Vax-Unvax by Robert F. Kennedy Jr

Author:Robert F. Kennedy Jr.
Language: eng
Format: epub
ISBN: 9781510766976
Publisher: Skyhorse
Published: 2023-08-16T00:00:00+00:00

Association between Influenza Infection and Vaccination during Pregnancy and Risk of Autism Spectrum Disorder

Figure 11.1 Hazard ratios for ASD incidence in the offspring of women receiving the flu vaccine in the first trimester and any trimester of pregnancy compared to the offspring of unvaccinated pregnant women (Zerbo et al. 2017).

Figure 11.1 shows results from the paper “Association between Influenza Infection and Vaccination during Pregnancy and Risk of Autism Spectrum Disorder,” published in the journal JAMA Pediatrics in 2017.8 The lead investigator, Dr. Ousseny Zerbo, is from the Division of Research at Kaiser Permanente in Oakland, California. Dr. Zerbo and his coauthors investigated the relationship between influenza infection and vaccination during pregnancy and autism spectrum disorder. They found that first-trimester influenza vaccination is associated with an increased risk of ASD, with a hazard ratio of 1.20 and a 95% confidence interval between 1.04 and 1.39.9 Investigators also observed that flu vaccination in any trimester was associated with autism spectrum disorder with a hazard ratio of 1.10 and a 95% confidence interval between 1.00 and 1.21.10

After obtaining these statistically significant results, the investigators applied the Bonferroni correction, which is sometimes used when multiple statistical tests are completed on the same data sample. When statisticians make many comparisons within a single study, the “false positive” rate, or the probability of finding associations that are not valid, can increase when the comparisons made are independent or unrelated.11 The Bonferroni correction adjusts for this. Erroneously using the correction, the authors raised the p-value from 0.01 to 0.1, which is above the threshold for the statistical significance (p-value less than 0.05). The researchers then claimed insignificant statistical relationships. This was countered by Dr. Alberto Donzelli12 and Dr. Brian Hooker,13 who wrote two separate letters to the editor of JAMA Pediatrics in 2017. They showed it was inappropriate for Dr. Zerbo14 to apply any correction for multiple testing because all associations made in the study were highly interdependent and not independent, which is required for correction. For example, Dr. Zerbo’s results for each trimester roll up into a total result, showing interdependency instead of independence. Therefore, any correction for a “false positive rate” would not apply.15 Thus, the actual p-value for the analysis was 0.01, which is statistically significant.16

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.