Background: Anorexia is an important cause of protein-energy malnutrition (PEM) in haemodialysis patients. We investigated whether self-reported appetite was associated with death and hospitalization in subjects enrolled in the Hemodialysis (HEMO) Study.
Methods: The HEMO Study was a 7-year, multicentre, randomized trial (N = 1846), which examined the effects of dialysis dose and membrane flux on mortality and morbidity. Three questions from the Appetite and Diet Assessment Tool (ADAT) were used to determine whether appetite had changed over time in the randomized treatment groups. The relations among ADAT scores, dietary protein and energy intakes, biochemical and anthropometric measures, and quality of life were assessed. We used Cox proportional hazards models to evaluate the relative risks of death and hospitalization associated with static and dynamic ADAT scores, adjusted for demographic factors, dose and flux assignments, and co-morbidity.
Results: The average length of follow-up was 2.84 years. After adjusting for demographic factors and randomized treatment assignments, there was a significant association between poorer self-reported appetite and death (RR 1.52, 95% CI 1.16-1.98); however, the association became non-significant with further adjustment for co-morbidity (RR 1.23, 95% CI 0.94-1.62). Poorer appetite was unequivocally associated with increased hospitalization rates (multivariable RR 1.35, 95% CI 1.13-1.61). The longitudinal effect of worsening appetite from baseline to 1 year was not associated with mortality or hospitalization rate after adjusting for co-morbidity.
Conclusions: The association between appetite and death was confounded by co-morbidity. Self-reported appetite was associated with hospitalization rate in haemodialysis patients and, thus, it may be a useful screening tool for this outcome. Patients who report poor or very poor appetites should be monitored, and they should receive more comprehensive nutritional assessments.