This ecological observational study examined relations between areas of environmental risk and hotspots of adverse perinatal health indicators in NYC. We found notable overlap in hotspots and areas of environmental risk in this urban setting: a majority of environmental risk areas were a hotspot for at least one adverse perinatal health indicator. We also found that environmental risk area designation was associated with hotspots of preterm birth, adolescent pregnancy, and multiple adverse indicators. These findings reflect our hypothesis, highlighting the interconnected nature of environmental threats and perinatal health. With worsening environmental risks [22], our results support the creation of policies directed at identifying high-risk environmental areas, given their demonstrated association with significant perinatal health indicators. More than half of all environmental risk areas were hotspots for at least one indicator, and, among these, most were hotspots for multiple indicators as seen in parts of the Bronx, Brooklyn, and northern Manhattan. These same areas are known to have higher relative economic inequalities and health disparities [23, 24]. The Poisson regression shows that identified areas of environmental risk were associated with 3.36 times the number of adverse perinatal health indicator hotspots. Stratified analysis based on component scores shows that these associations tend to be more driven by population characteristics and health vulnerabilities which may have collinearity with the health indicators considered in this study. Environmental and climate burden component score was positively associated with hotspots of adolescent pregnancy, providing insight into how neighborhood disinvestment may have wide-ranging effects on population health. These insights can inform future strategies to address high rates of adverse perinatal health indicators in communities related to structural disparities of environmental risk, which are more concentrated in urban cities like NYC.

We identified significant overlap and associations between hotspots of preterm births and environmental risk area designations. A majority of preterm birth hotspots were environmental risk areas, totaling 165 census tracts out of 259 classified as hotspots (63.7%). Previously conducted geographic studies have found greater frequency of preterm birth in communities that have been historically redlined, experience greater poverty, and have higher concentrations of Black and Hispanic communities [7, 25]. Individual environmental hazards (e.g., extreme heat, water contaminants, and air pollution) have been associated with an increased risk of preterm birth [4, 26], and our findings support that the cumulative influence of multiple environmental threats may be associated with higher rates of preterm birth on the population level.

We also found that 428 of the 457 adolescent pregnancy hotspot census tracts were areas of environmental risk (93.7%). Unlike preterm birth, there is no direct pathway between environmental exposures and rates of adolescent pregnancy elucidated in the literature. Thus, as hypothesized, while environmental risk areas were associated with adolescent pregnancy hotspots, the distribution of adolescent pregnancy hotspots differed from the hotspots of preterm birth. However, this was the only perinatal health indicator that showed an association with census tract environmental and climate burden component scores. Adolescent pregnancies in New York City show disparities by race/ethnicity and neighborhood disinvestment, aligned with prior literature demonstrating associations between adolescent pregnancy and higher social risk [27]. The overlaps and associations we detected may stem from the greater prevalence of adolescent pregnancies in communities with persistent poverty, where logistical and societal barriers to reproductive healthcare access are common [28]. Even with the use of socioeconomic composite indices, ecological studies have provided evidence that areas experiencing disadvantage face higher rates of adolescent pregnancy [29, 30]. These same communities often face disproportionate exposure to environmental threats, which our findings suggest may be associated with community risk for adverse perinatal outcomes.

We detected overlaps between areas of environmental risk and hotspots of pre-pregnancy obesity. Pre-pregnancy obesity—associated with preterm birth, congenital defects in offspring, and childhood obesity risk [15]—is the most prevalent of our three indicators. According to data from 2021, 31.3% of pregnant residents in New York State were classified as obese prior to pregnancy [31]. Among the 596 census tracts within hotspots of pre-pregnancy obesity, 401 of those census tracts were environmental risk areas (67.3%), aligning with prior evidence finding that obesity in the general population is associated with neighborhood food environments, walkability, and other socioeconomic factors [32, 33]. Notably, one study highlighted that this association was confounded by the percentage of Black residents in the neighborhood [32]. Specifically for pre-pregnancy obesity, an ecological study on metropolitan cities in the USA similarly found that an index of 11 indicators of area deprivation, among other city-level factors, was a significant risk factor for pre-pregnancy obesity prevalence [34]. Communities with higher exposure to environmental hazards may face increased risks of obesity additionally due to increased contact with endocrine-disrupting chemicals, which interfere with hormonal systems that regulate metabolism, fat storage, and energy balance [35].

The strengths of this study include the high-quality data from the Bureau of Vital Statistics, a large sample size of births, and the use of the comprehensive environmental risk area designation. Previous studies have examined the relation between various environmental hazards and climate-related events, but few have looked at the use of a composite of these exposures with multiple perinatal health indicators. This is crucial since environmental hazards are often simultaneously present. A similar metric, the former federal Climate and Economic Justice Screening Tool, assesses comparable burdens and has served as guidance for state, territory, and tribal governments in their approaches to environmental justice screening, last updated to have 35 factors in August 2024 [9]. The National Academies of Sciences, Engineering, and Medicine has called for validated geospatial and cumulative tools for fair environmental health, and our study provides some validity for this policy designation system in the context of perinatal health burden [9, 36]. The ecological nature of this study and scope of the data used are strengths for the surveillance of pregnant populations in NYC.

Due to the ecological nature of our study, however, our findings neither determine causality nor draw individual-level conclusions (i.e., assessing one’s individual risk of adverse indicators from living in an environmental risk area). Some initially considered environmental risk indicators that would have provided valuable information on community burden were dropped due to lack of usable data [16]. Additionally, six of the 45 environmental indicators were collected after perinatal health outcome assessment in 2020. These variables were not expected to have changed substantially as they were based on pre-existing neighborhood infrastructure. Evidence has shown that the validity of maternal weight data reported on vital records may be poor due to recall bias [37]. This health indicator was still chosen due to the previously identified relationship between environmental factors and body mass index. Lastly, the inclusion of data before 2020 may not be reflective of the current environment following the COVID-19 pandemic. This temporal adjustment calls for a reevaluation of this association and across other health domains once updated data are released.

As clinical and community-based shareholders develop strategies to connect individuals at risk for adverse perinatal outcomes to services, our findings contribute to evidence that integrating knowledge of residence in environmental risk areas could help guide efforts to mitigate exposures and deliver more targeted care [38]. The New York 1115 Medicaid Waiver amendment, running from 2024 to 2027, aims to improve health quality and outcomes in areas with historical health discrepancies and disengagement [39]. Environmental risk area classifications could help inform this goal by enabling clinicians to integrate geospatial data into screening efforts.

Structural conditions leading to heightened environmental threats have resulted in communities being at a greater risk of adverse exposures and subsequent disease burden [40, 41]. For leaders and policymakers, principles of environmental health and fairness should be considered when approaching health promotion, as infrastructural and societal barriers perpetuate environmental risk and burden. As expected, given that race and ethnicity frequently overlaps with social and economic disadvantage, we detected differences by environmental risk area designation in the proportion of pregnant residents who identified as Hispanic or Black. Previous studies have repeatedly identified disparities in environmental risk for predominantly Black and Hispanic neighborhoods in NYC [42, 43]. The use of environmental risk areas offers a more targeted and actionable basis for resource allocation and intervention and shifts the focus from individual characteristics and behaviors as drivers of perinatal risk to environmental conditions shaped by modifiable policies. This study represents a framework by which policy-derived designations can be tested for their relationships with health indicators to foster transparency, integrity, and social engagement.