Cranford, NJ - The newly released study on autism and the
measles-mumps-rubella vaccine (“A Population Based Study of Measles, Mumps and Rubella Vaccination and Autism.” New England Journal of Medicine, Vol 347, No 19; Nov 7, 2002: 1477-1483, by Kreesten Meldgaard,et al) is a welcome addition to autism epidemiology. Unfortunately, the study conclusions appear overreaching, claiming that this analysis is the final word on autism and vaccines and implying that more research on the topic is unnecessary. Safe Minds asserts that other vaccines besides MMR may be involved in autism, and that only biological research, not epidemiology, can answer the question of whether the MMR vaccine plays a role in autism.

      “It is important to note that the study only focused on the MMR
vaccine, and not vaccines also implicated in autism which contain the
mercury preservative thimerosal,” explains Sallie Bernard, executive
director of Safe Minds.  “The study also failed to investigate whether the
MMR vaccine might be interacting with the thimerosal from other vaccines to increase the severity of symptoms in children who already have autism.
Finally, the study did not differentiate between regressive autism, which is the type being linked to MMR vaccine, and the more prevalent early onset autism, which is the type being linked to thimerosal.”

      Safe Minds is an advocacy organization which focuses on the role of
mercury in neurodevelopmental disrorders, including autism. It was founded by parents of autistic children.  Thimerosal contains 50% ethylmercury and has been used in most recommended childhood vaccines, including the Diphtheria-Tetanus-Pertussis (DTP), Haemophilus influenzae type B (HiB), and Hepatitis B (Hep B) vaccines.

      Research studies have shown that mercury exposure in utero or during
early postnatal life – the time when thimerosal vaccines are being given –
can cause immune system abnormalities which predispose the child to ongoing viral infections. It is biologically plausible that this immune disruption may have allowed the live measles virus component in the MMR vaccine to persist in susceptible autistic children, making the symptoms of the disorder worse.  This connection would not be detected through an
epidemiology study like the Denmark one. Nor does the Denmark study have the power to detect differences in rates of regressive autism between vaccinated and unvaccinated children, since the number of regressive cases – estimated to be 10%-20% of all autism cases - would be too small.

      “The overreaching conclusion of the study should not obscure other
important findings from this extensive and well planned analysis from
Denmark,” continued Ms. Bernard.  “The authors report an increased
prevalence of autism in that country, and thus it supports other recent
studies that are also showing increases. This rise tells us that an
environmental agent is at work worldwide that is driving this trend. We
believe that thimerosal and environmental mercury – which are worldwide
pollutants – are behind the surge. Also, Denmark has had lower and later
exposures to thimerosal in vaccines, and the report shows that their rate of autism is lower than in the US, which is also consistent with a thimerosal connection.”

      Safe Minds is encouraged that the Centers for Disease Control
sponsored such an extensive study on autism, which shows that this terrible disease is finally getting the attention of public health officials. Safe Minds looks forward to increased support for autism research, especially at the biological level.

Positives
   ·  The CDC and public health authorities are investing dollars and
efforts into autism research.  These efforts should be applauded, and
expanded!

   ·  Shows steep rise in autism rates from 1980s to 1990s (from <2.0 to
>10.0 per 10,000). Increases are being reported in other countries, again
suggesting environmental influences at work, as the recent landmark MIND Institute epidemiology of California study did.

   ·  Supports thimerosal hypothesis?? (This assumes that we are able to
obtain the information which was requested from the Congressional Research Office on the level of thimerosal exposure in the Danish population during the time period 1986-1998.)

   ·  Acknowledges that previous attempts to refute MMR-autism hypothesis were too poorly designed to reach definitive conclusions. (p.1477, 2nd paragraph on right)

   ·  Has identified a rich database of information (i.e., Danish
registries) on which additional studies of autism can be based.

Cautions
   ·  Lower incidence of Danish autism group vs rates reported in US and UK limit applicability to other countries. [see table at end of document for
statistics]

   o  Same diagnostic criteria developed by CDC used in Brick, Atlanta, and
Denmark, so lower incidence not a factor of variation in diagnosis.

   o  Means other environmental factors, rates of factors, or combination of factors may be at work in Denmark vs US or UK.

   o  It is possible that MMR increases the rate of autism only when acting
in conjunction with another environmental factor, such as mercury. If that
factor’s prevalence is not controlled for among the study groups (vaccinated vs unvaccinated), it would obscure the role of MMR as a causative factor in the study.

   ·  Only psychiatric records were accessed, not medical records, so there
were no data on gastrointestinal symptoms and no taking of CSF or GI samples to detect presence or absence of measles virus. Cannot tell if measles persistence is impacting a subgroup of children, if any. Measles persistence may be increasing the severity of autism, even if it is not causing an increase in the number of cases.

   ·  There was no attempt to differentiate between regressive and
early-onset forms of autism. Since the regressive form comprises a minority of cases – 10%-20% - the power to detect whether there was a difference in regressive autism prevalence between MMR vaccinated and non-vaccinated is lacking in this study.

   o  The assertion that a relative risk of autism of less than one rules
out the possibility that there are important subgroups is false.

   ·  Although overall well designed, there appear to be some methodological problems with the study which need further elucidation from the investigators and raise questions about its conclusions of being the “definitive” MMR-Autism study.

   o  The study covered 8 birth cohorts, but two of these, those born in
1997 and 1998, were only 1 or 2 years old when the data records were
obtained. These age groups are too young in most cases to be diagnosed with autism or to be immunized with the MMR. This might have been fine if the impact applied equally to both vaccinated and unvaccinated groups. However, fully half (50.6%) of the unvaccinated group fell into these two younger birth cohorts, vs. just a fourth (27.7%) of the vaccinated group.  This issue involves approximately 32% of the sample.

   o  Children who were in fact vaccinated were assigned to the unvaccinated group if they were diagnosed with autism before they received the MMR. The reassigned cases comprise 10% of the unvaccinated autism cases (13 out of 130). This commingling blurs the distinction between vaccinated and unvaccinated. It is not clear what effect this would have on the results.

   o  A number of the measures used to arrive at the conclusion that autism and autism disorders were not associated with MMR vaccination are irrelevant.  Age of vaccination with MMR, time interval between receipt of MMR and diagnosis of autism, and year of MMR vaccination do not help
elucidate the hypothesized relationship between receipt of MMR and
development of measles-related symptoms and regressive autism.  The child’s age at time of diagnosis is arbitrary and can vary for many reasons, among them differences in severity of illness, access to care, and clinician skill and preference. Thus these measures cannot be used to refute the presence of a temporal relationship between MMR and onset of symptoms of measles-related illness and regressive autism.

      As the authors point out on page 1481, they had no information on the presence or absence of a family history of autism, which could explain the study’s negative findings only if families with a history of autism avoided MMR vaccination. It should be noted that in 1993, there was a widely reported news story in Denmark about a parent with autistic twins who asserted that their autism was caused by the MMR vaccine. It is entirely possible that parents with either (a) a family history of autism or (b) an infant or toddler with emerging symptoms of autism, would avoid vaccination at a higher rate than other parents. This would inflate the unvaccinated group with children of families predisposed to autism.

Reported Rate of Autism from Recent US, UK, and Denmark Studies

Study Group                                                        Rate per 10,000

Denmark MMR-Autism Study, 2002
  Broader autism (738 cases ¸ 537,303)
  among 1-8/9 year olds comprising study sample                    13.7
  Classic autism (316 cases ¸ 537,303)
  among 1-8/9 year olds comprising study sample                     5.9
  Other PDD 7.9 per 10,000 (422 cases ¸ 537,303)
  among 1-8/9 year olds comprising study sample                     7.9
  Classic autism among 5-9 year olds in 2000 (data not sho        >10.0
  Broader autism among 8 year olds in study sample                 29.9
  Classic autism among 8 year olds in study sample                  7.7
  Other PDD among 8 year olds in study sample                      22.2

Chakrabarti and Fombonne UK Study, 2001
  Broader autism among 2-6 year olds in UK                         63
  Classic autism among 2-6 year olds in UK                         17
  Other PDDs among 2-6 year olds in UK                             46

Brick Township, NJ, by CDC, 2000
  Broader autism among 3-10 year olds in 1998                      61
  Classic autism among 3-10 year olds in 1998                      31
  Other PDDs among 3-10 year olds in 1998                          30
  Broader autism among 6-10 year olds in 1998                      67
  Classic autism among 6-10 year olds in 1998                      40
  Other PDDs among 6-10 year olds in 1998                          27

Metro Atlanta, by CDC, 2001
 Classic autism among 3-10 year olds in 1996                       34

* * *
      For more information, contact:
      Sallie Bernard Executive Director Safe Minds 970 429-1460
      sbernard@nac.net