By Hal B. Jenson, MD, FAAP

Professor of Pediatric and Adolescent Medicine, and Dean, Western Michigan University Homer Stryker M.D. School of Medicine, Kalamazoo, MI

Dr. Jenson reports no financial relationships relevant to this field of study.

SYNOPSIS: A very large cohort study over 11 years failed to find an association between autism risk and maternal influenza infection or influenza vaccination during pregnancy. A low risk of autism was associated on initial analysis with first-trimester vaccination, but adjusting statistically for the multiplicity of hypotheses tested in the study showed that this association could be due to chance (P = 0.10).

SOURCE: Zerbo O, Qian Y, Yoshida C, et al. Association between influenza infection and vaccination during pregnancy and risk of autism spectrum disorder. JAMA Pediatr. Published online Nov. 28, 2016.

A cohort study was performed of 196,929 singleton children born in Northern California from Jan. 1, 2000, to Dec. 31, 2010, of at least 24 weeks’ gestational age and who were followed for at least the first two years of life. Data were collected on maternal influenza infection and vaccination from the date of conception to the date of delivery. Maternal influenza infection was defined by ICD-9 codes in medical records or positive influenza laboratory test results. Diagnosis of childhood autism spectrum disorder was defined by ICD-9 codes in medical records on at least two visits from birth through June 30, 2015.

Influenza was diagnosed during pregnancy in 1,400 (0.7%) mothers. There were 45,231 (23%) mothers who received influenza vaccination during pregnancy, from a low of 6% in 2002 to a high of 58% in 2010. (Beginning in 2004, influenza vaccination was recommended by the Advisory Committee on Immunization Practices and the American College of Obstetricians and Gynecologists for all women who are or would be pregnant during the influenza season, regardless of trimester of pregnancy.) Follow-up of children was from two to 15 years (median, 8.3 years).

There were 3,101 (1.6%) children diagnosed with autism spectrum disorder. Covariate analysis showed no association of autism with maternal influenza infection (adjusted hazard ratio [HR], 1.04; 95% confidence interval [CI], 0.68-1.58) or maternal influenza vaccination (adjusted HR, 1.10; 95% CI, 1.00-1.21). Trimester-specific analysis showed that first-trimester influenza vaccination was associated with increased autism risk (adjusted HR, 1.20; 95% CI, 1.04-1.39) but that this association could result from chance (P = 0.1) if adjusted statistically for the multiplicity of hypotheses (n = 8) tested using the Bonferroni correction.

COMMENTARY

Many genetic and environmental factors likely contribute to the etiology of autism spectrum disorder, and family and social factors appear to further modulate the clinical manifestations. Recent epidemiological studies report an increased risk of autism with maternal infections and fever during pregnancy, although the risk with specific infections and timing during pregnancy is not clear. Previous smaller studies of maternal influenza infection and possible association with autism have shown mixed findings.

This most recent study found no association between autism risk and influenza infection during pregnancy or influenza vaccination during the second and third trimesters of pregnancy. The initial analysis found that influenza vaccination in the first trimester was associated with a slightly increased risk of autism. However, adjusting statistically for the multiplicity of hypotheses tested in the study showed that this association could be due to chance (P = 0.10).

It is imperative not to over interpret the findings of a statistical association of first-trimester influenza vaccination with autism risk, especially when further statistical analysis found that the association may be due to chance. For perspective, in 1998 a small case series by Wakefield et al in Lancet reported a speculative association of behavioral regression and pervasive developmental disorder in children with measles, mumps, and rubella (MMR) vaccine. That uncontrolled study later was deemed fraudulent because of scientific misrepresentation of the data, by reporting selective data as being consecutive data. There were also concerns about that study because the authors had financial conflicts of interest, and the study was conducted without appropriate ethical oversight. Ten of 12 authors subsequently retracted their interpretation of the data, and the journal subsequently completely retracted the report. Nevertheless, the time and financial costs to conduct the appropriate epidemiological studies that conclusively refuted the results were significant. More importantly, the report undermined consumer confidence in vaccines. Parents around the world did not vaccinate their children out of the unfounded fear of increased risk of autism, which exposed their children to the risks of measles and its complications and resulted in measles outbreaks in the United Kingdom, United States, and Canada.

Both young children and pregnant women are at increased risk of severe illness and complications from influenza infection. One part of the results in this most recent study, the increased risk of autism with first-trimester maternal influenza vaccination, remains a statistical association under one analysis but not under additional, more rigorous analysis. Regardless of the interpretation of the statistical analyses, the effect of influenza on pregnant mothers and their offspring, especially in the first year of life before those children themselves can receive influenza vaccination, is a reality that is seen in physician offices and hospitals every year. Infants are among the most susceptible to the complications of influenza, but are not candidates for influenza vaccination. Universal vaccination of healthy adults, especially in households with young infants, is an important step to minimize the risk of influenza in young children.

Further studies are needed to specifically address the possible autism risk of first-trimester influenza vaccination. In the meantime, we should continue our diligence to ensure that all individuals are appropriately vaccinated against influenza, including pregnant women, to protect both them and their offspring.