The 24-h recall used in NHANES is less prone to bias compared to other dietary assessment methods, such as the food frequency questionnaire [
31]. This notion is further reinforced by earlier studies including NHANES I [
12], NHANES II [
11], and NHANES III [
32], which identified a comparable relationship between low sodium intake and a rise in all-cause mortality using the Food Frequency Questionnaires. These findings lend weight to our conclusions.
The analysis of the NHANES data in the present study demonstrated the associate between low sodium intake, low potassium intake, and a high sodium–potassium ratio with high risk in mortality in the univariate model among individuals aged 40 years and older. Intriguingly, this correlation remained significant even after accounting for other covariates. Nevertheless, a more detailed subgroup analysis uncovered further noteworthy findings.
Sodium intake
The NHANES data provide us with a wide range of age-related information, allowing us to analyze the impact of age on the sodium-mortality risk relationship. This study found that for people aged 40 and above, daily sodium intake is negatively correlated with all-cause mortality. Our study uncovers a negative association between daily sodium intake and all-cause mortality for individuals over 40, corroborating findings from previous research conducted by Messerli FH et al. [
5] and O'Donnell M et al. [
6]. While a reduced sodium intake may diminish peak blood pressure and lower the likelihood of hypertension-related cardiovascular events, an exceedingly low intake can potentially stimulate the renin–angiotensin–aldosterone system (RAAS) [
33,
34], impacting catecholamine and lipid metabolism and consequently raising the mortality risk. On conducting subgroup analyses, we observed a significant interaction between age and daily sodium intake. Consequently, we segmented the population based on age, which revealed a distinct negative relationship between daily sodium consumption and all-cause mortality among those aged 40–60 years. However, this correlation was not significant in the 60–80 age bracket, a finding that aligns with Kalogeropoulos AP et al.'s [
15] conclusion from their ten-year longitudinal study of 2,642 individuals aged 71–80. Notably, post-adjustment for covariates, the dose–response curve in the 60–80 age segment suggested a U-shaped relationship between the HR and daily sodium intake. This implies that both excessively high or low sodium intake could escalate the risk of all-cause mortality within this demographic. In addition, our study indicates that, compared to a low-sodium diet, men derive greater benefits from normal or high sodium intake. Currently, there is no literature reporting the reasons for this; however, we hypothesize that this may be attributed to men having higher energy requirements than women. A correlation between sodium and energy intake is evident, and low sodium intake may suggest a lower nutritional status.
Our study also reveals a correlation between higher sodium intake and a younger age demographic. However, this differential did not correspond to an anticipated increase in all-cause mortality risk. Given the dynamics of sodium metabolism, this result may align with reality. Typically, younger individuals exhibit eGFR and robust renal regulation of sodium metabolism [
35]. Thus, it is plausible to hypothesize significant variations in sodium metabolism inflection points across distinct age groups, partially accounting for the observed statistical differences in sodium intake among various age subgroups. For the elderly, the fragility of renal sodium metabolism regulation potentially escalates mortality risk associated with excessively high or low sodium intake. This necessitates further research to establish tailored daily sodium intake recommendations for older individuals. Additionally, we discerned a similar correlation curve between sodium intake and all-cause mortality in CKD patients, lending further credibility to our hypothesis (see Supplement Figure 8).
Potassium intake
Our study identified a significant increase in all-cause mortality when daily potassium intake falls below 3500 mg, a result aligning with O'Donnell M et al.'s findings [
6]. Prior studies [
36,
37] highlight the integral role of potassium in blood pressure regulation, which our subgroup analysis, focused on hypertensive individuals, corroborates. Notably, we discovered that elevating daily potassium intake considerably mitigates all-cause mortality among hypertensive individuals (Supplement Figure 5). The antihypertensive effect of potassium, attributed to a variety of mechanisms, is well-recognized. These mechanisms encompass stimulation of natriuresis, enhancement of endothelial function, release of nitric oxide (NO), increased Na–K pump function, amplified membrane potassium channel activity, resulting in vasodilation and subdued sympathetic nervous system activity, thereby inducing arterial muscle relaxation [
36].
Sodium–potassium ratio
The sodium-to-potassium ratio is a significant factor in managing blood pressure and overall health. It is more strongly associated with blood pressure outcomes than either sodium or potassium alone, particularly in hypertensive adult populations [
22]. What’s more? A study of healthy Greek adults found that food sodium intake was positively correlated with energy intake and food potassium intake [
38]. The correlations observed between sodium or potassium intake and all-cause mortality may potentially reflect influences of nutritional status or the intensity of physical activity. The residual effects of these factors, which may not be entirely mitigated by statistical adjustments. Given the strong correlation between sodium and potassium intake and energy intake, employing the sodium–potassium ratio as a monitoring measure might be fitting, Current researches indicate that the sodium-to-potassium ratio, as an indicator of sodium and potassium intake, has been shown to be unrelated to total energy intake [
39,
40]. From a statistical perspective, it is more effective in preventing biases introduced by reverse causation. Yet, there exist no formal recommendations within the United States advocating for such a use. Various studies indicate that cardiovascular event or composite risk heat maps demonstrate the lowest risk in the moderate sodium intake range of 3–5 g/d [
41], accompanied by a higher potassium intake. The WHO stipulates the optimal sodium–potassium ratio as less than1 mmol/mmol (below 0.6 mg/mg) [
42], but relevant research [
43] validating the association between this ratio and all-cause mortality remains scarce. Current studies [
43‐
45] concerning the sodium–potassium ratio primarily revolve around its connection with hypertension, with limited literature on its relation to all-cause mortality. Some research findings [
46,
47] point towards a correlation between a high sodium–potassium ratio and the onset of cardiovascular diseases. In our study, in our study's restricted cubic spline model, we observed that as the sodium-to-potassium ratio exceeded 1.203 mg/mg/day, the overall mortality rate showed an increasing trend (see Fig.
4). Therefore, integrating our results with previous studies, we posit that maintaining a balanced sodium and potassium intake ratio within an appropriate range is vital for controlling all-cause mortality in the population. However, verifying the reliability of this conclusion necessitates further clinical research.
The findings of our study suggest that the WHO current recommendations on daily sodium and potassium intake for the general population might be oversimplified. The advisable amounts of daily sodium and potassium intake should ideally differ among various age groups and individuals with distinct underlying medical conditions [
48]. Hence, the necessity for age-specific and individualized assessment techniques for evaluating sodium and potassium intake is evident. Nevertheless, focusing exclusively on a single nutrient indicator might yield inaccuracies, given that sodium and potassium intake can act as proxies for consumption patterns. Moreover, an excessively high or low sodium–potassium ratio might suggest poor dietary structure, resulting in an escalation in all-cause mortality. Recent reports [
49,
50] discovered that, during 2015–2016, primary contributors to sodium intake were processed foods or store-bought foods with added sodium, such as deli sandwiches, poultry, or sodium-enhanced vegetables. In contrast, potassium intake predominantly originated from naturally low-sodium foods, such as plain milk, fruits, and vegetables, along with processed foods. Past researches [
49,
51] indicate that individuals with higher income might maintain a more balanced daily dietary nutrient structure, thus achieving a sodium–potassium ratio closer to the optimum range. This could be one factor contributing to the disparity in all-cause mortality between high-income and low-income populations. However, even with similar daily dietary nutrient structures (similar sodium–potassium ratios), our study still identified substantial differences in all-cause mortality between high-income and low-income populations, as depicted in Supplement Figure 3, Supplement Figure 5, Supplement Figure 7, with the most conspicuous disparities occurring in the context of high sodium and potassium intake, This phenomenon could be attributed to disparities in healthcare access and conditions available to the two demographic groups in their daily lives.
Presently, the existing evidence surrounding dietary interventions is scant, with most conclusions extrapolated from the DASH study. Adjustments to other dietary structures, along with their beneficial or detrimental effects, cannot be exclusively ascribed to a singular electrolyte. Hence, additional controlled experiments are imperative to substantiate these conclusions [
52].
There exist some limitations to the findings of this study. Firstly, the dietary data and medical conditions were self-reported, rendering them susceptible to recall and social desirability biases. Furthermore, our estimations do not account for the sodium present in salt added at the table, which represent 5–6% [
53] of total intake, which may poorly reflect the variability in seasoning use. Secondly, the possibility of reverse causality could introduce a bias in the estimation of effects. Certain studies include populations with cardiovascular or other chronic diseases. Individuals from these groups might have been recommended to limit their sodium intake due to their potential diseases, thereby inadvertently associating reduced sodium intake with a marked rise in the risk of all-cause mortality among the low-level population [
16].
Our findings highlighted that low sodium intake is associated with an increased overall mortality rate. Additional evidence is required to determine the potential benefits of restricted sodium intake for young individuals or those with normal kidney function. Secondly, the benefits of sodium and potassium intake vary among different populations, necessitating personalized recommendations for each group. Lastly, focusing solely on sodium and potassium intake may be one-sided. A healthy diet should be based on reasonable sodium intake and include an appropriate sodium-to-potassium ratio.