Self‐Performed Dietary Sodium Reduction and Blood Pressure in Patients With Essential Hypertension: A Randomized Clinical Trial

Abstract

Background

Hypertension is the leading risk factor for cardiovascular disease worldwide. Patients with blood pressure (BP) response to dietary sodium reduction are referred to as “salt sensitive.” Salt sensitivity (SS) might be due to differences in sodium storage capacity and the erythrocyte SS examines this capacity of the red blood cells. This study aimed to test the effect of a self‐performed sodium reduced diet on BP in patients with essential hypertension and examine whether erythrocyte SS predicts SS.

Methods and Results

Seventy‐two patients with hypertension were included and randomized 2:1 to either sodium reduction or a control group for 4 weeks. Blood samples, 24‐hour BP measurement, and 24‐hour urine collection were performed before and after. The intervention group received advice on how to lower sodium intake. Urinary sodium excretion decreased 66 mmol (95% CI, −96 to −37 mmol) in the intervention group compared with the control group. Systolic 24‐hour BP decreased 9 mm Hg after low‐sodium diet compared with the control group (95% CI, −13 to −4 mm Hg). Similarly, the difference in reduction in diastolic BP between the groups was 5 mm Hg (95% CI, −8 to −1 mm Hg). We found no correlation between erythrocyte SS at baseline and decrease in 24‐hour BP, neither systolic nor diastolic (P=0.66 and P = 0.84).

Conclusions

Self‐performed sodium reduction was feasible and led to decrease in 24‐hour BP of 9/5 mm Hg compared with a control group. The erythrocyte SS did not correlate to the change in BP after lowering sodium intake.

Registration

URL: https://clinicaltrials.gov; Unique Identifier: NCT05165823.

Nonstandard Abbreviations and Acronyms

EGC

endothelial glycocalyx

ESS

erythrocyte sodium sensitivity

SBT

Salt Blood Test Mini

SS

salt sensitivity

TBW

total body water

Clinical Perspective.

What Is New?

• Self‐performed dietary sodium reduction is possible with the right guidance and leads to a clinically relevant decrease in blood pressure in patients with hypertension.

• Dietary sodium reduction might be a helpful tool in the treatment of hypertension that could lead to better blood pressure control.

What Are the Clinical Implications?

• More public information on the subject as well as better and more structured advice from health care professionals could be useful preventive measures in lowering the risks of cardiovascular morbidity.

Hypertension is considered the leading global risk factor for mortality. ¹ , ² , ³ , ⁴ An important cofactor in the pathogenesis of hypertension is high sodium intake and lowering dietary sodium leads to decrease in blood pressure (BP). ⁵ , ⁶ General guidelines suggest that salt (sodium chloride) intake should not exceed 6 g/day. However, in Denmark the average intake is ~11 g/day for men and ~8 g/day for women. ⁷ Reducing sodium intake is difficult as modern processed food contains excessive amounts of sodium. Perhaps therefore most studies on low sodium diet have used handout food and not self‐performed sodium reduction. A recent well‐designed crossover trial found a decrease in systolic BP of 8 mm Hg after 1 week of handout food containing only 500 mg sodium (equivalent to 1.25 g salt) daily. ⁸ However, handout food is not applicable in a clinical setting and thus additional studies on self‐performed sodium reduction are required. Recently, a small study of self‐performed sodium restriction in patients with treatment resistant hypertension was published and found that it was feasible and also lead to significant decreases in BP. ⁹

Studies have found that the individual BP response to changes in sodium intake varies substantially. Some experience no change in BP whereas others—referred to as “salt sensitive”—experience large changes in BP due to changes in sodium intake. However, no consistent definition of salt sensitivity (SS) exists. Depending on the study SS can be defined as either a specific absolute or relative change in BP or even just a significant change from baseline induced by reduced sodium intake. ⁸ , ¹⁰ , ¹¹ Patients with SS have an increased risk of developing target organ damage such as proteinuria, retinopathy, and ventricular hypertrophy. ¹² Discussions on the physiological mechanisms behind salt sensitivity are ongoing. It has been suggested that these patients have a higher volume load leading to suppressed renin levels, ¹³ whereas other theories focus on the ability to store sodium in the skin, for example. ¹⁴ , ¹⁵ Yet another actor in SS might be impaired endothelial function due to damaged endothelial glycocalyx (EGC). ¹¹ , ¹⁴ , ¹⁶ , ¹⁷ Because the EGC consists of negatively charged polymers that can bind sodium ions in a nonosmotic way, the EGC can contribute to storage of excessive sodium. ¹⁸ , ¹⁹ , ²⁰ Salt Blood Test Mini (SBT) has been developed as a measure of sodium buffer capacity of the EGC and it has been suggested that this relates to an individual's SS. ²¹ SBT is a specialized sedimentation reaction in which venous blood is mixed with a saline solution. The theory is that the healthier the EGC, the shorter the supernatant as the erythrocytes will not agglutinate as much. Standards are used to calculate the erythrocyte sodium sensitivity (ESS) based on the SBT and this value might be related to SS. Diagnosing which patients are salt sensitive and directing the dietary guidance toward them could contribute to better treatment of this high risk patient group.

In this study we aimed to test a clinically feasible self‐performed sodium reduction in patients with essential hypertension and examine the effects on systemic hemodynamics and body water. In addition, we tested the predictive value of ESS in identifying individuals with SS who would benefit from sodium restriction.

Methods

Anonymized data will be available from the corresponding author upon reasonable request from researchers for the purpose of reproducing results or replicating the trial design.

A randomized, controlled, nonblinded clinical trial was conducted in 72 participants at the University Clinic in Nephrology and Hypertension, Gødstrup Hospital, Denmark. Eligible participants were ≥18 years with a body mass index ≤35 m/kg², diagnosed with essential hypertension (defined as either being on BP medication or home BP measurements with average >135 mm Hg systolic or >85 mm Hg diastolic), receiving a maximum of three antihypertensive drugs, estimated glomerular filtration rate ≥15 mL/min per 1.73 m², and urine albumin/creatinine ratio ≤500 mg/g. Fertile women were to use contraceptives. Participants were included only if considered capable of adhering to a dietary regimen. Advertisements in local newspapers were the main method of recruitment, though a few participants were recruited from the outpatient clinic at the Department of Internal Medicine, Nephrology, Gødstrup Hospital, Denmark.

Participants with secondary hypertension, diabetes, clinically relevant liver or heart disease, stroke within the past 6 months, cancer other than nonmelanoma skin cancers, or participants on immunosuppressant therapy were excluded.

All participants were informed of the study purpose and methods and signed informed consent, where after medical history, physical examination by a medical doctor, office BP measurement on both arms, ECG, urine sample for albumin/creatinine ratio and urine dipstick (pH, glucose, ketone bodies, nitrite, leucocytes, protein, and blood), and blood samples (alanine transaminase, hemoglobin A1c, sodium, potassium, albumin, creatinine, leucocytes, hemoglobin, hematocrit, thrombocytes and international normalized ratio) was performed to uncover preexisting diseases. Urine‐human chorionic gonadotropin test was performed in all fertile women. During the entire study period all medications including BP medication remained unchanged and participants were all informed of the importance of adhering to their regular medications throughout the entire study period though adherence was not formally assessed.

Study data were managed using Research Electronic Data Capture hosted at Aarhus University and the Research Electronic Data Capture randomization module was used to randomize participants in a 2:1 ratio to either 4 weeks of sodium‐restrictive diet or usual diet. ²² , ²³ Participants were randomized in blocks of 12. Participants were informed of their randomization on the day of the baseline measurements. The primary outcome was change in systolic 24‐hour BP, which our power calculation was based upon. Thus, other presented outcomes must be considered secondary. All analyses performed are intention‐to‐treat‐analysis with the intervention group as the treated group.

Baseline and outcome measurements consisted of body weight, 24‐hour ambulatory blood pressure measurement (Mobil‐o‐graph PWA, I.E.M., Stolberg, Germany), 24‐hour urine collection (potassium, sodium, creatinine, urea, albumin), bioimpedance spectroscopy (Body Composition Monitor, Fresenius Medical Care, Bad Homburg, Germany) and blood sample collection (potassium, sodium, creatinine, hemoglobin and SBT) after 30 minutes' rest in supine position.

The intervention group received approximately 20 minutes of oral dietary advice from the same medical doctor and received written material on how to decrease sodium intake. The dietary advice was based on the participant's individual dietary patterns. Focus was on information on sodium content of groceries considered standard in a Danish context as well as concrete advice on which courses for each meal have low sodium content and how to lower the sodium content of the dishes they would usually eat. For more details, see Data S1. Because bread plays a large role in the Danish kitchen and store‐bought bread contains excessive amounts of salt, participants in the intervention group were offered sodium‐free bread and recommended to bake their own bread. To increase adherence, participants in the intervention group were contacted by phone once a week. Participants in the control group were instructed to continue eating as usual. Baseline measurements were performed no earlier than 1 week before the beginning of the intervention. After a 4‐week intervention (±3 days) follow‐up measurements were obtained.

Analysis of electrolytes in plasma and urine were performed by the Department of Clinical Biochemistry, Gødstrup Hospital as routine procedures. Estimated glomerular filtration rate was calculated using the Chronic Kidney Disease Epidemiology Collaboration formula.

SBT was performed using 50 μL venous K‐EDTA blood, mixing it with a 50 μL saline cocktail (CARE Diagnostica Lorreagenzeien, Voerde, Germany), shaking it by hand and filling it into a hematocrit tube. After 60 minutes the length of the supernatant was measured and ESS was calculated using the formula:

$$\mathrm{ESS}=\left(\frac{\text{length of supernatant}}{\text{standard}}\right)*100$$

The female standard was 26.1 mm and the male 21.4 mm. We did triple determination of SBT and used the mean to calculate ESS. If 3 measurements could not be obtained, we used the mean from the remaining. The normal value of ESS is defined to be around 100, values >120 are considered high SS, and values <80 low SS. ²¹ Variance of the SBT was evaluated by the proportion of participants with a difference in supernatant measurements of maximum 2 mm.

Body composition monitoring was performed twice on each examination day. Only measurements with a data quality >80% were used for further analysis. All participants underwent 24‐hour ambulatory BP measurements in relation to both examination days with 3 measurements per hour during both day and night time.

We chose the definition of salt sensitivity as a systolic BP change of 10 mm Hg in response to change in sodium intake.

Power Calculation

We considered a change in BP of 5 mm Hg to be significant and expected an SD of 8 mm Hg. With a 2‐sided α‐level of 0.05 and power of 80% 21 participants were needed in each group. The 2:1 ratio was chosen because earlier studies had shown it difficult for participants to adhere to a low sodium diet.

Statistical Analysis

All statistics were done in Stata version 18.0 (StataCorp LLC, College Station, TX, USA). P values <0.05 were considered statistically significant. Variables were tested for normality and equal variance before analysis. If logarithmic transformation led to normal distribution data are presented as median. In case of nonnormality, nonparametric tests were applied. Student's t test either with equal or nonequal variances was applied as appropriate on normally distributed data. For outcome comparisons between groups ANCOVA analysis was performed with adjustment for baseline values. Linear regression analyses were performed with change in systolic 24‐hour BP as dependent variable.

Approvals and Registration

The study was approved by the Central Denmark Region Committees on Health Research Ethics (Journal number 1‐10‐72‐241‐21) and was conducted according to the Declaration of Helsinki 2013. The study is registered at https://clinicaltrials.gov (NCT05165823).

Results

Inclusion began in February 2022 and examinations were completed in July 2023. As planned, 72 participants completed the study: 24 in the control group and 48 in the intervention group. For elaboration on screening, inclusion, and exclusion, see flow chart in Figure 1. Baseline characteristics are shown in Table 1.

Figure 1. Flow chart of screening, inclusion, and exclusion.

BMI indicates body mass index.

Table 1.

Baseline Characteristics

	Control group N=24	Intervention group N=48
Age, y*	64±9	66±8
Sex, female†	13 (54%)	20 (42%)
Body weight, kg*	82.9±13	80.4±12
Body mass index, kg/m2 *	28.5±4.2	26.9±3.2
Current smokers†	1 (4%)	2 (4%)
Former smokers†	11 (46%)	20 (42%)
Package years (former and current smokers)‡	15.0 (5–25)	13.5 (7–20)
P‐Na, mmol/L‡	141 (140–142.5)	141 (140–142)
P‐K, mmol/L‡	3.9 (3.8–4.0)	4.0 (3.7–4.2)
Estimated glomerular filtration rate, mL/min per 1.73 m2 *	85±14	83±16
U‐albumin/creatinine‐ratio‡	7.9 (5.7–10.3)	7.7 (5.6–13.1)
Antihypertensive drugs, no.‡	3 (2–3)	3 (2–3)
Angiotensin‐converting enzyme‐inhibitor or angiotensin II antagonist†	19 (79.2%)	35 (72.9%)
Ca‐antagonist†	9 (37.5%)	20 (41.7%)
Diuretics (Thiazid or Loop‐diuretics)†	7 (29.2%)	21 (43.8%)
Spironolactone†	2 (8.3%)	0 (0%)
Beta‐blockers†	4 (16.7%)	7 (14.6%)
Other antihypertensives†	1 (4.2%)	2 (4.2%)
No antihypertensives†	1 (4.2%)	1 (2.1%)

^*Mean±SD.

^† Number (percentage).

^‡ Median (interquartile range).

Urinary Sodium and Potassium Excretion

The 2 groups showed similar 24‐hour urinary sodium excretion at baseline with 138 mmol in the control group and 136 mmol in the intervention group. We found a significant change in sodium excretion in the intervention group of −62 mmol (95% CI, −82 to −42 mmol), which equals 3.6 g NaCl, that we did not see in the control group (5 mmol [95% CI, −12 to 21 mmol]) and the difference between groups was statistically significant (−66 mmol [95% CI, −96 to −37]).

Thirty‐five (73%) participants in the interventions group reduced sodium excretion to the level of the public recommendation of 6 g NaCl (103 mmol). This was seen in only 2 individuals (8%) in the control group.

The urinary excretion of potassium did not change in the control group (−5 mmol [95% CI, −13 to 4 mmol]); however, it did increase slightly but significantly in the intervention group (7 mmol [95% CI, 1–13 mmol]). The difference between groups was significant (P=0.023). The level of potassium in plasma did not change in any of the groups.

Changes in Hemodynamics

Hemodynamic measurements are presented in Table 2 and changes with 95% CIs are presented in Table 3. BPs in the 2 groups were similar at baseline (130/81 mm Hg in the intervention group and 130/80 mm Hg in the control group) and no significant change in BP was detected in the control group.

Table 2.

Hemodynamic Measurements in the 2 Groups at Baseline and Follow‐Up

	Baseline		Follow‐Up
	Control group N=24	Intervention group N=48	Control group N=24	Intervention group N=48	P value
Systolic 24‐h blood pressure	130±13	130±10	132±12	123±9	<0.001
Systolic daytime blood pressure	135±14	134±10	136±11	127±9	<0.001
Systolic night time blood pressure	121±15	122±12	122±15	115±11	0.018
Diastolic 24‐h blood pressure	80±10	81±9	81±10	76±8	0.051
Diastolic daytime blood pressure	84±11	85±10	84±9	81±8	0.082
Diastolic night time blood pressure	72±11	73±10	73±12	69±8	0.140

All blood pressure measurements are presented in mm Hg. All data shown are mean±SD.

Table 3.

Changes in Hemodynamic Parameters in the 2 Groups

Changes in hemodynamic parameters
	Control group N=24	Intervention group N=48	Comparison between groups
Systolic 24‐h blood pressure	2 (−2 to 6) P=0.436	−7 (−10 to −5) P<0.001	−9 (−13 to −5) P<0.001
Systolic daytime blood pressure	1 (−4 to 5) P=0.723	−7 (−10 to −4) P<0.001	−8 (−13 to −4) P<0.001
Systolic night time blood pressure	2 (−3 to 6) P=0.484	−7 (−10 to −5) P<0.001	−8 (−13 to −4) P<0.001
Diastolic 24‐h blood pressure	1 (−2 to 3) P=0.709	−4 (−6 to −2) P<0.001	−4 (−7 to −2) P=0.003
Diastolic daytime blood pressure	1 (−3 to 3) P=0.953	−4 (−6 to −2) P<0.001	−4 (−7 to −1) P=0.011
Diastolic night time blood pressure	1 (−3 to 4) P=0.777	−4 (−6 to −2) P<0.001	−4 (−7 to −1) P=0.020

Data are presented in mm Hg. All data shown are mean (95% CIs), P value.

The intervention group experienced a mean decrease in systolic 24‐hour BP of 7 mm Hg (95% CI, −10 to −5 mm Hg), which was significantly different from both baseline and from the change in the control group (P<0.001). Compared to the control group mean decrease in systolic 24‐hour BP was 9 mm Hg (95% CI, −13 to −5 mm Hg). Twenty‐four‐hour diastolic BP decreased 4 mm Hg (95% CI, −7 to −2 mm Hg) in the intervention group compared with the control group. Changes were seen in both daytime and nocturnal BP. Systolic BP in daytime decreased 8 mm Hg in the intervention group compared with the control group (95% CI, −13 to −4) and nocturnal systolic BP decreased 8 mm Hg (95% CI, −13 to −4), also compared with the control group.

For the 35 participants who lowered sodium intake to the recommendations of 103 mmol the change in BP in was −9 mm Hg (95% CI, −11 to −6 mm Hg) and in the remaining 13 (27%) participants the change was −3 mm Hg (95% CI, −8 to 1 mm Hg).

Changes in Body Composition

No significant change in body weight was found in the control group (−0.3 kg [95% CI, −0.8 to 0.2 kg]). There was a difference in the change in body weight between the 2 groups of −1.0 kg (95% CI, −1.6 to −0.4 kg) indicating a weight loss in the intervention group.

Data on body composition monitor‐measurements were available from 23 in the control group and 39 in the intervention group. These measurements suggest that weight loss was at least in part due to a significant change in total body water of −0.60 L (95% CI, −0.98 to −0.23 L) in the intervention group. This change was not found in the control group (−0.05 L [95% CI, −0.49 to 0.38 L]). The difference in loss of body water between groups was not statistically significant, however the tendency was present (P=0.0635).

As shown in Figure 2, we found a significant correlation between loss of body weight in the intervention group and decrease in systolic 24‐hour BP (P=0.02, R=0.335); however, we found no correlation between the loss of total body water and decrease in systolic 24‐hour BP (P=0.45, R=0.126).

Figure 2. Systolic 24‐hour blood pressure changes in the intervention group.

A, Correlation between change in systolic 24‐h blood pressure and change in body weight. B, Correlation between change in systolic 24‐h blood pressure and change in total body water. Black lines: Fitted values. Dotted lines: 95% CIs.

Salt Blood Test and Salt Sensitivity

The SBT was obtained from 70 participants. In 4 participants only 2 SBT measurements were obtained. In 64/70 (91%) participants the supernatant measurements varied maximum 2 mm. Mean values of the calculated ESS at baseline and follow‐up as well as comparison of the values between groups are given in Table 4.

Table 4.

Calculated Erythrocyte Sodium Sensitivity from Salt Blood Test Mini Measurements Presented as Mean±SD

	Control group N=24	Intervention group N=46	Comparison between groups
Baseline	74.1±25.6	73.8±30.7	0.3 (−14.3 to 15.0)
Follow‐up	77.8±28.8	76.4±28.7	0.9 (−13.3 to 15.0)
Difference	3.7 (−10.2 to 2.8)	2.6 (−6.4 to 1.1)	−1.1 (−7.7 to 5.6)

There were no significant differences between the groups at any of the time points nor were there any significant changes in ESS over time. In the intervention group we found no correlation between ESS at baseline and decrease in 24‐hour BP, neither systolic nor diastolic (P=0.31 and P=0.63).

We found 18 individuals in the intervention group defined to be salt sensitive.

Discussion

The long‐term acknowledgement of sodium's role in BP is supported by the results of our trial. The achieved mean decreases in BP after low‐sodium diet compared with a control group was found to be 9/4 mm Hg, which is a clinically relevant change and comparable to the results in the before‐mentioned American trial. ⁸ The participants in our trial reduced sodium intake 62 mmol, which corresponds to a 3.6 g/day lower NaCl‐intake, and 71% of participants were able to reach the recommended daily intake of NaCl. These participants were also the ones with the largest change in BP. The ESS was not correlated to the change in BP.

Worth noting in our study is that this clinically relevant decrease in BP was seen after a self‐performed dietary sodium reduction. This has not previously been established as most studies have been performed using preprepared handout food. Our study shows that the right information and motivation make it possible to guide patients to decrease their sodium intake—at least for the 4 weeks our trial lasted. We did not examine whether patients adhered to the dietary regimen after the trial. We are aware of selection bias as most of our participants contacted the trial personnel on their own initiative and thus might be more motivated to implement dietary changes than the general population. Trial results may thus not be directly generalizable to the entire patient population but confirms that introducing dietary changes is feasible and effective in the treatment of hypertension.

Because of the limited sample size, we did not investigate differences in BP changes in relation to BP medication and thus do not know whether certain drugs lead to larger or smaller changes in BP after sodium reduction. Nor do we know whether BP medication influences ESS or how the correlation would be in patients with untreated hypertension.

A strength of this study is the randomized controlled setup and that we were able to confirm a change in sodium intake in the intervention group using 24‐hour urine collections. Whether 1 24‐hour urine collection is enough to establish a true value of sodium intake has been thoroughly discussed. ²⁴ A recent study of the variability of sodium excretion on standardized diet concluded that steady state was reached within a few days and that 1 24‐hour urine collection was sufficient. ²⁵ We collected urine only before and after the intervention and not during, thus we cannot be sure the participants adhered to the restriction throughout the entire study period and not only in the past few days, though we attempted to increase adherence during all 4 weeks by telephoning participants once a week.

The urinary excretion of potassium increased slightly in the intervention group, and this could be due to either higher dietary intake of potassium or because the lower sodium intake leads to an increase in the renin‐angiotensin‐aldosterone‐system leading to higher excretion rates of potassium. Because diet was not held constant in all aspects other than sodium, the participants could have consumed more potassium in the study period, which in itself could contribute to a lower BP. ²⁶ , ²⁷ This uncertainty is avoided in more controlled setups. However, studies on the effects of real‐life interventions that are feasible as treatment of hypertension are necessary before implementation.

With SS defined as a decrease in systolic BP of 10 mm Hg, we found 18 (38%) participants with SS in the intervention group, which corresponds well to earlier studies. ²⁸ , ²⁹ SS is no doubt a clinically relevant condition to diagnose in patients with hypertension due to the prognostic value and to potentially improve treatment. The novel technique for determining SS tested in this study would markedly increase the ease of diagnosing and help the health care workers to direct the dietary advice toward the patients who would benefit the most. This is the first trial to evaluate the SBT and the calculated ESS in patients with hypertension and under randomized conditions. ESS did not turn out to be useful in predicting a BP response to change in sodium consumption. Similarly, no relation between ESS and change in BP in patients with treatment‐resistant hypertension after 2 weeks of sodium restriction was found in the aforementioned study. ⁹ ESS is a marker of erythrocyte sodium sensitivity and other studies have shown a relation between ESS and self‐reported sodium consumption, cerebrovascular disease, and end‐stage kidney disease. ³⁰ , ³¹ , ³² Thus, ESS might be related to vascular damage, though our study does not support it as a good measure of SS of BP.

The recommended daily NaCl intake of 6 g/day separated our participants well regarding change in BP. This should perhaps lead to an increased focus on and conversation about sodium balance in patients with hypertension. This could be introduced in hypertension clinics initially on a study basis to evaluate the effect in a more clinical setting. This would test the effect of sodium reduction on a broader patient population as well as for at longer duration. Mean decreases in sodium intake and thus BP in this group might not be as pronounced as in our study, but even a small change in BP in this population would lead to important epidemiological improvements in cardiovascular outcomes and thereby have a strong impact on public health.

Important learnings from our study are that in a developed country like Denmark it is possible to eat <6 g NaCl/day but also that it takes guidance and tangible advice. It is important to emphasize that 4 weeks with a low‐sodium diet is not enough to change the risk of cardiovascular disease—permanent adherence to this regimen would be necessary.

Conclusions

We found it possible to guide patients to decrease their sodium intake using our verbal and written information. The reduction in sodium intake led to a relevant change in BP of 9/5 mm Hg along with a loss of body weight and total body water. This study does not, however, support ESS as a valuable tool in identifying patients with SS. The mechanisms behind SS remain uncovered.

Perspectives

The results from this study add to previous knowledge on the BP‐lowering effect of sodium restrictive diet. We have shown that patients with hypertension can be guided to lower sodium intake and eat according to recommendations and that this has a beneficial and clinically relevant effect on their BP. More public information on the subject as well as better and more structured advice from health care professionals could be important in changing the rising curve of cardiovascular morbidity and mortality.

Sources of Funding

The study has been funded by the Augustinus Foundation, Health Research Foundation of Central Denmark Region, Gødstrup Hospital Research Foundation, Danish Society of Nephrology, the Danish Kidney Association, the Toyota Foundation and Arvid Nilsson's Foundation.

Disclosures

Frank Holden Mose has participated in advisory board and received speaker honoraria from Boehringer Ingelheim and AstraZeneca. Jesper Nørgaard Bech disclosed advisory board participation for Bayer, Boehringer Ingelheim, and AstraZeneca. Steffen Flindt Nielsen has had travel expenses covered by AstraZeneca. The remaining authors have no disclosures to report.

Supporting information

Data S1