Beyond Natural | Food
How to decode a food label
The information on a processed food label may seem obvious, but there's often more to decipher than first appears.
In the 19th Century, food shopping was a gamble. Tins, packets and containers tended to lack ingredient labels, let alone nutrition information. With no obligation to tell people what was in their products or what it meant for their health, manufacturers inserted all sorts of unpleasant substances.
As the journalist Deborah Blum writes in her book The Poison Squad, milk sellers in the US once added chalk, plaster dust, or dye to make their watered-down, bacteria-ridden batches look more palatable. Other food-makers sprinkled copper sulphate – a garden pesticide that can burn the skin – into tinned vegetables to make them appear greener. To extend shelf life, some manufacturers even added formaldehyde or borax – a laundry detergent – to meat and dairy products. Meanwhile in England, arsenic was used to colour green sugary sweets, while lead was added to red or yellow ones, as well as for colouring cheese.
To force more stringent labelling regulations on a reluctant food industry, it took decades of lobbying and research. To bolster the case, one US scientist even staged a controlled experiment to feed meals of contaminated food to a group of willing young men – Blum's "poison squad" – to demonstrate that it could harm their health. Many of the squad's healthy volunteers fell ill, a consequence that US lawmakers could no longer ignore.
At the turn of the 20th Century, adulterated milk was a concern in France as well as the US (Credit: Getty Images)
More than a century later, processed foods now come packaged with an abundance of information about their content. But that doesn't mean that this labelling is always easy to decipher. Peer closer at a packet or box, and you'll find various chemicals, codes, weights and percentages, while the front features statements and health claims that are not as straightforward as they first appear.Are there any tricks to deciphering a food label?
You may also like:
- How processed food changed our species
- Can processed food be better for you?
- What were the first processed foods?
To begin, it's worth acknowledging that food regulations differ around the world, so it would require more than a single article to describe every country's labelling conventions. The UN and WHO's Codex Alimentarius provides an international set of labelling standards, but this "food code" is voluntary, and is applied in different ways around the world.When it comes to nutrition labelling, most of the world's major economies make it mandatory, including the US, India, China, Japan, Australia and EU members. But for some it is voluntary unless a health claim is made, such as in Turkey, Singapore or South Africa.
Generally though, most labels feature an ingredient list and some information about the product's nutritional value: calories, fat, sugars, salt and so on. In recent decades, manufacturers have also started adding health and wellbeing benefits, aware that it helps sell their products.
All clear and obvious? Not quite.
For starters, consider the ingredient list. What manufacturers don't spell out on the front is that these are listed from most predominant to least predominant by weight. So far, so obvious, you might say. But it does lead to some labelling sleight of hand. For instance, if you look closer at a hazelnut spread label, past the pictures of nuts on the front, you'll notice that the first (and therefore biggest) ingredient is actually sugar, often followed by oil. You can find a similar pattern with breakfast cereal: even if a box advertises "wholegrain wheat" in big letters on the front, the ingredients often list sugar in a close second place. Some brands of frosted flakes, for example, have 37g of sugar for every 100g of cereal. That's roughly the same ratio as in a chocolate chip cookie.
A few, like the artificial colour E122 in cakes and sweets, may have adverse effects on children prone to hyperactivity
Lower down the ingredient list, you'll also come across names that are less recognisable than oats, sugar or nuts. In the EU, manufacturers use a system of short codes to describe additives called "E numbers", which over the years have acquired a controversial– and occasionally undeserved – reputation as dangerous and mysterious chemicals. A few, like the artificial colour E122 in cakes and sweets, may have adverse effects on children prone to hyperactivity. But others are good for you, or at least harmless: E300 is vitamin C, E948 is oxygen and E160c is paprika.
In the US, there is no such coding, and such additives are described with their chemical name. So, on a US label you'd read "sodium caseinate", rather than "E469". On the surface, that would seem clearer, but even that convention is a little obtuse about what the stuff actually is: sodium caseinate is used in foods like sausages or bread, and is the main ingredient in coffee creamer, but neither its E number nor its chemical name would tell you that it's a protein derived from milk.
Another sodium-based difference between countries to be aware of is that the US lists sodium levels on its products (specifically its nutritional labels, which we'll cover next), whereas the EU lists salt. Salt may be a type of sodium, but sodium is a category that also includes the caseinate additive we just mentioned, as well as other ingredients like bicarbonate of soda.
As your eye roves around a processed food packet, you will probably also come across some form of nutritional information too.
Some nations, like the UK, have a traffic light system for nutrition that expresses how healthy a processed food is in terms of fat, saturates, sugars and salt, using the colours red, amber and green. For example, a processed oven meal might have 7.7g of saturated fat, and so be labelled red. In some (but not all) cases, it also comes with a percentage, in this case 39%. The colour scheme was designed to be intuitively easy to understand, but how the percentages are worked out may not be immediately obvious. The 39% in that meal is calculated using the "Reference Intake", which is the maximum recommended amount. In Europe, this value has gradually been replacing "Guideline Daily Amounts" (GDA) on labelling, which differed by gender and age.
The traffic light system, which appears on the front of UK food, offers a quick and clear message for nutritional value (Credit: Alamy)
Unlike GDAs, the Reference Intakes are single figures – 90g sugars or 20g of saturated fat, for instance. What is perhaps less obvious to the average consumer is that they are the recommended maximum intake for an adult woman who does an average amount of physical activity. If the Reference Intake was based on the nutritional maximums recommended for the average adult man, which are sometimes higher, then it would risk women unwittingly eating too much.
So, if the saturated fat in the meal pictured above had been labelled based on UK health guidelines for men aged 19-64 (which is no more than 30g per day), the percentage would actually be more like 26% – clearly that's still not healthy, but it's lower.
However, it's not like British men are reading these labels and undereating – it's the opposite. And one recent survey found that a third of British men say they ignore nutritional labels, compared to one in six women.
The Nutrition Facts label was one of the most ambitious public health initiatives ever undertaken by the FDA
Decisions about how labels should account for sex differences – along with various other dilemmas – also played out in the US a few decades ago, when the Food and Drink Administration was designing its ubiquitous "Nutrition Facts", a label that legislators in 1990 mandated should appear on pretty much all food packaging in the US.
The US Nutrition Facts label was updated in 2016, but has been mandatory on foods since the 1990s (Credit: FDA)
Navigating opposition from foodmakers and politicians, the Nutrition Facts label was "one of the most ambitious public health initiatives ever undertaken by the FDA", its designers later wrote. It went through various iterations, and was scientifically tested on consumers to see how well they read and understood it.
"Nothing in the legislation specified what the label should look like, but we knew that it had to employ the science of graphic design to nudge behaviour. In the lingo of marketers, it had to 'pop'," wrote David Kessler, former commissioner of the FDA, in 2014. This included details such as black and white colouring to provide strong contrast, eliminating punctuation marks to improve readability, and even fiddling with the specific distance between the lines to ensure clarity.
Below you can see the design evolution up to the 1990s version, experimenting with pie charts, bars and other graphical presentations:
The pre-1990s label (top left), followed by four potential designs, and the one eventually approved (bottom right) (Credit: Jerold Mande/FDA/Burkey Belser/Erika Ritzer)
The FDA team also had to make a series of pivotal decisions about how and what nutrition information to display. The US version of Reference Intake was dubbed "Daily Value", and like in Europe, the FDA had to balance the needs of men and women. In the early 1990s, the recommendations of caloric intake for adults ranged from 1,900 calories per day for women over 51 to 3,000 for men 15-18 years of age. At first, the FDA toyed with the idea of using a population-weighted average of 2,350 calories, but they settled on 2,000 calories, arguing that overconsumption in women was the greater risk.
Another major set of decisions for the FDA involved the official definitions of "serving size", which would seem obvious at first, but is actually complex to define. Until the Nutrition Facts label arrived, each manufacturer got to decide how big a portion was. This led to all sorts of trickery: a cake could be presented as "light" by simply cutting the serving size on the label, rather than the fat or sugar content.
A cake could be presented as "light" by simply cutting the serving size on the label, rather than the fat or sugar content
The 1990 legislation had mandated that serving size should be the amount "customarily consumed and which is expressed in a common household measure that is appropriate to the food".But as the FDA team soon discovered, calculating serving sizes for the full smorgasbord of American food was quite the task, involving myriad national surveys to establish how much people ate.
So, for example, the serving size for cookies was fixed at 30g, or thereabouts (so two 18g cookies would still equal a serving size). Some categories, though, provided the researchers with a real challenge: for instance, they couldn't make a single serving rule for "breakfast cereal", because granola, for instance, is much heavier than cornflakes.
In the end, they defined 139 categories, which includes four types of cereal, four types of cheese, and a host of definitions for desserts and cakes, from "heavyweight cake" to "lightweight cake". (Cake frosting, incidentally, also has its own category, if you are partial to that – and it's two tablespoons.)
The US serving size for ice cream was revised recently, to reflect that Americans were eating larger portions (Credit: Carlo Allegri/Getty Images)
The US serving sizes changed again in 2016, notably for ice cream and for soda. Either the original estimates were inaccurate, or Americans now eat more ice cream and drink more fizzy drinks. Either way, the FDA changed a serving of ice cream from half a cup to two-thirds (one cup of ice cream is around 150g), and ditched the reference amount of 8oz (227ml) for a serving of soda. The 2016 change also accommodated the fact that bottle sizes are now bigger: so whether you pick up a 12oz (355ml) or 20oz (568ml) of sugary soda, both will be labelled as a single serving.
Crucially, if it's not already clear by now, what you ought to know when you see the words "serving size" on a label is that it is not a recommended amount, rather it is the typical portion that the average American actually consumes – and that's not always an amount that adds up to a healthy diet.
Health and nutrition claims
A final piece of label decoding involves health claims, which are regulated to avoid manufacturers over-promising. Laws differ internationally, but in the US, there are three types of claim that are relevant: "authorised", "qualified", and "structure-function" claims. Knowing which is which can help to gauge how much weight to give the claim when making diet choices.
An example of an authorised claim, on a low sodium packet, might be: "Diets low in sodium may reduce the risk of high blood pressure, a disease associated with many factors". To have this approved by the FDA, there must be "significant scientific agreement".
A qualified health claim has more nuance, reflecting the fact that the science is not settled. It might go something like: "Scientific evidence suggests, but does not prove, that whole grains, as part of a low-saturated-fat, low-cholesterol diet, may reduce the risk of diabetes".
Finally, a structure-function claim is intentionally vague, such as "calcium builds strong bones". The key with these assertions is that they don't (and can't) make any specific promises to improve health, and should be viewed only as marketing language.
A range of breakfast cereals aimed at children in Spain (Credit: Getty Images)
Using typography tricks and small print, many manufacturers – in the US and all over the world – try to amp up health benefits and fade back downsides. In one recent study of breakfast cereal in Australia, for example, 83% of children's cereals touted a nutrition or health claim, such as "high fibre" or "low salt", yet a majority were rated as unhealthy or nutritionally poor overall. Among other notorious examples on the supermarket shelf are the "low fat" yoghurts that should really be labelled "high sugar".
No matter which country you live in, then, it's prudent to do some background research about what you see on a label. While it might seem comprehensive and detailed, there's often more to the health claims, ingredients or nutritional information than can be displayed on the packet.The labelling of foods has come a long way since the days milk contained formaldehyde and candy came with arsenic – but that doesn't mean that a label tells you everything.
* Richard Fisher is a senior journalist for BBC Future. Twitter: @rifish.
Deborah Blum was Fisher's supervisor during a fellowship he completed between 2019-20 at Massachusetts Institute of Technology.
If you liked this story,sign up for the weekly bbc.com features newsletter, called “The Essential List”. A handpicked selection of stories from BBC Future, Culture, Worklife, and Travel, delivered to your inbox every Friday.