March 5, 2015
Influenza vaccine studies typically do not follow the established rules of medical science and evidence-based medicine. Specifically, influenza vaccine studies are not randomized or placebo controlled. The vaccine advocates say this would be unethical because it would be denying some study subjects a beneficial treatment.
The problem with this argument is that it presumes the influenza vaccine is beneficial. But this must be established by science; it cannot be assumed. Proof of benefit requires randomized, placebo-controlled, double-blind studies. So claims about influenza vaccine efficacy (and safety) are based on assumptions, not scientific evidence.
Influenza vaccine studies DO use unvaccinated controls, but they are not randomized or placebo-controlled or blinded. Consequently, all these studies have severe problems with selection bias and confirmation bias. The people getting the vaccine tend to have more “health seeking” behavior than controls, and the investigators know who got the vaccine and who didn’t. This issue is explained in detail here: Healthy User Bias: Why Most Vaccine Safety Studies Are Wrong
Additionally, the investigators often have a financial interest in making the vaccine look good.
A placebo-Controlled Study in Children
Following is a properly conducted (randomized, placebo-controlled, and double-blind) study of the influenza vaccine (the trivalent inactivated Vaxigrip by Sanofi Pasteur).
In this study 115 children (age 6-15 years) were randomized to receive Vaxigrip or placebo. Vaccines and placebos were administered in Nov and Dec, and the subjects were followed for 9 months. They were monitored for illnesses and viral infection. Laboratory analysis determined the type of virus present, in case of illness.
Full paper (Cowling, 2012): Increased risk of noninfluenza respiratory virus infections associated with receipt of inactivated influenza vaccine
The results were not good for the vaccine. The rate of influenza infection was almost exactly the same on both groups (4.3% vax vs 6.5% unvax). Most significant of all, the vaccinated group had a 4.4-fold higher risk (statistically significant, with P<0.01) of experiencing a non-influenza infection.
Here are the details:
Above: Trivalent inactivated vaccine (TIV) did not reduce influenza, and it caused non-influenza infection. Children that received the influenza vaccine were 4.4x more likely to suffer a non-influenza infection. From Cowling et al. 2012.
29% (20) of those receiving the vaccine got a non-influenza infection, compared to 6.5% (3) of the unvaccinated. These results were highly statistically significant, with P<0.01.
The most common non-influenza infection were due to rhinovirus and coxsackie virus. These viral illnesses are typically mild, but can cause serious harm in susceptible people. Also, other respiratory viruses were reported.
The authors state:
“We identified a statistically significant increased risk of noninfluenza respiratory virus infection among TIV recipients (Table 3), including significant increases in the risk of rhinovirus and coxsackie/echovirus infection….”
“Vaccine recipients may lack temporary non-specific immunity that protected against other respiratory viruses.”
The obvious concern here is that immune impairment caused by the vaccine can increase susceptibility to other infections that are much more dangerous than influenza, rhinovirus or coxsackie.
If its non-specific immunity thats impaired (hypothesized by the researchers), there will be increased susceptibility to a wide variety of viruses, and possibly other types of pathogens (bacteria, fungi). It is not known how broad or long-lasting the immune impairment is.
The non-influenza infections occurred mostly in January and March, 1 and 3 months after vaccination. This of course implies that the immune suppression from the vaccine lasts for at least 3 months. See Figure 1 in the above paper.
Influenza Vaccine in Young Children
An increased rate of non-influenza illness was observed in a separate study of the influenza vaccine, by a different research group in Sydney, Australia. Subjects (381 total) were children, aged 6-35 months. This was not a randomized placebo-controlled study like the Cowling study above; the subjects (i.e. their parents) chose whether to receive the influenza vaccine or not. Hence, the study was susceptible to selection bias (i.e. vaccinated and unvaccinated groups might not be matched). Subjects were assigned to three groups, depending on which vaccines they chose to receive:
Fully vaccinated (n=91): Typically received two doses
Partially vaccinated (n=52): Typically received one dose
Unvaccinated (n=238): Received no influenza vaccine
Subjects were tracked for 13 weeks. Fully and partially vaccinated subjects had increased risk for non-influenza infection. There was no difference in influenza. These results confirm the Cowling 2012 study above.
Full Study (Dierig et al, 2014): https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4181477/pdf/irv0008-0293.pdf
Key results are shown below.
Above: Full vaccination for influenza (2 doses) was associated with a 1.6x risk of non-influenza infection, and a 1.23x risk of virus detection (in sinus or throat) compared to controls. There was no difference in influenza infections. Subjects were age 6-35 months and followed for 13 weeks. The unvaccinated group had fewer past hospitalizations. The unvaccinated children were healthier both before and during the study. From Dierig et al, 2014.
Dierig et al describe their results:
“We did, however, unexpectedly find that non-influenza ILI (influenza-like illness) occurred about 1.6 times more commonly in children vaccinated with one or two doses of the influenza vaccine than in unvaccinated children. These results support the findings of a recent RCT (randomized controlled trial) reported by Cowling et al.”
“The non- influenza virus incident rate ratio (IRR) was higher in the Hong Kong study (4.4 versus 1.6), but there are some key differences to our study, including age of subjects, follow-up period, proportion of illnesses swabbed and proportion of swabs yielding viruses. As with all observational studies, bias must be considered.”
“Cowling et al” refers to Cowling 2012, cited above
Interestingly, unvaccinated children had significantly (P=0.01) lower hospitalization rate than fully or partially vaccinated children, even before the study started. This may be because children not receiving the influenza vaccine received fewer first-year vaccines (based on an assumption that those refusing the flu vaccine also refuse other vaccines). Hence, the observed differences in hospitalization may be explained by vaccine injury from first-year vaccines.
Dierig suggest the hospitalization rate difference may be because “families that vaccinate children have a prior preference for greater healthcare service usage.” But is it preference or need? This speculation by Dierig is contradicted by the data in the paper. Specifically, the paper states:
“We could find no evidence of different parental responses to ILIs (influenza-like illness) in vaccinated and unvaccinated children: parents of vaccinated children were no more likely to seek medical care during an ILI.“
In other words, the groups had the same tendency to seek medical care. Hence, the most reasonable explanation (for the higher illness rate among the vaccinated) is that the influenza vaccine caused non-influenza illnesses. And of course, this is consistent with the Cowling study, which was higher-quality since it was a placebo-controlled randomized trial. The Cowling study proves causation, because it was randomized.
The Dierig study is an excellent example of the ambiguities and selection bias inherent to observational (i.e. non-experimental) trials. Since there was no randomization, the groups are not matched and therefore one can interpret the results in numerous ways. The problems created by selection bias in vaccine safety research are described here: http://vaccinepapers.org/healthy-user-bias-why-most-vaccine-safety-studies-are-wrong/
Randomized, Placebo-Controlled Trial in Adults
An increased illness rate was also observed in a study in adults. This is a great study because it lasted for 2 influenza seasons. In the first year, circulating influenza was not well-matched (50% efficacy, which is typical); in the second year, circulating influenza was very well matched (86% efficacy, which is exceptionally high).
This study was large, with 1130 and 1178 participants in years 1 and 2. Half received the vaccine, and half received a saline placebo. Unfortunately, this study was not well-blinded because a vaccine produces irritation at the injection site that does not occur in those receiving saline placebo. So, many study subjects knew which group they were in.
The study found that in year 1 (97-98), when the vaccine had low (but typical effectiveness), the vaccine caused increases in illness and increases in lost workdays, lost hours and days ill. In year 2 (98-99), when the vaccine was well matched, the vaccine reduced illnesses, lost workdays and physician visits.
Full Paper (Bridges et al): Effectiveness and Cost-Benefit of Influenza Vaccination of Healthy Working Adults: A Randomized Controlled trial
Above: Results of randomized placebo-controlled trial of influenza vaccine, in two flu seasons. In the first year, the vaccine had poor efficacy and caused illnesses. In the second year, the vaccine was exceptionally well matched and reduced illnesses. Vaccine effectiveness for influenza was 50% and 86% for first and second year, respectively. The vaccine increases susceptibility to influenza when it is not well-matched to circulating influenza strains. This is consistent with animal experiments and the “original antigenic sin” effect, which is well known in immunology. Statistically significant (or almost significant) results are marked in red. From Bridges et al, 2000.
So the influenza vaccine increases illnesses when its poorly matched, and reduces illnesses when its well matched. This suggests that the flu vaccine may provide no net benefit when its used for several years. Flu vaccine efficacy is not predictable, and efficacy data is generally not available until flu season is well underway, so at the time when the vaccine is received, one does not know if it will cause or prevent illness.
This study was conducted by the CDC. The CDC is powerfully biased to promote vaccines and dishonestly downplay adverse effects. Consistent with this, the paper ignores the adverse effects caused by the vaccine in year 1. For example, the Bridges paper dishonestly characterizes the results like this:
“[In year 1] vaccination did not reduce ILI (influenza-like illness), physician visits, or lost workdays.”
“No other adverse effects, including fever, myalgia, headache, fatigue, rhinitis, or sore throat, were reported significantly more often by vaccine recipients, nor did they report significantly more lost workdays or physician visits.“
These statements are blatantly misleading or false. In year 1, the vaccine caused increases in lost workdays, hours lost to physician visits, and number of days ill. Incredibly, the paper does not discuss these findings.
In year 1 the vaccine increased increased illness severity. This is expected in view of the Cowling results above and the evidence of immune suppression by the influenza vaccine (more on this below).
The Bridges study only wants to talk about the results they like. They mention the non-significant increase in illnesses, but not the large increase in number of days ill.
Original Antigenic Sin
A poorly-matched influenza vaccine may cause illness (and increase risk of influenza illness) by a phenomenon in immunology known as “original antigenic sin” (OAS). First discovered in 1960, OAS is well known and firmly established. Its described in any immunology textbook. OAS occurs in this scenario:
1) Immunity is created by a first exposure to strain #1. The immune system learns and remembers how to respond to strain #1. Pathogen can be virus, bacteria etc.
2) A second exposure occurs, this time to a very similar but different pathogen strain, strain #2. Strains #1 and #2 are both of the same disease, such as two strains of dengue virus. Dengue virus is well known to cause OAS.
3) During the second exposure, the immune system responds as if strain #1 is attacking, because it learned how to do this. The problem is that such an immune response is not effective against strain #2 (the antibodies are not a good fit). The result is a defective (and delayed) immune response and severe, long-lasting illness with strain #2.
OAS is why a first dengue fever illness is almost always mild, but a second exposure (to a different dengue strain) can be very severe and long lasting. A second dengue infection can be fatal due to OAS.
By receiving an influenza vaccine that is poorly matched to circulating strains, the immune system is improperly trained, and this can be worse than no “training” at all. The OAS phenomenon may explain the results of the Bridges study. OAS may also be a consequence of receiving the papilloma virus (HPV) vaccine, since not all strains of HPV are included in the vaccine.
You can read more about OAS here: https://en.wikipedia.org/wiki/Original_antigenic_sin
For many years there have been many anecdotal reports of people becoming sick from the influenza vaccine. These anecdotes are supported by the Cowling, Dierig and Bridges studies. Anecdotal reports can be important and meaningful. Sometimes the anecdotes are correct and years ahead of the science.
More Science Showing Immune Impairment By Influenza Vaccine
There are several other studies showing that the influenza vaccine causes immune impairment.
CD+ T-cells are specific types of immune cells targeted for specific types of pathogens. CD8+ T cells are necessary for proper immune function.
In the following study by Bodowes, it was found that annual influenza vaccination was associated with a reduction in the number of CD8+ T-cells. This will reduce immunity. Main results are shown below.
Above: Children receiving annual influenza vaccines had significantly (P<0.05) lower numbers of CD8+ T cells, an important type of immune cell responsible for heterosubtypic immunity (HI).
Specifically, lower numbers of CD8+T cells will reduce “heterosubtypic immunity” (HI). “Heterosubtypic immunity” refers to immunity to a range of pathogens that are different from pathogens the immune system has already been exposed to. In other words, HI is when an exposure to one type of pathogen creates immunity to different but similar pathogens. HI is important for providing broad-spectrum immunity against viruses like influenza that change constantly. Bodowes states:
“…long-term annual vaccination using inactivated vaccines may hamper the induction of cross-reactive CD8+T cell responses by natural infections and thus may affect the induction of heterosubtypic immunity. This may render young children who have not previously been infected with an influenza virus more susceptible to infection with a pandemic influenza virus of a novel subtype.” (emphasis added)
HI is provided by natural infections, but not the influenza vaccine. Influenza vaccines reduce HI, apparently by reducing the number of CD8+ T cells.
The influenza vaccine reduces HI in those who receive the vaccine annually, and therefore can be expected to cause increased illness by pathogens different from the strains in the vaccine.
Full paper: Annual Vaccination against Influenza Virus Hampers Development of Virus-Specific CD8 T Cell Immunity in Children
Further, Bodowes published 2 papers on this same effect in controlled experiments in ferrets and mice. He found that natural infection provided HI, but influenza vaccination reduced HI. Influenza vaccination also dramatically increased the severity of a later acquired influenza infection. In the study on mice, Bodowes states:
“Here we show in a mouse model that the induction of protective heterosubtypic immunity by infection with a human A/ H3N2 influenza virus is prevented by effective vaccination against the A/H3N2 strain. Consequently, vaccinated mice were no longer protected against a lethal infection with an avian A/H5N1 influenza virus. As a result H3N2-vaccinated mice continued to lose body weight after A/H5N1 infection, had 100-fold higher lung virus titers on day 7 post infection and more severe histopathological changes than mice that were not protected by vaccination against A/H3N2 influenza.” (emphasis added)
“Prior infection with seasonal influenza viruses, which generally results in a self-limiting upper respiratory tract infection, may afford at least partial protection against potentially pandemic heterosubtypic influenza virus strains.”
“…it [vaccination] may interfere with the induction of heterosubtypic immunity against potentially pandemic strains of a novel subtype, e.g. H5N1, by creating an immunological ‘‘blind spot.’’
And in the study on ferrets, Bodowes states:
“The vaccinated ferrets suffered more from the subsequent infection with the highly pathogenic H5N1 influenza virus A/Ind/5/05 than did their unvaccinated counterparts. These findings in the ferret model are in concordance with those we obtained recently with the mouse model.”
Full paper: Vaccination against Human Influenza A:H3N2 Virus Prevents the Induction of Heterosubtypic Immunity against Lethal Infection with Avian Influenza A:H5N1 Virus
Ferrets were used because they are the best animal for studying influenza. Ferrets respond to influenza infection and influenza vaccines very similarly to humans.
Immune suppression from the influenza vaccine has been observed in human studies, and is supported by findings of reduced CD8+ T cells in vaccinated humans. And these effects have been confirmed in multiple animal studies, using two different types of animals.
It is clear that immune suppression from the influenza vaccine is a real effect.
For many years natural medicine advocates have warned that vaccines impair immune function. Natural medicine advocates have also claimed that natural infections provide immunity benefits that are not provided by vaccines. The above scientific results prove natural medicine advocates correct.
In view of the above results, the influenza vaccine likely causes more problems than it prevents. The benefits provided by a wall-matched vaccine are likely cancelled by the harm caused by a poorly matched vaccine. And of course this does not include the risk of severe neurological damage (e.g. Guillan-Barre syndrome, demyelinating disease), or the risk of autoimmune disease.
A much more effective, safer and scientifically-supported approach to protecting yourself from infectious diseases and influenza is to use vitamins A, D, and C, among others.
For further reading, see these two papers by Dr Peter Doshi of Johns Hopkins U.:
Influenza Vaccines Time for a Rethink, JAMA, 2013
Influenza: Marketing vaccine by Marketing Disease, British Medical Journal, 2013
Papers in this post:
- Vaccination against Human Influenza A:H3N2 Virus Prevents the Induction of Heterosubtypic Immunity against Lethal Infection with Avian Influenza A:H5N1 Virus
- Bodewes, Rogier. Kreijtz, Joost H C M. Baas, Chantal. Geelhoed-Mieras, Martina M. de Mutsert, Gerrie. van Amerongen, Geert. van den Brand, Judith M A. Fouchier, Ron A M. Osterhaus, Albert D M E. Rimmelzwaan, Guus F. PloS one 2009; 4(5):e5538
- Annual Vaccination against Influenza Virus Hampers Development of Virus-Specific CD8 T Cell Immunity in Children
- Bodewes, Rogier. Fraaij, Pieter L A. Geelhoed-Mieras, Martina M. van Baalen, Carel A. Tiddens, Harm A W M. van Rossum, Annemarie M C. van der Klis, Fiona R. Fouchier, Ron A M. Osterhaus, Albert D M E. Rimmelzwaan, Guus F. Journal of virology 2011; 85(22):11995-2000
- Increased risk of noninfluenza respiratory virus infections associated with receipt of inactivated influenza vaccine
- Cowling, Benjamin J. Fang, Vicky J. Nishiura, Hiroshi. Chan, Kwok-Hung. Ng, Sophia. Ip, Dennis K M. Chiu, Susan S. Leung, Gabriel M. Peiris, J S Malik. Clinical infectious diseases : an official publication of the Infectious Diseases Society of America 2012; 54(12):1778-83
- Influenza Vaccines Time for a Rethink
- Doshi, Peter. JAMA internal medicine 2013; 173(11):1014-6