When Emily Y. Chew, MD, was a child in a Chinese immigrant family in Canada, her mother would cook with goji berries—which, as it happens, contain the beneficial antioxidant lutein. “And before I had a tough exam, my father would make fish, because it was ‘brain food,’” Dr. Chew said during her Jackson Memorial Lecture at Sunday’s Opening Session.
These “Chinese food secrets,” as she called them, continue to have resonance today, given Dr. Chew’s pioneering research into age-related macular degeneration (AMD). As it has turned out, she said, in AMD—as with so many other health conditions—“You are what you eat.”
Dr. Chew’s lecture focused on three key areas of AMD research: nutrition, genetics, and artificial intelligence (AI).
The food we eat. Studies of the Mediterranean diet have found that it can reduce both the incidence of AMD1 and the risk of disease progression2 by as much as 20% to 40%, Dr. Chew said.
In a recent analysis of data on almost 8,000 patients in the Age-Related Eye Disease Studies (AREDS) database, NEI researchers found that those who had a high intake of foods in the nine categories of the Mediterranean diet had the lowest risk of progressing to late AMD. The results—a 29% reduction—are “nothing to sneeze at,” she said. Moreover, high intake of fish reduced the risk of progression by 31%.
Similar beneficial effects were found in patients who were in the earlier stages of the disease, Dr. Chew reported. Her overall conclusion: “It’s never too late a start a Mediterranean diet.”
The genes we inherit. Multiple genes have been implicated in the development of AMD, and genetic testing may help detect pathways of disease development and identify patients at risk of progression, Dr. Chew said. However, while genetic testing currently is important in the research arena, it is “much less so in clinical practice”—and, at this time, she said, “no prospective data support routine genetic testing” in the office (or home) setting.
On the potential interaction between nutrition and genetics, she said, “fish pops up again,” particularly with regard to geographic atrophy: A high intake of fish reduced risk of disease progression by 65% in those who had the highest genetic risk of progression.
The algorithms we use. AI and machine learning will continue to be brought to bear on detecting AMD features, classifying disease severity, and predicting risk of progression, Dr. Chew said. Earlier this year, she and her NEI colleagues published a paper on the DeepSeeNet model; they found that it “demonstrated high accuracy with increased transparency in the automated assignment of individual patients to AMD risk categories based on the AREDS Simplified Severity Scale.”3
Where do the rapid advances in AI leave clinicians? “We’ll never be out of a job. Deep learning will not replace us, but it will enhance our diagnostic skills and improve our clinical management,” Dr. Chew said. —Jean Shaw
1 Merle BM et al. Ophthalmology. 2019;126(3):381-390.
2 Merle BM et al. Am J Clin Nutr. 2015;102(5):1196-1206.
3 Peng Y et al. Ophthalmology. 2019;126(4):565-575.
Financial interests: Dr. Chew. None.
Watch three interviews with Dr. Chew. Sunir J. Garg, MD, FACS, interviews Dr. Chew about aspects of her Jackson Memorial Lecture:
Read more news from AAO 2019 and the Subspecialty Day meetings.