Congenital cytomegalovirus (cCMV) infection is the most common infectious cause of birth defects and the leading non-genetic cause of pediatric sensorineural hearing loss (SNHL) and single-sided deafness in the United States.^1–4 Although the majority of cCMV infections are asymptomatic during the neonatal period, both asymptomatic and symptomatic infants are at risk of long-term sequelae, including SNHL, vision impairment, and neurodevelopmental disabilities. Few U.S. studies have assessed cCMV prevalence based on newborn screening. In the United States, prior cCMV prevalence estimates were based on one multi-site screening study conducted between 2007 and 2012.⁵ Since the publication of those findings, it has been assumed that cCMV infection occurs in approximately 4.5 per 1,000 live births, with considerable variation observed by maternal race/ethnicity and age.^5–8 However, these estimates were not adjusted for sociodemographic distributions of annual live births in the United States (U.S.) and important differences may be obscured.

Despite the potential severity of disease, cCMV is not nationally notifiable in the U.S. and only 12 states currently conduct surveillance for infection and related conditions.^9,10 This has contributed to a limited understanding of the state-specific disease burden. As states consider implementing screening or surveillance for cCMV, the availability of more accurate and representative estimates of expected prevalence may inform resource allocation, priority setting, and policy- and decision-making for cCMV. Furthermore, CMV vaccines may come to market within the next five years and addressing this knowledge gap will be especially important to prepare for vaccine effectiveness and impact evaluation. Using cCMV prevalence reported from U.S. newborn cCMV screening studies, this study sought to estimate the expected national and state-specific live birth prevalence of cCMV infection from 2018–2022, adjusted for maternal race and ethnicity and maternal age group.

This study provides an updated estimate for national cCMV infection prevalence in the United States, as well as individual, state-specific estimates for 2018–2022. Our estimates account for the demographic makeup (maternal race and ethnicity or age group) of the United States and individual states, including data from multiple studies that represent 10 geographical and economically diverse states. At the state level, these estimates may help with resource allocation, program planning, and policy- and decision-making with regards to cCMV screening, surveillance, and care for children diagnosed with cCMV. Nationally, these estimates may be useful to inform future vaccine policy, including efforts centered on improving health equity. We estimated 15,800–17,600 infants born with cCMV infection annually, approximately 2,830–3,170 of whom might present with cCMV disease at birth or related sequelae. Some studies have estimated the number of annual infections to be 40,000 with approximately 8,000 children with permanent disabilities.^17,18 However, more recent U.S. screening studies demonstrate lower prevalence, suggesting that this number was likely an overestimate.^5,6,8,19 Placing the burden of cCMV in context with other congenital or perinatal causes of neurodevelopmental disabilities and impairments in children is helpful to increase public awareness. It has long been stated that more children are affected by serious cCMV-related disabilities than by several better-known childhood conditions, including Down Syndrome and fetal alcohol syndrome (FAS). However, those assertions have been limited by the methods of case ascertainment for each condition (passive versus active surveillance, universal screening, analysis of hospitalization data), the case definitions or spectrum of disease captured (all disease versus severe disease), and the representativeness of the populations studied. For example, infants born with Down Syndrome have distinctive physical characteristics that prompt immediate genetic testing. With a prevalence of 1.3 per 1,000,²⁰ Down Syndrome may lead to more cases of neurodevelopmental disabilities in U.S. children as compared to cCMV disease (assuming 18% of infections result in disease). In contrast, FAS encompasses the most severely affected cases of fetal alcohol spectrum disorder (FASD). Though FASD may not be readily identified, its prevalence ranges from 11.3–50.0 per 1,000, much higher than that of cCMV infection.^21,22 It is therefore fair to describe cCMV infection as the leading infectious cause of neurodevelopmental disabilities and impairments in U.S. children. Future state-based surveillance will be helpful in guiding education and prevention strategies, including potential opportunities for integrated approaches to birth defect prevention.

Although most states had an annual maternal race and ethnicity-adjusted prevalence similar to the national estimate (4.6 per 1,000), there were notable regional patterns. Overall, maternal race and ethnicity-adjusted burden was disproportionately higher in southeastern parts of the country, including DC, which consistently had the highest estimated race and ethnicity-adjusted prevalence. Compared to the rest of the country, these states have large minority populations – the average proportion of infants born to non-Hispanic Black mothers from 2018–2022 was 44.3% in DC, 42.1% in Mississippi, and 36.9% in Louisiana compared to 5.0% in California (the state with the lowest prevalence) and 14.5% across the US. Disparities in cCMV burden exist among racial and ethnic minorities, as reported in all included studies in this analysis: infants born to non-Hispanic Black and non-Hispanic AI/AN mothers consistently had higher cCMV prevalence estimates compared to those born to non-Hispanic White and non-Hispanic Asian mothers (prevalence observed among Hispanic infants was inconsistent).^5–8 Health disparities observed among racial and ethnic groups are often related to socioeconomic inequities, many of which – such as frequent and/or prolonged contact with children under six years and crowded living arrangements – directly increase the risk of cCMV infection.²³ Reasons for these disparities are likely multifactorial and should not be attributed to differential biological susceptibility. Understanding the effects of race and ethnicity and socioeconomic resources (or lack of) within the context of systemic racism and impacts on health outcomes will be important for guiding future research and developing informed prevention efforts.

In addition to disparities in disease burden, this study also demonstrated persistent inequities in the availability of public health data for racial and ethnic minorities, especially Indigenous populations such as AI/AN and HOPI persons. We were not able to estimate prevalence for infants born to HOPI mothers, and data for AI/AN infants was only available from two studies (which had to be obtained from study authors, as published data were collapsed into an “other” category – a common, albeit problematic, practice in public health research). Although pooled prevalence results in the current analysis showed increased prevalence among AI/AN infants compared to all other racial and ethnic groups, accurate enumeration of disparities among Indigenous infants is challenging because data from newborn cCMV screenings with a large proportion of Indigenous infants are lacking. One previous study reported that cCMV-coded diagnosis among AI/AN infants was significantly higher in Alaska compared to the Southern Plains region, which is consistent with our findings, but reasons for this geographic difference are unknown.²⁴ Future studies that include better (i.e., larger) representation of these populations will be helpful in assessing cCMV disease burden and potential health disparities.

Younger maternal age has also been shown to be associated with greater risk of cCMV. One study showed that the youngest mothers (<20 years) were the most likely to deliver an infant with cCMV (prevalence odds ratio [POR]: 4.5 per 1,000) compared to mothers aged 20–25 years (POR: 2.2 per 1,000) or mothers aged 25–29 years (POR: 1.1 per 1,000).²⁵ A more recent report from Michigan showed that approximately half of cCMV cases from 2004–2011 were born to mothers aged ≤24 years and that prevalence was highest among those born to mothers aged ≤20 years (14 per 1,000 live births).²⁶ The studies included in the pooled prevalence estimates all showed higher prevalence among infants born to the youngest mothers. However, when the pooled prevalence estimates were applied to the state-specific populations adjusted for maternal age group, prevalence was more homogenous across states compared to prevalence adjusted for race/ethnicity. Although “hot spots” were identified, differences were not as pronounced. It is likely that maternal age does not vary substantially across states and that finer scale (e.g., county-level) data are needed to identify prevalence differences by maternal age. It has been shown that maternal age at first birth is strongly driven by urbanicity²⁷ and that a larger proportion of teenage pregnancies occur in lower-income, rural settings.^28,29 Additionally, because of limits to reported and available data, we collapsed the youngest mothers (<20 years) with mothers aged 20–24 years. Therefore, synergistic effects of teenage pregnancies and living in a rural area would have been masked, limiting the potential diversity of cCMV prevalence estimated in the current analysis.

Currently, infection with cCMV is not a nationally notifiable condition and only 12 states conduct cCMV surveillance.^9,10 In 2023, the Council for State and Territorial Epidemiologists (CSTE) approved standardized case definitions for cCMV infection and disease, which went into effect January 2024. In August 2022, CDC began collaborating with seven jurisdictions (Illinois, Iowa, Minnesota, New Jersey, New York, Utah, and Virginia) to pilot cCMV surveillance through the Surveillance for Emerging Threats to Pregnant People and Infants Network (SET-NET).¹⁰ The ongoing pilot will inform methods for case ascertainment, as well as the feasibility of implementing the CSTE case definitions and conducting longitudinal data collection. These efforts are critical to advancing cCMV surveillance and findings from the SET-NET cCMV pilot are expected to guide the standardization of cCMV surveillance across the country. For example, Utah has conducted cCMV surveillance since 2013 and serves as a model for other jurisdictions; data collected over the past decade provides valuable insight into the impact of newborn screening protocols (hearing-targeted and later, expanded high-risk) on case ascertainment.³⁰ Surveillance in Minnesota, the first state in the U.S. to implement universal newborn cCMV screening, will provide additional information on equity in access to infant follow-up services.³¹ Across all piloting jurisdictions, a critical facilitator for collecting data has been the inclusion of cCMV in state-specific reporting rules. In the future, inclusion of cCMV in the list of nationally notifiable conditions by the CSTE may facilitate case ascertainment and data collection across the U.S. and improve representativeness. This analysis was subject to limitations. We relied on data from newborn cCMV screening studies conducted over the last two decades, representing over 15 years of data collection. However, reported prevalence was consistent between studies, suggesting cCMV prevalence is stable. Additionally, the categorization of persons into race and ethnicity and age groups differed between the included studies, limiting the data we were able to include in the pooled prevalence estimates for some groups. Furthermore, some studies combined certain groups (e.g., non-Hispanic AI/AN and Asian into “other”), which may misrepresent minority populations. We were not able to stratify by maternal race and ethnicity and age group, both of which are known to influence the risk of cCMV and likely act synergistically. Discrepancies in maternal age were noted between reported study enrollment and state-specific makeups observed in CDC WONDER, which may have impacted the age-adjusted analysis. Finally, prevalence of cCMV in 2020 might have been impacted by the COVID-19 pandemic, as reported in recent studies.^8,32

This analysis provides updated national cCMV prevalence estimates and new state-specific, adjusted prevalence estimates over a five-year period. It demonstrated notable regional patterns in cCMV prevalence adjusted for maternal race and ethnicity, which reflect underlying health disparities. These findings highlight the need to collect more robust sociodemographic data to be able to identify groups at the greatest risk of infection, improve our understanding of specific risk factors, and develop tailored, culturally relevant education and prevention strategies. Furthermore, these data may be useful for individual jurisdictions with resource allocation, program planning, and policy- and decision-making with regards to cCMV screening, surveillance, and care for children diagnosed with cCMV.