# Feeding Difficulties and Survival Outcomes in Advanced Dementia: A multicentre retrospective study from Oman

## TL;DR

This study from Oman found that feeding problems are common in advanced dementia and linked to shorter survival, but the link weakens after adjusting for other factors.

## Contribution

The study provides new insights into feeding difficulties and survival in advanced dementia patients in a Middle Eastern context.

## Key findings

- Feeding problems occurred in 71.8% of patients with advanced dementia.
- Patients with feeding problems had shorter unadjusted survival but not after adjusting for confounders.
- Male sex was an independent predictor of higher mortality in these patients.

## Abstract

Feeding problems are among the most common in patients with advanced dementia. This study aimed to assess the prevalence of feeding problems and evaluate survival outcomes following the onset of feeding problems in patients with advanced dementia.

A retrospective cohort study of adults with advanced dementia identified from the electronic patient record using diagnostic codes and keyword searches was conducted. This multicentre study took place at Sultan Qaboos University Hospital, Muscat, Oman and the Royal Hospital, Muscat, Oman, from January 2010 to December 2023.

A total of 202 patients were included. The median age was 78.5 years (interquartile range [IQR]: 72.0–83.0); 58.9% were male. Median follow-up was 25.9 months; 70.8% (n = 143) died and overall median survival was 33.3 months following the diagnosis of advanced dementia. Feeding problems developed in 71.8% (95% confidence interval [CI]: 65.1–77.6) of patients with a median time from diagnosis to onset of 5.9 months (IQR: 1.0–25.6). Patients with feeding problems were older (80 versus 75 years) and more often had prior stroke history (37.2% versus 19.3%). Survival was shorter with feeding problems (28.8 versus 48.0 months; log-rank P = 0.010). In the adjusted Cox models (C-index = 0.64), feeding problems were not an independent predictor of mortality (hazard ratio [HR] = 1.409, 95% CI: 0.878–2.260; P = 0.155). The stroke × feeding interaction was not significant (HR = 0.831, 95% CI: 0.280–2.463; P = 0.738).

Feeding problems identify a subgroup with poorer unadjusted survival, but the association with mortality was not significant after adjustment for confounders. Male sex independently predicted higher mortality. These findings support the need for prospective studies with standardised feeding assessments and nutritional measures.