Sodium supplementation strategies in preterm infants: effects on growth, electrolyte balance, and exploratory evidence on neurodevelopment — a systematic review and meta-analysis Authors: Yasmin Ahmed Hassan Mohamed Gad, Abeer Mohamed Hekal, Mohmmed Mostafa Ali Kashef, Muhamed Rafi, Fida Rafi, Mohamed Hafez Habeeb, Dina Mohamed Hafez Published: 2026 ABSTRACT This study evaluated the effects of sodium supplementation on growth, electrolyte balance, and neurodevelopmental outcomes in preterm infants. A systematic review and meta-analysis of published studies were conducted using data from PubMed, Embase, Scopus, and Google Scholar, including randomized controlled trials and observational studies involving infants born before 37 weeks of gestation. Twelve studies comprising more than 2,800 participants were analysed using random-effects models. Sodium supplementation was associated with improved weight gain and a significant reduction in the risk of hyponatremia (RR = 0.28, 95% CI: 0.08–0.96). Pooled incidence estimates indicated hypernatremia (32%) and intraventricular hemorrhage (29%) among supplemented cohorts; however, substantial heterogeneity was observed, and comparative risk estimates were limited. Only one study reported long-term neurodevelopmental follow-up, suggesting potential benefit; however, the available evidence remains limited and hypothesis-generating.These findings suggest that sodium supplementation may support growth and is associated with a lower risk of hyponatremia in preterm infants; however, hypernatremia and other morbidity estimates reported here are largely descriptive frequencies within supplemented cohorts rather than comparative treatment effects, and careful, individualized therapy with close monitoring is needed to minimize potential risks.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s40795-026-01464-4. Introduction Preterm birth, defined as delivery before 37 weeks of gestation, remains a significant global health concern, contributing to increased neonatal morbidity and mortality. Among the myriad challenges faced by preterm infants, achieving optimal postnatal growth and neurodevelopment is paramount. Despite advancements in neonatal care, postnatal growth failure persists as a common issue, often linked to long-term neurodevelopmental impairments. Emerging evidence suggests that sodium homeostasis plays a crucial role in the growth and development of preterm infants, yet it remains an underappreciated aspect of neonatal nutrition [1]. Sodium is the primary extracellular cation, essential for maintaining fluid balance, nerve impulse transmission, and muscle function. In preterm infants, the kidneys' immature tubular function leads to increased sodium losses, making them susceptible to sodium deficiency. Studies have indicated that inadequate sodium intake during the early postnatal period may contribute to growth failure in this vulnerable population. Animal models have long demonstrated the importance of sodium in growth, and recent human studies are beginning to elucidate similar associations. For instance, a review by Araya et al. highlights the relationship between sodium homeostasis and growth in preterm infants, emphasizing the need for increased awareness and potential benefits of dietary sodium supplementation [1]. Beyond somatic growth, sodium balance may also influence neurodevelopmental outcomes. The developing brain is highly sensitive to nutritional deficits, and sodium plays a role in neuronal function and cerebral blood flow regulation. A study by Skinner and Narchi discusses the direct role of nutrition, including sodium, on neurodevelopment, underscoring the importance of early nutritional interventions to support brain development in preterm infants. Furthermore, research by Segar et al. implemented a clinical practice algorithm to identify sodium deficiency in preterm infants, guiding supplementation based on urine sodium concentrations. Their findings suggest that such targeted approaches can improve growth outcomes without adversely affecting respiratory status [2, 3]. Despite these insights, there remains a lack of consensus on optimal sodium supplementation strategies for preterm infants. Variability in clinical practice, small trial sample sizes, and limited long-term follow-up contribute to this uncertainty, as highlighted in a systematic review assessing the quality and consistency of nutrition guidelines for preterm infants [4]. The most direct evidence addressing this uncertainty comes from randomized controlled trials, several of which were synthesized in the 2023 Cochrane review of higher versus lower sodium intake in preterm infants [5], which found that higher sodium intake reduced postnatal weight loss and hyponatremia while increasing hypernatremia risk, based on a small number of heterogeneous trials with predominantly unclear or high risk of bias. That review, restricted to randomized evidence, could not address several questions of ongoing clinical relevance: how supplementation practices and their electrolyte and morbidity associations manifest in contemporary, real-world neonatal cohorts; whether gestational age or birth weight modify electrolyte risk; and how individualized, biomarker-guided dosing strategies compare with fixed-dose regimens, an approach increasingly used in clinical practice but rarely evaluated in trials. The present review is intended to complement, rather than replace, the randomized-evidence synthesis provided by the 2023 Cochrane review. We incorporate more recent observational cohort studies to characterize real-world supplementation practice and associated electrolyte and morbidity patterns, and we present exploratory meta-regression and narrative analyses of gestational-age and dosing-strategy associations, including a preliminary, single-cohort comparison of urine-sodium-guided dosing, that were outside the scope of the Cochrane review. These observational and exploratory findings are intended to be hypothesis-generating rather than confirmatory, and are reported alongside, rather than in place of, the existing randomized-trial evidence base. This systematic review therefore aims to synthesize current evidence, both randomized and observational, on sodium supplementation in preterm infants, clearly distinguishing comparative treatment effects from single-arm descriptive estimates, in order to characterize growth, electrolyte, and morbidity associations and to identify priorities for future randomized research on individualized sodium dosing strategies. Methods This systematic review and meta-analysis were conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines and the methodological standards outlined in the Cochrane Handbook for Systematic Reviews of Interventions. Study measures The primary study measure was the estimation of the pooled effects of sodium supplementation on growth outcomes, electrolyte balance, and neurodevelopmental outcomes in preterm infants born at less than 37 weeks of gestation. Sodium supplementation was defined as the administration of sodium chloride or sodium acetate supplements in addition to standard nutritional support, as opposed to standard care without supplementation. The secondary study measures included: (1) identification of the determinants of hyponatremia and hypernatremia incidence; (2) assessment of neonatal morbidities including intraventricular haemorrhage (IVH), necrotizing enterocolitis (NEC), bronchopulmonary dysplasia (BPD), sepsis, and mortality; and (3) evaluation of long-term neurodevelopmental outcomes including motor function, cognitive performance, memory, language skills, and behavioural outcomes in sodium-supplemented preterm infants. Data sources This meta-analysis was guided by the 2020 Cochrane Handbook of Systematic Review and Meta-Analysis, with respect to the PRISMA checklist. To assess the effects of sodium supplementation in preterm infants, the search process started on 15 May 2024 and concluded on 31 May 2024 for studies that had been published up until 31 May 2024. Published studies in the following databases were included: Google Scholar (n = 511), PubMed (n = 13), Scopus (n = 45), and Embase (n = 22). Search terms were determined based on consultation with neonatal nutrition experts and included combinations of "sodium supplementation," "salt," "Na +," "preterm," "premature," "neonate," "infant," "growth," "neurodevelopment," "hyponatremia," and "hypernatremia." No restrictions were applied regarding publication date or language. Google Scholar was used as a supplementary search source rather than as a substitute for established bibliographic databases. Data extraction and study selection All studies investigating sodium supplementation in preterm infants born at less than 37 weeks of gestation were included with no language restriction. Eligible studies reported on at least one of the following outcomes: weight gain, growth parameters (linear or head circumference), serum or urine sodium levels, and neonatal morbidities such as IVH, NEC, BPD, sepsis, mortality, or neurodevelopmental outcomes. We included randomized controlled trials (RCTs), cohort studies (prospective or retrospective), and observational studies with comparator groups. Studies excluded focused exclusively on term infants, were conference abstracts without full-text availability, or were review articles, editorials, or case reports. The PRISMA flow chart for the different steps of the current meta-analysis is depicted in Fig. 1. All articles were imported into EndNote X8 for Windows (Thomson Research Soft, Stanford, CA, USA), to detect and remove duplicates. After the citation's exportation to a data extraction sheet containing the authors' names, publication year, journal name, study design, and outcome measures, the authors screened both the title and the abstract. This was followed by full-text screening to identify the eligible articles. Screening was performed independently by two authors (Y.A.H.M.G and AMH). The third author (D.M.H) resolved any disagreement. Furthermore, manual search for eligible citations was conducted through careful examination of the references of the included studies. All included articles were extracted to a standardized data extraction form with the following predefined variables: first author's name, publication year, study design, sample size, gestational age at birth, birth weight, sodium dose and route of administration, duration of supplementation, primary and secondary outcomes, and follow-up duration. Data were extracted independently by two reviewers and cross-verified. Any discrepancies were resolved through discussion. Where data were unclear or incomplete, we attempted to derive estimates from tables or figures; however, no authors were contacted for additional information.Fig. 1PRISMA flow diagram of study screening and selection PRISMA flow diagram of study screening and selection Data items and effect measures The primary outcomes of interest included, serum sodium concentration, incidence of hyponatremia and hypernatremia, and neurodevelopmental outcomes such as motor function, cognitive performance, and behaviour. Secondary outcomes included weight gain (both absolute and z-score), the incidence of neonatal morbidities: IVH (grades III–IV), NEC, BPD, sepsis, and mortality. Additional variables extracted included the timing of sodium initiation, cumulative dose, route of administration (enteral vs. parenteral), and any protocol for dose adjustment (e.g., based on urinary sodium levels). For dichotomous outcomes, pooled proportions with 95% confidence intervals (CIs) were calculated to describe outcome frequency within supplemented cohorts and do not represent comparative treatment effects. For outcomes with comparator groups, pooled relative risks (RRs) were calculated. All statistical syntheses were conducted using a random-effects model to account for anticipated clinical and methodological heterogeneity. Meta-analyses were performed using the meta package in R (version 4.4.0). Due to expected variability in sodium dose, timing of initiation, route of administration, and study design, pooled analyses were conducted using random-effects models and interpreted cautiously as exploratory syntheses rather than definitive causal estimates. Because randomized evidence was limited, both randomized and non-randomized studies were retained in the evidence synthesis. However, randomized evidence was prioritized when interpreting comparative treatment effects, whereas observational evidence was considered complementary and interpreted cautiously because of the potential for residual confounding. For outcomes with limited or unavailable randomized evidence, observational studies were retained to provide additional clinically relevant information. Pooled single-arm proportions that combined study designs were interpreted as descriptive estimates of event frequency within sodium-supplemented cohorts rather than causal treatment effects. Investigations of heterogeneity Cochrane's Q test (I2) was used to assess and measure heterogeneity between studies. DerSimonian and Laird random-effects models were applied to the pooled outcomes. Given anticipated clinical and methodological variability, pooled estimates were interpreted as exploratory rather than confirmatory treatment effects. We conducted subgroup analyses based on gestational age and birth weight. To evaluate the robustness of the results, leave-one-out sensitivity analyses were performed for key outcomes. Formal stratification of forest plots by study design was not undertaken because the number of studies contributing to several outcomes was small and, in multiple analyses, one design category was represented by only one or very few studies. Under these circumstances, design-specific pooled estimates were considered statistically unstable and potentially misleading. Publication bias Publication bias was assessed by visual inspection of the funnel plot and statistically by Egger's regression test when ten or more studies were available for analysis. However, due to the limited number of eligible studies per outcome, formal statistical testing was not performed for most outcomes. Quality assessment Quality assessment (QA) was based on tools appropriate for the study design. For randomized controlled trials, the Cochrane Risk of Bias 2 (RoB 2) tool was employed, which evaluates five domains: the randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome, and selection of the reported result. For cohort and observational studies, we used the Newcastle–Ottawa Scale (NOS), which examines bias across three broad perspectives: selection of study groups, comparability, and outcome assessment. The quality of studies was either very good (9–10 points), good (7–8 points), satisfactory (5–6 points), or unsatisfactory (0–4 points). The assessment was performed by two independent reviewers (Y.A.H.M.G. and A.M.H) and further checked by the senior author (D.M.H). No automation tools were used during the risk of bias evaluation. The overall certainty of the evidence was not formally graded using GRADE due to the heterogeneity in study designs and outcome definitions; however, findings were interpreted in light of study quality and consistency. Results Study selection Of 591 initial records, 73 duplicates were removed. Title and abstract screening of 518 records excluded 402 irrelevant studies. Full-text assessment of 116 articles excluded 104 for not reporting relevant outcomes (n = 14), being conference abstracts/editorials (n = 27), review articles (n = 18), or being outside the neonatal surgical domain (n = 15). Twelve studies met the eligibility criteria for inclusion, as shown in the PRISMA 2020 flow diagram (Fig. 1). Study characteristics The included studies encompassed 12 investigations with a combined sample size exceeding 2,800 preterm infants (< 37 weeks gestation). Study designs included randomized controlled trials (n = 3), cohort studies (n = 5), and prospective or retrospective observational studies (n = 4). Sodium supplementation dosages ranged from 3 to 10.9 mmol/kg/day, with administration routes including both enteral and parenteral modalities. Key study characteristics are presented in Table 1.Table 1Baseline Characteristics and Clinical Outcomes of Studies Evaluating Sodium Supplementation in Preterm InfantsStudyDesignSample SizeGroupsGA at BirthBirth WeightNa DoseRouteSerum NaUrine NaHartnoll et al., 2000 [6]RCT46Early Na + (n = 24) vs Delayed (n = 22)28 ± 0 vs 27 ± 3 wks945 g vs 946 g4 mmol/kg/dayParenteralHyponatremia: 12 vs 14; Hypernatremia: 15 vs 11NA Lee et al., 2015 [7]Retrospective cohort169Severe IVH (n = 32) vs Control (n = 137)25.4 vs 27.9 wks714 g vs 785 g4.8 ± 3.4 vs 2.1 ± 1.3 mEq/kg/dParenteral/Enteral144.7 max (Severe); 142.2 max (Control)NAModi & Hutton, 1990 [8]Prospective observational26All ventilated infants < 34w< 34 wks1170 g4 mmol/kg/dayParenteral139–140 mmol/L (resp. improvement)NA Eibensteiner et al., 2020 [9]Retrospective90All ELBW < 28wks, < 1000 g24 + 6 wks718 g10.2 mmol/kg/dayIV, fluids, antibioticsHypernatremia: 92% W1, 47% W2NA Segar et al., 2018 [2]Cohort comparison902014–15 (n = 50) vs 2016 (n = 40)27.8 vs 28.2 wks1090 g vs 1173 g3.2–4.4 mEq/kg/dayIndividualized NaCl/acetateNo hypernatremia > 14458% at 2 wks, 75% any time in 2016 groupVenkatraman et al., 2023 [10]RCT30Placebo vs Na + (n = 15 each)30.6 wks1102 g vs 1061 g0.5 mEq/mL 3% NaCl at D7Oral/IVD14; D35NA Späth et al., 2017 [11]Observational cohort592All < 27wks25.3 ± 1.1 wks765 g0–10.9 mmol/kg/dayParenteral/EnteralHypernatremia: 50%, Hyponatremia W2: 79%NA Isemann et al., 2016 [12]RCT, masked, placebo-controlled53Na + vs Placebo28.4 (24–31.9)1080 (551–1669)4Oral/IV140.4 vs 135.5NAAl-Dahhan et al., 2002 [13]Prospective cohort (historical controls)46Na + vs No Na31.3 (25–33) vs 30.2 (24.5–33)1538 vs 14734–5Oral141.5 vs 135.6NASkillman et al., 2012 [14]Pre-post observational study33Pre vs Post Na + 28.4 ± 1.31074 ± 1824.2 ± 1.1OralIncreased post Na + Increased after Na + Stalter et al., 2024 [15]Prospective GA-stratified cohort821Na + across GA strata26–29 vs 30–33936 (26–29 wks), 1411 (30–33 wks)Not reportedNot specifiedNot specifiedNA Petersen et al., 2024 [16]Retrospective cohort821Na + vs No Na26.5 ± 1.9 vs 29.4 ± 1.6849.9 ± 242.1 vs 1178.7 ± 219.54.2 ± 2.6Oral135 to 137 (p < 0.001) < 40 in 42% pre; increased in 81% post Table 1Baseline Characteristics and Clinical Outcomes of Studies Evaluating Sodium Supplementation in Preterm InfantsStudyDesignSample SizeGroupsGA at BirthBirth WeightNa DoseRouteSerum NaUrine NaHartnoll et al., 2000 [6]RCT46Early Na + (n = 24) vs Delayed (n = 22)28 ± 0 vs 27 ± 3 wks945 g vs 946 g4 mmol/kg/dayParenteralHyponatremia: 12 vs 14; Hypernatremia: 15 vs 11NA Lee et al., 2015 [7]Retrospective cohort169Severe IVH (n = 32) vs Control (n = 137)25.4 vs 27.9 wks714 g vs 785 g4.8 ± 3.4 vs 2.1 ± 1.3 mEq/kg/dParenteral/Enteral144.7 max (Severe); 142.2 max (Control)NAModi & Hutton, 1990 [8]Prospective observational26All ventilated infants < 34w< 34 wks1170 g4 mmol/kg/dayParenteral139–140 mmol/L (resp. improvement)NA Eibensteiner et al., 2020 [9]Retrospective90All ELBW < 28wks, < 1000 g24 + 6 wks718 g10.2 mmol/kg/dayIV, fluids, antibioticsHypernatremia: 92% W1, 47% W2NA Segar et al., 2018 [2]Cohort comparison902014–15 (n = 50) vs 2016 (n = 40)27.8 vs 28.2 wks1090 g vs 1173 g3.2–4.4 mEq/kg/dayIndividualized NaCl/acetateNo hypernatremia > 14458% at 2 wks, 75% any time in 2016 groupVenkatraman et al., 2023 [10]RCT30Placebo vs Na + (n = 15 each)30.6 wks1102 g vs 1061 g0.5 mEq/mL 3% NaCl at D7Oral/IVD14; D35NA Späth et al., 2017 [11]Observational cohort592All < 27wks25.3 ± 1.1 wks765 g0–10.9 mmol/kg/dayParenteral/EnteralHypernatremia: 50%, Hyponatremia W2: 79%NA Isemann et al., 2016 [12]RCT, masked, placebo-controlled53Na + vs Placebo28.4 (24–31.9)1080 (551–1669)4Oral/IV140.4 vs 135.5NAAl-Dahhan et al., 2002 [13]Prospective cohort (historical controls)46Na + vs No Na31.3 (25–33) vs 30.2 (24.5–33)1538 vs 14734–5Oral141.5 vs 135.6NASkillman et al., 2012 [14]Pre-post observational study33Pre vs Post Na + 28.4 ± 1.31074 ± 1824.2 ± 1.1OralIncreased post Na + Increased after Na + Stalter et al., 2024 [15]Prospective GA-stratified cohort821Na + across GA strata26–29 vs 30–33936 (26–29 wks), 1411 (30–33 wks)Not reportedNot specifiedNot specifiedNA Petersen et al., 2024 [16]Retrospective cohort821Na + vs No Na26.5 ± 1.9 vs 29.4 ± 1.6849.9 ± 242.1 vs 1178.7 ± 219.54.2 ± 2.6Oral135 to 137 (p < 0.001) < 40 in 42% pre; increased in 81% post Baseline Characteristics and Clinical Outcomes of Studies Evaluating Sodium Supplementation in Preterm Infants Overview of included studies This review analysed 12 studies exploring the impact of sodium supplementation in preterm infants. The studies varied in design and size, including RCTs (Isemann et al., 2016; Hartnoll et al., 2000; Venkatraman et al., 2023) [6, 10, 12], retrospective cohort studies (Petersen et al., 2024; Lee et al., 2015; Segar et al., 2018) [2, 7, 16], and observational designs (Eibensteiner et al., 2020; Modi et al., 1990; Skillman et al., 2012; Stalter et al., 2024; Aldahhan et al., 2002; Späth et al., 2017) [8, 9, 11, 13–15]. Sample sizes ranged widely from just 30 infants in small pilot trials to over 800 in large-scale cohort studies. Most studies administered sodium in the form of sodium chloride or acetate, either orally or intravenously, at doses between 3 to 5 mEq/kg/day. Supplementation typically began within the first week of life, usually between day 2 and day 7. Sodium supplementation and growth outcomes Weight gain and percentile maintenance Growth promotion was a consistent benefit observed in supplemented infants. Isemann et al. (2016) demonstrated that infants receiving 4 mEq/kg/day sodium maintained their weight-for-age percentile at 6 weeks in 79% of cases versus 13% in the placebo group (p = 0.002), with significantly higher weekly weight gain (p = 0.012) [12]. Similarly, Venkatraman et al. (2023) found that infants supplemented from day 4 achieved higher weights at day 35 (1283 g vs 1047 g), though the difference approached but did not reach statistical significance (p = 0.059) [10]. Segar et al. (2018) implemented a clinical algorithm based on urinary sodium and found significantly higher weight z-score change from 2 to 8 weeks in the 2016 cohort (0.32 ± 0.05 vs −0.01 ± 0.08, p < 0.01) [2]. Linear and head growth While weight gain showed a clear response to sodium, length and head circumference didn't follow the same pattern. Both Isemann et al. (2016) and Petersen et al. (2024) reported no significant differences in these measures between infants who received sodium and those who didn't [12, 16]. Serum and urinary sodium concentrations Serum sodium response All intervention groups exhibited higher serum sodium levels following supplementation. Isemann et al. (2016) reported significantly greater serum sodium concentrations at week 3 in the intervention group compared to placebo (p < 0.05), with similar post-intervention values [12]. Venkatraman et al. observed statistically significant increases in serum sodium at day 7 and day 14 (p < 0.005) [10]. Urine sodium monitoring and clinical utility Urine sodium (< 40 mEq/L) was used in Segar et al. [2] to guide initiation and titration of supplementation, leading to improved sodium intake and weight outcomes without increased risk of hypertension or hypernatremia. Petersen et al. (2024) also analyzed urinary sodium concentrations and reported improved sodium balance following supplementation [12, 16]. Neonatal morbidities and mortality Bronchopulmonary Dysplasia (BPD) Evidence on BPD is mixed. Petersen et al. reported a significantly higher incidence of BPD in supplemented infants (Grade 2/3 BPD in 47.1% vs 11.8%, p < 0.001). Conversely, Isemann and Hartnoll found no increase in oxygen dependency at 36 weeks PMA in supplemented groups [6, 12, 16]. Neurodevelopmental outcomes Studies on sodium supplementation in preterm infants reveals a complex balance between benefit and risk, with most studies focusing on short-term metabolic outcomes rather than long-term neurodevelopment. Isemann et al. (2016) and Hartnoll et al. (2000) acknowledged neurodevelopmental implications but provided no formal cognitive follow-up data. Critically, safety concerns are highlighted by retrospective analyses from Eibensteiner et al. (2020) and Lee et al. (2015) [7], which linked higher early sodium intake to an increased risk of severe intraventricular hemorrhage (IVH). Conversely, Segar et al. (2018) demonstrated that a guided supplementation protocol could minimize electrolyte disturbances, thereby reducing iatrogenic risks [2, 6, 9, 12]. The only long-term neurodevelopmental follow-up data available come from Al-Dahhan et al. (2002). In their prospective study with historical, non-randomized controls, preterm infants who received sodium supplementation were followed up at 10–13 years of age. This study reported improved motor and cognitive outcomes in supplemented infants; however, as this finding is based on a single historical cohort with a historical, non-randomized comparator group, it should be interpreted cautiously and considered hypothesis-generating [13]. Meta-analysis Where randomized and observational studies contributed to the same synthesis, study design was considered explicitly in the interpretation. Mixed-design pooled estimates, particularly single-arm proportions, were treated as descriptive and exploratory rather than as causal treatment effects. Hypernatremia A meta-analysis of six studies evaluated the incidence of hypernatremia in preterm infants receiving sodium supplementation (Fig. 2A). The pooled proportion of hypernatremia was 32% (95% CI: 15% to 55%). Statistical heterogeneity was substantial, with I2 = 97.6%, and a significant Chi2 test (χ2 = 205.24, df = 5, p < 0.0001). Given the very high heterogeneity, these pooled estimates should be interpreted cautiously and primarily reflect variability in clinical practice and supplementation strategies rather than a uniform treatment effect.Fig. 2Forest Plots of the Pooled estimates of A Hypernatremia Proportion, B RR for Hyponatremia in Sodium-Supplemented Versus Control Groups in Preterm Infants Among, C Hyponatremia Proportion Forest Plots of the Pooled estimates of A Hypernatremia Proportion, B RR for Hyponatremia in Sodium-Supplemented Versus Control Groups in Preterm Infants Among, C Hyponatremia Proportion Figure S1 presents the funnel plot assessing publication bias for the meta-analysis of hypernatremia proportions and demonstrates visual asymmetry. Despite subgroup analyses by birth weight and gestational age, substantial heterogeneity persisted across all subgroups (I2 ≥ 79%). Exploratory meta-regression suggested gestational age as a potential moderator of hypernatremia risk, with each additional week associated with a 44% lower odds of hypernatremia (OR = 0.56, 95% CI: 0.33–0.95, p = 0.03); birth weight was not significantly associated with hypernatremia risk (p = 0.70). Given the small number of studies contributing to this analysis (k = 6) and the study-level (ecological) nature of the covariate, this finding should be regarded as hypothesis-generating rather than a confirmed independent predictor effect (Table 2).Table 2Meta-Regression of Hypernatremia Risk on Birth Weight and Gestational AgeMeta-Regression Analysisk (studies)Pooled proportion (95% CI)I2 (%)—6——BW (per 100 g increase)—b = − 0.124 (p = 0.70)—GA (per 1 week increase)—OR = 0.56 (95% CI: 0.33–0.95), p = 0.03—Note. k = number of studies; CI = confidence interval; I2 = percentage of total variability attributable to heterogeneity Table 2Meta-Regression of Hypernatremia Risk on Birth Weight and Gestational AgeMeta-Regression Analysisk (studies)Pooled proportion (95% CI)I2 (%)—6——BW (per 100 g increase)—b = − 0.124 (p = 0.70)—GA (per 1 week increase)—OR = 0.56 (95% CI: 0.33–0.95), p = 0.03—Note. k = number of studies; CI = confidence interval; I2 = percentage of total variability attributable to heterogeneity Meta-Regression of Hypernatremia Risk on Birth Weight and Gestational Age Note. k = number of studies; CI = confidence interval; I2 = percentage of total variability attributable to heterogeneity Hyponatremia comparison For comparative analysis of hyponatremia incidence between preterm infants who received sodium supplementation and those who did not, the pooled analysis across four studies demonstrates a significant reduction in the risk of hyponatremia in the supplemented group, with a pooled relative risk RR of 0.28 (95% CI: 0.08 to 0.96) and low heterogeneity (I2 = 0%), indicating a 72% relative risk reduction (Fig. 2B). Hyponatremia pooled proportion The pooled proportion of hyponatremia among preterm infants who received sodium supplementation across seven studies (Fig. 2C) revealed a pooled proportion of 28% (95% CI: 10% to 59%). However, the results are marked by substantial heterogeneity (I2 = 92.2%). The leave-one-out sensitivity analysis revealed reducing heterogeneity (I2) from 92.2% to approximately 80% upon omitting Spath et al. 2017 (Supplementary Table S3). Exploratory subgroup and meta-regression analyses suggested that both lower birth weight and younger gestational age are associated with an increased risk of hyponatremia in preterm infants, with each 100-g increase in birth weight and each additional week of gestation associated with 64% and 49% lower odds of hyponatremia, respectively (p < 0.001 for both). Given the observational, study-level nature of this analysis, these associations should be interpreted as hypothesis-generating rather than confirmatory (Supplementary Table S4). Intraventricular Hemorrhage (IVH) Figure 3A presents a forest plot summarizing the pooled proportion of IVH among preterm infants who received sodium supplementation. The meta-analysis includes three studies, with a pooled IVH incidence of 29% (95% CI: 22% to 38%; I2 = 7.8%). This represents the observed incidence within supplemented cohorts. No pooled comparative risk estimates versus non-supplemented controls were available; therefore, the observed incidence cannot be interpreted as an increased risk attributable to supplementation.Fig. 3Forest Plot of the Pooled Proportion of IVH, Mortality, NEC, Sepsis Among Sodium-Supplemented Preterm Infants Forest Plot of the Pooled Proportion of IVH, Mortality, NEC, Sepsis Among Sodium-Supplemented Preterm Infants Mortality The forest plot of pooled mortality among preterm infants receiving sodium supplementation, based on six studies (Fig. 3B), yielded a pooled mortality estimate of 19% (95% CI: 10%–34%) with moderate heterogeneity (I2 = 65.9%). Mortality rates varied widely across studies, from 0% in Isemann et al., 2016 to 41% in the IVF group of Lee et al., 2015. The pooled mortality estimate represents the observed incidence within supplemented cohorts only. No comparative pooled data versus non-supplemented controls were available for this outcome; therefore, these rates cannot be attributed to the effect of sodium supplementation and primarily reflect the underlying baseline risk of the studied populations, which varied substantially by gestational age and birth weight. Subgroup analysis revealed a significant mortality gradient by birth weight (p = 0.007), with a pooled mortality of 17% in infants under 1000 g compared to no deaths in heavier infants (Supplementary S9). A similar trend for gestational age did not reach statistical significance (p = 0.23). (Supplementary S10). Exploratory univariate meta-regressions suggested an association between higher birth weight and lower mortality odds (p = 0.014), while greater gestational age showed a possible protective trend (p = 0.051). Given the small number of contributing studies and study-level covariates, these findings should be considered hypothesis-generating rather than confirmatory. (Supplementary Table S11). Necrotizing Enterocolitis (NEC) The meta-analysis of 854 preterm infants found a pooled NEC incidence of 3% (95% CI: 0.01–0.09), with significant heterogeneity (I2 = 81.5%, p < 0.0001). Subgroup analysis revealed a statistically significant decrease in NEC with higher gestational age (p = 0.001), with rates falling from 12% in infants < 26 weeks to 0% in those 30–33 weeks. While a trend showed higher NEC incidence in infants with birth weight < 1000 g (4%) versus ≥ 1000 g (1%), this subgroup difference was not statistically significant (p = 0.186). (Fig. 3C) As with mortality, the pooled NEC incidence reflects rates observed within supplemented cohorts and does not represent a comparative treatment effect of sodium supplementation. The subgroup gradient by gestational age reflects well-established baseline NEC risk differences rather than a sodium-specific effect, and should be interpreted accordingly. Sepsis pooled estimate In Fig. 3D, the forest plot presents the pooled incidence of sepsis among preterm infants who received sodium supplementation, incorporating data from four studies. The meta-analysis yields a pooled proportion of 6% (95% CI: 2% to 15%). The individual study estimates ranged from 3% in Stalter et al., 2024 (GA 26–29) to 13% in Venkatraman et al., 2023. Despite the overall low pooled estimate, the analysis revealed substantial heterogeneity (I2 = 79.7%). The highest contribution to the pooled estimate came from Stalter et al., 2024 (GA 30–33), accounting for approximately 35% of the weight due to its large sample size. Sepsis comparison In Fig. 4, the forest plot compares the risk of sepsis between preterm infants who received sodium supplementation and those who did not. Based on two studies, the pooled analysis shows a significant reduction in sepsis risk in the supplemented group, with a pooled RR of 0.18 (95% CI: 0.04 to 0.85). Individually, Venkatraman et al., 2023 reported a strong protective effect (RR = 0.17), while Isemann et al., 2016 also showed a favorable trend (RR = 0.22), although with a wider confidence interval crossing 1. The absence of heterogeneity (I2 = 0%) supports consistency between the studies.Fig. 4Forest Plot of the RR for Sepsis in Sodium-Supplemented Versus Control Groups in Preterm Infants Forest Plot of the RR for Sepsis in Sodium-Supplemented Versus Control Groups in Preterm Infants Serum sodium concentration post-supplement intake Forest plot presents a meta-analysis comparing serum sodium concentrations between preterm infants who received sodium supplementation and those who did not (Fig. 5). The pooled mean difference across three studies is –0.13 mmol/L (95% CI: –6.42 to 6.17) with high heterogeneity (I2 = 83.4%), indicating no statistically significant difference in serum sodium levels between the two groups. This heterogeneity is evident in the direction and magnitude of effect: Petersen et al., 2024 and Isemann et al., 2016 report lower sodium levels in the intervention group (–1.50 and –1.40 mmol/L, respectively), whereas Al-Dahan et al., 2002 reports a significantly higher level in the supplemented group (+ 3.10 mmol/L).Fig. 5Forest Plot of the Mean Difference in Serum Sodium Concentration Between Sodium-Supplemented and Control Preterm Infants Forest Plot of the Mean Difference in Serum Sodium Concentration Between Sodium-Supplemented and Control Preterm Infants Quality assessment of included cohort studies Table 3 provides the Newcastle–Ottawa Scale (NOS) assessment of the quality of cohort studies. Most studies (7/9) were rated as high quality, demonstrating strong selection criteria, adequate adjustment for confounders, and reliable outcome assessment. The remaining studies [13, 14] had moderate quality, primarily due to weaker comparability adjustments or less rigorous outcome follow-up.Table 3Newcastle–Ottawa Scale (NOS) Quality Assessment of Included Cohort StudiesStudySelection (max 4)Comparability (max 2)Outcome (max 3)Total Score (max 9)QualitySegar et al. (2018) [2]4239HighSkillman et al. (2012) [14]3126ModerateSpäth et al. (2017) [11]4239HighAl-Dahan et al. (2002) [13]3126ModerateEibensteiner et al. (2020) [9]3238HighLee et al. (2015) [7]3238HighModi et al. (1990) [8]4239HighPetersen et al. (2024) [16]4239HighStalter et al. (2024) [15]4239High Table 3Newcastle–Ottawa Scale (NOS) Quality Assessment of Included Cohort StudiesStudySelection (max 4)Comparability (max 2)Outcome (max 3)Total Score (max 9)QualitySegar et al. (2018) [2]4239HighSkillman et al. (2012) [14]3126ModerateSpäth et al. (2017) [11]4239HighAl-Dahan et al. (2002) [13]3126ModerateEibensteiner et al. (2020) [9]3238HighLee et al. (2015) [7]3238HighModi et al. (1990) [8]4239HighPetersen et al. (2024) [16]4239HighStalter et al. (2024) [15]4239High Newcastle–Ottawa Scale (NOS) Quality Assessment of Included Cohort Studies Assessment for included randomized controlled trials Risk of bias was assessed using the Cochrane Risk of Bias 2 (RoB 2) tool for randomized controlled trials. Two studies [6, 10] raised some concerns overall, primarily due to potential deviations from intended interventions and selective reporting. The study by Isemann et al. [12] was judged to have a high overall risk of bias due to concerns regarding selective reporting, despite otherwise low risk in most domains (Table 4).Table 4Cochrane Risk of Bias 2 (RoB 2) tool for randomized controlled trialsStudyRandomization ProcessDeviations from Intended InterventionsMissing Outcome DataMeasurement of OutcomeSelection of Reported ResultOverall Risk of BiasVenkatraman et al. (2023) [10]Low RiskSome ConcernsLow RiskLow RiskSome ConcernsSome ConcernsHartnoll et al. (2000) [6]Low RiskSome ConcernsLow RiskLow RiskSome ConcernsSome ConcernsIsemann et al. (2016) [12]Some ConcernsSome ConcernsLow RiskLow RiskHigh RiskHigh Risk Table 4Cochrane Risk of Bias 2 (RoB 2) tool for randomized controlled trialsStudyRandomization ProcessDeviations from Intended InterventionsMissing Outcome DataMeasurement of OutcomeSelection of Reported ResultOverall Risk of BiasVenkatraman et al. (2023) [10]Low RiskSome ConcernsLow RiskLow RiskSome ConcernsSome ConcernsHartnoll et al. (2000) [6]Low RiskSome ConcernsLow RiskLow RiskSome ConcernsSome ConcernsIsemann et al. (2016) [12]Some ConcernsSome ConcernsLow RiskLow RiskHigh RiskHigh Risk Cochrane Risk of Bias 2 (RoB 2) tool for randomized controlled trials Discussion Our systematic review and meta-analysis suggest that sodium supplementation in preterm infants is associated with improved weight gain and reduced hyponatremia risk. However, substantial clinical and methodological heterogeneity across included studies limits the certainty of pooled estimates. Several findings should therefore be interpreted cautiously, particularly for outcomes derived primarily from observational data. Importantly, the included studies differed considerably in sodium dosing strategies, timing of initiation, and routes of administration. Early high-dose sodium administration during the first postnatal week was evaluated in some cohorts, whereas other studies implemented delayed supplementation following the physiologic diuretic phase (typically after days 3–5). In addition, supplementation protocols ranged from fixed empirical dosing regimens to individualized approaches guided by urinary sodium concentrations. Routes of administration also varied between enteral and parenteral delivery. These differences likely contributed substantially to the observed heterogeneity and may partly explain variability in reported electrolyte and morbidity outcomes. Where formal quantitative stratification was limited by the small number of comparable studies, we therefore interpreted findings within the context of these clinical differences using structured narrative analysis. Interpretation also considered both study design and study-level methodological quality. Randomized evidence was limited to three trials, of which two were judged as having some concerns and one as having a high overall risk of bias using RoB 2. In contrast, seven of the nine non-randomized studies were rated as high quality and two as moderate quality using the Newcastle–Ottawa Scale. Nevertheless, favorable observational quality ratings do not eliminate the possibility of residual confounding; observational evidence was therefore treated as complementary rather than equivalent to randomized evidence. Accordingly, comparative randomized evidence was prioritized for causal interpretation where available, while observational studies were used to contextualize findings and to inform outcomes for which randomized evidence was sparse or unavailable. Our findings are consistent with existing evidence that sodium supplementation is associated with early postnatal weight gain, particularly in the randomized trials included in this review. Studies by Isemann et al. [12] and Segar et al. [2] showed improved weight trajectories in supplemented infants, consistent with the 2023 Cochrane review, which concluded that higher sodium intake reduces postnatal growth failure (Diller et al., 2023) [5]. The biological basis for this effect is sound-sodium facilitates extracellular fluid expansion [6], enhances nutrient transport, and supports anabolic processes during a critical phase of postnatal adaptation. In contrast, linear and head growth were not consistently improved. This aligns with Isemann et al. [12], who reported minimal gains in these measures despite sodium intervention. Sodium appears to support soft tissue and water retention more than skeletal or cranial growth, which are influenced by broader nutritional and hormonal factors. A key finding of this review was the association between sodium supplementation and reduced hyponatremia risk. The pooled analysis demonstrated a 72% relative risk reduction in hyponatremia among supplemented infants (RR = 0.28, 95% CI: 0.08–0.96), consistent with prior evidence from Modi et al. (1990), Petersen et al. [16], and the 2023 Cochrane review [5, 8, 16]; however, the wide confidence interval, which approaches unity, and the small number of contributing studies (k = 4) indicate that this estimate should be interpreted with caution. This effect reflects the known physiology of preterm infants, who are predisposed to sodium depletion due to immature renal tubular function and high urinary sodium losses. However, this potential benefit must be weighed against the substantial pooled hypernatremia incidence observed within supplemented cohorts (32%, 95% CI: 15–55%), although estimates varied substantially across studies. Because no untreated comparator group was available for this outcome in most included studies, this proportion describes the frequency of hypernatremia among infants who received supplementation and should not be interpreted as a treatment-attributable increase in risk. Larger cohorts, such as those by Petersen et al. [16] and studies summarized in the 2023 Cochrane review [5, 8, 16], reported hypernatremia rates exceeding 40%, particularly in settings where sodium was initiated early or given in higher doses. In contrast, lower rates in studies like Segar et al. [2] may reflect individualized protocols guided by urine sodium concentrations, although this observation is based on a single retrospective cohort comparison and should be considered preliminary. These findings reinforce the need for tailored sodium supplementation, ideally guided by biomarkers such as urinary sodium [2], and underscore the potential harms of rapid correction or empirical dosing without monitoring, especially in the first postnatal week. The pooled IVH incidence of 29% observed in this review reflects the frequency of IVH within supplemented cohorts rather than a demonstrated increase attributable to sodium supplementation. Although some observational studies reported associations between early high sodium intake and severe IVH, these findings are potentially confounded by gestational age, illness severity, fluid management, and baseline risk. Therefore, the current evidence does not establish a causal relationship between sodium supplementation and IVH. A reduction in sepsis risk was observed in two studies reporting comparative data; however, the small number of studies and limited sample size preclude firm conclusions. While biologically plausible mechanisms have been proposed, this finding should be considered preliminary and warrants further investigation in adequately powered trials [1]. Long-term neurodevelopmental evidence remains limited. Only one historical cohort study reported follow-up into adolescence, suggesting improved motor and cognitive outcomes among supplemented infants. As this finding is based on a single study with historical controls, it should be interpreted cautiously and considered hypothesis-generating. Most included studies did not assess standardized long-term neurodevelopmental outcomes. This aligns with preclinical and clinical evidence showing that sodium plays a critical role in neuronal signaling, myelination, and brain energy metabolism processes essential to neurodevelopment [1]. Recent cohort studies have further linked early-life sodium fluctuations and deficiencies in preterm infants with poorer neurodevelopmental outcomes, such as increased risk of death or impairment at 18 months corrected age, reinforcing the idea that sodium balance in the neonatal period may have long-term cognitive consequences [17]. Despite this, the majority of studies included in our review, such as those evaluating sodium supplementation protocols, did not assess neurodevelopmental outcomes beyond infancy, focusing primarily on growth parameters [12]. These findings suggest that individualized sodium supplementation strategies, including urinary sodium-guided approaches, may be promising; however, further validation in randomized controlled trials is required before firm clinical recommendations can be made. This strategy aligns with AAP (2023) and ESPGHAN (2022) guidelines, which recommend targeted dosing (3–5 mEq/kg/day) with close monitoring to prevent complications like IVH. However, key questions remain regarding the optimal dose, timing, and route of administration, with emerging data [16] suggesting potential benefits of enteral sodium for gut development. Additionally, the observed signal toward reduced sepsis risk introduces a preliminary and hypothesis-generating avenue for future research into the broader physiological effects of sodium in the neonatal period; however, this finding is based on only two studies and requires confirmation in adequately powered trials before any mechanistic conclusions can be drawn. Limitations The majority of included studies were observational. Although seven of nine non-randomized studies were rated as high quality and two as moderate quality using the Newcastle–Ottawa Scale, residual confounding by indication and baseline illness severity cannot be excluded. In particular, sicker and more premature infants are more likely to receive sodium supplementation and are simultaneously at higher baseline risk for adverse outcomes including IVH and mortality, which may inflate apparent associations. Selection bias cannot be excluded in the retrospective cohort studies, as unmeasured confounders such as illness severity scores, ventilatory support, and fluid management practices were not consistently reported or adjusted for across studies. Substantial heterogeneity in sodium dosing, timing, and route of administration limited formal subgroup meta-analysis. Some pooled estimates were based on small numbers of studies, and publication bias could not be formally assessed for most outcomes. In addition, long-term neurodevelopmental data were scarce and derived primarily from a single historical cohort study. Furthermore, this review pooled randomized and observational studies within several quantitative syntheses; although design was considered in the interpretation of findings, most pooled analyses were not formally stratified by study design, which may affect the precision and comparability of the resulting estimates. Several pooled estimates, including those for hypernatremia, intraventricular hemorrhage, necrotizing enterocolitis, sepsis, and mortality, represent single-arm proportions describing event frequency within supplemented cohorts rather than comparative treatment effects, and should not be interpreted as demonstrating increased or decreased risk attributable to sodium supplementation. Meta-regression analyses were based on a small number of studies (k = 6 or fewer) and study-level covariates, and are therefore exploratory and susceptible to ecological bias; they should not be regarded as confirmatory of independent predictor effects. The narrative comparison of urine-sodium-guided dosing was supported by a single retrospective cohort comparison [2] and should not be interpreted as established subgroup evidence. Finally, formal GRADE certainty rating was not performed given the heterogeneity in study designs and outcome definitions, and readers should weigh the certainty of each finding individually in light of the study design, precision, and consistency described above. This decision reflected the small number of studies within individual design strata, which precluded reliable design-specific meta-analysis for most outcomes. Conclusion Our systematic review suggests that sodium supplementation in preterm infants is associated with improved weight gain and, based on comparative analyses, a reduced risk of hyponatremia. However, substantial clinical and methodological heterogeneity across studies, together with the predominance of single-arm incidence data for several outcomes, limits the certainty of pooled estimates and precludes definitive causal conclusions. The observed frequencies of hypernatremia (32%) and intraventricular hemorrhage (29%) within supplemented cohorts, in the absence of consistent comparator data, highlight the need for careful dosing, individualized monitoring, and attention to gestational age and illness severity, rather than establishing these as treatment-attributable risks. Evidence regarding long-term neurodevelopment remains limited to a single historical cohort study and should be considered hypothesis-generating only. Individualized approaches, particularly those guided by urinary sodium measurements, appear promising in exploratory observational data but require further validation in adequately powered randomized trials before informing clinical practice. Future research should focus on dose–response relationships, standardized neurodevelopmental follow-up, and clarification of sodium's broader physiological effects in preterm infants, ideally through randomized designs that allow direct comparative risk estimation for hypernatremia, IVH, and other morbidities currently characterized only as single-arm proportions. Collectively, the current evidence supports cautious, individualized sodium management while underscoring the need for higher-quality randomized trials to guide definitive clinical recommendations. Supplementary Information Supplementary Material 1. Supplementary Material 1.