Survival in older adults with dementia and eating problems: To PEG or not to PEG?

Background & aims: Despite guidelines, long-term enteral nutrition (EN) through percutaneous endoscopic gastrostomy (PEG) is often prescribed to older individuals with dementia and eating problems (refusal to eat or dysphagia). The aim of this prospective observational non-randomized un-blinded study was to assess the role of this procedure on risk of mortality. Methods: 184 demented malnourished patients (58 M, age 82.2 ± 7.7) with eating problems, discharged from a hospital ward in Italy, were enrolled. Information on dementia type and staging (FAST and CDR scores), Charlson Comorbidity Score and setting of living (community vs nursing home) was collected. After an 18-month follow-up, a telephonic interview with caregivers was planned to assess mortality. Survival of patients discharged on EN by PEG vs oral nutrition (ON) was analyzed by Kaplan-Meier method. Multivariable Cox proportional regression models were also built to test the effects of EN over mortality. Results: EN was prescribed in 54 patients (15 M). At follow-up, mortality was higher in EN than in ON group (70% vs 40%, p = 0.0002). Survival was significantly shorter in the EN group (log-rank 17.259, p < 0.0001; average length 0.66 ± 0.09 vs 1.28 ± 0.08 years, p < 0.0001). At multivariate Cox proportional regression model, EN was a significant predictor of death (HR 1.82, 95% CI 1.09-3.02, p = 0.02) independent of age, dementia type, FAST, CDR, Charlson score and setting of living in the whole cohort, but not in those with CDR score ranking 4-5. Conclusions: In elderly individuals with dementia and eating problems, long-term PEG feeding increases the risk of mortality and should be discouraged.