Educational Technology and Change Journal

By Jim Shimabukuro (assisted by Perplexity and Gemini)
Editor

Li and colleagues’ Neurology¹ (27 July 2022) cohort analysis of UK Biobank participants concluded that higher ultraprocessed food (UPF) intake is associated with increased risk of all‑cause dementia, Alzheimer’s disease, and vascular dementia, and that substituting unprocessed or minimally processed foods for UPFs is linked to a lower dementia risk. Subsequent research from 2022–2025 has largely reinforced the direction of this association but has refined the magnitude of risk, highlighted age- and dose‑related nuances, and raised sharper questions about residual confounding and causality.(jamanetwork+7 5 Dec 2022)

Li et al. analyzed 72,083 dementia‑free participants aged 55 years or older from the UK Biobank, classifying foods by NOVA categories and linking baseline dietary data to incident dementia over a median follow‑up of about 10 years. In fully adjusted Cox models, each 10% increase in the proportion of dietary weight from UPFs was associated with a 25% higher risk of overall dementia (hazard ratio [HR] 1.25, 95% confidence interval [CI] 1.14–1.37), a 14% higher risk of Alzheimer’s disease (HR 1.14, 95% CI 1.00–1.30), and a 28% higher risk of vascular dementia (HR 1.28, 95% CI 1.06–1.55). The substitution analysis suggested that replacing 10% of UPF weight with minimally processed or unprocessed foods corresponded to a 19% lower dementia risk (HR 0.81, 95% CI 0.74–0.89), and larger modeled substitutions (for example, 20%) implied proportionally larger relative reductions, especially for vascular dementia. The authors stressed the observational nature of the study and the possibility of residual confounding, but nonetheless argued that limiting UPFs could be one component of dementia prevention strategies.(aan+2 27 July 2022)

In 2022, a JAMA Neurology cohort study from the Brazilian Longitudinal Study of Adult Health (ELSA‑Brasil) extended the UPF–brain health link from dementia incidence to trajectories of cognitive decline. Following 10,775 middle‑aged and older adults for a median of eight years, the investigators reported that individuals in the highest quartiles of UPF consumption experienced about 28% faster global cognitive decline and 25% faster executive function decline than those in the lowest quartile, after adjustment for demographic, lifestyle, and clinical covariates. Importantly, stratified analyses suggested that the association was evident in participants younger than 60 years, particularly in those with low overall diet quality, whereas it was attenuated or absent among older adults and those with higher healthy‑diet scores. This age‑ and diet‑quality interaction complicates a simple extrapolation from Li et al.’s dementia‑focused findings: whereas Li et al. treated mid‑to‑late life as a relatively homogeneous risk window, the Brazilian results imply that UPF‑related vulnerability may be greatest in midlife and may be partly buffered by healthier overall dietary patterns.(jamanetwork 5 Dec 2022)

By late 2023 and 2024, evidence syntheses began integrating Li et al. with other large cohorts to quantify the overall association between UPF intake and dementia and to interrogate dose–response patterns. A systematic review and meta‑analysis published in 2023–2024, which explicitly framed itself as the first to examine UPF and dementia, concluded that higher UPF consumption was consistently associated with greater dementia prevalence or incidence across included studies, echoing Li et al.’s core message. However, this review did not find a statistically robust dose–response gradient between moderate increments of UPF intake and dementia, suggesting that while very high consumption is clearly unfavorable, the risk curve may flatten at moderate intakes or be obscured by measurement error and confounding. In contrast, several of the underlying cohort studies, including Li et al., reported monotonic associations per 10% of energy or weight from UPFs, highlighting heterogeneity in analytical strategies and the difficulty of harmonizing exposure definitions and covariate adjustments across datasets.(pmc.ncbi.nlm.nih+1 13 Oct 2023)

Newer prospective analyses have also focused more explicitly on life‑course timing and dementia subtypes. A 2023–2024 study published in the Alzheimer’s Association journal Alzheimer’s & Dementia reported that UPF consumption in middle age and “young‑old” adulthood was associated with increased risk of all‑cause dementia, with especially strong associations for Alzheimer’s disease when UPF intake exceeded about 10 servings per day. Among participants younger than 68 years at baseline, each additional energy‑adjusted serving of UPFs per day corresponded to a 13% higher risk of Alzheimer’s disease (HR 1.13, 95% CI 1.03–1.25), and consuming at least 10 daily servings versus fewer than 10 was linked to a 2.7‑fold increase in Alzheimer’s disease risk (HR 2.71, 95% CI 1.18–6.24) after extensive adjustment. However, these associations weakened or disappeared in those who were 68 or older at baseline, paralleling the ELSA‑Brasil finding that UPFs may be a more salient risk factor earlier in the cognitive aging trajectory. Compared with Li et al., who reported somewhat similar hazard ratios for Alzheimer’s and vascular dementia and did not emphasize age stratification to the same degree, the newer work suggests a more complex pattern in which extreme UPF intake in midlife may be particularly relevant to Alzheimer’s risk, while vascular pathways may dominate in some older or more cardiovascularly burdened subgroups.(pmc.ncbi.nlm.nih+3 1 Jan 2025)

By 2025, large‑scale meta‑analyses of UPFs and neurodegenerative outcomes, encompassing dementia, Parkinson’s disease, multiple sclerosis, and cognitive impairment, further contextualized Li et al.’s dementia‑specific results. A dose‑response meta‑analysis of large‑scale cohorts found that persistent overconsumption of UPFs was associated with elevated risks across multiple neurodegenerative conditions, including approximately 17% higher risk of cognitive impairment and increased risks for dementia and Parkinson’s disease, while again underscoring that causal inference remains uncertain. Within this broader synthesis, the UK Biobank dementia analysis by Li et al. appears at the higher end of risk estimates, both because the exposure was expressed per 10% of total food weight rather than per serving and because the study could finely adjust for socioeconomic and lifestyle variables captured in the Biobank. Later cohorts with different dietary assessment instruments—typically food‑frequency questionnaires rather than repeated 24‑hour recalls—and shorter follow‑up sometimes reported more modest hazard ratios, which meta‑analysts interpreted as reflecting a mix of genuine population differences, regression dilution from measurement error, and variation in how aggressively studies adjusted for total diet quality, cardiometabolic status, and social determinants.(pubmed.ncbi.nlm.nih+5 16 May 2024)

Methodologically, post‑2022 studies have expanded on and, in some cases, critiqued aspects of Li et al.’s design. Li and colleagues relied on a single baseline dietary assessment, applied the NOVA classification to UK food codes, and used energy‑adjusted UPF weight to define exposure, with substitution models that replaced equal weights of UPFs with minimally processed foods. Subsequent work has experimented with alternative exposure metrics, including percentage of total energy from UPFs, servings per day, and latent dietary patterns that co‑vary with UPF intake, and has incorporated repeated dietary measures where available to reduce regression dilution. Several newer analyses have also devoted more attention to competing risks (for example, death from cardiovascular disease before dementia onset) and to fine‑grained sensitivity analyses excluding early cases to mitigate reverse causation, where prodromal dementia might alter dietary choices. While these refinements do not overturn Li et al.’s conclusions, they tend to slightly attenuate effect sizes and widen confidence intervals, reinforcing the notion that the true causal effect—if any—is likely smaller than the most dramatic observational estimates and operates in concert with other lifestyle and vascular risk factors.(alz-journals.onlinelibrary.wiley+6 25 Dec 2023)

Conceptually, Li et al. framed UPFs as a marker of poor diet quality and a potential direct driver of dementia via additives, packaging‑derived chemicals, advanced glycation end products, and effects on obesity, diabetes, and vascular health. Later research has preserved this vascular‑metabolic framing but has added nuance, particularly by examining effect modification by overall diet quality and by suggesting that UPFs might be especially detrimental when they crowd out nutrient‑dense foods rather than in isolation. For instance, the ELSA‑Brasil study’s finding that the UPF–cognitive decline link was most apparent in those with low healthy‑diet scores indicates that diets that are both high in UPFs and low in whole foods may be the highest‑risk pattern, which aligns with Mediterranean‑ and DASH‑style dietary research on dementia. Meta‑analysts have also emphasized that UPFs are socioeconomically patterned—being more common in individuals with lower income, education, and neighborhood resources—and that these social factors themselves shape dementia risk through health‑care access, comorbidities, and cumulative life stress. Li et al. adjusted for several of these variables but could not fully disentangle UPFs from the broader social environment, a limitation shared and more explicitly foregrounded in later work.(research.lu+6 27 July 2022¹)

Synthesizing Li et al. with the 2023–2025 literature, the overall narrative has shifted less in direction than in precision and caution. Across multiple cohorts, high UPF intake is consistently associated with faster cognitive decline, greater risk of all‑cause dementia, and specifically increased risk of Alzheimer’s disease and vascular dementia, particularly when consumption is very high and begins in midlife. At the same time, more recent studies and meta‑analyses emphasize that the association is modest in magnitude, potentially non‑linear, sensitive to how UPF is measured, and entangled with broader patterns of diet quality, cardiometabolic health, and socioeconomic conditions. Li et al.’s 2022 Neurology article thus remains a central and influential piece of evidence, but it now sits within a more differentiated evidence base in which UPF consumption is best understood as one modifiable component of a complex web of midlife and later‑life risk factors for dementia rather than as a singular causal agent.(neurologyadvisor+6 2 Sep 2022)

Additional Updates From Gemini

Li and colleagues’ 2022 landmark study laid a critical foundation for understanding how the industrial processing of food, beyond mere nutrient content, might impact neurodegeneration. As we move into 2025 and 2026, the scientific consensus has evolved from establishing a broad association to dissecting the specific “culprits” within the UPF category and identifying the biological mechanisms at play. While the 2022 study treated UPFs as a monolithic group, recent research from the Virginia Tech College of Liberal Arts and Human Sciences² and reports in the American Journal of Clinical Nutrition (2025)³ have begun to differentiate between various types of ultra-processed items. These latest findings suggest that not all UPFs are created equal; specifically, ultra-processed meats (like deli meats and hot dogs) and sugar-sweetened beverages show the most robust and consistent links to cognitive impairment and poor memory, whereas some other ultra-processed categories, such as certain whole-grain breads or dairy-based UPFs, may not carry the same degree of risk.

Another significant shift in recent research involves the focus on “global cognitive aging” and domain-specific decline. A 2026 report⁴ utilizing data from the Health and Retirement Study (HRS) at the NIH found that high UPF consumption is particularly detrimental to executive function—the mental skills used for self-regulation and focus—even before a formal dementia diagnosis is made. This indicates that the damage from UPFs may manifest as “accelerated cognitive aging,” adding years to a person’s brain age well before the onset of Alzheimer’s disease. Furthermore, 2024 research presented by the Alzheimer’s Association⁵ emphasized that the risk is not just about the presence of harmful additives like nitrites and high sodium, but also the “displacement” of protective nutrients like fiber and polyphenols, which are vital for maintaining the gut-brain axis.

The most critical difference between the 2022 study and the current 2025–2026 landscape is the move toward precision and mechanistic understanding. Earlier research focused on the “what” (UPFs are linked to dementia), whereas current studies are answering the “how” (inflammation, gut microbiome disruption, and vascular damage) and the “who” (identifying middle-aged adults as the most critical group for intervention). While Li et al.’s work remains a cornerstone of nutritional neuroscience, the latest research provides a more nuanced roadmap for public health, suggesting that policy and personal choices should prioritize the reduction of specific high-risk UPF subgroups—primarily processed meats and sodas—to most effectively protect the aging brain.

__________
¹ Huiping Li et al., “Association of Ultraprocessed Food Consumption With Risk of Dementia,” Neurology, 27 July 2022.
² Jenny Kincaid Boone, “Certain processed foods and beverages linked to declines in brain health,” Virginia Tech News, 29 Sep 2025.
³ Ciarán G Forde et al., “Eating rate has sustained effects on energy intake from ultraprocessed diets: a 2-week ad libitum dietary randomized controlled crossover trial,” American Journal of Clinical Nutriton, 26 Nov 2025.
⁴ Heejin Lee et al., “Ultra-processed food intake and impairment across multiple cognitive domains in nationally representative older U.S. adults,” NIH, 14 Jan 2026.
⁵ “Processed Red Meat Raises the Risk of Dementia; Swapping It For Nuts and Beans May Lower Risk,” Alzheimer’s Association, 31 July 2024.

[End]

Filed under: Uncategorized |