Harvard Study Predicts Environmental Risks for Alzheimer’s

A team of Harvard researchers built a model to predict risk of Alzheimer’s disease and related dementia from nearly 200 environmental exposures, according to a study published last month.

The study, published in the journal Alzheimer’s & Dementia last month, used data from the UK Biobank to show that environmental exposures from five domains — lifestyle, social, chemical, physical, and ecosystems — could predict risk for ADRD alongside genetic variants.

The researchers then converted the data into a tool that could predict an individual’s risk based on the environmental factors, finding that “higher risk score” individuals had nearly three times higher risk of dementia than “lower risk score” individuals.

Two of the researchers — Chenshuang Li, a postdoctoral fellow at the Harvard School of Public Health, and Chirag Patel, an associate professor at Harvard Medical School — said in a Wednesday interview that the study could help inform policies intended to prevent Alzheimer’s disease and dementia.

“If we ask the question about somebody’s environment, it can be just as important as a genetic variant for risk for Alzheimer’s disease,” Patel said. “That was what struck me.”

Research has previously shown that individuals with a genetic variant in their ApoE gene have a much greater risk for Alzheimer’s than people without that variant. But Li and Patel said they wanted to do something different: apply the methodologies used for genome-wide studies to assess the influence of environmental factors via an exposome-wide association study.

From the nearly 200 environmental factors they screened for, the researchers identified 76 exposures strongly linked to dementia and 34 to Alzheimer’s, including smoking, poor sleep, and exposure to nitrogen dioxide through air pollution. Exercise, living with a partner, and eating cheese and poultry were linked to lower ADRD risk.

While Patel and Li cautioned that the results proved correlation and not causation, they said the findings could push future researchers closer to identifying causal links and help target populations for screening or Alzheimer’s clinical trials.

The two said they were already working with others at HSPH on an experiment that could prove causation between certain exposures and manifestations of dementia.

“The work that Chenshuang is doing, while correlational, are still predictive tools that could help us figure out and screen the population of who might be at higher risk,” Patel said. “Once you have a screening tool, you can ask the question: ‘Okay, who might be eligible for a clinical trial for Alzheimer’s disease?’”

Patel said the collaboration arose from a shared interest in probing the causal nature of Alzheimer’s, and said it was “great happenstance” that he was able to find a partner who brought expertise in large data analysis and predictive modeling to the table.

“It was awesome to get a person who can actually work on this project who’s properly skilled to do so,” Patel said. “This research project was a marriage of the scientific question and having the appropriate skills to address it.”