Additional predictive value of nutritional status in the prognostic assessment of heart failure patients

Nutritional status (NS) is not routinely assessed in HF. We sought to evaluate whether NS may be additive to a comprehensive pre-discharge evaluation based on a clinical score that includes BMI (MAGGIC) and on an index of functional capacity (six minute walking test, 6mWT) in HF patients.

The CONUT (Controlling Nutritional Status) score (including serum albumin level, total cholesterol and lymphocyte count) was computed in 466 consecutive patients (mean age 61 ± 11 years, NYHA class 2.6 ± 0.6, LVEF 34 ± 11%, BMI 27.2 ± 4.5) who had pre-discharge MAGGIC and 6MWT. The endpoint was all-cause mortality. Mild or moderate undernourishment was present in 54% of patients with no differences across BMI strata. The 12-month event rate was 7.7%. Deceased patients had a more compromised NS (CONUT 2.8 ± 1.5 vs 1.7 ± 1.3, p < 0.0001), and a more advanced HF (MAGGIC 28.2 ± 6.0 vs 22.0 ± 6.6, p < 0.0001; 6MWT 311.1 ± 102.2 vs. 408.9 ± 95.9 m, p < 0.0001). The 12-month mortality rate varied from 4% for well-nourished to 11% for undernourished patients (p = 0.008). At univariate analysis, the CONUT was predictive for all-cause mortality with a Hazard Ratio of 1.701 [95% CI 1.363-2.122], p < 0.0001. Multivariable analysis showed that the CONUT significantly added to the combination of MAGGIC and 6MWT and improved predictive discrimination and risk classification (c-index 0.82 [95% CI 0.75-0.88], integrated discrimination improvement 0.028 [95% CI 0.015-0.081]).

In HF patients assessment of NS, significantly improves prediction of 12-month mortality on top of the information provided by clinical evaluation and functional capacity and should be incorporated in the overall assessment of HF patients.