Altered dietary salt intake for people with chronic kidney disease

Emma J McMahon; Katrina L Campbell; Judith D Bauer; David W Mudge; Jaimon T Kelly

doi:10.1002/14651858.CD010070.pub3

Altered dietary salt intake for people with chronic kidney disease

Authors' declarations of interest

Version published: 24 June 2021 Version history

https://doi.org/10.1002/14651858.CD010070.pub3

Collapse all Expand all

Abstract

available in

Background

Evidence indicates that reducing dietary salt may reduce the incidence of heart disease and delay decline in kidney function in people with chronic kidney disease (CKD). This is an update of a review first published in 2015.

Objectives

To evaluate the benefits and harms of altering dietary salt for adults with CKD.

Search methods

We searched the Cochrane Kidney and Transplant Register of Studies up to 6 October 2020 through contact with the Information Specialist using search terms relevant to this review. Studies in the Register are identified through searches of CENTRAL, MEDLINE, and EMBASE, conference proceedings, the International Clinical Trials Register (ICTRP) Search Portal and ClinicalTrials.gov.

Selection criteria

Randomised controlled trials comparing two or more levels of salt intake in adults with any stage of CKD.

Data collection and analysis

Two authors independently assessed studies for eligibility, conducted risk of bias evaluation and evaluated confidence in the evidence using GRADE. Results were summarised using random effects models as risk ratios (RR) for dichotomous outcomes or mean differences (MD) for continuous outcomes, with 95% confidence intervals (CI).

Main results

We included 21 studies (1197 randomised participants), 12 in the earlier stages of CKD (779 randomised participants), seven in dialysis (363 randomised participants) and two in post‐transplant (55 randomised participants). Selection bias was low in seven studies, high in one and unclear in 13. Performance and detection biases were low in four studies, high in two, and unclear in 15. Attrition and reporting biases were low in 10 studies, high in three and unclear in eight.

Because duration of the included studies was too short (1 to 36 weeks) to test the effect of salt restriction on endpoints such as death, cardiovascular events or CKD progression, changes in salt intake on blood pressure and other secondary risk factors were examined.

Reducing salt by mean ‐73.51 mmol/day (95% CI ‐92.76 to ‐54.27), equivalent to 4.2 g or 1690 mg sodium/day, reduced systolic/diastolic blood pressure by ‐6.91/‐3.91 mm Hg (95% CI ‐8.82 to ‐4.99/‐4.80 to ‐3.02; 19 studies, 1405 participants; high certainty evidence). Albuminuria was reduced by 36% (95% CI 26 to 44) in six studies, five of which were carried out in people in the earlier stages of CKD (MD ‐0.44, 95% CI ‐0.58 to ‐0.30; 501 participants; high certainty evidence). The evidence is very uncertain about the effect of lower salt intake on weight, as the weight change observed (‐1.32 kg, 95% CI ‐1.94 to ‐0.70; 12 studies, 759 participants) may have been due to fluid volume, lean tissue, or body fat. Lower salt intake may reduce extracellular fluid volume in the earlier stages of CKD (‐0.87 L, 95% CI ‐1.17 to ‐0.58; 3 studies; 187 participants; low certainty evidence). The evidence is very uncertain about the effect of lower salt intake on reduction in antihypertensive dose (RR 2.45, 95% CI 0.98 to 6.08; 8 studies; 754 participants). Lower salt intake may lead to symptomatic hypotension (RR 6.70, 95% CI 2.40 to 18.69; 6 studies; 678 participants; moderate certainty evidence). Data were sparse for other types of adverse events.

Authors' conclusions

We found high certainty evidence that salt reduction reduced blood pressure in people with CKD, and albuminuria in people with earlier stage CKD in the short‐term. If such reductions could be maintained long‐term, this effect may translate to clinically significant reductions in CKD progression and cardiovascular events. Research into the long‐term effects of sodium‐restricted diet for people with CKD is warranted.

PICOs

Population

Intervention

Comparison

Outcome

The PICO model is widely used and taught in evidence-based health care as a strategy for formulating questions and search strategies and for characterizing clinical studies or meta-analyses. PICO stands for four different potential components of a clinical question: Patient, Population or Problem; Intervention; Comparison; Outcome.

See more on using PICO in the Cochrane Handbook.

Plain language summary

available in

Altered dietary salt intake for adults with chronic kidney disease

What is the issue?

People with chronic kidney disease (CKD) have a reduced kidney function that persists over time. People with CKD are at increased risk of heart disease and worsening kidney function which can lead to the need for dialysis or kidney transplantation to survive. High salt intake is linked to risk factors for both heart disease and worsening kidney function, including high blood pressure, excess protein in the urine and fluid overload. Therefore reducing salt intake may help reduce risk of heart disease and preserve kidney function. We aimed to assess the benefits and harms of reducing salt intake for people with CKD.

What did we do?

We searched the evidence up to October 2020 to find randomised controlled trials comparing two or more levels of salt intake in adults with CKD, including those in the earlier stages of CKD, those treated with dialysis, and those who had received a kidney transplant.

What did we find?

We found 21 studies that included 1197 adults with CKD (725 in low salt and 725 people in high salt group). Study participants included adults who were in the early stages of CKD (779 people in 12 studies), adults who were on dialysis (363 people in seven studies), and adult kidney transplant recipients (55 people in two studies). The average study duration was seven weeks, ranging from one to 36 weeks. We did not find any studies that measured the effect on the incidence of death, heart disease, or need for dialysis or kidney transplantation. Instead, we found studies that measured risk factors for these outcomes.

We found that lowering salt intake reduced blood pressure and protein in the urine in people with CKD. Lowering salt intake may reduce extracellular fluid volume in the earlier stages of CKD. The evidence is very uncertain about the effects on body weight and reductions in blood pressure medications. The effect on kidney function measures was mixed. Lower salt intake may increase dizziness due to low blood pressure.

Conclusions

We found that reducing salt intake reduced risk factors for heart disease and worsening kidney function in people with CKD in the short term. We could not find evidence of the longer term effects of reducing salt intake, which meant we were unable to determine the direct effects of salt intake on death, heart disease, or need for dialysis or kidney transplantation. We need more high quality research in these areas.

Authors' conclusions

Implications for practice

We found a strong case for the benefits of salt restriction in people with CKD. We found that reducing dietary salt considerably reduced BP in people with CKD. We found consistent evidence that dietary salt restriction reduced proteinuria in people with earlier stage (non‐dialysed, non‐transplanted) CKD by 34% to 36%. If such reductions were maintained long‐term, this may translate to clinically significant reductions in ESKD and cardiovascular events.

Reduced salt intake may increase symptomatic hypotension. Data were sparse for other types of adverse events.

Current evidence‐based clinical guidelines recommend a sodium intake target of less than 6 g of salt (100 mmol; 2300 mg sodium) per day for people with CKD, although achieving longer‐term adherence to this target can be challenging for patients without regular and ongoing support to improve their motivation, knowledge and prevent behaviour change decay. These findings were based on studies with intervention durations up to 6 months. There are ongoing studies with longer intervention durations which will strengthen the evidence for longer‐term effects.

Implications for research

We found that salt reduction in people with CKD reduced BP considerably and consistently reduced proteinuria over a time‐frame of up to 6 months. We found a critical evidence gap in long‐term effects of salt restriction in people with CKD that meant we were unable to determine the direct effects of sodium restriction on critical outcomes such as death and progression to ESKD. If the reductions we found in short‐term studies could be maintained long‐term, these benefits may translate to clinically significant reductions in ESKD incidence and cardiovascular events. Research into longer‐term effects of dietary sodium restriction for people with CKD is warranted, along with investigation of adherence to a low salt diet.

Despite consistent data from observational and non‐randomised studies showing that salt restriction reduced fluid volume in people with CKD, high quality RCTs are lacking. Further research on the effect of salt restriction on other cardiac and vascular abnormalities such as arterial stiffness, left ventricular hypertrophy, inflammation and oxidative stress is warranted.

Future studies investigating salt restriction should employ methods that limit risk of bias due to dietary confounders where possible and should take care to adequately measure dietary intake of not only sodium, but other nutrients that may confound study results. Research into long‐term adherence to a sodium‐restricted diet may assist in translating these results into a practical setting.

Summary of findings

Open in table viewer

Summary of findings 1. Lower salt intake versus higher salt intake for participants with chronic kidney disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)
Lower salt intake versus higher salt intake for participants with CKD
Patient or population: participants with CKD Setting: any Intervention: lower salt intake Comparison: higher salt intake
Outcomes	Risk with higher salt intake	Risk with lower salt intake	Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)
Systolic blood pressure	The mean SBP in the lower salt intake group was 6.91 mm Hg lower (8.82 lower to 4.99 lower) than the higher salt intake group		‐	1261 (19)	⊕⊕⊕⊕ HIGH
Diastolic blood pressure	The mean DBP in the lower salt intake group was 3.91 mm Hg lower (4.8 lower to 3.02 lower) than the higher salt intake group		‐	1257 (19)	⊕⊕⊕⊕ HIGH
Albuminuria	The mean albuminuria in the lower salt intake group was 36%¹ lower (44% lower to 26% lower) than the higher salt intake group		‐	501 (6)	⊕⊕⊕⊕ HIGH
Extracellular fluid volume	The mean extracellular fluid volume in the lower salt intake group was 0.87 L lower (1.17 lower to 0.58 lower) than the higher salt intake group		‐	187 (3)	⊕⊕⊝⊝ LOW ^{2 3}
Weight	The mean weight in the lower salt intake group was 1.32 kg lower (1.94 lower to 0.7 lower) than the higher salt intake group		‐	759 (12)	⊕⊝⊝⊝ VERY LOW ^{2 4 5}
Reduction in antihypertensive dose	24 per 1,000	58 per 1,000 (23 to 144)	RR 2.45 (0.98 to 6.08)	664 (7)	⊕⊝⊝⊝ VERY LOW ^{6 7}
Symptomatic hypotension	8 per 1,000	56 per 1,000 (20 to 156)	RR 6.70 (2.40 to 18.69)	478 (6)	⊕⊕⊕⊝ MODERATE ³
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CKD: chronic kidney disease; CI: confidence interval; SBP: systolic blood pressure; DBP: diastolic blood pressure; RR: risk ratio
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹ Albuminuria was analysed as log‐transformed (mg/day) and expressed as a percentage reduction ² Evidence certainty downgraded one level for uncertain or high risk of bias ³ Evidence certainty downgraded one level for few events and/or participants across all studies ⁴ Evidence certainty downgraded one level for moderate heterogeneity ⁵ Evidence certainty downgraded one levels for indirectness of outcome ⁶ Evidence certainty downgraded two levels for a high risk of bias ⁷ Evidence certainty downgraded one level for wide confidence interval range that include benefit or harm

Background

Description of the condition

Chronic kidney disease (CKD) is a major global public health problem, affecting an estimated 10% of the global population (GBD 2019). According to the Global Burden of Disease, CKD was the 12th most common cause of death in 2017, accounting for 1.2 million deaths worldwide (GBD 2019). Overall CKD death increased by 34% from 2007 to 2017, making it one of the fastest rising major causes of death, alongside diabetes and dementia (GBD 2019). Diabetes and high blood pressure (BP) remain the two biggest risks for developing CKD, with 28% of CKD deaths being due to diabetes, and 28% due to high BP (GBD 2019).

CKD is a progressive condition, and risk of cardiovascular disease and death increases with declining glomerular filtration rate (GFR) (Go 2004). Healthcare costs of treating people with end‐stage kidney disease (ESKD), where dialysis or kidney transplantation is required to survive, have been estimated at about 10 times greater than the cost of CKD management (Hunsicker 2004). Because both cardiovascular disease and progression to ESKD may be delayed, or possibly prevented, effective strategies to reduce these outcomes are needed to improve patients' prognoses and reduce healthcare costs.

Description of the intervention

Excessive salt (sodium) intake is related to many risk factors for cardiovascular disease and CKD progression (Jones‐Burton 2006; Suckling 2010; Malta 2018). These include increased BP, fluid retention, proteinuria, inflammation, oxidative stress, and endothelial dysfunction (Al‐Solaiman 2009; Ritz 2009). The World Health Organisation recommends adults consume less than 5 g of salt (2 g or 87 mmol sodium) per day (WHO 2012). Data from populations around the world show that salt intakes far exceed this (Mozaffarian 2014; Trieu 2015). Estimates in people with CKD are commonly between 9 to 12 g (150 to 200 mmol sodium) (McMahon 2012a). National salt reduction strategies have been established in 75 countries; most of these aim to reduce population salt intake through reducing salt content of food and/or consumer education to shift dietary behaviour (Trieu 2015). Intervention trials comparing two or more levels of salt intake usually achieve a difference in salt intake by dietary education (e.g., individual or group counselling, online support), providing meals to participants and/or using salt supplements (McMahon 2012a).

How the intervention might work

Sodium is an essential nutrient, and has vital roles in regulating fluid homeostasis, blood volume and osmotic pressure (Elliot 2006). Excess sodium increases serum osmolality, triggering the release of vasopressin (arginine vasopressin (AVP), also known as antidiuretic hormone) which initiative a cascade of metabolic actions including activation of the renin‐angiotensin‐aldosterone system (RAAS) causing blood vessels to constrict and increasing water reabsorption from the urine (Qian 2018). Chronic high‐salt intake (> 7 days) leads to excessive AVP production and vasoconstriction. This can lead to kidney and cardiovascular damage through dysregulation of fluid volume homeostasis and BP. Non‐osmotic storage of salt in the skin and the muscle is thought to be linked to increased risk of cardiovascular and kidney disease (Garofalo 2018; Qian 2018).

High BP is common in CKD, and sustained elevations in BP hasten CKD progression (Bakris 2000). Studies in the general population have consistently demonstrated a link between dietary salt intake and BP (He 2013; Svetkey 1999). A 2017 Cochrane review found that reducing salt from 11.5 g/day to 3.8 g/day reduced systolic BP/diastolic BP in people with normotension by 1/0 mm Hg, and in people with hypertension about approximately 5.5/3 mm Hg (Graudal 2020). High salt intake may have a greater impact on BP in people with CKD than in those without CKD since CKD may reduce the ability to excrete the excess salt. A 2010 Cochrane review on reducing salt intake in people with diabetic kidney disease showed considerable BP reductions; systolic/diastolic BP was lowered by 7/3 mm Hg (Suckling 2010).

It has also been suggested that salt has adverse effects independent of BP. Todd 2010 found arterial stiffness measured by pulse wave velocity was significantly decreased independently of BP changes in hypertensive people on a low salt diet. Increased pulse wave velocity is a predictor of all‐cause and cardiovascular death (Guerin 2001). Proteinuria, a risk factor for both CKD progression and cardiovascular disease in people with CKD, has also shown to be reduced by salt restriction independent of BP (Verhave 2004).

Why it is important to do this review

Salt intake shows great promise as a modifiable risk factor for reducing cardiovascular risk and CKD progression even among people in the very early stages of the disease. However, clear consensus of the benefits of reducing salt for people with CKD and the optimal target salt intake for this population have yet to be established. This review updates a 2015 Cochrane review of altered dietary salt intake in people with CKD (McMahon 2015). Updating the review will ensure that best evidence is available for the benefits (and potential adverse effects) of reducing salt for people with CKD.

Objectives

This review evaluated the benefits and harms of altering dietary salt intake in adults with CKD.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) and quasi‐RCTs (RCTs in which allocation to treatment was obtained by alternation, use of alternate medical records, date of birth or other predictable methods) measuring the effect of low versus high salt intake in adults with CKD.

Types of participants

Inclusion criteria

Adults (≥ 18 years) with CKD were included. For this review update, we defined CKD as those who are receiving any form of kidney replacement therapy, have a functioning kidney transplant, have impaired kidney function defined as a reduced GFR (< 60 mL/min/1.73 m²) or the presence of other markers of kidney damage such as proteinuria (KDOQI stages 1‐5) or an elevated serum creatinine (SCr > 120 mmol/L) (KDIGO 2013), or as defined by study authors. This definition of CKD has been updated from that used in the initial review (McMahon 2015), where we defined CKD using the Kidney Disease Outcomes Quality Initiative (KDOQI) Clinical Practice Guidelines) at all disease stages (NKF 2002).

Exclusion criteria

Pregnant women
Children (aged up to 18 years).

Types of interventions

We planned to evaluate the following interventions.

Comparing two or more differing levels of sodium intake
Of at least one week duration
Where concomitant interventions such as antihypertensive medication or other dietary modifications were used during the study period, providing that these interventions were constant throughout the low and high salt interventions.

In the initial review (2015), we included studies that evaluated sodium intake using 24‐hour urinary sodium excretion (24‐hour UNa) with a minimum difference in 24‐hour UNa of 34 mmol/day (2 g salt/day) achieved between allocated interventions. No studies were excluded based on this criterion in the initial review. This eligibility criterion has been removed in this update, allowing the inclusion of studies measuring sodium intake using other methods (such as food records or 24‐hour recalls) and allowing any magnitude of difference in sodium intake between the allocated interventions.

Types of outcome measures

Primary outcomes

Cardiovascular death
Death (any cause)

Secondary outcomes

Cardiovascular disease (coronary artery disease, heart failure, cerebrovascular disease, peripheral vascular disease)
Progression to ESKD requiring dialysis or transplantation
24‐hour UNa
Change in BP: clinic and 24‐hour measurement
Change in arterial stiffness: pulse wave velocity and augmentation index)
Change in kidney function measures: creatinine clearance (CrCl), SCr, proteinuria, GFR
Change in markers of fluid overload: brain natriuretic peptide (NT‐proBNP), weight, bio‐impedance analysis
Change in markers of oxidative stress or inflammation: C‐reactive protein, adipokines
Adverse events: hypotensive episodes, undesirable change in blood lipids (low density lipoprotein, high‐density lipoprotein).

Search methods for identification of studies

Electronic searches

Monthly searches of the Cochrane Central Register of Controlled Trials (CENTRAL)
Weekly searches of MEDLINE OVID SP
Searches of kidney and transplant journals, and the proceedings and abstracts from major kidney and transplant conferences
Searching of the current year of EMBASE OVID SP
Weekly current awareness alerts for selected kidney and transplant journals
Searches of the International Clinical Trials Register (ICTRP) Search Portal and ClinicalTrials.gov.

Studies contained in the Register are identified through searches of CENTRAL, MEDLINE, and EMBASE based on the scope of Cochrane Kidney and Transplant. Details of search strategies, as well as a list of handsearched journals, conference proceedings and current awareness alerts, are available on the Cochrane Kidney and Transplant website.

See Appendix 1 for search terms used in strategies for this review.

Searching other resources

Reference lists of review articles, relevant studies, and clinical practice guidelines
Contacting relevant individuals/organisations seeking information about unpublished or incomplete studies.

Data collection and analysis

Selection of studies

2021 update

The search results were screened and studies included or excluded based on the selection criteria listed above. Excluded studies from the 2015 review were reassessed based on the updated eligibility criteria. Screening was done independently by two authors.

2015 review

The search strategy described was used to obtain titles and abstracts of studies that may be relevant to the review. The titles and abstracts were screened independently by two authors, who discarded studies that were not applicable; however, studies and reviews that might have included relevant data or information on studies were retained initially. Two authors independently assessed the retrieved abstracts, and if necessary, the full text of these studies to determine which studies satisfied the inclusion criteria.

Data extraction and management

Data extraction for studies new to this 2021 update was done using standard data extraction forms. In the 2015 review, data extraction was carried out independently by two authors. Studies reported in non‐English language journals were translated before assessment. Where more than one publication of one study exists, reports were grouped together and the publication with the most complete data was used in the analyses. Where relevant outcomes were only published in earlier versions these data were used. Any discrepancies between published versions were highlighted.

Assessment of risk of bias in included studies

The following items were independently assessed by two authors using the risk of bias assessment tool (Higgins 2011) (see Appendix 2).

Was there adequate sequence generation (selection bias)?
Was allocation adequately concealed (selection bias)?
Was knowledge of the allocated interventions adequately prevented during the study?
- Participants and personnel (performance bias)
- Outcome assessors (detection bias)
Were incomplete outcome data adequately addressed (attrition bias)?
Are reports of the study free of suggestion of selective outcome reporting (reporting bias)?
Was the study apparently free of other problems that could put it at a risk of bias? Two additional domains were addressed:
- Carry‐over effect
- Confounding factors (bias due to differences between groups other than sodium intake that are likely to impact the effect estimate (in either direction) including differential changes in diet/medication (potassium intake, body weight changes, anti‐hypertensive medication dosage) and baseline differences in the outcome, where likely to impact the effect estimate)

Measures of treatment effect

For dichotomous outcomes (cardiovascular death, death (any cause), progression to ESKD, cardiovascular disease, symptomatic hypotension, reduction in anti‐hypertensive dose) results were expressed as risk ratios (RR) with 95% confidence intervals (CI). Where continuous scales of measurement were used to assess the effects of treatment (BP, pulse wave velocity, augmentation index, CrCl, SCr, proteinuria, GFR, brain natriuretic peptide (NT‐proBNP), weight, bio‐impedance analysis, C‐reactive protein, adipokines) the mean difference (MD) was used, or the standardised mean difference (SMD) if different scales were used. Studies analysing change scores were included in meta‐analysis along with studies including endpoint data only.

Unit of analysis issues

For cross‐over studies, we determined the mean difference in outcomes as the difference between the end of low salt and high salt periods. Standard deviations for cross‐over studies were imputed from paired analysis CIs or P values. Where these were not reported, SD of the values at the end of each period was used.

For parallel studies, we calculated the treatment effect as the difference between treatment groups' change in outcomes from baseline. Where these were not reported, or could not be imputed from reported values, values at the end of the intervention were used.

Dealing with missing data

Any further information required from the original author was requested by written correspondence (e.g. emailing corresponding author) and any relevant information obtained in this manner was included in the review. Evaluation of important numerical data such as screened, randomised subjects as well as intention‐to‐treat, as‐treated and per‐protocol population were performed. Attrition rates, for example drop‐outs, losses to follow‐up and withdrawals were investigated. Issues of missing data and imputation methods (e.g. last‐observation‐carried‐forward) were critically appraised (Higgins 2011).

Assessment of heterogeneity

We first assessed the heterogeneity by visual inspection of the forest plot. We then quantified statistical heterogeneity using the I² statistic, which describes the percentage of total variation across studies that is due to heterogeneity rather than sampling error (Higgins 2003). A guide to the interpretation of I² values was as follows.

0% to 40%: might not be important
30% to 60%: may represent moderate heterogeneity
50% to 90%: may represent substantial heterogeneity
75% to 100%: considerable heterogeneity.

The importance of the observed value of I² depends on the magnitude and direction of treatment effects and the strength of evidence for heterogeneity (e.g. P value from the Chi² test, or a CI for I²) (Higgins 2011).

Assessment of reporting biases

If possible, funnel plots were to be constructed to assess for the potential existence of small study bias (Higgins 2011). There were insufficient data to enable construction of funnel plots for this review.

Data synthesis

Data were pooled using the random‐effects model but the fixed‐effect model was also used to ensure robustness of the model chosen and susceptibility to outliers.

Subgroup analysis and investigation of heterogeneity

Subgroup analysis was used to explore possible sources of heterogeneity. Heterogeneity among participants could be related to age, ethnicity, stage of CKD and presence of comorbidities (e.g. hypertension and diabetes).

Sensitivity analysis

Where necessary, we performed sensitivity analyses in order to explore the influence of the following factors on effect size.

repeating the analysis excluding unpublished studies
repeating the analysis taking account of risk of bias, as specified
repeating the analysis excluding any very long or large studies to establish how much they dominate the results
repeating the analysis excluding studies using the following filters: diagnostic criteria, language of publication, source of funding (industry versus other), and country.

Summary of findings and assessment of the certainty of the evidence

We presented the main results of the review in 'Summary of findings' tables. These tables present key information concerning the quality of the evidence, the magnitude of the effects of the interventions examined, and the sum of the available data for the main outcomes (Schunemann 2011a). The 'Summary of findings' tables also include an overall grading of the evidence related to each of the main outcomes using the GRADE (Grades of Recommendation, Assessment, Development and Evaluation) approach (GRADE 2008; GRADE 2011). The GRADE approach defines the quality of a body of evidence as the extent to which one can be confident that an estimate of effect or association is close to the true quantity of specific interest. The quality of a body of evidence involves consideration of within‐trial risk of bias (methodological quality), directness of evidence, heterogeneity, precision of effect estimates and risk of publication bias (Schunemann 2011b). We presented the following outcomes in the 'Summary of findings' tables.

Systolic BP
Diastolic BP
Albuminuria
Extracellular fluid volume
Weight
Reduction in antihypertensive dose
Symptomatic hypotension.

Results

Description of studies

See Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies.

Results of the search

Figure 1 details the study search results. The initial version of this review (McMahon 2015) included eight studies (24 reports) with 224 participants in the low salt group and 218 participants in the high salt group (442 combined; 261 unique). Nine studies were excluded, four were ongoing, and two studies were awaiting assessment. None of the previously excluded studies were eligible when reassessed against the updated eligibility criteria.

Figure 1

Study flow diagram

For this update (2021), we identified 67 new reports. We included 13 new included studies (45 reports) (BalanceWise‐HD 2013; BalanceWise‐PD 2011; Catto 1973; de Vries 2016; Doulton 2007; ESMO 2017; ESPECIAL 2014; Kwakernaak 2014; Mulhauser 1996; Power 2010; Rodrigues Telini 2014; Saran 2017; ViRTUE‐CKD 2016); three of these were previously ongoing studies (BalanceWise‐HD 2013; Doulton 2007; Saran 2017). We also identified six new reports of five previously included studies; seven new excluded studies (14 reports); four new reports of two existing excluded studies; and six reports for five new ongoing studies. One study previously listed as ongoing was withdrawn in late 2015 and has been moved to excluded studies (Reyes 2013). All non‐randomised studies and RCTs in non‐CKD patients have been removed from this update.

A total of 21 studies (72 reports) have been included and 11 studies (26 reports) have been excluded from this review update. There are five ongoing studies (6 reports) which will be assessed in a future update of this review.

Included studies

See Characteristics of included studies.

We included 21 studies (1197 randomised participants), 12 in the earlier stages of CKD (779 randomised participants), seven in dialysis (363 randomised participants) and two in post‐transplant (55 randomised participants). These studies reported the results of 725 participants in the low salt arm and 725 participants in the high salt arm. Power 2010 (31 participants) did not report the sample size for each group and therefore the results could not be included in the quantitative synthesis.

Six studies were conducted in the Netherlands (de Vries 2016; DUAAAL 2011; ESMO 2017; Kwakernaak 2014; ViRTUE‐CKD 2016; Vogt 2008), four in the USA (BalanceWise‐HD 2013; BalanceWise‐PD 2011; Fine 1997; Saran 2017), four in the UK (de Brito‐Ashurst 2013; Catto 1973; Doulton 2007; Power 2010), and one study in Australia (LowSALT CKD 2012), Brazil (Rodrigues Telini 2014), Germany (Mulhauser 1996), Japan (Konishi 2001), Korea (ESPECIAL 2014), Spain (Ruilope 1992a), and Turkey (Keven 2006).

Of the 21 included studies, 12 were cross‐over studies (Catto 1973; de Vries 2016; Doulton 2007; DUAAAL 2011; Fine 1997; Konishi 2001; Kwakernaak 2014; LowSALT CKD 2012; Ruilope 1992a; Saran 2017; ViRTUE‐CKD 2016; Vogt 2008) and nine were parallel design studies (BalanceWise‐HD 2013; BalanceWise‐PD 2011 de Brito‐Ashurst 2013; ESMO 2017; ESPECIAL 2014; Keven 2006; Mulhauser 1996; Power 2010; Rodrigues Telini 2014). Values from paired analyses were reported for most cross‐over studies, although not always for all outcomes. Konishi 2001 and Ruilope 1992a did not report paired analyses and values from the end of the intervention were used. Four of the parallel studies (BalanceWise‐HD 2013; BalanceWise‐PD 2011; de Brito‐Ashurst 2013; ESPECIAL 2014) reported change from baseline for each group, and these data were used for analysis. Mulhauser 1996 reported change from baseline for BP but not for other outcomes. Four of the parallel studies (ESMO 2017; Keven 2006; Power 2010; Rodrigues Telini 2014) did not present data on change from baseline, and because there were no appropriate data available to impute values, values at the end of the intervention were used.

Nine studies enrolled participants with ESKD, seven in dialysis (haemodialysis (HD): BalanceWise‐HD 2013; Catto 1973; Doulton 2007; Power 2010; Rodrigues Telini 2014; peritoneal dialysis (PD): BalanceWise‐PD 2011; Fine 1997) and two post‐transplant (Keven 2006; de Vries 2016). Twelve studies enrolled participants in earlier stages of CKD (de Brito‐Ashurst 2013; DUAAAL 2011; ESMO 2017; ESPECIAL 2014; Konishi 2001; Kwakernaak 2014; LowSALT CKD 2012; Mulhauser 1996; Ruilope 1992a; Saran 2017; ViRTUE‐CKD 2016; Vogt 2008).

Median study duration was seven weeks, ranging from one to 36 weeks. A cut‐off of four weeks was used to classify studies according to intervention duration (short‐term: fewer than four weeks; longer‐term: four weeks or more). Three studies were classified as short‐term, all of which were cross‐over studies with two arms and enrolled participants in the earlier stages of CKD: Konishi 2001 and Ruilope 1992a had one‐week interventions and LowSALT CKD 2012 had two‐week interventions. We classified 15 studies as longer‐term (range: four to 36 weeks); these included all ESKD studies (BalanceWise‐HD 2013; BalanceWise‐PD 2011; Catto 1973; de Vries 2016; Doulton 2007; Fine 1997; Keven 2006; Power 2010; Rodrigues Telini 2014) and nine studies in earlier CKD (de Brito‐Ashurst 2013; DUAAAL 2011; ESMO 2017; ESPECIAL 2014; Kwakernaak 2014; Mulhauser 1996; Saran 2017; ViRTUE‐CKD 2016; Vogt 2008).

Thirteen studies used dietary counselling to achieve a difference in sodium intake. Of these, 11 studies compared sodium restriction achieved through dietary counselling in the low salt intervention to usual diet or usual care in the high salt intervention (BalanceWise‐PD 2011;de Brito‐Ashurst 2013; de Vries 2016; DUAAAL 2011; ESMO 2017; Keven 2006; Kwakernaak 2014; Rodrigues Telini 2014; Saran 2017; ViRTUE‐CKD 2016; Vogt 2008). Two studies compared against a less intensive education strategy to reduce dietary sodium: ESPECIAL 2014 compared intensive education to reduce sodium in the high salt intervention versus brief education in the low salt intervention, and BalanceWise‐HD 2013 provided individualised dietary counselling paired with personal digital assistant‐based self‐monitoring and feedback on sodium intake to participants in the low salt intervention and six educational modules to both groups (attention control for the high salt group). Six studies used sodium supplements to achieve difference in sodium intake with sodium supplements in the high salt group compared against placebo (Catto 1973; Doulton 2007; Fine 1997; LowSALT CKD 2012; Mulhauser 1996) or a lower dose of sodium supplementation (Ruilope 1992a) in the low salt group. Power 2010 did not specify means for achieving difference in sodium intake. Konishi 2001 provided all food for participants.

Five studies included concomitant interventions of medications that remained stable throughout the high and low salt phases. DUAAAL 2011 and Kwakernaak 2014 started all participants on lisinopril 40 mg/day, ESPECIAL 2014 used olmesartan medoxomil 40 mg/day, Ruilope 1992a used verapamil 240 mg/day, and ViRTUE‐CKD 2016 used ramipril 10 mg/day. The protocol for Konishi 2001 included cessation of all medications one week before the study.

Outcome reporting in included studies

Where multiple markers of sodium intake were collected, 24‐hour UNa excretion was used preferentially in analyses. Fifteen studies reported 24‐hour UNa excretion as a marker of sodium intake (de Brito‐Ashurst 2013;de Vries 2016;DUAAAL 2011;ESMO 2017;ESPECIAL 2014;Fine 1997;Keven 2006;Konishi 2001;Kwakernaak 2014;LowSALT CKD 2012;Mulhauser 1996; Ruilope 1992a;Saran 2017;ViRTUE‐CKD 2016;Vogt 2008). Fine 1997, whose participants were receiving PD, added 24‐hour UNa excretion to 24‐hour dialysate sodium to achieve a total value for sodium excretion, and this value was used for analysis. Four studies collected self‐reported dietary sodium intake measurements (values were converted to mmol/day prior to analysis); BalanceWise‐HD 2013 and BalanceWise‐PD 2011 used three, 24‐hour recalls, Rodrigues Telini 2014 used three‐day diet records, while Catto 1973, Doulton 2007 and Power 2010 did not report sodium intake or excretion.

All studies except BalanceWise‐HD 2013 and Power 2010 measured and reported BP. Where more than one BP measurement was reported, 24‐hour ambulatory BP was used preferentially in analyses, and clinic‐assessed BP was used preferentially over self‐assessed BP measurements. Seven studies measured 24‐hour ambulatory BP (de Brito‐Ashurst 2013; Doulton 2007; ESMO 2017; Konishi 2001; LowSALT CKD 2012; Ruilope 1992a; Saran 2017), 10 used clinic‐assessed BP (Catto 1973; de Vries 2016; DUAAAL 2011; ESPECIAL 2014; Keven 2006; Kwakernaak 2014; Mulhauser 1996; Rodrigues Telini 2014; ViRTUE‐CKD 2016; Vogt 2008); BalanceWise‐PD 2011 and Fine 1997 used self‐assessed BP. If more than one position was reported, standing BP was the first preference, except in BalanceWise‐PD 2011 where sitting BPs were used due to there being more missing data for the standing measurements.

Eleven studies reported changes in urinary protein or albumin (de Vries 2016; DUAAAL 2011; ESMO 2017; ESPECIAL 2014; Konishi 2001; Kwakernaak 2014; LowSALT CKD 2012; Mulhauser 1996; Saran 2017; ViRTUE‐CKD 2016; Vogt 2008). Pooled analyses were performed on log‐transformed values (mg/day) to give log mean (95% CI of log mean). The exponential of these values were also reported to indicate percentage change for ease of understanding (Bland 1996).

Nine studies reported 24‐hour urinary protein excretion (de Vries 2016; DUAAAL 2011; ESMO 2017; Konishi 2001; Kwakernaak 2014; LowSALT CKD 2012; Mulhauser 1996;ViRTUE‐CKD 2016; Vogt 2008). Three studies (DUAAAL 2011; Kwakernaak 2014; ViRTUE‐CKD 2016) reported geometric mean (95% CI) which was transformed to log mean and SD using the methods specified in the Cochrane Handbook. Two studies (ESMO 2017; Vogt 2008) reported values as mean and standard error which were converted to log mean and SD using the methods specified in Higgins 2008. LowSALT CKD 2012 provided raw data and these were log‐transformed. Two studies (de Vries 2016; Konishi 2001) reported data as median and inter‐quartile range and could not be entered into pooled analysis. Mulhauser 1996 reported values as mean and range and from these values it appeared that the data were not normally distributed therefore could not be entered into pooled analysis.

Six studies reported 24‐hour urinary albumin excretion (de Vries 2016; ESPECIAL 2014; Kwakernaak 2014; LowSALT CKD 2012; Saran 2017; ViRTUE‐CKD 2016). de Vries 2016 and Saran 2017 reported values as change in log mean and 95% CI. Two studies (Kwakernaak 2014; ViRTUE‐CKD 2016) reported geometric mean (95% CI) which was transformed to log mean and SD using the methods specified in the Cochrane Handbook. Two studies (LowSALT CKD 2012; ESPECIAL 2014) provided raw data which were log‐transformed for analysis.

Thirteen studies reported change in body weight (BalanceWise‐HD 2013 de Brito‐Ashurst 2013; de Vries 2016; Doulton 2007; DUAAAL 2011; ESMO 2017; Fine 1997; Kwakernaak 2014; LowSALT CKD 2012; Ruilope 1992a; Saran 2017; ViRTUE‐CKD 2016; Vogt 2008).

Objective markers of fluid status were reported by six studies; three studies reported extracellular fluid volume (LowSALT CKD 2012; Rodrigues Telini 2014; Saran 2017), de Brito‐Ashurst 2013 reported total body water, two studies reported presence of oedema (DUAAAL 2011; ViRTUE‐CKD 2016), and three reported intradialytic weight gain (BalanceWise‐HD 2013; Doulton 2007; Rodrigues Telini 2014). Doulton 2007 reported only mean values at the end of each intervention arm and the P value for the mean difference for body weight and inter‐dialytic weight gain, therefore SDs were imputed from P values using the methods described in the Cochrane Handbook.

Excluded studies

We excluded 11 studies (26 reports) that did not meet our inclusion criteria. Reasons for exclusion were non‐CKD population (Li 2013f; Suckling 2016), concomitant intervention that was not stable between interventions (DD 2017; Esnault 2005; Kauric‐Klein 2012; HHK 2018) not comparing two or more levels of sodium intake (Chanwikrai 2012; Imanishi 1997), or less than one week duration (Clark‐Cutaia 2016; De Nicola 2000).

Ongoing studies

Five studies are ongoing and will be assessed once they have been completed (NCT03373500; OxCKD1 2012; STICK 2015; SUBLIME 2020; SUPER 2016).

Risk of bias in included studies

Figure 2 and Figure 3 summarise risk of bias assessment for the included studies.

Figure 2

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Figure 3

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Studies were frequently assessed as having unclear or high risk of bias for the risk of bias study domains with selection bias, performance bias, detection bias, and confounding bias domains having the largest proportion of unclear/high risk of bias.

Allocation

Random sequence generation

All included studies were randomised. Further details were provided about the method of randomisation for 12 studies and these studies were judged to be at low risk of bias (BalanceWise‐HD 2013; BalanceWise‐PD 2011; de Brito‐Ashurst 2013; de Vries 2016; DUAAAL 2011; ESMO 2017; ESPECIAL 2014; Fine 1997; Kwakernaak 2014; LowSALT CKD 2012; Vogt 2008; ViRTUE‐CKD 2016). The remaining nine studies were judged to have unclear risk of bias.

Allocation concealment

Allocation concealment was not described for 11 studies and these were judged to be low risk of bias (Catto 1973; de Brito‐Ashurst 2013; de Vries 2016; DUAAAL 2011; Keven 2006; Konishi 2001; Kwakernaak 2014; Power 2010; Rodrigues Telini 2014; Ruilope 1992a; Vogt 2008). Saran 2017 was considered to be at high risk of bias due to allocation concealment. Randomisation was concealed using sealed envelopes but it was unclear how this process was monitored. This study had considerable violation of the protocol at one of the two study sites (6/21 given treatment in the reverse order for unknown reasons, 4/21 recommended to continue low salt instead of crossing over to high salt diet) (Saran 2017). The remaining nine studies were judged to have unclear risk of bias.

Blinding

Most outcomes were objective and therefore at lower risk of being influenced by performance bias. Five studies (Catto 1973; Doulton 2007, Fine 1997;LowSALT CKD 2012;Mulhauser 1996) were double blinded. Clinicians, but not participants, were blinded in ESPECIAL 2014. Investigators, but not participants or study coordinators, were blinded in Saran 2017, and the lack of blinding of study coordinators may have influenced the outcomes, as evidenced by the protocol deviations described previously.

Twelve studies had open‐label dietary interventions (BalanceWise‐HD 2013; BalanceWise‐PD 2011; de Brito‐Ashurst 2013; de Vries 2016; DUAAAL 2011; ESMO 2017; Keven 2006; Kwakernaak 2014; Power 2010; Rodrigues Telini 2014; ViRTUE‐CKD 2016; Vogt 2008).

Konishi 2001 and Ruilope 1992a did not describe blinding.

Incomplete outcome data

There was a high risk of attrition bias due to incomplete outcome data for three studies; BalanceWise‐PD 2011 reported 19% attrition and data were missing for a large proportion of participants for BP, weight and urine outcomes; Power 2010 had 38% attrition and did not report reasons or attrition by allocation; Saran 2017 had a proportion of participants with missing data for some outcomes (e.g. only 10/21 participants at one of the sites included in urinary albumin data, but 21/21 included in urinary creatinine data), and reasons for this were not provided. The authors were contacted to provide further information, but none was provided.

Konishi 2001 and Ruilope 1992a did not report participant attrition. Doulton 2007 specified a target enrolment of 20 and reported 13 completers but did not report actual enrolment to determine attrition. Fine 1997 reported a large degree of attrition (12/32 participants; 37%) that was even among the two groups, however may have introduced bias as some participants withdrew due to concern about values when self‐measuring BP. Rodrigues Telini 2014 reported 26% attrition which was higher in the high salt (9/30; 30%) than the low salt group (5/23; 21%).

Selective reporting

Risk of reporting bias was unclear for four studies for which trial registration or study protocol could not be located (Keven 2006; Konishi 2001; Ruilope 1992a) and for Saran 2017 with unexplained missing outcome data. Risk of reporting bias was high for two studies where most or all outcomes could not be meta‐analysed due to the way that they were reported (Mulhauser 1996; Power 2010).

Other potential sources of bias

Carry‐over effect in cross‐over studies

Carry‐over effect may have introduced bias in Konishi 2001 and Ruilope 1992a; both were of one week duration with no washout between interventions. Ruilope 1992a introduced a new antihypertensive medication at day 1 of the first intervention, increasing risk of treatment order effect.

Bias from cofounders

Bias from confounders was classified as unclear or high risk for nearly all studies, mostly due to lack of measurement or failing to adequately report or account for the impact of changes in medications and/or dietary changes such as potassium intake, protein intake and/or weight loss between interventions. Dietary changes were considered more likely to introduce bias in studies comparing a sodium restricted to usual diet, and less so those using sodium supplements to achieve differences in salt intake. Mean systolic/diastolic BP was 6‐7/3 mm Hg higher in low salt versus high salt group at baseline in both Keven 2006 and Rodrigues Telini 2014. While this difference may not have been statistically significant, this may have impacted the clinical significance of the results, as both studies did not report results as change from baseline. This information was sought from authors but was not provided.

We assessed that two studies were at high risk of bias in relation to funding sources (Fine 1997; Saran 2017).

Effects of interventions

See: Summary of findings 1 Lower salt intake versus higher salt intake for participants with chronic kidney disease

Duration of the included studies was too short to test the efficacy of salt restriction on endpoints such as death, cardiovascular events, or progression to ESKD. Therefore, changes in salt intake on BP and risk factors for cardiovascular disease and ESKD were considered in evaluating the evidence for this review.

Sodium intake/excretion

Low salt intake resulted in sodium intake/excretion being reduced in included studies (Analysis 1.1 (18 studies, 1383 participants): MD ‐73.51 mmol/day, 95% CI ‐92.76 to ‐54.27; high certainty evidence) although heterogeneity was high (I² = 93%).

CKD stage

In the earlier stages of CKD, reducing salt intake reduced sodium intake/excretion (Analysis 1.1.1 (12 studies, 1048 participants): MD ‐83.81 mmol/day, 95% CI ‐104.54 to ‐63.08; I² = 92%). In studies in dialysis patients, reducing salt intake reduced sodium intake/excretion (Analysis 1.1.2 (4 studies, 259 participants): MD ‐27.17 mmol/day, 95% CI ‐48.59, ‐5.76; I² = 70%. In studies in post‐transplant recipients, reducing salt intake reduced sodium intake/excretion (Analysis 1.1.3 (2 studies, 76 participants): MD ‐96.99 mmol/day, 95% CI ‐148.10 to ‐45.89; I² = 55%).

The test for subgroup differences showed a difference between the earlier stages of CKD and dialysis patients (Chi² = 13.87, df = 1 (P = 0.0002), I² = 92.8%) and dialysis patients and transplant recipients (Chi² = 6.10, df = 1 (P = 0.01), I² = 83.6%) but not between the earlier stages of CKD and transplant recipients (Chi² = 0.22, df = 1 (P = 0.64), I² = 0%).

Duration

In short‐term studies (< 4 weeks), reducing salt intake reduced sodium intake/excretion (Analysis 2.1.1 (3 studies, 142 participants): MD ‐111.29 mmol/day, 95% CI ‐134.25 to ‐88.32; I² = 77%;). In longer‐term studies (≥ 4 weeks), reducing salt intake reduced sodium intake/excretion ( Analysis 2.1.2 (15 studies, 1241 participants): MD ‐64.53 mmol/day, 95% CI ‐83.29, ‐45.78; I² = 90%).

The test for subgroup differences showed a difference between short‐ and longer‐term studies (Chi² = 248.56, df = 1 (P = 0.00001), I² = 93%).

Blood pressure

Reducing salt intake reduced both systolic (Analysis 1.2 (19 studies, 1261 participants): MD ‐6.91 mm Hg, 95% CI ‐8.82 to ‐4.99; I² = 50%; high certainty evidence) and diastolic BP (Analysis 1.3 (19 studies, 1257 participants): MD ‐3.91 mm Hg, 95% CI ‐4.80 to ‐3.02; I² = 21%; high certainty evidence).

Power 2010 reported intradialytic BP stability as being significantly improved when sodium restriction was combined with low dialysate (P < 0.05), but did not report results for sodium restriction alone.

CKD stage

In the earlier stages of CKD, reducing salt intake reduced both systolic (Analysis 1.2.1 (12 studies, 1036 participants): MD ‐6.10 mm Hg, 95% CI ‐8.11 to ‐4.08; I² = 48%) and diastolic BP (Analysis 1.3.1 (12 studies, 1032 participants): MD ‐3.47 mm Hg, 95% CI ‐4.28 to ‐2.65; I² = 0%).

In dialysis patients, reducing salt intake reduced both systolic (Analysis 1.2.2 (5 studies, 149 participants): MD ‐6.32 mm Hg, 95% CI ‐11.04 to ‐1.60; I² = 0%) and diastolic BP (Analysis 1.3.2 (5 studies, 149 participants): MD ‐3.46 mm Hg, 95% CI ‐6.39 to ‐0.54; I² = 0%).

In post‐transplant recipients, reducing salt intake reduced both systolic (Analysis 1.2.3 (2 studies, 76 participants): MD ‐11.94 mm Hg, 95% CI ‐15.77 to ‐8.11; I² = 12%) and diastolic BP (Analysis 1.3.2 (2 studies, 76 participants): MD ‐7.13 mm Hg, 95% CI ‐9.46 to ‐4.79; I² = 0%).

For systolic BP, the test for subgroup differences showed a difference between the earlier stages of CKD and post‐transplant recipients (Test for subgroup differences: Chi² = 7.01, df = 1 (P = 0.008), I² = 85.7%) but not between the earlier stages of CKD and dialysis patients (Chi² = 0.01, df = 1 (P = 0.93), I² = 0%) or dialysis patients and post‐transplant recipients (Chi² = 3.28, df = 1 (P = 0.07), I² = 69.6%).

For diastolic BP, the test for subgroup differences showed a difference between earlier stages of CKD and post‐transplant recipient (Chi² = 8.38, df = 1 (P = 0.004), I² = 88.1%), but not for earlier stages of CKD and dialysis patients (Chi² = 0.00, df = 1 (P = 1.00), I² = 0%) or dialysis patients and post‐transplant recipients (Chi² = 3.68, df = 1 (P = 0.06), I² = 72.8%).

Duration

In short‐term studies (< 4 weeks), reducing salt intake reduced both systolic (Analysis 2.2.1 (3 studies, 144 participants): MD ‐7.91 mm Hg, 95% CI ‐11.53 to ‐4.28; I² = 0%) and diastolic BP (Analysis 2.3.1 (3 studies, 144 participants): MD ‐3.70 mm Hg, 95% CI ‐5.74 to ‐1.67; I² = 0%). In longer‐term studies (≥ 4 weeks), reducing salt intake reduced both systolic (Analysis 2.2.2 (16 studies, 1117 participants): MD ‐6.82 mm Hg, 95% CI ‐9.00 to ‐4.64; I² = 56%) and diastolic BP (Analysis 2.3.2 (16 studies, 1113 participants): MD ‐3.99 mm Hg, 95% CI ‐5.03 to ‐2.94; I² = 32%; .2).

The tests for subgroup differences showed no difference between short‐ and longer‐term studies for both systolic (Chi² = 0.25, df = 1 (P = 0.6155), I² = 0%) and diastolic BP (Chi² = 0.06, df = 1 (P = 0.8073), I² = 0%).

Urinary protein excretion

Reducing salt intake reduced 24‐hour urinary protein excretion in the earlier stages of CKD (Analysis 1.4 (6 studies, 436 participants): MD ‐0.41 ln mg/day, 95% CI ‐0.58 to ‐0.25; I² = 78%) corresponding to an average percentage difference of ‐33.9% (95% CI ‐44.1 to ‐21.9) from the high salt to the low salt intervention.

Three studies could not be entered into the pooled analysis (de Vries 2016; Konishi 2001; Mulhauser 1996). Konishi 2001 reported a reduction in median protein excretion with salt restriction (P = 0.004, 38 CKD participants) equivalent to a 27% reduction. de Vries 2016 found no effect on urinary protein excretion (P = 0.9, 22 post‐transplant participants). Mulhauser 1996 reported no significant difference in proteinuria between the low salt group and the high salt group (P > 0.05 using analysis of covariance adjusted for baseline values), although the analysis method may have been inappropriate as it is likely that data were not normally distributed (mean (range) values g/day end of run‐in low salt: 1 (0.06 to 1.92) High salt: 0.71 (0.06 to 2.31); end of intervention: low salt 0.79 (0.1 to 1.39) high salt 1.14 (0.1 to 5.06)).

Reducing salt intake reduced 24‐hour urinary albumin excretion (Analysis 1.5 (6 studies, 501 participants): MD ‐0.44 ln mg/day, 95% CI ‐0.58 to ‐0.3; I² = 29%; high certainty evidence) corresponding to an average percentage difference of ‐35.6 (95% CI ‐44.2 to ‐25.7) from the high salt to the low salt intervention.

All six studies reporting 24‐hour urinary protein (520 participants) were in the earlier stages of CKD and all but one (LowSALT CKD 2012) were of longer‐term duration.

Most studies reporting 24‐hour urinary albumin were in the earlier stages of CKD (except de Vries 2016 which is in post‐transplant) and most were of longer‐term duration (except LowSALT CKD 2012).

Measures of kidney function

Measures of kidney function reported by the included studies included eGFR, SCr, CrCl, effective renal plasma flow, and filtration fraction. Mulhauser 1996 reported no significant effect on renal parameters including eGFR, CrCl, effective renal plasma flow, filtration fraction or renal vascular resistance (P > 0.05 using analysis of covariance adjusted for baseline values) but data could not be pooled as values were reported as mean and range.

Estimated glomerular filtration rate

Reducing salt intake made little or no difference to eGFR (Analysis 1.6 (7 studies, 559 participants): MD ‐0.82 mL/min/1.73 m², 95% CI ‐2.31 to 0.66; I² = 0%). Of the seven studies that reported eGFR, all but one (de Vries 2016, post‐transplant) were in the earlier stages of CKD and all but one (LowSALT CKD 2012) were of longer‐term duration. Results were similar if including only earlier stages of CKD or in longer‐term studies only. Results did not change when analysed using fixed effects.

Creatinine clearance

Reducing salt intake reduced CrCl (Analysis 1.7 (7 studies, 529 participants): MD ‐4.70 mL/min, 95% CI ‐6.94 to ‐2.47; I² = 0%). DUAAAL 2011 (52 participants) analysed log‐transformed CrCl and reported the results as geometric means with a significant reduction from 72 (62 to 84) mL/min in high salt to 66 (57 to 76) mL/min in low salt (P = 0.002).

The test for subgroup differences showed no difference between short‐ and longer‐term studies in CrCl (Chi² = 0.96, df = 1 (P = 0.33), I² = 0%). Results did not change when analysed using fixed versus random effects.

Serum creatinine

Reducing salt intake increased SCr (Analysis 1.8 (7 studies, 519 participants): MD 4.46 μmol/L, 95% CI 0.08 to 8.83; I² = 24%). de Vries 2016 analysed log transformed SCr and found no effect with salt restriction (MD ln ‐0.02 mg/dL, 95% CI ‐0.03 to 0.07 P = 0.4). LowSALT CKD 2012 reported a significant increase in SCr (μmol/L) from median 149 (IQR 135–230) with high salt intake to 172 (157–276) with low salt intake (P < 0.05).

Effective renal plasma flow

Konishi 2001 (41 participants) reported no significant difference in effective renal plasma flow between low and high salt intake.

Filtration fraction

Reducing salt intake may make little or no difference to filtration fraction in the earlier stages of CKD (Analysis 1.9 (2 studies, 160 participants): MD ‐0.40%, 95% CI ‐0.87 to 0.07; studies = 2; participants = 160; I² = 0%).

Body weight

The evidence is very uncertain about the effect of reducing salt intake on body weight (Analysis 1.10 (12 studies, 759 participants): MD ‐1.32 kg, 95% CI ‐1.94 to ‐0.70; I² = 54%; very low certainty evidence).

CKD stage

In the earlier stages of CKD, reducing salt intake reduced body weight (Analysis 1.10.1 (8 studies, 634 participants): MD ‐1.47 kg, 95% CI ‐2.42 to ‐0.53; I² = 62%). In studies in dialysis patients, reducing salt intake may make little or no difference to body weight (Analysis 1.10.2 (3 studies, 81 participants): MD ‐0.67 kg, 95% CI ‐1.37 to 0.04; I² = 0%). In post‐transplant recipients de Vries 2016 reported reducing salt intake reduced body weight (Analysis 1.10,3 (44 participants): MD ‐2.00 kg, 95% CI ‐3.00 to ‐1.00). The test for subgroup differences showed no differences between the stages of CKD (Chi² = 5.01, df = 2 (P = 0.0819), I² = 60%).

Duration

In both short‐term studies (< 4 weeks) (Analysis 2.6.1 (2 studies, 68 participants): MD ‐0.41 kg, 95% CI ‐0.81 to ‐0.01; I² = 0%) and longer‐term studies (≥ 4 weeks) (Analysis 2.6.2 (10 studies, 691 participants): MD ‐1.53 kg, 95% CI ‐2.02 to ‐1.04; I² = 11%) reducing salt intake reduced body weight. The test for subgroup differences showed no difference between short‐ and longer‐term studies for body weight (Chi² = 12.09, df = 1 (P = 0.0005), I² = 91.7%)

Fluid status

Reducing salt intake may reduce extracellular fluid volume in the earlier stages of CKD (Analysis 1.12 (3 studies, 187 participants): MD ‐0.87 L, 95% CI ‐1.17 to ‐0.58; I² = 0%; low certainty evidence). de Brito-Ashurst 2013 reported a significant reduction in mean total body water in the low salt group of 0.50 L (P < 0.01) and no change in the control group (0.26 L, not statistically significant).

Reducing salt intake did not change interdialytic weight gain (Analysis 1.11 (2 studies, 187 participants): MD ‐0.11 kg, 95% CI ‐0.28 to 0.06; I² = 0%). Power 2010 reported intradialytic weight gain was significantly reduced when sodium restriction was combined with low dialysate but did not report results for sodium restriction alone. Catto 1973 did not report intradialytic weight gain; however differences between mean pre‐ and post‐dialysis weights taken over 15 dialysis sessions were similar in the low salt (2.17 kg) and the high salt (2.21 kg) arms.

Reducing salt intake reduced oedema in the earlier stages of CKD (Analysis 1.13 (2 studies, 191 participants): RR 0.52, 95% CI 0.31 to 0.90; I² = 0%).

Renin‐angiotensin‐aldosterone system and N‐terminal pro‐brain natriuretic peptide stimulation

RAAS stimulation was reported as plasma renin activity, plasma renin, plasma aldosterone and serum aldosterone.

In the earlier stages of CKD reducing salt intake increased plasma renin activity (Analysis 1.14 (2 studies, 142 participants): MD 1.08 ng/mL/h, 95% CI 0.51 to 1.65; studies = 2; participants = 142; I² = 0%) and increased aldosterone (plasma or serum) (Analysis 1.15 (2 studies, 142 participants): MD 6.20 ng/dL, 95% CI 3.82 to 8.58; I² = 0%). Some data on renin and aldosterone could not be pooled; de Vries 2016 reported a significant increase in log transformed plasma aldosterone (MD 0.5 ln pmol/L; 95% CI 0.3 to 0.8; P < 0.001), corresponding to a percentage difference in geometric mean of 65% (35% to 123%) and no significant effect on plasma renin (MD 0.2 ln IU/mL; 95% CI ‐0.1 to 0.5; P = 0.1) in post‐transplant recipients. LowSALT CKD 2012 reported an increase in plasma renin of median 48 pmol/L (interquartile range (IQR) 23.5 to 70.5) and plasma aldosterone by 53.8 mU/L (IQR 4.8 to 74.7) with salt restriction (P < 0.001 for both analyses) in the earlier stages of CKD.

NT‐proBNP results could not be pooled. Two studies in earlier stages of CKD reported a significant reduction in NT‐proBNP with salt restriction; LowSALT CKD 2012 reported that salt restriction reduced NT‐proBNP by 125 pg/mL from median 330 (IQR 167–793) to 205 (124–528) pg/mL (P < 0.05); Vogt 2008 reported salt restriction significantly decreased NT‐proBNP from geometric mean 91 pg/mL (IQR 60 – 137) 62 (IQR 41 – 93) to (P = 0.005). de Vries 2016 reported no significant effect of salt restriction on log‐transformed NT‐proBNP (‐0.1 ln ng/L,95% CI ‐0.4 to 0.1; P = 0.3) in post‐transplant participants.

Change in antihypertensive regimen

Seven studies (664 participants) reported incidence of reduction in antihypertensives with salt restriction (de Vries 2016; DUAAAL 2011; ESPECIAL 2014; Keven 2006; LowSALT CKD 2012; Saran 2017; ViRTUE‐CKD 2016). Three of these reported no changes in either low salt or high salt groups (DUAAAL 2011; ESPECIAL 2014; ViRTUE‐CKD 2016).

The evidence is very uncertain about the effect of reducing salt intake on antihypertensive medication dosage (Analysis 1.16 (7 studies, 664 participants): RR 2.45, 95% CI 0.98 to 6.08; I² = 12%; very low certainty evidence). Antihypertensive medication dosage were reported by four studies (228 participants; de Vries 2016; Keven 2006; LowSALT CKD 2012; Saran 2017). This may have reduced the effect size of sodium restriction on BP in these studies. There were 20 instances of reduction in antihypertensive dosage in low salt group versus eight in the high salt group. Six of the eight instances in the high salt arm occurred in Saran 2017, and these were mostly in the second arm of the cross‐over study. Saran 2017 also experienced protocol violations with respect to treatment order; 10 of the 56 patients were not receiving the correct treatment in the second arm (six in the first arm), therefore this may have impacted anti‐hypertensive dose. If Saran 2017 is excluded from the analysis, the effect size is larger (RR 5.38, 95% CI 1.45 to 20.01 12 events in low salt and two in high salt arms).

ESMO 2017 reported the number of antihypertensive medications was not significantly changed with salt restriction (MD 0.0 95% ‐0.3 to 0.2) but did not report change in dosage. Rodrigues Telini 2014 reported there was no significant difference in the number of antihypertensive medications between the groups at follow‐up.

Some studies did not report change in antihypertensive medications but it is reasonable to assume medication changes did not occur due to their study protocols; the study protocols for Kwakernaak 2014 and Vogt 2008 specified keeping medications stable; Konishi 2001 refrained from the use of antihypertensive medications during the study; Ruilope 1992a specified that no medications were given during run‐in and it is likely that this continued (besides the co‐intervention of 240 mg verapamil); Mulhauser 1996 excluded participants taking antihypertensive medications.

Fine 1997 reported that there were no increases to medication in the 10 hypertensive patients during the high salt intervention, but it is unclear whether changes occurred in the low salt intervention or in the other 10 participants. de Brito‐Ashurst 2013 study protocol included salt restriction “in addition to standard hypertension medication management” therefore there may have been changes to medications but measurement of this was not specified. Power 2010 measured antihypertensive dosage but did not report results.

Adverse effects

Potential adverse effects reported included symptomatic hypotension, cramps, and serum cholesterol.

Reducing salt intake may increase symptomatic hypotension (Analysis 1.17 (4 studies, 478 participants): RR 6.70, 95% CI 2.40 to 18.69; I² = 0%; moderate certainty evidence). These were mostly considered mild (ViRTUE‐CKD 2016) or were resolved by lowering the dose of antihypertensive medication (de Vries 2016; LowSALT CKD 2012). This is a potential adverse effect associated with salt restriction, although one that could be rectified by reducing the antihypertensive dose.

Power 2010 reported that sodium restriction increased frequency, but not severity, of cramps by 21% (P = 0.04). Catto 1973 reported that administration of slow sodium during dialysis (versus placebo) resulted in a 26% reduction in the frequency (P < 0.05) and 30% in the severity (P < 0.02) of muscle cramps recorded over 15 dialysis sessions.

Reducing salt intake did not change total cholesterol (Analysis 1.18 (7 studies, 473 participants): MD ‐0.11 mmol/L, 95% CI ‐0.31 to 0.10; I² = 0%)

Sensitivity analysis

The funnel plots of analyses for sodium intake/excretion (Figure 4), systolic BP (Figure 5) and diastolic BP (Figure 6) were investigated. Studies giving rise to asymmetry for systolic BP (BalanceWise‐PD 2011) and diastolic BP (BalanceWise‐PD 2011; Rodrigues Telini 2014) were removed with no change in direction or effect size.

Figure 4

Funnel plot of comparison: 1 CKD stage, outcome: 1.1 Sodium intake/excretion [mmol/d].

Figure 5

Funnel plot of comparison: 1 CKD stage, outcome: 1.2 Systolic blood pressure [mm Hg].

Figure 6

Funnel plot of comparison: 1 CKD stage, outcome: 1.3 Diastolic blood pressure [mm Hg].

Studies with high risk of bias in more than one domain (BalanceWise‐PD 2011; Saran 2017) were removed. The differences in analyses were mostly minimal except when Saran 2017 was eliminated from antihypertensive dosage reduction analysis as discussed previously (analyses not shown).

Discussion

Summary of main results

We included 21 studies (1197 randomised participants), 12 in the earlier stages of CKD (779 randomised participants), seven in dialysis (363 randomised participants) and two in post‐transplant (55 randomised participants). Included studies had a median study duration of seven weeks (range 1 to 36 weeks) and compared two levels of sodium intake achieved via dietary counselling to reduce salt (13 studies); sodium supplements (six studies) providing food (one study) or unspecified methods (one study).

Reducing salt reduced BP in all stages of CKD. We found that reducing salt intake by approximately 4.2 g/day (73.52 mmol or 1690 mg sodium) lowered BP by 7/4 mm Hg in people with CKD and proteinuria/albuminuria by 34% to 36% in people in the earlier stages of CKD. This is a clinically significant reduction in BP, comparable to expectations of administering a single antihypertensive drug (Law 2009).

In non‐dialysed, non‐transplanted people with CKD, reducing salt intake by 4.8 g/day (84 mmol or 1930 mg sodium) for a median of six weeks duration reduced BP by 6.1/3.5 mm Hg. In dialysed people with CKD reducing salt by 1.6 g/day (27 mmol or 620 mg sodium) for a median of 16 weeks duration reduced BP by 6.3/3.5 mm Hg. In people who had undergone transplantation (two studies 6 and 12 week duration), reducing salt by 5.6 g/day (97 mmol or 2230 mg sodium) reduced BP by 11.9/7.1 mm Hg.

In studies of less than four weeks duration, reducing salt by 6.4 g/day (111 mmol or 2550 mg sodium) reduced BP by 7.9/3.7mm Hg. In studies of four weeks or more duration, reducing salt by 3.7 g/day (64.53 mmol or 1480 mg sodium) reduced BP by 6.8/4.0 mm Hg.

The evidence is very uncertain about the effect of lower salt intake on weight. We found body weight was reduced by 1.3 kg with salt reduction, with a larger reduction in studies over 4 weeks (1.5 kg) than studies under 4 weeks duration (0.4 kg). However, it was not possible to determine the extent to which this weight loss was due to fluid volume, lean tissue, or body fat. Salt restriction resulted in reduced extracellular fluid volume by 0.87 L and a 48% reduced risk of oedema in non‐transplanted people with CKD. Reducing salt intake did not significantly change inter‐dialytic weight gain. Evidence for effects on fluid status was not available for transplanted people with CKD. Reductions in body weight were seen in all stages of CKD with 1.5 kg weight loss in non‐dialysed, non‐transplanted CKD, 0.7 kg in dialysed CKD and 2 kg in transplanted CKD.

Reducing salt intake reduced 24‐hour urinary protein excretion in the earlier stages of CKD by ‐0.41 ln mg/day (95% CI ‐0.58 to ‐0.25) which corresponded to an average percentage difference of ‐33.9% (95% CI ‐44.1 to ‐21.9). Reducing salt intake may reduce extracellular fluid volume and oedema in the earlier stages of CKD but did not alter interdialytic weight gain.

It was found that eGFR was not significantly changed by salt restriction, but salt restriction decreased CrCl by 4.7 ml/min and increased SCr by 4.5 μmol/L. These effects were more pronounced in studies under four weeks duration than those four or more weeks duration, but the difference by duration was not statistically significant.

While reduced salt intake may increase symptomatic hypotension, the evidence is very uncertain about the effect of reducing salt intake on antihypertensive medication dosage.

Overall completeness and applicability of evidence

We aimed to evaluate the benefits and harms of altering dietary salt intake for people with CKD. We could not assess the effect of restricting salt intake on endpoints such as death, cardiovascular events of progression to ESKD in people with CKD because there were no RCTs of adequate size or duration to examine these outcomes. This limitation has been noted in previous reviews in non‐CKD populations (Hooper 2002; Suckling 2010), and may be also in part due to the resources and practical aspects of providing the ongoing and regular support that is needed to achieve long‐term adherence to a sodium‐restricted diet (McMahon 2012a). In practice, supporting patients to achieve this requires an individualised approach tailored to support patients to improve their motivation, knowledge and develop personal coping skills to prevent behaviour change decay over time. More research on these elements are needed. In this review, we identified some ongoing studies with longer intervention durations meaning that we may be able to assess the longer‐term effects on salt restriction in people with CKD in future updates of this review.

While nearly all included studies reported effects of salt restriction on dietary salt intake/excretion and BP, there were limited evidence for effects on kidney function, proteinuria/albuminuria, and fluid status. Subgroup analyses were not possible for many outcomes due to the small number of studies reporting these outcomes. We examined results by stage of CKD and study duration; we were unable to explore other potential sources of heterogeneity such as presence of diabetes mellitus or albuminuria/proteinuria, primary kidney disease, or antihypertensive regimen.

There was a substantial degree of heterogeneity among study results for some outcomes, including sodium intake/excretion, systolic BP, and weight, although this was reduced in subgroup analyses in most cases. A contributor to heterogeneity was inclusion of people across the stages of CKD, including early stage CKD (non‐dialysis), dialysis, and transplant populations. Sodium handling in people with mild CKD is likely to differ considerably from those with severe kidney dysfunction, people on dialysis, and kidney transplant recipients. We analysed outcomes by CKD stage where possible, however this subgroup analysis was limited due to the small number of included studies, particularly in the more advanced stages of CKD. Only two studies included post‐transplant participants and seven in dialysis (two with people receiving PD, four with people receiving HD, one dialysis modality not specified). Other salt restriction studies have been conducted in people undergoing dialysis, but these are either observational (Kayikcioglu 2009), non‐randomised (Ang 1999; Osanai 2002) or used a concomitant intervention in the sodium restricted group (Rupp 1978) and were not eligible for inclusion in this review.

There were important differences in methodology between short‐ (fewer than four weeks) and longer‐term studies (four weeks and more) that limited subgroup analysis according to study duration. Short‐term studies achieved a larger reduction in salt (6.4 g/day; 111 mmol or 2550 mg sodium) than studies of four weeks or longer reducing salt by (3.7 g/day; 64.53 mmol or 1480 mg sodium) but achieved similar reductions in BP (8/4 and 7/4 mm Hg respectively). However short‐term studies used either supplemental sodium (Konishi 2001; LowSALT CKD 2012) or full meal provision (Ruilope 1992a), therefore dietary confounders may have been more tightly controlled than in longer‐term studies which most commonly compared dietary education to reduce salt versus usual diet. A limitation in two of the three short‐term studies was that neither employed a washout period (Konishi 2001; Ruilope 1992a). Considering that these studies had intervention durations of only one week, carry‐over effect may have influenced study results. Ruilope 1992a, a short‐term study, began a new antihypertensive medication on day one of the study, further increasing risk of carry‐over effect.

It is likely that differing magnitude of salt restriction impacted heterogeneity. A previous analysis showed a dose‐response relationship between salt reduction and BP reduction (He 2003). We could not examine differential effects by magnitude of sodium reduction achieved as studies did not report results separately by level of sodium reduction achieved.

Quality of the evidence

We assessed the quality of study evidence using standard risks of bias domains within the Cochrane tool together with GRADE methodology. Confidence in evidence was high for systolic and diastolic BP and albuminuria. Confidence in evidence was low for sodium intake/excretion (substantial heterogeneity), extracellular fluid volume (uncertain or high risk of bias; few participants across all studies) and symptomatic hypotension (few participants across all studies; wide confidence interval range that include benefit or harm). Confidence in evidence was very low for weight (uncertain or high risk of bias; moderate heterogeneity; indirectness of outcome) and reduction in antihypertensive dose (high risk of bias; wide confidence interval range that include benefit or harm).

Potential bias from confounders were mostly due to lack of measurement or failing to adequately report or account for the impact of changes in medications and/or dietary changes such as potassium intake, protein intake and/or weight loss between interventions. Reduction in anti‐hypertensive medications may have led to underestimations of the effect of salt restriction on BP and proteinuria. Studies that provided supplementary sodium (Catto 1973; Doulton 2007; Fine 1997; LowSALT CKD 2012; Mulhauser 1996; Ruilope 1992a) to manage sodium intake were at lower risk of bias from dietary confounders because other dietary factors were likely to remain stable over the study period (McMahon 2012a). Previous research has found that when dietary advice is given about reducing sodium intake, other factors such as energy and potassium intake can also change (Korhonen 2000). This means that studies relying on dietary advice to manage sodium intake (BalanceWise‐HD 2013; BalanceWise‐PD 2011; de Brito‐Ashurst 2013; de Vries 2016; DUAAAL 2011; ESMO 2017; ESPECIAL 2014; Keven 2006; Kwakernaak 2014; Rodrigues Telini 2014; Saran 2017; ViRTUE‐CKD 2016; Vogt 2008) may be at a higher risk of bias from dietary confounders.

It is widely accepted that potassium intake affects BP. Nine of the 19 studies that reported BP as an outcome measured change in potassium intake either directly or indirectly and reported the findings. Seven studies (de Brito‐Ashurst 2013; de Vries 2016; ESMO 2017; Kwakernaak 2014; LowSALT CKD 2012; Mulhauser 1996; ViRTUE‐CKD 2016) reported potassium intake/excretion as unchanged with sodium restriction. DUAAAL 2011 and Ruilope 1992a reported small, but significant, reductions in urinary potassium excretion in the sodium‐restricted groups (3 to 4 mmol/day), which is likely to reflect reduction in urinary volume.

Body weight has also been reported to affect BP and proteinuria (Siebenhofer 2016). Given that most studies did not report change in fluid status, we could not determine the degree to which body weight change observed in included studies was due to reduction in fluid volume or body fat. The latter could introduce bias by overestimating the effect of salt restriction on BP and proteinuria.

Saran 2017 had high risk of bias across several domains due to protocol violation meaning that a large proportion of participants did not receive treatments in the allocated order or did not receive both treatments, incomplete outcome data and potential conflict of interest (due to two study authors holding stock in a company that makes the tool used for primary outcome measurement). Results for were similar when Saran 2017 was excluded for all outcomes anti‐hypertensive dosage reduction; the magnitude of effect was increased when Saran 2017 was excluded.

Potential biases in the review process

This review was conducted according to a pre‐specified protocol and used a highly sensitive search strategy. Inclusion/exclusion and risk of bias was conducted by two independent review authors and we considered evidence certainty in interpretation of the results. All efforts were made to minimise bias inherent in the review process. Study inclusion and risk of bias assessment were carried out by two authors working independently.

In the 2015 review, we contacted all study authors for additional information to inform our risk of bias assessment and received further information for five of the eight included studies (de Brito‐Ashurst 2013; DUAAAL 2011; Konishi 2001; LowSALT CKD 2012; Vogt 2008). Of the 13 additional studies included in this review update, we did not require further information for one study (Kwakernaak 2014); we could not find current contact details to request further information for three studies (Catto 1973; Doulton 2007; Mulhauser 1996); and we contacted study authors of the remaining nine studies and received further information for five studies (BalanceWise‐HD 2013; BalanceWise‐PD 2011; ESMO 2017; ESPECIAL 2014; ViRTUE‐CKD 2016).

We applied a search strategy to include both published and unpublished studies, and were able to include one study where main findings were unpublished (BalanceWise‐PD 2011) and two studies where main findings were published as conference abstracts only (Doulton 2007; Power 2010).

Agreements and disagreements with other studies or reviews

Previous version of this review

This review update included 21 studies with 725 participants in the low salt group and 725 in the high salt group. This is a considerable increase compared with our 2015 review which included eight studies with 224 participants in the low salt group and 218 in the high salt group (McMahon 2015), and we were able to include additional outcomes in this update (proteinuria, albuminuria, inter‐dialytic weight gain). A further difference from the 2015 review is that we included results from paired analyses where available which in most instances led to narrower confidence intervals than when results from unpaired analyses were included. Some key differences in the results of this update compared to the 2015 review are discussed below.

Our 2015 review included six studies in the earlier stages of CKD, one in dialysis, and one post‐transplant. This update includes an additional six studies in the earlier stages of CKD (12 studies total), six in dialysis (seven studies total) and one post‐transplant (two studies total). The additional studies in ESKD in this update improves generalisability of the findings to participants at this stage of kidney disease.

We found a smaller reduction on sodium intake/excretion in included studies in this update (mean 74 mmol/day) compared with the 2015 review (mean 106 mmol/day). This is likely due to an increase in the number of studies examining effectiveness of behavioural and education‐based approaches to reducing salt intake, rather than the efficacy of salt reduction in a controlled setting. Included studies also had a longer intervention duration in this update compared with the 2015 review. Four of the eight included studies in the 2015 review used education to reduce sodium intake compared to 13 of the 21 studies in this update. Of these studies, those in the 2015 review tended to have higher baseline sodium intakes (mean values ranging from 177 to 261 mmol/day) than those added in this update (mean values ranging from 98 to 175 mmol/day), however target sodium intakes for the low salt group were similar, ranging between 50 to 100 mmol/day. This likely contributed to the smaller reduction in sodium seen in this review update versus the 2015 review. The inclusion of studies with lower sodium reduction also likely explains the slightly lower effect on systolic BP. We originally demonstrated that reducing salt intake by approximately 6 g/day (100 mmol or 2300 mg sodium/day) lowered BP by 9/4 mm Hg in people with CKD. In this update, we found that reducing salt intake by approximately 4.2 g/day (73.52 mmol or 1690 mg sodium) lowered BP by 7/4 mm Hg.

In the 2015 review, we did not meta‐analyse the effects on proteinuria and albuminuria due to heterogeneity in outcome reporting. Instead we reported the range of effect in the four studies reporting proteinuria and one reporting albuminuria. In this update, we meta‐analysed the effect of salt restriction on proteinuria (six studies) and albuminuria (six studies) mostly in the earlier stages of CKD, and our findings were consistent with those we reported in the 2015 review.

This update included three additional studies (four total) reporting on extracellular fluid, one (two total) reporting on presence of oedema and four (four total) reporting on intra‐dialytic weight gain. We found significant reductions on extracellular fluid volume and presence of oedema which is an important finding as fluid overload is a strong predictor of cardiovascular events and kidney function decline in CKD (Essig 2008). We also found significant changes in CrCl and body weight in this update, while these were not significantly altered in the 2015 review, although these changes are likely due to changes in BP, proteinuria/albuminuria and fluid status and are unlikely to be clinically significant.

Other reviews in CKD

There have been three reviews of trials of salt restriction in people with CKD published since the 2015 review, including: 1) a narrative review of seven RCTs in non‐dialysed, non‐transplanted CKD participants (Nerbass 2018); 2) a systematic review of 12 randomised and non‐randomised trials achieving at least one gram salt reduction in dialysis participants, and meta‐analysis of the four RCTs (Cole 2019); and 3) a systematic review and meta‐analysis of 11 RCTs in non‐dialysed CKD patients (Garofalo 2018). Agreements and disagreements with the findings of the meta‐analyses are discussed below.

Cole 2019 reported that salt restriction in four RCTs in dialysis reduced systolic/diastolic BP by 8/4. mm Hg. This is slightly higher than the difference in BP in our meta‐analysis of five RCTs in dialysis (8/3 mm Hg), likely due to our inclusion of the BalanceWise‐PD 2011 Study which did not achieve a significant reduction in sodium intake or BP.

Garofalo 2018 reported that a mean 80 mmol/day reduction in sodium excretion reduced ambulatory BP by 6/2 mm Hg and clinic‐assessed BP by 5/3 in 11 RCTs in non‐dialysed CKD patients, similar to our findings of a 6/3 reduction in 12 RCTs in the earlier stages of CKD, and resulted in a 29% mean reduction in proteinuria, similar to our results (34% reduction), and 14% reduction in albuminuria, whereas we found a 37% reduction in our analysis of the same five studies. These are likely due to differences in meta‐analysis methods. We analysed albuminuria as log‐transformed values (which in some cases was made possible only by provision of raw data by study authors), because these data are usually not normally distributed and effect is highly dependent on their baseline values. Garofalo 2018 converted medians and inter‐quartile ranges to means and SD and pooled log‐transformed with untransformed data. Three studies were weighted very low due to wide CI suggesting no effect on albuminuria (Kwakernaak 2014; LowSALT CKD 2012; ViRTUE‐CKD 2016); however these studies all reported clinically and statistically significant effects on albuminuria (P < 0.001). It is likely that the study by Saran 2017 was weighted heavily because log‐transformed values were used, leading to narrower confidence intervals.

Other reviews in non‐CKD

Previous reviews investigating the effect of salt restriction have been conducted in people with normal kidney function, and these consistently show that reducing dietary sodium intake reduces BP, although magnitudes vary (Graudal 2020; He 2013; Hernandez 2018; Hooper 2002). Across these reviews, dietary salt reduction generally has a greater effect on people who are hypertensive (Graudal 2020; He 2013; Hooper 2002). We could not conduct subgroup analyses by hypertensive status as most participants in the included studies were hypertensive and it was rare to report findings separately for hypertensive and normotensive participants. We found a reduction in BP of approximately 7/4 mm Hg with salt reduction of approximately 74 mmol. This is a similar to that found in hypertensive people without CKD; Graudal 2020 reported mean reduction of 6/3 mm Hg with median salt reduction of 145 mmol; Hernandez 2018 found that low‐sodium salt substitutes decreased systolic BP by 8/4 mm Hg, with a 27 mmol decrease in urinary sodium and 13 mmol increase in urinary potassium.

Graudal 2020 also identified that increased serum cholesterol level was an adverse effect in a meta‐analysis of sodium restriction. We found no significant change in total cholesterol in relation to salt restriction in people with CKD.

In a pooled meta‐analysis of people with diabetes, Suckling 2010 reported that CrCl was significantly reduced (‐6.33 mL/min, 95% CI ‐10.47 to ‐2.19) with salt restriction; eGFR was not significantly changed (MD ‐1.92; 95% CI ‐4.49 to 0.64). Our update found similar, with CrCl significantly reducing by 4.46 and that eGFR did not change significantly with salt restriction. It is thought that reductions in CrCl with salt restriction occur as a result of hyperfiltration paradoxically decreasing risk of kidney disease (Allen 1997).

Suckling 2010 found mixed results for the effects of salt restriction on proteinuria in people with diabetes. Aburto 2013 reported urinary protein excretion and albuminuria was reduced with lower sodium intake in a systematic review including two studies in the general population. Our review demonstrated proteinuria to significantly reduce by 34% following salt restriction; this difference was expected as people with CKD are more susceptible to proteinuria.

Figure 1

Study flow diagram

Navigate to figure in ReviewOpen in new tab

Figure 2

Risk of bias summary: review authors' judgements about each risk of bias item for each included study.

Navigate to figure in ReviewOpen in new tab

Figure 3

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.

Navigate to figure in ReviewOpen in new tab

Figure 4

Funnel plot of comparison: 1 CKD stage, outcome: 1.1 Sodium intake/excretion [mmol/d].

Navigate to figure in ReviewOpen in new tab

Figure 5

Funnel plot of comparison: 1 CKD stage, outcome: 1.2 Systolic blood pressure [mm Hg].

Navigate to figure in ReviewOpen in new tab

Figure 6

Funnel plot of comparison: 1 CKD stage, outcome: 1.3 Diastolic blood pressure [mm Hg].

Navigate to figure in ReviewOpen in new tab

Analysis 1.1

Comparison 1: CKD stage, Outcome 1: Sodium intake/excretion

Navigate to figure in ReviewOpen in new tab

Analysis 1.2

Comparison 1: CKD stage, Outcome 2: Systolic blood pressure

Navigate to figure in ReviewOpen in new tab

Analysis 1.3

Comparison 1: CKD stage, Outcome 3: Diastolic blood pressure

Navigate to figure in ReviewOpen in new tab

Analysis 1.4

Comparison 1: CKD stage, Outcome 4: Proteinuria [ln mg/d]

Navigate to figure in ReviewOpen in new tab

Analysis 1.5

Comparison 1: CKD stage, Outcome 5: Albuminuria [ln mg/d]

Navigate to figure in ReviewOpen in new tab

Comparison 1: CKD stage, Outcome 6: eGFR [mL/min/1.73 m2]

Analysis 1.6

Comparison 1: CKD stage, Outcome 6: eGFR [mL/min/1.73 m²]

Navigate to figure in ReviewOpen in new tab

Analysis 1.7

Comparison 1: CKD stage, Outcome 7: Creatinine clearance

Navigate to figure in ReviewOpen in new tab

Analysis 1.8

Comparison 1: CKD stage, Outcome 8: Serum creatinine

Navigate to figure in ReviewOpen in new tab

$Comparison 1: CKD stage, Outcome 9: Filtration fraction$

Analysis 1.9

Comparison 1: CKD stage, Outcome 9: Filtration fraction

Navigate to figure in ReviewOpen in new tab

Analysis 1.10

Comparison 1: CKD stage, Outcome 10: Weight

Navigate to figure in ReviewOpen in new tab

Analysis 1.11

Comparison 1: CKD stage, Outcome 11: Inter‐dialytic weight gain

Navigate to figure in ReviewOpen in new tab

Analysis 1.12

Comparison 1: CKD stage, Outcome 12: Extracellular fluid volume

Navigate to figure in ReviewOpen in new tab

Analysis 1.13

Comparison 1: CKD stage, Outcome 13: Presence of oedema

Navigate to figure in ReviewOpen in new tab

Analysis 1.14

Comparison 1: CKD stage, Outcome 14: Plasma renin activity

Navigate to figure in ReviewOpen in new tab

Analysis 1.15

Comparison 1: CKD stage, Outcome 15: Aldosterone (plasma or serum)

Navigate to figure in ReviewOpen in new tab

Analysis 1.16

Comparison 1: CKD stage, Outcome 16: Reduction in antihypertensive dose

Navigate to figure in ReviewOpen in new tab

Analysis 1.17

Comparison 1: CKD stage, Outcome 17: Symptomatic hypotension

Navigate to figure in ReviewOpen in new tab

Analysis 1.18

Comparison 1: CKD stage, Outcome 18: Total cholesterol

Navigate to figure in ReviewOpen in new tab

Analysis 2.1

Comparison 2: Duration, Outcome 1: Sodium intake/excretion

Navigate to figure in ReviewOpen in new tab

Analysis 2.2

Comparison 2: Duration, Outcome 2: Systolic blood pressure

Navigate to figure in ReviewOpen in new tab

Analysis 2.3

Comparison 2: Duration, Outcome 3: Diastolic blood pressure

Navigate to figure in ReviewOpen in new tab

Analysis 2.4

Comparison 2: Duration, Outcome 4: Creatinine clearance

Navigate to figure in ReviewOpen in new tab

Analysis 2.5

Comparison 2: Duration, Outcome 5: Serum creatinine

Navigate to figure in ReviewOpen in new tab

Analysis 2.6

Comparison 2: Duration, Outcome 6: Weight

Navigate to figure in ReviewOpen in new tab

Summary of findings 1. Lower salt intake versus higher salt intake for participants with chronic kidney disease

Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)
Lower salt intake versus higher salt intake for participants with CKD
Patient or population: participants with CKD Setting: any Intervention: lower salt intake Comparison: higher salt intake
Outcomes	Risk with higher salt intake	Risk with lower salt intake	Relative effect (95% CI)	No. of participants (studies)	Certainty of the evidence (GRADE)
Systolic blood pressure	The mean SBP in the lower salt intake group was 6.91 mm Hg lower (8.82 lower to 4.99 lower) than the higher salt intake group		‐	1261 (19)	⊕⊕⊕⊕ HIGH
Diastolic blood pressure	The mean DBP in the lower salt intake group was 3.91 mm Hg lower (4.8 lower to 3.02 lower) than the higher salt intake group		‐	1257 (19)	⊕⊕⊕⊕ HIGH
Albuminuria	The mean albuminuria in the lower salt intake group was 36%¹ lower (44% lower to 26% lower) than the higher salt intake group		‐	501 (6)	⊕⊕⊕⊕ HIGH
Extracellular fluid volume	The mean extracellular fluid volume in the lower salt intake group was 0.87 L lower (1.17 lower to 0.58 lower) than the higher salt intake group		‐	187 (3)	⊕⊕⊝⊝ LOW ^{2 3}
Weight	The mean weight in the lower salt intake group was 1.32 kg lower (1.94 lower to 0.7 lower) than the higher salt intake group		‐	759 (12)	⊕⊝⊝⊝ VERY LOW ^{2 4 5}
Reduction in antihypertensive dose	24 per 1,000	58 per 1,000 (23 to 144)	RR 2.45 (0.98 to 6.08)	664 (7)	⊕⊝⊝⊝ VERY LOW ^{6 7}
Symptomatic hypotension	8 per 1,000	56 per 1,000 (20 to 156)	RR 6.70 (2.40 to 18.69)	478 (6)	⊕⊕⊕⊝ MODERATE ³
*The risk in the intervention group (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CKD: chronic kidney disease; CI: confidence interval; SBP: systolic blood pressure; DBP: diastolic blood pressure; RR: risk ratio
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹ Albuminuria was analysed as log‐transformed (mg/day) and expressed as a percentage reduction ² Evidence certainty downgraded one level for uncertain or high risk of bias ³ Evidence certainty downgraded one level for few events and/or participants across all studies ⁴ Evidence certainty downgraded one level for moderate heterogeneity ⁵ Evidence certainty downgraded one levels for indirectness of outcome ⁶ Evidence certainty downgraded two levels for a high risk of bias ⁷ Evidence certainty downgraded one level for wide confidence interval range that include benefit or harm

Summary of findings 1. Lower salt intake versus higher salt intake for participants with chronic kidney disease

Navigate to table in Review

Comparison 1. CKD stage

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Sodium intake/excretion Show forest plot	18	1383	Mean Difference (IV, Random, 95% CI)	‐73.51 [‐92.76, ‐54.27]

1.1.1 CKD	12	1048	Mean Difference (IV, Random, 95% CI)	‐83.81 [‐104.54, ‐63.08]
1.1.2 Dialysis	4	259	Mean Difference (IV, Random, 95% CI)	‐27.17 [‐48.59, ‐5.76]
1.1.3 Post‐transplant	2	76	Mean Difference (IV, Random, 95% CI)	‐96.99 [‐148.10, ‐45.89]
1.2 Systolic blood pressure Show forest plot	19	1261	Mean Difference (IV, Random, 95% CI)	‐6.91 [‐8.82, ‐4.99]

1.2.1 CKD	12	1036	Mean Difference (IV, Random, 95% CI)	‐6.10 [‐8.11, ‐4.08]
1.2.2 Dialysis	5	149	Mean Difference (IV, Random, 95% CI)	‐6.32 [‐11.04, ‐1.60]
1.2.3 Post‐transplant	2	76	Mean Difference (IV, Random, 95% CI)	‐11.94 [‐15.77, ‐8.11]
1.3 Diastolic blood pressure Show forest plot	19	1257	Mean Difference (IV, Random, 95% CI)	‐3.91 [‐4.80, ‐3.02]

1.3.1 CKD	12	1032	Mean Difference (IV, Random, 95% CI)	‐3.47 [‐4.28, ‐2.65]
1.3.2 Dialysis	5	149	Mean Difference (IV, Random, 95% CI)	‐3.46 [‐6.39, ‐0.54]
1.3.3 Post‐transplant	2	76	Mean Difference (IV, Random, 95% CI)	‐7.13 [‐9.46, ‐4.79]
1.4 Proteinuria [ln mg/d] Show forest plot	6		Mean Difference (IV, Random, 95% CI)	‐0.41 [‐0.58, ‐0.25]

1.4.1 CKD	6		Mean Difference (IV, Random, 95% CI)	‐0.41 [‐0.58, ‐0.25]
1.5 Albuminuria [ln mg/d] Show forest plot	6		Mean Difference (IV, Random, 95% CI)	‐0.44 [‐0.58, ‐0.30]

1.5.1 CKD	5		Mean Difference (IV, Random, 95% CI)	‐0.47 [‐0.60, ‐0.34]
1.5.2 Post‐transplant	1		Mean Difference (IV, Random, 95% CI)	‐0.03 [‐0.63, 0.57]
1.6 eGFR [mL/min/1.73 m²] Show forest plot	7	559	Mean Difference (IV, Random, 95% CI)	‐0.82 [‐2.31, 0.66]

1.6.1 CKD	6	515	Mean Difference (IV, Random, 95% CI)	‐1.00 [‐2.85, 0.85]
1.6.2 Post‐transplant	1	44	Mean Difference (IV, Random, 95% CI)	‐0.50 [‐3.00, 2.00]
1.7 Creatinine clearance Show forest plot	7	529	Mean Difference (IV, Random, 95% CI)	‐4.70 [‐6.94, ‐2.47]

1.7.1 CKD	6	485	Mean Difference (IV, Random, 95% CI)	‐5.06 [‐7.40, ‐2.72]
1.7.2 Post‐transplant	1	44	Mean Difference (IV, Random, 95% CI)	‐1.00 [‐8.50, 6.50]
1.8 Serum creatinine Show forest plot	7	519	Mean Difference (IV, Random, 95% CI)	4.46 [0.08, 8.83]

1.8.1 CKD	6	487	Mean Difference (IV, Random, 95% CI)	4.86 [0.95, 8.76]
1.8.2 Post‐transplant	1	32	Mean Difference (IV, Random, 95% CI)	‐11.00 [‐32.30, 10.30]
1.9 Filtration fraction Show forest plot	2	160	Mean Difference (IV, Random, 95% CI)	‐0.40 [‐0.87, 0.07]

1.9.1 CKD	2	160	Mean Difference (IV, Random, 95% CI)	‐0.40 [‐0.87, 0.07]
1.10 Weight Show forest plot	12	759	Mean Difference (IV, Random, 95% CI)	‐1.32 [‐1.94, ‐0.70]

1.10.1 CKD	8	634	Mean Difference (IV, Random, 95% CI)	‐1.47 [‐2.42, ‐0.53]
1.10.2 Dialysis	3	81	Mean Difference (IV, Random, 95% CI)	‐0.67 [‐1.37, 0.04]
1.10.3 Post‐transplant	1	44	Mean Difference (IV, Random, 95% CI)	‐2.00 [‐3.00, ‐1.00]
1.11 Inter‐dialytic weight gain Show forest plot	2	187	Mean Difference (IV, Random, 95% CI)	‐0.11 [‐0.28, 0.06]

1.11.1 Dialysis	2	187	Mean Difference (IV, Random, 95% CI)	‐0.11 [‐0.28, 0.06]
1.12 Extracellular fluid volume Show forest plot	3	187	Mean Difference (IV, Random, 95% CI)	‐0.87 [‐1.17, ‐0.58]

1.12.1 CKD	2	148	Mean Difference (IV, Random, 95% CI)	‐0.89 [‐1.20, ‐0.59]
1.12.2 Dialysis	1	39	Mean Difference (IV, Random, 95% CI)	‐0.30 [‐1.87, 1.27]
1.13 Presence of oedema Show forest plot	2	191	Risk Ratio (M‐H, Random, 95% CI)	0.52 [0.31, 0.90]

1.13.1 CKD	2	191	Risk Ratio (M‐H, Random, 95% CI)	0.52 [0.31, 0.90]
1.14 Plasma renin activity Show forest plot	2	142	Mean Difference (IV, Random, 95% CI)	1.08 [0.51, 1.65]

1.14.1 CKD	2	142	Mean Difference (IV, Random, 95% CI)	1.08 [0.51, 1.65]
1.15 Aldosterone (plasma or serum) Show forest plot	2	142	Mean Difference (IV, Random, 95% CI)	6.20 [3.82, 8.58]

1.15.1 CKD	2	142	Mean Difference (IV, Random, 95% CI)	6.20 [3.82, 8.58]
1.16 Reduction in antihypertensive dose Show forest plot	7	664	Risk Ratio (M‐H, Random, 95% CI)	2.45 [0.98, 6.08]

1.16.1 CKD	5	588	Risk Ratio (M‐H, Random, 95% CI)	1.53 [0.62, 3.78]
1.16.2 Post‐transplant	2	76	Risk Ratio (M‐H, Random, 95% CI)	7.52 [1.45, 39.04]
1.17 Symptomatic hypotension Show forest plot	6	478	Risk Ratio (M‐H, Random, 95% CI)	6.70 [2.40, 18.69]

1.17.1 CKD	5	433	Risk Ratio (M‐H, Random, 95% CI)	6.26 [2.08, 18.81]
1.17.2 Post‐transplant	1	45	Risk Ratio (M‐H, Random, 95% CI)	10.54 [0.62, 180.07]
1.18 Total cholesterol Show forest plot	7	473	Mean Difference (IV, Random, 95% CI)	‐0.11 [‐0.31, 0.10]

1.18.1 CKD	5	390	Mean Difference (IV, Random, 95% CI)	‐0.06 [‐0.30, 0.19]
1.18.2 Dialysis	1	39	Mean Difference (IV, Random, 95% CI)	‐0.42 [‐0.99, 0.15]
1.18.3 Post‐transplant	1	44	Mean Difference (IV, Random, 95% CI)	‐0.08 [‐0.54, 0.39]

Comparison 1. CKD stage

Navigate to table in Review

Comparison 2. Duration

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Sodium intake/excretion Show forest plot	18	1383	Mean Difference (IV, Random, 95% CI)	‐73.51 [‐92.76, ‐54.27]

2.1.1 Short‐term (< 4 weeks)	3	142	Mean Difference (IV, Random, 95% CI)	‐111.29 [‐134.25, ‐88.32]
2.1.2 Longer‐term (≥ 4 weeks)	15	1241	Mean Difference (IV, Random, 95% CI)	‐64.53 [‐83.29, ‐45.78]
2.2 Systolic blood pressure Show forest plot	19	1261	Mean Difference (IV, Random, 95% CI)	‐6.91 [‐8.82, ‐4.99]

2.2.1 Short‐term (< 4 weeks)	3	144	Mean Difference (IV, Random, 95% CI)	‐7.91 [‐11.53, ‐4.28]
2.2.2 Longer‐term (≥ 4 weeks)	16	1117	Mean Difference (IV, Random, 95% CI)	‐6.82 [‐9.00, ‐4.64]
2.3 Diastolic blood pressure Show forest plot	19	1257	Mean Difference (IV, Random, 95% CI)	‐3.91 [‐4.80, ‐3.02]

2.3.1 Short‐term (< 4 weeks)	3	144	Mean Difference (IV, Random, 95% CI)	‐3.70 [‐5.74, ‐1.67]
2.3.2 Longer‐term (≥ 4 weeks)	16	1113	Mean Difference (IV, Random, 95% CI)	‐3.99 [‐5.03, ‐2.94]
2.4 Creatinine clearance Show forest plot	7	529	Mean Difference (IV, Random, 95% CI)	‐4.70 [‐6.94, ‐2.47]

2.4.1 Short‐term (< 4 weeks)	2	104	Mean Difference (IV, Random, 95% CI)	‐5.88 [‐9.14, ‐2.63]
2.4.2 Longer‐term (≥ 4 weeks)	5	425	Mean Difference (IV, Random, 95% CI)	‐3.65 [‐6.72, ‐0.58]
2.5 Serum creatinine Show forest plot	7	519	Mean Difference (IV, Random, 95% CI)	4.46 [0.08, 8.83]

2.5.1 Short‐term (< 4 weeks)	1	28	Mean Difference (IV, Random, 95% CI)	17.68 [‐47.51, 82.87]
2.5.2 Longer‐term (≥ 4 weeks)	6	491	Mean Difference (IV, Random, 95% CI)	4.40 [‐0.41, 9.20]
2.6 Weight Show forest plot	12	759	Mean Difference (IV, Random, 95% CI)	‐1.32 [‐1.94, ‐0.70]

2.6.1 Short‐term (< 4 weeks)	2	68	Mean Difference (IV, Random, 95% CI)	‐0.41 [‐0.81, ‐0.01]
2.6.2 Longer‐term (≥ 4 weeks)	10	691	Mean Difference (IV, Random, 95% CI)	‐1.53 [‐2.02, ‐1.04]

Comparison 2. Duration

Navigate to table in Review

Cochrane Review language

Website language

Abstract

Background

Objectives

Search methods

Selection criteria

Data collection and analysis

Main results

Authors' conclusions

PICOs

PICOs

Population

Intervention

Comparison

Outcome

Plain language summary

Altered dietary salt intake for adults with chronic kidney disease

Visual summary

Authors' conclusions

Implications for practice

Implications for research

Summary of findings

Background

Description of the condition

Description of the intervention

How the intervention might work

Why it is important to do this review

Objectives

Methods

Criteria for considering studies for this review

Types of studies

Types of participants

Inclusion criteria

Exclusion criteria

Types of interventions

Types of outcome measures

Primary outcomes

Secondary outcomes

Search methods for identification of studies

Electronic searches

Searching other resources

Data collection and analysis

Selection of studies

2021 update

2015 review

Data extraction and management

Assessment of risk of bias in included studies

Measures of treatment effect

Unit of analysis issues

Dealing with missing data

Assessment of heterogeneity

Assessment of reporting biases

Data synthesis

Subgroup analysis and investigation of heterogeneity

Sensitivity analysis

Summary of findings and assessment of the certainty of the evidence

Results

Description of studies

Results of the search

Included studies

Outcome reporting in included studies

Excluded studies

Ongoing studies

Risk of bias in included studies

Allocation

Random sequence generation

Allocation concealment

Blinding

Incomplete outcome data

Selective reporting

Other potential sources of bias

Carry‐over effect in cross‐over studies

Bias from cofounders

Effects of interventions

Sodium intake/excretion

CKD stage

Duration

Blood pressure

CKD stage