Background: Dietary patterns (DP) are associated with health outcomes in younger adults but there is a lack of evidence in the very old (aged 85+) on DP and their association with sociodemographic factors, lifestyle, health and functioning measures. Higher socioeconomic status (SES) has been linked with healthier DP but it is not known whether these associations are sustained in the very old.
Objective: We aimed to (a) characterise DP in the very old and (b) assess the relationships between three SES indicators (education, occupational class and area-deprivation index [IMD]) and DP.
Methods: Complete dietary data at baseline (2006/07) for 793 participants in the Newcastle 85+ Study were established through 24-hr multiple pass recall. We used Two-Step clustering and 30 food groups to derive DP, and multinomial logistic regression models to assess the association with SES.
Results: We identified three distinct DP (characterised as 'High Red Meat', 'Low Meat', and 'High Butter') that varied with key sociodemographic, health and functioning measures. 'Low Meat' participants were more advantaged (i.e. higher education and occupational class, and lived in more affluent areas in owned homes), were least disabled, cognitively impaired, and depressed, and were more physically active than those in the other DP. After adjusting for other lifestyle factors, cognitive status and BMI, lower educational attainment remained a significant predictor of 'High Red Meat' and 'High Butter' membership compared with 'Low Meat' ('High Red Meat': OR [95% CI] for 0-9 and 10-11 years of education vs. ≥12 years: 5.28 [2.85-9.79], p<0.001 and 3.27 [1.65-6.51], p = 0.001, respectively; 'High Butter': 3.32 [1.89-5.82], p<0.001 and 2.83 [1.52-5.28], p = 0.001).
Conclusions: In this cohort of very old adults, we detected a favourable DP ('Low Meat'), which was associated with better health and functioning and higher SES.