Abstract
Cross-sectional studies have reported that high sodium sensitivity is more common among individuals with hypertension. Experimental studies have also reported various animal models with sodium-resistant hypertension. It is unknown, however, whether sodium sensitivity and resistance precede the development of hypertension. We conducted a feeding study, including a 7-day low-sodium diet (1,180 mg/day) followed by a 7-day high-sodium diet (7,081 mg/day), among 1,718 Chinese adults with blood pressure (BP) <140/90 mmHg. We longitudinally followed them over an average of 7.4 years. Three BP measurements and 24-hour urinary sodium excretion were obtained on each of three days during baseline observation, low-sodium and high-sodium interventions, and two follow-up studies. Three trajectories of BP responses to dietary sodium intake were identified using latent trajectory analysis. Mean (standard deviation) changes in systolic BP were −13.7 (5.5), −4.9 (3.0), and 2.4 (3.0) mmHg during the low-sodium intervention, and 11.2 (5.3), 4.4 (4.1), and −0.2 (4.1) mmHg during the high-sodium intervention (P<0.001 for group differences) in high sodium-sensitive, moderate sodium-sensitive, and sodium-resistant groups, respectively. Compared to individuals with moderate sodium sensitivity, multiple-adjusted odds ratios (95% confidence intervals) for incident hypertension were 1.43 (1.03–1.98) for those with high sodium sensitivity and 1.43 (1.03–1.99) for those with sodium resistance (P=0.006 for non-linear trend). Furthermore, a J-shaped association between systolic BP responses to sodium intake and incident hypertension was identified (P<0.001). Similar results were observed for diastolic BP. Our study indicates that individuals with either high sodium sensitivity or sodium resistance are at an increased risk for developing hypertension.
Keywords: Sodium sensitivity, sodium resistance, blood pressure, hypertension, cohort studies, feeding study, polygenic risk scores
Graphical Abstract
Hypertension is a leading preventable risk factor for cardiovascular disease and premature death.1.2 Animal experiments, epidemiological studies, and clinical trials have documented a causal relationship between high dietary sodium intake and elevated blood pressure (BP).3–5 A recent meta-analysis of 48 randomized clinical trials reported that an average 42 mmol/day sodium reduction lowered systolic BP (SBP) by 3.23 mmHg (95% confidence interval [CI]: 2.41–4.06) and diastolic BP (DBP) by 2.24 mmHg (1.61–2.96).5 However, BP responses to dietary sodium intake vary considerably among individuals, a phenomenon known as sodium sensitivity.6,7 It has been found that blacks, women, older people, and those with hypertension, obesity, or chronic kidney disease are more sensitive to dietary sodium intake.6–9 Several cross-sectional studies have reported that high sodium sensitivity is more common among individuals with hypertension compared to those with normal BP.6–9 However, there are sparse data on the prospective association between sodium sensitivity and subsequent incidence of hypertension.10,11 It is unclear whether sodium sensitivity is a cause or consequence of hypertension.6,7
Several genetic sodium-resistant forms of hypertension have been identified in animal models.12,13 Lopez and colleagues reported a salt-resistant hypertension mouse model in which the GC-A gene mutations resulted in the loss of the guanylyl cyclase-A receptor for atrial natriuretic peptide.12 Rieg and colleagues developed a salt-resistant hypertension mouse model by knockout of the P2Y2 receptors gene.13 In these mouse models, elevated BP did not change during high and low salt diet.12,13 Sodium resistance has been well documented in human studies and the proportion of sodium-resistant individuals varied considerably according to the definition of sodium resistance.8,9,14 However, the relationship between sodium resistance and risk of hypertension has not been studied in human populations.
We examined the prospective association of sodium sensitivity and sodium resistance with the incidence of hypertension among normotensive individuals from the Genetic Epidemiology Network of Salt Sensitivity (GenSalt) study.
Methods
Study Participants
The present article adheres to the American Heart Association Journals’ implementation of the Transparency and Openness Promotion Guidelines. The data that support the findings of this study are available from the corresponding author upon reasonable request. Preregistration of the study can be found at https://www.clinicaltrials.gov (ClinicalTrials.gov Identifier: NCT00721721). Written informed consents were obtained from all participants, and the GenSalt study was approved by the Institutional Review Boards at all participating institutions.
The GenSalt study was a diet-feeding study with low- and high-sodium interventions conducted in six rural communities in northern China. The GenSalt study aimed to test BP responses to dietary sodium intake and investigate the genetic and environmental determinants of BP sodium sensitivity in the general Chinese population.15 A population-based BP screening was conducted among residents aged 18–60 years in the study communities to identify potential participants. Individuals with a mean SBP of 130–160 mmHg and/or DBP of 85–100 mmHg based on nine BP measures from three baseline visits were recruited, along with their family members. Persons who were >60 years old; had BP >160/100 mmHg, secondary hypertension, or a history of clinical cardiovascular disease, diabetes, or chronic kidney failure; or were using antihypertensive medication, pregnant, heavy alcohol drinkers, or currently on a low-sodium diet were excluded from the study. Of 1,906 dietary intervention participants, 1,718 (90.1%) with BP <140/90 mmHg at baseline were eligible for this analysis. Additionally, 20 participants who did not complete the 2-week dietary sodium intervention were further excluded. Therefore, 1,698 participants were included in this analysis.
Dietary Intervention
After a 3-day baseline observation period on their usual diet, participants received a low-sodium diet (3 g salt or 51.3 mmol sodium/day [1,180 mg]) for 7 days, and then received a high-sodium diet (18 g salt or 307.8 mmol sodium/day [7,081 mg]) for another 7 days. During the dietary sodium intervention, potassium and other nutrient intake remained unchanged. Although dietary sodium intake was the same for all participants, dietary total energy intake varied according to their baseline energy intake. All foods were cooked without salt, and prepackaged salt was added to individual participants’ meals when they were served by study staff. To ensure participants’ compliance with the intervention, they were required to have breakfast, lunch, and dinner at the study kitchen under the supervision of study staff during the entire study period. Participants were instructed to avoid consuming any food and beverages (except water) that were not provided by the study. In addition, three timed urinary specimens (one 24-hour and two overnight) were collected during the 3-day baseline observation and in the last 3 days of each intervention phase to monitor participants’ compliance with the dietary intervention. The overnight urinary sodium and potassium excretions were converted to 24-hour values based on formulas developed in the GenSalt study.9
Data Collection during Baseline and Intervention
The GenSalt study baseline examinations and dietary interventions were conducted in 2003–2005. Standard questionnaires were administered by trained staff at baseline to collect information on demographic characteristics, medical history, and lifestyle risk factors.15 Three BP measurements were obtained each morning of the 3-day baseline observation, and on days 2, 5, 6, and 7 of each intervention period, by trained and certified observers and were averaged for analyses. Baseline BP was calculated as the mean of nine measurements from the 3-day observation. BP was measured with participants seated after 5 minutes of rest using a random-zero sphygmomanometer and one of four cuff sizes (pediatric, regular, large, or thigh) based on the participant’s arm circumference.16 In addition, participants were required to avoid alcohol, cigarette smoking, coffee/tea, and exercise for at least 30 minutes prior to their BP measurements. All BP observers were blinded to the dietary intervention. Body weight and height were measured twice in light indoor clothing without shoes. Overnight fasting blood samples were drawn by venipuncture at baseline, and serum creatinine was measured at a central laboratory.15 The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation was used to calculate estimated-glomerular filtration rate (eGFR).17
Data Collection during Follow-up Study
The GenSalt Follow-up Study was conducted in 2008–2009 and 2011–2012 to examine longitudinal changes in BP and the incidence of hypertension among the GenSalt study participants. The median duration from baseline to the final follow-up visit was 7.4 years (inter-quartile range 7.0–7.8 years). At each follow-up study, medical history questionnaires were administered by trained staff to collect information on newly diagnosed hypertension and use of antihypertensive medication. Three BP measurements were obtained every day at each 3-day visit by trained and certified observers with the same protocol used in the baseline examination.16 BP levels were calculated as the mean of nine measurements from the 3-day visits at each follow-up study. Hypertension was defined as mean SBP ≥140 mmHg and/or DBP ≥90 mmHg or use of antihypertensive medication.
Latent Trajectory Analysis
Latent trajectory analysis was conducted to identify subgroups sharing a similar underlying trajectory of BP responses to low-sodium and high-sodium interventions and implemented by the SAS Proc Traj procedure.18 BP changes from baseline to low-sodium intervention and changes from low-sodium to high-sodium intervention on days 2, 5, 6, and 7 of each intervention period were used for modeling. We fit models up to the optimal number of trajectories by comparing the Bayesian Information Criterion for each set of trajectories. The model with 3 trajectories fit best, and participants were classified into trajectory groups with excellent discrimination. See Methods Supplement for details. The mean probabilities of group membership of high sodium sensitivity, moderate sodium sensitivity, and sodium resistance were 91.7%, 91.8%, and 89.1% for SBP and 90.2%, 93.2%, and 90.5% for DBP, respectively.
Polygenic Risk Score (PRS)
As part of the NHLBI TOPMed program, GenSalt whole-genome sequencing was performed by the Baylor Human Genome Sequencing Center to a median coverage of ≥30× using Illumina HiSeq X technology.19 Details regarding the laboratory methods, data processing, and quality control are described on the TOPMed website (https://www.nhlbiwgs.org/). We performed genome-wide analyses to identify genetic variants associated with trajectories of BP responses to dietary sodium intake using the moderate sodium-sensitive trajectory as a reference group in the GenSalt study participants, while controlling for age, sex, and the first 3 genetic ancestry principal components. A total of 3,214, 4,121, 1,918, and 615 independent variants (r2<0.3) had a P-value <0.0001 for SBP high sodium-sensitive trajectory, SBP sodium-resistant trajectory, DBP high sodium-sensitive trajectory, and DBP sodium-resistant trajectory, respectively. These variants were used to generate PRSs by summing risk alleles for their corresponding sodium-sensitive or resistant BP trajectories. In addition, genome-wide significant SNPs from 1,436 loci for SBP, DBP, and/or pulse pressure were retrieved from four large-scale genome-wide studies among populations of multiple ancestries.20–23 PRSs were calculated by summing SBP- and DBP-increasing alleles, respectively.
Statistical Analysis
Logistic regression models were used to examine the associations between trajectories and incidence of hypertension during follow-up. Mixed-effects linear models were used to compare mean BP change over follow-up by trajectories. Several sets of covariables were adjusted in these models. First, age, sex, and baseline BP were adjusted. Next, physical activity, alcohol consumption, smoking, 24-hour sodium excretion, and 24-hour potassium excretion at baseline were added into multivariable models. Furthermore, eGFR was adjusted because reduced kidney function has been suggested as a mechanism for sodium sensitivity.24 Finally, PRSs for BP sodium sensitivity and resistance and PRSs for usual BP were adjusted in multivariable models, separately, because genetic factors might play an important role in sodium sensitivity and resistance.
Possible non-linear relationships between BP responses to sodium intervention and incidence of hypertension were examined with restricted cubic splines.25 BP responses were defined as mean BP differences between high-sodium intervention and low-sodium intervention. Three knots were located at 5%, 50%, and 95% percentiles corresponding to SBP responses of −4.0, 4.7, and 15.1 mmHg and DBP responses of −6.8, 1.8, and 10.8 mmHg, respectively. Tests for non-linearity comparing a model with only the linear term to a model with the linear and restricted cubic spline terms were conducted using likelihood ratio tests. All analyses were conducted using SAS version 9.4, with the significance level set at a 2-sided P<0.05.
Results
Based on BP changes during low-sodium and high-sodium interventions, three trajectories of SBP responses (Figure 1A) and three trajectories of DBP responses (Figure 1B) were identified. BP sharply decreased during the low-sodium intervention and rapidly increased during the high-sodium intervention among high sodium-sensitive individuals, while BP did not decrease during the low-sodium intervention and did not increase during the high-sodium intervention among sodium-resistant individuals. The majority of individuals were high or moderate sodium-sensitive based on SBP (18.0% and 58.3%) and DBP (12.0% and 63.7%) responses, respectively (Figures 1C–F). Compared to sodium-resistant participants, those with high SBP sodium sensitivity were older and had greater baseline waist circumference, BP, fasting glucose, PRS for sodium sensitivity, and PRS for BP, but lower eGFR at baseline and PRS for sodium resistance (Table 1). In addition, urinary excretion of sodium and potassium was higher in participants with moderate sodium sensitivity. Those with high DBP sodium sensitivity were less likely to be men, smoke cigarettes, drink alcohol, or engage in physical activity, and had higher baseline BP, potassium excretion, PRS for sodium sensitivity, and PRS for BP, but lower PRS for sodium resistance (Table S1).
Figure 1. Mean and distribution of blood pressure responses according to three trajectories of high sodium sensitivity, moderate sodium sensitivity, and sodium resistance.
A. Mean (95% CI) systolic blood pressure responses by sodium-sensitive trajectories; B. Mean (95% CI) diastolic blood pressure responses by sodium-sensitive trajectories; C. Distribution of systolic blood pressure responses during low-sodium intervention by sodium-sensitive trajectories; D. Distribution of diastolic blood pressure responses during low-sodium intervention by sodium-sensitive trajectories; E. Distribution of systolic blood pressure responses during high-sodium intervention by sodium-sensitive trajectories; F. Distribution of diastolic blood pressure responses during high-sodium intervention by sodium-sensitive trajectories. L2, L5, L6, and L7 refer to low-sodium intervention days 2, 5, 6, and 7. H2, H5, H6, and H7 refer to high-sodium intervention days 2, 5, 6, and 7.
Table 1.
Baseline Characteristics of Study Participants According to Trajectories of Systolic Blood Pressure Responses to Dietary Sodium Intake
| Characteristics | High Sodium-Sensitive (n=306) |
Moderate Sodium-Sensitive (n=990) |
Sodium-Resistant (n=402) |
P-values for Group Differences |
|---|---|---|---|---|
| Age, years | 42.7 (8.6) | 37.7 (9.4) | 36.0 (9.5) | <0.001 |
| Male, n (%) | 152 (49.7) | 515 (52.0) | 223 (55.5) | 0.29 |
| High school education, n (%) | 39 (12.7) | 144 (14.5) | 61 (15.2) | 0.63 |
| Current cigarette smoking, n (%) | 98 (32.0) | 294 (29.7) | 129 (32.1) | 0.58 |
| Current alcohol drinking, n (%) | 81 (26.5) | 284 (28.7) | 121 (30.1) | 0.57 |
| Physical activity, METs/wk | 24.6 (10.8) | 23.9 (11.5) | 24.6 (12.4) | 0.46 |
| Body-mass index, kg/m2 | 23.4 (3.0) | 23.2 (3.0) | 22.9 (3.2) | 0.05 |
| Waist circumference, cm | 81.1 (9.4) | 79.6 (9.6) | 79.0 (9.7) | 0.01 |
| SBP, mmHg | 121.9 (9.9) | 114.3 (11.5) | 108.6 (11.3) | <0.001 |
| DBP, mmHg | 76.5 (7.3) | 71.7 (8.4) | 68.9 (8.9) | <0.001 |
| Plasma fasting glucose, mg/dL | 88.7 (12.4) | 86.4 (12.2) | 86.4 (13.6) | 0.02 |
| Serum LDL-cholesterol, mg/dL | 96.3 (26.0) | 94.7 (26.5) | 92.2 (27.8) | 0.12 |
| Serum HDL-cholesterol, mg/dL | 51.5 (12.3) | 51.1 (11.0) | 50.6 (10.8) | 0.54 |
| Serum triglycerides, mg/dL | 126.6 (76.9) | 118.3 (71.4) | 119.9 (65.5) | 0.21 |
| Estimated-glomerular filtration rate, mL/min | 88.4 (16.5) | 95.3 (19.8) | 94.8 (18.2) | <0.001 |
| Urinary sodium excretion, mmol/24-hr* | 236.7 (63.4) | 245.0 (69.0) | 230.7 (57.4) | 0.001 |
| Urinary potassium excretion, mmol/24-hr* | 35.7 (9.8) | 37.8 (10.1) | 35.7 (8.1) | <0.001 |
| SBP sodium-sensitivity PRS | 5260.2 (98.5) | 5085.4 (134.5) | 5140.3 (124.7) | <0.001 |
| DBP sodium-sensitivity PRS | 465.1 (136.8) | 426.6 (126.9) | 420.2 (126.3) | <0.001 |
| SBP sodium-resistance PRS | 6951.4 (121.4) | 6898.6 (130.8) | 7053.9 (72.3) | <0.001 |
| DBP sodium-resistance PRS | 175.6 (17.6) | 177.8 (20.1) | 185.1 (22) | <0.001 |
| SBP PRS | 728.6 (17.5) | 723.5 (18.3) | 719.1 (18.4) | <0.001 |
| DBP PRS | 462.2 (10.5) | 459.8 (11.3) | 458.2 (11.5) | <0.001 |
Values are means (standard deviations) for continuous characteristics and numbers (percentages) for categorical variables. SBP=systolic blood pressure; DBP=diastolic blood pressure; PRS=polygenic risk score; HDL=high-density lipoprotein; LDL=low-density lipoprotein; METs=metabolic equivalents.
To convert sodium from mmol to mg, multiply by22.9898, and to convert potassium from mmol to mg, multiply by 39.0983.
Average 24-hour urinary sodium excretion significantly increased from 47.4 mmol during the low-sodium intervention to 244.1 mmol during the high-sodium intervention (P<0.001) consistently across all SBP response trajectories (Table 2). Mean SBP and DBP significantly decreased by 13.7 and 6.2 mmHg among individuals with high SBP sodium sensitivity and by 4.9 and 2.5 mmHg among individuals with moderate SBP sodium sensitivity, but increased by 2.4 and 0.7 mmHg among individuals with SBP sodium resistance, respectively, during the low-sodium intervention (P<0.001 for group difference). Likewise, SBP and DBP significantly increased by 11.2 and 4.7 mmHg among individuals with high SBP sodium sensitivity and by 4.4 and 1.8 mmHg among individuals with moderate SBP sodium sensitivity, but slightly decreased by 0.2 and 0.5 mmHg among individuals with SBP sodium resistance, respectively, during the high-sodium intervention (P<0.001 for group difference). Similar patterns were observed for BP responses to the low-sodium and high-sodium interventions according to DBP response trajectories (Table S2). However, urinary excretion of sodium was significantly higher among individuals with high DBP sodium sensitivity during the low-sodium and high-sodium interventions.
Table 2.
Urinary Sodium and Potassium Excretion and Blood Pressure Responses According to Trajectories of Systolic Blood Pressure Responses to Dietary Sodium Intake
| Characteristics | High Sodium-Sensitive (n=306) |
Moderate Sodium-Sensitive (n=990) |
Sodium-Resistant (n=402) |
P-Values for Group Differences |
|---|---|---|---|---|
| 24-hour electrolyte excretion during low-sodium intervention* | ||||
| Urinary sodium, mmol/24-hr‡ | 47.0 (14.6) | 47.7 (15.8) | 47.0 (17.9) | 0.67 |
| Urinary potassium, mmol/24-hr‡ | 31.1 (7.3) | 31.8 (8.1) | 31.1 (7.3) | 0.21 |
| 24-hour electrolyte excretion during high-sodium intervention* | ||||
| Urinary sodium, mmol/24-hr‡ | 246.5 (40.1) | 243.2 (36.9) | 244.5 (39.0) | 0.41 |
| Urinary potassium, mmol/24-hr‡ | 35.3 (6.9) | 35.9 (7.8) | 35.6 (7.2) | 0.52 |
| Blood pressure responses to low-sodium intervention† | ||||
| SBP, mmHg | −13.7 (5.5) | −4.9 (3.0) | 2.4 (3.0) | <0.001 |
| DBP, mmHg | −6.2 (5.0) | −2.5 (4.5) | 0.7 (5.3) | <0.001 |
| Mean arterial BP§, mmHg | −8.7 (4.5) | −3.3 (3.4) | 1.3 (4.0) | <0.001 |
| Blood pressure responses to high-sodium intervention† | ||||
| SBP, mmHg | 11.2 (5.3) | 4.4 (4.1) | −0.2 (4.1) | <0.001 |
| DBP, mmHg | 4.7 (5.6) | 1.8 (4.9) | −0.5 (4.9) | <0.001 |
| Mean arterial BP§, mmHg | 6.8 (4.9) | 2.6 (4.1) | −0.4 (4.0) | <0.001 |
Values are means (standard deviations).
Mean of three 24-hour excretions on days 5, 6, and 7.
BP responses to low-sodium intervention=mean BP during low-sodium intervention ‒ mean BP at baseline; and BP responses to high-sodium intervention=mean BP during high-sodium intervention ‒ mean BP during low-sodium intervention. BP levels at baseline and during the intervention were calculated as the mean of nine measurements from three days at baseline or on days 5, 6, and 7 of each intervention phase.
To convert sodium from mmol to mg, multiply by 22.9898, and to convert potassium from mmol to mg, multiply by 39.0983.
Mean arterial pressure=(SBP+2×DBP) / 3.
Among 1,698 GenSalt participants included in this analysis, 1,604 (94.5%) took part in the two follow-up studies. Over an average follow-up of 7.4 years, 514 (32.0%) participants developed incident hypertension. The age-sex-baseline BP-adjusted incidence of hypertension was 36.1%, 29.2%, and 36.0% for SBP high sodium sensitivity, moderate sodium sensitivity, and sodium resistance, and 35.6%, 29.0%, and 38.2% for DBP high sodium sensitivity, moderate sodium sensitivity, and sodium resistance, respectively (Table 3). Compared to those with moderate SBP sodium sensitivity, individuals with high sodium sensitivity and sodium resistance both had significantly increased risk of incident hypertension. After adjustment for multiple covariables in model 1, the odds ratios (95% CI) were 1.49 (1.08–2.07) and 1.45 (1.05–2.02) for individuals with high SBP sodium sensitivity and resistance, respectively. After additional adjustment for eGFR, the odds ratios were slightly diminished but still significant: 1.43 (1.03–1.98) for high SBP sodium sensitivity and 1.43 (1.03–1.99) for SBP sodium resistance. After further adjustment for sodium-sensitivity and resistance PRS in model 3 or usual BP PRS in model 4, the odds ratio of incident hypertension associated with SBP sodium resistance became insignificant (Table 3). Likewise, the multiple-adjusted odds ratios (95% CI) were 1.45 (1.01–2.07) and 1.69 (1.25–2.28) for individuals with high DBP sodium sensitivity and resistance, respectively. After further adjustment for PRS for sodium sensitivity and sodium resistance, or for usual BP, the odds ratio of incident hypertension associated with high DBP sodium sensitivity became insignificant. Although PRSs were significantly associated with the incidence of hypertension, only SBP sodium-resistant PRS remained significantly associated with incidence of hypertension in multiple-adjusted models. For example, a change in SBP sodium-resistant PRS by one standard deviation (SD=134.09) was associated with an odds ratio of 1.29 (95% CI 1.10–1.52) in model 3 and 1.30 (1.11–1.54) in model 4 for incident hypertension.
Table 3.
Risk of Incident Hypertension According to Systolic and Diastolic Blood Pressure Responses to Dietary Sodium Intake
| BP response trajectories | No. of cases/ No. of participants | Age, gender, and baseline blood pressure adjusted incidence, % | Odds Ratio (95% CIs) | ||||
|---|---|---|---|---|---|---|---|
| Age, gender, and baseline blood pressure adjusted | Multiple-adjusted model 1* | Multiple-adjusted model 2† | Multiple-adjusted-model 3‡ | Multiple-adjusted-model 4§ | |||
| Trajectories of SBP responses to dietary sodium intake | |||||||
| High sodium-sensitive | 144/283 | 36.1 | 1.42 (1.03–1.94) | 1.49 (1.08–2.07) | 1.43 (1.03–1.98) | 1.63 (1.11–2.41) | 1.63 (1.10–2.40) |
| Moderate sodium-sensitive | 274/940 | 29.2 | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| Sodium-resistant | 96/381 | 36.0 | 1.49 (1.08–2.06) | 1.45 (1.05–2.02) | 1.43 (1.03–1.99) | 1.14 (0.78–1.68) | 1.11 (0.75–1.63) |
| P-values for nonlinear trend | 0.003 | 0.003 | 0.003 | 0.006 | 0.05 | 0.06 | |
| Trajectories of DBP responses to dietary sodium intake | |||||||
| High sodium-sensitive | 88/192 | 35.6 | 1.42 (0.99–2.02) | 1.45 (1.01–2.07) | 1.43 (1.00–2.05) | 1.23 (0.81–1.85) | 1.22 (0.81–1.85) |
| Moderate sodium-sensitive | 297/1019 | 29.0 | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) | 1.00 (ref) |
| Sodium-resistant | 129/393 | 38.2 | 1.73 (1.29–2.32) | 1.69 (1.25–2.28) | 1.67 (1.23–2.26) | 1.57 (1.10–2.24) | 1.56 (1.09–2.24) |
| P-values for nonlinear trend | <0.001 | <0.001 | <0.001 | <0.001 | 0.03 | 0.04 | |
Model 1 adjusted for age, gender, body-mass index, SBP/DBP, physical activity, alcohol consumption, smoking, antihypertensive medication usage, 24-hour sodium excretion, and 24-hour potassium excretion at baseline.
Model 2 adjusted for all covariables in model 1 plus estimated-glomerular filtration rate.
Model 3 adjusted for all covariables in model 2 plus SBP/DBP-sodium sensitivity and resistance polygenic risk scores.
Model 4 adjusted for all covariables in model 2 plus SBP/DBP polygenic risk scores.
Compared to those with moderate SBP sodium sensitivity, individuals with high sodium sensitivity or resistance had significantly greater increases in BP from baseline during an average 7.2-year follow-up after adjustment for multiple covariables (Table S3). For example, the multiple-adjusted SBP increases were 9.5, 8.4, and 10.3 mmHg for high SBP sodium sensitivity, moderate sodium sensitivity, and sodium resistance; and 9.0, 8.4, and 10.7 mmHg for high DBP sodium sensitivity, moderate sodium sensitivity, and sodium resistance, respectively (all P<0.01 for non-linear trends). However, BP changes became not significant after further adjustment for PRS.
A J-shaped association was identified between SBP and DBP responses to sodium intervention and incidence of hypertension (P<0.001) after adjustment for multiple covariables (Figure 2). The lowest risk of incident hypertension was at approximately zero change of SBP (0.67 mmHg) and DBP (0.0 mmHg) from low-sodium to high-sodium intervention.
Figure 2. Multiple-adjusted incidence of hypertension by blood pressure responses from low-sodium to high-sodium intervention.
A. Multiple-adjusted incidence of hypertension by systolic blood pressure responses; B. Multiple-adjusted incidence of hypertension by diastolic blood pressure responses. Multiple-adjusted models included age, gender, body-mass index, systolic/diastolic blood pressure, physical activity, alcohol consumption, smoking, 24-hour urinary sodium and potassium excretion, and estimated-glomerular filtration rate at baseline.
Discussion
The GenSalt study was the largest diet-feeding study to test BP sodium sensitivity and resistance in a human population. It has contributed novel knowledge on dietary sodium intake and hypertension in several important aspects. Our findings show that sodium sensitivity is prospectively associated with incidence of hypertension and suggest that sodium sensitivity is a cause, not a consequence, of hypertension. In addition, the GenSalt study was the first human study to document that individuals with sodium resistance had a significantly higher risk for hypertension. Genetic factors partially explain this increased risk of hypertension associated with sodium sensitivity and resistance. These findings suggest that sodium sensitivity and resistance might play important roles in the pathogenesis of hypertension. Furthermore, the GenSalt study indicated that more than three-quarters of normotensive individuals were high (12.0%−18.0%) or moderate (58.3%−63.7%) sodium-sensitive. Our study findings support recommendations from randomized controlled trials that sodium reduction should be an important strategy for hypertension prevention in the general population.5
In the general population, BP responses to dietary sodium intake are a continuous, normally distributed quantitative trait.6,7 Accordingly, there have been controversies over whether sodium sensitivity and resistance are reliable phenotypes or simply the extremes of the gaussian distribution of random BP responses to dietary sodium intake. Several studies testing the reliability of sodium sensitivity reported that BP response to dietary sodium intake is a long-term reproducible and stable characteristic in the general population.6,26,27 In the GenSalt Study, the repeated low-sodium (1,180 mg/day) and high-sodium (7,081 mg/day) interventions 4.5 years apart were conducted among 487 Chinese adults using identical study protocols.26 BP responses to dietary sodium intervention in the initial and repeated studies were significantly correlated, with correlation coefficients of 0.37 during low-sodium and 0.37 during high-sodium interventions for SBP (both P<0.001). The current analysis has identified three distinct subgroups of individuals with high sensitivity, moderate sensitivity, and resistance to sodium intake who were at significantly different risk for incident hypertension. These data provide further evidence to support sodium sensitivity and resistance as important clinical phenotypes.
Sodium sensitivity was associated with an increased risk of cardiovascular disease and all-cause mortality in two small cohort studies.28,29 In a retrospective cohort study of 156 patients with hypertension, sodium sensitivity was associated with a three-fold increase in cardiovascular disease risk (relative risk 3.05; 95% CI 1.34–6.89).28 Sodium sensitivity was defined as a ≥10% difference in BP between one-week low-sodium (1–3 g salt/day) and one-week high-sodium (12–15 g salt/day) diets. In another cohort study of 596 normal or hypertensive participants, sodium sensitivity was associated with increased all-cause mortality (odds ratio 1.73, 95% CI 1.02–2.94).29 Sodium sensitivity was defined as a decrease in mean arterial BP ≥10 mmHg during a diuretic-induced sodium and volume depletion (three 40-mg doses of furosemide over eight hours) after a rapid intravenous saline loading (2 liters over 4 hours).
A few cross-sectional studies have reported that high sodium sensitivity is more common among individuals with hypertension compared to those with normal BP.6–9 However, it is unknown whether sodium sensitivity precedes the development of hypertension. In the Olivetti Heart Study, 16 participants with high sodium sensitivity, defined as a mean arterial pressure decrease of >5.7% during a three-day low-salt diet (69 mmol sodium/day), had higher incidence of hypertension compared to 20 participants with low sodium sensitivity during a 15-year follow-up.10 However, this increased risk became insignificant after adjustment for baseline BP. In the Hanzhong Adolescent Hypertension Study, sodium sensitivity was defined as a mean arterial pressure difference of >10 mmHg during oral saline loading and furosemide sodium-volume depletion.11 The unadjusted incidence of hypertension was 15.5% in 71 participants with sodium sensitivity and 6.3% in 152 participants without sodium sensitivity (P<0.05) over an 18-year follow-up. Our study found that individuals with high sodium sensitivity had a significantly increased risk of hypertension, independent of baseline BP and established risk factors for hypertension. Given the GenSalt study’s large sample size and high-quality data, these findings provide strong evidence that sodium sensitivity is an independent risk factor for the development of hypertension.
Our study identified a large proportion of normotensive individuals (23.7% SBP and 24.3% DBP) resistant to dietary sodium intervention. The proportion of sodium-resistant individuals in a population varies according to the definition of sodium resistance.8,9,14 The GenSalt study has provided the first evidence that individuals with sodium resistance are at an increased risk for incident hypertension compared to those with moderate sodium sensitivity. This finding contradicts conventional beliefs that sodium-resistant normotensives are at a lower risk for hypertension.6,7 Additionally, experimental studies reported various inbred strains of animal models of sodium-resistant hypertension and discovered underlying genetic mechanisms for some types of sodium-resistant hypertension.12,13,30 For example, the GC-A gene mutations resulted in the loss of the guanylyl cyclase-A receptor for atrial natriuretic peptide, increased peripheral resistance, and sodium-resistant hypertension.12 P2Y2 receptor knockout mice displayed multiple alterations in renal sodium and water reabsorption and developed sodium-resistant hypertension.13 Several potential mechanisms for sodium-resistant hypertension were proposed, including increased systemic vascular resistance, endothelial dysfunction, insulin resistance, and abnormalities in hormones, sodium and water transporters, and intracellular messengers.12,13,30–32 Future studies are warranted to explore the biological mechanisms of sodium-resistant hypertension in diverse populations.
Previous studies showed that age, sex, baseline BP, physical activity, alcohol consumption, smoking, sodium and potassium intake, and eGFR were associated with sodium sensitivity and risk of hypertension.3–9 Our study indicated that sodium sensitivity and resistance independently predicted the risk of hypertension and long-term changes in BP among individuals with normal BP. However, after further adjustment for PRSs, some of these associations became non-significant. These analyses suggested that genetic factors partially explained sodium-sensitivity and resistance related risk of hypertension.
In addition to the large sample size, the compliance to the low- and high-sodium interventions among the GenSalt study participants was excellent. Multiple BP measurements on each of multiple days during baseline, intervention, and follow-up were obtained using stringent quality control procedures. Therefore, our study provides reliable data to examine sodium sensitivity and resistance of BP. However, we were not able to collect data on sodium metabolism, vascular function, or other biomarkers in this large diet-feeding study, which might provide insight into the underlying mechanisms of sodium sensitivity and resistance. In addition, the GenSalt study was conducted exclusively in a Chinese population, which might limit the generalizability of the results to other racial and ethnic groups.
Perspectives
Our study found that normotensive individuals with either high sodium sensitivity or sodium resistance were at an increased risk for developing hypertension. These findings indicate that sodium sensitivity and resistance are potential causes, rather than consequences, of hypertension. In addition, these findings support personalized preventive interventions, including dietary sodium reduction, for the primary prevention of hypertension. Additional research is warranted to better understand the underlying mechanisms of sodium-sensitive and sodium-resistant hypertension and develop novel interventions for prevention and treatment. Genomics research to identify both rare and common genetic variants will help identify individuals who are predisposed to sodium-sensitive or sodium-resistant hypertension. Furthermore, discovery and validation of biomarkers to identify individuals with sodium sensitivity or resistance using proteomics and metabolomics will increase accuracy and facilitate implementation in clinical and public health settings.
Supplementary Material
Novelty and Significance.
What Is New?
Based on blood pressure responses to a 7-day low-sodium and a 7-day high-sodium intervention, three subgroups of individuals with high sodium sensitivity, moderate sodium sensitivity, and sodium resistance were identified.
Compared to individuals with moderate sodium sensitivity of systolic blood pressure, multiple-adjusted odds ratios for incident hypertension were 1.44 for those with high sodium sensitivity and 1.42 for those with sodium resistance.
What Is Relevant?
Our study indicates that normotensive individuals with either high sodium sensitivity or sodium resistance are at an increased risk for developing hypertension.
Summary
These findings suggest that sodium sensitivity and resistance may play an important role in the pathogenesis of hypertension.
Acknowledgments
We thank Ms. Katherine Obst for her editorial support of this work.
Sources of Funding
This work was supported by research grants U01HL072507, R01HL087263, and R01HL090682 from the National Heart, Lung, and Blood Institute, and P20GM109036 from the National Institute of General Medical Sciences, Bethesda, Maryland.
Footnotes
Disclosures
None
References
- 1.Mills KT, Bundy JD, Kelly TN, Reed JE, Kearney PM, Reynolds K, Chen J, He J. Global Disparities of Hypertension Prevalence and Control: A Systematic Analysis of Population-Based Studies From 90 Countries. Circulation. 2016;134:441–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Forouzanfar MH, Liu P, Roth GA, Ng M, Biryukov S, Marczak L, Alexander L, Estep K, Hassen Abate K, Akinyemiju TF, et al. Global Burden of Hypertension and Systolic Blood Pressure of at Least 110 to 115 mm Hg, 1990–2015. JAMA. 2017;317:165–182. [DOI] [PubMed] [Google Scholar]
- 3.Elliott P, Stamler J, Nichols R, Dyer AR, Stamler R, Kesteloot H, Marmot M. Intersalt revisited: further analyses of 24 hour sodium excretion and blood pressure within and across populations. Intersalt Cooperative Research Group. BMJ. 1996;312:1249–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sacks FM, Svetkey LP, Vollmer WM, Appel LJ, Bray GA, Harsha D, Obarzanek E, Conlin PR, Miller ER, Simons-Morton DG, Karanja N, Lin PH. Effects on blood pressure of reduced dietary sodium and the Dietary Approaches to Stop Hypertension (DASH) diet. DASH-Sodium Collaborative Research Group. N Engl J Med. 2001;344:3–10. [DOI] [PubMed] [Google Scholar]
- 5.Newberry SJ, Chung M, Anderson CAM, Chen C, Fu Z, Tang A, Zhao N, Booth M, Marks J, Hollands S, Motala A, Larkin J, Shanman R, Hempel S. AHRQ Comparative Effectiveness Reviews Sodium and Potassium Intake: Effects on Chronic Disease Outcomes and Risks Rockville (MD): Agency for Healthcare Research and Quality (US); 2018. [PubMed]
- 6.Liu Y, Shi M, Dolan J, He J. Sodium sensitivity of blood pressure in Chinese populations. J Hum Hypertens. 2020;34:94–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Elijovich F, Weinberger MH, Anderson CA, Appel LJ, Bursztyn M, Cook NR, Dart RA, Newton-Cheh CH, Sacks FM, Laffer CL. Salt Sensitivity of Blood Pressure: A Scientific Statement From the American Heart Association. Hypertension. 2016;68:e7–e46. [DOI] [PubMed] [Google Scholar]
- 8.Wright JT Jr., Rahman M, Scarpa A, Fatholahi M, Griffin V, Jean-Baptiste R, Islam M, Eissa M, White S, Douglas JG. Determinants of salt sensitivity in black and white normotensive and hypertensive women. Hypertension. 2003;42:1087–92. [DOI] [PubMed] [Google Scholar]
- 9.He J, Gu D, Chen J, Jaquish CE, Rao DC, Hixson JE, Chen JC, Duan X, Huang JF, Chen CS, Kelly TN, Bazzano LA, Whelton PK. Gender difference in blood pressure responses to dietary sodium intervention in the GenSalt study. J Hypertens. 2009;27:48–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Barba G, Galletti F, Cappuccio FP, Siani A, Venezia A, Versiero M, Della Valle E, Sorrentino P, Tarantino G, Farinaro E, Strazzullo P. Incidence of hypertension in individuals with different blood pressure salt-sensitivity: results of a 15-year follow-up study. J Hypertens. 2007;25:1465–71. [DOI] [PubMed] [Google Scholar]
- 11.Mu J, Zheng S, Lian Q, Liu F, Liu Z. Evolution of blood pressure from adolescents to youth in salt sensitivies: a 18-year follow-up study in Hanzhong children cohort. Nutr J. 2012;11:70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Lopez MJ, Wong SK, Kishimoto I, Dubois S, Mach V, Friesen J, Garbers DL, Beuve A. Salt-resistant hypertension in mice lacking the guanylyl cyclase-A receptor for atrial natriuretic peptide. Nature. 1995;378:65–8. [DOI] [PubMed] [Google Scholar]
- 13.Rieg T, Bundey RA, Chen Y, Deschenes G, Junger W, Insel PA, Vallon V. Mice lacking P2Y2 receptors have salt-resistant hypertension and facilitated renal Na+ and water reabsorption. FASEB J. 2007;21:3717–26. [DOI] [PubMed] [Google Scholar]
- 14.Weinberger MH, Miller JZ, Luft FC, Grim CE, Fineberg NS. Definitions and characteristics of sodium sensitivity and blood pressure resistance. Hypertension. 1986;8:Ii127–34. [DOI] [PubMed] [Google Scholar]
- 15.GenSalt: rationale, design, methods and baseline characteristics of study participants. J Hum Hypertens. 2007;21:639–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Pickering TG, Hall JE, Appel LJ, Falkner BE, Graves J, Hill MN, Jones DW, Kurtz T, Sheps SG, Roccella EJ. Recommendations for blood pressure measurement in humans and experimental animals: Part 1: blood pressure measurement in humans: a statement for professionals from the Subcommittee of Professional and Public Education of the American Heart Association Council on High Blood Pressure Research. Hypertension. 2005;45:142–61. [DOI] [PubMed] [Google Scholar]
- 17.Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J. A new equation to estimate glomerular filtration rate. Ann Intern Med. 2009;150:604–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Jones BL, Nagin DS, Roeder K. A SAS Procedure Based on Mixture Models for Estimating Developmental Trajectories. Sociol Methods Res. 2001;29:374–393. [Google Scholar]
- 19.Taliun D, Harris DN, Kessler MD, Carlson J, Szpiech ZA, Torres R, Taliun SAG, Corvelo A, Gogarten SM, Kang HM, et al. Sequencing of 53,831 diverse genomes from the NHLBI TOPMed Program. bioRxiv. 2019:563866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Evangelou E, Warren HR, Mosen-Ansorena D, Mifsud B, Pazoki R, Gao H, Ntritsos G, Dimou N, Cabrera CP, Karaman I, et al. Genetic analysis of over 1 million people identifies 535 new loci associated with blood pressure traits. Nat Genet. 2018;50:1412–1425. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Giri A, Hellwege JN, Keaton JM, Park J, Qiu C, Warren HR, Torstenson ES, Kovesdy CP, Sun YV, Wilson OD, et al. Trans-ethnic association study of blood pressure determinants in over 750,000 individuals. Nat Genet. 2019;51:51–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hoffmann TJ, Ehret GB, Nandakumar P, Ranatunga D, Schaefer C, Kwok PY, Iribarren C, Chakravarti A, Risch N. Genome-wide association analyses using electronic health records identify new loci influencing blood pressure variation. Nat Genet. 2017;49:54–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Warren HR, Evangelou E, Cabrera CP, Gao H, Ren M, Mifsud B, Ntalla I, Surendran P, Liu C, Cook JP, et al. Genome-wide association analysis identifies novel blood pressure loci and offers biological insights into cardiovascular risk. Nat Genet. 2017;49:403–415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hall JE. Renal Dysfunction, Rather Than Nonrenal Vascular Dysfunction, Mediates Salt-Induced Hypertension. Circulation. 2016;133:894–906. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Durrleman S, Simon R. Flexible regression models with cubic splines. Stat Med. 1989;8:551–61. [DOI] [PubMed] [Google Scholar]
- 26.Gu D, Zhao Q, Chen J, Chen JC, Huang J, Bazzano LA, Lu F, Mu J, Li J, Cao J, Mills K, Chen CS, Rice T, Hamm LL, He J. Reproducibility of blood pressure responses to dietary sodium and potassium interventions: the GenSalt study. Hypertension. 2013;62:499–505. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Sharma AM, Schattenfroh S, Kribben A, Distler A. Reliability of salt-sensitivity testing in normotensive subjects. Klin Wochenschr. 1989;67:632–4. [DOI] [PubMed] [Google Scholar]
- 28.Morimoto A, Uzu T, Fujii T, Nishimura M, Kuroda S, Nakamura S, Inenaga T, Kimura G. Sodium sensitivity and cardiovascular events in patients with essential hypertension. Lancet. 1997;350:1734–7. [DOI] [PubMed] [Google Scholar]
- 29.Weinberger MH, Fineberg NS, Fineberg SE, Weinberger M. Salt sensitivity, pulse pressure, and death in normal and hypertensive humans. Hypertension. 2001;37:429–32. [DOI] [PubMed] [Google Scholar]
- 30.Kimura G, Frem GJ, Brenner BM. Renal mechanisms of salt sensitivity in hypertension. Curr Opin Nephrol Hypertens. 1994;3:1–12. [PubMed] [Google Scholar]
- 31.DuPont JJ, Greaney JL, Wenner MM, Lennon-Edwards SL, Sanders PW, Farquhar WB, Edwards DG. High dietary sodium intake impairs endothelium-dependent dilation in healthy salt-resistant humans. J Hypertens. 2013;31:530–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.St Lezin EM, Pravenec M, Wong AL, Liu W, Wang N, Lu S, Jacob HJ, Roman RJ, Stec DE, Wang JM, Reid IA, Kurtz TW. Effects of renin gene transfer on blood pressure and renin gene expression in a congenic strain of Dahl salt-resistant rats. J Clin Invest. 1996;97:522–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.