Tropical cyclones increase infant mortality in developing countries

Tropical cyclones, including storms below hurricane and typhoon strength, were associated with a sharp rise in infant mortality in low- and middle-income countries during the first two decades of this century, according to new research published in Science Advances. The findings point to a critical need for stronger disaster response and child health protections in vulnerable regions, especially as climate change increases the frequency and severity of these storms.

Infants in these regions exposed to tropical cyclones before they were born or during their first year of life were significantly more likely to die: Infant mortality rose an average of 11% over baseline rates – an increase of 4.4 deaths per 1,000 live births.

The risk was greatest in the first year after a storm and did not appear to persist beyond two years after.

Surprisingly, the increase in mortality could not be explained by reduced access to prenatal care or worsening nutrition, two commonly cited health risks following natural disasters. "The fact that health care use and undernutrition were not affected by tropical cyclone exposure suggests that the mortality effects are driven by other factors that we could not directly study," said lead author Zachary Wagner, associate professor (research) of economics at the USC Dornsife College of Letters, Arts and Sciences and senior economist at the College's Center for Economic and Social Research. "We have a lot more research to do to uncover these main drivers."

The study team – which included researchers from RAND Corporation, Stanford University, Johns Hopkins University and Belgium's UCLouvain – found that the increased mortality followed not just the most intense storms but also lower-intensity tropical storms, which are far more common.

"Some of the countries had only a few Category 3 or higher storms during the period we studied," said Zetianyu Wang, a PhD student under Wagner at RAND and first author on the report. This made it more difficult to detect clear links between the most intense storms and infant deaths. "But that doesn't mean the impact of larger storms is absent. As the planet warms, we risk more tragedies across the globe if measures aren't taken to protect children in the poorest countries."

Storm effects varied widely among countries

The researchers analyzed nearly 1.7 million child records from seven economically disadvantaged countries: Madagascar, India, Bangladesh, Cambodia, the Philippines, the Dominican Republic and Haiti.

While the average increase in infant mortality across all of these countries was 11%, the impact of storms varied significantly from country to country. In Bangladesh, Haiti and the Dominican Republic, cyclones were followed by increases of more than 10 deaths per 1,000 births. In contrast, little to no increase in mortality was observed in India, the Philippines, Cambodia and Madagascar. The researchers say it remains unclear why some countries fare better than others. The differences may reflect varying levels of disaster preparedness, geographic vulnerability, or underlying public health conditions.

Some countries may be helped by mountains while others have more flood-prone areas. Some countries have better systems in place for evacuation efforts, or they may have sturdier housing while others rely on thatched roofs. And in some places, children may already be malnourished or in poor health from malaria and other diseases, which increases vulnerability."

Zachary Wagner, associate professor (research) of economics, USC Dornsife College of Letters, Arts and Sciences

Understanding the reasons behind the differences, Wagner says, will be a key focus of future research.

"If we want to protect children from the growing threat of climate-linked disasters," he said, "we need to understand not just where risk is greatest, but why."

About this study

In addition to Wagner and Wang, authors on the study include Renzhi Jing, Sam Heft-Neal and Eran Bendavid of Stanford University; Aaron Clark-Ginsberg of RAND; and Debarati Guha-Sapir of UC Louvain and Johns Hopkins University.