A major, USC-led study of more than 214,000 older adults across 14 countries and regions finds that the most common controllable risk factors for dementia — such as low education, high blood pressure and smoking — vary widely from country to country, meaning a one-size-fits-all approach to prevention won’t work everywhere.

The findings appeared Sunday in the journal The Lancet Healthy Longevity.

Most of what scientists know about preventing dementia comes from research in wealthy countries such as the United States and those in Western Europe. But this study, led by researchers at USC along with colleagues at Brown University and Johns Hopkins University, set out to see whether those same patterns hold true in low- and middle-income countries.

Differences — and similarities — in dementia risk factors

The differences were striking — but so were the similarities:

• Differences: Low education affected 86% of older adults in China but only 12% in the United States, while high BMI (a measure of excess body weight) affected 45% of Americans compared to just 13% of people in India.
• Similarities: Certain risk factors tended to cluster together in similar patterns worldwide, such as cardiovascular risks (for example, high cholesterol and hypertension) or risky behaviors (such as smoking and drinking).

Lead author Emma Nichols, a research scientist with the Center for Economic and Social Research at the USC Schaeffer Institute for Public Policy & Government Service, says the consistency of these clusters was the most unexpected part of the findings.

“I was less surprised by the differences and more surprised by some of the similarities, particularly in the ways these risks are patterned across settings,” Nichols said. “That has real implications for how we design prevention strategies and interventions, because some things are more consistent across places than we might expect.”

For the study, researchers from the Gateway to Global Aging Data team combined harmonized survey data from long-running aging studies in 14 places — the United States, Brazil, China, England, India, Ireland, Korea, Malaysia, Mexico, Northern Ireland and four regions of Europe — collected between 2009 and 2023. (Jinkook Lee of the Center for Economic and Social Research at the USC Schaeffer Institute is principal investigator of the Gateway to Global Aging Data project as well as the Longitudinal Aging Study in India.)

They analyzed 12 modifiable risk factors identified by the Lancet Commission on dementia — things such as hearing loss, depression, physical inactivity and social isolation, comparing how common each factor was; how they varied by age, gender and education level; and how often multiple risk factors showed up together in the same person.

Dementia risk factors help guide decision-making

The researchers say the findings should help guide decision-makers and health organizations in designing dementia-prevention strategies tailored to their own populations. For example, a program that connects people to care for diabetes could be redesigned to address the entire cluster of related cardiometabolic risks, such as high cholesterol and hypertension, at the same time.

For the average person, Nichols adds, the takeaway is that dementia risk is not fixed or fated: “Risk for these late-life outcomes isn’t predetermined. These are risk factors you experience over the life course, and you can have an impact on changing your own risk — while also recognizing the ways broader societal factors shape that risk, too.”

Future work may also expand to include newer risk factors such as poor sleep and additional countries as more harmonized data becomes available; new data collection is already underway in additional countries, including Kenya and Egypt.

---

About the study: In addition to Nichols, other authors include senior author Jinkook Lee, Michael Markot, Drystan Phillips and Jenny Wilkens, all of the Gateway to Global Aging Data team at the Center for Economic and Social Research at the USC Schaeffer Institute; co-first author Zachary Kunicki of the Warren Alpert Medical School of Brown University; and Alden Gross of the Johns Hopkins Bloomberg School of Public Health.

This research was supported by the National Institutes of Health (grant R01AG030153).

As a heat dome drives dangerous temperatures across much of the United States and renews concerns about extreme heat, USC researchers have developed a new, freely available AI tool that could help cities better understand one of their best defenses against rising temperatures: trees.

The USC-developed tool combines free aerial imagery with AI to create detailed tree canopy maps without costly lidar surveys, which use lasers to produce highly detailed 3D landscape maps, or commercial satellite imagery that many existing systems require. The technology could help communities make smarter, targeted investments in urban forests as extreme heat becomes more common.

“Trees provide a wide range of benefits, including helping reduce the health risks associated with rising temperatures in cities,” said John Wilson, founding director of the Spatial Sciences Institute at the USC Dornsife College of Letters, Arts and Sciences and the study’s corresponding author. “But to plant trees where they’ll make the biggest difference, cities first need a clear picture of their existing tree canopy.”

The study, published in Remote Sensing, is the latest example of how USC researchers are applying artificial intelligence — one of the university’s strategic priorities — to solve real-world problems, in this case helping cities make smarter, more cost-effective decisions about where to invest in urban trees.

A cheaper way to map the urban tree canopy

Unlike many high-accuracy tree-mapping systems, which rely on expensive lidar surveys or commercial satellite imagery, the USC-developed tool works with free aerial photographs collected nationwide every two to three years through the U.S. Department of Agriculture’s National Agriculture Imagery Program. By combining those publicly available images with AI, the researchers dramatically reduced the cost of producing detailed tree canopy maps, making the technology practical for cities and communities that may not have the resources for specialized surveys.

“We need fine-scale data to know where to plant new trees that provide the best return on investment,” said Wilson, who is also a professor of spatial sciences and sociology at USC Dornsife, with adjunct appointments in the USC School of Architecture, the Department of Population and Public Health Sciences at the Keck School of Medicine of USC, and the departments of computer science and civil and environmental engineering at the USC Viterbi School of Engineering.

“Our work shows how we can use free, publicly accessible data to map tree canopy over time — data cities can use to guide planting plans at every scale, from a single street block to an entire county,” Wilson said.

Putting the tool to the test

The researchers developed and tested the system in Boyle Heights and City Terrace, two densely populated, majority-Latino neighborhoods east of downtown Los Angeles that have historically had less tree cover than wealthier parts of the city.

The canopy-mapping model accurately identified tree cover, while the individual tree detection model — a more challenging task because tree crowns appear small and often overlap in aerial imagery — performed competitively with far more expensive lidar-based approaches.

To test whether the approach could work beyond Southern California, the researchers applied the trained models, without additional retraining, to neighborhoods in San Francisco and Phoenix. Despite the cities’ different climates and urban layouts, the tool produced consistently strong results, suggesting communities elsewhere may be able to use the model and avoid building one from scratch.

Interest in the technology is already growing. The ArcGIS deep learning package developed through the project has been downloaded more than 12,900 times from Esri’s Living Atlas platform over the past six months, according to Wilson.

The research, code and a ready-to-use ArcGIS deep learning model are freely available online, making the tool accessible to municipalities without in-house machine learning expertise.

The work also feeds directly into USC’s efforts to green Los Angeles. Wilson’s geospatial research helps inform the USC Urban Trees Initiative, a USC Dornsife Public Exchange program that has worked since 2020 with the city of Los Angeles, local nonprofits and community groups to identify where trees are most needed and to guide their planting across neighborhoods including Boyle Heights, South Los Angeles and the Eastside.

What’s next

Wilson said the team’s next step is to pair its AI tool with freely available lidar data that captures the height and three-dimensional structure of tree canopies.

“Knowing both the height and extent of the canopy will allow us to estimate the shade trees provide today — and model how much additional shade new plantings could create,” Wilson said.

The researchers plan to begin by analyzing individual street blocks, school playgrounds and parks before expanding the approach to neighborhoods, cities and counties, giving communities even more precise information for planning cooler, healthier and more resilient urban environments.

---

About the study: Co-authors include Yi Qi, Isaac Ashe-McNalley and Beau MacDonald of USC’s Spatial Sciences Institute, and Jooyoung Yoo of Emory University.

The research was supported by the Climate-related Exposures, Adaptation, and Health Equity (CLIMA) Center, funded by the National Heart, Lung, and Blood Institute (grant P20HL176204); the Southern California Environmental Health Sciences Center (SCEHSC), funded by the National Institute of Environmental Health Sciences (grant P30ES007048); and the Bezos Earth Fund.