Objective: To investigate how total lymphocyte count (TLC) and the Geriatric Nutritional Risk Index (GNRI) are associated with short-term nutritional-related complications (death, infections, bedsores) in institutionalised elderly.
Methods: 220 home-care resident elderly (age +/- SD; 80.7 +/- 7.9, range: 67-98 years) were studied (anthropometry, biochemistry, food intake) and prospectively followed over a period of 3 months for the occurrence of health complications. Nutritional risk was assessed by GNRI. Patients were categorized according to GNRI (<92, 92-98, >98) and TLC (<900, 900-1499, >or=1500/mm3).
Results: GNRI was significantly associated with TLC according to both simple and adjusted correlation models (p < 0.001) and to multiple stepwise regression analysis (p < 0.005). TLC < 900 revealed a higher specificity (87.8%) than sensitivity (30.6%) in identifying "at-risk" patients (GNRI < 92). Adjusted multiple logistic regression revealed a significant association between overall 3-month health outcomes and both TLC and food intake. TLC was the only significant predictor for infections, while death was independently associated with GNRI and food intake. When a GNRI < 92 and a TLC < 900 were considered together, the sensitivity was 0.83 (95% confidence interval, C.I.95%: 0.66-1.0) and 0.89 (C.I.95%: 0.68-1.00) for overall complications (Odds ratio: 22.1; C.I.95%: 5.1-96.1) and infections (Odds ratio: 20.8; C.I.95%: 2.6-168.8), respectively. The association of a GNRI > 98 with a TLC >or= 1500 was able to exclude health complications.
Conclusions: In the institutionalised elderly patients, GNRI confirmed its predictive value even for short-term health complications, particularly when death was considered. However, the use of TLC might improve the evaluation of nutritional risk and the identification of patients at risk of infections. Nutrition study should be considered to confirm possible risk reduction.