What JAMA and the Press Said
(all passages in this column are direct quotes; italics are ours, for emphasis)

What Careful Reading of the
JAMA Article, and the Authors'
Subsequent Letter, Reveals

Relative to an estimated BAC [blood-alcohol concentration] of less than 0.02, the adjusted odds ratio of bicycling injury was 5.6 [the risk of injury increased 5.6-fold] for a BAC of 0.02 or higher.

The JAMA construct "BAC of 0.02 or higher" is so broad (from one drink to extreme intoxication) as to be meaningless: for the cases studied by the JAMA authors, blood alcohol contents ranged from zero to 0.35 - a level equivalent to eighteen drinks (!)

U.S. News: Just one drink, the [researchers] calculated, multiplied the risk of serious injury or death six times.

Wrong. Imbibing one to eighteen drinks multiplies the risk to the cyclist 5.6 times.

Reuters: The [JAMA] report added that just one drink increased the risk of a fatal or serious injury about fivefold.

See paragraph above.

NY Times: The researchers "estimate that a blood-alcohol level of 0.08 " increases the risk of a bicyclist's death by about 6 times and the risk of injury by about 20 times.

The JAMA authors found that a BAC of 0.08 or higher increases the risk of a cyclist's death six-fold, and injury 20-fold. But the "0.08 or higher" group ranges up to 0.35, making the "finding" meaningless, not to mention misleading.

Reuters, quoting JAMA lead author Guohua Li: "Riding a bike requires a higher level of psychomotor skills and physical coordination than driving a car, so alcohol has an even stronger effect on bicyclists than [on] drivers."

In their letter, the authors concede that BAC's of 0.02 to 0.07 (one to several drinks) are associated with an increased risk of only 40%. This appears to be on a par with the added risk of driving after one to several drinks. Of course, a drunk cyclist puts only himself at risk, whereas a drunk driver poses a threat to others.

U.S. News: Roughly 800 bikers die each year on the roads, "usually when they tangle with cars."

An interesting formulation. Note that it is the cyclist who tangles with the car, not the car that hits the cyclist. In reality, driver speeding, passing too closely, failure to yield, etc. are the principal causes of cyclist fatalities in NYC (where we have studied the question) and probably elsewhere as well.

Blood Alcohol Concentration (BAC) Guide (figures are approximate):

0.02 = 1 drink

0.08 = 4 drinks and is the legal threshold for intoxication in some states

0.10 = 5 drinks and is the legal threshold for intoxication in other states

0.20 = 10 drinks and is at least twice the legal threshold for intoxication

0.35 = 17-18 drinks (that a person with this BAC could cycle is remarkable!)

See Right Of Way's The Only Good Cyclist, for documentation of our finding that driver misconduct outweighs cyclist error 3-fold as the principal cause of bicyclist fatalities in New York City. Motorists' improper turns, red-light running, speeding, and unsafe passing each kill more cyclists than do cyclist red-light running and wrong-way riding.

Charles Komanoff may be reached at kea@igc.org