Why Aspartame does not worry me: Critique of Apartame Disease: An Ignored Epidemic

Aspartame

Critique of  Apartame Disease:  An Ignored Epidemic, H.J. Roberts, MD, Sunshine Sentinel Press, West Palm Beach, FL,  2001.

In general, the book was entertaining, nicely peppered with quotations drawn from a variety of sources, and seemed to make a point (with some serious caveats - see below) that aspartame may induce in some people a sensitivity reaction, the most common form of which was headache and cognitive clouding, and that was entirely reversible upon discontinuation of the aspartame.  It was unclear what proportion of those exposed to aspartame suffered these reactions; we only know the proportion of the responders to his survey (and those who responded to the FDA) who had these reactions.  Nevertheless, even in this group, most did not suffer headaches or confusion (although as many as half may have).  Assuming the book is not an outright fraud (and I have no reason to believe it is, although the self-publication without peer review and the many references to his own work make verification and replication of his findings impossible), questioning patients who have unexplained headaches, memory problems, or visual disturbances about their use of products containing aspartame may be helpful.  At any rate, it is a harmless exercise to discontinue the exposure, which gives an answer within hours or days (according to the author).   I think his point that in a nation of 275 million people, 70% of whom may be exposed to aspartame on a daily basis, the idea that 1200 (about .000436% of the population) may have developed some adverse reaction associated with the exposure warrants at least some suspicion in susceptible patients.  Even low probability events should be considered.

What seems clear from the author's data and the peer-reviewed FDA data, is the following:

• Americans have been exposed to aspartame since the early 1980's;
• The organizations who reviewed aspartame and deemed it safe as a dietary supplement include the World Health Organization, the FDA (Dec, 1980; then 1983 for carbonated beverages), the Centers for Disease Control (November, 1984), and the American Medical Association (July, 1985);
• The organizations who oppose aspartame and believe it is unsafe:  H.J. Roberts, M.D., the Aspartame Victims and Their Friends (295 people), Mission Possible, a volunteer organization.
• Some Americans have reported their belief that aspartame caused various symptoms, the most common being headache and confusion, entirely reversible upon discontinuation of aspartame;
• According to the author's database, this number is 1,200 (.0004% of the American population);
• The FDA has received 7,232 consumer complaints regarding aspartame through April, 1995 (page 71), representing .000026% of the American population, or, more generously, .000038% if one uses 70% as the proportion of the population exposed to aspartame (a number the author cites elsewhere).  Assuming that these complaints are cumulative, correcting for the number of years of exposure reveals that about .000004% of the American population has complained about an alleged aspartame to the FDA each year.   Assuming that the adverse effect would have to affect .1% of the population to be significant, the ratio of those who experienced adverse effects and wrote to the FDA to those who experienced adverse effects and did not would have to be 1: 26,315 (meaning that over 26,000 people experienced adverse effects and didn't contact the FDA for every one who did).
• The vast majority (89.4%) of those who complained to the FDA were classified as having a "mild to moderate" reaction.  Only 10.6% were classified as having a "severe" reaction.  This means that only .0000038% of the American population exposed to aspartame had a severe reaction and wrote to the FDA.
• The most common FDA-reported symptom was headache (28%) followed by dizziness (11%) and change in mood (10%).  No control data are given for the background rates of these complaints in the general age-adjusted population.
• Note that these data do NOT indicate that 28% of those exposed to aspartame develop headache, only that 28% of those who reported to the FDA about side effects (.000038% of the American population exposed to aspartame) experienced headache.   One can assume of course that the rate of headache among aspartame-consumers in the general population is higher or lower, but since we do not have controlled data, we can only speculate.  We could reason that those who complain in general would have a higher complaint rate about any specific symptom than the general population.

Those are the basic facts, as presented by the author.  We do NOT know:

• The rates of headache, confusion, dizziness, etc., in the general population;
• Whether those rates are rising post-1984 (when aspartame was introduced en masse in soft drinks) - if they are not, and 70% of the population is exposed, then the population-level effect if any must be weak;
• Whether even if those rates, on an age-adjusted basis, were rising, aspartame is the proximate cause of these ailments.

In general, there are several serious flaws in the book that would make me hesitant about changing my personal or clinical practice:

1. The book was not peer-reviewed.  It was not, by his admission, even edited.   A professional scholarly editor at least could have pointed out obvious flaws and identified potential errors.  It appears to have been self-published (I have not heard of Sunshine Sentinel Press; not to be snooty, but if it had been published in by John Wiley & Sons, let's say, it would have had to go through several editors before it would even have been considered for publication.  As an author, I know that that process entails considerable review by multiple sets of eyeballs that help improve accuracy and eliminate hyperbole.)
2. He mentions repeatedly that he is "corporate-neutral."  He does not disclose, however, his own personal stake in promoting his theories, either by selling more copies of his book, being invited to more television appearances, or recruiting more patients for his practice.   Although I applaud his independence, he also has an agenda, one that no doubt is positively reinforced.  This does not per mean his ideas are without merit, but it also does not mean that the corporate-sponsored research he dismisses is worthless.  Biased, yes, worthless, no.  The FDA is far from perfect, but aspartame had substantially more review than did his book.
3. The book is based almost entirely on anecdotal evidence.  Although, as his quotations indicate, case reports almost always open the door to more research so should not be dismissed out of hand, the key here is that in other examples (such as with digitalis, for example), rigorous, controlled trials followed initial observations.  These trials were controlled and the observer, who has a vested interest in a particular outcome, is blinded to both exposure.   We know very little about his questionnaire, who it was sent to (how was the sample selected?), and the critical response rate.  Generally, a response rate below 80-85% indicates a serious risk of selection bias (something about the respondents differs significantly from those who don't respond, meaning any conclusions about this group cannot be generalized to the larger population).  We know, for example, that 43% of his sample of 1,200 responders experienced headaches.  What we don't know is what percentage of the total population to whom he mailed (or gave) the surveys experienced headaches.  In a general outpatient practice, headache is a very prevalent condition.  Is this statistically significant?  We just don't know because he has not shared these data.  He also, without explanation or elaboration, uses 649 for the denominator of some symptoms, lumping them in with the rest of the responders.  If he is splitting the sample in 2, I would like to see both raw and adjusted proportions and why he chose one versus the other.  If he changed methodologies in the middle of his study, then he has to explain why.  (pages 68-71)
4. He has a tendency to cite himself perhaps excessively, but his publications (if you dig through the sources (pages 1010-11)) were mostly written in the 1960's and mostly in local or non-peer-reviewed journals.  I spotted one NEJM publication.  Once again, the importance of peer review can't be overstated for quality assurance.  If his case is so strong, why is the consensus (and the FDA and the WHO) against him?  Why can he not find a home for his work in journals edited and read by his colleagues?  He mentions at one point his explanation (part of the vast conspiracy of corporations and biased editors) which overlooks another explanation (that his hypotheses, although passionately held, failed to rise above the level of personal conviction to the plain of scientific proof).
5. He cites the Internet as some kind of propagator of truth, of the little guy winning out over corporate interests.  His opinion of the Internet as a source of legitimate information is much higher than mine.  Anyone with an Internet connection and a browser can throw up a web site and it is impossible to verify anyone's credentials or background.  The propagation of misinformation is much easier.  There is no peer review, no checks and balances, no quality assurance.  Anyone who uses the Internet as the primary source of medical information is being reckless.
6. He has a tendency to make sweeping conclusions from limited theoretical research findings.  For example, he states that aspartame caused increased parathyroid concentrations in two aspartame reactors (page 437 - note this equals .16% of his 1200 responders).  He does not tell us whether the increase was significant or what the rate of increase of parathyroid hormone in a control sample of 1200 would be expected to be, but this does not prevent him from extrapolating that that aspartame may bind to calcium-ion receptors, leading to this finding, which he associates with gynecomastia.
7. He takes the patients' self-report at face value.  A mailed in survey tells us only two things:  that the respondent endorsed the symptoms and (in this case) had some aspartame exposure.  Was there some attempt to quality check this self-report, perhaps by sampling a statistically significant sub-group on some objective, measurable criterion, then extrapolating from this?
8. He assumes that the direction of reporting bias must be negative (toward underreporting).  The exact opposite could be true, that is, that an information cascade (that he elsewhere uses to discount other faddish illnesses such as the "Persian Gulf Syndrome" or chronic fatigue syndrome) could be at play leading to overreporting.  It is noteworthy that he does not consider this, even as he uses at one point a caller into a radio talk show program as a "representative sample"; a message beamed to a large audience is going to generate responses.
9. He fails to differentiate between what seem to me to be at least 2 distinct populations in his sample (of course, we know very little about his sample anyway, but this is the most egregious error):  those who assume extraordinary quantities of aspartame (aspartame abusers, by his count perhaps 5% of his sample) and those who experience severe adverse, idiosyncratic, dose-independent responses (aspartame sensitive).    Most of the severe adverse "outcomes" he reports, such as seizures, occurred among people consuming a liter of soft drink a day.  This seems to represent an independent problem, as does the sensitivity issue, from the larger public health question:  can most people most of the time be exposed to aspartame with few adverse side effects?  Neither his data nor those of the FDA prove that they can't.  An analogy with alcohol could be helpful here:  we know that most people consume alcohol responsibly, but some, perhaps 5% of the population, abuse it or become dependent.  A much smaller proportion, perhaps far less than 1%, cannot tolerate it all, getting an anaphylactic type of reaction with flushing, sweating, headache, etc.   One cannot reason therefore that because some of the population abuses or is sensitive to alcohol that it should be banned or that it poses a major public health risk to all Americans, although that, it seems, is what the author is trying to do with aspartame.  We know that about 15-20,000 Americans die each year on the highway from alcohol-related deaths.  We have no such data on aspartame; all we know is that a small proportion of those exposed report headaches and mild, reversible symptoms, and that a much tinier proportion report more severe symptoms that appear idiosyncratic and dose-independent (by his description).   If the substance should be banned based on this, then so too should Tylenol, aspirin, and ibuprofen, all of which are associated with much more prevalent health risks.  Also, a strong argument could be made for the banning of fresh fruits and vegetables or poultry and meat since thousands of Americans die each year from food poisoning.
10. The author demonstrates that powerful, wealthy corporations have a vested interest in insuring that aspartame is not yanked from the market.  In essence, he subscribes to a vast conspiracy theory propagated by a profit-hungry industry that does not care about the well-being of the general public.   However, it takes more than a motive to establish a crime.  You must also demonstrate that a crime has been committed (i.e, that there is demonstrable harm that rises to the challenge of a public health risk), and explain why so many disinterested third parties, such as the FDA, the WHO, and the CDC, would sign off on such a deal.   Why would a salaried government official sign off on something that might theoretically benefit a distant CEO?  Is he asserting that there was outright bribery?  If so, where is the smoking gun, and why has no one "blown the whistle"?  It takes only one informant to ruin a conspiracy and he posits that this conspiracy involves multiple agencies in multiple countries.  Also, if the federal government is so corrupt (or inept) then why did this conspiracy succeed when others failed (the list of medications and food additives that do not get FDA approval or are yanked from the market is much longer than the short list of those that do make it, and those that do are subjected to critical ongoing scrutiny.  If anything, the government overreacts in a draconian fashion, pulling products that have merit and save lives (such as DDT in the developing world where far more people die of malaria than of pesticide toxicity)).
11. I object strongly to the use of the word "Epidemic" in the title.  One of his quotations refers to the swine flu epidemic of the early 20th century, a disease then killed millions, more than died in World War I (whose final year overlapped with the epidemic).  Even if the author's worst concerns are true, all we know is that aspartame may induce uncomfortable symptoms in a subset of those exposed; he most certainly does not establish the case for millions of deaths.  If so, then why is life expectancy rising in a linear fashion since the introduction of aspartame?
12. He makes some rather loose associations between aspartame and e.g., the rise in teen suicide.  He does not discuss the role of substance abuse or alcohol, which we know are independent risk factors for suicide, or why he feels a Diet Coke would be more likely to disrupt mood than crack cocaine or alcohol.   In refutation he offers an anecdote about 1 of his respondents who said it was subjectively harder to wean himself off of aspartame than alcohol.  No vital signs or objective evidence of withdrawal was offered.
13. Although I agree strongly that the "cult of thinness" is overdone in this country, I also understand that obesity poses a far greater risk to the population (about 30% of Americans are morbidly obese) than does anorexia (an illness that affects less than 5% of the population, and usually transiently).   I agree also with the author that bouncing your weight up and down is probably associated with long-term health consequences, but do not agree with his idea that aspartame leads to weight gain.  His own sample data do not support this assertion.  As many respondents (3%) had severe weight gain as weight loss, and we do not know if either is statistically significant.
14. The vast majority of new cases of diabetes the author cites on page 756 are Type II diabetes (adult onset) secondary to obesity and a sedentary lifestyle.  The role, if any, of aspartame seems remote and theoretical.
15. He completely fails to account for confounders throughout.  The most glaring examples:  the failure to distinguish between caffeine dependence, which is real and demonstrable, and aspartame dependence, which he posits, but is  seemingly impossible to separate, since as he admits, most of the products that are abused by his respondents have both aspartame and caffeine.   Another glaring example:  the risk of osteoporosis from consumers of carbonated beverages (??; page 757).  We KNOW that young women who eat a normal diet in America get far too little calcium; those who diet are much more likely to get even less calcium. Obviously, as he admits, most of those who consume aspartame-containing foods are dieting (calorie-restricting), so how can he assume that of all the variables at play aspartame is the causative agent of osteoporosis?  It defies logic to leap to this conclusion.  He accuses aspartame of causing seizures, yet fails to note that caffeine potently lowers seizure threshold (and is therefore used in ECT to help induce seizures).
16. He has very dramatic headings followed by very little in the way of substance, such as "International Experience" (citing India - his sole example - forcing people to add sugar to soft drinks), and "Plane Accidents and Near-Accidents" (p. 790) followed by a some anecdotes about pilots and seizures and perceptions of poor eyesight.  No data is offered to support his idea that aspartame might be linked with accidents (such as the rate of aspartame consumption among pilots who were in accidents versus the rate of aspartame consumption in those who were not, as a start).  Since pilots as a group are probably more likely than the general population to use aspartame-containing products, one would have expected an increase in the passenger-mile fatality rate since 1984; the opposite has occurred.   He informs us that there are many near misses, but does not make the link to aspartame, nor does he inform of what the number of near misses prior to 1984, adjusted for flight frequency.
17. His links to severe illness are weak; for example, he cites a colleague (page 802) who estimates an odds ratio of 1.24 among patient with brain tumors who consumed low calorie drinks versus controls, but no p is given.  An OR of 1.24 is nothing to write home about, and may be due to chance.
18. The author neglects, discounts, or simply fails to mention any data that may refute his ideas about aspartame.  Nothing in public health or medicine is so black and white.  There is always contradictory, noisy data.  Patients have multiple exposures and multiple possible causes of a given adverse outcome.  His examples are too neat, too clean.  What about his failures, the patients whom he suspected of having aspartame sensitivity but didn't?  He includes them in his survey but does not include them in his "representative" examples.  How do we know they are representative - because he tells us?  What about negative studies?  To show evidence of corporate bias (who else would spend millions sponsoring studies to get approval of aspartame but the corporation with something to gain from its approval?) is not enough to refute the studies in question.  What were their sample sizes?  How were the samples selected?  What was his critique of their methodology or data?  He does not give any concrete criticism, only general condemnation.  After all, he is "corporate-neutral."
19. He discusses at length why he believes aspartame causes "Gulf War syndrome" even though the weight of large-scale epidemiological data indicates that Gulf War veterans had lower rates of complaints of fatigue, irritability, headache, depression, etc., than age-adjusted controls who were not Gulf War veterans.  This illustrates precisely the problem that any population-level study of complaints by people making an association between exposure and outcome runs into:  there is a background rate of unexplained somatic and psychiatric complaints in the general population.  To claim someone has a headache and drinks aspartame is not proof that aspartame causes the headache.  If you can demonstrate that 3 times as many volunteers who consumed aspartame suffered headaches as controls who consumed a placebo, then perhaps you are onto something.  But if you don't know the rate of headache (or symptom x) in the general population then you can't even begin to generalize.
20. He assumes, somewhat grandiosely, that because he is taking an unpopular stand, and that people (he cites Gandhi and Confucius among others) who later proved to be right in the past took unpopular stands in their time, that therefore he must ultimately be proven right.  History is full of unpopular ideas that were not only unpopular but wrong (such as phrenology).  Simply because his ideas are not widely accepted does not mean that he is onto some hidden truth.
21. Similarly, the observation that a belief is wide-spread does not disprove it, such as his rather wholesale condemnation of "the new deity of clinical medicine", evidence-based medicine (page 860).  What would he offer in its place, a collection of colorful anecdotes?  He needs to read the struggles of another unpopular (in his time) physician, Semmelweis, who argued the exact opposite - that through the first controlled trial in the history of medicine that physicians should wash their hands before delivering babies and that it was hand-washing, not bad miasma, that critical to preventing puerperal fever.  Only after Semmelweis had been dead a century did modern medicine finally get it and realize that the majority of the concoctions they thought were helpful were actually less than worthless.  Only through rigorous, blinded, controlled trials could they demonstrate the efficacy (or lack thereof) of medications (or environmental exposures).
22. He cites (p. 871) the ethical lapse at the NEJM as proof of his theories on aspartame.  The term "true, true, and unrelated" comes to mind.  Yes, the staff was wrong, but this not make Dr. Roberts right.
23. Finally, he doth protest too much.  Frankly, the "all aspartame all the time" theme grew tiresome after a few hours (earlier, actually).  I think he could have accomplished his mission in far fewer pages.  I get it already:  he and many others feel misunderstood and neglected by the corporate fat cats and inept regulators who have a vested interest in shutting him up and discrediting him.   To be honest, at certain points in the book he sounds shrill.  He gives great detail about canceled interviews and rejected articles as "proof" of corporate malfeasance and collusion against him as defender of the little guy.   Enough already!  As an example of his shrill language, he equates "suffer[ing adverse] reactions]" with "being victimized!" [His exclamation mark, p. 907]  His description of Mission Possible (p. 918) is equally hyperbolic and alarmist:  [Mission Possible] rose out of a desperate need to warn the world.  It is estimated that 3 out of 5 people using aspartame have the symptoms or some disease."  ??  Even his own data, of those who endorsed symptoms they associated with aspartame, do not support this assertion.  Less than half have the most prevalent symptom, headache.  On page 922, he seems to support the idea that aspartame is "not simply a national scandal - it's genocidal!"

In summary, I must agree with Dr. Arturo R. Rolla, who slammed the author's 1988 book on Aspartame in the NEJM:  "This type of book… only adds to public misinformation, confusion, and mistrust" (p. 875).  It is hard enough to get patients to make good, sensible decisions about reducing risk factors to illness and death that have been unquestionably demonstrated, such as:

• avoiding tobacco, especially cigarette smoking;
• avoiding excess alcohol and substance abuse;
• avoiding drinking while driving;
• maintaining a healthy, balanced lifestyle with a sensible diet and reasonable exercise regimen;
• treating chronic conditions such as diabetes, hypertension, and major depression.

We know that heart disease, drunk drivers, and tobacco kill hundreds of thousands each year in total just in this country alone.   With limited clinical resources, we should be focusing on the many barriers to people modifying their exposure to these known and overwhelming risk factors (clear and present danger, if you will).  As mental health professionals, we should be helping clients overcome the extraordinary stimgatizing and mistrust that prevents people from seeking the latest in treatment, both psychotherapeutic and psychopharmacological.  This is a tall order without throwing in additional tasks over less demonstrable, but arguably important risks.   In an ideal world, we could do it all, but our time and energy is limited, and the attention span of our patients is also limited.  Focusing compliance on a few key areas probably would be much more helpful than trying to do too much, especially if it threatens to undermine our credibility as clinicians and scientists.  We may have strong, personal convictions about the choices we make, but I think we have an ethical duty from imposing those personal convictions on our patients unless we can demonstrate empirically that they are significant and evidence-based.


Methyl alcohol:
The author states that the average daily intake from natural sources is < 10 mg; aspartame beverages contain 55 mg/ liter (no source is given).  Therefore a can of Diet Coke (240 cc or 8 fluid oz.) would contain roughly 13 mg of methanol, or about 30% more than the average daily methanol intake.

Weight gain:
The author notes (p. 389) that 3% of his sample of positive responders gained excessive weight; however this number (40) was identical to the 3% who LOST excessive weight.   After spending some time explaining why at the theoretical level aspartame might cause anorexia, he then turns around and explains why it may cause weight gain.