February 11, 2007

What Is The Best and Healthiest Coffee To Drink?

A short digression on my fourth favorite subject.

Does coffee raise blood pressure?  Does coffee elevate cholesterol?  Does coffee hurt your liver?  Does coffee taste delicious?

At the outset, you need to know that not all coffee preparations are the same.   The diterpenes cafestol and kahweol are the alleged cuplrits in the negative effects of coffee, especially raising cholesterol and increasing risk of coronoary artery disease.  However, these are lipid soluble and are almost entirely filtered out by paper filters. Mix coffee grounds and water in a pot, and boil.  Pour off the cofffee into a glass.  Drink.  Now look at the glass.  That oily residue is-- well, oily residue.  You don't get that with a filter.

For example, here is the breakdown of lipids in coffee: filtered coffee: 7mg/cup.  Boiled and  unfiltered (Turkish): 60-160mg/cup.  Metal screener (french press): 50mg/cup.  The types of lipids in each were the same (no selectivity in lipid filtration.)   So how you make your coffee matters.

Blood pressure: as long as you're a regular drinker, don't worry.

Reports of coffee elevating blood pressure are misleading, because they aren't done they way we drink coffee: daily.  Going from nothing to a triple espresso raises blood pressure; but chronic coffee drinking eventually allows for normalized blood pressure.

For example, the much repeated finding "unfiltered boiled coffee causes a significant elevation in blood pressure, especially in women" is misleading:  the study actually found that if you switch exclusively to boiled unfltered coffee from filtered coffe, your systolic blood pressure rises about 4mm Hg.  However, switching from filtered coffee to abstinence did not have any effect on blood pressure or heart rate.  Another study found a trivial change in blood pressure (-3.4mm Hg) after two months of abstinence (afgter 5 cups/d.)

Interestingly, a metaanalysis of 16 studies found that chronic caffeine (400mg/d) raised systolic blood pressure by 4 mm Hg. while 5 cups coffee/d (>500mg caffeine ) only raised it 1.2mm Hg.)  This was corroborated by another study finding >5 cups lead to 1.35mm Hg increase.


Cholesterol: raised slightly by unfiltered coffee, and possibly with filtered. 

Initial reports had found that drinking unfiltered coffee was associated with higher triglycerides and cholesterol levels than filtered coffee, because the filter removed almost all (80%) of the causative substance.  Another study found unfiltered caused higher cholesterol (but not TG) than filtered; filtered coffee had no effect on lipids over no coffee at all.

These findings were slightly contradicted in a recent study: Abstaining after 4 cups/d reduced cholesterol by about 12mg/dl.  Drinking filtered coffee raised cholesterol by about 11 mg/dL.  For perspective, 4 cups/d of whole milk would raise cholesterol by about 14mg/dL.

The question t ask here would be, how good was the paper filter? 

Coronary/heart disease: no. 

Retrospective analyses find that >4 cups/d, but not <2/d, had almost double the risk of coronary disease;  however, prospective studies found no increased risk.

A review identified possible explanations for an increased risk of heart disease in coffee drinkers  including a genetic predisposition to slower caffeine metabolism in some people, and the presence of diterpenes (which raise cholesterol) in unfiltered coffee .  However, the same review found several studies indicating a protective effect of moderate coffee drinking, which they conclude is related to the  antioxidants.

One study found heart attacks more frequent in coffee drinking women than abstainers: but was only usefully relevant at <7 cups/d, which doubled the heart attack risk.   However, a gigantic 85000 middle aged women prospective 10 year study found no effect of 6 or more cups coffee/d on coronary heart disease. 

For what it's worth, unfiltered coffee seems to be associated with higher rates of heart attacks than filtered. 

But it pays to wait: a week after I initially posted this, an 8 year  prospective study in the elderly found a dose dependent (i.e. greatest >4 servings) protective effect of caffeine in cardivascular mortality (reduced by 50%) (but, oddly, no effect on cerebrovascular mortality).  Importantly, these were normotensive individuals.


Suicide: Opposite of smoking: drink up.

Gigantic 10 year prospective study of 86626 female middle aged nurses:  suicide rate was reduced by 60-70% in those who drank more than 3 coffees/d, all other factors controlled.

A Finnish study of 43000 people over 14 years-- 216 suicides-- found that 2-5 cups/d moderately (30%) reduces suicide risk, while >8cups increases risk 1.5 times.  (For reference: "heavy drinking" (weirdly: 2 drinks/d) or smoking had about the same risk.)

A 1993 study looking at death from any cause found a reduction in suicide risk (RR 0.87 per cup) with increasing coffee. 

Liver cancer and cirrhosis: can't hurt, may help, especially if you're an alcoholic. 

The same 1993 study above also found a lower risk of cirrhosis (RR 0.77/cup).  The same authors, in a more recent study, again find such a reduction in risk, and find lower levels of liver enzymes ALT and AST. An Italian study found coffee reduced the risk of hepatocellular carcinoma from any cause (Hep B, C, alcohol, etc); same in the Japanese, and in the Japanese in a prosepctive trial.  And in Americans chronic liver disease rates were half in 2 cups/d drinkers.

Recent evidence suggests that this may be partly due to caffeine, but also to phenolic acid antioxidants which are not present in tea.  The authors cite reports of such ingredients' protective effect sagainst various forms of liver damage (including Tylenol.)  

So if you're going to drink coffee, there are two prudent things to do.  1) drink filtered coffee, made with a good filter.  2) drink medium roast, not dark roast.  The roasting process burns off volatile chemicals such as caffeine and the antioxidants. 

For this reason, my vote for best, healthiest, and most delicious coffee to drink is Dunkin' Donuts.