Artificial food colorings and additives in ADHD
I get the question somewhat routinely – “Does Red 40 really make ADHD worse?” The controversy around Red 40 and other food additives is certainly not new. Artificial food colorings have been around for centuries, but their use (and consumption) has grown dramatically over the past 7 decades. According to one study, in 1950, people averaged around 12 mgs per day of estimated food dye consumption. In 2010, that amount quintupled, to around 60 mgs per day. Color additives are nearly everywhere, and different countries have different rules about their use. For the purposes of this review, though, we will focus on the major artificial (not plant based) food additives used in the U.S.
What does the US Food & Drug Administration (FDA) say about color additives?
The FDA regulates color additives used in the US, and these include dyes used in food (and dietary supplements), drugs, cosmetics, and medical devices. The FDA uses a standard of evidence called “reasonable certainty of no harm” to determine if a product can be used. If the FDA establishes that “reasonable certainty” criterion after reviewing safety data, they approve the product for its intended purpose in public use.
There are two main categories on FDA’s list of permitted color additives: certifiable and exempt. “Certifiable” color additives are man-made, usually derived from petroleum and coal sources. After a product is approved, it must then go through “certification,” where the manufacturer sends a batch to the FDA, which the FDA then analyses to determine it meets requirements for composition and purity. If it does, then the product gets “certified” and can then be used legally. Certified color names have a special prefix, such as FD&C, D&C, or Ext. D&C followed by a color, and then a number. So, FD&C Yellow No. 6, often found in cereals, ice cream, and baked goods, is an example of a certified color. The full name for Red 40, the color I am most frequently asked about, is FD&C Red 40.
“Exempt” products are colors obtained from plant, animal, or mineral sources, and can include caramel color and grape color extract. These do not have to go through the certification process, even though they are still artificial color additives and thus still must comply with FDA regulations. A few years ago, when I worked as a researcher, I was the principal investigator on a study for a medical device. In my case, the device I was researching was placed in the “Exempt” category by FDA. What this meant was that I still had to follow all FDA rules in my study for labeling my device, reporting participants, and filing paperwork. Basically, it meant a LOT more paperwork for me as the researcher. The regulations weren’t easy. So, I can tell you from personal experience, exempt does not mean the product has not been rigorously reviewed. It just means the review is slightly less rigorous than the “certified” process.
The FDA states that reactions to color additives are rare. However, not everyone agrees with this statement (as we will see in a moment). Some people have an allergy to color additives. For example, Yellow No. 5 is a widely used additive that may cause itching and hives in some people. That’s why the FDA requires all products containing Yellow No. 5 to identify it on their labels. Sometimes, Yellow No. 5 is also called tartrazine. If you have or think you have an allergy to a food dye, ask your doctor and be sure you know all the names that dye goes by, so you can avoid it.
More information from the FDA can be found at the link below. But, in a nutshell, if an artificial dye is in use in the US, it has to be certified by the FDA. So, if you ask the FDA if a color additive like Red 40 is safe, they will tell you that for the overwhelming majority of people, it is.
What do the “critics” say about artificial food additives?
Despite the FDA’s stance, there are still plenty of critics. The majority of studies on artificial food colors cover toxicity and childhood behavior complaints. For this review, I’ll focus on the latter. However, I’ll also mention that a 2012 review found that “all nine of the currently US-approved [food] dyes raise health concerns of varying degrees” of toxicology (Kobylewski and Jacobson, 2012). Most focused on contamination with carcinogens. Others have viewed genotoxicity (toxic to DNA), and others have studied hypersensitivity reactions. The FDAs stance on these studies is that these contaminants do not occur at a high enough rate to warrant disapproval from the FDA. But toxicology of artificial dyes is another topic and argument altogether, beyond the scope of this review. Again, here, we focus on behavioral changes associated with food dyes.
In 2011, an FDA advisory committee convened to hear, among other things, the then-current status of evidence regarding ADHD and artificial food coloring. A scientific summary of the critical recommendations is available in a scientific publication (Arnold, Lofthous, & Hurt, 2012). My review here uses this publication as a primary source, though I did consult a number of other publications too. To summarize, Arnold and colleagues determined from their review of 35+ years of research that artificial food colorings have a clear, albeit small, link with childhood behavioral concerns, regardless of whether or not the child has ADHD. Food colorings are not a cause of ADHD, and this is clear. However, studies suggest that food colors can contribute to behavioral problems in general, at least to a small degree, which parents can associate with “worsening” of ADHD.
When we say food dyes “worsen” ADHD, though, it should be specified that this is generally only true on parent report of symptoms. The difference in symptoms has not shown up on objective tests or on clinic observations. In other words, really only parents notice the difference. This doesn’t mean that parents are wrong or that the objective tests are wrong; it only means that the noted differences are not reliable across raters. It’s also important to note that the “effect size” – which is a statistical measure of how much of a difference any one thing makes on a particular outcome, in this case, of food dyes on ADHD – is small. A popular effect size statistic, called Cohen’s d, can be positive or negative, and has no range limitations, meaning the number theoretically could be infinitely high. A positive d means that the outcome increases due to the variable being tested. Again, in this case, it means symptoms of ADHD go up as artificial food dye consumption goes up. Even though Cohen’s d can go to infinity, interpretive guidelines tell us that 0.1 is a small effect size, 0.3 is medium, and >0.5 is large. If d is <0.1, this means there is really “no effect.” Effect sizes in high quality studies that inspect the effects of food dyes on ADHD symptoms have ranged from 0.12 – 0.27, depending on the study and on the outcome. Thus, there is a clear effect, but the effects are small.
Nonetheless, the findings of these studies were compelling enough for the United Kingdom that they resulted in the UK government requesting that food manufacturers avoid artificial additives in favor of natural food colors and flavors. In fact, if you look at the difference in ingredients across foods that are available in both the UK and US, you will find that US foods contain artificial colorings, whereas UK foods do not, in many cases. Shortly after the UK action, the US also began to consider the evidence. Still, in the end, that 2011 FDA advisory committee ultimately voted (though it was close) to not make any changes to food additives or even to warn the public about potential connections. The committee believed that there was not enough evidence, and that the effects were too small, to justify regulatory action.
Another difficulty in deciphering the evidence is in the way ADHD is defined in different studies. ADHD is hard to diagnose accurately, and in general practice, proper steps are rarely taken to explain all the possible causes of ADHD-like symptoms. Many studies have not completed due diligence in formally diagnosing ADHD using gold-standard practices. In many cases, they rely only a parent-rating forms, which is very useful information, but not, in and of itself, diagnostic of ADHD. In children, inattention and hyperactivity, in fact, are among the least specific symptoms available, and can occur in practically all childhood psychiatric/behavioral conditions. Many studies forgo a thorough diagnostic interview because those take a lot of time (in research settings, such interviews usually take 2-5 hours) and require the usage of expertly trained staff. These demands can prevent researchers from being able to carry out the gold-standard approach, so they rely on alternative measures. The problem is that these alternative measures may not be as specific or as sensitive as the gold-standards. In other words, they may misdiagnose some kids as having ADHD when they do not, and they also may miss some kids that do not have ADHD, when they actually do. One way around this issue is to just recruit such a huge number of kids in your study that these effects “wash out” eventually. But still, that’s not perfect, and it’s also quite difficult to recruit a large enough sample to fully negate these confounding effects.
It’s also important to know diagnosis of ADHD is a “threshold” diagnosis. It’s not as simple as, say, running a blood test to see if a person has a virus or not. ADHD requires you to have a certain amount of symptoms (6 out of 9, in each of two domains) to get a diagnosis. That means that if you have 5 symptoms, you don’t get the classic diagnosis, officially. Still, it feels, to most, like there is a difference between children who have 0-1 symptoms versus those who have 5. But, is there really a difference between a person with 6 symptoms versus one with 7? How about one with 2 and one with 3? What about one with 5 and one with 7? When it comes to using basic symptom checklists, we really don’t know what is a meaningful difference. This is a major problem in interpreting a lot of studies.
There are ways, of course, to improve results. By using norm-referenced tests (tests where, for example, we already know how much kids’ scores change over time), we can identify what is an actual “meaningful” difference (a remarkable change) and what falls within the range of “normal” change. Let’s use blood pressure as an analogy. If you measure your blood pressure across time, you’re going to get a different reading most of the time. That’s because BP has normal variation, and doctors don’t get worried if it varies within the normal range. What does worry doctors is if there is a higher than normal change in BP, either up or down. But to know what’s abnormal variation, you first have to know what is normal variation. Norm-referenced tests of ADHD are ones in which we know the normal variation. Once we know normal variation, we can use relatively simple (but regrettably not often used) statistics to see if change over time exceeds the normal variation (the statistic I like is called the Reliable Change Index). Unfortunately, not many people use this. The Reliable Change Index is similar to Cohen’s d, but it’s much better suited for the types of studies we’re talking about here, because it accounts for normal variation much better than Cohen’s d, and it gives us more valuable information. Regrettably, I’m not aware of any food dye studies that use the RCI to represent the level of change in ADHD symptoms over time. A lot of food dye studies also don’t use norm-referenced tests. They use simple symptom measures. All of this makes interpretation of the results very difficult. And these are only some of the reasons that I think the FDA ultimately decided not to take action yet. The majority of studies just don’t meet a high enough standard of evidence.
Nonetheless, within the FDA, there was (and is still) a great amount of support for further research on the connection between food dyes and behavioral changes in children. Currently, much more research is underway. It will take years, I expect, though for the public to hear about any results from new research. Nonetheless, the authors of the review that I cited above claimed that the research is still “too substantial to dismiss,” and I happen to agree. Here is why.
Do the benefits outweigh the potential risks?
In my opinion, the benefits of food dyes do not outweigh risk. Food colors are added for two reasons: 1) to make food look more attractive, and 2) to allow people to distinguish between foods that otherwise would have the same color (e.g. telling the difference between flavors of Skittles). Artificial food dyes provide only an aesthetic benefit, and they are almost exclusively used in processed foods and candies. Thus, there is, in my opinion, nothing useful about artificial food dyes, except that they make us think (wrongly) that “natural” foods are less attractive – they don’t look as “yummy.” Now, having said that, I’ve had plenty of Doritos, Starbursts, cakes, and other processed foods in my life – and they do provide a qualitatively different experience to eat than natural foods do. Plenty of dyed foods are absolutely delicious, and part of the joy of life is letting ourselves go every once in a while to enjoy. However, none of that joy comes from a chemical benefit of food dyes. Food dyes do not add taste, chemically speaking. The taste benefit that we do enjoy is completely psychological. We expect it to taste great, because it looks great, and so our minds trick us into actually liking it better. This has gotten worse over time, because food manufacturers have packed dyed foods with other rich flavors, which give us a quick kick, but usually with low nutritional benefit. So we enjoy it, and we crave more, because it’s not nutritionally satisfying. It’s great for marketing; not so great for our bodies.
Still, having a psychologically enjoyable experience while eating is an important part of being human, and food dyes have become, in our culture, a key part of our psychological dining experience. This makes the change to non-artificial food coloring a difficult one for many to make, because we are so used to it. Nonetheless, in reality, “natural” food dyes (the “exempt” ones) often provide just as much of an aesthetic benefit as artificial ones. They are just a little more expensive to make, which means that foods containing natural colors are a often a bit more expensive at the grocery store. The price tag of “natural” foods often deters people from committing to purchase them. Rao and colleagues (2013) conducted a study exploring price difference in “healthy” versus “less healthy” foods. Summarizing 27 studies across 10 countries, they found that price per 2000 calories was $1.56 more for healthy foods, versus non-healthy foods. However, when they compared nutrient-based patterns (how many nutrients you get), price per day was not significantly different ($0.04). In other words, healthier foods are more nutritious foods, and the differences in nutrition account for all of the difference in price. Overall, the study summarized that healthier food choices cost about $1.50 per day, which translates to ~$550 per year in food costs per person. For low-income families, this is a burden. But the total cost burden is actually more complicated if you also account for health challenges that can be attributed at least in part to dietary choices. A recent study, for example, estimated that suboptimal diet accounted for 14% of all disability-adjusted life years in the USA, which corresponds roughly to $1200/year for every US-person. Thus, theoretically, healthier diet choices can potentially lead to ~$650 per year in savings per person. So, for cost alone, it can be argued that the extra cost is worth the health benefit, and may, in fact, reduce costs in the long run, by reducing healthcare costs.
It is also possible to enjoy eating in other ways. There are many aesthetic means to enhance your eating experience. A lot of these are outlined in an interesting book (Spence, 2017 – see references below) that outlines the science of eating. I encourage you to check it out for lots of great insights into how you can make eating as enjoyable as possible, while still eating healthy.
For me though, overall, the choice is pretty simple. There are enough studies that suggest artificial food dyes MIGHT contribute to behavioral problems in children, and there are no studies that suggest they provide any non-psychological benefit that cannot be replaced with other psychological factors of enjoyable eating. Thus, in my family, I choose to avoid artificial dyes and eat healthy foods as often as possible. This doesn’t mean we never eat unhealthy foods – we do, and they’re still pretty good, albeit honestly fairly unsatisfying most of the time – but since making the change, we rarely crave sugar-rich foods, and dyed foods often, to me, look “fake,” and I just don’t crave them anymore. Shopping for healthier foods is easier when you don’t have to check the labels. There are grocery stores, such as Earth Fare and Whole Foods, for example, who have made pledges not to carry foods with artificial food colorings. So, by shopping at one of those stores, you don’t even have to check labels. If you prefer another grocery store, put in some work up front to check your labels. Once you’ve identified your favorite artificial-dye-free brands, you can just keep buying them – you don’t have to check labels every time.
At the same time, you may not notice a difference at all in your child’s behaviors by going artificial dye free. Dietary changes are not an evidence-based treatment for ADHD or any other childhood psychiatric condition. Alone, making changes in your child’s diet is unlikely to cure them of any mental health condition. So, if your child is struggling with difficult behaviors or emotions, schedule a checkup with a mental health professional for a consult. We can often provide good insights to help round out whatever you’re already doing to help. But for many, dietary changes may be just that thing that does help you round out all the other things you’re doing, including therapy, medication, and/or good parenting strategies, to help your child thrive at his or her best.
Arnold, L. E., Lofthous, N., & Hurt, E. (2012). Artificial food colors and Attention-Deficit/Hyperactivity symptoms: Conclusions to dye for. Neurotherapeutics, 9(3), 599-609.
Kobylewski, S., Jacobsen, M.F. (2012). Toxicology of food dyes. International Journal of Occupational and Enironmental Health, 18(3), 220-246.
Rao, M., Afshin, A., Singh, G., & Mozaffarian, D. (2013). Do healthier foods and diet patterns cost more than less healthy options? A systematic review and meta-analysis. British Medical Journal Open, 3(12). doi: 10.1136/bmjopen-2013-004277
Spence, C. (2017). Gastrophysics: The new science of eating. New York: Penguin Books.