Lifetime effects of the 1918–1919 influenza pandemic

The 1918 Influenza pandemic (Influenza A, H1N1 subtype) had multiple health consequences for the cohort that was in utero during the pandemic.Reference Almond ^{1 ,} Reference Mazumder, Almond, Park, Crimmins and Finch ² The U.S. cohort in utero during the peak pandemic period, experienced more depression after age 50, diabetes and ischemic heart disease after age 60, and mortality after age 65, with some variation in estimated effects.Reference Mazumder, Almond, Park, Crimmins and Finch ^{2 –} Reference Schnittker ⁵ The Taiwanese cohort exposed in utero in 1919 had more renal disease, circulatory and respiratory morbidities, and diabetes after age 60.Reference Lin and Liu ⁶ The corresponding Swedish cohort had higher morbidity at ages 54–87, indicated by excess hospitalization and male mortality from heart disease and cancer.Reference Helgertz and Bengtsson ⁷ The effect on late life health varied by trimester of development at the pandemic peak. Elevated heart disease in the U.S. was associated with second trimester exposure at the peak for the 1918–1919 cohort; while diabetes was associated with third trimester exposure at the peak.

No data on birth weight are available for the 1918 Influenza cohort; however, there are many indications that early life physical health and neurodevelopment were directly affected either by in utero exposure or indirectly by its multifarious stresses. Men in the 1918–1919 U.S. flu cohort were shorter at WWII enlistment than flanking cohorts,Reference Mazumder, Almond, Park, Crimmins and Finch ² while children and adolescents in Taiwan were shorter than surrounding cohorts.Reference Lin and Liu ⁶ U.S. women exposed in utero married men with less education.Reference Fletcher ³ Educational deficits in the U.S. and Taiwan for these cohorts could be related to early cognitive deficits.Reference Almond ^{1 ,} Reference Lin and Liu ⁶ Lower socioeconomic status (SES) as adults for those in utero during the 1918 epidemic was also confirmed in the parents of the Wisconsin Longitudinal Study participants,Reference Cook, Fletcher and Forgues ⁸ but was not found in Sweden.Reference Helgertz and Bengtsson ⁷

These historical associations for the cohort in utero in 1918–1919 are based solely on birth dates and cannot identify individuals whose mothers actually experienced infections. Thus, the analysis is restricted to population data on timing of infections and birth dates related to outcomes many decades after the pandemic. Thus we cannot readily distinguish direct effects on long-term health and well-being due to exposure to the infection itself versus the stress effects due to exposure to the pandemic environment. These socio-environmental stressors include potentially increased levels of uncertainty, undernutrition, under- and unemployment, poverty, and loss of loved ones. Despite this, the findings from the 1918 flu pandemic of lasting consequences for those exposed prenatally warrants study of similar programming effects from the current pandemic. Unlike the earlier pandemic, the current pandemic provides opportunity to study individuals with known in utero exposure to SARS-CoV-2 throughout their life based on known maternal infection and socio-environmental stresses. We can add further critical comparisons to individuals exposed in utero to the pandemic environment but with no maternal infection, which potentially allows disentanglement of the effects of exposure to infection versus pandemic environment.

Long-term consequences of maternal and fetal exposure to viral infections

Maternal viral infections can affect the fetus or neonate through multiple pathways. Maternal infections can directly infect the fetus as well as the placenta.Reference Raciot and Mor ⁹ Viral species differ markedly in their impact on pregnant women and fetuses. For instance, HIV and hepatitis B virus are vertically transmitted in up to 40% of cases without the successful interventions that have almost eliminated direct transmission in the U.S. Fortunately, maternal transmission is rare for the coronaviruses MERS-CoV (2012) and SARS-CoV (2002)Reference Rodríguez-Blanco, Vegara-Lopez, Aleo-Giner and Tuells ¹⁰ and the 2009 H1N1 influenza virus.Reference Mosby, Rasmussen and Jamieson ¹¹ And, no severe birth defects have been associated with prior coronaviruses, MERS-CoV (2012) and SARS-CoV (2002). In general, pregnancy increases virus-associated morbidity and mortality, as shown for influenza H1N1 in 2009 and SARS-CoV.Reference Alfaraj, Al-Tawfiq and Memish ^{12 ,} Reference Giles, Krishnaswamy, Macartney and Cheng ¹³ These infections also caused more preterm delivery and low birth weight.Reference Giles, Krishnaswamy, Macartney and Cheng ¹³

Virus-related maternal inflammatory processes can impact the placenta and disturb in utero metabolism to cause intrauterine growth restriction (IUGR).Reference Maldonado ¹⁴ Long-term consequences include impaired glucose and insulin metabolism and hyperactivity of the hypothalamic–pituitary–adrenal (HPA) axis.Reference Kesavan and Devaskar ¹⁵ Historically, Barker and colleagues showed that IUGR and low birth weight which could be related to maternal infection or poor nutrition are associated with later life cardiovascular disease and obesity.Reference Barker, Osmond, Kajantie and Eriksson ¹⁶ In the Helsinki Birth Cohort (1934–1944), which suffered a suboptimal nutritional environment, smaller placentas predicted larger childhood body mass index (BMI) and increased adult cardiovascular disease.Reference Thornburg and Marshall ¹⁷ Maternal immune response to viruses and bacterial coinfections can impact fetal development with multi-generational behavioral and cognitive consequences.Reference Pollak and Weber-Stadlbauer ¹⁸

Effects of COVID-19 on pregnant women and neonates

We summarize the main points of the COVID-19 literature published thus far, bearing in mind that these findings are preliminary as they are based on small numbers, specific locations, and, often, more severely ill patients.

Ongoing/future studies

The COVID-19 cohort provides a natural experiment to examine the effects of maternal viral infection while infants were in utero. Such effects may span the lifecycle of these babies; during this time it may be possible to intervene to prevent later life problems. Cohort studies include the UCSF PRIORITY: Pregnancy Coronavirus Outcomes Registry and the American Academy of Pediatrics Society of Neonatal and Perinatal Medicine (AAP SoNPM) – National Perinatal COVID-19 (NPC-19) Registry, a collaboration of the AAP SoNPM with the Vermont Oxford Network (VON). Neither of these registries are currently supported for the long-term follow-up which is warranted by important scientific and clinical reasons.

We suggest that to capture the consequences of viral exposure in utero for childhood development and adult health COVID-19 birth cohort studies consider immediate collection of data from the mother, fetus, neonate, and placenta. These initial data should be followed by analysis of child growth and development and lifelong study of health, behavioral patterns, and cognitive functioning. Table 1 suggests multiple indicators for immediate measurement and measurement across the lifespan. Initial indications of disease presence, exposure, and severity are important including inflammatory cytokines in the mother and infant. Clinical data on the size and development of the infant both before and at birth and the placenta at birth should be routinely collected. Because relatively simple anthropometric measures routinely taken at birth serve only as a very crude proxy measure of the prenatal environment, we encourage studies using longitudinal ultrasound measurements of fetal development. Because hypocortisolemia was common in SARS-CoV-1, HPA axis activity should be studied in severe COVID-19 mothers (ChiCTR20000301150).Reference Pal ³⁷ Maternal genetics may also influence vulnerability to infection and impact on the fetus, shown for adult carriers of ApoE alleles.Reference Finch and Kulminski ³⁸ Molecular studies could include the placental transcriptome for genetic variants of gene expression (eQTLs) that predict childhood obesity and BMI.Reference Peng, Deyssenroth and Di Narzo ³⁹ Epigenomic methylation of DNA and histones should be monitored across the lifespan of the infant. DNA methylation measures represented as epigenetic clocks that represent “aging” can be estimated in pediatric and then adult years.Reference McEwen, O’Donnell and McGill ⁴⁰ Socioeconomic status and educational status of the mother and child should also be assessed, as causes of adverse outcomes can be social as well as biological. Additionally, COVID-19 infection appears to be more common and severe in racial and ethnic minorities in the U.S.Reference Kopel, Perisetti, Roghani, Aziz, Gajendran and Goyal ⁴¹ Moreover, preexisting conditions increase the severity of COVID-19 infection, which may have implications for transmission and neonatal outcomes. Our recommendations are oriented toward the United States as the situation may differ in low- and middle-income settings.

Table 1.

Domains for evaluating effects of in utero exposure to COVID-19

The child should be followed for indicators of growth, motor, and neurological development milestones. Throughout life, the standard blood-based indicators should be collected. Composite measures derived from clinical chemistry and physical exams can be used starting in childhood to categorize the “Pace of Aging.”Reference Belsky, Caspi and Houts ⁴² Follow-up should include information on development of both pediatric (e.g., leukemia, asthma) and adult diseases (e.g., heart disease, cancer, stroke).

The COVID-19 pandemic also provides an opportunity to study the effects of in utero exposure to the pandemic environment and the effects of different regional and national policies to address viral spread. The COVID-19 pandemic has increased levels of stress, unemployment, food insecurity, and domestic violence, and diminished or disrupted prenatal care. These factors may have programming effects and/or direct effects on neonatal health. The pandemic-related lockdown policies may benefit some individuals; pregnant women staying home who may rest more, and have decreased exposure to other infections, and women with job stability, may benefit from less stress from work and commuting time. These benefits are evidenced by reports of decreased rates of preterm birth for the general population during the pandemic.

Studies would include mother and child dyads and a selection of mothers with symptomatic and nonsymptomatic COVID-19 diagnosis, as well as a group of mothers and babies with no in utero exposure. We suggest the desirability of enrolling an additional child born either shortly before or shortly after the COVID-19 pandemic to the same mother. These study groups would allow comparison between infants who were exposed to maternal infection and those who were prenatally exposed to the pandemic without having experienced maternal infection. These infants can also be compared to those who were born before or conceived after the pandemic, providing further insight on the effects of the pandemic environment. The inclusion of information on social and economic stresses will allow comparisons between countries taking different measures to reduce spread of the virus, These types of comparisons may give us further insights beyond the effects of COVID, such as socioeconomic and social policies that may decrease risk of preterm birth, which has eluded the maternal-infant health community for decades. This should be balanced by further investigating reports of generally increased stillbirth and potential contributors.