Benchmarking Diet Quality to Assess Nutritional Risk in Hemodialysis Patients: Applying Adequacy and Moderation Metrics of the Hemodialysis-Healthy Eating Index

Objectives: This study modified Healthy Eating Index (HEI) based on hemodialysis (HD)-specific nutritional guidelines and investigated associations between the diet quality (DQ) and nutritional risk in HD patients. Methods: The HD-HEI tool adapted the Malaysian Dietary Guidelines 2010 framework according to HD-specific nutrition guidelines. This HD-HEI was applied to 3-day dietary records of 382 HD patients. Relationships between HD-HEI scores and nutritional parameters were tested by partial correlations. Binary logistic regression models adjusted with confounders were used to determine adjusted odds ratio (adjOR) with 95% confidence interval (CI) for nutritional risk based on HD-HEI scores categorization. Results: The total HD-HEI score (51.3 6 10.2) for this HD patient population was affected by ethnicity (Ptrend < .001) and sex (P = .003). No patient achieved "good"DQ (score: 81-100), while DQ of 54.5% patients were classified as "needs improvement"(score: 51-80) and remaining as "poor"(score: 0-51). Total HD-HEI scores were positively associated with dietary energy intake (DEI), dietary protein intake (DPI), dry weight, and handgrip strength, but inversely associated with Dietary Monotony Index (DMI) (all P < .05). Individually, scores for refined grain, total protein, and animal protein were positively associated with DEI (all P < .05), while total, animal, fish, and vegetable proteins indicated positive associations with DPI (all P < .05). Moderating metrics for convenience meals, saturated fats, sodium, and fluid negatively correlated toward DEI with similar trends for DPI excepting convenience meals and fluids. "Poor"DQ was associated with DMI >= 29.2 (adjOR 18.83, 95% CI 9.36-37.86, P < .001), Malnutrition Inflammation Score >= 5 (adjOR 1.78, 95% CI 1.01-3.15, P = .045), and protein energy wasting (adjOR 1.96, 95% CI 1.14-3.34, P = .031), but became nullified with covariate adjustments. "Poor"DQ was also associated with low lean tissue mass (<32.6 kg) in men (adjOR 2.38, 95% CI 1.01-5.58, P = .046) but not women. Conclusion: "Poor"DQ was associated with poor nutritional status in Malaysian HD patients, who should be targeted for nutritional counseling.