Higher Potassium Intake and Lower Sodium Intake May Help in Reducing CVD Risk by Lowering Salt Sensitivity of Blood Pressure in the Han Chinese Population

Round 1

Reviewer 1 Report

I have two main comments:

  Methods: clarify how the Na:K ratio is calculated- it says weight but this is not the usual approach, where the molar concentration in urine is used.    Discussion: I thought that this was somewhat superficial in the coverage of possible mechanistic explanation. This review article will help: PMID: 35058650

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The authors present a nice study investigating the impact of SSBP and intake of sodium and potassium on CV risk as predicted by the Framingham equation.

Comments

Use of a single 24 hour urine to estimate Na and K intake is not ideal as intake may vary from day to day.  This limitation should be acknowledged. Additionally volume is not a robust method for assessing adequacy. Use of the ratio mitigates these issues. 

The 24 hour urine Na seems typical but the urine K is quite low.  Is the potassium value incorrectly listed as mg instead of mmol or meq? Higher Potassium intake has been reported to have its greatest effect when coupled with a higher Na intake.  Would these findings translate to other populations with higher intake of potassium and/or sodium?

SS Hypertensive subjects respond more robustly to sodium loading and depletion and higher thresholds to define SS are recommended in those with hypertension (see Kurtz et al JAMA 2017).  Please comment and consider reclassifying patients if applicable and running subanalysis.

 

In methods please correct the criteria for stage I hypertension as the > sign is not correct – should be 140 to < 160 and 90 to < 100

In results page 4 line 154  please correct to say p< 0.05 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Xie et al. investigated the association between urine Na/K ratio and salt sensitivity of blood pressure and cardiovascular risk in 2057 Han Chinese. The authors report an association between urine Na/K ratio and estimated cardiovascular risk, which is, in part, mediated by salt sensitivity of blood pressure. Due to lack of important baseline data and additional correction for baseline blood pressure in the mediation analysis, the data is currently difficult to interpret.

Major points

-          Table 1 states that potassium intake was 41-42 mg/day. I guess that these numbers are not correct as this is basically means no potassium intake

-          Table 1: the authors should add baseline blood pressure to this table

-          Table 1: the authors should add antihypertensive medication use to this table (% of different antihypertensive drug classes)

-          Table 1: the authors should add eGFR to this table

-          Can the Framingham risk score be used in the Chinese population as this was not developed in the Chinese population? Does it predict CVD in Chinese people?

-          Should the authors add baseline blood pressure to the mediation analysis as this is an important covariate that affects SSBP and CVD risk, and is affected by urine Na/K?

-          When was the 24-hour urine sample collected. Was this collected during days prior to the actual salt loading and furosemide administration?

 

Minor points

-          The initial measurements were performed at least 6 years ago. Is it possible to relate SSBP and urine Na/K ratio to the cardiovascular events that actually occurred?

-          People with blood pressure >160/100 mmHg were excluded. The authors should limit their conclusions to patients with normotension or stage 1 hypertension

-          The authors excluded subjects with kidney disease. As kidney disease is an important determinant of salt sensitivity, the authors should provide the definition of kidney disease

-          Did subjects that were classified salt sensitive in response to salt loading have different associations compared to subjects that were classified salt sensitive in response to diuresis shrinkage?

-          Line 154: I don’t understand the p-value <0,05 when the distributions are equal

-          Why did the authors investigate SSBP in quartiles and not as a continuous variable. An analysis with a continuous variable has more power

-          Table 2: can the authors add the estimated baseline 10-year cardiovascular risk and additional risk in the other quartiles? This helps the readers to interpret the findings

-          Figure 2: why not make a linear regression analysis with delta BP vs Na – K – Na/K in 1 figure. This would enable an analysis of continuous outcomes on both the x-axis and y-axis

-          Why did the authors not investigate the relationship between urine Na/K ratio and salt sensitivity of BP in a separate analysis?

 

-          Line 305/317: the authors should provide additional information on how urine Na/K ratio was assessed because spot urine samples are not considered a for estimation of dietary Na/K ratio and the use of multiple 24-hour urine samples may result in best estimates

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Round 2

Reviewer 3 Report

1.      Table 2: can the authors add the estimated baseline 10-year cardiovascular risk and additional risk in the other quartiles? This helps the readers to interpret the findings

RE: Thanks for the comments. We have made corresponding changes to table 2 and table 3.

--> I would suggest to add the absolute estimated 10-year risk to be able to interpret the findings

 

 

 

1.      Figure 2: why not make a linear regression analysis with delta BP vs Na – K – Na/K in 1 figure. This would enable an analysis of continuous outcomes on both the x-axis and y-axis

RE: Thanks for the comments. We have added the results of the analysis using linear regression in table 4.

--> I would suggest to add a figure summarized these findings in a single figure

 

 

The authors should critically revise the number of decimals they are using through-out the manuscript, for example many risk estimates and p-values do not need 3 decimals.

 

 

The findings for the MAP change 1 and 2 (salt loading vs shrinkage) are much different. The authors should provide an explantation for this finding

Author Response

Please see the attachment.

Author Response File: Author Response.docx