A diet-related life skills scale predicts self-care and nutritional risk in patients with diabetes undergoing hemodialysis: establishing a clinically relevant cutoff value

Abstract Background Poor self-management in older patients with diabetes undergoing hemodialysis contributes to adverse outcomes, including malnutrition. However, practical tools to identify patients who require targeted support are lacking. This study aimed to establish and validate a clinical cutoff value for a Diet-Related Life Skills Scale (DRLSS) to predict poor self-care and nutritional risk in this population. Methods This multicenter, cross-sectional study included 140 patients with diabetes undergoing hemodialysis aged 65 years or older. We assessed diet-related life skills, self-care behavior, nutritional status (using the Geriatric Nutritional Risk Index [GNRI]), and functional/psychosocial status. Receiver operating characteristic (ROC) curve analysis was performed to determine a cutoff value for predicting poor self-care. The validity of this cutoff was tested against GNRI-defined nutritional risk, including a sensitivity analysis using an alternative GNRI threshold (< 91.2). Additionally, age-stratified ROC analyses were conducted to examine performance differences between early old (65–74 years) and late-old (≥ 75 years) groups. A multiple linear regression analysis was conducted to identify predictors of self-care. Results The DRLSS demonstrated fair discriminatory power for predicting poor self-care (area under the curve (AUC) 0.760). A clinical cutoff of 116 was established, yielding a sensitivity of 0.726 and a specificity of 0.723. Stratified analysis revealed superior discriminatory performance in the early old group (AUC 0.815). Validation against nutritional risk identified a nearly identical cutoff (116; AUC 0.660), and the sensitivity analysis using GNRI < 91.2 produced a similar pattern. The high life skills group (> 116) demonstrated significantly better self-care ( p < 0.001) but paradoxically had a higher prevalence of moderate or higher nutritional risk ( p = 0.009). Multiple regression analysis identified the total life skills score ( β = 0.53, p < 0.001) and depression ( β = −0.198, p = 0.007) as significant independent predictors of self-care behavior. Conclusions The DRLSS is a valuable tool for assessing self-management capabilities in older patients with diabetes undergoing hemodialysis, particularly in the early old population. A clinical cutoff of 116 may help identify individuals with complex needs, including poor self-care or hidden nutritional risk, and support the design of targeted, multidisciplinary interventions. The paradoxical association between higher life skills and poorer nutritional status warrants further investigation into potential over-restriction.