Prevalence and mortality of cardiovascular-kidney-metabolic syndrome in US adults, 1999–2018

Abstract

Background

Cardiovascular-kidney-metabolic (CKM) syndrome is a novel staging construct that reflects the pathophysiological interplay between these individual conditions. The long-term trends in the prevalence of CKM syndrome in the US have not been well characterized.

Methods

This serial cross-sectional study used data from the National Health and Nutrition Examination Survey (1999–2018) to define CKM stages from 0 (no risk factors) to 4 (clinical CVD). Participants younger than 20 years, pregnant patients, and those with missing mortality data and NHANES fasting subsample weights were excluded. We calculated the age- and sex-adjusted prevalence of each stage and its trends. Survey-weighted Cox regression was used to estimate the mortality risk with each stage.

Results

The prevalence of CKM syndrome at each stage was as follows: stage 0, 11.9%; stage 1, 24.9%; stage 2, 49.1%; stage 3, 4.6%; and stage 4, 9.5%. The prevalence of CKM stages 0 and 4 decreased between 1999 and 2018, but that of CKM stages 1 and 3 increased. The risk of all-cause and CVD mortality showed a graded increase from stage 1 to 4 compared to stage 0 (HRs for all-cause death in stage 4 = 2.87, 95% CI: 1.98–4.17, and CVD death = 4.68, 95% CI: 1.95–11.26).

Conclusions

Nearly 90% of US adults have Stage 1 or higher CKM syndrome, and CKM stage is associated with graded mortality risk. Metabolic abnormalities explain the largest share of the mortality gradient, suggesting that population and clinical strategies focused on metabolic risk reduction should be prioritized to reduce CKM-related mortality.

Graphical abstract

Highlights

• •We examined the prevalence and mortality of cardiovascular-kidney-metabolic (CKM) syndrome stages in US adults between 1999 and 2018.
• •Almost 90% of US adults had stage 1 or higher CKM syndrome.
• •The prevalence of CKM stages 0 and 4 decreased between 1999 -2018, but that of CKM stages 1 and 3 increased.
• •The risk of all-cause and CVD mortality showed a graded increase from stage 1 to 4 compared to stage 0.
• •Metabolic factors drive most of the excess mortality.

1. Introduction

Cardiovascular-kidney-metabolic (CKM) syndrome is a novel disease staging system, defined in a statement from the American Heart Association (AHA) [1], that reflects the pathophysiological interplay between cardiovascular, kidney, and metabolic health. CKM syndrome has five stages, ranging from Stage 0 (no risk factors) to Stage 4 (clinical cardiovascular [CVD] disease) [1]. Poor CKM health is a key driver of premature mortality, particularly CVD mortality and excess morbidity, resulting in a substantial burden on healthcare systems [2].

Estimating the burden of CKM syndrome presents opportunities for the prevention and management of its risk factors, contributory diseases, and adverse clinical outcomes. Prior studies have attempted to characterize the prevalence and mortality of CKM syndrome in the US, with the reported prevalence of each CKM stage remaining stable between 2011 and 2020 [3]. However, the long-term temporal trends in the prevalence of CKM syndrome have not been well-studied. Using a nationally representative cohort of US adults, we aimed to assess the prevalence and mortality of CKM syndrome stages between 1999 and 2018. Importantly, by studying trends from 1999 through 2018, we capture earlier population-level changes in adiposity and metabolic health that predate many contemporary analyses that began in the 2010s. This extended timeline enables detection of early inflection points in CKM stage prevalence that shorter surveillance windows may miss, and it provides a pre-COVID-19 baseline against which future changes can be compared.

2. Methods

The National Health and Nutrition Examination Survey (NHANES) is approved by the NCHS Research Ethics Review Board, and its participants provide written informed consent. Since publicly available, deidentified NHANES data were used, the ethical approval for this study was waived.

2.1. Data and sample

Using data from NHANES (1999–2018), we defined CKM stages, ranging from 0 (no risk factors) to 4 (clinical CVD), as described in the presidential advisory from the AHA. Briefly, the stages are defined as follows: stage 0 (no CKM risk factors), stage 1 (excess or dysfunctional adiposity), stage 2 (metabolic risk factors or kidney disease), stage 3 (subclinical CVD), or stage 4 (established CVD). Stage 4 is further divided into two substages: Stage 4a, which includes patients without renal failure, and Stage 4b, which includes patients with renal failure. Advanced CKM syndrome stages were defined as stages 3 or 4. We excluded participants younger than 20 years, pregnant patients, and those with missing mortality data and NHANES fasting subsample weights. The 10-year PREVENT (Predicting Risk of Cardiovascular Disease Events) risk equation from the AHA was used to estimate CVD risk [4]. Mortality was calculated through linkage with the U.S. National Death Index through December 2019. Cardiovascular mortality was defined as death from diseases of the heart (International Classification of Diseases [ICD]-10 codes I00-I09, I11, I13, I20-I151) and cerebrovascular diseases (ICD-10 codes I60-I69).

2.2. Statistical analysis

We calculated the prevalence and all-cause and CVD mortality rates of each stage using estimates adjusted for age and sex to the 2010 US Census. The rates were compared across the subgroups of sex, age, and ethnicity using log-linear Wald tests. To estimate the temporal trends in prevalence, first, we used polynomial logistic regression to assess for nonlinearity [5]. If nonlinearity was detected, the proportions and standard errors were transformed to the log-odds scale before fitting a joinpoint regression model using Joinpoint software to identify the points where trend changes occurred [5]. Finally, survey-weighted logistic regression was used to estimate the trends for each CKM stage. In the case of nonlinearity, a logistic spline regression model was fit.

Survey-weighted Cox regression was used to estimate the risk of all-cause and CVD mortality with each stage. Hazard ratios (HRs) with corresponding 95% confidence intervals (CIs) were calculated. To investigate the association between CKM syndrome and premature all-cause and CVD mortality, we used log-binomial regression to generate risk ratios (RRs). Premature mortality was defined as the death of a person aged <65 years [6]. All analyses were adjusted for survey year, age, sex, and self-reported race or ethnicity in Model 1. The analyses were repeated by building Model 2, which was adjusted for body mass index (BMI), waist circumference, total cholesterol, statin use, antihypertensive use, hemoglobin A1c levels, and urinary albumin-to-creatinine ratio (UACR). To explore which CKM components contributed most to the CKM–mortality gradient, we fitted a series of sequential Cox models. Model A (demographics) adjusted for survey year, age, sex, and race/ethnicity; Model B (+adiposity) added BMI and waist circumference; Model C (+metabolic) additionally included diabetes, hypertension, total cholesterol, HDL/triglycerides, statin use, antihypertensive use, and smoking; and Model D (full) further added kidney measures (eGFR and urinary albumin-to-creatinine ratio). For each CKM stage, the percent attenuation in the association with mortality was calculated sequentially across models. Specifically, we compared the log-hazard ratio from the baseline model (demographics only) with those from subsequent models that added adiposity, metabolic, and kidney measures. The percent attenuation reflects the proportion of the CKM–mortality association explained by the added set of variables. Stata 17.0 was used for all analyses, except for joinpoint regression.

3. Results

The final study population included 21,334 adults. The mean age was 50.6 (SD 17.8), and 50.7% were females. Over a median length of follow-up of 9.0 years (IQR 4.9 to 13.7), there were 3284 deaths (1048 cardiovascular deaths).

Between 1999 and 2018, the prevalence of CKM stages was as follows: stage 0, 11.9% (95% CI: 11.2–12.6%); stage 1, 24.9% (95% CI: 24.1–25.7%); stage 2, 49.1% (95% CI: 48.2–50.0%), stage 3, 4.6% (95% CI: 4.4–4.9%), stage 4, 9.5% (95% CI: 9.0%–10.1%) (Table 1). The prevalence of CKM stages 4a and 4b was 9.45% (95% CI: 8.90–10.02%) and 0.09% (95% CI: 0.05–0.13%), respectively. Compared to men, women had a significantly higher prevalence of stage 0 (15.0% vs 8.7%) and a lower prevalence of stage 4 (8.3% vs 11.0%). Older adults (≥65 years and 45–64 years) were more likely to have higher CKM syndrome stages 3 and 4 as compared to 20–44-year-old adults. Compared to White adults (8.9%), Black adults had a higher prevalence (11.0%), but Asian adults had a lower prevalence of stage 4 (4.7%). The percentage of White and Asian adults with stage 3 was similar (4.5% and 5.3%, respectively), but Black adults had a higher prevalence (7.1%). Stages 3 and 4 had the highest proportion of all-cause mortality (28.2% and 22.8%, respectively; Table 1). The average predicted mortality rates per 1000 person-years increased with advancing CKM stage: 28.8 in stage 0, 40.1 in stage 1, 99.0 in stage 2, 437.6 in stage 3, and 458.6 in stage 4. The proportion of CVD deaths also exhibited a graded increase across CKM syndrome stages, ranging from 0.8% in stage 0 to 8.7% in stage 4.

Table 1.

Prevalence of cardiovascular-kidney-metabolic syndrome stages and related deaths among US adults, 1999–2018.

Characteristic	Prevalence, % (95% CI)					Wald (log-linear) P-value
	Stage 0	Stage 1	Stage 2	Stage 3	Stage 4
Total	11.88 (11.23–12.56)	24.90 (24.11–25.71)	49.05 (48.16–49.95)	4.63 (4.36–4.91)	9.54 (8.99–10.11)	–
Sex						<0.001
Male	8.73 (7.99–9.52)	24.89 (23.78–26.04)	49.93 (48.68–51.19)	5.42 (5.03–5.84)	11.03 (10.26–11.84)
Female	14.98 (13.98–16.03)	24.91 (23.73–26.12)	47.77 (46.53–49.01)	4.00 (3.68–4.35)	8.35 (7.71–9.03)
Age group						<0.001
20–44 years	19.36 (18.25–20.52)	33.99 (32.82–35.18)	44.55 (43.16–45.95)	0.18 (0.11–0.27)	1.92 (1.60–2.29)
45–64 years	7.29 (6.49–8.18)	21.05 (19.83–22.33)	60.68 (59.20–62.15)	1.28 (1.03–1.60)	9.70 (8.79–10.69)
≥65 years	2.06 (1.56–2.71)	9.26 (7.96–10.75)	35.87 (34.03–37.76)	23.26 (21.85–24.72)	29.55 (27.86–31.30)
Race/ethnicity						<0.001
Mexican American	10.04 (8.79–11.45)	29.46 (27.56–31.45)	48.95 (46.78–51.13)	3.17 (2.58–3.90)	8.38 (7.20–9.73)
Other Hispanic	10.59 (9.06–12.34)	27.68 (25.45–30.01)	49.98 (47.20–52.76)	3.68 (2.79–4.84)	8.06 (6.71–9.65)
Non-Hispanic White	13.58 (12.73–14.48)	25.55 (24.45–26.68)	47.76 (46.45–49.07)	5.28 (4.88–5.71)	7.83 (7.17–8.56)
Non-Hispanic Black	5.33 (4.45–6.36)	18.96 (17.45–20.56)	51.31 (49.07–53.54)	5.92 (5.03–6.96)	18.49 (16.87–20.24)
Non-Hispanic Asian	19.69 (17.27–22.36)	24.91 (21.81–28.29)	45.84 (41.98–49.75)	2.10 (1.51–2.90)	7.46 (5.97–9.29)
Other Race / Multi-Racial	9.46 (7.90–11.28)	25.76 (22.74–29.02)	45.73 (41.89–49.62)	7.41 (5.80–9.41)	11.65 (9.51–14.20)
All-cause death	6.54 (4.566, 9.294)	5.80 (4.801, 7.000)	9.33 (8.691, 10.01)	28.21 (19.61, 38.76)	22.83 (20.26, 25.62)	<0.001
Cardiovascular death	0.81 (0.3743, 1.732)	1.47 (1.009, 2.138)	2.23 (1.938, 2.573)	7.77 (5.16, 11.53)	8.66 (7.165, 10.43)	<0.001

The prevalence of CKM syndrome stage 0 decreased from 17.2% during 1999–2000 to 9.5% during 2017–2018 (Fig. 1). Joinpoint analysis identified an inflection point in the early 2000s, with a faster decline in stage 0 prevalence from 1999 to 2000 through 2003–2004 (P for trend = 0.009), followed by a slower decrease through 2017–2018 (P for trend = 0.019). The proportion of participants with stage 1 increased sharply from 18.8% in 1999–2000 to 28.7% in 2017–2018 (P for trend <0.001). The prevalence of stage 2 remained stable over the study period (P for trend >0.05). Between 1999–200 and 2017–2018, the prevalence of stage 3 increased slightly from 4.2% to 4.9% (P for trend = 0.008), while stage 4 declined gradually from 10.4% to 9.3% (P for trend = 0.009).

Fig. 1.

Age- and sex-adjusted prevalence of cardiovascular-kidney-metabolic syndrome stages, 1999–2018.

Compared to stage 0, stage 4 CKM syndrome was associated with an increased risk of all-cause (HR 2.87, 95% CI: 1.98–4.17) and CVD mortality (HR 4.68, 95% CI: 1.95–11.26). There was a graded increase in the risk of all-cause and CVD mortality from stages 1 to 4 compared to stage 0 (Table 2). Sequential adjustment (Table 3) revealed that the addition of adiposity measures produced only small reductions in the CKM–mortality associations (e.g., % attenuation for Stage 4 = 3.8%), whereas metabolic factors produced the largest attenuation (e.g., Stage 2 attenuation 82.3%; Stage 3 attenuation 67.3%; Stage 4 attenuation 43.6%). Inclusion of kidney measures in the full model yielded modest additional attenuation, most notable for Stage 4 (full attenuation 46.7%). Findings were consistent for cardiovascular mortality: adiposity accounted for a small share of attenuation (e.g., Stage 4 attenuation 8.4%), metabolic factors accounted for the largest share (e.g., Stage 2 attenuation 96.3%; Stage 3 attenuation 72.8%), and kidney measures produced additional but more modest attenuation for advanced stages (Stage 4 full attenuation 50.9%).

Table 2.

Association between cardiovascular-kidney-metabolic syndrome stages and mortality.

Outcome	Model	Hazard ratios (95%CI)					P-value
		Stage 0	Stage 1	Stage 2	Stage 3	Stage 4
All-cause death	Model 1a	Referent	1.09 (0.75, 1.59)	1.60 (1.16, 2.20)	2.59 (1.82, 3.68)	3.56 (2.48, 5.13)	<0.001
	Model 2b	Referent	1.03 (0.71, 1.51)	1.45 (1.04, 2.01)	1.95 (1.34, 2.83)	2.87 (1.98, 4.17)	<0.001
Cardiovascular death	Model 1a	Referent	1.74 (0.69, 4.36)	2.28 (1.01, 5.15)	4.51 (1.79, 11.39)	8.20 (3.38, 19.90)	<0.001
	Model 2b	Referent	1.38 (0.58, 3.28)	1.62 (0.71, 3.67)	2.55 (1.03, 6.31)	4.68 (1.95, 11.26)	<0.001

^aModel 1 was adjusted for survey year, age, sex, and self-reported race or ethnicity.

^bModel 2 was adjusted for survey year, age, sex, self-reported race or ethnicity, body mass index, waist circumference, total cholesterol, statin use, antihypertensive use, hemoglobin A1c levels, and urinary albumin-to-creatinine ratio.

Table 3.

Hazard ratios and percent attenuation by CKM stage for all-cause and cardiovascular mortality.

All-cause mortality
CKM Stage	HR (95% CI) Base	HR (95% CI) + Adiposity	% Attenuation	HR (95% CI) + Metabolic	% Attenuation	HR (95% CI) Full (+Kidney)	% Attenuation
Stage 1	1.09 (0.75–1.59)	1.07 (0.73–1.58)	14.6%	0.77 (0.45–1.33)	402.8%	0.77 (0.45–1.34)	407.0%
Stage 2	1.60 (1.16–2.20)**	1.53 (1.09–2.14)*	9.8%	1.09 (0.71–1.65)	82.3%	1.11 (0.72–1.69)	78.6%
Stage 3	2.59 (1.82–3.68)***	2.54 (1.74–3.70)***	2.0%	1.37 (0.79–2.35)	67.3%	1.32 (0.77–2.26)	70.7%
Stage 4	3.57 (2.48–5.13)***	3.40 (2.31–5.01)***	3.8%	2.05 (1.22–3.45)**	43.6%	1.98 (1.15–3.37)*	46.7%
CVD mortality
Stage 1	1.09 (0.75–1.59)	1.07 (0.73–1.58)	14.6%	0.77 (0.45–1.33)	402.8%	0.77 (0.45–1.34)	407.0%
Stage 2	1.60 (1.16–2.20)**	1.53 (1.09–2.14)*	9.8%	1.09 (0.71–1.65)	82.3%	1.11 (0.72–1.69)	78.6%
Stage 3	2.59 (1.82–3.68)***	2.54 (1.74–3.70)***	2.0%	1.37 (0.79–2.35)	67.3%	1.32 (0.77–2.26)	70.7%
Stage 4	3.57 (2.48–5.13)***	3.40 (2.31–5.01)***	3.8%	2.05 (1.22–3.45)**	43.6%	1.98 (1.15–3.37)*	46.7%

^⁎p < 0.05, ^⁎⁎p < 0.01, ^⁎⁎⁎p < 0.001.

Model adjustments:

Base (Demographics): adjusted for survey year, age, sex, and self-reported race/ethnicity.

+Adiposity: Base model plus body mass index (BMI) and waist circumference.

+Metabolic: Adiposity model plus diabetes, hypertension, total cholesterol, HDL cholesterol, triglycerides, statin use, antihypertensive use, and smoking status.

Full (+Kidney): Metabolic model plus estimated glomerular filtration rate (eGFR) and urinary albumin-to-creatinine ratio (UACR).

Participants with advanced CKM syndrome (stages 3 or 4) compared to early CKM syndrome (stages 0, 1, or 2) were more likely to be associated with an increased risk of premature all-cause death (RR 1.67, 95% CI: 1.13–2.46) and cardiovascular death (RR 2.50, 95% CI: 1.17–5.32). Likewise, participants with any CKM syndrome stage were associated with increased risk of premature all-cause death (RR 1.85, 95% CI: 1.07–3.20) and cardiovascular mortality (RR 4.02, 95% CI: 1.07–14.98) compared to those with stage 0 CKM syndrome.

4. Discussion

In this nationally representative cohort study of US adults spanning from 1999 to 2018, we observed a rapid decline in stage 0 CKM syndrome prevalence in the initial years, followed by a slower decrease from 2003 to 2004 through 2017–2018. This was accompanied by a constant increase in the prevalence of stages 1 and 3, although the increase in stage 3 was comparatively minor. In contrast, the prevalence of stage 2 remained stable, while that of stage 4 decreased slightly over the study period. Furthermore, we observed a graded risk increase in all-cause and cardiovascular mortality across the CKM syndrome stages. A similar, although attenuated, association was observed for premature all-cause and CVD death.

The early rapid decline in stage 0 (1999–2000 through 2003–2004) and concurrent rise in stage 1 are temporally consistent with the well-documented increases in population adiposity during the early 2000s, which likely contributed to the shift from “no risk” to “excess adiposity” stages. National estimates from NHANES document increasing obesity prevalence across the study era [7]. Later therapeutic and guideline changes, such as the introduction and uptake of SGLT2 inhibitors following initial approvals in 2013–2014 and evolving KDIGO guidance for CKD evaluation, could plausibly affect CKM stage prevalence and outcomes in subsequent years [8], [9]; however, such changes post-date or overlap the latter portion of our study period and cannot explain the earlier inflection observed in the early 2000s. We therefore interpret early and late period trends separately and cautiously.

Importantly, CKM stage was strongly prognostic for mortality. We observed a dose–response relationship between advancing CKM stage and both all-cause and cardiovascular death. These associations persisted after adjustment for individual metabolic, anthropometric, lipid, renal, and treatment-related factors. The persistence of elevated risk for stages 2–4 in fully adjusted models indicates that CKM staging captures cumulative and synergistic risk beyond the sum of individual risk factors. By contrast, the attenuation of risk for stage 1 after multivariable adjustment is consistent with its conceptual role as an exposure phase of excess adiposity or early metabolic dysfunction rather than established organ damage or high predicted cardiovascular risk. Sequential attenuation analyses indicate that the excess mortality associated with higher CKM stages is driven predominantly by metabolic factors, with only small contributions from adiposity and modest additional contributions from kidney dysfunction, suggesting that population-level efforts to control metabolic risk may yield the greatest reductions in CKM-related mortality.

We also found that advanced CKM stages were associated with substantially increased risk of premature all-cause and cardiovascular death. Combined with the substantial prevalence of stages 1 and 2 among younger adults, these findings emphasize the public health importance of early detection and prevention across the adult lifespan. At a population level, expanding early-stage prevalence increases the number of individuals at risk for progression to advanced and high-mortality states.

Our attenuation results suggest that focusing on metabolic risk control will likely yield the largest mortality benefit. A pragmatic two-tiered approach is warranted: population-level prevention (nutrition, activity, access to evidence-based weight-management) to reduce progression into Stage 1–2, and clinic-level opportunistic CKM staging to trigger targeted care: lifestyle programs for Stage 1; aggressive risk-factor control and consideration of cardiorenal agents for Stage 2; and integrated cardiorenal specialty care for Stages 3–4. These recommendations follow from our observational attenuation analyses and should be evaluated in implementation studies.

To operationalize CKM staging in practice, the minimum data set should include BMI (or waist), blood pressure, diabetes status/HbA1c, a basic lipid panel, eGFR and urine ACR, data commonly available in electronic health records. EHR-based staging with tiered decision support (lifestyle referral for Stage 1; intensified medical therapy for Stage 2; expedited specialty referral for Stage 3–4) may be feasible and scalable. Prioritization should focus on younger adults with Stage 1–2 and higher-burden subgroups (for example, Black adults), while ensuring equitable access to interventions.

These results highlight the potential value of incorporating CKM staging into future guideline-based risk calculators, such as the AHA's Life's Essential 8, to better stratify risk across the adult lifespan. Tailoring interventions by CKM stage may inform evidence-based updates to cardiovascular and renal prevention recommendations, guiding resource allocation for early-stage prevention versus advanced-stage management. It is important to highlight that formal incorporation into guidelines requires prospective validation and cost-effectiveness data; our paper provides epidemiologic rationale and prioritization hypotheses that guideline developers could test.

Our findings contrast with a previous study, which observed stable trends in CKM syndrome stages between 2011 and 2020 [3]. This discrepancy is likely due to our use of more time points to investigate long-term trends from 1999 to 2018; this may have helped discern previously undetected trends. In addition, a significant strength of our study is its close adherence to the National Center for Health Statistics (NCHS) Guidelines for Analysis of Trends [5]. Specifically, we first used recommended methods to assess nonlinearity and, if nonlinearity was identified, deployed joinpoint regression. Moreover, a key advantage of our timeline is that it ends before the onset of the COVID-19 pandemic, thereby censoring its impact, which is closely linked to CKM outcomes, including mortality.

Limitations of our analysis include the potential underestimation of CKM syndrome stages 3 or 4 due to the unavailability of data that could help define these stages, such as cardiac biomarkers, coronary catheterization, computed tomography angiography, and a history of atrial fibrillation or peripheral artery disease. Additionally, CVD diagnosis was based on self-reported data. Finally, the NHANES design is repeated cross-sectional; therefore, we cannot estimate individual-level transitions (person-level stage progression) directly. Estimating transition probabilities would require longitudinal repeated measures or linkage to cohort studies, which were beyond the scope of the present analysis.

In conclusion, nearly 90% of US adults have Stage 1 or higher CKM syndrome, and CKM stage is associated with graded mortality risk. Metabolic abnormalities explain the largest share of the mortality gradient, suggesting that population and clinical strategies focused on metabolic risk reduction should be prioritized to reduce CKM-related mortality.

CRediT authorship contribution statement

Huzaifa Ahmad Cheema: Writing – original draft, Methodology, Investigation, Data curation, Conceptualization. Abia Shahid: Writing – original draft, Investigation, Formal analysis, Data curation, Conceptualization. Saad Ur Rehman: Writing – original draft, Investigation, Data curation, Conceptualization. Khurram Nasir: Writing – review & editing, Visualization, Validation, Project administration. Jaideep Patel: Writing – review & editing, Visualization, Validation. Anandita Kulkarni: Writing – review & editing, Validation, Investigation. Janani Rangaswami: Writing – review & editing, Visualization, Investigation. Sarju Ganatra: Writing – review & editing, Visualization, Validation, Supervision. Sourbha S. Dani: Writing – review & editing, Supervision, Project administration, Methodology, Conceptualization.

Financial support

No financial support was received for this study.

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.