For More Information GO TO Aspartame (NutraSweet) Toxicity Home Page: http://www.holisticmed.com/aspartame/ 3. Serious NutraSweet Research Flaws Before we discuss individual studies, it is important to list common and very serious flaws in all of the research funded by NutraSweet. This is by no means meant to be a comprehensive list of flaws -- simply the most common serious flaws. The flaws I will discuss in this section related to studies after aspartame was approved (post- approval). The pre-approval studies bordered on criminally fraudulent activity in my opinion and will be discussed in a later section. a. Test Material In most of the NutraSweet-funded studies, the test material used was fresh, encapsulated aspartame. This is a major flaw for the following reasons: 1. The chemical makeup of the fresh aspartame used is almost 100% pure aspartame and differs significantly from what is being ingested by the general public. This is discussed thoroughly in the "By-Products and Breakdown constituents" section above. This means that DKP, beta-aspartame, free methanol, and other possible breakdown products are not being tested in these experiments. 2. In 1987, Stegink tested the effect of aspartame taken in liquid as opposed to capsules on plasma phenylalanine, phenylalanine/LNAA, tyrosine, and aspartate (Stegink 1987a). The difference was striking. The plasma phenylalanine and aspartate levels rose very quickly to extremely high levels when ingesting the liquid aspartame mixtures, but the plasma amino acid levels only rose moderately when ingesting encapsulated aspartame. While this experiment compared the effect of aspartame in liquid vs. capsules, it did not test real world liquid aspartame-containing products which would contain significant amounts of DKP, methanol, free amino acids, and other possibly dangerous chemicals. The rise in plasma amino acid levels may be even more striking and sudden with such products due to even faster absorbtion. 3. The experiment conducted by Stegink (1987a) showed that capsule administration of aspartame significantly delayed absorption of aspartic acid and phenylalanine. In fact, with liquid administration, the peak amino acid levels were reached within 32 minutes (average), yet capsule administration led to a gradual rise in amino acid levels and took approximately 2.5 times longer to reach much lower peak levels. Not only is the enormous difference in the plasma amino acid spikes important as discussed above, but the sudden spike that occurs in liquid administration that may also be very important. When a substance is gradually absorbed in a way that causes it to be slightly toxic, the body has a chance to adjust and mount a defense. Sudden absorption of single, potentially neurotoxic amino acids does not give the body a chance to mount a defense. It is also very important to note that delaying the absorption of methanol as would happen when ingesting encapsulated aspartame may reduce the methanol toxicity somewhat since food in the stomach, which also delays methanol absorption, seems to reduce methanol toxicity (Posner 1975). It is interesting to note that as early as 1973, the FDA told the manufacturer of aspartame that there is "No pharmacokinetic data . . . on absorption, excretion, metabolism, half-life; nor bioavailability of capsule vs. food additive administration" (Freeman 1973). It wasn't until 1987, 14 years later, that NutraSweet finally got around to testing capsule administration as compared to liquid administration! There was a striking difference as described above. To this day, there has been no tests comparing the administration of various real-world, aspartame-containing products to capsule administration. The large difference in biochemical reactions produced when ingesting real world aspartame-containing products as opposed to capsules given in the laboratory totally negates the results from experiments which used such capsules and found no adverse effects. When using aspartame-containing capsules, 1) much less aspartame gets absorbed (Stegink 1987a), 2) the absorption is much slower causing the increase in blood levels of aspartame by-products to be much more gradual (Stegink 1987a), and 3) other by-products and breakdown constituents do not get absorbed as they do in real-world products (Tsang 1985). Had real-world, liquid products been used, the number and severity of negative reactions due to aspartame would likely have been much greater. b. Test Product Administration In order to test the effects of aspartame on health it is important to simulate the way the product is taken by the general public. Sometimes aspartame is ingested with full meals. More frequently, however, aspartame is ingested by itself (e.g., diet colas) or with a sugary snack. It is important to test both methods of administration. It is obvious that the biochemical effect of the three original components of aspartame, aspartic acid, phenylalanine, and methanol will be much greater when it is ingested separate from a full meal. (This will be discussed in more detail in later sections.) When taking aspartame with meals the following things occur: i) The aspartame will not be absorbed as quickly leading to less of a rise in plasma aspartate and phenylalanine levels; ii) The other amino acids absorbed from the food will keep the plasma aspartate and phenylalanine levels from rising as high as they would normally; iii) The plasma phenylalanine to large neutral amino acid (LNAA) ratio will not be as large due to the LNAA's in the food. (This will be discussed in detail in a later section.); iv) The food may serve as a protective factor reducing the methanol toxicity as will be discussed in the Methanol section. It is obvious that the acute and chronic effects of aspartame ingestion will be slightly less when it is ingested with full meals. All previous experiments that tested aspartame ingestion with full meals, tested the best-case scenerio and not what is most common in the real world. Therefore, all such research of aspartame with full meals should be regarded as interesting, but not very useful. On the other hand, Yokogoshi (1984) has shown that aspartame ingestion with caloric sweeteners significantly raises to phenylalanine/LNAA ratio to even greater heights than by simply ingesting aspartame alone. Even a NutraSweet-funded, short study on healthy persons using a relatively small amount of aspartame showed a significant further increase in plasma phenylalanine/LNAA ratio when aspartame was ingested with a caloric sweetener (Wolf-Novak 1990). However, it is unclear whether the ingestion of carbohydrates along with aspartame renders the phenylalanine part of aspartame more dangerous. Carbohydrate ingestion lowers the levels of Large Neutral Amino Acids (LNAAs) and therefore the neutral amino acid trasport cites may become unsaturated causing a smaller change in brain chemistry than would otherwise happen with the phenylalanine alone. We will discuss this in more detail in a later section. Therefore, not only should more experiments be done using real-world aspartame products on persons not eating full meals, but experiments should be done on the combination of real-world liquid aspartame (at FDA ADI levels or greater) and caloric sweeteners. Only then will we begin to approach what happens in real-world aspartame ingestion. It is important to note, however, that the ingestion of aspartame with sugar reduces the possible negative effects from the aspartic acid part of aspartame (as will be discussed in the Aspartic Acid section). Such experiments (with caloric sweeteners plus aspartame are very useful, but lack of negative effects does not rule out possible negative effects from the aspartic acid or the aspartic acid plus another breakdown product (i.e., synergy). c. Short Experiments The majority of NutraSweet-funded experiments on humans tested aspartame for one day or less. Although there is a wide variation in when adverse reactions begin, it is usually several weeks or months after use begins before adverse reactions are noticed (Roberts 1990a, page 70). After the adverse reactions begin, regular aspartame use usually causes more frequent adverse reactions. Studies which are not funded by NutraSweet are usually much longer because the researchers are actually interested in testing aspartame. A quality study would be at least six months long, and preferably as long as one year or more. This way, the researchers will be testing adverse reactions that occur due to ongoing use of aspartame -- real world use. One to two year experiments would be ideal. James Scala, the former director of Health Sciences for General Foods Corporation said that most of the early NutraSweet research consisted of short-term studies that ignored possible subtle, long-term effects. Pediatrician and Geneticist Dr. Reubon Matalon stated "Let us say cigarettes were invented today, and you give 20 people two packs a day and after six weeks, no one has cancer, would you say that it is safe? That's what they did with NutraSweet." (Gordon 1987, page 486 of US Senate 1987) All single day or single challenge tests, usually conducted by NutraSweet-connected researchers, should be disregarded. No one is claiming that a single ingestion of even real-world aspartame-containing products represents an imminent health hazard. A single dose of aspartic acid, phenylalanine, methanol, DKP, etc., from aspartame is not acutely toxic in the majority of cases. It is the regular use that represents a extreme hazard. These concerns are not addressed in one-day studies. No reputable researcher would try to extrapolate a lack of negative effects from aspartame in a one-day experiment to a declaration of safety for onging use throughout a lifetime. Unfortunately, this is commonly done by some NutraSweet-supported researchers. It should be noted that some very subtle adverse reactions may be noticed in a single day experiment, especially in vulnerable populations. While a much longer test is necessary, any significant adverse reactions after only a single day test should be a cause for extreme concern. On the other hand, lack of adverse reations after only a single day test does not prove anything. d. Small Test Population The smaller the test population, the more difficult it is to get a statistically significant difference in amino acid and methanol/formate levels in the blood and urine. NutraSweet-funded research usually has such a ridiculously small test population that it is virtually impossible to have a statistically significant difference in measurements. This is especially true when this flaw is combined with other flaws such as capsule administration of aspartame. Another way a small test population can be a problem is if a particular symptom (e.g., seizures) appears in a small percentage of aspartame users -- let's say one out of every 100 regular users, then experiments with small numbers of people would usually not have a seizure victim. This problem is called "lack of statistical power" of the studies and will be dealt with in more detail when the cancer issue is discussed. The technique of using ridiculously small test populations helps to guarantee that reactions that are occurring in the general population do not occur in the test population. e. Irrelevant or Faulty Tests There are numerous NutraSweet-funded research projects which used irrelevant and faulty tests to draw their conclusions. An irrelevant or faulty test is useful only for press releases and convincing scientists who are not intimately familiar with the scientific issues surrounding aspartame ingestion. Performing many such irrelevant or faulty tests allows the researcher to proclaim, "Look! We tested aspartame, performed a whole battery of tests, and found no adverse effects!" When the protocol is examined closely, however, it becomes clear that many of these tests were meaningless or conducted improperly. Many of the improper testing methods appear to be deliberately created and used to avoid negative results. There are many cases of such irrelevant and faulty tests. I will discuss some of those when I get into the details of the research cited in the article. f. Dosage Tested In a number of NutraSweet-funded studies, the aspartame dosage used was much less than it should have been. As discussed in the previous section, NutraSweet's estimates of aspartame intake are obviously flawed. Frey (1976) showed that children can ingest as much as 76 mg/kg/day. In three studies on obese adults, Porikos (1984) showed that the average daily intake of aspartame varied from 8 to 36 mg/kg/day. The FDA's current Acceptable Daily Intake (ADI) is 50 mg/kg/day. NutraSweet researchers have made a significant effort to convince other researchers and the general population that it is okay to test small doses of aspartame. Much of their argument is based on average intake values presented in their ridiculous intake surveys. The fact is that hundreds of thousands of people are consuming amounts of aspartame approaching the FDA's Acceptable Daily Intake level. If the NutraSweet Company actually believes that the FDA's ADI is a safe amount of aspartame, then that is the minimum that should be tested. Either "put up, or shut up" so to speak. In order to have a safety margin, it is preferable to test at least double the ADI using read-world aspartame- containing products in long-term experiments. If they do not want to test at levels at or above the FDA's ADI (using real-world aspartame of course), then we should 1) lower the ADI back to 20 mg/kg/day, 2) label the amount of aspartame on each "food" product that it is in, 3) put a warning on the label stating that no more than 20 mg/kg/day (9 mg/lbs./day) should be consumed, and 4) start proper safety testing in humans at 20 mg/kg/day (or more) using real-world aspartame- containing products. g. Reaction Time Some research ignored certain adverse reactions if they did not occur soon after aspartame ingestion. This is one common way that food industry scientists significantly reduce the number of adverse reactions recorded during the experiment. They simply imply that the suspected reactions are "allergic" (i.e., IgE- mediated) and therefore must occur quickly after ingestion. The fact of the matter is that many food intolerance or toxicity reactions can occur as much as 48 hours after ingestion (Carroll 1992). This use of this flaw has occurred in several NutraSweet-funded experiments. The studies of MSG funded by the International Glutamate Association use this flaw quite often. Independent researchers usually design the experimental protocols to take into account the fact that people often experience delayed reactions. h. Average Values Shown In the publication of many NutraSweet-funded research projects only average values were shown in tables and plotted on graphs. This is fine in studies where there are a large number of participants and the substance being studied has similar biochemical effects on all people. However, most of these studies have very few subjects and it is well-known that there is a wide variation in the biochemical changes caused by methanol, phenylalanine, and aspartic acid. This may also be true with DKP and other breakdown products. If an experiment has six subjects, for example, and two of the subjects show biochemical changes that would be of concern, the significance of those changes would get lost in a listing of averages. Human studies of aspartame conducted by independent researchers often show measurements on an individual basis (not averages) so as not to obscure the possibility of individual susceptibilities (Koehler 1988, Matalon 1988, Van Den Eeden 1994, Walton 1993). Another related technique used by NutraSweet to help hide negative test results is to combine all of the subjects' measurements when they should not be combined. Suppose for a moment that we measure a blood plasma level of aspartate for the first 60 minutes after ingesting aspartame. Below are example values for five different subjects followed by the average values . 0 min 15 min 30 min 45 min 60 min Subject 1 6.0 38.0 32.0 16.0 9.0 Subject 2 5.0 14.0 28.0 51.0 50.0 Subject 3 7.0 8.0 12.0 41.0 12.0 Subject 4 5.0 23.0 41.0 21.0 19.0 Subject 5 6.0 9.0 24.0 37.0 53.0 Mean Values 5.8 18.4 27.4 33.2 28.6 In almost all NutraSweet funded experiments, all that would be shown in publication are the "Mean Values." Notice how the mean values only rise from 5.8 at 0 minutes to a maximum of 33.2 at 45 minutes. When going back to the actual data all of the rises in the subjects were much more extreme than what is apparent by looking only at the mean values. For example, the plasma aspartate level of Subject 2 rose from 5.0 to 51.0 at 45 minutes. Showing only the means values in a chart or a graph will obscure the true rise in blood plasma levels of the substance being measured. This is because some of the subjects reach their peak values at different times, so that when Subject 1 has a peak value of 38.0 at 15 minutes, Subject 3 has a value of only 8.0 at that time -- which brings the mean value at the 15 minutes time period down considerably. The appropriate thing to do would be to list all of measurements for each subject and list the mean starting values and mean peak values: Mean Starting Values = (6.0 + 5.0 + 7.0 + 5.0 + 6.0) / 5 = 5.8 Mean Peak Values = (38.0 + 51.0 + 41.0 + 41.0 + 53.0) / 5 = 44.8 If increases in certain measurements are only moderate, the NutraSweet Company researcher's use of mean values for each time period can sometimes enable the researcher to "prove" that the changes were not statistically significant. Therefore, NutraSweet-sponsored studies which show only mean values for each time period may be hiding some of the negative effects with their statistical games. i. Types of Volunteers Since serious health problems from aspartame seem to develop gradually in many people, it makes sense to conduct very long tests. The NutraSweet Company wants people to swallow this junk for their entire life. Experiments which last a lifetime are obviously impractical. Therefore, it is prudent to conduct relatively long experiments on susceptible populations first so as to get clues as to what will happen to healthy populations after years of use. Most of the studies funded by NutraSweet appears to have been conducted on healthy volunteers (for a very short test period). These experiments had numerous flaws. It seems that the experiments in which unhealthy volunteer were used, the number and seriousness of the experimental flaws increased significantly. The NutraSweet Company wants you to believe that out of the population that is currently allowed to consume aspartame, PKU Heterozygotes (persons who do not have phenylketonuria -- PKU, but have a single gene of PKU) would be the most susceptible to any possible negative effects from aspartame. While this population may be more susceptible than a healthy population, I submit that persons with the following illnesses will be much more susceptible: - Fibromyalgia - Chronic Fatigue Syndrome (CFS) - Chronic Depression - Multiple Chemical Sensitivities - Multiple Sclerosis There are other possibilities, but this is a good start for legitimate tests on susceptible populations. Please note, however, that healthy persons are very susceptible to the damage that can be caused by aspartame, just not quite as susceptible, on average, as persons with the above-mentioned illnesses. j. Animal Tests All three main ingredients in aspartame, methanol, aspartic acid, and phenylalanine have been shown to have much greater toxic effects in humans than in rodents. Methanol and aspartic acid is much more toxic in humans than in monkeys. Methanol tests in rodents are worthless. Methanol tests in rheusus monkeys are also worthless or guesswork at best as discussed in the Methanol section below. Methanol is much more toxic in humans than any other species. Aspartic acid is 5 times more toxic in humans than in rodents and at least 20 times more toxic in humans than in monkeys as discussed in the Aspartic Acid section. Phenylalanine tests in rodents are guesswork at best and probably worthless as discussed in the Phenylalanine section below. Phenylalanine is much more dangerous in humans than in rodents. It is unknown whether DKP, beta-aspartame, or racemized amino acids have different effects in humans as opposed to laboratory animals. Therefore, aspartame tests in animals, especially those which tested for the effects of methanol and phenylalanine in rodents, or methanol and aspartic acid effects in monkeys should be ignored -- as the negative effect in humans would likely have been much greater. It is important to point out that these flaws are not news to the NutraSweet Company. Many people have been pointing out these flaws for years and pushing for legitimate experiments instead of press releases disguised as research. It is also important to note that NutraSweet can make the following claims about their research: "We have studied aspartame in healthy individuals." "We have studied aspartame in diabetics." "We have studied aspartame in persons with liver disease." "We have studied aspartame in adults." "We have studied aspartame in children." "We have conducted acute-dosing tests." "We have conducted chronic use tests [a few weeks only]." etc., etc. However, each one of their experiments had multiple serious flaws, making it virtually useless for anything but a press release. For example, long-term studies often use capsules which were taken with meals and contained numerous irrelevant or poorly conducted tests. Given aspartame's history of what some people consider pre-approval fraud, it does not surprise me that the NutraSweet Company continues to flood the scientific community and the news services with results from badly flawed studies.