ABSTRACT
Background and Aims
To examine the association between sugar‐sweetened beverage (SSB) intake and all‐cause mortality among U.S. adults with and without metabolic syndrome (MetS), and to identify potential nonlinear dose‐response relationships and high‐risk subpopulations.
Methods
We analyzed data from 26,885 U.S. adults aged 20–79 years who participated in NHANES 2003–2018. MetS was defined based on the NCEP ATP III criteria. SSB intake was assessed using a single 24‐h dietary recall and quantified as total beverage weight (g/day), energy intake (kcal/day), and sugar intake (g/day). Cox proportional hazards models and restricted cubic splines were used to assess the association of SSB intake and mortality in MetS and non‐MetS groups.
Results
Among 8,656 participants with MetS and 18,229 without, total weight and energy intake from SSBs showed non‐linear association with the risk of all‐cause mortality in the MetS participants (P non‐linear < 0.05), with inflection points of approximately 950 g/day and 90 kcal/day, respectively. Above this level, each standard deviation (SD) increment was associated with 14% (HR: 1.14, 95% CI: 1.01–1.29) and 18% (HR: 1.18, 95% CI: 1.08–1.30) higher mortality risk. For SSB‐derived sugar, the non‐linear trend missed statistical significance (p = 0.07), though risk nadir appeared at ~20 g/day, with subsequent SD increments linked to a 19% risk increase (HR: 1.19, 95% CI: 1.09–1.30). No linear or non‐linear association was presented among individuals without MetS.
Conclusion
High SSB intake is associated with increased all‐cause mortality in adults with MetS, with nonlinear threshold effects.
Keywords: all‐cause mortality, metabolic syndrome, sugar‐sweetened beverages
1. Introduction
The global burden of metabolic syndrome (MetS), a cluster of metabolic abnormalities including central obesity, dyslipidemia, hypertension, and hyperglycemia, has been rising steadily in both developed and developing countries, now affecting approximately one in four adults worldwide [1, 2]. As a major risk factor for cardiovascular disease, type 2 diabetes, and premature mortality, MetS has emerged as a critical target for preventive interventions in dietary and public health research [3].
Among the modifiable lifestyle factors contributing to MetS, sugar‐sweetened beverages (SSBs) are of particular concern. As a leading global source of added sugars, excessive SSB consumption has been consistently linked to adverse metabolic outcomes, including obesity, insulin resistance, and type 2 diabetes [4, 5, 6]. Several large‐scale studies have identified a positive association between SSB intake and the development of MetS, suggesting that frequent consumption of sugary beverages may initiate or exacerbate metabolic dysregulation [3, 7, 8]. Because SSBs may affect multiple interrelated metabolic processes rather than a single marker alone, MetS provides a clinically meaningful phenotype for evaluating combined metabolic vulnerability associated with SSB exposure. Although SSB consumption has declined in the United States (U.S.), NHANES data from 2017 to 2018 showed that 65.3% of non‐Hispanic White adults and 77.8% of Hispanic adults consumed SSBs on a given day [9], and recent NHIS 2022 data reported past‐month SSB consumption in 82.69% of U.S. adults [10]. Notably, SSB intake was not merely occasional.
While the role of SSBs in the development of MetS is well documented, less is known about the relationship between SSB intake and mortality among individuals who already have MetS. Most prior research has focused on general populations or specific high‐risk subgroups, such as individuals with inflammatory bowel disease or chronic kidney disease [11, 12, 13]. However, evidence regarding mortality outcomes in those with established metabolic abnormalities is limited. For example, a recent U.S. study reported that high SSB intake was associated with a nearly three‐fold increase in cardiovascular mortality among adults with metabolic dysfunction‐associated steatotic liver disease, but no significant link was observed with all‐cause mortality [14]. Moreover, emerging evidence suggests that the association between SSB consumption and mortality may be nonlinear, characterized by threshold or inflection points beyond which health risks escalate markedly [5]. This possibility may be particularly relevant in individuals with MetS, a population in whom dietary modification is an important component of cardiometabolic risk management [15].
To address these gaps, we conducted a nationally representative prospective cohort study using data from the U.S. National Health and Nutrition Examination Survey (NHANES) to evaluate the associations between SSB consumption and all‐cause mortality among individuals with and without MetS. We further explored dose‐response relationships, potential nonlinear thresholds, and effect modification across key demographic and metabolic subgroups.
2. Methods
2.1. Study Design and Participants
The NHANES is an ongoing, nationally representative survey designed to monitor the health and nutritional status of the U.S. population [16]. It collects comprehensive information through demographic interviews, physical examinations, laboratory assessments, and dietary surveys, with data released biennially [17]. The NHANES study protocol was approved by the Research Ethics Review Board of the National Center for Health Statistics, and all participants provided written informed consent.
This study utilized data from NHANES cycles conducted between 2003 and 2018. We included participants aged 20 to 79 years who had complete 24‐h dietary recall data. As shown in Figure 1, among 44,790 individuals aged 20–79 years, 9715 participants without valid dietary recall data and 8190 participants with missing information required to define MetS status or without mortality follow‐up data were excluded. The final analytic sample included 26,885 participants, comprising 8656 individuals with MetS and 18,229 without MetS.
Figure 1.
Flowchart of study participant selection from the NHANES cohort. MetS, metabolic syndrome. Note: MetS classification accounted for partial missing data to minimize selection bias: (1) MetS was confirmed if ≥ 3 components were present, regardless of missing items; (2) Non‐MetS was confirmed if the sum of existing positive components plus all missing items could not mathematically reach 3. Remaining indeterminate cases were excluded.
2.2. Definition of Metabolic Syndrome
MetS was defined according to the criteria of the National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) as the presence of at least three of the following five components [18]: (1) waist circumference ≥ 88 cm in women and ≥ 102 cm in men (≥ 80 cm in women and ≥ 90 cm in men if Asian ancestry); (2) triglyceride level ≥ 150 mg/dL; (3) high‐density lipoprotein cholesterol (HDL‐C) < 40 mg/dL in men or < 50 mg/dL in women; (4) blood pressure ≥ 130/85 mmHg or use of antihypertensive medication; and (5) fasting glucose ≥ 100 mg/dL. Participants with partial missing data in MetS components were strictly evaluated; they were classified and retained only if their available data were mathematically sufficient to confirm or rule out the diagnosis, while those with indeterminate status were excluded.
2.3. Assessment of Sugar‐Sweetened Beverage Intake
Dietary intake was assessed using a 24‐h dietary recall interview. SSBs were defined according to standard NHANES protocols and included soft drinks, fruit drinks, nutritional drinks, sweetened teas and coffees, sports and energy drinks, smoothie, and carbonated or bottled water. Excluded beverages included 100% fruit juice, alcoholic beverages, fluid replacement, flavored milks, and self‐sweetened beverages [19]. Three key exposure variables were derived: (1) total sugar intake from SSBs (g/day); (2) total gram of SSBs consumed (g/day); (3) total energy intake from SSBs (kcal/day).
3. Mortality Outcomes and Covariates
The primary outcome was all‐cause mortality, determined through linkage to the National Death Index. Follow‐up time was calculated from the date of the NHANES interview to the date of death or December 31, 2019, whichever came first. Associations between SSB intake and all‐cause mortality were assessed separately among participants with and without MetS. Covariates included age, sex, race/ethnicity, educational attainment, marital status, household income, smoking status, alcohol consumption, body mass index (BMI), medical history (including congestive heart failure, coronary heart disease, stroke, hypertension, diabetes, and cancer), and NHANES cycle year.
4. Statistical Analysis
All analyses incorporated NHANES‐provided sample weights to account for the complex, multistage survey design. Continuous variables were expressed as medians and interquartile ranges and compared using the Wilcoxon rank‐sum test. Categorical variables were reported as weighted percentages with 95% confidence intervals (CIs) and compared using the Rao‐Scott adjusted chi‐square test to account for complex survey design.
Weighted Cox proportional hazards models were used to estimate hazard ratios (HRs) and 95% CIs for the association between each SSB intake indicators and all‐cause mortality in MetS and non‐MetS groups. Restricted cubic spline (RCS) was employed to assess potential nonlinear relationships between each SSB intake indicators and all‐cause mortality. In the case of a nonlinear relationship (Pnon‐linear < 0.05), a threshold was identified at the point where the slope of the spline curve changed. Subsequently, a two‐piecewise Cox proportional hazards model was fitted to examine potential threshold effects. Stratified analyses were conducted to explore effect modification by age, sex, and individual MetS components. All multivariable models were adjusted for the aforementioned covariates.
Statistical analyses were performed using R (version 4.2.2; R Foundation for Statistical Computing, Vienna, Austria) and SAS version 9.4 (version 9.4; SAS Institute, Cary, NC, USA). A two‐sided p < 0.05 was considered statistically significant.
5. Results
5.1. Baseline Characteristics
Compared with participants without MetS, those with MetS were older and had a greater cardiometabolic burden, with higher median age (54.59 vs. 40.85 years), BMI (31.90 vs. 25.51 kg/m2), waist circumference (109.17 vs. 90.86 cm), fasting blood glucose (108.80 vs. 94.99 mg/dL), and HbA1c (5.66% vs. 5.27%). The prevalence of hypertension (84.55% vs. 30.09%) and diabetes (30.48% vs. 4.73%) was markedly higher in the MetS group. During a median follow‐up of 8.33 years, MetS participants also demonstrated a significantly elevated risk of all‐cause mortality (13.35% vs. 5.94%). In addition, they were more likely to have lower educational attainment and lower household income (Table 1).
Table 1.
Characteristics of study participants by metabolic syndrome status.
| Variables | MetS | No MetS | p value |
|---|---|---|---|
| Age, years | 54.59 (43.45–65.38) | 40.85 (29.14–54.42) | < 0.01 |
| Male, % | 4063 [48.42 (46.95, 49.89)] | 9478 [50.44 (49.60, 51.29)] | 0.02 |
| Race, % | 0.01 | ||
| Mexican American | 1514 [8.20 (6.74–9.67)] | 2962 [8.53 (7.30–9.77)] | |
| Other Hispanic | 816 [4.75 (3.81–5.68)] | 1683 [5.54 (4.74–6.34)] | |
| Non‐Hispanic White | 4087 [71.34 (68.68–73.99)] | 8233 [68.82 (66.46–71.17)] | |
| Non‐Hispanic Black | 1550 [9.39 (8.13–10.64)] | 3634 [10.25 (9.06–11.44)] | |
| Other Race | 689 [6.33 (5.43–7.22)] | 1717 [6.86 (6.08–7.64)] | |
| Education, % | < 0.01 | ||
| High school or less | 4747 [46.67 (44.54–48.80)] | 8246 [36.79 (35.01–38.58)] | |
| Some college or higher | 3898 [53.33 (51.20–55.46)] | 9971 [63.21 (61.42–64.99)] | |
| Family income, % | < 0.01 | ||
| Low | 2610 [21.34 (19.82–22.86)] | 5005 [19.56 (18.29–20.83)] | |
| Medium | 3167 [35.70 (34.12–37.29)] | 31.69 [31.69 (30.29–33.10)] | |
| High | 2150 [36.50 (34.31–38.69)] | 5602 [42.40 (40.53–44.27)] | |
| Marital status, % | 0.01 | ||
| Married/Cohabiting | 5381 [65.21 (63.63–66.78)] | 11041 [62.65 (61.19–64.12)] | |
| Single/Divorced/Widowed | 3273 [34.79 (33.22–36.37)] | 7182 [37.35 (35.88–38.81)] | |
| BMI, kg/m2 | 31.90 (28.65–36.54) | 25.51 (22.67–28.79) | < 0.01 |
| WC, cm | 109.17 (101.72–118.98) | 90.86 (82.20–99.43) | < 0.01 |
| Smoking, % | 4395 [52.57 (51.03–54.12)] | 8255 [45.97 (44.55–47.39)] | < 0.01 |
| Alcohol consumer, % | 5197 [67.36 (65.47–69.25)] | 12661 [78.64 (77.36–79.92)] | < 0.01 |
| SBP, mmHg | 129.60 (118.20–139.91) | 115.39 (107.51–123.90) | < 0.01 |
| DBP, mmHg | 73.27 (64.56–81.73) | 69.27 (62.57–75.58) | < 0.01 |
| FBG, mg/dL | 108.80 (101.83–121.92) | 94.99 (89.92–100.95) | < 0.01 |
| HbA1c, % | 5.66 (5.35–6.16) | 5.27 (5.03–5.50) | < 0.01 |
| Insulin, pmol/L | 86.79 (57.53–133.28) | 42.40 (28.22–65.22) | < 0.01 |
| TC, mg/dL | 193.21 (164.92–224.27) | 190.63 (164.83–218.57) | < 0.01 |
| TG, mg/dL | 159.09 (107.90–216.05) | 87.38 (63.10–118.19) | < 0.01 |
| HDL‐C, mg/dL | 40.85 (35.20–47.91) | 55.57 (46.96–66.41) | < 0.01 |
| LDL‐C, mg/dL | 112.93 (89.28–138.62) | 110.47 (89.20–134.30) | < 0.01 |
| Hypertension, % | 7374 [84.55 (83.29–85.80)] | 6018 [30.09 (29.14–31.04)] | < 0.01 |
| Diabetes, % | < 0.01 | ||
| Normal | 1451 [19.96 (18.56–21.36)] | 12182 [72.81 (71.75–73.88)] | |
| Pre‐diabetes | 4068 [49.55 (48.04–51.07)] | 4735 [22.46 (21.48–23.44)] | |
| Diabetes | 3134 [30.48 (29.08–31.89)] | 1302 [4.73 (4.31–5.15)] | |
| Cancer, % | 1106 [13.62 (12.47–14.77)] | 1319 [7.47 (6.94–8.00)] | < 0.01 |
| Congestive heart failure, % | 509 [4.81 (4.20–5.42)] | 299 [1.16 (0.98–1.34)] | < 0.01 |
| Coronary heart disease, % | 622 [6.69 (5.97–7.42)] | 463 [1.98 (1.67–2.29)] | < 0.01 |
| Stroke, % | 533 [5.10 (4.49–5.72)] | 412 [1.70 (1.48–1.91)] | < 0.01 |
| All‐cause mortality, % | 1434 [13.35 (12.38–14.32)] | 1601 [5.94 (5.43–6.45)] | < 0.01 |
Note: Continuous variables are presented as weighted interquartile Range, Categorical variables were presented as unweighted number and weighted percentage with 95% confidence interval.
Abbreviations: BMI, body mass index; DBP, diastolic blood pressure; FBG, fasting blood glucose; HbA1c, hemoglobin A1c; HDL‐C, high‐density lipoprotein cholesterol; LDL‐C, low‐density lipoprotein cholesterol; MetS, metabolic syndrome; SBP, systolic blood pressure; TC, total cholesterol; TG, triglycerides; WC, waist circumference.
With respect to SSB intake, those with MetS consumed significantly greater total beverage weight (867.41 g vs. 790.40 g, p < 0.01). However, Median energy intake and sugar intake from SSBs was slightly lower in the MetS group (91.32 kcal vs. 117.83 kcal; 18.34 g vs. 22.84 g), but the difference was not statistically significant (p = 0.19; p = 0.80).
5.2. Association of SSB Intake With All‐Cause Mortality
RCS analyses revealed varying non‐linear patterns between specific domains of SSB consumption and all‐cause mortality within the MetS cohort (Figure 2). For total weight and energy from SSBs, distinct threshold effects were identified at approximately 950 g/day and 90 kcal/day, respectively (P non‐linear < 0.05). Above this level, each standard deviation (SD) increment was associated with 14% and 18% increase in mortality risk, respectively (HR = 1.14, 95% CI: 1.01–1.29, p = 0.04; HR = 1.18, 95% CI: 1.08‐1.30, p < 0.01). A non‐significant, non‐linear trend was observed for sugar intake from SSBs (p = 0.07); visually, however, the risk reached its nadir at ~ 20 g/day, past each SD increment was associated with heightened mortality (HR = 1.19; 95% CI: 1.09–1.30) (Table 2).
Figure 2.
Non‐linear associations of SSB consumption with all‐cause mortality among participants with metabolic syndrome. Restricted cubic splines within weighted Cox proportional hazards regression models describe the multi‐variable adjusted Hazard Ratios (solid lines) and 95% confidence intervals (shaded areas) for (A) total gram weight of SSBs, (B) energy intake from SSBs, and (C) sugar from SSBs. The green dashed lines indicate the respective inflection points (thresholds). Models were adjusted for age, sex, race, education level, marital status, household income, smoking status, alcohol consumption, body mass index, medical history (congestive heart failure, coronary heart disease, stroke, hypertension, diabetes, and cancer), and NHANES cycle. NHANES, National Health and Nutrition Examination Survey; SSB, sugar‐sweetened beverage.
Table 2.
Analysis of the threshold effect of sugar‐sweetened beverages and mortality risk among US Adults.
| Per 1 SD | Gram weight of SSBs, g | Energy from SSBs, kcal | Sugar intake from SSBs, g | |||
|---|---|---|---|---|---|---|
| Adjusted HR (95% CI) | p value | Adjusted HR (95%CI) | p value | Adjusted HR (95%CI) | p value | |
| MetS | ||||||
| Fitting by the standard Cox model | 1.05 (0.95–1.15) | 0.38 | 1.14 (1.05–1.23) | < 0.01 | 1.13 (1.05–1.22) | < 0.01 |
| p value for nonlinearity | < 0.01 | 0.03 | 0.07 | |||
| Inflection point | 950 | 90 | 20 | |||
| Intake ≤ Inflection point | 0.94 (0.87–1.02) | 0.14 | 0.95 (0.85–1.06) | 0.35 | 0.97 (0.88–1.08) | 0.60 |
| Intake > Inflection point | 1.14 (1.01–1.29) | 0.04 | 1.18 (1.08–1.30) | < 0.01 | 1.19 (1.09–1.30) | < 0.01 |
| No MetS | ||||||
| Fitting by the standard Cox model | 1.07 (0.99–1.16) | 0.08 | 1.09 (0.99–1.21) | 0.08 | 1.07 (0.97–1.17) | 0.16 |
| p value for nonlinearity | 0.20 | 0.42 | 0.61 | |||
Note: HR was calculated after adjusting for age, sex, race, education level, marital status, household income, smoking status, alcohol, BMI, medical history (congestive heart failure, coronary heart disease, stroke, hypertension, diabetes, cancer), and NHANES cycle.
Abbreviations: CI, confidence interval; HR, hazard ratio; MetS, metabolic syndrome; NHANES, National Health and Nutrition Examination Survey; SD, standard deviation; SSBs, sugar‐sweetened beverages.
In contrast, standard Cox models for participants without MetS yielded a marginal upward trend for both SSB gram weight and energy intake, though neither reached statistical significance (Table 2). And no significant nonlinear associations were identified (Supporting Information S1: Figure 1).
5.3. Subgroup Analyses Among Mets Participants
Participants aged < 65 years old and males reported higher SSB intake (Supplementary Table 1). However, mortality risks were more pronounced among older adults (HR for energy = 1.15; sugar = 1.13) and females (HR for energy = 1.23; sugar = 1.23; all p < 0.05). Significant interactions were observed for sex (p = 0.02 for sugar), indicating a stronger relative effect of sugar from SSBs on mortality in females.
Across MetS components, individuals with metabolic abnormalities, such as abdominal obesity, low HDL, and hypertriglyceridemia consumed higher total beverage weight (p < 0.05), despite variations in sugar and energy intake. These groups consistently showed elevated mortality risks associated with SSB consumption: abdominal obesity (HR for energy = 1.11; sugar = 1.10), low HDL (HR for energy = 1.20; sugar=1.16), impaired glucose tolerance (HR for energy = 1.13; sugar = 1.13), all p < 0.05. Interestingly, participants without hypertension (HR for weight = 1.12; energy = 1.14; sugar = 1.13) or without impaired glucose (HR for energy = 1.10) also showed significant risk, suggesting that harmful effects of SSBs may manifest before clinical diagnosis. Significant interactions were observed in low HDL cholesterol subgroup analysis (P for interaction = 0.03) (Table 3).
Table 3.
Subgroup analysis of the association between sugar‐sweetened beverages and mortality risk.
| Per 1 SD | Gram weight of SSBs, g | Energy from SSBs, kcal | Sugar intake from SSBs, g | ||||||
|---|---|---|---|---|---|---|---|---|---|
| HR (95% CI) | p value | p interaction | HR (95% CI) | p value | p interaction | HR (95% CI) | p value | p interaction | |
| Age | 0.32 | 0.80 | 0.86 | ||||||
| ≥ 65 years | 1.08 (0.99–1.19) | 0.09 | 1.15 (1.04–1.28) | 0.01 | 1.13 (1.01–1.27) | 0.04 | |||
| < 65 years | 1.04 (0.96–1.13) | 0.36 | 1.06 (0.97–1.15) | 0.21 | 1.04 (0.96–1.12) | 0.33 | |||
| Sex | 0.63 | 0.06 | 0.02 | ||||||
| Male | 1.04 (0.96–1.13) | 0.35 | 1.06 (0.97–1.16) | 0.20 | 1.04 (0.96–1.12) | 0.38 | |||
| Female | 1.09 (1.00–1.18) | 0.05 | 1.23 (1.11–1.37) | < 0.01 | 1.23 (1.11–1.37) | < 0.01 | |||
| Abdominal obesity | 0.63 | 0.80 | 0.63 | ||||||
| Yes | 1.07 (0.98–1.15) | 0.13 | 1.11 (1.02–1.21) | 0.01 | 1.10 (1.02–1.20) | 0.02 | |||
| No | 1.05 (0.95–1.15) | 0.35 | 1.09 (0.97–1.22) | 0.17 | 1.06 (0.95–1.18) | 0.28 | |||
| Low HDL cholesterol | 0.13 | 0.03 | 0.06 | ||||||
| Yes | 1.10 (1.00–1.21) | 0.06 | 1.20 (1.10–1.30) | < 0.01 | 1.16 (1.07–1.25) | < 0.01 | |||
| No | 1.04 (0.95–1.13) | 0.39 | 1.04 (0.95–1.13) | 0.37 | 1.03 (0.95–1.12) | 0.43 | |||
| Hypertriglyceridemia | 0.59 | 0.70 | 0.57 | ||||||
| Yes | 1.01 (0.88–1.14) | 0.94 | 1.13 (0.99–1.28) | 0.06 | 1.12 (0.99–1.26) | 0.07 | |||
| No | 1.07 (0.97–1.18) | 0.16 | 1.11 (0.97–1.28) | 0.13 | 1.09 (0.96–1.23) | 0.19 | |||
| Hypertension | 0.28 | 0.83 | 0.70 | ||||||
| Yes | 1.02 (0.95–1.10) | 0.56 | 1.11 (1.02–1.20) | 0.01 | 1.08 (1.01–1.16) | 0.03 | |||
| No | 1.12 (1.01–1.23) | 0.03 | 1.14 (1.02–1.27) | 0.02 | 1.13 (1.02–1.25) | 0.02 | |||
| Impaired fasting glucose | 0.15 | 0.99 | 0.67 | ||||||
| Yes | 1.02 (0.93–1.11) | 0.72 | 1.13 (1.01–1.27) | 0.04 | 1.13 (1.01–1.26) | 0.03 | |||
| No | 1.08 (0.99–1.17) | 0.07 | 1.10 (1.01–1.20) | 0.02 | 1.08 (0.99–1.17) | 0.07 | |||
Note: HR was calculated after adjusting for age, sex, race, education level, marital status, household income, smoking status, alcohol, BMI, medical history (congestive heart failure, coronary heart disease, stroke, hypertension, diabetes, cancer), and NHANES cycle.
Abbreviations: CI, confidence interval; HR, hazard ratio; NHANES, National Health and Nutrition Examination Survey; SD, standard deviation; SSBs, sugar‐sweetened beverages.
5.4. High SSB Intake and Mortality Risk in MetS Subgroups
Among individuals exceeding intake thresholds (weight > 950 g/day, enereg > 90 kcal/day, sugar > 20 g/day), mortality risks remained elevated across key subgroups. Older adults (HR for weight = 1.33; energy = 1.41; sugar = 1.45), females (HR for energy = 1.31; sugar = 1.30), and those with low HDL (HR for weight = 1.17; energy = 1.19; sugar = 1.20) or impaired fasting glucose (HR for energy = 1.26; sugar = 1.23) were particularly vulnerable. Additionally, those without abdominal obesity (HR for weight = 1.53; energy = 1.59; sugar = 1.87) or without hypertension still exhibited strong associations, reinforcing that metabolic damage from SSBs may occur before overt MetS is detected (Supporting Information S1: Table S2).
6. Discussion
In this nationally representative cohort study of over 26,000 U.S. adults, we found that high SSB intake was significantly associated with increased all‐cause mortality in individuals with MetS. Importantly, this association appeared nonlinear, with potential inflection points above which mortality risk increased. Specifically, in the MetS population, mortality risk increased when SSB intake exceeded approximately 950 g/day of beverage weight, 90 kcal/day of SSBs‐derived energy, or 20 g/day of SSBs‐derived sugar, suggesting a threshold‐dependent pattern of risk. However, the nonlinear association for SSB‐derived sugar intake was not statistically significant, and this potential inflection point should be interpreted as exploratory. We also observed that individuals with MetS consumed a greater total weight of SSBs than those without MetS, despite similar SSBs‐derived sugar and energy intake, which may indicate differences in beverage consumption patterns that are not fully captured by sugar or energy content alone.
From a clinical and public health perspective, our findings add to existing evidence by identifying adults with MetS as a metabolically vulnerable group in whom higher SSB intake may be particularly relevant to mortality risk. In routine care, clinicians may consider assessing SSB intake more explicitly among patients with MetS, especially older adults, females, and individuals with abdominal obesity, low HDL cholesterol, or impaired fasting glucose. This may help inform future risk communication and dietary counseling, although the potential inflection points should not be interpreted as established intake thresholds or used as definitive clinical cutoffs. For public health practitioners, these findings may support prioritizing SSBs‐reduction strategies in adults with combined metabolic risk. However, because this study is observational, the identified thresholds should be interpreted as risk markers requiring further validation rather than definitive causal cutoffs for guidelines development.
While previous studies have consistently demonstrated a positive linear association between SSB intake and adverse health outcomes including all‐cause mortality, liver steatosis, and fibrosis, few have investigated potential nonlinear dose‐response relationships [5, 20, 21]. For instance, NHANES‐based research by Leung et al. reported a dose‐dependent increase in liver stiffness and steatosis with higher SSB consumption, although the analysis was based on linear trend models [20]. Similarly, in two large U.S. prospective cohorts, Malik et al. reported a dose‐response association between SSB consumption and total mortality, with pooled HRs increasing from 1.06 (95% CI: 1.03– 1.09) for two to six servings per week to 1.21 (95% CI: 1.13–1.28) for two or more servings per day, compared with less than one serving per month [22]. These studies underscore the metabolic burden associated with high SSB intake but do not address potential intake thresholds or inflection points. In contrast, our study used RCS to examine nonlinear associations and identified potential inflection points among individuals with MetS. Mortality risk appeared to increase above these levels, although the apparent threshold for SSB‐derived sugar intake remains uncertain given the nonsignificant nonlinear association. Notably, our findings align with a recent NHANES‐based CKD cohort study that also identified a nonlinear association between SSB intake and mortality, further supporting the relevance of threshold‐based dietary guidance in metabolically vulnerable populations [13].
SSBs‐related mortality risk may differ by age and sex among adults with MetS. In our subgroup analyses, older adults and females showed stronger associations between SSB intake and all‐cause mortality, despite generally lower reported intake levels. This finding should be interpreted cautiously, but it is consistent with NHANES‐based studies linking SSB intake to mortality in metabolically vulnerable adults, including those with metabolic dysfunction‐associated steatotic liver disease and chronic kidney disease [13, 14]. The California Teachers Study also reported a positive association between SSB intake and mortality among females [23], while a dose‐response meta‐analysis suggested greater SSBs‐related mortality risk in individuals aged ≥ 60 years [21]. These differences may reflect variations in metabolic reserve, body composition, hormonal status, and cumulative cardiometabolic burden, although the underlying mechanisms require further study.
Although the association of SSB consumption and mortality risk did not reach strict statistical significance among participants without MetS, a distinct directional consistency toward a higher mortality risk was still present, indicating potential population‐wide harm but with heightened sensitivity among those with MetS [24]. Moreover, participants without hypertension or impaired fasting glucose also demonstrated elevated mortality risks at higher SSB intake levels. These findings suggest that the adverse effects of excessive SSB consumption may precede overt metabolic disease. Supporting this notion, prior research has found that high SSB intake contributes to subclinical inflammation, ectopic fat storage, and insulin resistance even in individuals considered metabolically healthy. Stanhope et al. demonstrated that consuming fructose‐sweetened beverages, rather than glucose, led to significant increases in hepatic de novo lipogenesis and visceral adiposity, key markers of metabolic dysfunction [25]. Additionally, dietary sugar intake has been linked to systemic inflammation through increased free fatty acid synthesis in the liver, leading to oxidative stress and inflammatory mediators. Collectively, these data underline the importance of early dietary interventions targeting SSB consumption across the population, not solely in clinically at‐risk individuals.
Strengths of this study include its large, representative sample; objective linkage to mortality records; and comprehensive adjustment for confounders. We also employed spline regression to capture nonlinear patterns, which are often overlooked in dietary studies. However, the study has several limitations. First, dietary intake was assessed using self‐reported 24‐h dietary recall data, which may be subject to recall bias and may not fully capture usual long‐term intake. Second, residual confounding may persist despite multivariable adjustment. Third, cause‐specific mortality was not assessed in this analysis. Lastly, dietary intake was assessed at a single time point, which limits causal inference and may not capture changes in SSB consumption over time.
7. Conclusion
Our findings demonstrate a nonlinear, threshold‐dependent relationship between higher SSB intake and increased all‐cause mortality in adults with MetS. Notably, this risk is particularly pronounced among females. The observed inflection points may provide useful reference points for risk communication and dietary counseling.
Author Contributions
Tingting Lv: writing – original draft, conceptualization, methodology, data curation. Xintian Ren: conceptualization, software, data curation, methodology. Yanfang Tan: conceptualization, methodology. Min Li: conceptualization, writing – review and editing, funding acquisition, project administration, supervision, formal analysis, data curation, software, methodology.
Ethics Statement
The NHANES study protocol was approved by the Research Ethics Review Board of the National Center for Health Statistics, and all participants provided written informed consent.
Conflicts of Interest
The authors declare no conflicts of interest.
Transparency Statement
The corresponding author Min Li affirms that this manuscript is an honest, accurate, and transparent account of the study being reported; that no important aspects of the study have been omitted; and that any discrepancies from the study as planned (and, if relevant, registered) have been explained.
Supporting information
Supporting File
Acknowledgments
This work was supported by the National Natural Science Foundation of China under Grant (No. 82103902) and Grant (No. 82100633). Beijing Friendship Hospital, Capital Medical University (Contract grant number: yyqcjh2023‐9). The funding source had no involvement in study design, data collection, analysis, interpretation, writing the manuscript, or the decision to submit for publication.
Data Availability Statement
The data supporting the findings of this study are openly available in NHANES. These data were derived from the following resources available in the public domain: https://wwwn.cdc.gov/nchs/nhanes/Default.aspx.
References
- 1. Russell J. B. W., Koroma T. R., Sesay S., et al., “Prevalence and Correlates of Metabolic Syndrome Among Adults in Freetown, Sierra Leone: A Comparative Analysis of NCEP ATP III, IDF and Harmonized ATP III Criteria,” International Journal of Cardiology Cardiovascular Risk and Prevention 20 (2024): 200236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Tang H., Zhang X., Luo N., et al., “Temporal Trends in the Planetary Health Diet Index and Its Association With Cardiovascular, Kidney, and Metabolic Diseases: A Comprehensive Analysis From Global and Individual Perspectives,” Journal of Nutrition, Health and Aging 29, no. 5 (2025): 100520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Muñoz‐Cabrejas A., Guallar‐Castillón P., Laclaustra M., Sandoval‐Insausti H., and Moreno‐Franco B., “Association Between Sugar‐Sweetened Beverage Consumption and the Risk of the Metabolic Syndrome: A Systematic Review and Meta‐Analysis,” Nutrients 15, no. 2 (2023): 430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Hamner H. C., Dooyema C. A., Blanck H. M., et al., “Fruit, Vegetable, and Sugar‐Sweetened Beverage Intake Among Young Children, by State ‐ United States, 2021,” MMWR. Morbidity and Mortality Weekly Report 72, no. 7 (2023): 165–170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Meng Y., Li S., Khan J., et al., “Sugar‐ and Artificially Sweetened Beverages Consumption Linked to Type 2 Diabetes, Cardiovascular Diseases, and All‐Cause Mortality: A Systematic Review and Dose‐Response Meta‐Analysis of Prospective Cohort Studies,” Nutrients 13, no. 8 (2021): 2636. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Heo G. Y., Koh H. B., Park J. T., et al., “Sweetened Beverage Intake and Incident Chronic Kidney Disease in the UK Biobank Study,” JAMA Network Open 7, no. 2 (2024): e2356885. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Park S., Kim D. S., and Kang S., “Gene‐Diet Interactions in Carbonated Sugar‐Sweetened Beverage Consumption and Metabolic Syndrome Risk: A Machine Learning Analysis in a Large Hospital‐Based Cohort,” Clinical Nutrition ESPEN 64 (2024): 358–369. [DOI] [PubMed] [Google Scholar]
- 8. Ong J., Roem J., Ducharme‐Smith K., et al., “Association of Sodium and Sugar‐Sweetened Beverage Intake With Cardiovascular Disease Risk Factors in Adolescents and Young Adults With Obesity,” Clinical Pediatrics 63, no. 5 (2024): 669–679. [DOI] [PubMed] [Google Scholar]
- 9. Dai J., Soto M. J., Dunn C. G., and Bleich S. N., “Trends and Patterns in Sugar‐Sweetened Beverage Consumption Among Children and Adults by Race And/Or Ethnicity, 2003‐2018,” Public Health Nutrition 24 (2021): 2405–2410. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Sengupta S., Datta B. K., and Coughlin S. S., “Medical Debt Is Associated With Higher Sugar‐Sweetened Beverage Consumption Among U.S. Adults,” Journal of Epidemiology and Population Health 73 (2025): 203123. [DOI] [PubMed] [Google Scholar]
- 11. Jung H., Choi Y., and Kim B., “The Longitudinal Effect of Sugar‐Sweetened Beverage Consumption and Risk of All‐Cause and Cause‐Specific Mortality in Korean Adults: Results From the Health Examinees Study,” Journal of the American Nutrition Association 44 (2025): 468–477. [DOI] [PubMed] [Google Scholar]
- 12. Dan L., Fu T., Sun Y., et al., “Associations of Sugar‐Sweetened Beverages, Artificially Sweetened Beverages, and Natural Juices With Cardiovascular Disease and All‐Cause Mortality in Individuals With Inflammatory Bowel Disease in a Prospective Cohort Study,” Therapeutic Advances in Gastroenterology 16 (2023): 17562848231207305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Cai X. Y., Zhang N. H., Cheng Y. C., Ge S. W., and Xu G., “Sugar‐Sweetened Beverage Consumption and Mortality of Chronic Kidney Disease: Results From the US National Health and Nutrition Examination Survey, 1999‐2014,” Clinical Kidney Journal 15 (2022): 718–726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Zhou J., Liu C., Liu L., and Li L., “Sugar‐Sweetened Beverage Intake and Long‐Term Mortality in Individuals With Metabolic Dysfunction‐Associated Steatotic Liver Disease: A Longitudinal Analysis of the National Health and Nutrition Examination Survey Database,” European Journal of Gastroenterology & Hepatology 36 (2024): 608–614. [DOI] [PubMed] [Google Scholar]
- 15. Peterseim C. M., Jabbour K., and Kamath Mulki A., “Metabolic Syndrome: An Updated Review on Diagnosis and Treatment for Primary Care Clinicians,” Journal of Primary Care & Community Health 15 (2024): 21501319241309168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Chen T. C., Clark J., Riddles M. K., Mohadjer L. K., and Fakhouri T., “National Health and Nutrition Examination Survey, 2015‐2018: Sample Design and Estimation Procedures,” Vital and Health Statistics. Series 2, Data Evaluation and Methods Research 2, no. 184 (2020): 1–35. [PubMed] [Google Scholar]
- 17. Wang W., Sun P., Lv T., and Li M., “The Impact of Modifiable Health Metrics on Mortality for Older Adults With Low Cognitive Function,” Frontiers in Public Health 12 (2024): 1304876. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Chen X., Li A., and Ma Q., “Association of Estimated Glucose Disposal Rate With Metabolic Syndrome Prevalence and Mortality Risks: A Population‐Based Study,” Cardiovascular Diabetology 24 (2025): 38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Moss M. E., Luo H., Rosinger A. Y., Jacobs M. M., and Kaur R., “High Sugar Intake From Sugar‐Sweetened Beverages Is Associated With Prevalence of Untreated Decay in US Adults: NHANES 2013‐2016,” Community Dentistry and Oral Epidemiology 50 (2022): 579–588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Leung C. W. and Tapper E. B., “Sugar‐Sweetened Beverages Are Associated With Increased Liver Stiffness and Steatosis Among Apparently Healthy Adults in the United States,” Clinical Gastroenterology and Hepatology 20 (2022): 959–961.e1 e951. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Wang Y., Zhao R., Wang B., Zhao C., Zhu B., and Tian X., “The Dose‐Response Associations of Sugar‐Sweetened Beverage Intake With the Risk of Stroke, Depression, Cancer, and Cause‐Specific Mortality: A Systematic Review and Meta‐Analysis of Prospective Studies,” Nutrients 14 (2022): 777. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Malik V. S., Li Y., Pan A., et al., “Long‐Term Consumption of Sugar‐Sweetened and Artificially Sweetened Beverages and Risk of Mortality in US Adults,” Circulation 139 (2019): 2113–2125. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Pacheco L. S., J. V. Lacey, Jr. , Martinez M. E., et al., “Association Between Sugar‐Sweetened Beverage Intake and Mortality Risk in Women: The California Teachers Study,” Journal of the Academy of Nutrition and Dietetics 122 (2022): 320–333.e6 e326. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Doherty A. M., Lacko A. M., and Popkin B. M., “Sugar‐Sweetened Beverage (Ssb) Consumption Is Associated With Lower Quality of the Non‐SSB Diet in US Adolescents and Young Adults,” American Journal of Clinical Nutrition 113 (2021): 657–664. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Stanhope K. L., Schwarz J. M., Keim N. L., et al., “Consuming Fructose‐Sweetened, Not Glucose‐Sweetened, Beverages Increases Visceral Adiposity and Lipids and Decreases Insulin Sensitivity in Overweight/Obese Humans,” Journal of Clinical Investigation 119 (2009): 1322–1334. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supporting File
Data Availability Statement
The data supporting the findings of this study are openly available in NHANES. These data were derived from the following resources available in the public domain: https://wwwn.cdc.gov/nchs/nhanes/Default.aspx.