffee is administered to some participants and not to others, while controlling for confounding lifestyle factors. This would be costly and impractical since the response variable is longevity. We would have to wait for participants to die to analyze the results. However, a few older and smaller studies have attempted to do exactly this and these do shed some light on the new studies.
Previous studies have attempted to address this. One indicated that antioxidants in coffee might prevent a lipid reaction in the cardiovascular system that leads to inflammation and atherosclerosis. However, several studies have shown that there is a genetic component also at work here; people with certain variations of a specific gene do not see any benefit from coffee drinking, while other genotypes do.
Beyond changes in lipid profiles, some studies have shown that female coffee drinkers may have longer telomeres than nondrinkers. Increased caffeine consumption seems to have some effect on increased telomere length. Telomeres essentially allow cells to divide without losing genes; longer telomeres are associated with longevity. Which brings me to another important point. All of the results we have seen for coffee, including the genetic aspect, also apply to tea drinkers.
At this point we only know that coffee drinkers seem to gain some benefit, however marginal, against some diseases. We have no idea why. The new studies differ from previous work in that they were very large, multinational and multiethnic. That is important. It suggests that the results, although general and correlative, are valid and that they hold across many ethnic and racial groups.
The next logical step, which I believe is ongoing in labs across many countries, is to identify the chemical components of coffee and tea that might be responsible for the results we see. Clinical trials to examine the effects of phenolics or caffeine on lipid profiles are one possible avenue of research—and much of this research is currently underway.
New biochemical technologies are emerging that allow us to assess the effect of dietary components on human metabolism (the field of metabolomics) and how these compounds interact with different genotypes (the field of nutrigenomics). The current correlative studies will be very useful in guiding the design of metabolic and nutrigenomic studies to pinpoint what compounds are metabolically active and who will gain the most from using them. I would hazard a guess that many of these studies are in progress now.
