Oral Presentation 2025 Joint Meeting of the COSA ASM and IPOS Congress

Mortality Outcomes of Eating As Treatment: 5 year survival following a multi-centre trial of a psychological intervention to reduce malnutrition in head and neck cancer patients (126687)

Ben Britton 1 , Amanda L Baker 2 , Luke Wolfenden 1 , Christopher Wratten 3 , Judy Bauer 4 , Alison k Beck 5 , Kristen Mccarter 1 , Tonelle Handley 1 , Gregory Carter 1
  1. University of Newcastle, Callaghan, NSW, Australia
  2. National Drug and Alcohol Centre, University of NSW, Randwick, NSW, Australia
  3. Radiation Oncology, Calvary Mater Newcastle, Waratah, NSW, Australia
  4. Department of Nutrition, Dietetics & Food, Monash University, Melbourne, VIC, Australia
  5. School of Psychology, University of Wollongong, Wollongong, NSW, Australia

Objectives/Purpose

Malnutrition affects up to 80% of head and neck cancer (HNC) patients, leading to poorer outcomes and higher mortality. The “Eating As Treatment” (EAT) behavioral intervention previously improved nutrition, depression, and quality of life in this group. This study investigates the effect of the EAT intervention on 5-year mortality rates among trial participants.

Sample and Setting

A multicenter, stepped-wedge, randomized controlled trial was conducted across five Australian hospitals, including HNC patients undergoing radiation therapy. This design facilitated a phased rollout of the intervention.

Procedures

Dietitians trained in motivational interviewing and cognitive behavior therapy delivered the EAT intervention. A secondary, intention-to-treat analysis was performed to assess survival benefits. Differences in 5-year all-cause mortality between the control and EAT groups were analysed using multivariable logistic regression, while Cox proportional hazards regression was used for 5-year survival rates. Analyses were adjusted for study duration, hospital site, and baseline nutritional status.

Results

Over five years, 64 deaths occurred. The mortality rate was 24% (36 of 150) in the control group versus 18% (28 of 156) in the EAT group. Logistic regression showed a significant reduction in the odds of death for the EAT group (odds ratio, 0.33), representing a 17% absolute and 55% relative risk reduction. The number needed to treat was six. Survival analysis confirmed the EAT intervention significantly reduced death risk (hazard ratio, 0.39).

Conclusion and Clinical Implications

The EAT intervention provides a statistically and clinically significant 5-year survival benefit for HNC patients, likely due to improved nutrition during radiation therapy. These findings support the integration of this targeted behavioral intervention into standard clinical care to improve HNC outcomes. Further research using stepped-wedge designs to guide broader implementation is recommended.