Abstract
Obesity is not only an important part of metabolic syndrome, but also closely related to the pathogenesis of a variety of malignancies. However, there are differing views on the relationship between obesity and the risk of digestive malignancies. In this context, it is necessary to assess the true cancer risk in people with different obesity states. In the past, some researchers proposed to use the metabolic syndrome severity z score to predict the risk of certain diseases, but since its introduction, there has been a lack of relevant studies in the Chinese population. Based on the Kailuan cohort, the association between metabolic syndrome severity z score and the risk of digestive system malignancy in the Chinese population was explored. In this study, 48,205 participants who had undergone 3 consecutive physical examinations since 2006 were collected. Cumulative metabolic syndrome severity z score (cMets-Z) was calculated using parameters such as fasting blood glucose, total cholesterol, high-density lipoprotein, systolic blood pressure, and body mass index. Participants were categorized into 4 groups based on cMetS-Z (Q1–Q4), and a Cox regression model was used to evaluate the risk of new digestive system malignancies. The mean age of participants was 48.99 ± 11.75 years. Over a median follow-up period of 11.03 years, 749 new cases of digestive cancers were identified, including esophageal, gastric, colorectal, liver, bile duct, and pancreatic cancers. Cox proportional hazards modeling revealed adjusted hazard ratios for gastrointestinal cancers in the Q2 to Q4 groups as 1.19 (0.96–1.48), 1.36 (1.08–1.71), and 1.72 (1.33–2.23). In site-specific analyses, we observed that this risk was more pronounced in esophageal cancer. Subgroup analyses showed that cMetS-Z was associated with digestive system malignancies, particularly in male participants with metabolic syndrome. Restricted cubic spline regresion results showed that cMets-Z had a linear relationship with the risk of digestive system tumors. We found that cMetS-Z was associated with an increased risk of gastrointestinal cancer. cMetS-Z can be used as a prospective tool to predict the risk of gastrointestinal malignancies in Chinese people.
Keywords: cMets-z, cohort study, digestive system cancer, obesity, risk factor
1. Introduction
Gastrointestinal (GI) cancers include malignancies of the esophagus, stomach, small bowel, colorectum, liver, pancreas, extrahepatic bile duct, and gallbladder. According to Global Cancer Statistics 2023, 3 of the 5 leading causes of cancer mortality worldwide are digestive system cancers – colorectal (9%), pancreatic (8%), and liver (6%).[1] In China, liver cancer ranks as the 2nd leading cause of cancer death, followed by gastric, esophageal, and colorectal cancers.[2] Although risk factors vary by anatomic site, several are shared across GI cancers, including obesity, alcohol consumption, smoking, diabetes,[3] and chronic inflammatory states.[4]
Obesity is a growing global health challenge. A recent US study reported an obesity prevalence of 21.5% among young people aged 19 to 21 years.[5] In China, the prevalence of overweight and obesity among adults is 34.8% and 14.1%, respectively.[6] Obesity is a core component of metabolic syndrome (MetS)[7] and is strongly linked to major chronic diseases – hypertension, diabetes, dyslipidemia, and coronary heart disease – as well as to the incidence of multiple malignancies.[8–10] Prior research has also shown robust associations between obesity and cancers of the digestive system,[11] though the strength (and, in some sites, the direction) of these associations remains debated.[12] Accordingly, better risk stratification across obesity phenotypes is needed.
Because conventional MetS definitions provide only coarse risk categories, investigators introduced the metabolic syndrome severity z-score (MetS-Z) – a continuous index calculated from waist circumference (WC), high-density lipoprotein (HDL), systolic blood pressure (SBP), triglycerides (TGs), and fasting blood glucose (FBG) using sex- and age-specific equations – to enhance prediction and enable longitudinal tracking. US studies suggest that MetS-Z provides additional prognostic information for certain outcomes,[13] yet its application in Chinese populations remains limited. Leveraging the Kailuan cohort, we therefore aimed to evaluate the association between MetS-Z and the risk of GI cancers.
2. Research methods
This study is based on the Kailuan cohort study and adopts a prospective research method; the Kailuan study has been described in detail before.[14] Briefly, this is a longitudinal cohort study that began in 2006. This study was approved by the Ethics Committee of Kailuan Hospital with a batch number of 200605. The clinical registration number is ChiCTR-TNC-11001489. All subjects signed the informed consent form.
2.1. Study population
We initially identified 57,927 adults in the Kailuan cohort who completed 3 consecutive routine health examinations beginning in 2006. We excluded individuals with a prior history of cancer, those missing essential baseline information, and those with incomplete laboratory test data (n = 9722), leaving 48,205 participants for analysis. The study flow is summarized in Figure 1.
Figure 1.
Flow chart for subjects of the study.
2.2. Data collection and related definitions
A biennial questionnaire survey was used to collect basic information. Demographic information such as age and gender, lifestyle habits such as smoking, drinking, and physical exercise, and history of hypertension, diabetes, and dyslipidemia were collected. In the study, the body mass index (BMI) was calculated using the formula BMI = weight (kg)/ height2 (m2). To measure blood pressure, a calibrated mercury sphygmomanometer with a suitable cuff was used after at least 5 minutes of sitting and resting. WC was measured using a tape measure, with the midpoint of the line between the lower edge of the rib cage and the upper edge of the hip joint as the measurement point. Fasting blood samples from the examiners were kept in a vacuum tube containing ethylene diamine tetraacetic acid or a coagulant for blood biochemical assays. The concentrations of FBG, total cholesterol, TG, low-density lipoprotein (LDL), HDL, and C-reactive protein were detected using an Autoanalyzer (Hitachi 747, Hitachi, Tokyo, Japan). Cirrhosis, nonalcoholic fatty liver disease, cholelithiasis, and gallbladder polyps are based on previously clinically established criteria[15,16] or abdominal ultrasound diagnosis through the medical records of the Tangshan medical insurance system.
The occurrence of malignant tumors of the digestive tract was obtained through questionnaires. In addition, cancer cases are further identified annually through the medical system in conjunction with the Provincial Vital Statistics, the Tangshan Medical Insurance System, and the Kailuan Group Guarantee Information System. Almost all of the health information of the insured persons is included in the medical insurance system of Tangshan City and the social security system of Kailuan. All cancer cases were reconfirmed based on specific clinical features or positive histopathological findings from the hospital where the patient was treated for malignancy. When pathology results are not available, the potential case is further evaluated by 2 oncologists. A cancer case can only be identified if 2 clinicians make the same diagnosis. For patients with multiple tumors and no histopathological findings, each tumor is recorded when it is difficult to determine whether the tumor is primary or metastatic. Cases of cancer were recorded according to the International Classification of Diseases, 10th edition: liver (C22.0), gallbladder or extrahepatic bile ducts (C23 and C24), stomach (C16), pancreas (C25), small intestine (C17), esophageal cancer (C15), and colorectal cancer (C18-C21).[17]
Hypertension,[18] diabetes mellitus,[19] and hyperlipidemia[20] are defined according to the guidelines. MetS is defined according to the 2005 United States Heart Association criteria: WC ≥ 102 cm (male), 88 ≥ cm (female); TG ≥ 150 mg/dL; HDL < 40 mg/dL (male); <50 mg/dL (women); SBP ≥ 130 or DBP ≥ 85 mm Hg; FBG > 110 mg/dL (including type 2 diabetes mellitus). In this study, 3 or more of 5 conditions can be defined as MetS[21].
In this study, MetS-Z was calculated as follows: MetS-Z = ‐4.8316 + 0.0315 × BMI ‐0.0272 × HDL + 0.004 × SBP + 0.8018 × lnTG + 0.0101 × FBG (male) and MetS-Z = ‐6.5231 + 0.0523 × BMI -0.0138 × HDL + 0.0081 × SBP + 0.6125 × lnTG + 0.0208 × FBG (female).[22] The cumulative MetS-Z (cMetS-Z) was calculated, and the formula cMetS-Z = (MetS-Z2006 + MetS-Z2008) × (time2008-time2006)/2 + (MetS-Z2008 + MetS-Z2010) × (time2010-time2008)/2 was used to evaluate the cumulative MetS-Z during the follow-up period.
2.3. Follow-up
The morbidity and mortality of malignant tumors of the digestive system were followed up by physical examination and questionnaire surveys twice a year and a visit to the Tangshan health care system once a year. The starting point of follow-up was the 2010 physical examination, and the end event of follow-up was the occurrence of a malignant tumor of the digestive system, death, or December 31, 2021.
2.4. Statistical methods
The analysis was performed using SAS 9.4 statistical software. The measurement data of normal distribution were expressed as mean ± SD, and one-way analysis of variance was used for comparison between groups. Measurement data of skew distribution were represented by median (Q1, Q3), and comparison between groups was performed by the Kruskal–Wallis rank sum test. Statistical data were expressed as absolute numbers and percentages, and the χ2 test was used for comparison between groups. Cumulative incidence was calculated by the Kaplan–Meier method, and a survival curve was drawn. Comparison of cumulative incidence between groups was performed by the log-rank test. The hazard ratio and 95% confidence interval for new digestive system tumors in different groups were analyzed by the COX proportional hazard model. We applied a conventional 3-step adjustment strategy: model 1 was unadjusted (crude); model 2 additionally adjusted for age and sex; and model 3 further adjusted for TGs, high-sensitivity C-reactive protein (hsCRP), marital status, alcohol consumption, smoking, household income, educational attainment, high-salt diet, sedentary behavior, physical exercise, high-fat diet, tea-drinking habits, and use of glucose-lowering, lipid-lowering, and antihypertensive medications. To address reverse causality and medication confounding, sensitivity analyses excluded participants who developed cancer within the first 5 years of follow-up and, separately, excluded users of lipid-lowering, glucose-lowering, and antihypertensive drugs. We then used restricted cubic spline regression to characterize the dose–response relationship between MetS-Z and GI cancer risk, with site-specific models; in addition to the above covariates, liver cancer models further adjusted for HBV infection, cirrhosis, and fatty liver, while gallbladder and extrahepatic cholangiocarcinoma models additionally adjusted for gallstones. Finally, subgroup analyses were stratified by age, sex, hypertension, diabetes, hyperlipidemia, MetS status, and individual MetS components, with 2-sided P values <.05 considered statistically significant.
3. Results
3.1. Basic information
At baseline, the mean age was 48.99 ± 11.75 years. Participants were stratified into quartiles by MetS-Z; the mean (±SD) MetS-Z values were Q1: −4.38 ± 0.51, Q2: −3.61 ± 0.14, Q3: −3.17 ± 0.12, and Q4: −2.59 ± 0.31. There were significant differences in age, sex, hs-CRP, WC, FBG, TG, HDL, and BMI among the 4 groups. In addition, educational background, marital status, income level, physical activity, sedentary lifestyle, recent smoking, recent alcohol consumption, history of tea consumption, high-fat diet, high salt intake, family history of cancer, lipid-lowering drugs, and blood pressure medications differed significantly among the 4 predesignated groups. However, no differences were observed between the group of subjects with gallstones and those taking hypoglycemic drugs. During a median follow-up of 11.03 years, 749 cases of malignant tumors of the digestive system were found, including 76 cases of esophageal cancer, 152 cases of gastric cancer, 16 cases of small bowel cancer, 284 cases of colorectal cancer, 151 cases of liver cancer, 21 cases of gallbladder and extrahepatic bile duct cancer, and 49 cases of pancreatic cancer. Overall, participants in the Q4 group had higher blood pressure levels and higher blood biochemical concentrations, including FBG, total cholesterol, TG, and C-reactive protein. At the same time, rates of MetS were highest, with 45.8% of subjects in the Q4 group having MetS. Details of subjects are shown in Table 1.
Table 1.
The baseline characteristics of the study population.
| Variables | Total (N = 48,205) | Q1 (N = 12,085) | Q2 (N = 11,958) | Q3 (N = 12,099) | Q4 (N = 12,063) | P-value |
|---|---|---|---|---|---|---|
| Age (yr) | 48.99 ± 11.75 | 50.08 ± 12.54 | 51.01 ± 12.08 | 48.42 ± 11.52 | 46.48 ± 10.22 | <.001 |
| Men, N (%) | 37,266 (77.3) | 5170 (42.8) | 9233 (77.2) | 11,101 (91.8) | 11,762 (97.5) | <.001 |
| FBG (mmol/L) | 5.42 ± 1.55 | 5.14 ± 1.26 | 5.35 ± 1.49 | 5.48 ± 1.56 | 5.71 ± 1.77 | <.001 |
| CRP (mg/L) | 0.71 (0.28–2.00) | 0.58 (0.20–1.63) | 0.70 (0.26–2.00) | 0.76 (0.30–2.00) | 0.90 (0.36–2.25) | <.001 |
| TG (mmol/L) | 1.29 (0.90–1.96) | 0.90 (0.67–1.23) | 1.11 (0.82–1.53) | 1.36 (1.05–1.87) | 2.23 (1.53–3.49) | <.001 |
| TC (mmol/L) | 4.93 ± 1.15 | 4.83 ± 1.00 | 4.91 ± 1.03 | 4.90 ± 1.16 | 5.08 ± 1.35 | <.001 |
| LDL (mmol/L) | 2.32 ± 0.87 | 2.19 ± 0.86 | 2.32 ± 0.87 | 2.40 ± 0.83 | 2.39 ± 0.91 | <.001 |
| Hdl (mmol/L) | 1.55 ± 0.39 | 1.59 ± 0.39 | 1.55 ± 0.39 | 1.54 ± 0.39 | 1.51 ± 0.40 | <.001 |
| BMI (kg/m2) | 25.10 ± 3.47 | 23.25 ± 3.09 | 24.59 ± 3.23 | 25.53 ± 3.11 | 27.05 ± 3.29 | <.001 |
| WC (cm) | 86.49 ± 9.97 | 81.57 ± 10.20 | 85.68 ± 9.25 | 87.68 ± 9.03 | 91.02 ± 8.93 | <.001 |
| SBP (mm Hg) | 128.95 ± 20.08 | 121.73 ± 18.73 | 128.26 ± 19.84 | 130.92 ± 19.24 | 134.89 ± 20.16 | <.001 |
| DBP (mm Hg) | 82.94 ± 11.47 | 78.20 ± 10.24 | 81.94 ± 10.75 | 84.22 ± 10.95 | 87.42 ± 11.85 | <.001 |
| cMets-Z | ‐3.44 ± 0.72 | ‐4.38 ± 0.51 | ‐3.61 ± 0.14 | ‐3.17 ± 0.12 | ‐2.59 ± 0.31 | <.001 |
| Marital status (married, N [%]) | 45,830 (95.1) | 11,335 (93.8) | 11,303 (94.5) | 11,556 (95.5) | 11,636 (96.5) | <.001 |
| Current drinker, N (%) | 8582 (17.8) | 1187 (9.82) | 2139 (17.9) | 2385 (19.7) | 5527 (48.8) | <.001 |
| Current smoker, N (%) | 14,698 (30.5) | 1922 (15.9) | 3742 (31.3) | 4209 (34.8) | 4825 (40.0) | <.001 |
| Reported income (≥1000¥, N [%]) | 3306 (6.86) | 1007 (8.33) | 821 (6.87) | 712 (5.88) | 766 (6.35) | <.001 |
| Educational background (middle school or above, N [%]) | 3582 (7.43) | 1253 (10.04) | 843 (7.05) | 749 (6.19) | 737 (6.11) | <.001 |
| Salt intake high (>10 g/d, N [%]) | 5081 (10.5) | 1085 (8.98) | 1280 (10.7) | 1273 (10.5) | 1443 (12.0) | <.001 |
| Sedentary lifestyle (>8 h/d, N [%]) | 1521 (3.16) | 486 (4.02) | 361 (3.02) | 319 (2.64) | 355 (2.94) | <.001 |
| Physical exercise (regularly, N [%]) | 6923 (14.4) | 2126 (17.6) | 2043 (17.1) | 1530 (12.7) | 1224 (10.2) | <.001 |
| High-fat diet, (regularly, N [%]) | 4081 (8.56) | 1190 (10.00) | 1088 (9.22) | 973 (8.10) | 830 (6.95) | <.001 |
| Tea consumption (regularly, N [%]) | 11,319 (23.8) | 2446 (20.6) | 2837 (24.0) | 2898 (24.1) | 3138 (26.3) | <.001 |
| Mets | 11,776 (24.4) | 1282 (10.6) | 2101 (17.6) | 2866 (23.7) | 5527 (45.8) | <.001 |
| HBV infection, N (%) | 1290 (2.68) | 308 (2.55) | 355 (2.97) | 351 (2.90) | 276 (2.29) | .007 |
| Live cirrhosis, N (%) | 42 (0.09) | 18 (0.15) | 9 (0.08) | 12 (0.10) | 3 (0.02) | .026 |
| Gallstone disease, N (%) | 915 (1.90) | 222 (1.84) | 259 (2.17) | 230 (1.90) | 204 (1.69) | .106 |
| NAFLD, N (%) | 12,640 (26.2) | 1664 (13.8) | 2492 (20.8) | 3326 (27.5) | 5158 (42.8) | <.001 |
| Family history of cancer, N (%) | 1884(3.95) | 567 (4.77) | 511 (4.33) | 402 (3.35) | 404 (3.38) | <.001 |
| lipid-lowering drugs, N (%) | 470 (0.98) | 121 (1.00) | 119 (1.00) | 108 (0.89) | 122 (1.01) | <.001 |
| Antihypertensive drugs, N (%) | 4819 (10.0) | 907 (7.51) | 1267 (10.6) | 1275 (10.5) | 1370 (11.4) | <.001 |
| Antidiabetic drugs, N (%) | 1045 (2.17) | 225 (1.86) | 292 (2.44) | 276 (2.28) | 252 (2.09) | .168 |
BMI = body mass index, cMetS-Z = cumulative metabolic syndrome severity z score, CRP = C-reactive protein, DBP = diastolic blood pressure, FBG = fasting blood glucose, HBV = hepatitis B virus, HDL = high-density lipoprotein, LDL = low-density lipoprotein, MetS = metabolic syndrome, NAFLD = nonalcoholic fatty liver disease, SBP = systolic blood pressure, TC = total cholesterol, TG = triglyceride, WC = waist circumference.
3.2. Cumulative incidence and log-rank test of subjects
The cumulative incidence of GI cancers in groups Q1 to Q4 increased gradually from group Q1 to group Q4, and the difference by log-rank test was statistically significant (P < .05) (Fig. 2).
Figure 2.
The cumulative incidence of GI cancers in different groups. GI = gastrointestinal.
3.3. The associations between MetS-Z and the risk of GI cancers
We examined the association between MetS-Z and incident GI cancers using Cox proportional hazards models (Table 2). In the crude model (model 1), risk increased across Q2 to Q4 versus Q1. After adjusting for age and sex (model 2) and then additionally for TG, ALT, hsCRP, marital status, alcohol intake, smoking, household income, education, high-salt diet, sedentary behavior, physical activity, high-fat diet, tea consumption, and use of glucose-lowering, lipid-lowering, and antihypertensive medications (model 3), the hazard ratios (95% confidence intervals) for Q2 to Q4 were 1.19 (0.92–1.47), 1.36 (1.08–1.71), and 1.72 (1.33–2.23), respectively (Q1 as reference). Findings were robust in sensitivity analyses excluding participants on antihypertensive, glucose-lowering, or lipid-lowering therapy at baseline (model 4; n = 5567 excluded) and excluding cancers diagnosed within the first 5 years of follow-up (model 5; n = 56 excluded), which produced estimates similar to model 3.
Table 2.
The HRs and 95% CIs of GI cancers based on different cMetS-Z.
| Group | Q1 | Q2 | Q3 | Q4 |
|---|---|---|---|---|
| Case/total | 154/12085 | 193/11958 | 194/12099 | 208/12063 |
| Model 1 | Ref. | 1.27 (1.03–1.57) | 1.29 (1.04–1.59) | 1.37 (1.12–1.69) |
| Model 2 | Ref. | 1.25 (1.01–1.55) | 1.42 (1.15–1.76)) | 1.75 (1.42–2.16)) |
| Model 3 | Ref. | 1.19 (0.96–1.48) | 1.36 (1.08–1.71) | 1.72 (1.33–2.23) |
| Model 4 | Ref. | 1.15 (0.92–1.47) | 1.16 (0.92–1.49) | 1.16 (0.88–1.53) |
| Model 5 | Ref. | 1.13 (0.91–1.41) | 1.11 (0.88–1.40) | 1.17 (0.90–1.53) |
Model 1 was nonadjusted model.
Model 2 adjusted for age and sex.
Model 3 further adjusted for family income, educational background, marital status, salt consumption, current smoker, current drinker, tea consumption, high-fat diet, sedentary lifestyle, physical activity, antihypertensive drug use, antidiabetic drug use, lipid-lowering drug use, and diastolic blood pressure.
Model 4 excluded participants using antihypertensive drugs, antidiabetic drugs, and lipid-lowering drugs.
Model 5 excluded participants who had arterial stiffness within the first 5 years of the follow-up.
Bold values are statistically significant.
CI = confidence interval, cMetS-Z = cumulative metabolic syndrome severity z score, GI = gastrointestinal, HR = hazard ratio.
3.4. RCS curve of cMets-Z in the risk of malignant tumors of the digestive system
To further explore the relationship between cMets-Z and GI cancers, we used cMets-Z as a continuous variable and plotted a restricted cube bar plot (Fig. 3). The risk of GI cancers increased with the increase of cMets-Z, and the cutoff for the increased risk of GI cancers was ‐3.38.
Figure 3.
The risk of GI cancers based on cMetS-Z values. cMetS-Z = cumulative metabolic syndrome severity z score, GI = gastrointestinal.
3.5. Risk analysis of specific sites of malignant tumors of the digestive system
In the analysis of tumor-specific sites, subjects in the Q4 group at all sites showed the highest risk for group 4. Except for gastric cancer and colorectal cancer, there was no statistical association (Table 3).
Table 3.
The association of MetS-Z with the risk of specific sites of GI cancers.
| Cancer type | Cases/participants | Q1 | Q2 | Q3 | Q4 |
|---|---|---|---|---|---|
| Esophageal cancer | 76/48,205 | Ref. | 1.36 (0.62–2.99) | 1.77 (0.78–4.00) | 2.36 (0.94–5.97) |
| Stomach cancer | 152/48,205 | Ref. | 1.66 (1.00–2.78) | 1.71 (1.03–3.04) | 2.36 (1.30–4.28) |
| Small intestine cancer | 16/48,205 | Ref. | 0.11 (0.01–0.93) | 0.22 (0.04–1.19) | 0.98 (0.21–4.50) |
| Colorectal cancer | 284/48,205 | Ref. | 1.07 (0.75–1.52) | 1.34 (0.93–1.92) | 1.50 (1.00–2.30) |
| Liver cancer* | 151/48,205 | Ref. | 1.10 (0.68–1.77) | 1.25 (0.76–2.05) | 1.49 (0.83–2.68) |
| Gallbladder or extrahepatic bile duct cancer† | 21/48,205 | Ref. | 1.47 (0.36–5.98) | 1.49 (0.41–5.46) | 2.01 (0.43–9.48) |
| Pancreatic cancer | 49/93,928 | Ref. | 0.62 (0.23–1.66) | 1.06 (0.49–2.32) | 1.15 (0.40–3.36) |
All models were adjusted for age, sex, family income, educational background, marital status, salt consumption, current smoker, current drinker, tea consumption, high-fat diet, sedentary lifestyle, physical activity, antihypertensive drug use, antidiabetic drug use, lipid-lowering drug use, and diastolic blood pressure.
Bold values are statistically significant.
cMetS-Z = cumulative metabolic syndrome severity z score, GI = gastrointestinal.
Further adjusted for liver cirrhosis and fatty liver disease.
Further adjusted for gallstone disease.
3.6. Stratified risk analysis of malignant tumors of the digestive system
Subgroup analysis was performed based on age, sex, hypertension, diabetes, hyperlipidemia, BMI, MetS, and MetS composition. Overall, participants in the Q4 group had the highest risk, a trend that was observed in all subgroup analyses except those with hypertension and those with a MetS degree = 1. In conclusion, Mets-Z can be used to predict the risk of malignant tumors of the digestive system (Fig. 4).
Figure 4.
Subgroup analysis about HRs and 95%CIs of GI cancers based on different cMetS-Z. BMI = body mass index, CI = confidence interval, cMetS-Z = cumulative metabolic syndrome severity z score, DM = diabetes mellitus, GI = gastrointestinal, HLP = hyperlipidemia, HR= hazard ratio, HT = hypertension, MetS = metabolic syndrome.
4. Discussion
Our results indicate that a higher continuous MetS-Z (cMetS-Z) is independently associated with an increased risk of GI cancers after multivariable adjustment. In stratified analyses, the association was evident across subgroups and appeared particularly pronounced among men and among participants with established MetS. Dose–response modeling using restricted cubic splines supported an approximately linear increase in risk with rising cMetS-Z.
Most prior work has computed MetS-Z using WC; in this study, we used BMI to derive MetS-Z. The rationale is 2-fold. First, both BMI and WC are accepted adiposity indices and correlate with metabolic risk.[23,24] Second, in routine clinical practice in China, BMI is more consistently recorded and more readily obtainable than WC, enabling broader applicability and reduced missingness. Using a BMI-based MetS-Z also facilitates longitudinal tracking from standard health records and may improve generalizability for population screening. We acknowledge that WC can better reflect central (visceral) adiposity, which is biologically relevant to GI carcinogenesis; therefore, future work should compare BMI- and WC-based MetS-Z directly in Chinese cohorts and evaluate whether incorporating both measures (when available) further refines risk prediction.
Individuals with obesity frequently have comorbid hypertension and diabetes. In our cohort, higher BMI was likewise associated with a greater prevalence of hypertension, diabetes, hyperlipidemia, and MetS, consistent with prior reports on obesity.[25,26] A growing literature indicates that MetS-Z improves prediction of future disease risk: Jackson et al showed that higher MetS-Z strongly forecast incident diabetes[27]; elevated MetS-Z was linked to an increased risk of chronic kidney disease in a study using WC as the adiposity metric[28]; and MetS-Z was associated with atherosclerosis in work that used BMI as the obesity indicator.[29] Together, these findings support the clinical utility of MetS-Z – and, by extension, its applicability to Chinese populations – for risk stratification and early prevention.
Our study demonstrates that higher continuous MetS-Z (cMetS-Z) is associated with an increased risk of digestive system cancers. While prior work has not directly evaluated cMetS-Z and cancer, numerous studies link MetS itself to malignancy risk. Our previous research showed that MetS is associated with the development of colorectal and liver cancers, with greater risk at earlier ages of onset.[30] In the UK Biobank, a prospective cohort of 366,046 participants, MetS correlated positively with digestive cancer risk in both men and women.[17] A large matched cohort likewise found higher cancer risk among younger and middle-aged adults (<65 years) with MetS compared with older adults (≥65 years).[18] Not all evidence is uniform, however: in a study, MetS was not associated with increased overall cancer incidence.[19]
Our findings differ from previous studies describing MetS-Z with a mean of 0 and a standard deviation of 1 for MetS-Z[31]. The mean BMI in the Q1 group in our population was less than normal, and the cMets-Z was <0 in all groups. In our study, participants in the Q2 group had a normal mean BMI, but further analysis still showed a significantly increased risk of digestive malignancy compared with the Q1 group, where cMets-Z was the lowest. Subgroup analyses also showed an increased risk of digestive malignancy in participants who did not meet the diagnostic criteria for MetS and had normal blood glucose, lipids, and blood pressure. Even though the BMI was higher in the 4th group, with the exception of small bowel cancer, the risk of developing malignancies of the digestive system increased as the BMI increased. According to a report by the World Cancer Research Fund and the International Agency for Research on Cancer, high levels of BMI increase the risk of breast cancer (especially in postmenopausal women), ovarian cancer, prostate cancer, renal cell carcinoma, and GI-related cancers.[32] Of course, there are also reports to the contrary, such as Suzuki et al’s[33] Mate analysis results show that obesity in women before menopause can not only reduce the risk of breast cancer but also improve the prognosis of malignant tumors. This may be related to the increase in body fat content when the BMI is normal, such as ascites formation, edema, and so forth, which can overestimate BMI.[34] BMI is an indicator of systemic obesity, and it is not well represented by abdominal obesity and visceral fat deposits. A series of previous studies have suggested that the use of other measurement indicators other than BMI can reflect the degree of obesity of the body, such as bioelectrical impedance analysis, dual-energy X-ray absorptiometry, and so forth.[35,36] The results of a prospective cohort study conducted by Gonzalez et al[12] showed that the “obesity paradox” existed only when BMI was used to assess the level of obesity in cancer patients, but not when bioelectrical impedance analysis was used. In our study, different outcomes were found in women, participants with diabetes, hyperlipidemia, hypertension, and those who did not meet the diagnostic criteria for MetS, but the risk was still highest in the 4th group. These studies suggest that focusing on only 1 risk factor for digestive malignancy may underestimate the true risk, so the role of MetS-Z in predicting the risk of GI cancers in Chinese should be highlighted. We also found that participants with cMetS-Z > ‐3.38 had a further increased risk of developing GI cancers. This may suggest that all participants should have a regular MetS score. However, there are fewer studies on MetS-Z in Chinese, so more research is necessary to confirm this finding. Considering our findings in these subjects, we believe that comorbidities in clinical studies may require more attention. In the future, we will conduct further analysis in subjects with metabolic disease and hypertension.
To our knowledge, this is the 1st study to evaluate continuous MetS-Z (cMetS-Z) as a predictor of digestive system malignancies. Leveraging a large, prospective cohort, the design reduces selection bias relative to case-control studies. Additional strengths include comprehensive adjustment for potential confounders and near-complete follow-up (~100%) over up to 12 years, achieved via biennial health examinations and annual case ascertainment through multiple healthcare systems (death certificates, medical records, and health-insurance records). Nonetheless, several limitations merit consideration. First, the cohort’s occupational composition (Kailuan’s industrial workforce) results in a male-predominant sample, which may introduce sex-related bias; future analyses focusing on women in this cohort are warranted. Second, as a single regional cohort from northern China, generalizability to other Chinese populations – and to non-Chinese populations – remains to be established and will require external validation in diverse settings.
5. Conclusion
In conclusion, higher cMetS-Z was independently associated with an increased risk of GI cancers. MetS-Z may serve as a practical, prospective tool for risk prediction and stratification in GI malignancies.
Acknowledgments
I would like to thank the Kailuan Institute staff and participants for their important contributions.
Author contributions
Conceptualization: Wanchao Wang, Kuan Liu.
Methodology: Kuan Liu.
Project administration: Kuan Liu.
Software: Kuan Liu, Yang Zhang, Jiaxing Li.
Validation: Kuan Liu, Chao Ma.
Visualization: Kuan Liu, Yang Zhang, Chao Ma.
Formal analysis: Yang Zhang.
Supervision: Yang Zhang, Jiaxing Li, Chao Ma, Shouling Wu.
Resources: Jiaxing Li.
Data curation: Shouling Wu.
Writing – original draft: Wanchao Wang, Kuan Liu.
Writing – review & editing: Wanchao Wang, Kuan Liu, Siqing Liu.
Abbreviations:
- BMI
- body mass index
- cMets-Z
- cumulative metabolic syndrome severity z score
- DBP
- diastolic blood pressure
- FBG
- fasting blood glucose
- GI
- gastrointestinal
- HDL
- high-density lipoprotein
- HLP
- hyperlipidemia
- hsCRP
- high-sensitivity C-reactive protein
- HTN
- hypertension
- LDL
- low-density lipoprotein
- MetS
- metabolic syndrome
- Mets-Z
- metabolic syndrome severity z score
- SBP
- systolic blood pressure
- TG
- triglyceride
- WC
- waist circumference.
The authors have no funding and conflicts of interest to disclose.
The project was approved by the Ethics Committee of Kailuan Hospital, batch number: 200605, registration number: ChiCTR-TNC-11001489, and all subjects signed the informed consent form.
SL and SW contributed to this article equally.
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
How to cite this article: Wang W, Liu K, Zhang Y, Li J, Ma C, Wu S, Liu S. BMI-based metabolic syndrome Z-score and gastrointestinal cancers: A cohort study. Medicine 2026;105:26(e49460).
WW and KL contributed to this article equally.
Contributor Information
Wanchao Wang, Email: 758869521@qq.com.
Kuan Liu, Email: siqingliu@163.com.
Yang Zhang, Email: 2246036376@qq.com.
Jiaxing Li, Email: 1969635777@qq.com.
Chao Ma, Email: 1004565691@qq.com.
Shouling Wu, Email: drwusl@163.com.
References
- [1].Siegel RL, Miller KD, Wagle NS, Jemal A. Cancer statistics, 2023. CA Cancer J Clin. 2023;73:17–48. [DOI] [PubMed] [Google Scholar]
- [2].Xia C, Dong X, Li H, et al. Cancer statistics in China and United States, 2022: profiles, trends, and determinants. Chin Med J (Engl). 2022;135:584–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Qin SS, Pan GQ, Meng QB, Liu JB, Tian ZY, Luan SJ. The causal relationship between metabolic factors, drinking, smoking and intrahepatic cholangiocarcinoma: a Mendelian randomization study. Front Oncol. 2023;13:1203685. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Heikkila K, Ebrahim S, Lawlor DA. A systematic review of the association between circulating concentrations of C reactive protein and cancer. J Epidemiol Community Health. 2007;61:824–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [5].Hu K, Staiano AE. Trends in obesity prevalence among children and adolescents aged 2 to 19 years in the US from 2011 to 2020. JAMA Pediatr. 2022;176:1037–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Chen K, Shen Z, Gu W, et al. ; Meinian Investigator Group. Prevalence of obesity and associated complications in China: a cross-sectional, real-world study in 15.8 million adults. Diabetes Obes Metab. 2023;25:3390–9. [DOI] [PubMed] [Google Scholar]
- [7].Després JP, Lemieux I. Abdominal obesity and metabolic syndrome. Nature. 2006;444:881–7. [DOI] [PubMed] [Google Scholar]
- [8].Abdelaal M, le Roux CW, Docherty NG. Morbidity and mortality associated with obesity. Ann Transl Med. 2017;5:161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Zhang X, Ha S, Lau HC, Yu J. Excess body weight: novel insights into its roles in obesity comorbidities. Semin Cancer Biol. 2023;92:16–27. [DOI] [PubMed] [Google Scholar]
- [10].Maffetone PB, Rivera-Dominguez I, Laursen PB. Overfat adults and children in developed countries: the public health importance of identifying excess body fat. Front Public Health. 2017;5:190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Stokes A, Collins JM, Grant BF, et al. Prevalence and determinants of engagement with obesity care in the United States. Obesity (Silver Spring). 2018;26:814–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Gonzalez MC, Pastore CA, Orlandi SP, Heymsfield SB. Obesity paradox in cancer: new insights provided by body composition. Am J Clin Nutr. 2014;99:999–1005. [DOI] [PubMed] [Google Scholar]
- [13].DeBoer MD, Filipp SL, Gurka MJ. Use of a metabolic syndrome severity Z score to track risk during treatment of prediabetes: an analysis of the diabetes prevention program. Diabetes Care. 2018;41:2421–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Wu S, Huang Z, Yang X, et al. Prevalence of ideal cardiovascular health and its relationship with the 4-year cardiovascular events in a northern Chinese industrial city. Circ Cardiovasc Qual Outcomes. 2012;5:487–93. [DOI] [PubMed] [Google Scholar]
- [15].Esterson YB, Grimaldi GM. Radiologic imaging in nonalcoholic fatty liver disease and nonalcoholic steatohepatitis. Clin Liver Dis. 2018;22:93–108. [DOI] [PubMed] [Google Scholar]
- [16].Saverymuttu SH, Joseph AE, Maxwell JD. Ultrasound scanning in the detection of hepatic fibrosis and steatosis. Br Med J (Clin Res Ed). 1986;292:13–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Liu T, Song C, Zhang Y, et al. Hepatitis B virus infection and the risk of gastrointestinal cancers among Chinese population: a prospective cohort study. Int J Cancer. 2022;150:1018–28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].[National guideline for hypertension management in China (2019)]. Zhonghua Xin Xue Guan Bing Za Zhi. 2020;48:10–46. [DOI] [PubMed] [Google Scholar]
- [19].2. Classification and diagnosis of diabetes: standards of medical care in diabetes-2019. Diabetes Care. 2019;42(Suppl 1):S13–s28. [DOI] [PubMed] [Google Scholar]
- [20].[2016 Chinese guideline for the management of dyslipidemia in adults]. Zhonghua Xin Xue Guan Bing Za Zhi. 2016;44:833–53. [DOI] [PubMed] [Google Scholar]
- [21].Grundy SM, Cleeman JI, Daniels SR, et al. ; American Heart Association. Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation. 2005;112:2735–52. [DOI] [PubMed] [Google Scholar]
- [22].Gurka MJ, Filipp SL, Musani SK, Sims M, DeBoer MD. Use of BMI as the marker of adiposity in a metabolic syndrome severity score: derivation and validation in predicting long-term disease outcomes. Metabolism. 2018;83:68–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Ross R, Neeland IJ, Yamashita S, et al. Waist circumference as a vital sign in clinical practice: a consensus statement from the IAS and ICCR working group on visceral obesity. Nat Rev Endocrinol. 2020;16:177–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Lee WS. Body fatness charts based on BMI and waist circumference. Obesity (Silver Spring). 2016;24:245–9. [DOI] [PubMed] [Google Scholar]
- [25].Piché ME, Tchernof A, Després JP. Obesity phenotypes, diabetes, and cardiovascular diseases. Circ Res. 2020;126:1477–500. [DOI] [PubMed] [Google Scholar]
- [26].Wiley JF, Carrington MJ. A metabolic syndrome severity score: a tool to quantify cardio-metabolic risk factors. Prev Med. 2016;88:189–95. [DOI] [PubMed] [Google Scholar]
- [27].Gurka MJ, Golden SH, Musani SK, et al. Independent associations between a metabolic syndrome severity score and future diabetes by sex and race: the atherosclerosis risk in communities study and Jackson Heart Study. Diabetologia. 2017;60:1261–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [28].Wu M, Shu Y, Wang L, et al. Metabolic syndrome severity score and the progression of CKD. Eur J Clin Invest. 2022;52:e13646. [DOI] [PubMed] [Google Scholar]
- [29].Wang M, Wang C, Zhao M, Wu S, Xue H, Liu H. BMI-based metabolic syndrome severity score and arterial stiffness in a cohort Chinese study. Nutr Metabol Cardiovascular Diseases. 2024;34:1761–8. [DOI] [PubMed] [Google Scholar]
- [30].Liu T, Liu CA, Zhang QS, et al. Association of the age of onset of metabolic syndrome with the risk of all cancer types. Diabetes Metab Syndr. 2023;17:102896. [DOI] [PubMed] [Google Scholar]
- [31].Amouzegar A, Honarvar M, Masoumi S, Khalili D, Azizi F, Mehran L. Trajectory patterns of metabolic syndrome severity score and risk of type 2 diabetes. J Transl Med. 2023;21:750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [32].Rotondi M, Castagna MG, Cappelli C, et al. Obesity does not modify the risk of differentiated thyroid cancer in a cytological series of thyroid nodules. Eur Thyroid J. 2016;5:125–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [33].Suzuki R, Orsini N, Saji S, Key TJ, Wolk A. Body weight and incidence of breast cancer defined by estrogen and progesterone receptor status – a meta-analysis. Int J Cancer. 2009;124:698–712. [DOI] [PubMed] [Google Scholar]
- [34].Kim S, Kyung C, Park JS, et al. Normal-weight obesity is associated with increased risk of subclinical atherosclerosis. Cardiovasc Diabetol. 2015;14:58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Prado CM, Lieffers JR, McCargar LJ, et al. Prevalence and clinical implications of sarcopenic obesity in patients with solid tumours of the respiratory and gastrointestinal tracts: a population-based study. Lancet Oncol. 2008;9:629–35. [DOI] [PubMed] [Google Scholar]
- [36].Hakimi AA, Furberg H, Zabor EC, et al. An epidemiologic and genomic investigation into the obesity paradox in renal cell carcinoma. J Natl Cancer Inst. 2013;105:1862–70. [DOI] [PMC free article] [PubMed] [Google Scholar]