Lactose: How Low Is Zero?

The Advertising Standards Authority in the UK gets to judge advertising claims and their truthfulness. In April, they had to judge a case involving lactose, as shown in this summary by Susan Barty and Susie Carr on Mondaq.com.

Arla Foods Ltd, 23 April 2008

A TV ad, for Lactofree milk, stated "Lactofree, the full taste of real milk, just without the lactose". Text on the packaging stated "with less than 0.05%" lactose.

Complaint / Decision

Two viewers claimed that the ad was misleading and potentially harmful, because they considered that the product contained 0.05% lactose.

Complaint not upheld: The advertiser produced UK-accredited tests, which it had undertaken, to demonstrate that Lactofree contained 0% lactose but at a detection level of 0.05%. The ASA understood that it was generally accepted that the lowest amount of lactose that could trigger symptoms in an individual who was lactose intolerant was 5g. Lactofree was well below the 5g margin and was therefore unlikely to have any effect on someone who suffered from lactose intolerance the ASA concluded that the ad was unlikely to mislead or cause harm to viewers.

That summary confused me the first time I read it, so let's take it apart piece by piece.

First, we have to look at how nutritional information is presented on packaging. Information has two parts: accuracy and precision. Recipes, even for standardized commercial goods, are hard to perfectly regulate. One batch may have slightly varying amounts of fats or vitamins or calories than another. That makes accuracy an issue, with the amount of any individual item usually an average taken over many test batches. Say that the average amount of fat in a cookie is 5 grams. How precise is that number? Is it really 5.0 grams? 5.019093873975499383983983900498409808 grams? 4.99976 grams? There's a limit to how many decimal places can be measured by the best equipment, and no need to use the best equipment in the first place. In practice nobody can tell the difference between those three amounts. Trying for that level of precision is silly and confusing. In the U.S. amounts are always rounded to the nearest half gram.

What happens, then, if the amount is less than a half (0.5) gram? In the U.S. manufacturers are allowed to round down to 0. That's right, many times when you see a listing of 0 grams per serving of an item you can't assume that nothing is there. A famous example occurs when cheeses prominently list 0 grams per serving of lactose. There may truly be 0 grams. But there also may be, say, 0.4 grams. For a one-ounce serving that's 1.4% lactose. Not much, but not insignificant.

Other types of rounding are also allowed. A product with fewer than 5 calories per serving, say, a sugar substitute, is legally allowed to bill itself in the U.S. as having 0 calories when it really might have a calorie or four.

In the U.S., lactose has a further twist. You can report 0 grams of lactose per serving and not have zero lactose, but if you use the words "lactose free" then it better be 100% free of lactose. Or at least 100% free within the precision of whatever testing equipment is used.

That brings us back to Lactofree. In the U.K., apparently, the testing precision for lactose in milk allows for up to a 0.05% variance.

That's not much. A liter of cow's milk has about 40 grams of lactose, so 0.05% of that is 0.02 gram (or about 0.005 gram in a standard 8-ounce American glass). Absolutely minimal.

The ASA said that was within standards and too low to affect even those who were lactose intolerant. They dismissed the complaint.

I agree that this is a reasonable ruling. One sentence does worry me, though. "it was generally accepted that the lowest amount of lactose that could trigger symptoms in an individual who was lactose intolerant was 5g." Really? That's saying that anybody can have 1/8 of a liter (about three ounces) of milk without symptoms. Most people can, true. But everybody? I'd like to know where they got that standard. It's nothing I've ever seen before.

I accept that Lactofree milk has as little remaining lactose in it as their manufacturing process can make it. It's not as clear whether little is the same as zero or just equal to "very small." A very tiny number of people who say they can't have even the slightest amount of lactose might want to stay away from Lactofree. On the other hand, if your symptoms aren't too bad you might want to take a test drink and see whether the "not even the slightest amount" clause is really true or whether you're underestimating your capacity.

The broader lesson for all of us is that just reading the nutrition numbers on the side of a package is nowhere near sufficient. You have to research deeper to understand where the numbers come from, what they mean, and what variants are legally allowed. You can bet that the manufacturers know the rules down to the last letter and are manipulating them to their advantage every chance they get. You have to be as smart as they are. Never let up your guard. They are always trying to slip a new twist past us. Don't ever let them.

If you have questions about what the numbers on packaging mean, drop me a line at stevecarper@cs.com. I'll try my best to sort the numbers out from a to z.