Welcome to DrPHealth

Please leave comments and stimulate dialogue. For those wanting a bit more privacy or information, email drphealth@gmail.com. Comments will be posted unless they promote specific products or services, or contain inappropriate material or wording. Twitter @drphealth.

Thursday, 7 June 2012

Coffee, Tea or ??

Products that are widely consumed are excellent targets for health claims, but the very nature of consumption patterns make epidemiological investigation challenging, and if absolute associations are weak, then identification of potential benefit or risks can be a lifelong career.  The identification of the health effects of tea and coffee are prime examples.  

Since the mid-1970’s numerous publications have sought the definitive relationship, and many of these studies were published in no less an esteemed journal as the New England Journal of Medicine.  The wonderful series of publications makes an excellent course in critical appraisal of health publications as conflicting results appear to arise from similarly executed investigations. The major challenge being the association between stimulate consumption and certain other behaviours such as smoking, caloric intake, alcohol consumption and socioeconomic status. 

The latest in the series of hundreds of North American and European coffee studies is also published in this esteemed journal and looks at some 5 million person-years of follow-up and death events in 33,000 men and 19,000 women.   NEJM study on coffee and mortality

Tea drinkers fear not, the literature is just as rife with more interest from predominately tea drinking countries.  The subject received a rigorous review that explored some 40 studies also covering a few million person-years of exposure and a variety of epidemiological study methods.  The analysis was more specifically focused on cardiovascular disease and diabetes outcomes  tea and health outcomes review article.  

So the results, or at least what is available and comparable.

Coffee - male
Coffee - female
Tea (black)
Unadjusted relationship
>3 cups associated with up to 60% increase in mortality
Similar to males with up to 50% increase in high consumption

Adjustment for identified associated variables (confounders)
Up to a 10% reduction in mortality noted at 2 or more cups per day
More benefit up to 16% and also in groups of >2 cups/day

Cancer adjusted for confounders
Slight increase of 4 and 8% respectively for 4-5 cups and >6 cups
Non-statistical differences.

Heart disease
Reduction of 7-14% with greatest benefit at 2-3 cups per day
Greater benefit of up to 28%
Several studies showing up to 40% reductions in cardiovascular death.  Benefit to women also seems slightly greater
Respiratory disease
Dose response relationship with up to 19% reduction at >6 cups
Similar positive benefit of up to 35%

A notable reduction of up to 30% in all consumption ranges except for unadjusted >6 cups per day
Benefit only noted at >2 cups per day
Numerous studies reporting benefit of up to 40% reduction with greater benefit for women
An apparent benefit for all consumption groups, adjusted and unadjusted
Similar broad benefits of up to 43% noted
Benefits in the range of up to 40% noted.  

Perhaps even more reflective of the great debate globally, no study mentions adjusting for consumption the other common beverage.  The assumption that both coffee and tea drinkers are exclusive in their habits is reflective of the biases within the literature.  Also notable is the extensive discussion of what constituents of tea are perceived as beneficial, with the absence of such debate in the coffee literature. 

Such investigations will continue to populate the medical literature for decades to come until someone convinces a naive group of youth to be randomly allocated to cohorts of coffee, tea or neither and subject them to decades without allowing for their choice of stimulant.  Stated otherwise, the question and the subject will continue to percolate and simmer. 

1 comment:

  1. I've seen a number of studies that adjust for genetic variabilities in caffeine metabolism that suggest that fast metabolizers may benefit from some caffeine (re: stroke, hypertension) and slow metabolizers may be at risk above 1 cup of coffee (or caffeine equivalent) for stroke, hypertension, anxiety. This seems to jive better with our "common sense" than studies that do not adjust for genetic variables. Basically, if you can't sleep easily if you have coffee after noon, then you are a slow metabolizer and should be careful.

    I would love to see pregnancy studies that include this genetic variable.