Inverse association between dietary potassium-to-sodium ratio and systemic inflammatory markers: A cross-sectional analysis of the NHANES 2021–2023 data

Abstract

Systemic chronic inflammation (SCI) is associated with a range of noncommunicable diseases. Emerging evidence indicates that the dietary potassium-to-sodium (K/Na) ratio may predict cardiovascular and inflammatory health more accurately than potassium or sodium alone. However, its relationship with established systemic chronic inflammation markers (SCI markers) has not yet been thoroughly investigated. This study was conducted to examine the association between dietary K/Na ratio and serum concentrations of alpha-1-acid glycoprotein (AGP) and C-reactive protein (CRP), and to explore the shape of these dose-response relationships. We performed a cross-sectional analysis of 892 adults aged ≥ 20 years using data from the National Health and Nutrition Examination Survey (NHANES) 2021 to 2023. Dietary intake was assessed via two 24-hour recalls, from which individual potassium and sodium intakes were used to calculate the K/Na ratio. Serum alpha-1-acid glycoprotein (AGP) and high-sensitivity C-reactive protein (hs-CRP) were measured using standardized immunoturbidimetric assays. We applied survey-weighted multivariable linear regression models to examine adjusted associations between the continuous K/Na ratio and each inflammatory marker, controlling for age, sex, race/ethnicity, education, income, smoking, physical activity, body mass index, hypertension, diabetes, and lipid profiles. Generalized additive models (GAMs) with penalized splines were used to assess potential nonlinear dose-response relationships. Sensitivity analyses excluded participants with CRP > 10 mg/L and explored subgroup effects by sex and race/ethnicity. After full adjustment, each 1-unit increase in dietary K/Na ratio was associated with a 0.068 mg/dL decrease in AGP (SE = 0.021; P = .001) and a 0.248 decrease in ln-transformed hs-CRP (SE = 0.095; P = .009). GAM analyses revealed a significant nonlinear inverse relationship between K/Na ratio and AGP (effective degrees of freedom = 1.49; P = .007), whereas the association with ln-CRP remained linear (edf = 1.33; P = .499). Findings were consistent across sensitivity and subgroup analyses. In this nationally representative US sample, a higher dietary K/Na ratio is independently associated with lower systemic inflammation, supporting public health strategies that promote improved potassium-to-sodium balance to mitigate chronic disease risk.

1. Background

Systemic chronic inflammation (SCI) is a persistent, low-grade inflammatory state recognized as a core pathological process in most noncommunicable diseases, which represent the leading cause of death globally.^[1] The World Health Organization identifies chronic inflammatory diseases as the most significant cause of death worldwide, with conditions such as cardiovascular disease, cancer, type 2 diabetes, chronic kidney disease, and neurodegenerative disorders all sharing an inflammatory basis.^[2] This health-damaging phenotype is not an inevitable consequence of aging but is profoundly influenced by modifiable environmental and lifestyle factors, including physical inactivity, psychological stress, and, most notably, diet. The modern dietary pattern, characterized by high intakes of saturated fats, refined sugars, and imbalanced electrolyte profile, is a potent driver of SCI.^[3,4] Given that SCI represents a common mechanistic pathway for a wide range of human diseases, identifying and characterizing modifiable dietary factors that can influence this process is of paramount importance for public health.

For decades, nutritional researches have focused on the individual roles of dietary sodium and potassium in health, particularly regarding blood pressure regulation and cardiovascular health. Extensive evidence has established the adverse effects of high sodium intake, which promotes hypertension, endothelial dysfunction, and increased cardiovascular morbidity and mortality.^[5] Conversely, adequate potassium intake has been shown to attenuate these effects, contributing to blood pressure reduction and a lower risk of stroke and cardiovascular disease.^[6] However, a critical paradigm shift has occurred in recent years, with growing recognition that the dynamic interrelationship between these 2 electrolytes is more biologically significant than the absolute intake of either 1 in isolation. Mechanistically, the body responds to the relative balance of sodium and potassium rather than to absolute levels of either nutrient alone, as these cations exert opposing biological effects through complex physiological interactions at both cellular and systemic levels.^[7] High dietary sodium induces vasoconstriction and increases blood pressure, whereas elevated potassium counteracts these effects by promoting natriuresis through downregulation of the thiazide-sensitive sodium-chloride cotransporter(NCC) and modulating the renin-angiotensin-aldosterone system.^[8] These mechanistic interactions explain why the sodium-to-potassium ratio (or its inverse, the potassium-to-sodium [K/Na] ratio) has been consistently demonstrated to be a stronger and more reliable predictor of hypertension, cardiovascular events, and all-cause mortality than either nutrient alone.^[9]

C-reactive protein (CRP) and α1-acid glycoprotein (AGP) are well-established indicators of systemic inflammation, providing complementary information on both acute and chronic inflammatory states. CRP is primarily produced by hepatocytes in response to pro-inflammatory cytokines, particularly interleukin-6 (IL-6), with interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) serving as secondary inducers. The IL-6–mediated STAT3 signaling pathway plays a central role in CRP gene regulation. CRP levels rise rapidly within 48 hours during acute inflammation and are strongly associated with increased risks of coronary heart disease, stroke, all-cause mortality, and cardiovascular-related death, independent of traditional risk factors.^[10–12]

AGP, another major positive acute-phase protein regulated byIL-6, IL-1β, and TNF-α, exhibits slower kinetics and remains elevated for prolonged periods, thus reflecting both acute and chronic inflammatory processes. Beyond its role as a passive marker, AGP exerts immunomodulatory effects on neutrophil activation, migration, and adhesion, and distinct AGP glycoforms differentially modulate platelet aggregation and intracellular signaling. The immunoregulatory function of AGP is critically determined by its glycosylation pattern, which undergoes substantial alterations depending on the nature and stage of inflammation. Beyond its role as a passive marker, AGP exerts immunomodulatory effects on neutrophil activation, migration, and adhesion, and distinct AGP glycoforms differentially modulate platelet aggregation and intracellular signaling. The immunoregulatory function of AGP is critically determined by its glycosylation pattern, which undergoes substantial alterations depending on the nature and stage of inflammation.^[13,14]

Combined measurement of CRP and AGP provides a more comprehensive evaluation of systemic inflammation by capturing different temporal and functional aspects of the inflammatory response. This multiparametric approach allows classification of inflammation into distinct stages and enhances prognostic accuracy and risk stratification across various clinical conditions, including sepsis, acute exacerbations of chronic diseases, and other inflammatory disorders.

While substantial evidence links a favorable K/Na ratio to improved clinical outcomes, particularly for cardiovascular and renal health, the association with underlying systemic inflammation is less well characterized. A potential antiinflammatory effect of high K/Na diet could represent a crucial mechanistic link explaining these broad health benefits. To date, no study has comprehensively examined the relationship between the dietary K/Na ratio and inflammatory markers including both AGP and CRP within a large, nationally representative sample of US population.

Therefore, the primary objective of this study was to determine the association between the dietary K/Na ratio and serum concentrations of AGP and CRP using recent data from the National Health and Nutrition Examination Survey (NHANES). The secondary objective was to explore the shape of these associations (linear vs nonlinear) to provide detailed characterization of the dose-response relationship between dietary electrolyte balance and systemic inflammation.

2. Materials and methods

This study used data from the National Health and Nutrition Examination Survey (NHANES), a program designed to assess the health and nutritional status of adults and children in the United States. NHANES is a continuous, cross-sectional survey conducted by the National Center for Health Statistics (NCHS) of the Centers for Disease Control and Prevention (CDC). It employs a complex, stratified, multistage probability sampling design to select a nationally representative sample of the civilian, non-institutionalized US population. The NHANES protocol was reviewed and approved by the National Center for Health Statistics (NCHS) Research Ethics Review Board (Protocol #2018-01). All participants provided written informed consent prior to participation. As this analysis used publicly available, de-identified secondary data, additional Institutional Review Board (IRB) approval was not required in accordance with the U.S. Department of Health and Human Services regulations (45 CFR 46.104(d)(4)).^[15] Data collection includes demographic, socioeconomic, dietary, and health-related questionnaires; physical examinations; and laboratory tests administered by highly trained medical personnel.

For this analysis, we used data from the most recently available continuous cycle, NHANES 2021–2023. The study protocol was approved by the NCHS Research Ethics Review Board, and all participants provided written informed consent. The initial sample included 11,933 participants of all ages. Of these, 7809 adults aged ≥ 20 years were identified. We excluded participants without valid data for dietary potassium or sodium intake (n = 3011), and those with missing serum AGP or CRP concentrations (n = 3006). Finally, we excluded individuals with missing covariate data required for multivariable models (n = 900). After applying these exclusion criteria, the final analytical sample comprised 892 participants (Fig. 1).

Figure 1.

Flow diagram of participants meeting the inclusion and exclusion criteria.

2.1. Exposure assessment: dietary potassium-to-sodium (K/Na) ratio

Dietary intake data in NHANES were collected through two 24-hour dietary recalls. The first recall was conducted in person by a trained interviewer at the Mobile Examination Center (MEC), and the second was conducted via telephone 3 to 10 days later. Both recalls were administered using the NCHS Automated Multiple-Pass Method (AMPM) to ensure data accuracy and completeness. Nutrient intakes were calculated using the US Department of Agriculture’s Food and Nutrient Database for Dietary Studies.

The primary exposure variable was the dietary potassium-to-sodium (K/Na) ratio. The average daily intake of potassium (in mg) and sodium (in mg) was calculated from the two 24-hour recalls. The K/Na ratio was then computed for each participant by dividing average daily potassium intake by average daily sodium intake. For descriptive analyses, participants were categorized into quartiles based on the distribution of their K/Na ratio. For regression analyses, the K/Na ratio was treated as a continuous variable to maximize statistical power and assess dose-response relationship.

2.2. Outcomes: inflammatory biomarkers

The primary outcomes were the serum concentrations of 2 systemic inflammatory markers: alpha-1-acid glycoprotein (AGP) and C-reactive protein (CRP). Blood samples were collected from participants at the MEC, processed, stored, and shipped to designated laboratories for analysis.

Serum CRP was measured using a high-sensitivity latex-enhanced immunoturbidimetric assay. This assay has a lower detection limit suitable for quantifying low-grade inflammation associated with chronic disease risk. Due to the characteristic right-skewed distribution of CRP values in the general population, CRP data were natural log-transformed (log-CRP) before regression analysis to approximate a normal distribution and satisfy assumptions of linear modeling.

Serum AGP was quantified using a standardized immunoturbidimetric assay. AGP concentrations were analyzed as a continuous variable in original units (mg/dL).

2.3. Covariates

Potential confounding variables was selected based on established literature and biological plausibility. All covariate data were obtained from NHANES questionnaires, physical examinations, and laboratory components. Selected covariates were as follows:

1. Demographic characteristics: Age (years, continuous) and race/ethnicity (categorized as non-Hispanic White, non-Hispanic Black, Mexican American, Hispanic, and other race);

2. Socioeconomic Status: Educational attainment (categorized as <9th grade, 9th-11th grade, high school graduate, some college or associate degree, and college graduate or above) and poverty-to-income ratio (PIR), a continuous variable calculated by dividing family income by US Census Bureau poverty thresholds;

3. Lifestyle Factors: Smoking status (categorized as never, former, or current smoker) and alcohol consumption (categorized based on standard NHANES definitions);

4. Anthropometric and Clinical Measures: Body mass index (BMI), calculated as weight in kilograms divided by height in meters squared (kg/m²), treated as a continuous variable. Hypertension was defined as a self-reported diagnosis, use of antihypertensive medication, or average systolic blood pressure ≥ 140 mm Hg or diastolic blood pressure ≥ 90 mm Hg. Diabetes mellitus status was defined based on self-reported diagnosis, use of insulin or oral hypoglycemic agents, fasting plasma glucose level ≥ 126 mg/dL, or a hemoglobin A1c ≥ 6.5%.

2.4. Statistical analysis

All statistical analyses were performed accounting for the complex, multistage probability sampling design of NHANES. This was accomplished by using appropriate sample weights, strata, and primary sampling units (PSUs) as specified in the NCHS analytic guidelines. This approach ensures that results are nationally representative of the US civilian noninstitutionalized adult population.

Descriptive statistics were used to summarize study population characteristics across quartiles of the dietary K/Na ratio. Continuous variables were presented as mean ± standard deviation (SD), and categorical variables were presented as frequency counts and percentages (n, %). To test for trends across quartiles, weighted linear regression was used for continuous variables (modeling the median value of each quartile) and the Rao-Scott chi-square test was used for categorical variables. Standardized mean differences (SMDs) were calculated to compare characteristics of the highest quartile (Q4) to the lowest quartile (Q1), with SMD > 0.1 considered a meaningful difference.

Multivariable linear regression models were constructed to examine the independent association of the continuous K/Na ratio with serum AGP and log-CRP concentrations. A series of nested models were built to assess the impact of confounding. The final, fully adjusted models included age, race/ethnicity, PIR, education level, BMI, smoking status, alcohol use, hypertension, and diabetes mellitus.

To investigate potential nonlinear relationships, we employed generalized additive models (GAMs). GAMs extend linear models by allowing nonparametric smooth functions of predictor variables. We fitted GAMs with penalized cubic regression spline for the K/Na ratio, adjusting for all covariates included in the final linear models. The statistical significance of the nonlinear component was assessed by examining the P-value for the smooth term. An effective degree of freedom (edf) value >1 suggests departure from linearity.

Model diagnostics were performed for final linear regression models. Variance inflation factors (VIFs) were calculated for all predictor variables to assess multicollinearity. VIF values < 2.0 were considered to indicate no significant collinearity issues. Data analysis was conducted using R version 4.3.1 (R Foundation, Vienna, Austria). For all analyses, 2-sided P-value of <.05 was considered statistically significant.

3. Results

3.1. Characteristics of the study population

The analysis included 892 adults, representing the US population aged 20 years and older. The baseline characteristics of the study participants, stratified by quartiles of the dietary K/Na ratio, are presented in Table 1. Several significant trends were observed across the quartiles. Individuals in the highest K/Na ratio quartile (Q4) were, on average, older than those in the lowest quartile (Q1) (37.5 ± 7.6 years vs 33.6 ± 8.2 years, P < .001). A higher K/Na ratio was associated with a higher socioeconomic status, as indicated by a significant trend towards higher education levels and a higher poverty-to-income ratio (PIR) across the quartiles (PIR: 3.04 ± 1.65 in Q4 vs 2.34 ± 1.59 in Q1, P < .001).

Table 1.

Participant characteristics by dietary potassium-to-sodium (K/Na) ratio quartiles.

	Q1 (n = 223)	Q2 (n = 223)	Q3 (n = 223)	Q4 (n = 223)	SMD	P-value
Age (years)	33.6 ± 8.18	35.39 ± 7.91	35.55 ± 8.05	37.51 ± 7.59	0.25	<.001
Race					0.211	.212
Mexican American	16 (7.2)	21 (9.4)	26 (11.7)	18 (8.1)
Hispanic	24 (10.8)	31 (13.9)	29 (13.0)	33 (14.8)
Non-Hispanic White	110 (49.3)	107 (48.0)	118 (52.9)	116 (52.0)
Non-Hispanic Black	44 (19.7)	32 (14.3)	20 (9.0)	33 (14.8)
Other Race	29 (13.0)	32 (14.3)	30 (13.5)	23 (10.3)
Education level					0.25	.037
< 9th grade	4 (1.8)	6 (2.7)	3 (1.3)	7 (3.1)
9–11th grade	13 (5.8)	16 (7.2)	9 (4.0)	12 (5.4)
High school graduate	41 (18.4)	29 (13.0)	29 (13.0)	30 (13.5)
Some college or AA degree	84 (37.7)	67 (30.0)	77 (34.5)	54 (24.2)
College graduate or above	81 (36.3)	105 (47.1)	105 (47.1)	120 (53.8)
Poverty Income Ratio	2.34 ± 1.59	2.66 ± 1.68	2.83 ± 1.73	3.04 ± 1.65	0.228	<.001
BMI (kg/m2)	31.83 ± 8.64	30.51 ± 9.57	29.36 ± 8.10	29.55 ± 7.80	0.164	.01
Alcohol	18 (9.4)	16 (7.9)	15 (7.3)	16 (8.3)	0.127	.413
Smoking					0.096	.663
current	60 (26.9)	59 (26.5)	56 (25.1)	64 (28.7)
never	163 (73.1)	162 (72.6)	167 (74.9)	159 (71.3)
Hypertension, n (%)	44 (19.7)	29 (13.0)	32 (14.3)	28 (12.6)	0.138	.251
Diabetes, n (%)	15 (6.7)	10 (4.5)	12 (5.4)	8 (3.6)	0.127	.413
AGP (mg/dL)	0.84 ± 0.24	0.80 ± 0.26	0.77 ± 0.24	0.75 ± 0.24	0.213	<.001
CRP (mg/dL)	5.99 ± 9.28	4.28 ± 6.33	5.13 ± 11.98	3.53 ± 5.69	0.166	.019

^*Data are presented as mean ± SD for continuous variables and n (%) for categorical variables. p-values are for trend across quartiles.

SMD = standardized mean difference.

Conversely, participants with a higher K/Na ratio tended to have a healthier clinical profile. Body Mass Index (BMI) was significantly lower in the higher quartiles (29.6 ± 7.8 kg/m2 in Q4 vs 31.8 ± 8.6 kg/m2 in Q1, P = .01). While the trends for hypertension and diabetes prevalence did not reach statistical significance in the descriptive analysis, there was a pattern of lower prevalence in the higher K/Na ratio groups.

Most importantly, a clear and statistically significant inverse relationship was observed between the K/Na ratio and the primary inflammatory outcomes. Mean serum AGP concentrations progressively decreased from the lowest to the highest quartile (0.84 ± 0.24 mg/dL in Q1 to 0.75 ± 0.24 mg/dL in Q4, P < .001). A similar significant trend was observed for serum CRP levels (5.99 ± 9.28 mg/L in Q1 to 3.53 ± 5.69 mg/L in Q4, P = .019).

3.2. Multivariable linear association of K/Na ratio with inflammatory markers

The results of the multivariable linear regression analyses are detailed in Table 2. After adjusting for a comprehensive set of demographics, socioeconomic status, lifestyle, and clinical confounders, the dietary K/Na ratio remained a significant and independent predictor of both inflammatory markers.

Table 2.

Multivariable linear regression models for the association of K/Na ratio with AGP and log (CRP).

Model	Variable	β Estimate	Std. Error	t-value	P-value
AGP	K/Na Ratio	−0.067	0.021	−3.24	.001
	Race (ref: White)
	└ Black	0.006	0.034	0.18	.859
	└ Hispanic	0.044	0.029	1.5	.135
	└ Other	0.004	0.034	0.13	.894
	└ Asian	−0.03	0.034	−0.88	.378
	Smoking (ref: Never)
	└ Current	0.041	0.017	2.378	.018
	Diabetes (ref: No)
	└ Yes	0.043	0.036	1.18	.240
	Hypertension (ref: No)
	└ Yes	0.098	0.151	0.65	.517
	PIR	−0.015	0.005	−3.13	.002
	BMI	0.015	0.001	15.5	<.001
log(CRP)	K/Na Ratio	−0.248	0.095	−2.62	.009
	Age	−0.0004	0.0046	−0.09	.932
	Race (ref: White)
	└ Black	-0.097	0.153	−0.63	.527
	└ Hispanic	−0.215	0.133	−1.62	.105
	└ Other	−0.237	0.151	−1.57	.117
	└ Asian	−0.227	0.155	−1.46	.144
	Smoking (ref: Never)
	└ Current	−0.979	0.688	−1.424	.155
	Diabetes	−0.001	0.164	−0.01	.995
	Hypertension	−0.075	0.1	−0.75	.455
	Education	−0.014	0.043	−0.32	.751
	PIR	−0.005	0.025	−0.21	.833
	BMI	0.085	0.004	20.1	<.001

Models are fully adjusted for age, race, education, PIR, BMI, smoking, alcohol, hypertension, and diabetes. Only select covariates are shown for brevity.

AGP = alpha-1-acid glycoprotein, BMI = body mass index, CRP = C-reactive protein, PIR = poverty-to-income ratio.

In the fully adjusted model for AGP, a higher K/Na ratio was significantly associated with lower serum AGP concentrations. Specifically, for each 1-unit increase in the K/Na ratio, there was an associated decrease of 0.068 mg/dL in AGP levels (β = −0.068, SE = 0.021, P = .001). In this model, BMI was a strong positive predictor of AGP (β = 0.015, P < .001), while PIR showed a significant inverse association (β = −0.015, P = .002). The full model explained 29.6% of the variance in AGP levels (Adjusted R2 = 0.296).

A similar significant inverse association was observed in the model for log-CRP. In the fully adjusted model, a higher K/Na ratio was associated with lower log-transformed CRP levels (β = −0.248, SE = 0.095, P = .009). BMI was again the most powerful predictor, showing a strong positive association with log-CRP (β = 0.085, P < .001). This model explained 37.2% of the variance in log-CRP (Adjusted R2 = 0.372). For all models, variance inflation factors were below 2.0, indicating no concerns regarding multicollinearity.

3.3. Non-linear associations explored with generalized additive models

To explore the shape of the dose-response relationship more deeply, generalized additive models (GAM) with penalized splines were fitted (Table 3 and Figure 2).

Table 3.

Generalized additive models (GAM) for the association between AGP, CRP, and K/Na ratio.

Model	Term	Estimate (SE)	edf	F	P-value
AGP	Intercept	0.60 (0.07)	–	–	<.001
	Age	0.0081 (0.0015)	–	–	<.001
	BMI	–0.0018 (0.0017)	–	–	.286
	s(K/Na Ratio)	–	1.49	4.8	.007
CRP	Intercept	0.66 (0.07)	–	–	<.001
	Age	0.0081 (0.0015)	–	–	<.001
	BMI	–0.0038 (0.0016)	–	–	.019
	s(K/Na Ratio)	–	1.33	0.45	.499

Models are adjusted for all covariates.

edf = effective degrees of freedom, s() = smooth term.

Figure 2.

Penalized spline plots illustrating the relationship of K/Na ratio with AGP and CRP.

The GAM analysis for AGP revealed a significant non-linear component in the relationship with the K/Na ratio. The spline term for the K/Na ratio was statistically significant (edf = 1.49, F = 4.80, P = .007). The visual representation of this relationship in Figure 2 shows a distinct curvilinear inverse association. The plot demonstrates a relatively steep decline in AGP levels as the K/Na ratio increases from its lowest values, with the curve gradually flattening out at higher K/Na ratios, suggesting a potential saturation or threshold effect.

In contrast, the GAM analysis for log-CRP did not support a non-linear relationship. The spline term for the K/Na ratio was not statistically significant (edf = 1.33, F = 0.45, P = .499). The corresponding plot in Figure 2 shows a relatively straight, downward-sloping line across the range of K/Na ratios, consistent with the linear association identified in the regression analysis.

4. Discussion

This study, conducted in a large nationally representative sample of US adults, provides robust evidence for an independent and significant inverse association between the dietary potassium-to-sodium (K/Na) ratio and systemic inflammation. After rigorous adjustment for a comprehensive range of potential confounders-including socioeconomic factors, lifestyle behaviors, and chronic disease status-a higher K/Na ratio was significantly associated with lower serum concentrations of 2 established inflammatory biomarkers, alpha-1-acid glycoprotein (AGP) and C-reactive protein (CRP). These findings support the hypothesis that a favorable dietary electrolyte balance contributes to a reduced inflammatory burden. Of particular note, this analysis revealed a differential pattern in these associations: the relationship between the K/Na ratio and AGP was non-linear and curvilinear, whereas the association with log-transformed CRP (log-CRP) followed a linear trend.

These findings significantly advance our understanding of the health effects of dietary electrolyte balance. While previous studies have linked a favorable K/Na ratio to reduced risk of critical clinical outcomes, including hypertension,^[16] cardiovascular disease,^[7] and chronic kidney disease,^[17] the present study provides compelling evidence for a key underlying biological mechanism: the modulation of systemic inflammation. This observation serves as a mechanistic bridge linking dietary intake to disease pathogenesis, suggesting that the well-established benefits of a diet high in potassium and low in sodium are mediated, at least in part, through attenuation of the low-grade chronic inflammatory state that drives these conditions. Our results are consistent with the broader concept that diet is a powerful modulator of immune function, and that dietary patterns characterized by a high K/Na ratio: typically rich in fruits and vegetables and low in processed foods: can foster an antiinflammatory internal environment.

The observed epidemiological association is strongly supported by biological plausibility. The influence of the dietary K/Na ratio on systemic inflammation is explainable via at least 2 distinct, yet potentially synergistic, mechanistic pathways: a direct cellular pathway involving the regulation of the NLRP3 inflammasome and a systemic, hormonal pathway mediated by the renin-angiotensin-aldosterone system (RAAS).

The NOD-, LRR- and pyrin domain-containing protein 3 (NLRP3) inflammasome is a pivotal multi-protein complex within the innate immune system. Activation of the NLRP3 inflammasome occurs in response to diverse pathogen-associated molecular patterns (PAMPs) and damage-associated molecular patterns (DAMPs), but these agonists converge upon potassium efflux and a decrease in intracellular potassium concentration ([K+]).^[18] This mechanism provides a direct molecular link between the dietary K/Na ratio and cellular inflammation. A chronically low K/Na ratio would create a physiological state that facilitates potassium efflux from cells, thereby lowering the threshold of the NLRP3 inflammasome activation. In such a state, even minor, subclinical stimuli could trigger a robust inflammatory response. Conversely, a diet high in potassium maintains higher intracellular [K+], acting as a continuous inhibitory signal that suppresses excessive NLRP3 inflammasome activation and subsequent production of IL-1β and other proinflammatory mediators, thereby reducing chronic inflammation.^[18,19]

The RAAS tightly regulates blood pressure, fluid volume, and electrolyte homeostasis in response to dietary sodium and potassium intake.^[20] Low sodium delivery to the macula densa stimulates renin release, while elevated serum potassium suppresses it, making renin the rate-limiting step of the cascade.^[21] Angiotensin II (Ang II), acting via AT₁ receptors, increases reactive oxygen species production, upregulates vascular and intercellular adhesion molecules (VCAM-1, ICAM-1), and promotes pro-inflammatory cytokine and chemokine synthesis.^[22] Aldosterone similarly induces inflammation and fibrosis.^[23] Chronic RAAS activation therefore creates a systemic pro-inflammatory milieu that directly stimulates hepatic synthesis of acute-phase reactants such as CRP.^[24]

A low dietary K/Na ratio leads to sustained RAAS activation, which in turn fosters a systemic pro-inflammatory state, contributing to the elevated levels of CRP and AGP observed in our study participants. Conversely, a high K/Na ratio would naturally suppress the RAAS, thereby reducing the systemic inflammatory drive and exerting a broad antiinflammatory effect. The discovery of a differential dose-response relationship: a non-linear, curvilinear association for AGP vs a linear 1 for log-CRP: is a subtle but highly significant finding. It implies that these 2 acute-phase proteins are not simply redundant markers of the same underlying inflammatory process but may be responding to the dietary electrolyte balance through partially distinct mechanisms or may reflect different facets of the inflammatory response.

One hypothesis to explain this difference lies in the functional complexity of AGP. While CRP functions primarily as a sensitive, dose-dependent marker of the acute-phase response,^[25] AGP has been shown to be an active immunomodulator with context-dependent antiinflammatory and pro-inflammatory properties.^[26] The curvilinear relationship observed might reflect a biological threshold effect. At very low K/Na ratios, the pro-inflammatory stimulus (e.g., from RAAS activation and NLRP3 priming) is strong, leading to a steep increase in AGP levels.^[27] However, as the K/Na ratio improves beyond a certain point, the primary antiinflammatory effect may be sufficient to substantially quell the initial inflammatory triggers, leading to a plateau in the AGP response where further increases in the ratio yield diminishing returns. An alternative hypothesis is that CRP synthesis is governed by a linear RAAS-driven pathway. In contrast, AGP synthesis or its posttranslational modification (glycosylation), which affects its function and clearance, might be more sensitive to the more threshold-dependent, “all-or-none” activation characteristics of cellular pathways like the NLRP3 inflammasome.^[28,29] This differential sensitivity to the proposed dual pathways could manifest as the distinct linear and non-linear associations observed in our data.

This study possesses several notable strengths. First, the use of NHANES data enabled analysis in a large, well-characterized, and nationally representative sample, enhancing the generalizability of our findings to the US adult population. Second, simultaneous measurement of 2 distinct biomarkers of inflammation, CRP and AGP, provided a more comprehensive picture of systemic inflammation than prior studies focusing on a single marker. Third, the statistical models included rigorous adjustment for a broad range of demographic, socioeconomic, lifestyle, and clinical confounders. Lastly, the use of generalized additive models permitted sophisticated assessment of non-linear dose-response relationships, an advance over traditional regression techniques.

Nevertheless, certain limitations must be acknowledged. First, due to the cross-sectional study design, causality cannot be inferred, and the potential for residual confounding remains despite extensive adjustment. Second, dietary intake was estimated using two 24-hour dietary recalls on nonconsecutive days. While this approach follows the standard NHANES 2021–2023 protocol, recent evidence suggests that dietary assessment using three or more days of recalls, particularly when including at least 1 weekend day, may provide more reliable estimates of usual nutrient intake and reduce within-person variability.^[30,31] The use of only two 24-hour recalls introduces measurement error and may not fully capture the extent of day-to-day dietary variability. Additionally, 24-hour dietary recall methods are subject to recall bias and social desirability bias, which are inherent limitations of self-reported dietary assessment tools, regardless of the number of recall days. Participants may underreport socially undesirable foods (e.g., high-sodium processed foods) or overreport healthier options (e.g., fruits and vegetables high in potassium), potentially leading to misclassification of the K/Na ratio. This could potentially result in non-differential measurement error that would tend to attenuate the true strength of the associations observed in this study. Nevertheless, the use of mean values from 2 nonconsecutive recalls represents standard NHANES methodology and provides a nationally representative snapshot of dietary patterns. Future prospective studies using multiple-day dietary assessment methods may provide more precise characterization of the relationship between the K/Na ratio and inflammatory markers. Third, inflammatory biomarkers were measured at a single time point, potentially limiting the ability to capture long-term inflammatory status. Furthermore, data collection overlapped with the COVID-19 pandemic, which may have influenced both dietary behaviors and systemic inflammation in ways not fully captured by our analyses. Fourth, the generalizability of findings warrants careful consideration. While NHANES data are nationally representative of the US non-institutionalized civilian population, the observed associations may not be directly applicable to populations with substantially different dietary sodium environments or baseline K/Na ratios. For instance, East Asian populations typically consume significantly higher amounts of sodium (primarily from fermented foods, soy sauce, and pickled vegetables) compared to Western populations, resulting in markedly different baseline K/Na ratios. Similarly, Mediterranean populations following traditional dietary patterns may exhibit different electrolyte intake profiles characterized by high potassium intake from vegetables, fruits, and legumes combined with moderate sodium intake. The dose-response relationship between K/Na ratio and inflammatory markers may differ in these populations due to genetic variation in sodium sensitivity, differences in gut microbiome composition influenced by traditional dietary patterns, and varying background rates of chronic inflammatory diseases. Furthermore, the observed associations may not be directly extrapolable to clinical cohorts with active inflammatory diseases (e.g., rheumatoid arthritis, inflammatory bowel disease, systemic lupus erythematosus) where baseline inflammatory marker levels are substantially elevated and may be less responsive to dietary modulation. Future research in diverse populations with different dietary sodium environments and in clinical cohorts with chronic inflammatory conditions is needed to determine whether the inverse association between K/Na ratio and systemic inflammation is universal or population-specific. Finally, the generalizability of findings may be limited by the demographic and temporal context of the NHANES 2021–2023 cycle. Future prospective cohort studies are needed to clarify temporal dynamics and assess the persistence of observed associations. Ultimately, randomized controlled trials investigating the impact of dietary interventions targeting the K/Na ratio on inflammatory biomarkers would provide more robust evidence for causality.

5. Conclusions

This study demonstrates a significant and independent inverse association between the dietary potassium-to-sodium ratio and markers of systemic inflammation in a nationally representative sample of US adults. A higher dietary K/Na ratio: reflecting increased dietary potassium and decreased sodium intake: is associated with a substantially attenuated systemic inflammatory response. These findings elucidate a critical biological mechanism that may help explain the protective effect of optimal electrolyte intake against cardiovascular and other chronic diseases. Promotion of dietary patterns optimizing the K/Na ratio represents a practical and impactful public health strategy for reducing the burden of chronic inflammatory diseases.

Acknowledgments

We thank the National Center for Health Statistics and the participants of NHANES 2021–2023 for providing high-quality public data.

Author contributions

Conceptualization: Yeo Ju Sohn, Heewon Bae.

Methodology: Yeo Ju Sohn, Heewon Bae.

Project administration: Yeo Ju Sohn, Heewon Bae.

Resources: Yeo Ju Sohn.

Supervision: Yeo Ju Sohn, Heewon Bae.

Writing – original draft: Yeo Ju Sohn.

Writing – review & editing: Yeo Ju Sohn, Heewon Bae.

Data curation: Heewon Bae.

Formal analysis: Heewon Bae.

Funding acquisition: Heewon Bae.

Investigation: Heewon Bae.

Software: Heewon Bae.

Validation: Heewon Bae.

Visualization: Heewon Bae.