乌鸦传媒

A diet higher in ultra-processed foods was linked with a higher risk of dementia, but substituting unprocessed or minimally processed foods for ultra-processed ones dropped dementia risk, a longitudinal study showed.

Among more than 72,000 people in the U.K. Biobank cohort, dementia risk rose as consumption of ultra-processed foods increased in fully adjusted models (HR for a 10% increase in ultra-processed foods 1.25, 95% CI 1.14-1.37), reported Yaogang Wang, PhD, of Tianjin Medical University in China, and colleagues.

Replacing 10% of ultra-processed food weight with an equivalent proportion of unprocessed or minimally processed foods, however, led to an estimated 19% lower risk of dementia (HR 0.81, 95% CI 0.74-0.89), Wang and co-authors wrote in .

Ultra-processed foods -- which are high in added sugar, fat and salt, and low in protein and fiber -- include soft drinks, salty and sugary snacks, ice cream, sausage, deep-fried chicken, yogurt, canned baked beans and tomatoes, ketchup, mayonnaise, and packaged guacamole and hummus, the researchers noted.

"These foods may also contain food additives or molecules from packaging or produced during heating, all of which have been shown in other studies to have negative effects on thinking and memory skills," co-author Huiping Li, PhD, also of Tianjin Medical University, said in a statement.

"Our results also show increasing unprocessed or minimally processed foods by only 50 grams a day, which is equivalent to half an apple, a serving of corn, or a bowl of bran cereal, and simultaneously decreasing ultra-processed foods by 50 grams a day, equivalent to a chocolate bar or a serving of fish sticks, is associated with 3% decreased risk of dementia," Li added.

Nutrition research has become increasingly focused on food processing in recent years, but categorizing foods as unprocessed, minimally processed, processed, or ultra-processed remains a challenge, noted Maura Walker, PhD, and Nicole Spartano, PhD, both of Boston University, in an .

"For example, foods like soup would be classified differently if canned versus homemade," Walker said. "Plus, the level of processing is not always aligned with diet quality. Plant-based burgers that qualify as high quality may also be ultra-processed."

"As we aim to understand better the complexities of dietary intake, we must also consider that more high-quality dietary assessments may be required," Walker pointed out.

Wang and co-authors evaluated 72,083 people in the study who were 55 or older and dementia-free at baseline. Participants had at least two valid 24-hour dietary assessments using the questionnaire during 2009-2012. Researchers applied the , a system of classifying foods by the amount of industrial processing they undergo, to the dietary questionnaire data and grouped participants into quartiles.

Ultra-processed foods were a median of 8.6% of a participant's daily diet in the lowest quartile and 27.8% in the highest quartile. Food groups that contributed most to high ultra-processed food intake were beverages (34%), sugary products (21%), ultra-processed dairy products (17%), and salty snacks (11%).

Mean age of participants was 62; about 53% were women and 93% were white. Median follow-up was 10 years. In that period, 518 people developed dementia, including 287 with Alzheimer's disease and 119 with vascular dementia.

In the lowest quartile, 105 people developed dementia, compared with 150 people in the highest quartile. An increase of 10% in ultra-processed foods raised the risk of all-cause dementia by 25%, vascular dementia by 28%, and Alzheimer's disease by 14%. Results persisted after adjusting for covariates like total calorie intake, healthy diet scores, and dementia-related factors.

Replacing 20% of ultra-processed food weight with an equivalent proportion of unprocessed or minimally processed food led to a 34% lower risk of dementia and 39% lower risk of vascular dementia, but did not significantly affect Alzheimer's disease risk.

In secondary analyses, ultra-processed meat, fish, and eggs were associated with Alzheimer's disease. But the researchers included ham and bacon as ultra-processed foods when these meat products might be considered processed, noted the editorialists. "While this may seem like a trivial difference, misclassifying these meat products may in part account for the observed relation with Alzheimer's disease," Walker and Spartano wrote.

The analysis had several limitations, the researchers noted. Misclassification may have occurred; some foods like yogurt can have more or less processing, for example. Incident dementia was ascertained by inpatient data and death registries, not primary care records. In addition, residual or unmeasured confounding may have influenced results.

Comment