Here are two questions I would have gotten wrong before the release of a new Institute of Medicine report on sodium intake and health:

1. True or False: The American Heart Association's recommendation of 1,500 mg/day is still quite salty compared to diets our ancestors ate.

2. True or False: The average intake of 3,400 mg/day is skyrocketing over the consumption of a few decades ago.

Both are false, according to Brian Strom, a Penn professor of Public Health and Preventative Medicine and chair of the Institute of Medicine report.

The 1500 mg the heart association recommends, he said, is actually the minimal amount of salt it's possible to consume and still eat enough food to maintain yourself. That's because sodium is infused in food even if nobody adds any salt at all. We need some salt. So why do the heart association and other groups, such as the Center for Science in the Public Interest, recommend such a drastically low intake?

Strom says it's based on an inference. Lowering salt up to a point lowers blood pressure and lower blood pressure, up to a point, reduces heart attack and stroke risk. So the doctors who recommend that 1500 mg limit are extrapolating those connections to assume that the less salt you eat the better. But there's no actual data showing any direct benefit from going that low.

As for the second question, Strom learned that people ate the same 3400 mg of salt 50 years ago as they do now. Another interesting fact is that the major source of salt in the American diet is bread – not because bread is salty but because we eat so much of it, and all our food has some salt.
Salt is one place where scientific consensus is falling apart.

The Institute of Medicine panel examined the data we do have on salt and health. They narrowed that down to 25 relevant scientific papers. They found no evidence that lowering salt below about 2300 mg a day would benefit anyone's health and some suggestions it might cause harm. The harm is hard to asses, he said, because it could be being sick causes people to stop eating which would cause them to get much less sodium.

There are a number of limitations in our data on salt and health. Food diaries are notoriously inaccurate. You can measure people's sodium by collecting all their urine, but there's only so long people will subject themselves to this process and it's long term rather than short-term consumption that matters.

In talking to Strom, I left realizing that there's much we don't know. There's evidence that very salty diets might contribute to high blood pressure and therefore to heart disease, but it's not clear how much is too much. And who should be worried.

And in this case, the perfect may well be the enemy of the good. I asked Strom what he aims for in his own diet. He said his wife is a vegan and prepares the food in his household and that they eat about as low-salt a diet as may be possible – which adds up to about 2300 mg. If people who eat home cooked vegan dishes can't get it lower than 2300 mg, who can?

The Heart Association advocates haven't backed down. In a letter to the New York Times this week, Chief Executive Nancy Brown insisted that indeed the evidence supports their continued recommendation that we all try for 1500 mg a day – which equates to as little salt as humanly possible.

Americans seem to take an almost religious attitude toward certain health-related practices and prohibitions. No matter what studies say, some people insist that all women over 40 must get a mammogram every year. Others assumed for ages that all fats are evil. There's a pervasive belief we can all live forever if we eat right.

The take away lesson for me is that we need to ask the medical establishment harder questions about how they arrive at the recommendations they do. Whether we're talking about cancer screening or dieting or an hour of cardio exercise, calcium supplements, or vitamin pills, there should be some evidence that it will have the promised benefits and do no harm.