The use of synthetic dyes in food has been a contentious issue for quite some time, bristling hairs in the political world, the food industry, and throughout the public sector.
In the middle of 2010, the Center for Science in the Public Interest teamed up with UCLA doctoral candidate Sarah Kobylewski to release a review of safety-related studies performed on nine food dyes called Food Dyes - A Rainbow of Risks.
While the review seemed quite comprehensive, the conclusions drawn from the data didn't really seem to make sense.
CSPI is a rather inflammatory organization, often on the extreme end of conservatism when it comes to food regulations (meaning they favor heavy regulation).
Given CSPI's history of overstatement and fear-mongering, Ms.
Kobylewski's paper read almost as if she had written the review portion and then CSPI had come along afterward and written (or rewritten) the conclusions.
Because this report has gotten so much attention in the media since its release, I feel it's important to cover some of the more questionable aspects of the paper and give my thoughts on some of the major points.
A little rationality can go a long way when it comes to data interpretation.
Unfortunately, Ms.
Kobylewski's paper often errs on the side of hyperbole and paranoia, much to the detriment of the public at large.
 The first point that must be made is that only six of the nine dyes reviewed are used in any appreciable quantity.
The other three are either defunct or used in such small quantities that their effect on humans is almost assuredly nil.
Citrus Red 2 is a dye used to color the peels of some oranges.
While it might raise some concern if used in processed foods or other consumed products, its presence on the peel is benign.
In addition, its use is federally regulated to a maximum of 2ppm (~0.
000001g/lb of fruit), which is an incredibly small amount in any sort of application.
Green 3 is next on the list of irrelevant dyes.
Registering in at a miniscule 0.
1% of total yearly FDA-certified dye production, green 3 is very rarely used.
When a green color is needed, 99% of the food industry chooses a combination of blue 1 and yellow 5.
In addition, green 3 is known to be poorly absorbed, further reducing its effect on the body.
Mice studies yielded no evidence against green 3 and rat studies produced quite inconclusive data at very high treatment levels (1.
25-5% of the total diet as green 3!).
With almost no negative data to its name, even if you are especially paranoid, green 3 is a non-factor because it is so easy to avoid.
The final meaningless dye in this review is Orange B.
Approved for use only in sausage casings, Orange B is no longer used and hasn't even had a batch approved for use in over a decade.
 Now onto the relevant dyes, starting with blue 1 (a.
k.
a Brilliant Blue).
Blue 1 comprises 4.
7% of the total yearly FDA-certified dye production.
Part of this low percentage is derived from the fact that blue 1 is an intensely powerful colorant and is therefore generally used in minute quantities, even relative to other major dyes.
No published studies on blue 1 produced usable data pointing to toxicity or carcinogenicity (cancer-causing action).
A lone unpublished study (suspicious? Yes.
) showed some rise in rates of kidney tumors, but a dose-response relationship could not be established, making the claim of carcinogenicity quite suspect.
Two out of nine studies assessing the genotoxicity of blue 1 produced positive results in chromosomal aberration tests.
However, one study was listed without a dose of the active ingredient and the other used a dose of 5mg/ml, which is literally insanely high when compared to human consumption levels.
 The only interesting result regarding blue 1 comes from a single neurotoxicity study in which blue 1, in combination with L-glutamic acid, partially inhibited the development of neurites, outcroppings from neurons that are associated with neuronal and cognitive development.
Studies have found significant (if sometimes loosely defined) correlations between the consumption of food dyes and hyperactivity in children.
It's possible that there is some connection there, but given the level of current evidence for causality, it's certainly not established.
So, if your child is very young (<1 year) or has a problem with hyperactivity, it may be worth the effort to avoid blue 1.
Otherwise, as usual, it appears to do basically nothing.
 Blue 2 is the other certified blue dye and makes up 3.
7% of certified dye production.
It is extremely poorly absorbed, even more so than blue 1 and green 3.
Ten out of eleven studies looking at the potential genotoxicity of blue 2 found no effect.
Studies investigating chronic toxicity of blue 2 found nothing conclusive to point towards a negative impact on humans.
Activist groups have made loud noises over a finding in one study that showed an apparent increase in brain gliomas (a tumor that arises from glial cells).
However, the data failed to show a number of characteristics that would indicate blue 2 acted as a carcinogen.
In addition, when taken in context with other studies on the same type of rats, the incidence of the brain gliomas was not unusual, even in untreated populations.
Despite the scientific evidence against the viability of the brain glioma data, it is still touted by opponents of food dyes as a reason to ban blue 2.
 Now that we've covered the blues, it's time to take a look at our two options for red color.
The first is red 3.
Only making up 1.
4% of total certified dye production, red 3 is by far the lesser-used red dye.
Its one major place in the diet is in maraschino cherries, though it can also be found in canned fruits, candies, oral drugs, and a few other minor products.
Suspicions of genotoxicity regarding red 3 were raised when a few studies found positive evidence.
However, two of the four positive studies were performed on isolated yeast cells, which are hardly indicative of mammalian cells within a body.
One study, using the most informative protocol of all the referenced experiments, found positive results after three hours of treatment but negative results after 24 hours.
While interesting, the contradictory nature of the results makes any claims based on that data somewhat dubious.
That leaves one useful positive result out of 11 reviewed studies.
Hardly convincing.
 In studies investigating potential toxicity of red 3, no effect was found at doses up to 4% of the total diet.
However, there was significant evidence pointing towards a carcinogenic effect at the highest dose.
It seems a bit ridiculous to grossly generalize results from data based on consuming 4% of the diet as dye when a human may normally consume on the order of a few milligrams per day (if even that), especially when data on lesser doses showed no effect.
However, this is one of those cases where, if you are especially paranoid, red 3 is easy enough to cut out of your diet.
 Red 3's big brother is red 40.
Red 40 is the most heavily-consumed dye and comprises 41.
3% of yearly certified-dye production.
Genotoxicity data is limited, but one study found evidence of DNA damage due to red 40 at very high doses.
However, a single positive study among all the negative results provides little insight into the truth of the matter.
With the data currently available, I feel that the genotoxicity of red 40 is quite inconclusive.
Toxicity and carcinogenicity testing found no reliable effect.
Opponents of dyes often claim that one study found that red 40 accelerated the development of a certain type of tumor in mice.
However, that suspicion was preliminary and was raised in the middle of the study.
By the end of the experiment, the researchers had found no acceleration of tumor appearance.
In addition, a second confirmatory study was performed to assess the risk of red 40 specifically in regards to the tumor type seen in the first trial.
That experiment also found no effect of red 40 on tumor generation.
 Despite the non-issue of red 40 toxicity, other valid concerns over the dye exist.
The first is hypersensitivity reactions seen in a very small percentage of the population.
However, one's reaction to red 40 would be apparent and for the vast majority of people will not be a concern.
The most serious issue surrounding red 40, as well as the two yellow dyes to be discussed shortly, is contamination with potentially carcinogenic impurities.
While the contaminants historically found in red 40 likely pose little risk to humans at the levels consumed, in recent years more and more dyes have begun to be imported from foreign producers like China.
Given China's decidedly suspect history with chemical contamination and adulteration, along with the general difficulty of producing completely contaminant-free red 40, more in-depth inspections by the FDA of imported dyes is warranted.
However, that issue will be handled through policy and regulatory changes if it ever comes to pass in the future and should be of little concern to the day-to-day user.
 The last two dyes discussed in the CSPI review are the yellows.
Yellow 5 is the second most popular dye used in food and cosmetic products.
It produces an intense neon-type yellow color and is often used in combination with red 40 or blue 1 to produce orange and purple colors.
A few studies showed some evidence for concern of genotoxicity with yellow 5.
However, a majority of studies showed no evidence for such effects and only one of the positive studies was performed in vivo, limiting the viability of genotoxicity claims.
Studies found no evidence indicating carcinogenicity of yellow 5 even at extremely high doses.
 However, one negative aspect of yellow 5 that is well-established is hypersensitivity.
A small percentage of the population is allergic to the dye.
Interestingly, there is a large crossover between those allergic to yellow 5 and those allergic to aspirin.
If an individual reacts to one compound, they are likely to react to the other.
In the end though, the vast majority of consumers are not reactive to yellow 5 at all.
 Finally, there is the concern over contamination, similar to that raised with red 40.
The contaminant of major concern in the CSPI report, and the one that is continually mentioned throughout the text, is benzidine.
However, when consulting the original studies used by CSPI to bolster their claims of apparently widespread benzidine contamination in yellow dyes, the data does not appear to support their arguments.
The two studies were performed in the early to mid 1990s and examined the amount of free and bound benzidine found in samples of certified yellow 5 and yellow 6.
The FDA only tests for free benzidine and CSPI claims that bound benzidine is also dangerous because it is liberated to its free form in the human gut.
A number of holes in CSPI's statements become apparent after reading the original documents.
First, the theory that bound benzidine is freed during digestion is simply postulated by the authors and not supported by any data, whatsoever.
Second, 90% of the lots found to be contaminated with benzidine from both studies combined came from a single producer.
So, in essence the study didn't uncover a widespread contamination problem, but a localized issue with one company.
 Last but not least in our review is yellow 6, the "oranger" cousin of yellow 5.
In genotoxicity trials, yellow 6 was seen to produce no effect in 8 out of ten experiments.
In addition, the only in vivo study found no effect.
Therefore, the evidence for genotoxicity in the case of yellow 6 is quite suspect.
The only evidence for carcinogenicity of yellow 6 comes from a single study on rats.
However, the FDA reviewed the study and found that the data lacked a number of important features that would indicate that yellow 6 was acting as a carcinogen.
Given the lack of other studies pointing to yellow 6 as a carcinogen, the data supporting an argument against the dye is extremely weak.
 Yellow 6 shares the same valid concerns over hypersensitivity and contamination that pertain to yellow 5.
Reactions to the dye are extremely rare and shouldn't affect its use in the food industry considering labeling requirements and openness of ingredient information.
Contamination is a potential (but not current) problem that can easily be fixed through more stringent tests by the FDA.
Especially considering the growth in dye importation, increased ingredient security is a priority worth pursuing.
 Dyes play an important role in the food industry.
The argument that natural colorants could simply be substituted for the synthetics isn't a feasible option for many companies considering consumer demand for low food prices.
In addition, some natural colorants produce similar hypersensitivity reactions to those rarely seen with synthetics.
Importation of dyes is a valid concern and should be addressed by the FDA through enhanced certification testing.
The truth about food dyes is that there is little reason for concern.
The amount of dye we consume on a regular basis is incredibly small and, even considering the gigantic doses seen in many animal studies, the risks posed by these dyes are largely negligible.
CSPI is an organization that has a history of overstatement, hyperbole, and fear-mongering.
The information they put out is at the very least misleading and is most definitely untrustworthy.
Too much of anything is a bad idea and the same is true for dyes.
There's no reason to be especially fearful of dyes and avoid them altogether, but consuming enormous amounts of color every day is also likely not the best plan.
Moderation is key in most issues related to the intersection of food and human health.
The subject of dyes is just another perfect example.
In the middle of 2010, the Center for Science in the Public Interest teamed up with UCLA doctoral candidate Sarah Kobylewski to release a review of safety-related studies performed on nine food dyes called Food Dyes - A Rainbow of Risks.
While the review seemed quite comprehensive, the conclusions drawn from the data didn't really seem to make sense.
CSPI is a rather inflammatory organization, often on the extreme end of conservatism when it comes to food regulations (meaning they favor heavy regulation).
Given CSPI's history of overstatement and fear-mongering, Ms.
Kobylewski's paper read almost as if she had written the review portion and then CSPI had come along afterward and written (or rewritten) the conclusions.
Because this report has gotten so much attention in the media since its release, I feel it's important to cover some of the more questionable aspects of the paper and give my thoughts on some of the major points.
A little rationality can go a long way when it comes to data interpretation.
Unfortunately, Ms.
Kobylewski's paper often errs on the side of hyperbole and paranoia, much to the detriment of the public at large.
 The first point that must be made is that only six of the nine dyes reviewed are used in any appreciable quantity.
The other three are either defunct or used in such small quantities that their effect on humans is almost assuredly nil.
Citrus Red 2 is a dye used to color the peels of some oranges.
While it might raise some concern if used in processed foods or other consumed products, its presence on the peel is benign.
In addition, its use is federally regulated to a maximum of 2ppm (~0.
000001g/lb of fruit), which is an incredibly small amount in any sort of application.
Green 3 is next on the list of irrelevant dyes.
Registering in at a miniscule 0.
1% of total yearly FDA-certified dye production, green 3 is very rarely used.
When a green color is needed, 99% of the food industry chooses a combination of blue 1 and yellow 5.
In addition, green 3 is known to be poorly absorbed, further reducing its effect on the body.
Mice studies yielded no evidence against green 3 and rat studies produced quite inconclusive data at very high treatment levels (1.
25-5% of the total diet as green 3!).
With almost no negative data to its name, even if you are especially paranoid, green 3 is a non-factor because it is so easy to avoid.
The final meaningless dye in this review is Orange B.
Approved for use only in sausage casings, Orange B is no longer used and hasn't even had a batch approved for use in over a decade.
 Now onto the relevant dyes, starting with blue 1 (a.
k.
a Brilliant Blue).
Blue 1 comprises 4.
7% of the total yearly FDA-certified dye production.
Part of this low percentage is derived from the fact that blue 1 is an intensely powerful colorant and is therefore generally used in minute quantities, even relative to other major dyes.
No published studies on blue 1 produced usable data pointing to toxicity or carcinogenicity (cancer-causing action).
A lone unpublished study (suspicious? Yes.
) showed some rise in rates of kidney tumors, but a dose-response relationship could not be established, making the claim of carcinogenicity quite suspect.
Two out of nine studies assessing the genotoxicity of blue 1 produced positive results in chromosomal aberration tests.
However, one study was listed without a dose of the active ingredient and the other used a dose of 5mg/ml, which is literally insanely high when compared to human consumption levels.
 The only interesting result regarding blue 1 comes from a single neurotoxicity study in which blue 1, in combination with L-glutamic acid, partially inhibited the development of neurites, outcroppings from neurons that are associated with neuronal and cognitive development.
Studies have found significant (if sometimes loosely defined) correlations between the consumption of food dyes and hyperactivity in children.
It's possible that there is some connection there, but given the level of current evidence for causality, it's certainly not established.
So, if your child is very young (<1 year) or has a problem with hyperactivity, it may be worth the effort to avoid blue 1.
Otherwise, as usual, it appears to do basically nothing.
 Blue 2 is the other certified blue dye and makes up 3.
7% of certified dye production.
It is extremely poorly absorbed, even more so than blue 1 and green 3.
Ten out of eleven studies looking at the potential genotoxicity of blue 2 found no effect.
Studies investigating chronic toxicity of blue 2 found nothing conclusive to point towards a negative impact on humans.
Activist groups have made loud noises over a finding in one study that showed an apparent increase in brain gliomas (a tumor that arises from glial cells).
However, the data failed to show a number of characteristics that would indicate blue 2 acted as a carcinogen.
In addition, when taken in context with other studies on the same type of rats, the incidence of the brain gliomas was not unusual, even in untreated populations.
Despite the scientific evidence against the viability of the brain glioma data, it is still touted by opponents of food dyes as a reason to ban blue 2.
 Now that we've covered the blues, it's time to take a look at our two options for red color.
The first is red 3.
Only making up 1.
4% of total certified dye production, red 3 is by far the lesser-used red dye.
Its one major place in the diet is in maraschino cherries, though it can also be found in canned fruits, candies, oral drugs, and a few other minor products.
Suspicions of genotoxicity regarding red 3 were raised when a few studies found positive evidence.
However, two of the four positive studies were performed on isolated yeast cells, which are hardly indicative of mammalian cells within a body.
One study, using the most informative protocol of all the referenced experiments, found positive results after three hours of treatment but negative results after 24 hours.
While interesting, the contradictory nature of the results makes any claims based on that data somewhat dubious.
That leaves one useful positive result out of 11 reviewed studies.
Hardly convincing.
 In studies investigating potential toxicity of red 3, no effect was found at doses up to 4% of the total diet.
However, there was significant evidence pointing towards a carcinogenic effect at the highest dose.
It seems a bit ridiculous to grossly generalize results from data based on consuming 4% of the diet as dye when a human may normally consume on the order of a few milligrams per day (if even that), especially when data on lesser doses showed no effect.
However, this is one of those cases where, if you are especially paranoid, red 3 is easy enough to cut out of your diet.
 Red 3's big brother is red 40.
Red 40 is the most heavily-consumed dye and comprises 41.
3% of yearly certified-dye production.
Genotoxicity data is limited, but one study found evidence of DNA damage due to red 40 at very high doses.
However, a single positive study among all the negative results provides little insight into the truth of the matter.
With the data currently available, I feel that the genotoxicity of red 40 is quite inconclusive.
Toxicity and carcinogenicity testing found no reliable effect.
Opponents of dyes often claim that one study found that red 40 accelerated the development of a certain type of tumor in mice.
However, that suspicion was preliminary and was raised in the middle of the study.
By the end of the experiment, the researchers had found no acceleration of tumor appearance.
In addition, a second confirmatory study was performed to assess the risk of red 40 specifically in regards to the tumor type seen in the first trial.
That experiment also found no effect of red 40 on tumor generation.
 Despite the non-issue of red 40 toxicity, other valid concerns over the dye exist.
The first is hypersensitivity reactions seen in a very small percentage of the population.
However, one's reaction to red 40 would be apparent and for the vast majority of people will not be a concern.
The most serious issue surrounding red 40, as well as the two yellow dyes to be discussed shortly, is contamination with potentially carcinogenic impurities.
While the contaminants historically found in red 40 likely pose little risk to humans at the levels consumed, in recent years more and more dyes have begun to be imported from foreign producers like China.
Given China's decidedly suspect history with chemical contamination and adulteration, along with the general difficulty of producing completely contaminant-free red 40, more in-depth inspections by the FDA of imported dyes is warranted.
However, that issue will be handled through policy and regulatory changes if it ever comes to pass in the future and should be of little concern to the day-to-day user.
 The last two dyes discussed in the CSPI review are the yellows.
Yellow 5 is the second most popular dye used in food and cosmetic products.
It produces an intense neon-type yellow color and is often used in combination with red 40 or blue 1 to produce orange and purple colors.
A few studies showed some evidence for concern of genotoxicity with yellow 5.
However, a majority of studies showed no evidence for such effects and only one of the positive studies was performed in vivo, limiting the viability of genotoxicity claims.
Studies found no evidence indicating carcinogenicity of yellow 5 even at extremely high doses.
 However, one negative aspect of yellow 5 that is well-established is hypersensitivity.
A small percentage of the population is allergic to the dye.
Interestingly, there is a large crossover between those allergic to yellow 5 and those allergic to aspirin.
If an individual reacts to one compound, they are likely to react to the other.
In the end though, the vast majority of consumers are not reactive to yellow 5 at all.
 Finally, there is the concern over contamination, similar to that raised with red 40.
The contaminant of major concern in the CSPI report, and the one that is continually mentioned throughout the text, is benzidine.
However, when consulting the original studies used by CSPI to bolster their claims of apparently widespread benzidine contamination in yellow dyes, the data does not appear to support their arguments.
The two studies were performed in the early to mid 1990s and examined the amount of free and bound benzidine found in samples of certified yellow 5 and yellow 6.
The FDA only tests for free benzidine and CSPI claims that bound benzidine is also dangerous because it is liberated to its free form in the human gut.
A number of holes in CSPI's statements become apparent after reading the original documents.
First, the theory that bound benzidine is freed during digestion is simply postulated by the authors and not supported by any data, whatsoever.
Second, 90% of the lots found to be contaminated with benzidine from both studies combined came from a single producer.
So, in essence the study didn't uncover a widespread contamination problem, but a localized issue with one company.
 Last but not least in our review is yellow 6, the "oranger" cousin of yellow 5.
In genotoxicity trials, yellow 6 was seen to produce no effect in 8 out of ten experiments.
In addition, the only in vivo study found no effect.
Therefore, the evidence for genotoxicity in the case of yellow 6 is quite suspect.
The only evidence for carcinogenicity of yellow 6 comes from a single study on rats.
However, the FDA reviewed the study and found that the data lacked a number of important features that would indicate that yellow 6 was acting as a carcinogen.
Given the lack of other studies pointing to yellow 6 as a carcinogen, the data supporting an argument against the dye is extremely weak.
 Yellow 6 shares the same valid concerns over hypersensitivity and contamination that pertain to yellow 5.
Reactions to the dye are extremely rare and shouldn't affect its use in the food industry considering labeling requirements and openness of ingredient information.
Contamination is a potential (but not current) problem that can easily be fixed through more stringent tests by the FDA.
Especially considering the growth in dye importation, increased ingredient security is a priority worth pursuing.
 Dyes play an important role in the food industry.
The argument that natural colorants could simply be substituted for the synthetics isn't a feasible option for many companies considering consumer demand for low food prices.
In addition, some natural colorants produce similar hypersensitivity reactions to those rarely seen with synthetics.
Importation of dyes is a valid concern and should be addressed by the FDA through enhanced certification testing.
The truth about food dyes is that there is little reason for concern.
The amount of dye we consume on a regular basis is incredibly small and, even considering the gigantic doses seen in many animal studies, the risks posed by these dyes are largely negligible.
CSPI is an organization that has a history of overstatement, hyperbole, and fear-mongering.
The information they put out is at the very least misleading and is most definitely untrustworthy.
Too much of anything is a bad idea and the same is true for dyes.
There's no reason to be especially fearful of dyes and avoid them altogether, but consuming enormous amounts of color every day is also likely not the best plan.
Moderation is key in most issues related to the intersection of food and human health.
The subject of dyes is just another perfect example.
SHARE
