A new analysis of population surveys shows that exposure to a chemical commonly used to make plastics more flexible may have led to about 1.97 million premature births in 2018 alone, or more than 8 percent of the world’s total. Researchers estimate that the chemical was also linked to the deaths of 74,000 newborns.
The toxin, di-2-ethylhexylphthalate (DEHP), is part of a group of chemicals called phthalates, which appear in cosmetics, detergents, bug repellents, and other household products. Experts have found that these substances can break down into microscopic particles and enter the body through food, air and dust.
Led by NYU Langone Health researchers, the new study focuses on premature birth, which is a major risk factor for permanent learning and developmental issues and a leading cause of infant death, according to the World Health Organization.
According to the authors, the new analysis provides the first global estimate of preterm births associated with DEHP exposure and identifies which parts of the world are most affected. A report on the findings was published online March 31 in the journal eClinicalMedicine.
“By estimating how much phthalates may contribute to preterm birth worldwide, our findings suggest that reducing exposure, particularly in vulnerable areas, may help prevent early birth and the health problems that often occur.”
Sarah Hyman, MS, study lead author
Hyman, an associate research scientist at the NYU Grossman School of Medicine, said previous studies have linked DEHP exposure to cancer, heart disease and infertility, as well as many other health concerns. There is also a large body of research linking this chemical to premature birth.
According to the new work, DEHP exposure has contributed to 1.2 million years lived with disability, a measure of all the years people have lived or will live with diseases, injuries and other health problems due to being born prematurely.
Hyman said that while phthalates are in widespread use, some regions are projected to have a much larger share of health impacts than others, with the Middle East and South Asia representing 54 percent of the estimated disease burden from premature birth. These areas are home to a rapidly growing plastics industry and high levels of global plastic waste.
In Africa, where 26 percent of health problems associated with DEHP are due to preterm birth, the proportion of deaths is disproportionate to the overall prematurity incidence. The researchers said this reflects the high number of deaths due to premature birth in the region.
For the study, the research team estimated DEHP exposure in 200 countries and territories in 2018 by pulling data from large national surveys in the United States, Europe and Canada. They also used estimates from earlier investigations to fill in areas that did not have their own data.
The team then drew on earlier research that assessed how phthalate exposure could affect preterm birth and combined those findings with their global risk estimates. Finally, they combined this information with worldwide data on premature births and deaths to determine how much of these outcomes could be linked to DEHP.
Scientists repeated these steps for another phthalate, called disononyl phthalate (DiNP), a common replacement for DEHP. According to the results, DiNP may pose similar risks to DEHP, which has contributed to approximately 1.88 million premature births worldwide. The financial costs associated with the deaths of newborns range from millions to hundreds of billions of dollars for both phthalates.
“Our analysis makes it clear that regulating phthalates one at a time and swapping in poorly understood replacements is unlikely to solve the larger problem,” said the study’s senior author Leonardo Trasande, MD, MPP, the Jim G. Hendricks, MD, professor of pediatrics at the NYU Grossman School of Medicine. “We are playing a dangerous game of whack-a-mole with dangerous chemicals, and these findings highlight the urgent need for stronger, category-wide monitoring of plastic additives to avoid repeating the same mistakes.”
Dr. Trasande, who is also a professor in the Department of Population Health and director of the Department of Environmental Pediatrics and the Center for the Investigation of Environmental Hazards, cautions that the investigation was not designed to establish whether DEHP and DiNP directly or alone cause premature birth, nor did it take into account other types of phthalates.
Additionally, because there is some uncertainty in the data, researchers looked at a range of possible values rather than just a single estimate. This uncertainty range revealed that the actual impact of DEHP may be four times smaller or slightly larger than the main estimate. Even under the most conservative estimates, the results point to a substantial health burden, Hyman said.
Despite the limitations of such global modeling, this work lays an important foundation for future studies to confirm and refine these results and begins to fill a major gap in understanding the extent to which plastic chemicals affect preterm birth around the world, Hyman said.
Funding for the study was provided by National Institutes of Health grant P2CES033423 and Beyond Petrochemicals.
Dr. Trasande has received support for travel or meetings from the Endocrine Society, World Health Organization, United Nations Environment Programme, Japan Ministries of Environment and Health, and the American Academy of Pediatrics. She has also received royalties and licenses from Houghton Mifflin Harcourt, Audible, Paidos, and Kobunsha and has served in leadership or fiduciary roles at Beautycounter, Ahimsa, Grassroots Environmental Education, and Footprint. None of these activities were relevant to the present study. The terms and conditions of all these relationships are managed by NYU Langone Health.
Along with Hyman and Dr. Trasande, co-investigator was Jonathan Acevedo, MPH, of NYU Langone.
Source:
Journal Reference:
Hyman, S., and others. (2026). Premature birth due to exposure to chemicals used in plastic materials: a global estimate. eClinicalMedicine. doi: 10.1016/j.eclinum.2026.103842. https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(26)00089-1/fulltext
