Go to Host Home Page
Go to Center for Empirical Medicine Home Page


SIDS and Seizures

by Harris L. Coulter, PhD

"Crib death" was so infrequent in the pre-vaccination era that it was not even mentioned in the statistics, but it started to climb in the 1950s with the spread of mass vaccination against diseases of childhood. It became a matter of public and professional concern and even acquired a new name, "sudden infant death of unknown origin,." or, for short, SIDS. This name is significant, in the light of subsequent controversies, since "of unknown origin" means exactly that. So, when the medical establishment assures us that SIDS is unrelated to vaccinations, the obvious response is, How do you know?, if it is defined as "of unknown origin"? At this (as with most common-sense questions about vaccinations) the medical establishment prefers to retire from the debate in dignified silence.

So we have witnessed a steady rise in the incidence of SIDS, closely following the growth in childhood vaccinations. But information on the progress of this epidemic has been radically suppressed in the official literature. Whereas in earlier decades - up to the end of the 1950s - the medical establishment could recognize the fact of death after vaccination, more recently, as the official position has hardened, the earlier concessions have been withdrawn, and vaccinations of all kinds are now declared absolutely safe at all times and in all places. This has required some fancy footwork with the epidemiologic statistics, as we will see below. And since no physician or scientist with a normal IQ could really believe this "epidemiology," one is forced to conclude that the medical establishment, in its wisdom, has decided that 7000-8000 cases of crib death every year are a reasonable price to pay for a nice steady flow of vaccines with all their concomitant benefits for the public health (except, of course, for these same 7000-8000 babies each year who have already enjoyed all the possible advantages of childhood vaccines).

After all, they say to themselves, you can't make an omelette without breaking eggs. But the eggs being broken are small, helpless, and innocent babies, while the omelette is being enjoyed by the pediatricians and vaccine manufacturers. Death after whooping-cough vaccination was first described by a Danish physician in 1933. Two Americans in 1946 described the deaths of identical twins within 24 hours of a DPT shot (on the background and history of SIDS see H. Coulter and B. Fisher, DPT: A Shot in the Dark). E. M. Taylor and J. L. Emery in 1982 wrote: "we cannot exclude the possibility of recent immunisation being one of several contributory factors in an occasional unexpected infant death." But the early 1980s were a turning-point in the official line. In that same year of 1982 matters came to a crisis when William C. Torch, M.D., Director of Child Neurology, Department of Pediatrics, University of Nevada School of Medicine, at the 34th Annual Meeting of the American Academy of Pediatrics, presented a study linking the DPT shot with SIDS. Torch concluded: "These data show that DPT vaccination may be a generally unrecognized major cause of sudden infant and early childhood death, and that the risks of immunization may outweigh its potential benefits. A need for reevaluation and possible modification of current vaccination procedures is indicated by this study."

Torch's report provoked an uproar in the American Academy of Pediatrics. At a hastily arranged press conference he was soundly chastised for using "anecdotal data," meaning (will you believe it?) that he actually interviewed the families concerned! This mistake was not made again. Gerald M. Fenichel, MD, chairman of the Department of Neurology at Vanderbilt University Medical Center, in 1983 published an article on vaccinations entitled "the danger of case reports," and the pro-vaccination literature produced in profusion in later years and decades has generally steered away from and around any such thing as a "case report." These researchers will examine with minute precision hospital card files, medicare cover sheets, even physicians' records, but God preserve us from contact with the children themselves or their families! Another sign of the hardening official position was a two-part article by Daniel Shannon, M.D., in a 1982 issue of the New England Journal of Medicine. Shannon was Director of the Pediatric Pulmonary Unit at the Massachusetts General Hospital and a "principal investigator" of SIDS.

His article on the causes of SIDS (financed by the U.S. Public Health Service) never mentioned vaccination even though, at a 1979 FDA meeting on "The Relation between DPT Vaccines and Sudden Infant Death Syndrome," Shannon had described 200 infants with severe breathing difficulties after a DPT shot, such that they required resuscitation. In 1979 he had said: "We do have all this data. It is all recorded on tabular sheets, and we have it on nearly 200 infants that we have evaluated this way. It is in a capacity that it can be pulled," but in 1982 he preferred not to "pull" this information after all. When Barbara Fisher and I queried him on this in a 1982 letter, he replied: "I did not mention DPT shots in my review article on SIDS in the New England Journal of Medicine because there are no data collected in a scientific way [no anecdotal data, if you please!] that support an association. This includes Dr. Torch's report."

So the cat was let out of the bag by Dr. Torch, who has been effectively silenced by his colleagues since that memorable date. In his editorial attacking "case reports" as a basis for evaluating vaccine damage, Gerald Fenichel alluded to an ongoing study by the NIH on "risk factors" in sudden infant death syndrome which, Fenichel asserted, "excluded DPT as a causal factor in sudden infant death syndrome." Let us take a look at this study, published some years later as "Diphtheria-Tetanus-Pertussis Immunization and Sudden Infant Death: Results of the National Institute of Child Health and Human Development Cooperative Epidemiological Study of Sudden Infant Death Syndrome Risk Factors," coauthored by: Howard J. Hoffman, Jehu Hunter, Karla Damus, Jean Pakter, Donald R. Peterson, Gerald van Belle, and Eileen G. Hasselmeyer (Pediatrics 79:4 [April, 1987], 598-611.

This "retrospective case-controlled study" involved finding 838 children whose deaths had been classified as SIDS by the attending physician and/or the coroner and comparing them with 1514 "controls." The 800 "cases" were selected from among all children who died with a diagnosis of SIDS between October, 1978, and December, 1979, at or near certain designated centers. Excluded from the group were: (1) those on whom an autopsy was not performed or was performed with deviations from the standard protocol, (2) Those younger than 14 days or older than 24 months, (3) those who died after more than 24 hours in a hospital, and (4) those for whom the parents refused permission to perform an autopsy. The selection was made by a panel or panels of pathologists who examined the records of the children's deaths and autopsies and who decided whether or not the child had really died of SIDS or from some other cause.

There are two major objections to this procedure. The first is that the "case" group contained some children who were vaccinated and some who were not. The second is that we are not given the criteria by which the panel of pathologists decided whether or not to include a child as one of the "cases." On the first objection, the investigators are searching for a tie with vaccination in a group of 800+ infants, some vaccinated and others not. This is contrary to common sense. Why water down the sample with babies who were never vaccinated? At this point the whole methodology for determining whether a previous vaccination may or may not have contributed to the SIDS death in question rapidly becomes incoherent. This leads to objection #2, which is that we are not given the criteria according to which children were accepted as "cases" by the panel of pathologists, and we cannot judge whether or not this was done correctly.

A typical SIDS post-vaccination case would be the baby with a slight bacterial or viral infection who is vaccinated and then dies of the infection. These cases are invariably classified by attending physicians and coroners as "death from an infection" without taking into account the fact that vaccinations are known to lower resistance momentarily (for a day or two). In this state of lowered immunity the baby might well die from the infection which would otherwise have been innocuous. So such a case would not even be classified as SIDS (since the infectious "cause" is known), and certainly not as "SIDS after a vaccination," even though the baby would not have died in the absence of a vaccination. How many such cases were rejected by the "panel of pathologists"? We are not told.

The combination of (1) mixing vaccinated and unvaccinated babies with (2) failure to provide the criteria for acceptance into the "case" group taints this same "case" group irredeemably and, in itself, should prevent any further consideration of this study. The next step in the investigation was to select two live "controls" for each "case." Control A was "matched" for age with the corresponding "case," meaning that he or she was born as close as possible to the same day. Control B was "matched" not only for date of birth but also for birth weight and race. Again, as with the "cases," these "controls" were mixed with respect to vaccination status, some yes and some no. The obvious criticism here is that date of birth is simply not relevant to whether or not a baby is vulnerable to the effects of a vaccine (unless the selection is being made on astrological grounds!). Birth weight and race are slightly more relevant, since children of low birthweight and black children (who are more often of low birthweight than white children) are more likely to be affected adversely by vaccination.

However, sex was not included as a criterion, even though males die of SIDS, and are adversely affected by vaccinations, five times more frequently than females. This was a peculiar oversight. The only comment to be made about this "control" group is that it was selected on entirely incomprehensible grounds. It stands to reason that, when one group is being compared with another group, the two groups must be "matched" with respect to the variable being studied. In this case the variable being studied is "tendency to die after receiving a vaccination." Date of birth has nothing at all to do with this variable, whereas weight and race are only marginally related to it. Sex of the baby, which is related, was not included in the analysis.

Even though these two groups are not comparable, Drs. Hoffman et al. compared them anyway, finding that "only" 39.8% of the "cases" had received at least one DPT shot, while 55% of Control A infants and 53.2% of Control B infants had received at least one DPT shot. Since fewer "cases" than "controls" had received the shot, the authors concluded that "DTP immunization is not a significant [what do they mean by "significant?"] factor in the occurrence of SIDS." This sort of attempted comparison can only be described as a shambles, a grotesque imitation of scientific method designed to fool the public (and the journalists who are supposed to be monitoring precisely this sort of intellectual dishonesty). It would have made as much sense to interview the first 1600 people they could pick up in the Greyhound Bus Station and ask them about their vaccination status.

But this article had its effect. Dr. Torch was effectively silenced, and for years this pseudo-science has been cited as one of the medical establishment's principal weapons in its drive to extend childhood vaccination programs. How do you react when your own government lies to you systematically about life-and-death questions? As I have noted earlier, the answer is political action in the state legislatures, and one weapon in the hands of the public is an understanding of the pseudo-science and pseudo-epidemiology represented by articles like this one.

Another article on the SIDS-vaccination relationship, fortunately of far superior quality, is Larry J. Baraff, Wendy J. Ablon, and Robert C. Weiss, "Possible Temporal Association Between Diphtheria-Tetanus Toxoid-Pertussis Vaccination and Sudden Infant Death Syndrome." (Pediatric Infectious Diseases 2:1 [January, 1983], 7-11). The authors adopted a simpler, intuitively obvious method of investigation and concluded that there is, indeed, a "temporal association" between the DPT shot and sudden infant death. They found that 382 cases of SIDS were recorded in Los Angeles County between January 1, 1979, and August 23, 1980, and they simply interviewed the parents of 145 of these cases, either in person or by telephone. They asked: 1) the baby's sex, 2) the age at death, 3) the last visit to a physician or nurse prior to death, 4) the date of the last vaccination, 5) the name and telephone number of the physician or nurse, and 6) the type of immunization given.

They found a statistically significant excess of deaths in the first day and the first week after vaccination, i.e., a "temporal association." They rejected the use of a "control group," and instead relied on the intuitively obvious assumption that "there should be no temporal association between DPT immunization and SIDS were there no causal relationship between these two events." I have not found any criticism of this article for relying on "anecdotal evidence." This study was not financed by the US Government but apparently by the UCLA School of Medicine and the Los Angeles County Department of Health Services.

Another respectable study of the SIDS-vaccination connection is "Diptheria-Tetanus-Pertussis Immunization and Sudden Infant Death Syndrome" by Alexander M. Walker, Hershel Jick, David R. Perera, Robert S. Thompson, and Thomas A. Knauss, published in the American Journal of Public Health 77:8 [August, 1987], 945-951.

This study supports a link between the DPT shot and "sudden infant death syndrome." The authors examined the records of all children born in the Group Health Cooperative of Puget Sound between 1972 and 1983 to see how many had died of SIDS. Total births recorded during this period were 35,581, but of them only 26,500 were eligible for the study. Not all deaths of infants during this period were considered to be SIDS. "All deaths which on the basis of death certificate diagnosis, hospital discharge data, and pharmacy use taken together could be clearly ascribed to causes not related to immunization were excluded." Ultimately, "SIDS was defined as any death for which no cause could be discerned among infants of normal birthweight and without predisposing medical conditions." But, despite these exclusions and restrictions, the authors found "the SIDS mortality rate in the period 0-3 days following a DPT shot to be 7.3 times that in the period beginning 30 days after immunization." They called the results of this study "worrisome" but consoled themselves with the thought that "only a small proportion of SIDS cases in infants with birthweights greater than 2500 grams could be associated with DPT." A particular criticism to be made of this study is that children with "predisposing medical conditions" were excluded and their deaths were not considered to be SIDS, whereas in actuality children with "predisposing medical conditions" are routinely vaccinated.

Another study by the same group, of "neurologic events" following vaccination, is slightly more ambiguous than the preceding one but nonetheless raises a red flag about vaccines. Alexander M. Walker, Hershel Jick, David R. Perera, Thomas A. Knauss, and Robert S. Thompson. "Neurologic Events Following Diphtheria-Tetanus-Pertussis Immunization." (Pediatrics 81:3 [March, 1988], 345-349) was an investigation of the same 35,581 children, born between 1972 and 1983, as in the previous study. The attempt was made to identify "new neurologic conditions" in this group, not by interviewing the families, as might have been expected, but by examining hospitalization records and prescription records for the drugs typically used to treat seizures. Since the pharmacy was "on line" only on July 1, 1976, any drug purchases made prior to that date by families who left the Group Health Cooperative before July 1, 1976, would have been missed, as well as "any child neither hospitalized not treated with drug therapy."

Also excluded from the study were children with "uncomplicated first febrile seizures," because these "are not likely to have been hospitalized or treated with drugs." Also excluded from the study were children whose first seizure occurred prior to 30 days of age - presumably because no vaccinations were given in the first 30 days of life (although this is not stated). Also excluded from the study were children in the category "seizure with possible predisposing cause," such as "trauma, asphyxia, congenital malformation, disorders of metabolism, birth weight less than 2500g, central nervous system infection, and neonatal sepsis." Also excluded were children for whom it was not possible to identify from the available records a clear date of onset of illness.

Ultimately, the group was reduced by 25% - to 26,600. Of course, when studies such as this exclude whole categories of children - presumably those who are particularly vulnerable to vaccine damage - the question immediately arises whether the study is truly a representative sample, since in the "real world" all of the above excluded categories are routinely vaccinated. And if the sample is not "representative," the study itself has no predictive value. The authors found 239 seizures without an apparent predisposing cause among the children in the target population. One case, in particular, is worth describing: "The single seizure that occurred within three days of a DPT was in an 11-month old white girl who suffered a 2 ½ hour generalized tonic-clonic seizure on the evening of her third DPT-oral poliovirus vaccination. Her temperature during the seizure was 39 degrees C. (102.2 degrees F.). Results of CSF studies were normal. There was a transient left hemiparesis and right sixth nerve paresis. She was treated with phenobarbitol. At 6 years of age, while still taking phenobarbitol, she was experiencing rare focal left-sided seizures in the absence of fever and continued to have abnormal EEG tracings." However, this and the other 238 cases were explained away by the authors as part of the "expected incidence" of seizures in this population, a "background" incidence, as it were.

If a "background incidence" is stipulated, one would assume that it had been ascertained in a non-vaccinated population. Instead, somewhat surprisingly, the "background incidence" is defined as the incidence in the vaccinated population later than 30 days after a vaccination. The assumption seems to be that any seizure provoked by a vaccination will necessarily occur within the first 30 days after a vaccination; those occurring later than 30 days post-vaccination are thought to be God-given, a part of Nature, as it were. However, there is no evidence for this. No study of natural seizure incidence, or natural crib-death incidence, in an unvaccinated group of Americans has ever been performed, as far can be determined. Mass vaccination began in the late 1940s, and the medical establishment became concerned about vaccine damage only in the 1970s. Thus they were vaccinating children for over thirty years before they got interested in statistical comparisons; today it is difficult or impossible to locate a group of unvaccinated children sufficiently large to have any statistical value.

Also there seems to be the feeling that not vaccinating a child is "unethical," and that medical research should not venture into "unethical" areas. If that is how they feel, well and good, but they then should not discourse glibly about the "background incidence" of this or that disease or neurologic condition. These sorts of unfounded assertions about the "natural" or "background" incidence of seizures or other kinds of vaccine reactions bedevil nearly every study of this subject. Another trick used by the medical establishment to manipulate public opinion is to cite some study as supporting its arguments when, in actuality, the study came up with contrary conclusions. Sometimes one finds a conflict within the article itself - for instance, the summary or the abstract will make claims which are not supported in the body of the article. Both of these criticisms can be levelled at: W. Donald Shields, Claus Nielsen, Dorte Buch, Vibeke Jacobsen, Peter Christenson, Bengt Zachau-Christiansen, and James D. Cherry. "Relationship of Pertussis Immunization to the Onset of Neurologic Disorders: a Retrospective Epidemiologic Study." J. Pediatrics 1988; 113, 801-805.

This, conducted in Denmark, was of two groups of children who received pertussis and other immunizations at different ages, to see if this affected the dates of onset of neurological conditions. Before April, 1970, Danish children got the DPT shot (together with the Salk polio vaccine) at 5, 6, 7, and 15 months of age. After this date children received the monovalent pertussis vaccine at 5 weeks, 9 weeks, and 10 months of age, and the diphtheria, tetanus, and Salk polio vaccines at 5 months, 6 months, and 15 months. At the time of the change the potency of the pertussis vaccine was reduced by 20%, and the aluminum adjuvant (a frequent cause of reactions) was removed.

This study compared 82,518 births in the 1967-1968 period with 73,390 in the 1972-1973 period. Records of all hospital admissions for seizure disorders and related conditions were examined and "patients whose cases were appropriate for the study were entered into the computer data base." This is the first criticism to be made: the authors do not give further information on the criteria of inclusion. The authors found that the incidence of neurological diseases increased with the new vaccine schedule: epilepsy went from 0.35% (286 cases) to 0.37% (268 cases); febrile convulsions went from 1.01% (830 cases) to 1.87% (1369 cases), and central nervous system infections rose from 0.16% (136 cases) to 0.29% (214 cases).

This could not have been a very welcome finding, and it had to be explained away somehow. Take CNS infections, which almost doubled. The authors write: "there was no relationship between the time of the scheduled administration of pertussis vaccine" and these infections, whereas the accompanying table shows that there was a relationship. They then state that it "appeared to represent a change in the referral pattern" but gave no further details. Furthermore, in the "Discussion" section at the end, the authors went from "appeared to represent" to "was due to": "for CNS infections the change in rate was due to a change in referral patterns." This appears to be simple prevarication. The same occurred with respect to epilepsy. The authors write: "there was no relationship between the age of onset of epilepsy and the scheduled age of administration of pertussis vaccine," whereas the table on the very same page shows that there was such a relationship.

With respect to febrile seizures, they admitted a statistical correlation between the occurrence of first febrile seizures and the scheduled date of pertussis vaccination (p = 0.004). This occurred at the time of the third shot in the 1967-1968 cohort and the fourth shot in the 1972-1973 cohort. They note: "Thus at each period after the usual age of onset of febrile seizures, there was a significant increase in the incidence of febrile seizures in the group receiving pertussis immunization ... 5.9% of all children who developed a first febrile seizure between 28 days and 24 months of age had it as a consequence of fever caused by pertussis immunization." Then they soften the impact of this finding by claiming: "the majority of convulsions that occur within a few days of pertussis immunization are febrile seizures and therefore are only rarely associated with long-term seizure disorders." What does "only rarely" mean? This study has, of course, been cited numerous times in the subsequent literature in support of the total innocuousness of the pertussis vaccine.

Return to the Top


Copyright 1996 by Harris L. Coulter, Center For Empirical Medicine, 4221 45th Street NW, Washington, DC 20016, United States. Phone: 1-202-364-0898. Fax: 1-202-362-3407. Email: hlcoulter@msn.com. URL: http://home.earthlink.net/~emptherapies

Reproduction and dissemination of this article is encouraged but written permission is required.