Anti-Inflammatory Diet and Dementia in Older Adults With Cardiometabolic Diseases

Abstract

Importance: Inflammation has been proposed as a mechanism linking cardiometabolic diseases (CMDs) to increased risk of dementia. However, whether an anti-inflammatory diet can support brain and cognitive health among people with CMDs is unclear.

Objective: To examine CMD status and dietary inflammatory potential in association with dementia risk and brain magnetic resonance imaging (MRI) measures using joint effect analysis.

Design, setting, and participants: The UK Biobank is an ongoing community-based cohort study with baseline assessments conducted between March 13, 2006, and October 1, 2010. The present study included 84 342 dementia-free older adults (≥60 years), who were followed up until January 20, 2022 (maximum, 15 years). A subsample (n = 8917) underwent brain MRI scans between May 2, 2014, and March 13, 2020.

Exposures: Baseline CMDs (including type 2 diabetes, heart disease, and stroke) were ascertained from medical records. Dietary Inflammatory Index scores (anti-inflammatory [≤-1.5 points], neutral [>-1.5 to <0.5 points], or proinflammatory [≥0.5 points]) were calculated from participants' average intake of 31 nutrients, assessed up to 5 times using the Oxford WebQ, a web-based, 24-hour dietary assessment.

Main outcomes and measures: Incident dementia was identified through linkage to medical records. Regional brain volumes were collected from brain MRI scans.

Results: The study included 84 342 participants (mean [SD] age, 64.1 [2.9] years; 43 220 [51.2%] female). At baseline, 14 079 (16.7%) had at least 1 CMD. Over a median follow-up of 12.4 (IQR, 11.8-13.1) years, 1559 individuals (1.9%) developed dementia. With the use of joint effect analysis, the hazard ratio of dementia was 2.38 (95% CI, 1.93-2.93) for people with CMDs and a proinflammatory diet and 1.65 (95% CI, 1.36-2.00) for those with CMDs and an anti-inflammatory diet (reference: CMD-free, anti-inflammatory diet). Dementia risk was 31% lower (hazard ratio, 0.69; 95% CI, 0.55-0.88; P = .003) among people with CMDs and an anti-inflammatory diet. On brain MRI, participants with CMDs and an anti-inflammatory diet compared with a proinflammatory diet additionally had significantly larger gray matter volume (β = -0.15; 95% CI, -0.24 to -0.06 vs β = -0.27; 95% CI, -0.38 to -0.16) and smaller white matter hyperintensity volume (β = 0.05; 95% CI, -0.04 to 0.14 vs β = 0.16; 95% CI, 0.05-0.27).

Conclusions and relevance: In this cohort study, people with CMDs and an anti-inflammatory compared with proinflammatory diet had a significantly lower hazard ratio of dementia, larger gray matter volume, and smaller white matter hyperintensity volume.

Figure 1.. Association Between Dietary Inflammatory Potential and Dementia Risk

Gray bars represent the distribution of Dietary Inflammatory Index score in the study population. The association between the DII score and dementia risk is modeled using restricted cubic splines adjusted for age, sex, educational level, energy intake, race and ethnicity, socioeconomic status, body mass index, smoking status, physical activity, hypertension, apolipoprotein E ε4 carrier status, and cardiometabolic disease status. The blue line and blue shaded areas represent the hazard ratio and 95% CI of dementia as a function of the DII score. The reference level is set as the median DII score in the population (−1 pts).

Figure 2.. Joint Effect Analysis of Cardiometabolic Disease (CMD) Status and Dietary Inflammatory Potential in Dementia Risk, Time to Dementia Diagnosis, and Brain Structural Differences

Cox proportional hazards regression models for dementia risk (A) and Laplace regression models for time to dementia diagnosis (B) were adjusted for age, sex, educational level, energy intake, race and ethnicity, socioeconomic status, body mass index, smoking status, physical activity, hypertension, and apolipoprotein E ε4 carrier status. Linear regression models for brain structural differences (C) were additionally adjusted for time between baseline and magnetic resonance imaging scan, assessment center, and head and table position within the scanner. In all models, significant differences between the CMD with anti-inflammatory diet and CMD with proinflammatory diet groups were assessed by repeating the models using the CMD with proinflammatory diet group as the reference. GMV indicates gray matter volume; HV, hippocampal volume; TBV, total brain volume; WMHV, white matter hyperintensity volume; and WMV, white matter volume.