Food packages inevitably present detailed tables purporting to give us information about the nutritional value of their contents. Unfortunately, much of this label information is incorrect or useless.
Consumers with excellent eyesight or good corrective eyeglasses might be able to read the details on a box that proclaims, say, 62.2mg of some ingredient. But food is a biological product of variable composition.
For foods that are not chemically pure (like sugar or salt), I would guess a typical variability of at least 5 per cent. So that "62.2" really means a range between 60 and 64. A label claim of "62 mg" would indicate a true value somewhere between 61 and 63, which is probably correct. On the other hand, a printed claim of "62.2mg" asserts a real value somewhere between 62.1 and 62.3. That is almost certainly a lie.
That value of "62.2" is not merely false, it takes up space in the composition tables, which are almost always in minuscule print. For consumers in a crowded supermarket, trying to compare three or four products, the mental chore of taking in four characters (including the decimal point) is twice that of reading "62". Lots of extra pain, but no gain.
Does any consumer need three-figure accuracy? We could almost manage with just "low, medium, high" values. The advantage of numerical figures is that we can set our own thresholds. For me, a food with fat content less than 10% is "low", while 62% is "high". Some readers might want much lower values, but no-one cares whether a margarine has 62.1% versus 62.2%.
Some food packages wisely declare that their composition tables are based on averages, yet they still retain the falsely precise numbers. This is particularly hilarious when "average portion size" is used. Very few people actually weigh out exactly 45 grams for a bowl of cereal? So why claim that an average portion contains precisely 33.8 grams of carbohydrate?
Label proclamations are often of dubious validity. Consider "protein". The standard laboratory way to measure protein is to boil the food with concentrated sulphuric acid, which converts all nitrogenous ingredients to ammonia. The ammonia percentage is then multiplied by 6.25 to provide a "protein" value. Intentional adulteration of food with melamine, a nitrogen-rich chemical, was done in the knowledge that it would falsely yield a high result for "protein" when analytical laboratories used the boiling-acid method.
Melamine is a lot cheaper than genuine protein, but is toxic. Describing such adulteration as "contamination", which implies non-intentional actions, let's the perpetrators off too lightly.
In New Zealand (and most of the Western world), no food manufacturer would ever add melamine to his product. Does that mean the food labels for protein are correct? Not at all. The factor to convert ammonia into protein depends on the source of the protein. Standard practice is to assume 16% nitrogen, which equates with a conversion factor of 6.25. But protein compositions vary: the correct factor can be as low as 5.26 and as high as 7.69. The accurate conversion of ammonia to protein concentration depends on whether we're talking wheat, meat, peanuts or artichokes.
There are other ways to measure protein that are not affected by adulterants such as melamine, and that don't require imprecise correction factors. Were regulatory agencies to insist on these more expensive methods, would that help ordinary consumers? No! All proteins are not created equal. Proteins that are rich in scarce amino acids such as lysine and methionine are, in many cases, more valuable than other proteins. Yet we certainly don't want still more confusing details such as "lysine content" to clutter our food labels.
Because of these limitations, we should treat food labels as providing only rough estimates of protein content. No-one should base their nutritional practices on precise calculations from imprecisely estimated "protein" values.
With a few admirable exceptions, most local (New Zealand) food manufacturers do not provide potassium levels. The missing information might change our eating habits entirely! For instance, potatoes are an excellent source of potassium, and certain brands of potato crisps sold in Australia have more than twice as much potassium as sodium.
Such products are beneficial to our health, not deleterious (if we ignore the high fat content of many chips).
Foods that only proclaim their sodium but not their potassium contents don't give consumers any chance to pick healthier foods. For instance, many high-salt potato products may well contain more than enough potassium to compensate. What, then, is the function of the composition tables? Obviously they fulfil statutory obligations. They are so technical, so filled with hard-to-read and impossibly precise numbers, that they may reassure many consumers that the manufacturers have high professional standards, but often excessive precision makes the real information harder to ferret out.
Hint: An excellent table of typical potassium (K) contents of a wide range of foods was published by the Auckland District Health Board:
But if composition tables were really meant to educate the consumer, to give useful information in a concise way, they would become considerably simpler and easier to use.
Published in The Press (Christchurch, New Zealand) on Friday, 26 September 2008. Reprinted with permission, but modified for Internet readers. My modifications are not the responsibility of The Press.
Created on ... December 02, 2008