Eight studies (257 participants) were included. An overall pooled analysis was not performed due to the different interventions used and different outcomes measured. Selection bias was not reported (unclear) in six studies and was adequately reported (low) for random sequence generation in two studies. Participant/personnel blinding was adequately reported in one study and unclear in seven. Incomplete outcome reporting bias was low in six studies and high in two. Selective reporting was low in six studies, unclear in one study, and high in one study. No other biases were detected. There was a significant increase in recovery rate for AKI (RR 1.70, 95% CI 1.70 to 2.79) and survival in dialysed patients (RR 3.56, 95% CI 0.97 to 13.08) for intravenous essential L-amino acids (EAA) compared to hypertonic glucose alone. Compared to lower calorie-total parenteral nutrition (TPN), higher calorie-TPN did not improve estimated nitrogen balance, protein catabolic rate, or urea generation rate; but increased serum triglycerides, glucose, insulin need and nutritional fluid administration. There was no difference between groups in estimated nitrogen balance, but there were differences between urea nitrogen appearance (MD 0.98, 95% CI 0.25 to 1.71) and net protein utilisation (MD 21.50%, 95% CI 0.39 to 42.61). Urea nitrogen appearance was lower in the low nitrogen intake group than in the high nitrogen intake group. There was no significant difference in death between EAA and general amino acids (GAA) (RR 1.52, 95% CI 0.63 to 3.68). High dose amino acids did not improve cumulative water excretion, furosemide requirement, nitrogen balance or death compared to normal dose amino acids. Glucose+EAA+histidin had better nitrogen balance than glucose+GAA; glucose+nitrogen+fat significantly increased serum creatinine compared with glucose+GAA; glucose+EAA+histidin significantly improved nitrogen balance, U/P urea and serum creatinine, but increased plasma urea compared to glucose+nitrogen+fat.