Researchers from the University of California at Berkeley argue in an article published yesterday in the British Medical Journal recommending the "Parachute approach to evidence based medicine ," that there are times when it is unethical to wait for controlled trials of a health intervention. People die before the intervention is tested, they say, citing as their most impressive evidence the case of Oral Rehydration Therapy (ORT). They explain:

In 1980 childhood diarrhoea was killing an estimated 4.6 million children annually. Treatment with an intravenous drip is life saving but requires health facilities. Studies from 1977 onwards showed that infant diarrhoea could be treated with oral rehydration. The World Health Organization initiated a highly successful programme of oral replacement therapy in 1981 after it became obvious that the treatment saved lives and no alternative home based treatment was possible. Randomised controlled trials were later conducted in health facilities, confirming that oral replacement therapy was as effective as intravenous therapy. The initiation of large scale programmes for oral replacement therapy before the randomised trials meant that by 2000 there were three million fewer deaths from diarrhoea annually.

Treating diarrhea with a mixture of salt, sugar and water has, indeed, saved millions of lives. The authors advocate the "parachute" principle, explaining:

Evidence based medicine and randomised controlled trials are not synonymous. The parachute approach can be the most appropriate, especially in situations of high mortality and low resources, when a simple intervention can have a large impact. Randomised controlled trials are essential in many other settings and they have defined many life saving strategies and corrected some important mistakes. They are often needed when mortality has reached a low level because new treatments require large investment for relatively small improvements in therapy that may be difficult to distinguish.

All that may be well and good, but BMJ has also editorialized that they feel confident in recommending universal sodium reduction. Perhaps a caveat is needed before we call in the airborne assault on salt. In the case of ORT, the small scale studies produced strong consensus of a health benefit. While there is no such concensus that reducing salt would reduce blood pressure to a meaningfull degree, because the effect is so heterogeneous, the UC-Berkeley authors and BMJ editorial staff should be chastened by the consensus of the observational studies of low-salt diets -- there is no benefit in terms of reducing heart attacks or reducing mortality.

Let's not "parachute" in to reduce dietary salt -- at least until the clouds clear and we can see the landing area.

Used to be all dietitians would chant the mantra of varied diets where all foods could find a place. "We need to focus on the question of "good diets" and "bad diets," they'd say, and avoid labeling individual foods as "good foods" or "bad foods."

The science is unchanged, but the mantra is gone. When nutrition experts gather these days, they're shaking their heads and wringing their hands: the public doesn't "get it." While consumers pretty well understand the concepts of the dietary guidelines, they don't buy into them in terms of personal eating decisions. What to do?

Demonize foods, say some like the Center for Science in the Public Interest which has pushed the "good food/bad food" dichotomy for 30 years. Make people feel that the foods they eat are poisoning them. Ostracize foods with "bad" nutrients and limit diet choices to "good foods" with plentiful "good" nutrients.

The food industry is buying into the "good foods/bad foods" story too -- for marketing reasons. Food companies want to deliver what their customers want. If you can put a "healthy" label on your foods, it makes a difference in product placement and sales -- if you can make it taste good!

A new study reported by the European Food Information Council sums up this way:

There is widespread interest for nutrition information on food packages. Consumers generally understand the link between food and health, and many are interested in using information about the nutritional properties of the food they eat. However, the degree of interest differs between consumers and varies across situations and products. In addition, it can conflict with other interests in food, notably taste, traditional eating, and indulgence.

Consumers like the idea of simplified front-of-pack information but differ in their liking for the various formats. These include health logos, 'traffic lights', GDA-based systems and energy labels. Differences can be related to conflicting preferences for ease of use, being fully informed, and not being pressurised into behaving in a particular way. For example, many consumers like colour coding, but some regard reds and greens on food products as too coercive.

Most consumers understand the most common signposting formats in the sense that they themselves believe that they understand them and they can replay key information presented to them in an experimental situation.

There is still virtually no insight into how labelling information is, or will be, used in a real world shopping situation, and how it will affect consumers' dietary patterns.

The real question is will food buyers follow the red-yellow-green stoplight the same way they follow traffic speed limits -- by applying their own judgment in the absence of an officer writing speeding tickets? As EUFIC points out: "There is still virtually no insight into how labelling information is, or will be, used in a real world shopping situation." Are we ready to buy another set of unintended consequences?