Longitudinal trajectories of serum uric acid and incident cardiovascular–kidney–metabolic syndrome components: Outcome-specific mediating role of adiposity in a large community-based cohort

Abstract

Objective

Serum uric acid links cardiovascular, kidney, and metabolic health; however, its role within the cardiovascular–kidney–metabolic syndrome framework is poorly understood, especially its long-term dynamics. We aimed to identify serum uric acid trajectories and investigate their outcome-specific associations with incident cardiovascular–kidney–metabolic components to clarify the mediating role of adiposity.

Methods

We analyzed data from 6052 participants (age ≥40 years) initially free from cardiovascular–kidney–metabolic components from a community-based cohort. Latent class growth analysis identified serum uric acid trajectories over a median 8-year follow-up. Generalized linear mixed models and causal mediation analysis were employed to assess associations with incident hypertension, hyperglycemia, and hypercholesterolemia as well as the role of body mass index.

Results

We identified “low-stable” (64.8%) and “high-increasing” (35.2%) serum uric acid trajectories. Compared with the “low-stable” group, the “high-increasing” group had elevated risks of incident hypertension (adjusted odds ratio: 1.26, 95% confidence interval: 1.15–1.37), hyperglycemia (adjusted odds ratio: 1.53, 95% confidence interval: 1.22–1.93), and hypercholesterolemia (adjusted odds ratio: 1.31, 95% confidence interval: 1.12–1.52). Mediation analysis revealed distinct pathways; the effect on hyperglycemia was predominantly mediated by body mass index (60.5%, p < 0.001), with the direct effect rendered non-significant. In contrast, body mass index was a partial mediator for hypertension (37.5%) but a suppressor for hypercholesterolemia (−18.1%), masking a stronger direct association.

Conclusions

Long-term serum uric acid trajectories are powerful predictors of incident cardiovascular–kidney–metabolic components; however, their impact is channeled through distinct, adiposity-dependent and adiposity-independent pathways. Monitoring serum uric acid dynamics, not just static levels, is crucial for early, stratified risk assessment of cardiovascular–kidney–metabolic syndrome.

Keywords

Serum uric acid trajectory analysis hypertension hypercholesterolemia hyperglycemia mediation analysis cohort study

Introduction

Cardiovascular–kidney–metabolic (CKM) syndrome, a systemic disorder characterized by the vicious cycle of crosstalk between heart, kidney, and metabolic dysfunction, represents a paramount global health challenge.^1–3 Within this complex network, serum uric acid (SUA), the final product of purine metabolism, is emerging as a critical nexus molecule rather than merely a bystander.^4–6 The role of SUA in CKM syndrome is uniquely complex due to its dual identity; it is both a consequence and a cause of renal dysfunction, thereby creating a potent self-amplifying loop of injury.

Impaired kidney function results in reduced SUA excretion, positioning elevated SUA as a sensitive, albeit nonspecific, marker of early renal decline. In contrast, a large body of evidence now implicates SUA as a direct pathogenic driver of renal damage. By inducing crystal-dependent and crystal-independent inflammation, afferent arteriolopathy, and endothelial dysfunction within the kidneys, SUA can initiate and accelerate nephropathy.^7–9 This SUA-driven renal injury, in turn, amplifies systemic risk by promoting hypertension and metabolic derangements, thereby fueling the progression of CKM syndrome. This dual role makes SUA a compelling but challenging therapeutic and prognostic target.

However, significant gaps and inconsistencies persist in our understanding. For instance, the association between SUA and blood pressure components varies across studies,¹⁰ and its relationship with hyperglycemia appears to be strongly modified by sex.^11,12 Most importantly, the vast majority of previous studies have relied on static, single-point SUA measurements at baseline.¹³ This approach fails to capture the dynamic nature of SUA levels over an individual’s life course, potentially overlooking the cumulative risk conferred by sustained or progressively increasing SUA concentrations.¹⁴

Recognizing this limitation, recent studies have explored the prognostic value of longitudinal SUA changes. However, simply observing change over time is not sufficient. A critical gap remains in understanding the deeper prognostic meaning of different longitudinal patterns identified through data-driven methods. For instance, it is unclear whether a “high-increasing” trajectory represents a more potent marker of persistent metabolic derangement or underlying genetic susceptibility than a single high measurement, which could be transient and less informative. This distinction is pivotal, as different trajectories may signify distinct pathophysiological processes requiring different clinical approaches.

Furthermore, it is unknown whether these specific SUA trajectories confer risk through a uniform mechanism or if their impact is modified by other key factors, such as adiposity, in an outcome-specific manner. A crucial, unanswered question is whether the mediating role of body mass index (BMI) is consistent across the core CKM components, such as hypertension, hyperglycemia, and hypercholesterolemia, or if SUA dynamics exert their influence through divergent biological pathways. Therefore, using a large, longitudinal community-based cohort, we aimed to fulfill the following objectives: (a) identify distinct SUA trajectories over a multi-year follow-up; (b) determine their association with incident hypertension, hyperglycemia, and hypercholesterolemia; and (c) elucidate and quantify the mechanistic role of BMI in these pathways. We hypothesized that a high-risk SUA trajectory would not be a monolithic predictor, but rather that its impact and underlying mechanisms would diverge significantly across different metabolic outcomes.

Methods

Study design and data source

We conducted a retrospective longitudinal cohort study by performing a secondary analysis of a deidentified dataset publicly available from the Dryad Digital Repository (DOI: 10.5061/dryad.z08kprrk1). The source data originate from the Chinese National “Public Basic Health Services Project,” an ongoing community-based program conducting biennial (every 2 years), standardized health examinations for adults in Hangzhou, China. The rich longitudinal structure of this dataset, with multiple serial measurements, provided a unique opportunity to apply advanced trajectory modeling techniques. This study adheres to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.¹⁵

Study population

The source dataset included the data from 12,498 consecutively recruited adults (aged ≥40 years) examined between May 2010 and December 2018. To establish a disease-free inception cohort for prospective analysis, we applied the following pre-specified exclusion criteria: (a) fewer than three completed health examinations (n = 2618); (b) missing data on SUA levels, blood pressure, fasting blood glucose (FBG) levels, or total cholesterol levels (n = 636); and (c) prevalent hypertension, hyperglycemia, or hypercholesterolemia at the baseline visit (n = 3125). To mitigate immortal-time bias, participants who developed an outcome at the first follow-up visit (n = 67) were also excluded. This resulted in a final analytical cohort of 6052 participants who were followed up for up to 8 visits (Figure S1). Notably, because the original dataset did not include medication inventory records, we could not ascertain the use of drugs that might influence SUA levels (e.g. diuretics and urate-lowering agents) or cardiovascular risk. However, participants with self-reported use of antihypertensive, antidiabetic, or lipid-lowering medications were excluded from the disease-free inception cohort. This study involved secondary analysis of deidentified data from a publicly available repository (Dryad Digital Repository, DOI: 10.5061/dryad.z08kprrk1). The original data collection process was approved by the relevant institutional review board as part of the Chinese National “Public Basic Health Services Project.” As our analysis used only anonymized, publicly accessible data with no possibility of identifying individual participants, additional institutional review board approval was not required according to the ethical guidelines for secondary data analysis. Written informed consent was obtained from all participants in the original study. The Institutional Review Board of the Affiliated Hospital of Liaoning University of Traditional Chinese Medicine confirmed that the secondary analysis of this deidentified public dataset was exempt from ethical approval. This approach is consistent with international standards for the use of deidentified public datasets in health research. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2024).

Measurements and definitions

Standardized protocols were used for all measurements. Data on demographic characteristics, medical history, and anthropometric parameters were collected by trained personnel. BMI was calculated as weight (kg)/height (m²). Blood pressure was measured twice using a mercury sphygmomanometer after a 5-min rest, and the average was calculated. After an overnight fast (≥8 h), venous blood was assayed for SUA, FBG, total cholesterol, and serum creatinine. Estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease (CKD) Epidemiology Collaboration (CKD–EPI) equation.

The primary outcomes were incident cases of the following, among participants without the respective condition at baseline:

Hypertension. Systolic blood pressure (BP) ≥140 mmHg, diastolic BP ≥90 mmHg, or self-reported use of antihypertensive medication;

Hyperglycemia. FBG level ≥7.0 mmol/L or a physician diagnosis of diabetes;

Hypercholesterolemia. Total cholesterol level ≥6.22 mmol/L.

Statistical analyses

All analyses were performed using R (Version 4.3.1). A two-sided p-value <0.05 was considered statistically significant.

Identification of SUA trajectories

We employed latent class growth analysis (LCGA), a data-driven approach superior to static, single-point measurements, to identify distinct longitudinal SUA patterns. Using the ‘lcmm’ package, we modeled SUA as a function of age (as the time scale), which aligns individuals by biological time and accounts for age-related changes more effectively than study time. We first tested quadratic models; however, these produced convergence failures for ≥3 classes, indicating over-parameterization. We therefore adopted linear specifications. The optimal model was selected based on the lowest Bayesian information criterion (BIC), average posterior probability >0.70 for each class, and a class size ≥5% of the total cohort. A two-class linear model provided the most parsimonious and interpretable fit, yielding “low-stable” and “high-increasing” trajectories (Table S1).

Primary analysis: association and mediation

Our primary analysis followed a pre-specified, two-step causal framework. First, we estimated the total effect of the SUA trajectory on each incident outcome (hypertension, hyperglycemia, and hypercholesterolemia) using generalized linear mixed models (GLMMs) with a random intercept for each participant. These models were adjusted for a set of a priori baseline confounders (Table 1) but deliberately excluded the potential mediator, baseline BMI. The “low-stable” group served as the reference.

Table 1.

Odds ratios for incident cardiometabolic diseases according to serum uric acid trajectory groups.

Outcome	aOR (95% CI)	p-value
Incident hypertension	1.26 (1.15–1.37)	0.001
Incident hyperglycemia	1.53 (1.22–1.93)	<0.001
Incident hypercholesterolemia	1.31 (1.12–1.52)	<0.001

The “low-stable” trajectory group served as the reference category. Models represent the total effect, adjusted for baseline confounders and excluded the potential mediator (BMI). Covariates are as follows: baseline age, sex, fasting blood glucose (FBG), and estimated glomerular filtration rate (eGFR) for hypertension; baseline age, sex, systolic blood pressure (SBP), and eGFR for hyperglycemia; and baseline age, sex, SBP, FBG, and eGFR for hypercholesterolemia.

aOR: adjusted odds ratio; CI: confidence interval.

Second, to elucidate the mechanistic role of adiposity, we decomposed this total effect using causal mediation analysis (‘mediation’ package). This quantified the proportion of the total effect mediated through baseline BMI (average causal mediation effect (ACME)) and the proportion that remained independent of this pathway (average direct effect (ADE)). We derived 95% confidence intervals (CIs) from 1000 non-parametric bootstrap iterations.

Secondary and sensitivity analyses

We conducted several additional analyses to explore heterogeneity and assess the robustness of our findings. To test for potential effect modification, we introduced multiplicative interaction terms between the SUA trajectory and baseline sex, age (<65 vs. ≥65 years), and BMI (<24 vs. ≥24 kg/m²) into the main GLMMs.

To address the critical issue of causal directionality, we performed a key sensitivity analysis using reverse mediation, which inverted the model to test the plausibility of the alternative pathway where baseline BMI was the exposure and the SUA trajectory was the mediator. Furthermore, to mitigate against potential protopathic bias, we conducted another sensitivity analysis by repeating our main models after excluding participants who developed an outcome within the first 2 years of follow-up.

To further assess the robustness of our findings against alternative SUA definitions, we repeated the total effect models using baseline categorical hyperuricemia defined by the following two established cutoffs: (a) the classic threshold (male: >7.0 mg/dL (approximately 420 µmol/L); female: >6.0 mg/dL (approximately 360 µmol/L)) and (b) the lower uric acid right for heart health (URRAH) threshold (≥5.6 mg/dL (approximately 334 µmol/L) for both sexes), which has been associated with increased cardiometabolic risk.¹⁶ Associations with incident hypertension, hyperglycemia, and hypercholesterolemia were estimated using the same GLMMs and covariate sets as in the primary total effect analysis.

Finally, cumulative incidence curves for each outcome were constructed using discrete-time life-table methods for visualization, with 95% CIs calculated using Greenwood’s formula.

Results

Participant characteristics and SUA trajectories

In total, 6052 participants without cardiometabolic disease at baseline were included in the analysis. Using LCGA, we identified two distinct longitudinal trajectories of SUA (Figure 1), including a “low-stable” group (n = 3921; 64.8%) and a “high-increasing” group (n = 2131; 35.2%). Over the median 8-year follow-up, the mean (SD) observed SUA increased from 376.4 (67.9) µmol/L to 390.2 (72.1) µmol/L in the “high-increasing” group, whereas it remained essentially unchanged in the “low-stable” group (265.8 (54.8) µmol/L at baseline vs. 261.5 (56.3) µmol/L at the final visit).

Figure 1.

Latent class trajectories of serum uric acid (SUA). Two distinct SUA trajectories were identified using latent class growth analysis, including a “low-stable” group (n = 3921; 64.8%) and a “high-increasing” group (n = 2131; 35.2%).

As detailed in Table 2, baseline characteristics differed significantly between the two groups. Compared with those in the “low-stable” group, participants in the “high-increasing” group were older, predominantly male, and had a higher BMI (all p < 0.001). In addition, the “high-increasing” group exhibited a more adverse cardiometabolic profile, characterized by significantly higher systolic and diastolic blood pressure, FBG level, and lower eGFR (all p < 0.001).

Table 2.

Baseline characteristics of study participants, stratified by serum uric acid trajectory group.

Characteristics	Overall (n = 6052)	High-increasing (n = 2131)	Low-stable (n = 3921)	p-value
Demographic characteristics
Age (years)	65.75 (9.41)	66.57 (9.13)	65.31 (9.53)	<0.001
Female, n (%)	4040 (66.8)	941 (44.2)	3099 (79.0)	<0.001
Anthropometric characteristics
Body mass index (kg/m²)	23.23 (3.06)	24.22 (3.04)	22.70 (2.93)	<0.001
Cardiometabolic profile
Systolic blood pressure (mmHg)	123.81 (10.50)	125.46 (9.62)	122.92 (10.85)	<0.001
Diastolic blood pressure (mmHg)	75.29 (7.14)	76.72 (6.74)	74.52 (7.24)	<0.001
Fasting blood glucose (mmol/L)	5.32 (0.60)	5.41 (0.61)	5.27 (0.58)	<0.001
Total cholesterol (mmol/L)	4.72 (0.75)	4.73 (0.76)	4.71 (0.74)	0.461
Renal function markers
Serum creatinine (μ/mol/L)	74.88 (15.64)	74.87 (15.69)	74.88 (15.62)	0.993
eGFR (mL/min/1.73 m²)	87.54 (26.94)	84.28 (25.72)	89.32 (27.42)	<0.001
Exposure marker
Serum uric acid	304.71 (79.78)	376.38 (67.93)	265.76 (54.84)	<0.001
Incident events during follow-up
Hypertension, n (%)	2958 (48.9)	1151 (54.0)	1807 (46.5)	<0.001
Hyperglycemia, n (%)	824 (13.6)	362 (17.0)	462 (11.9)	<0.001
Hypercholesterolemia, n (%)	1421 (23.5)	543 (25.5)	878 (22.8)	0.018

Data are presented as mean (SD) for continuous variables and n (%) for categorical variables.

p-values were calculated using unpaired Student’s t-test for continuous variables and the χ² test for categorical variables.

eGFR: estimated glomerular filtration rate.

SUA trajectories and total effect on incident cardiometabolic diseases

We first estimated the total effect of SUA trajectories on the incidence of each cardiometabolic disease by adjusting for baseline confounders and excluding the potential mediator, BMI (Table 1). In these total effect models, compared with the “low-stable” trajectory group, the “high-increasing” group showed a significantly elevated risk for incident hypertension (adjusted odds ration (aOR) = 1.26, 95% CI: 1.15–1.37), hyperglycemia (aOR = 1.53, 95% CI: 1.22–1.93), and hypercholesterolemia (aOR = 1.31, 95% CI: 1.12–1.52).

The cumulative incidence curves visually support these findings (Figure 2). A clear and persistent separation was observed for both hypertension and hyperglycemia, with the “high-increasing” group showing a consistently higher risk over the entire follow-up period. By the final visit, the cumulative incidence of hypertension was 54.0% in the “high-increasing” group versus 46.5% in the “low-stable” group; for hyperglycemia, the corresponding incidences were 17.0% versus 11.9%. The curves for hypercholesterolemia, however, were less distinct, with final incidences of 25.5% and 22.8% in the “high-increasing” and “low-stable” groups, respectively.

Figure 2.

Cumulative incidence of cardiometabolic diseases stratified by SUA trajectory. Cumulative incidence curves for (a) hypertension, (b) hyperglycemia, and (c) hypercholesterolemia were generated using discrete-time life-table methods. Shaded areas represent 95% confidence intervals. SUA: serum uric acid.

Mechanistic pathways: Mediation by adiposity

The mediation analysis was conducted on a risk difference scale to estimate the absolute contribution of the direct and indirect pathways (Table 3). For incident hypertension, BMI acted as a significant partial mediator, explaining 37.5% of the total effect (ACME, p < 0.001). A significant direct effect remained after accounting for the pathway through BMI (ADE, p < 0.001), indicating that the “high-increasing” trajectory influences hypertension risk through both BMI-dependent and BMI-independent pathways.

Table 3.

Mediation analysis of BMI in the association between uric acid trajectory and incident cardiometabolic diseases.

Outcome	Estimate (95% CI)	P-value	Proportion mediated (%)
Hypertension
Total effect	0.080 (0.052, 0.108)	<0.001	100%
Direct effect (ADE)	0.050 (0.022, 0.077)	<0.001	62.5%
Indirect effect (ACME)	0.030 (0.023, 0.038)	<0.001	37.5%
Hyperglycemia
Total effect	0.033 (0.016, 0.051)	<0.001	100%
Direct effect (ADE)	0.013 (−0.004, 0.030)	0.152	39.5%
Indirect effect (ACME)	0.020 (0.016, 0.024)	<0.001	60.5%
Hypercholesterolemia
Total effect	0.038 (0.015, 0.060)	0.004	100%
Direct effect (ADE)	0.045 (0.021, 0.069)	<0.001	118.1%
Indirect effect (ACME)	−0.007 (−0.012, −0.003)	0.004	−18.1%

All effects are presented on a risk difference scale, representing the change in the probability of developing the outcome when comparing the “high-increasing” and “low-stable” groups. These estimates are derived from the same underlying models used for the total effect in Table 1, but are reported on a different scale.

Models were adjusted for baseline age, sex, and the same disease-specific covariates as in Table 1’s Model 2.

Proportion mediated was calculated as ACME/total effect. For hypercholesterolemia, the negative proportion indicates a suppressor effect. The proportion for direct effect (ADE/total effect) has also been added for completeness.

ADE: average direct effect; ACME: average causal mediation effect; CI: confidence interval.

For incident hyperglycemia, the mediating role of BMI was even more pronounced, accounting for 60.5% of the total association (ACME, p < 0.001). Notably, the direct effect of the SUA trajectory on hyperglycemia became non-significant (ADE, p = 0.152) after accounting for the BMI pathway. This suggests that the total effect of SUA trajectories on hyperglycemia risk is predominantly, if not entirely, driven by adiposity.

A contrasting mechanism was observed for hypercholesterolemia. Here, we identified a significant negative indirect effect, indicating that BMI acts as a suppressor (proportion mediated, −18.1%; ACME, p = 0.004). This implies that baseline BMI was masking an even stronger direct relationship between the “high-increasing” trajectory and the risk of hypercholesterolemia.

Reverse mediation analysis to test causal direction

In our key sensitivity analysis testing the reverse causal pathway (BMI → SUA Trajectory → Outcome), we found no evidence of significant mediation by the SUA trajectory for any of the three outcomes (Table S2). For hypertension, hyperglycemia, and hypercholesterolemia, the indirect effect (ACME) mediated through the SUA trajectory was non-significant (p = 0.312, 0.394, and 0.252, respectively). Most of the total effect of BMI on these outcomes was attributable to the direct pathway (proportion mediated by SUA trajectory: 1.2%, 0.4%, and −2.0%, respectively). These results provide strong evidence against the alternative causal model and support the plausibility of our primary hypothesized pathway where SUA trajectories influence cardiometabolic risk, with BMI acting as a downstream mediator.

Heterogeneity and robustness of findings

Subgroup analyses revealed that the associations were modified by age and sex. The effects on hypertension and hypercholesterolemia were stronger in participants aged <65 years (p for interaction = 0.050 and 0.006, respectively). The risk of hyperglycemia associated with the “high-increasing” trajectory was significant only in female participants (p for interaction < 0.001) (Table S3).

In a key sensitivity analysis that excluded participants who developed an outcome within the first 2 years of follow-up, the primary findings remained robust, reinforcing the stability of our results (Table S4).

Sensitivity analysis using categorical hyperuricemia definitions

At baseline, the prevalence of hyperuricemia was 10.2% (n = 617) according to the classic cutoff and 28.7% (n = 1737) according to the URRAH cutoff. In fully adjusted models (excluding BMI), classic hyperuricemia was significantly associated with increased risks of incident hypertension (aOR = 1.24, 95% CI: 1.12–1.38) and hypercholesterolemia (aOR = 1.23, 95% CI: 1.01–1.47); however, the association with hyperglycemia was not statistically significant (aOR = 1.10, 95% CI: 0.83–1.43). In contrast, URRAH-defined hyperuricemia was significantly associated with all three outcomes: (a) hypertension (aOR = 1.21, 95% CI: 1.11–1.32); (b) hyperglycemia (aOR = 1.36, 95% CI: 1.09–1.73); and (c) hypercholesterolemia (aOR = 1.29, 95% CI: 1.12–1.46) (Supplementary Table S5). These findings align with our trajectory-based results, demonstrating that even modestly elevated SUA levels confer incremental risk for CKM components and highlight the potential utility of lower diagnostic thresholds for early risk identification.

Discussion

This study provides new evidence for understanding the early pathophysiological mechanisms of CKM syndrome by revealing outcome-specific associations between SUA longitudinal trajectories and the core components of CKM. In this large-scale longitudinal analysis, we found that a “high-increasing” SUA trajectory was associated with an elevated total risk of incident hypertension and hypercholesterolemia over an 8-year follow-up. Causal mediation analysis revealed that these associations are driven by distinct, outcome-specific mechanisms. The total effect on hyperglycemia appears to be predominantly mediated by adiposity, whereas the relationship with hypercholesterolemia was characterized by a strong direct effect partially masked by a suppressor effect exerted by BMI. These findings move beyond static biomarker assessment, highlighting the complex interplay between long-term SUA dynamics, adiposity, and cardiometabolic risk.

A primary challenge in interpreting the relationship between SUA, adiposity, and metabolic disease is the uncertain causal direction. Although it is well-established that higher BMI can lead to elevated SUA levels, our study was designed to test the prognostic value of long-term SUA dynamics. A major strength of our analysis is the formal testing of this directionality through reverse causal mediation. Our sensitivity analysis revealed that the pathway from baseline BMI to cardiometabolic outcomes was not significantly mediated by an individual’s subsequent SUA trajectory. This finding, although not precluding the influence of BMI on SUA levels, strongly suggests that the long-term patterns of SUA we identified are not merely passive markers for adiposity but represent a distinct pathophysiological process with its own prognostic implications for cardiometabolic health. This provides a firmer methodologic foundation for our primary finding that BMI acts as a key mediator in the path from SUA trajectories to disease.

The pathophysiological mechanisms linking SUA trajectories to CKM components are multifaceted, with the kidneys acting as a central battleground. Within the renal microenvironment, SUA exerts direct toxic effects through several pathways. Soluble SUA enters vascular smooth muscle cells, inducing oxidative stress and proliferative responses that lead to afferent arteriolopathy and glomerulosclerosis, contributing to both hypertension and a decline in renal function.¹⁷ Furthermore, SUA can precipitate into microcrystals within the renal tubules, triggering nucleotide-binding domain, leucine-rich repeat family, pyrin domain-containing protein 3 (NLRP3) inflammasome activation and subsequent tubulointerstitial inflammation and fibrosis.¹⁸ Critically, SUA also activates the intrarenal renin–angiotensin–aldosterone system (RAAS), a key pathway that not only exacerbates local renal injury but also drives systemic hypertension, thereby forging a direct link between renal-specific damage and a core cardiovascular component of CKM.^19,20

Beyond the kidney, this sustained, high-risk SUA profile promotes systemic endothelial dysfunction by reducing nitric oxide bioavailability, further contributing to elevated blood pressure and vascular damage.²¹ In metabolic tissues, such as adipocytes, SUA impairs insulin signaling and promotes a proinflammatory state, fostering the adiposity-driven metabolic disturbances we observed.^22,23 Our trajectory-based approach captured the cumulative burden of these interconnected pathological processes over time, providing a more robust prognostic signal than a single static measurement. This pattern of a modest but sustained rise in SUA levels among a substantial subgroup aligns with recent longitudinal evidence from the Pressioni Arteriose Monitorate E Loro Associazioni (PAMELA) study, which has reported a progressive increase in population SUA levels over 25 years and identified baseline SUA and metabolic factors as key predictors of incident hyperuricemia.²⁴ Our trajectory-based approach extends beyond previous cross-sectional studies^25,26 because it captures the cumulative burden of SUA exposure over time. Consistent with a Korean cohort study that has demonstrated that both baseline and increasing uric acid levels independently predict hypertension,²⁷ we found that individuals with a “high-increasing” trajectory exhibited a 15% increased hypertension risk after comprehensive adjustment. Similarly, a Chinese study has reported stronger associations between SUA and arterial stiffness in younger individuals without antihypertensive treatment, aligning with our subgroup findings.²⁸ However, our work uniquely demonstrated that these associations vary fundamentally by outcome, suggesting that SUA operates through distinct pathophysiological pathways rather than a unified mechanism.

Our sensitivity analysis using categorical hyperuricemia definitions further corroborates these observations. Notably, the URRAH cutoff, which is lower than the classic threshold, captured a larger at-risk subgroup and yielded significant associations across all three CKM components, including hyperglycemia where the classic cutoff was not predictive. This is consistent with recent evidence that the URRAH threshold better identifies individuals at risk of metabolic syndrome components and underscores the clinical importance of recognizing even mild, sustained elevations in SUA, a concept inherently captured by our trajectory approach.¹⁶

The complex interplay between SUA, adiposity, and glucose metabolism has been extensively debated.²⁹ Although Chinese data have shown that the hyperuricemia prevalence reaches 69.8% in individuals with obesity, the directionality and independence of these relationships remain unclear.³⁰ Our mediation analysis provides critical mechanistic clarity; SUA-incident hyperglycemia, one of the most central metabolic disturbances within CKM syndrome, was entirely mediated by BMI (60.5%), suggesting that elevated SUA levels are a biomarker for adiposity-related metabolic dysfunction rather than an independent diabetogenic factor. In other words, within the CKM framework, lowering SUA levels alone may provide limited benefit for preventing CKM phenotypes characterized by hyperglycemia unless weight and fat distribution are concurrently addressed. This mechanistic interpretation also helps explain previous inconsistencies in epidemiologic findings regarding SUA and diabetes risk.

The most striking finding of our study is the suppressor effect of BMI on the SUA–hypercholesterolemia relationship, a phenomenon whereby adjusting for adiposity paradoxically strengthened rather than attenuated the association. This seemingly paradoxical phenomenon suggests that within the lipid-metabolic pathway of CKM syndrome, SUA and adiposity may act through partially independent or even oppositely directed mechanisms.³¹ Although adiposity promotes hypercholesterolemia through insulin resistance and altered hepatic lipid production, our data indicate that SUA may independently impair lipid metabolism through alternative mechanisms, possibly involving xanthine oxidase-mediated oxidative stress or via direct effects on hepatic lipid synthesis.^32,33 Recent evidence linking adipokines such as angiopoietin-like protein 2 (ANGPTL2) to both hyperuricemia and lipid dysregulation³⁴ further supports the existence of a complex and potentially nonlinear “uric acid–adiposity–lipid” regulatory network. Taken together, these findings suggest that in CKM-related lipid dysregulation, SUA serves as a long-underestimated direct pathogenic factor rather than merely a byproduct of obesity.

Our stratified analyses also provide important insight for refined CKM syndrome risk assessment. Age-stratified analyses showed that the associations between SUA trajectories and CKM components were strongest among individuals aged ≤65 years, indicating that SUA dynamic patterns carry particular prognostic value for CKM risk stratification in midlife. This aligns with previous meta-analytic evidence demonstrating stronger SUA–hypertension associations in younger populations,³⁵ possibly reflecting greater vascular plasticity and more reversible endothelial dysfunction before substantial atherosclerotic burden develops.³⁶ In addition, we observed significant associations between SUA trajectories and incident hyperglycemia only among women. This sex-specific pattern is consistent with prior evidence that women may be more susceptible to SUA-related metabolic disturbances,^37,38 potentially due to estrogen-regulated differences in uric acid handling, fat distribution, and insulin sensitivity. Collectively, these results indicate that both age and sex modify the magnitude and direction of SUA’s effects on CKM components and underscore the need to integrate SUA longitudinal dynamics with demographic characteristics to achieve more individualized CKM risk prediction and targeted intervention strategies.

Limitations

Certain study limitations must be acknowledged. First, as an observational study, our findings indicate strong associations but cannot definitively establish causality. Although our reverse mediation analysis provides robust support for the hypothesized causal direction, the possibility of residual confounding from unmeasured factors remains. Specifically, our dataset lacked information on key variables such as medication use (e.g. diuretics and allopurinol), detailed dietary patterns (e.g. purine or fructose intake), and physical activity, which could have introduced bias. For instance, diuretic use can elevate SUA levels and is linked to hypertension, potentially confounding the true association. Second, our analysis was conducted in a specific community-dwelling Chinese population, which may limit the generalizability of our findings to other ethnicities. Third, the definition of hypercholesterolemia was based solely on total cholesterol, as a full lipid panel including low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), and triglycerides was unavailable; this precluded a more granular analysis of specific atherogenic lipid fractions such as LDL-C. Fourth, the biennial health examinations may not have captured more rapid, short-term fluctuations in SUA levels or other risk factors that could influence disease development. Finally, our study did not include direct renal endpoints such as incident CKD or albuminuria. However, this limitation paradoxically highlights a key strength of our study; by focusing on the incident components of CKM in a population initially free of overt disease, we captured the prognostic value of SUA dynamics in the very early stages of the syndrome before the manifestation of irreversible renal damage. Our findings thus provide a window into the preclinical phase of CKM syndrome, offering valuable clues for primordial and primary prevention.

Conclusion

This study moves beyond static measurements to reveal that longitudinal SUA trajectories may serve as informative predictors of incident CKM syndrome components; however, their prognostic impact is mediated through distinct, outcome-specific mechanistic pathways. Although the link between a high-risk SUA trajectory and hyperglycemia is largely dependent on adiposity, its relationships with hypertension and hypercholesterolemia involve significant BMI-independent pathways. These findings underscore the importance of viewing SUA not as a monolithic risk factor, but as a dynamic entity, the prognostic meaning of which is deeply intertwined with adiposity and the specific CKM component in question. Clinically, monitoring SUA trajectories rather than single values, may aid in early identification and stratified management of individuals at high-risk of CKM syndrome.

Supplemental Material

sj-pdf-1-imr-10.1177_03000605261452233 - Supplemental material for Longitudinal trajectories of serum uric acid and incident cardiovascular–kidney–metabolic syndrome components: Outcome-specific mediating role of adiposity in a large community-based cohort

Supplemental material, sj-pdf-1-imr-10.1177_03000605261452233 for Longitudinal trajectories of serum uric acid and incident cardiovascular–kidney–metabolic syndrome components: Outcome-specific mediating role of adiposity in a large community-based cohort by Xiaoqing Chen, Jianhua Gu, Ye Xu and Fang Yuan in Journal of International Medical Research

Footnotes

Acknowledgments

The authors thank the participants and staff of the Chinese National Public Basic Health Services Project for their contributions to the original data collection and the Dryad Digital Repository for making the dataset publicly available.

Authors’ contributions

Y.X and F.Y. conceptualized and designed the study. X.C. and J.G. performed the data analysis, interpreted the data, and drafted the initial manuscript. F.Y. provided supervision and project administration. All authors contributed to the critical revision of the manuscript for important intellectual content. All authors read and approved the final manuscript.

Availability of data and materials

The dataset analyzed during the current study is publicly available in the Dryad Digital Repository. It can be accessed directly via the Digital Object Identifier (DOI): 10.5061/dryad.z08kprrk1.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Funding

This work was supported by the Science and Technology Program of Shenyang City (Grant No. 22-321-34-04) and the Basic Scientific Research Project of the Education Department of Liaoning Province (Grant No. LJ212510162010).

Statement of ethics

Ethical approval was not required for this study as it only used data extracted from the Dryad Digital Repository.

Supplemental material

Supplemental material for this article is available online.

ORCID iD

Fang Yuan

References

Larkin

Here's what to know about cardiovascular-kidney-metabolic syndrome, newly defined by the AHA. JAMA 2023; 330: 2042–2043.

Marassi

Fadini

GP.

The cardio-renal-metabolic connection: a review of the evidence. Cardiovasc Diabetol 2023; 22: 195.

Claudel

Schmidt

Waikar

, et al. Cumulative incidence of mortality associated with cardiovascular-kidney-metabolic (CKM) syndrome. J Am Soc Nephrol 2025; 36: 1343–1351.

Kittelson

Junior

Fillmore

, et al. Cardiovascular-kidney-metabolic syndrome-an integrative review. Prog Cardiovasc Dis 2024; 87: 26–36.

Gigante

Assanto

Brigato

, et al. Clinical outcomes in patients with cardiorenal multimorbidity: the role of serum uric acid/serum creatinine ratio. High Blood Press Cardiovasc Prev 2025; 32: 209–216.

Maloberti

Mengozzi

Russo

; Working Group on Uric Acid and Cardiovascular Risk of the Italian Society of Hypertension (SIIA)et al. The results of the URRAH (uric acid right for heart health) project: a focus on hyperuricemia in relation to cardiovascular and kidney disease and its role in metabolic dysregulation. High Blood Press Cardiovasc Prev 2023; 30: 411–425.

Johnson

Lanaspa

Gaucher

EA.

Uric acid: a danger signal from the RNA world that may have a role in the epidemic of obesity, metabolic syndrome, and cardiorenal disease: evolutionary considerations. Semin Nephrol 2011; 31: 394–399.

Kimura

Tsukui

Kono

Uric acid in inflammation and the pathogenesis of atherosclerosis. Int J Mol Sci 2021; 22: 12394.

Gherghina

Peride

Tiglis

, et al. Uric acid and oxidative stress-relationship with cardiovascular, metabolic, and renal impairment. Int J Mol Sci 2022; 23: 3188.

10.

Zhang

Guo

, et al. The interaction between serum uric acid and triglycerides level on blood pressure in middle-aged and elderly individuals in China: result from a large national cohort study. BMC Cardiovasc Disord 2020; 20: 174.

11.

Kivity

Kopel

Steinlauf

, et al. The association between serum uric acid and diabetes mellitus is stronger in women. J Womens Health(Larchmt) 2013; 22: 782–789.

12.

Kawamoto

Tabara

Kohara

, et al. Serum uric acid is more strongly associated with impaired fasting glucose in women than in men from a community-dwelling population. PLOS One 2013; 8: e65886.

13.

Sundström

Sullivan

D'Agostino

, et al. Relations of serum uric acid to longitudinal blood pressure tracking and hypertension incidence. Hypertension 2005; 45: 28–33.

14.

Lin

Chen

, et al. Poor serum uric acid control increases risk for developing hypertension: a retrospective cohort study in China. Front Endocrinol (Lausanne) 2024; 15: 1343998.

15.

Von Elm

Altman

Egger

; STROBE Initiativeet al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147: 573–577.

16.

Maloberti

Tognola

Garofani

, et al. Uric acid and metabolic syndrome: importance of hyperuricemia cut-off. Int J Cardiol 2024; 417: 132527.

17.

Zong

, et al. Hyperuricemia and its related diseases: mechanisms and advances in therapy. Signal Transduct Target Ther 2024; 9: 212.

18.

Areda

Gizaw

Berdida

, et al. Evaluation of serum lipid profiles, uric acid, and high sensitivity C-reactive protein levels between pregnancy-induced hypertension and normotensive pregnant women attending Ambo University Referral Hospital, Ambo, Ethiopia, 2020: a case-control study. Health Sci Rep 2022; 5: e806.

19.

Sattui

Singh

Gaffo

AL.

Comorbidities in patients with crystal diseases and hyperuricemia. Rheum Dis Clin North Am 2014; 40: 251–278.

20.

Prasad Sah

Qing

YX.

Associations between hyperuricemia and chronic kidney disease: a review. Nephrourol Mon 2015; 7: e27233.

21.

Johnson

Bakris

Borghi

, et al. Hyperuricemia, acute and chronic kidney disease, hypertension, and cardiovascular disease: report of a scientific workshop organized by the National Kidney Foundation. Am J Kidney Dis 2018; 71: 851–865.

22.

Sautin

Nakagawa

Zharikov

, et al. Adverse effects of the classic antioxidant uric acid in adipocytes: NADPH oxidase-mediated oxidative/nitrosative stress. Am J Physiol Cell Physiol 2007; 293: C584–C596.

23.

Sun

, et al. Metformin alleviates hyperuricaemia-induced serum FFA elevation and insulin resistance by inhibiting adipocyte hypertrophy and reversing suppressed white adipose tissue beiging. Clin Sci (Lond) 2020; 134: 1537–1553.

24.

Maloberti

Dell'Oro

Bombelli

, et al. Long-term increase in serum uric acid and its predictors over a 25 year follow-up: results of the PAMELA study. Nutr Metab Cardiovasc Dis 2024; 34: 223–229.

25.

Zhao

, et al. Serum uric acid levels as a causal factor in hypertension: insights from Mendelian randomization analysis. Clin Exp Hypertens 2025; 47: 2496514.

26.

Kuwabara

Kodama

, et al. Update in uric acid, hypertension, and cardiovascular diseases. Hypertens Res 2023; 46: 1714–1726.

27.

Sung

Byrne

Ryu

, et al. Baseline and change in uric acid concentration over time are associated with incident hypertension in large Korean cohort. Am J Hypertens 2017; 30: 42–50.

28.

Wang

Liu

, et al. High serum uric acid is a risk factor for arterial stiffness in a Chinese hypertensive population: a cohort study. Hypertens Res 2024; 47: 1512–1522.

29.

Lima

Martins-Santos

Chaves

VE.

Uric acid as a modulator of glucose and lipid metabolism. Biochimie 2015; 116: 17–23.

30.

Cheang

Law

Ren

, et al. Prevalence of hyperuricemia in patients with severe obesity and the relationship between serum uric acid and severe obesity: a decade retrospective cross-section study in Chinese adults. Front Public Health 2022; 10: 986954.

31.

Hsieh

Chang

SJ.

Metabolic syndrome, diabetes, and hyperuricemia. Curr Opin Rheumatol 2013; 25: 210–216.

32.

Wei

Wang

Dalbeth

, et al. Weight loss after receiving anti-obesity medications and gout among individuals with overweight and obese: a population-based cohort study. Arthritis Rheumatol 2025; 77: 335–345.

33.

Maloberti

Vanoli

Finotto

, et al. Uric acid relationships with lipid profile and adiposity indices: impact of different hyperuricemic thresholds. J Clin Hypertens (Greenwich) 2023; 25: 78–85.

34.

Yang

Yin

Shan

, et al. Adipocyte-secreted ANGPTL2 promotes hyperuricemia through inhibiting AKT/ABCG2 signaling. Free Radic Biol Med 2025; 233: 209–225.

35.

Grayson

Kim

Lavalley

, et al. Hyperuricemia and incident hypertension: a systematic review and meta-analysis. Arthritis Care Res (Hoboken) 2011; 63: 102–110.

36.

Thews

Schmid

Kluttig

, et al. Physiological serum uric acid concentrations correlate with arterial stiffness in a sex-dependent manner. BMC Med 2025; 23: 356.

37.

Chiou

Wang

Huang

, et al. The relationship between serum uric acid level and metabolic syndrome: differences by sex and age in Taiwanese. J Epidemiol 2010; 20: 219–224.

38.

Wakabayashi

Kato

Tanaka

; Japan Obesity Metabolic Syndrome Study (JOMS) Groupet al. Novel pathological implications of serum uric acid with cardiovascular disease risk in obesity. Diabetes Res Clin Pract 2023; 205: 110919.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.35 MB

0.00 MB