It is estimated that approximately one in three patients admitted to the Intensive Care Unit (ICU) develops acute kidney injury (AKI) during their stay.1 AKI is described as a multisystemic syndrome associated with high morbidity and mortality; in particular, patients requiring renal replacement therapy (RRT) have a mortality rate close to 50%,2 a figure that, unfortunately, has not decreased substantially in recent years.3

Several risk factors for the development of AKI have been identified, including chronic kidney disease (CKD), diabetes mellitus, arterial hypertension, as well as age and sex, among others.4,5 Furthermore, patients who survive an episode of AKI are at increased risk of progression to CKD and of developing long-term cardiovascular events.6

In the absence of a specific treatment, AKI management has historically focused on managing its complications, primarily those related to volume status, electrolyte disturbances, and uremia.7 However, this approach tends to overlook the nutritional consequences associated with loss of renal function. A significant increase in energy expenditure8 and alterations in protein metabolism that increase the risk of malnutrition9 have been documented. Additionally, up to 34% of patients may present with malnutrition at the time of AKI development, largely attributable to pre-existing comorbidities.10

During AKI, marked muscle mass loss occurs, attributable to multiple pathogenic mechanisms. Among the most relevant are fever and immobility, which are common during ICU stays due to the use of sedation and analgesia, invasive mechanical ventilation, and multiple intravenous accesses that hinder body mobilization.11

The etiologies of AKI are usually associated with high metabolic demand. The most frequent include sepsis, shock, the postoperative period, oncological diseases, rhabdomyolysis, and cardiorenal syndromes.3 These conditions promote the development of uremia, which negatively impacts protein metabolism.12

Pharmacological treatment also contributes to nutritional deterioration in patients with AKI. Some drugs, such as steroids, promote muscle catabolism, while antihyperglycemic therapies, including insulin, can additionally alter metabolic balance.13,14

Although there are no specific reports on the impact of nutritional therapy in patients with AKI, it has been documented that, in critically ill patients, suboptimal provision of energy and protein is associated with functional deterioration at hospital discharge and contributes to the development of ICU-acquired weakness.15 These findings underscore the importance of early and individualized nutritional intervention, combined with early mobilization strategies, which act synergistically.16

Currently, there are no robust nutritional recommendations based on AKI etiology or algorithms adjusted to the patient's baseline nutritional status. Considering the impact of AKI on nutritional status and metabolic demands, the aim of this review is to analyze the available evidence on nutritional therapy in critically ill patients with AKI undergoing some modality of RRT, in order to contribute to a more favorable clinical outcome.

Critical illness, combined with AKI, significantly increases nutritional risk.17,18 One of its most relevant manifestations is protein-energy wasting (PEW), reported in up to 82% of ICU patients.19 This phenomenon is favored by multiple factors, including the presence of comorbidities, inadequate nutritional therapy prescription, metabolic acidosis, fluid overload, loss of macro- and micronutrients during RRT, and prolonged immobilization.19,20

Several nutritional indicators have been associated with increased risk of AKI. Body mass index (BMI) is one of the most studied. A meta-analysis of 16 cohort studies showed that underweight (BMI < 18.5 kg/m2), overweight (BMI 25–29.9 kg/m2), and obesity (BMI ≥ 30 kg/m2) are all associated with an increased risk of AKI, with risk ratios (RR) of 1.08 (95% CI: 1.00–1.17), 1.12 (95% CI: 1.03–1.22), and 1.32 (95% CI: 1.18–1.47), respectively.21 In obesity, this relationship could be explained by alterations in renal hemodynamics.22

Malnutrition has also been identified as an independent risk factor for AKI, assessed using nutritional screening tools. In patients with acute coronary syndrome, a score ≥ 3 on the Nutritional Risk Screening 2002 (NRS-2002) was associated with a higher risk of AKI (OR 1.64, 95% CI: 1.24–2.17) and higher mortality (HR 2.02, 95% CI: 1.49–2.75).23 Similarly, malnutrition assessed using the Controlling Nutritional Status (CONUT) scale increased the risk of contrast-induced AKI (OR 2.04, 95% CI: 1.28–3.38).24 Concordant results have been described in elderly adults undergoing percutaneous coronary intervention using this same tool.25,26

In critically ill elderly populations, high nutritional risk assessed by the Nutrition Risk in Critically Ill (NUTRIC score) was associated with higher AKI incidence.27 In patients undergoing non-cardiac surgery, the Geriatric Nutritional Risk Index (GNRI) showed an increased risk of postoperative AKI (OR 1.88, 95% CI: 1.11–3.20), probably mediated by greater susceptibility to nosocomial infections, as well as alterations in cardiac contractility and intravascular volume that lead to inadequate renal perfusion.28

The relationship between malnutrition and AKI is bidirectional; both conditions potentiate each other. PEW, assessed by Subjective Global Assessment (SGA), was associated with higher mortality in patients with AKI (RR 1.99, 95% CI: 1.36–2.91).19 Severe malnutrition, according to SGA, increases in-hospital mortality (OR 4.4, 95% CI: 1.39–14.03).29

In sepsis-related AKI, a lower GNRI score was associated with higher 28-day mortality.30 Likewise, patients on continuous renal replacement therapy (CRRT) with high nutritional risk according to the Nutritional Risk Index (NRI) showed greater dialysis dependence at 90 days (OR 7.08, 95% CI: 1.10–41.39).31

Body composition analysis has gained growing interest due to its association with relevant clinical outcomes.32 In critically ill patients with AKI, both bioelectrical impedance analysis (BIA) and muscle ultrasound (MUS) have demonstrated adequate reproducibility and clinical utility.

BIA has been established as a useful technique for assessing the degree of overhydration in critically ill patients.33 Overhydration, quantified using various indicators depending on the BIA device used, has been associated with adverse clinical outcomes, including higher mortality and dialysis dependence.34–37 Additionally, some studies have used BIA to guide ultrafiltration volume, reporting significant improvements in overhydration status.38

BIA has shown excellent agreement with techniques considered reference standards, such as dual-energy X-ray absorptiometry (DXA). However, in critically ill patients, it has been documented that intravenous solution administration significantly affects the regression equations used by BIA devices,39 leading to an overestimation of fat-free mass (FFM) proportional to the degree of overhydration, expressed as the extracellular water/total body water (ECW/TBW) ratio.40 Consequently, it is essential to interpret both hydration parameters and FFM estimates cautiously or, alternatively, to apply adjustments for overhydration, such as the equations proposed by Chamney et al.,41 which have been shown to reduce the overestimation of muscle reserves.

Overhydration hinders the quantification of muscle mass; in this context, muscle ultrasound is a promising alternative.42 Measurement of quadriceps muscle thickness has shown good agreement with muscle mass quantification obtained by computed tomography.43 This technique is feasible even in patients undergoing RRT, with no significant differences in measurements before and after treatment, regardless of the ultrafiltrated volume.44

Using MUS, an approximate 15% loss of muscle thickness during the first days of ICU stay has been described, making it a useful tool for monitoring catabolism in this patient group.45 More recently, it has been documented that patients on CRRT show a 10% decrease (95% CI: 3–20%) in rectus femoris thickness and a 19% decrease (95% CI: 12–22%) in cross-sectional area of this muscle during the first seven days of treatment.46

Energy requirements. Indirect calorimetry is considered the reference standard for determining energy expenditure in critically ill patients; however, despite international consensus recommendations,18 its availability in clinical practice remains limited.17 When this technique is unavailable, it is suggested to use oxygen consumption (VO2) measurement via a pulmonary artery catheter or carbon dioxide production (VCO2) derived from the mechanical ventilator.17,18 Although CRRT has been shown to influence CO2 production, this effect generates minimal variation in energy expenditure measurement.47 Furthermore, energy expenditure does not differ significantly according to the RRT modality used.48

In the absence of direct measurements, simplified weight-based equations are recommended, prescribing a range of 20–30 kcal/kg/day. More specifically, 20–25 kcal/kg/day is suggested during the catabolic phase and 25–30 kcal/kg/day during the anabolic phase.49 Energy expenditure can also be estimated using artificial intelligence-based models; however, these still require further validation.50

Several studies have evaluated the accuracy of different predictive equations for energy expenditure; among the most commonly used are the Harris-Benedict, Mifflin–St. Jeor, Penn State University, American College of Chest Physicians, and Faisy equations. These studies have documented poor agreement between predictive models and indirect calorimetry measurements (Table 1), with a systematic tendency to underestimate actual energy requirements.50–53 Providing adequate amounts of energy is essential, as suboptimal intake is associated with greater muscle mass loss, delayed wound healing, and increased infection incidence, among other adverse outcomes. Conversely, excessive energy provision is associated with increased carbon dioxide (CO2) production, hepatic alterations, hyperglycemia, and greater difficulty in weaning from invasive mechanical ventilation.54 To avoid overfeeding, it is important to consider non-nutritional energy sources derived from pharmacological and dialytic therapy.

Moreover, it has been reported that patients with AKI often receive inadequate macronutrient intakes, with lower carbohydrate oxidation compared to prescribed and administered amounts, as well as increased lipid oxidation.55 Although no studies have evaluated different energy doses in AKI patients on RRT, in critically ill patients, hypocaloric nutrition has been shown to reduce the duration of mechanical ventilation (MD –1.85 days, 95% CI –3.44 to –0.27), adverse gastrointestinal events (RR 0.79, 95% CI 0.69–0.90), and mortality risk (RR 0.90, 95% CI 0.81–0.99).56

Non-nutritional energy sources. Critically ill patients undergoing RRT are at persistent risk of overfeeding, related to energy intake from medication regimens (such as dextrose and propofol infusions), from dialysate fluids containing lactate or glucose, and from the energy derived from citrate used as an anticoagulant in CRRT. The calories provided by these molecules are: 3 kcal/g for 2.2% citrate, 3.4 kcal/g for dextrose, and 3.62 kcal/g for lactate.18,57

Several studies have evaluated the absorption of these non-nutritional calories during RRT, showing clinically relevant amounts. In a prospective study of 10 critically ill patients on continuous venovenous hemofiltration (CVVH) anticoagulated with citrate and with dialysate solutions containing dextrose, an absorption of 512 ± 32 kcal was documented.58 Another study in nine critically ill patients on CVVH reported absorptions of 498 (339–681) kcal/day with low-dose citrate, 262 (56–565) kcal/day with high-dose citrate, and a loss of –189 (92–298) kcal/day with low dose without citrate.59 A constant of 13.3 kcal/h during CRRT with citrate anticoagulation has also been proposed.60 Likewise, in a cohort of 45 patients on CVVH, a positive energy balance of 439 ± 124 kcal/day was observed with citrate, and a negative balance of –327 ± 82 kcal/day without its use.61 In high-volume peritoneal dialysis (PD), glucose absorption can reach 35%, equivalent to an intake of up to 900 kcal/day.62

Quantification of these non-nutritional calories is essential when prescribing nutritional therapy (Fig. 1) and avoiding the complications of overfeeding.

Figure 1.

Energy contributions related to RRT. RRT: renal replacement therapy; CRRT: continuous renal replacement therapy; PD: peritoneal dialysis.

Refeeding syndrome prevention. Refeeding syndrome constitutes a persistent risk during the implementation of nutritional therapy in malnourished patients or those with prolonged fasting. It is characterized by electrolyte disturbances, primarily involving phosphorus, potassium, and magnesium, thiamine deficiency, and changes in fluid compartments, which may predispose to fluid overload and potentially fatal complications.17,63

Although modalities such as CVVH and PD may be associated with a higher risk of refeeding syndrome due to increased phosphorus and potassium losses compared to intermittent hemodialysis (IHD),64 there are no specific data on its incidence in critically ill patients on RRT.

The American Society for Parenteral and Enteral Nutrition (ASPEN) recommends initiating nutritional therapy with 100–150 g of dextrose or 10–20 kcal/kg during the first 24 h, gradually increasing calories over the following days. In the presence of low electrolyte levels, the initiation or escalation of feeding should be delayed until correction or supplementation. Additionally, administration of 100 mg of thiamine before initiating nutrition is suggested in cases of severe starvation or suboptimal intake present for more than five to seven days.64

Lipids. These constitute an essential component of energy provision in both enteral nutrition (EN) and parenteral nutrition (PN), as well as being a relevant source of essential fatty acids and fat-soluble vitamins.57 In AKI, alterations in lipolysis occur, characterized by decreased lipoprotein lipase and hepatic lipase activity, as well as reduced clearance of lipid emulsions.18

It has been described that lipid oxidation is increased in these patients and that, as part of nutritional therapy, the amounts of lipids prescribed and administered are often lower than required.55 In this context, a lipid prescription of less than 1.5 g/kg/day has been suggested, considering both nutritional and non-nutritional lipid sources.17,65

Currently, there are insufficient data on the impact of lipid emulsion type on clinical outcomes of critically ill patients with AKI on RRT. Nevertheless, the use of mixed lipid emulsions containing medium-chain triglycerides, olive oil, and fish oil is recommended.66 The European Society for Clinical Nutrition and Metabolism (ESPEN) further suggests the use of emulsions enriched with omega-3 fatty acids.17,67 In cases where triglyceride levels exceed 400 mg/dL, lipid infusion should be discontinued65.

Protein. During RRT, significant amino acid losses occur,68 which vary according to the modality used.69–71 These losses depend on the clearance mechanism: convection, as in CVVH; diffusion, as in IHD; or a combination of both, as in sustained low-efficiency dialysis (SLED). A comparative study reported greater losses with CVVH (18.69 ± 3.04 g), followed by SLED (8.21 ± 4.07 g) and IHD (5.13 ± 3.1 g).69

On the other hand, high-volume PD is associated with protein losses of 4.2 ± 6.1 g/24 h, which increase in the presence of peritonitis.70

These losses justify, at least in part, the increased protein intake in patients on RRT. Therefore, the ESPEN guidelines recommend adjusting protein prescription according to the protein catabolic rate. When urea kinetics cannot be measured to estimate this rate, a protein intake of 1.3–1.5 g/kg/day is suggested for patients on conventional intermittent RRT and 1.5–1.7 g/kg/day during prolonged intermittent or continuous RRT.18

A post hoc analysis of the EFFORT trial in patients who developed AKI compared high versus usual protein intakes (1.5 ± 0.5 g/kg/day versus 0.9 ± 0.3 g/kg/day). Results showed that the higher protein group had a longer hospital stay and higher 60-day mortality (RR 1.4, 95% CI: 1.1–1.8). However, this negative effect was not observed in patients who received RRT.72

Micronutrients. During RRT, micronutrients with a molecular weight below 50,000 daltons can be lost through the effluent,73 negatively impacting their serum levels.68,74,75 The magnitude of these losses depends, in addition to molecular weight, on multiple factors, such as charge, solubility, protein binding, lipophilicity or hydrophilicity, the patient's nutritional status, and the RRT modality used.73

The amounts lost vary according to the RRT technique. In CVVH, zinc and copper losses are greater compared to SLED and IHD, while no differences have been observed in B-vitamin elimination between the different modalities.69 However, according to Fah et al., micronutrient deficiency may differ in patients with high nutritional risk: copper and carnitine were the most affected in those treated with CRRT, while zinc deficiency was more frequent in those who did not receive CRRT.75

Other studies have documented that vitamin C, selenium, and zinc levels are decreased at the onset of CRRT, with significant reductions during and after therapy.76 Similarly, Ostermann et al. observed that more than 30% of patients on CRRT had deficiencies in zinc, iron, selenium, vitamin D3, and vitamin C during the first six days of treatment.74

In addition to elimination by RRT, several factors contribute to the decrease in serum micronutrients, including insufficient nutritional provision, gastrointestinal malabsorption, inflammation, diuretic use, and gastrointestinal losses associated with vomiting or diarrhea.73

Despite documentation of these losses during CRRT, specific micronutrient requirements in this context have not yet been established.68,77 ESPEN has issued guidelines for micronutrient prescription in critically ill patients, which are presented in Table 2.78

Enteral nutrition. This feeding modality has been shown to be a safe strategy in patients with AKI,79 and it is therefore recommended to initiate it within the first 48 h of hospitalization.18 There are no specific recommendations regarding the EN administration modality (intermittent/bolus, cyclic, or continuous).

In a study conducted in patients undergoing CRRT, groups with high and low energy intake through early EN were compared, reporting that the group with lower energy intake showed significant improvements in serum albumin and prealbumin concentrations, in addition to a reduction in ICU length of stay and CRRT duration.80

Regarding formula type, it is suggested not to use specialized renal formulas as first-line therapy instead of standard formulas. Nutritional supplement selection should be based on individual energy and protein requirements. However, in the presence of electrolyte disturbances or positive fluid balances, renal-specific formulas may constitute a more appropriate alternative.18

Parenteral nutrition. In patients with contraindications for the enteral route, parenteral initiation is recommended between three and seven days of hospitalization. This should be early and progressive, especially in subjects with severe PEW considering the risk of developing refeeding syndrome.18

Additionally, parenteral amino acid supplementation can be used to meet the increased protein requirements during RRT.20 When prescribing PN, aspects such as fluid overload and electrolyte disturbances should be considered; however, the composition and volume of PN can be adjusted according to the patient's clinical condition and the type of RRT administered.49

When PN is used for a short period, it is feasible to administer it peripherally, always considering macronutrient concentration and total volume. In patients with fluid overload, fluid restriction is prioritized; in these cases, and to ensure nutritional requirement fulfillment, the central route is preferable whenever the patient’s clinical condition allows.20

RRT exerts a negative impact on nutrient balance in critically ill patients with AKI, contributing to the development of PEW. CRRT is associated with particularly significant amino acid and micronutrient losses.

When implementing and monitoring nutritional therapy, it is essential to consider factors such as the type of RRT used, the patient's comorbidities, the course of critical illness, the availability of feeding access (enteral or parenteral), and the energy intake derived from medications and dialysis or hemofiltration solutions. These elements are key to individualizing nutritional intervention and optimizing clinical outcomes.

AKI in critically ill patients requiring RRT causes increased energy expenditure, protein catabolism, muscle mass loss, and malnutrition.

Tools such as NRS-2002, CONUT, NUTRIC, and the GNRI, along with body composition techniques such as bioelectrical impedance analysis and muscle ultrasound, allow identification of patients at nutritional risk and monitoring of changes in muscle mass and overhydration, facilitating early and personalized interventions.

Critically ill patients on RRT receive additional energy from medications and dialytic solutions, including dextrose infusions, propofol, lactate in dialysate fluids, and citrate used as an anticoagulant. These sources provide significant calories that must be considered in nutritional prescription.

Adequate nutrition, along with early mobilization strategies, contributes to preserving muscle mass, preventing PEW-associated complications, and improving clinical outcomes in critically ill patients with AKI. Evidence is still limited regarding micronutrients and specific recommendations according to RRT modality, underscoring the need for future studies.

All authors (IAOP, FGYE, and JSCI) contributed to the conception, development, and final review of the article.

The authors declare that they did not use artificial intelligence in the writing process.

The authors declare no financial support related to the manuscript.