Association between protein intake and functional capacity in critically ill patients: A retrospective cohort study

Abstract

Background

Intensive care unit (ICU) protein benchmarks are based on mortality and morbidity; whether these targets also support functional recovery is unknown. We assessed whether different protein doses influenced patients' functional capacity, measured by the Chelsea Physical Assessment score (CPAx).

Methods

Single-center retrospective cohort study on ICU survivors with length of stay ≥7 days admitted between October 2014 and September 2020. Eligible patients were divided according to protein intake (g/kg/day): low (<0.8), medium (0.8–1.19), high (1.2–1.5), and very high (>1.5). Protein dose effect on CPAx was assessed at ICU discharge with analysis of covariance adjusting for age, illness severity, hospital length of stay before ICU admission, time to start nutrition support, and mechanical ventilation duration. We also investigated effect modification by energy intake and nutrition status.

Results

Enrolled patients (n = 531) were similar for age, nutrition status, and illness severity across groups. CPAxs were nonlinearly associated with protein doses and similar among low, medium, and very high groups. The CPAx for the high group was statistically different (P = 0.014), indicating that the data of three groups could be pooled. Mean CPAx difference remained statistically significant after adjusting for confounding variables (3.9 ± 1.8, P = 0.029 in the four-group model, and 2.7 ± 0.9, P = 0.003 in the pooled two-group model). Energy intake was equivalent between groups and did not modify CPAx. The high group had superior CPAx in both well-nourished and malnourished patients, indicating nutrition status was not an effect modifier.

Conclusion

Protein dose 1.2–1.5 g/kg/day was associated with superior functional capacity at ICU discharge compared with other doses. Neither energy intake nor nutrition status modified functional capacity across groups; therefore, the results appear to be influenced by 1.2–1.5 g/kg/day.