We have been calculating nutrition panels for just short of 10 years. As a food technologist consultant, this is everyday work. In that time, we have never sent a product away to a lab to get tested for a nutrition panel.
Most nutrition panels can be calculated. The nutrition calculations aren’t often simple, but a test costs about $400. If you have a range of products, then the price can become considerable very quickly. Furthermore, if you change your recipe, you will need to retest. If you have calculated your values, then a formulation change may only require a quick update to the calculations.
There are a few situations where calculations aren’t an option. Three of these situations are below.
Separation of Components
In most food products (e.g. cake), you start with a recipe, combine everything, and out comes a product. You know what went into the formulation and can weigh the start and finish weight to determine if there was any moisture loss. For instance, as cake cooks, it gives off a little steam.
Some products don’t get made this way though.
Consider almond milk. You soak the almonds, macerate them, then filter out the solids. What’s left is something that looks milky. A calculation won’t be an option because you know what you have added (water and almonds), but you don’t know the characteristics of what you removed accurately.
Cheesemaking or tofu manufacturing are two other examples of this type of situation because both release liquid such as whey.
Still, in reality, all three of these items have table data that you can use, which makes lab results not necessary – but may desirable and a lot more accurate.
There is a situation with cheesemaking that might call for lab data. Let us say you have developed a very low-fat cheese, and the actual fat level is much lower than table data. You would then want to use a lab to validate your low-fat claim.
Fermentation
If you ferment your product, then it is likely that some of the carbs and sugars will convert into something else like alcohol or a food acid. You won’t know the final sugar content and data tables won’t be much use here. Examples include beer, kombucha, yoghurt, kimchi and sauerkraut – these all need a lab test for accurate results (yoghurt may have table data).
Ingredient Preparation
I’ve never come across a situation where this issue cannot be resolved somehow, but it is worth mentioning.
When you look at the data tables, there are some precise descriptions. Take almonds for example – they may be raw, blanched or roasted, skin on or off, salted or unsalted, and the roasting might be dry roasting or with oil. Your exact combination may not exist in the tables. As the number of factors increases, the number of factor combinations increases exponentially. I’ve never come across a situation where this issue hasn’t been resolved, but it is worth mentioning.
When Should You Use a Lab When Table Data is Available?
Table values for cheese are really useful when adding it to a formulation. It probably isn’t overly accurate, but if it is only 10% of a formulation, that error becomes less significant. If you are using table data for any type of cheese, it may be wildly inaccurate and shouldn’t be used.
Cost of Calculations
For most other situations, we can calculate and format a panel much quicker and cheaper than a lab. A simple nutrition panel is usually about 30 minutes worth of work, with the more complicated ones with multiple components and layering taking up to an hour.
Doesn’t sound right to say food product twice, but I can’t think of another word.