Abstract
Objective: To elucidate the relationships between metabolic syndrome (MetS)-related factors and all-cause mortality over a 28-year follow-up in O City, Ehime Prefecture, Japan.
Materials and Methods: Using the Basic Resident Registration System, we analyzed data from 3,931 participants aged 40–74 years (1,360 men and 2,571 women) who underwent annual medical examinations based on the Geriatric Health Care Law between 1996 and 1998. Cox proportional hazards regression was used to assess associations between all-cause mortality and nine variables, adjusted for sex and age: MetS; body mass index (BMI: underweight, <18.5 kg/m2; normal weight, 18.5 to <25.0 kg/m2; overweight, ≥25.0 kg/m2); height; waist circumference abnormality; elevated blood pressure/hypertension; dyslipidemia; prediabetes/diabetes; alcohol consumption; and smoking.
Results: Of the participants, 1,938 died. In both men and women, MetS was not associated with all-cause mortality. Underweight status (hazard ratio [HR]=1.93, 95% confidence interval [CI]: 1.43–2.60 for men; HR=1.51, 95% CI: 1.16–1.97 for women), prediabetes/diabetes (HR=1.35, 95% CI: 1.13–1.62 for men; HR=1.36, 95% CI: 1.12–1.66 for women), and smoking (HR=1.45, 95% CI: 1.25–1.67 for men; HR=1.95, 95% CI: 1.48–2.58 for women) were significantly associated with all-cause mortality.
Conclusion: Avoiding smoking, underweight status, and prediabetes/diabetes may help extend life expectancy.
Keywords: lifestyle factors related to disorders, metabolic syndrome, all-cause mortality, cohort survey, rural Japan
Introduction
Recently, cancer was reported as the leading cause of death in Japan, followed by cardiovascular disease. Cancer accounts for 24.3% of all deaths in Japan, whereas cardiovascular disease accounts for 14.7%1). The major causes of both cancer and cardiovascular disease are lifestyle related. Since 2008, local governments in Japan have recommended specific health examinations to detect metabolic syndrome (MetS) and provided health guidance focusing on its prevention. MetS is diagnosed by the presence of abdominal obesity, defined as a high waist circumference, and at least two of the following risk factors: (1) elevated blood pressure/hypertension, (2) dyslipidemia, and (3) prediabetes/diabetes2). Because central obesity is the principal diagnostic criterion for MetS, waist circumference is routinely measured as part of these health examinations.
Currently, MetS has been reported to be associated with an increased risk of developing various diseases. Van Hoang et al.3) conducted a follow-up study on severe cancer events through March 31, 2020, among 70,875 workers (59,950 men and 10,925 women) aged 20–59 years who had undergone health check-ups in 2011 and 2014. Van Hoang et al.3) reported that in Japanese workers, MetS was associated with an increased risk of severe cancer events, particularly those linked to obesity.
In the O City Cohort Study, Saito et al.4) carried out an 18.6-year follow-up of 4,068 participants who had undergone baseline medical health examinations in O City between 1996 and 1998. Of these, 3,969 participants who had not experienced a stroke at baseline were followed from the initial examination period through December 2018. Saito et al.4) reported a weak association between MetS and total stroke, hemorrhagic stroke, and cerebral infarction, concluding that although the relationship was not statistically significant, MetS might have a minor effect on stroke incidence. However, in Japan, few long-term follow-up studies have examined the relationships between all-cause mortality and MetS, along with its associated risk factors such as body mass index (BMI), height, alcohol consumption, and smoking.
The objective of the present study was to clarify the relationships between MetS and its related factors—waist circumference abnormality, elevated blood pressure/hypertension, dyslipidemia, prediabetes/diabetes, BMI, height, alcohol consumption, and smoking—and all-cause mortality, based on the results of a 28-year follow-up (1996–2024) of residents in O City, Ehime Prefecture, Japan, who had undergone detailed health examinations, including waist circumference measurement.
Materials and Methods
Ethical considerations pertaining to privacy issues and data access
Written informed consent was obtained from all individuals before using any confidential information from medical examinations or health-related questionnaires, in accordance with the Japanese Privacy Protection Law. Approval to conduct this epidemiological study was granted by the Research Ethics Committee of the Faculty of Education, Ehime University (approval No. R4-34-1), which also authorized access to electronically stored personal data from the Basic Resident Registration System maintained by the O City local government.
Survey area
O City is located in the southwestern part of Ehime Prefecture and has a population of approximately 50,000. The population increased following the municipal merger of O City with two neighboring towns and one village in January 2005. O City comprises mountainous, urban, and coastal areas, and its industrial structure includes primary industries such as agriculture and fishery, as well as manufacturing and commerce. There are currently four community health centers in O City: one located in the area of the original O City and three in the districts incorporated during the 2005 merger. These centers provide health management, community health, and nursing care services to residents.
Survey participants
The O City Cohort I Study was established between 1996 and 1998. Questionnaire surveys and medical health examinations were conducted, and participants were followed until December 31, 2024. Participants in O City Cohort I represented approximately 46.3% (4,972 individuals aged 29–94 years) of the 10,730 people aged 20–94 years who were enrolled in the 1996 National Health Insurance scheme in the former O City.
The analyzed cohort comprised 3,957 residents of O City (1,370 men and 2,587 women) who underwent annual medical examinations under the Geriatric Health Care Law in the former O City between 1996 and 1998. In accordance with the eligibility criteria requiring specific medical examinations to detect MetS and provide health guidance, the target age range for survey participants was defined as 40–74 years.
Among individuals aged 40–74 years covered by the 1996 National Health Insurance scheme in the former O City (n=6,408; 3,052 men and 3,356 women), there were 2,172 individuals (1,098 men and 1,074 women) aged 40–49 years, 1,625 (774 men and 851 women) aged 50–59 years, 1,856 (849 men and 1,007 women) aged 60–69 years, and 754 (330 men and 424 women) aged 70–74 years.
The baseline population consisted of 3,931 participants (1,360 men and 2,571 women), after excluding 26 individuals who died within one year of enrollment. These 3,931 participants accounted for approximately 36.6% of all 10,730 individuals aged 20–94 years and 61.3% of all 6,408 individuals aged 40–74 years enrolled in the 1996 National Health Insurance scheme in the former O City.
Baseline data
Baseline data for the 3,931 participants included sex, age, height, weight, blood pressure, serum total cholesterol, high-density lipoprotein (HDL) cholesterol, triglycerides, blood glucose, and waist circumference. BMI5) was calculated to classify participants as underweight (<18.5 kg/m2), normal weight (18.5 to <25.0 kg/m2), or overweight (≥25.0 kg/m2). Blood samples were collected from participants in the fasting state, at least 8 hours after their last meal. Height was included because it has been reported to be associated with mortality from cardiovascular disease, coronary heart disease, and stroke6).
Participants diagnosed with MetS7) met the criterion of a high waist circumference (≥85 cm in men and ≥90 cm in women) plus at least two of the following three abnormal health conditions: (1) elevated blood pressure/hypertension (≥130/85 mmHg), (2) dyslipidemia (triglycerides ≥150 mg/dL or HDL cholesterol <40 mg/dL), and (3) hyperglycemia (fasting glucose ≥110 mg/dL). In cases where blood samples were collected in a non-fasting state, dyslipidemia (triglycerides ≥175 mg/dL)8) and hyperglycemia (glucose ≥140 mg/dL) were used instead. Participants receiving medication for hypertension, dyslipidemia, or diabetes mellitus were classified under the corresponding abnormal health condition category.
Lifestyle habits related to alcohol consumption and smoking were also assessed. Participants were categorized into three groups: never drank alcohol, quit drinking alcohol, and currently drink alcohol. Based on responses to the lifestyle questionnaire on alcohol consumption, participants were further divided into two groups: (1) the drinking group (currently drink alcohol or quit drinking) and (2) the non-drinking group (never drank alcohol). Similarly, smoking status was classified into three categories: never smoked, quit smoking, and currently smoke. Participants were also grouped as (1) the smoking group (currently smoke or quit smoking) and (2) the non-smoking group (never smoked).
Follow-up period
All participants were followed annually until December 31, 2024, and their residency status in O City was verified. Using the Basic Resident Registration System, we confirmed whether each participant was alive, had moved out of the area, or had died. In the case of deceased participants, the follow-up duration (in years and days) was calculated from the date of the medical examination conducted between 1996 and 1998 until the date of death. For those still alive, the follow-up duration was calculated from the date of the medical examination to either December 31, 2024, or the date of relocation from O City, whichever occurred first.
Statistical analysis
First, a joint test was performed to confirm that there was no interaction between age group and MetS.
To assess the statistical significance of relationships between each risk factor and all-cause mortality, differences in means, medians, and proportions were examined using the t-test, Wilcoxon rank-sum test, and χ2 test, respectively. A Cox proportional hazards regression model was then applied to identify significant risk factors, and hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated.
In Model 1, a Cox proportional hazards regression model was used to individually analyze the association between all-cause mortality and nine factors, adjusted for both sex and age or for age alone: (1) MetS, (2) BMI (underweight: <18.5 kg/m2, normal weight: 18.5 to <25.0 kg/m2, overweight: ≥25.0 kg/m2), (3) height, (4) waist circumference abnormality, (5) elevated blood pressure/hypertension, (6) dyslipidemia, (7) prediabetes/diabetes, (8) alcohol consumption, and (9) smoking. Because a confounding relationship existed between waist circumference abnormality and being overweight (BMI ≥25.0 kg/m2), BMI was used as an indicator of physical size in this analysis. In Model 2, a Cox proportional hazards regression model was also used to simultaneously analyze the association between all-cause mortality and seven factors, adjusted for both sex and age or for age alone: (2) BMI (underweight: <18.5 kg/m2, normal weight: 18.5 to <25.0 kg/m2, overweight: ≥25.0 kg/m2), (3) height, (5) elevated blood pressure/hypertension, (6) dyslipidemia, (7) prediabetes/diabetes, (8) alcohol consumption, and (9) smoking—excluding (1) MetS and (4) waist circumference abnormality.
To further examine the relationship between BMI and prediabetes/diabetes, participants were classified into four BMI categories: non-severe thinness (BMI ≥16.0 kg/m2), non-moderate thinness (BMI ≥17.0 kg/m2), non-underweight (BMI ≥18.5 kg/m2), and overweight (BMI ≥25.0 kg/m2). The association between BMI and prediabetes/diabetes, along with all-cause mortality and five additional factors adjusted for age—(3) height, (5) elevated blood pressure/hypertension, (6) dyslipidemia, (8) alcohol consumption, and (9) smoking—was analyzed simultaneously.
Statistical significance was set at P<0.05. All statistical analyses were performed using SAS software (version 9.4; SAS Institute Inc., Cary, NC, USA).
Results
Of the 3,931 participants, data were obtained for 1,938 individuals who died during the 28-year follow-up period.
For the joint test, participants aged 40–74 years were divided into 5-year age groups (40–44, 45–49, 50–54, 55–59, 60–64, 65–69, and 70–74 years) by sex. The joint test assessing the interaction between 5-year age groups and MetS showed no significant interaction in either men (P=0.59) or women (P=0.15). Therefore, the hazard ratio was confirmed to be consistent across age groups.
Table 1 presents the baseline characteristics of male and female participants according to survival status. In the univariate analysis among men, significant associations were observed between all-cause mortality and age, underweight, height, systolic blood pressure, triglycerides, blood glucose, elevated blood pressure/hypertension, prediabetes/diabetes, alcohol consumption, and smoking. Among women, significant associations were found between all-cause mortality and age, BMI, underweight, normal weight, overweight, height, systolic and diastolic blood pressure, HDL cholesterol, triglycerides, blood glucose, MetS, waist circumference abnormality, elevated blood pressure/hypertension, dyslipidemia, prediabetes/diabetes, and alcohol consumption.
Table 1. Baseline characteristics of male and female participants who survived or died.
| Characteristic | Study participants, n=3,931 | ||||||
|---|---|---|---|---|---|---|---|
| Male participants, n=1,360 | Female participants, n=2,571 | ||||||
| All-cause mortality | Survivors | P-value | All-cause mortality | Survivors | P-value | ||
| Participants, number (%) | 867 (63.8) | 493 (36.3) | 1,071 (41.7) | 1,500 (58.3) | |||
| Age, years, mean (SD) | 66.0 ± 6.3 | 55.4 ± 8.9 | <0.001 | 65.8 ± 6.3 | 55.9 ± 8.6 | <0.001 | |
| Body mass index, kg/m2, mean (SD) | 23.1 ± 2.9 | 23.5 ± 2.7 | 0.011 | 23.4 ± 3.3 | 23.2 ± 3.0 | 0.276 | |
| Body weight categoriesa | |||||||
| Underweight, number (%) | 48 (5.5) | 13 (2.6) | 0.013 | 60 (5.6) | 50 (3.3) | 0.005 | |
| Normal weight, number (%) | 595 (68.6) | 340 (69.0) | 0.897 | 696 (65.0) | 1,064 (70.9) | 0.001 | |
| Overweight, number (%) | 224 (25.8) | 140 (28.4) | 0.305 | 315 (29.4) | 386 (25.7) | 0.039 | |
| Height, cm, mean (SD) | 159.8 ± 5.9 | 162.8 ± 6.5 | <0.001 | 147.4 ± 5.5 | 151.0± 5.6 | <0.001 | |
| Systolic blood pressure, mmHg, mean (SD) | 135.8 ± 17.3 | 128.7 ± 16.4 | <0.001 | 135.2 ± 18.0 | 126.8 ± 17.6 | <0.001 | |
| Diastolic blood pressure, mmHg, mean (SD) | 79.2 ± 10.6 | 79.8 ± 10.7 | 0.381 | 77.4 ± 10.2 | 75.9 ± 10.4 | <0.001 | |
| HDL Cholesterol, mg/dL, mean (SD) | 56.2 ± 15.8 | 56.5 ± 15.8 | 0.783 | 59.3 ± 16.0 | 62.2 ± 15.5 | <0.001 | |
| Triglyceride, mg/dL, median (IQR) | 105 (74–150) | 115 (79–157) | 0.022 | 111 (80–156) | 98 (71–141) | <0.001 | |
| Blood glucose, mg/dL, median (IQR) | 98 (90–112) | 95 (89–108) | 0.002 | 97 (90–107) | 94 (88–104) | <0.001 | |
| Metabolic syndrome, number (%) | 147 (17.0) | 66 (13.4) | 0.082 | 83 (7.8) | 59 (3.9) | <0.001 | |
| Components of the metabolic syndrome | |||||||
| Waist circumference abnormality, number (%) | 333 (38.4) | 184 (37.3) | 0.692 | 221 (20.6) | 213 (14.2) | <0.001 | |
| Elevated blood pressure/hypertensionb, number (%) | 634 (73.1) | 272 (55.2) | <0.001 | 768 (71.7) | 735 (49.0) | <0.001 | |
| Dyslipidemiac, number (%) | 256 (29.5) | 154 (31.2) | 0.509 | 286 (26.7) | 304 (20.3) | <0.001 | |
| Prediabetes/diabetesd, number (%) | 154 (17.8) | 63 (12.8) | 0.016 | 111 (10.4) | 95 (6.3) | <0.001 | |
| Lifestyle factors | |||||||
| Alcohol consumption, number (%) | 537 (61.9) | 333 (67.6) | 0.038 | 66 (6.2) | 152 (10.1) | <0.001 | |
| Smoking, number (%) | 571 (65.9) | 293 (59.4) | 0.018 | 55 (5.1) | 58 (3.9) | 0.122 | |
Significant differences (of 5% or less) between means, medians and percentages were calculated using a: t-test, Wilcoxon test, or χ2 test, respectively. aBody weight (BMI) categories were underweight (<18.5 kg/m2), normal weight (18.5 to <25.0 kg/m2), and overweight (≥25.0 kg/m2). bElevated blood pressure/hypertension was defined as systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥85 mm Hg. If participants had been receiving medication for hypertension, they were classified as being within the abnormal category. cDyslipidemia was defined as serum triglycerides ≥150 mg/dL or HDL cholesterol <40 mg/dL. In the case of a non-fasting blood sample, blood triglycerides ≥175 mg/dL was used instead. If participants had been receiving medication for dyslipidemia, they were classified as being within the abnormal category. dPrediabetes/diabetes was defined as fasting glucose ≥110 mg/dL. In the case of a non-fasting blood sample, glucose ≥140 mg/dL was used instead. If participants had been receiving medication for diabetes mellitus, they were classified as being within the abnormal category. BMI: body mass index; HDL: high-density lipoprotein; IQR: interquartile range.
Table 2 presents the associations between nine factors and all-cause mortality, analyzed using Models 1 and 2 of the Cox proportional hazards regression model for men and women combined. In Table 2, for Model 1, the Cox proportional hazards regression model was used to individually analyze the associations between all-cause mortality and nine factors, adjusted for both sex and age: (1) MetS, (2) BMI (underweight: <18.5 kg/m2, normal weight: 18.5 to <25.0 kg/m2, overweight: ≥25.0 kg/m2), (3) height, (4) waist circumference abnormality, (5) elevated blood pressure/hypertension, (6) dyslipidemia, (7) prediabetes/diabetes, (8) alcohol consumption, and (9) smoking. In Model 1, significant associations were observed with underweight (HR=1.69, 95% CI: 1.39–2.06), elevated blood pressure/hypertension (HR=1.10, 95% CI: 1.00–1.22), prediabetes/diabetes (HR=1.36, 95% CI: 1.19–1.55), and smoking (HR=1.54, 95% CI: 1.35–1.75). In Table 2, for Model 2, the Cox proportional hazards regression model was used to simultaneously analyze the associations between all-cause mortality and seven factors, adjusted for both sex and age: (2) BMI (underweight: <18.5 kg/m2, normal weight: 18.5 to <25.0 kg/m2, overweight: ≥25.0 kg/m2), (3) height, (5) elevated blood pressure/hypertension, (6) dyslipidemia, (7) prediabetes/diabetes, (8) alcohol consumption, and (9) smoking-excluding (1) MetS and (4) waist circumference abnormality. In Model 2, significant associations were observed with underweight (HR=1.69, 95% CI: 1.39–2.06), elevated blood pressure/hypertension (HR=1.12, 95% CI: 1.01–1.24), prediabetes/diabetes (HR=1.36, 95% CI: 1.19–1.55), and smoking (HR=1.55, 95% CI: 1.36–1.76).
Table 2. Association between metabolic syndrome and eight related factors with all-cause mortality based on Models 1* and 2† using Cox proportional hazards regression analysis.
| Study participants, n=3,931 | ||||||
|---|---|---|---|---|---|---|
| All-cause mortality, n | Person-years | Model 1* | Model 2† | |||
| HR (95% CI) | HR (95% CI) | |||||
| Category 1 | Metabolic syndrome | No | 1,708 | 80,642.6 | 1.00 | - |
| Yes | 230 | 71,88.2 | 1.12 (0.97–1.29) | - | ||
| Category 2 | Obesitya | Underweight | 108 | 3,350.6 | 1.69 (1.39–2.06) | 1.69 (1.39–2.06) |
| Normal weight | 1,291 | 60,377.1 | 1.00 | 1.00 | ||
| Overweight | 539 | 24,103.3 | 1.02 (0.92–1.13) | 0.99 (0.90–1.10) | ||
| Height | - | - | - | 0.99 (0.98–1.00) | 0.99 (0.98–1.00) | |
| Waist circumference abnormality | No | 1,384 | 67,457.9 | 1.00 | - | |
| Yes | 554 | 20,373.0 | 1.04 (0.94–1.15) | - | ||
| Elevated blood pressure/hypertensionb | No | 536 | 35,940.7 | 1.00 | 1.00 | |
| Yes | 1,402 | 51,890.2 | 1.10 (1.00–1.22) | 1.12 (1.01–1.24) | ||
| Dyslipidemiac | No | 1,396 | 65,913.3 | 1.00 | 1.00 | |
| Yes | 542 | 21,917.6 | 1.03 (0.93–1.14) | 1.02 (0.92–1.13) | ||
| Prediabetes/diabetesd | No | 1,673 | 79,309.2 | 1.00 | 1.00 | |
| Yes | 265 | 8,521.7 | 1.36 (1.19–1.55) | 1.36 (1.19–1.55) | ||
| Alcohol consumption | No | 1,335 | 64,659.8 | 1.00 | 1.00 | |
| Yes | 603 | 23,171.1 | 1.00 (0.89–1.13) | 0.95 (0.84–1.08) | ||
| Smoking | No | 1,312 | 68,372.1 | 1.00 | 1.00 | |
| Yes | 626 | 19,458.8 | 1.54 (1.35–1.75) | 1.55 (1.36–1.76) | ||
*Adjusted nine items by sex and age respectively. †Furthermore adjusted seven items simultaneously for sex and age. aObesity categories were underweight (<18.5 kg/m2), normal weight (18.5 to <25.0 kg/m2), and overweight (≥25.0 kg/m2). bElevated blood pressure/hypertension was defined as systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥85 mm Hg. If participants had been receiving medication for hypertension, they were classified as being within the abnormal category. cDyslipidemia was defined as serum triglycerides ≥150 mg/dL or HDL cholesterol <40 mg/dL. In the case of a non-fasting blood sample, blood triglycerides ≥175 mg/dL was used instead. If participants had been receiving medication for dyslipidemia, they were classified as being within the abnormal category. dPrediabetes/diabetes was defined as fasting glucose ≥110 mg/dL. In the case of a non-fasting blood sample, glucose ≥140 mg/dL was used instead. If participants had been receiving medication for diabetes mellitus, they were classified as being within the abnormal category. CI: confidence interval; HR: hazard ratio
Figures 1-1, 1-2, 1-3, and 1-4 present Kaplan–Meier survival curves illustrating the associations between all-cause mortality and underweight, elevated blood pressure/hypertension, prediabetes/diabetes, and smoking among 2,300 participants aged 40–64 years (688 men and 1,612 women). A proportional hazards relationship was observed between these four factors and all-cause mortality. Log-rank tests also indicated significant differences for (1) underweight (P<0.001), (2) elevated blood pressure/hypertension (P<0.001), (3) prediabetes/diabetes (P<0.001), and (4) smoking (P<0.001).
Figure 1.
Survival probabilities for underweight, elevated blood pressure/hypertension, prediabetes/diabetes, and smoking status among 2,300 individuals (688 men and 1,612 women) aged 40–64 years.
P-value based on the log-rank test.
Figures 2-1, 2-2, 2-3, and 2-4 show Kaplan–Meier survival curves depicting the associations between all-cause mortality and the same four factors among 1,631 participants aged 65 years and older (672 men and 959 women). Log-rank tests revealed significant differences for (1) underweight (P=0.003), (3) prediabetes/diabetes (P<0.001), and (4) smoking (P<0.001). Conversely, (2) elevated blood pressure/hypertension (P=0.126) showed no significant association with all-cause mortality.
Figure 2.
Survival probabilities for underweight, elevated blood pressure/hypertension, prediabetes/diabetes, and smoking status among 1,631 individuals (672 men and 959 women) aged 65 years and older.
P-value based on the log-rank test.
Tables 3 and 4 present the associations between nine factors and all-cause mortality using Models 1 and 2 of the Cox proportional hazards regression analysis, stratified by sex. In Tables 3 and 4, for Model 1, the Cox proportional hazards regression model was used to individually analyze the associations between all-cause mortality and nine factors, adjusted for age: (1) MetS, (2) BMI (underweight: <18.5 kg/m2, normal weight: 18.5 to <25.0 kg/m2, overweight: ≥25.0 kg/m2), (3) height, (4) waist circumference abnormality, (5) elevated blood pressure/hypertension, (6) dyslipidemia, (7) prediabetes/diabetes, (8) alcohol consumption, and (9) smoking. In Model 1, for both men and women, significant associations with all-cause mortality were observed for underweight (men: HR=1.94, 95% CI: 1.44–2.60; women: HR=1.52, 95% CI: 1.17–1.98), prediabetes/diabetes (men: HR=1.33, 95% CI: 1.12–1.59; women: HR=1.40, 95% CI: 1.15–1.70), and smoking (men: HR=1.44, 95% CI: 1.25–1.66; women: HR=1.96, 95% CI: 1.49–2.57). In Tables 3 and 4, for Model 2, the Cox proportional hazards regression model was used to simultaneously analyze the associations between all-cause mortality and seven factors adjusted for age: (2) BMI (underweight: <18.5 kg/m2, normal weight: 18.5 to <25.0 kg/m2, overweight: ≥25.0 kg/m2), (3) height, (5) elevated blood pressure/hypertension, (6) dyslipidemia, (7) prediabetes/diabetes, (8) alcohol consumption, and (9) smoking-excluding (1) MetS and (4) waist circumference abnormality. In Model 2, for both men and women, significant associations with all-cause mortality were again observed for underweight (men: HR=1.93, 95% CI: 1.43–2.60; women: HR=1.51, 95% CI: 1.16–1.97), prediabetes/diabetes (men: HR=1.35, 95% CI: 1.13–1.62; women: HR=1.36, 95% CI: 1.12–1.66), and smoking (men: HR=1.45, 95% CI: 1.25–1.67; women: HR=1.95, 95% CI: 1.48–2.58).
Table 3. Association between metabolic syndrome and eight related factors with all-cause mortality based on Models 1* and 2† using Cox proportional hazards regression analysis in male participants.
| Male participants, n=1,360 | ||||||
|---|---|---|---|---|---|---|
| All-cause mortality, n | Person-years | Model 1* | Model 2† | |||
| HR (95% CI) | HR (95% CI) | |||||
| Category 1 | Metabolic syndrome | No | 720 | 23,377.1 | 1.00 | - |
| Yes | 147 | 4,181.5 | 1.16 (0.97–1.39) | - | ||
| Category 2 | Obesitya | Underweight | 48 | 983.7 | 1.94 (1.44–2.60) | 1.93 (1.43–2.60) |
| Normal weight | 595 | 18,925.0 | 1.00 | 1.00 | ||
| Overweight | 224 | 7,649.9 | 1.03 (0.88–1.20) | 0.99 (0.85–1.17) | ||
| Height | - | - | - | 0.99 (0.98–1.01) | 0.99 (0.98–1.00) | |
| Waist circumference abnormality | No | 534 | 16,999.7 | 1.00 | - | |
| Yes | 333 | 10,559.0 | 1.01 (0.88–1.15) | - | ||
| Elevated blood pressure/hypertensionb | No | 233 | 9,750.1 | 1.00 | 1.00 | |
| Yes | 634 | 17,808.5 | 1.09 (0.93–1.26) | 1.10 (0.94–1.28) | ||
| Dyslipidemiac | No | 611 | 19,163.1 | 1.00 | 1.00 | |
| Yes | 256 | 8,395.6 | 1.06 (0.91–1.22) | 1.04 (0.90–1.21) | ||
| Prediabetes/diabetesd | No | 713 | 23,579.1 | 1.00 | 1.00 | |
| Yes | 154 | 3,979.6 | 1.33 (1.12–1.59) | 1.35 (1.13–1.62) | ||
| Alcohol consumption | No | 330 | 9,681.4 | 1.00 | 1.00 | |
| Yes | 537 | 17,877.3 | 0.98 (0.85–1.12) | 0.94 (0.82–1.08) | ||
| Smoking | No | 296 | 10,549.9 | 1.00 | 1.00 | |
| Yes | 571 | 17,008.8 | 1.44 (1.25–1.66) | 1.45 (1.25–1.67) | ||
* Adjusted nine items by age respectively. †Furthermore adjusted seven items simultaneously by age. aObesity categories were underweight (<18.5 kg/m2), normal weight (18.5 to <25.0 kg/m2), and overweight (≥25.0 kg/m2). bElevated blood pressure/hypertension was defined as systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥85 mm Hg. If participants had been receiving medication for hypertension, they were classified as being within the abnormal category. cDyslipidemia was defined as serum triglycerides ≥150 mg/dL or HDL cholesterol <40 mg/dL. In the case of a non-fasting blood sample, blood triglycerides ≥175 mg/dL was used instead. If participants had been receiving medication for dyslipidemia, they were classified as being within the abnormal category. dPrediabetes/diabetes was defined as fasting glucose ≥110 mg/dL. In the case of a non-fasting blood sample, glucose ≥140 mg/dL was used instead. If participants had been receiving medication for diabetes mellitus, they were classified as being within the abnormal category. CI: confidence interval; HR: hazard ratio.
Table 4. Association between metabolic syndrome and eight related factors with all-cause mortality based on Models 1* and 2† using Cox proportional hazards regression analysis in female participants.
| Female participants, n=2,571 | ||||||
|---|---|---|---|---|---|---|
| All-cause mortality, n | Person-years | Model 1* | Model 2† | |||
| HR (95% CI) | HR (95% CI) | |||||
| Category 1 | Metabolic syndrome | No | 988 | 57,265.5 | 1.00 | - |
| Yes | 83 | 3,006.7 | 1.04 (0.83–1.31) | - | ||
| Category 2 | Obesitya | Underweight | 60 | 2,366.8 | 1.52 (1.17–1.98) | 1.51 (1.16–1.97) |
| Normal weight | 696 | 41,452.1 | 1.00 | 1.00 | ||
| Overweight | 315 | 16,453.3 | 1.01 (0.89–1.16) | 0.99 (0.87–1.14) | ||
| Height | - | - | - | 0.99 (0.98–1.00) | 0.98 (0.97–1.00) | |
| Waist circumference abnormality | No | 850 | 50,458.2 | 1.00 | - | |
| Yes | 221 | 9,814.0 | 1.07 (0.92–1.24) | - | ||
| Elevated blood pressure/hypertensionb | No | 303 | 26,190.6 | 1.00 | 1.00 | |
| Yes | 768 | 34,081.7 | 1.11 (0.97–1.27) | 1.13 (0.98–1.30) | ||
| Dyslipidemiac | No | 785 | 46,750.2 | 1.00 | 1.00 | |
| Yes | 286 | 13,522.0 | 1.00 (0.87–1.14) | 1.00 (0.87–1.15) | ||
| Prediabetes/diabetesd | No | 960 | 55,730.2 | 1.00 | 1.00 | |
| Yes | 111 | 4,542.1 | 1.40 (1.15–1.70) | 1.36 (1.12–1.66) | ||
| Alcohol consumption | No | 1,005 | 54,978.4 | 1.00 | 1.00 | |
| Yes | 66 | 5,293.8 | 1.09 (0.85–1.40) | 1.01 (0.78–1.30) | ||
| Smoking | No | 1,016 | 57,822.2 | 1.00 | 1.00 | |
| Yes | 55 | 2,450.0 | 1.96 (1.49–2.57) | 1.95 (1.48–2.58) | ||
*Adjusted nine items by age respectively. †Furthermore adjusted seven items simultaneously by age. aObesity categories were underweight (<18.5 kg/m2), normal weight (18.5 to <25.0 kg/m2), and overweight (≥25.0 kg/m2). b Elevated blood pressure/hypertension was defined as systolic blood pressure ≥130 mm Hg or diastolic blood pressure ≥85 mm Hg. If participants had been receiving medication for hypertension, they were classified as being within the abnormal category. cDyslipidemia was defined as serum triglycerides ≥150 mg/dL or HDL cholesterol <40 mg/dL. In the case of a non-fasting blood sample, blood triglycerides ≥175 mg/dL was used instead. If participants had been receiving medication for dyslipidemia, they were classified as being within the abnormal category. dPrediabetes/diabetes was defined as fasting glucose ≥110 mg/dL. In the case of a non-fasting blood sample, glucose ≥140 mg/dL was used instead. If participants had been receiving medication for diabetes mellitus, they were classified as being within the abnormal category. CI: confidence interval; HR: hazard ratio.
Tables 5 and 6 present the associations between BMI and prediabetes/diabetes. Participants were classified into four BMI categories: non-severe thinness (BMI ≥16.0 kg/m2), non-moderate thinness (BMI ≥17.0 kg/m2), non-underweight (BMI ≥18.5 kg/m2), and overweight (BMI ≥25.0 kg/m2). Analyses were conducted separately by sex and adjusted for age, (3) height, (5) elevated blood pressure/hypertension, (6) dyslipidemia, (8) alcohol consumption, and (9) smoking. Among men, significant associations with prediabetes/diabetes were observed in the non-severe thinness (BMI ≥16.0 kg/m2), non-moderate thinness (BMI ≥17.0 kg/m2), and non-underweight (BMI ≥18.5 kg/m2) groups, with HRs and 95% CIs of 1.35 (95% CI: 1.13–1.61), 1.34 (95% CI: 1.13–1.61), and 1.36 (95% CI: 1.14–1.63), respectively. Among women, significant associations with prediabetes/diabetes were observed in the non-severe thinness (BMI ≥16.0 kg/m2), non-moderate thinness (BMI ≥17.0 kg/m2), non-underweight (BMI ≥18.5 kg/m2), and overweight (BMI ≥25.0 kg/m2) groups, with HRs and 95% CIs of 1.37 (95% CI: 1.12–1.67), 1.37 (95% CI: 1.13–1.67), 1.39 (95% CI: 1.13–1.70), and 1.52 (95% CI: 1.07–2.18), respectively.
Table 5. Association between BMI, prediabetes, and diabetes with all-cause mortality based on Cox proportional hazards regression analysis in male participants.
| Male participants, n=1,360 | |||
|---|---|---|---|
| Prediabetes/diabetes, n | Person-years | HR (95% CI) | |
| Non-severe thinness (BMI ≥16.0 kg/m²), n=1,357 | 217 | 27526.4 | 1.35 (1.13–1.61) |
| Non-moderate thinness (BMI ≥17.0 kg/m²), n=1,342 | 212 | 27278.0 | 1.34 (1.13–1.61) |
| Non-underweight (BMI ≥18.5 kg/m²), n=1,299 | 210 | 26574.9 | 1.36 (1.14–1.63) |
| Overweight (BMI ≥25.0 kg/m²), n=364 | 77 | 7649.9 | 1.31 (0.95–1.79) |
Prediabetes/diabetes was defined as fasting glucose ≥110 mg/dL. In the case of a non-fasting blood sample, glucose ≥140 mg/dL was used instead. If participants had been receiving medication for diabetes mellitus, they were classified as being within the abnormal category. Adjusted by age, height, elevated blood pressure/hypertension, dyslipidemia, prediabetes/diabetes, alcohol consumption, smoking. CI: confidence interval; HR: hazard ratio; BMI: body mass index.
Table 6. Association between BMI, prediabetes, and diabetes with all-cause mortality based on Cox proportional hazards regression analysis in female participants.
| Female participants, n=2,571 | |||
|---|---|---|---|
| Prediabetes/diabetes, n | Person-years | HR (95% CI) | |
| Non-severe thinness (BMI ≥16.0 kg/m²), n=2,565 | 205 | 60168.0 | 1.37 (1.12–1.67) |
| Non-moderate thinness (BMI ≥17.0 kg/m²), n=2,546 | 205 | 59775.8 | 1.37 (1.13–1.67) |
| Non-underweight (BMI ≥18.5 kg/m²), n=2,461 | 198 | 57905.4 | 1.39 (1.13–1.70) |
| Overweight (BMI ≥25.0 kg/m²), n=701 | 63 | 16453.3 | 1.52 (1.07–2.18) |
Prediabetes/diabetes was defined as fasting glucose ≥110 mg/dL. In the case of a non-fasting blood sample, glucose ≥140 mg/dL was used instead. If participants had been receiving medication for diabetes mellitus, they were classified as being within the abnormal category. Adjusted by age, height, elevated blood pressure/hypertension, dyslipidemia, prediabetes/diabetes, alcohol consumption, smoking. CI: confidence interval; HR: hazard ratio; BMI: body mass index.
These outcomes did not substantially change after excluding a 5-year buffer period at the beginning of the follow-up (data not shown) for both men and women.
Discussion
This study was a 28-year cohort investigation comprising 87,831 person-years of follow-up, examining the relationships among MetS, alcohol consumption, and smoking with life expectancy (for all-cause mortality).
In this study, underweight status, prediabetes/diabetes, and smoking were identified as factors significantly associated with all-cause mortality. Hypertension and diabetes are well-established predictors of all-cause mortality and cardiovascular disease-related deaths in the Japanese population9). In our cohort, when data from men and women were combined, elevated blood pressure/hypertension was associated with all-cause mortality in the Cox proportional hazards model.
However, when data were analyzed separately by sex using the same model, elevated blood pressure/hypertension was not significantly associated with all-cause mortality. The mean age of male and female participants was 60.7 ± 7.6 years and 60.9 ± 7.4 years, respectively, indicating that participants were relatively older. Therefore, based on the results of the log-rank test conducted separately for the two age groups (40–64 years and 65–74 years), it is possible that, among the older participants, blood pressure was not associated with all-cause mortality.
Fujiyoshi et al.10) reported on the long-term cardiovascular disease risk and its variation according to blood pressure categories and age groups. They categorized participants into three age groups—middle-aged (40–64 years), elderly (65–74 years), and very elderly (75–89 years)—and found that among the younger groups, the association between blood pressure category and cardiovascular disease risk was more strongly positive. In other words, mortality in older adults may be more strongly influenced by non-cardiovascular diseases rather than hypertension-related cardiovascular disease. Therefore, in our study, although the association between elevated blood pressure/hypertension and all-cause mortality was not statistically significant when analyzed separately by sex, the hazard ratios remained relatively high.
Furthermore, in our study, prediabetes/diabetes was associated with all-cause mortality in both men and women. Kato et al.11) reported that diabetes was associated with an increased risk of death, particularly from circulatory diseases such as cerebrovascular and cardiovascular disorders. Similarly, Inoue et al12). reported that the presence of diabetes was associated with an increased risk of cancer-related death.
In our study, smoking was associated with all-cause mortality in both men and women. Hara et al.13) reported that participants who smoked—even a small number of cigarettes, whether currently or in the past—had a higher risk of death than non-smokers. Hara et al.13) also found that the cumulative number of cigarettes smoked strongly influenced cancer-related mortality, and that smoking increased the risk of death from cardiovascular disease. Furthermore, Hirakawa et al.14) reported that men may accumulate a greater number of cardiovascular risk factors, such as smoking, regardless of diabetes status, which could lead to a higher incidence of coronary heart disease even among non-diabetic men. Therefore, smoking can be considered a strong risk factor for all-cause mortality.
In our study, MetS was not statistically significantly associated with all-cause mortality in either men or women. However, a positive trend (HRs ≥1.0) was observed between MetS and all-cause mortality. Saito et al.9) reported in the Japan Public Health Center-based Prospective Study that MetS was associated with a modest increase in all-cause mortality. In the JPHC study by Saito et al.9), MetS was diagnosed using BMI instead of waist circumference, as waist measurements were unavailable. Therefore, differences in diagnostic criteria for MetS—specifically, the BMI-based definition used by the National Cholesterol Education Program Adult Treatment Panel III15) and the waist circumference-based definition adopted in Japan (≥85 cm for men and ≥90 cm for women)7) may have influenced our results. Furthermore, Saito et al.9) suggested that a MetS definition requiring obesity might fail to identify non-overweight individuals, who are more prevalent than those with MetS but may have a high mortality risk. In our study, we found that underweight status was a factor influencing all-cause mortality. Individuals with low BMI may have lower resistance to various diseases, including infectious, respiratory, or inflammatory disorders16). In addition, because prediabetes/diabetes was associated with all-cause mortality in both men and women who were not underweight, underweight status and prediabetes/diabetes may represent independent risk factors. Therefore, in our study, no significant association was found between combined underweight status and prediabetes/diabetes and all-cause mortality.
Among our study participants, the medical factors included in the diagnostic criteria for MetS were measured more than 10 years before the implementation of annual medical examinations to detect MetS introduced by the Ministry of Health, Labour and Welfare. The present study involved a 28-year follow-up period. To our knowledge, few studies in Japan have examined the relationship between MetS and all-cause mortality over a follow-up period of 28 years or longer.
Furthermore, during the period from 1996 to 1998, 61.3% of eligible participants underwent the baseline medical health examination. In O City, by 2012, only 21.3% of individuals aged 40–74 years participated in specific medical examinations for MetS detection17). The participation rate in the 1996–1998 baseline survey, which included only individuals aged 40–74 years, was therefore higher than the 2012 participation rate for the same age group. One possible explanation for the higher baseline participation rate is that the Department of Public Health, Ehime University School of Medicine; the Ehime Prefecture Public Health Department (formerly located in O City); and the O City Public Health Center actively encouraged residents to undergo medical examinations to collect baseline data. Recently, it has remained unclear whether annual medical examinations to detect MetS and preventive efforts focusing on central obesity contribute to increased life expectancy18). To further clarify the association between MetS and all-cause mortality in Japanese populations, large-scale, long-term prospective studies are warranted.
Study limitations
This study has several limitations. First, we did not assess the severity of MetS, the specific degree of BMI, or the frequency of alcohol consumption and smoking. Second, because the endpoint of this study was all-cause mortality, we could not determine whether deaths related to cancer or cerebro-cardiovascular disease were associated with lifestyle-related risk factors. Third, given the 28-year observational period, it is likely that participants (1) experienced changes in lifestyle habits and (2) received medical treatment for lifestyle-related diseases during follow-up. Our study design did not include analyses of lifestyle changes or the effects of medical interventions. To capture such changes, a follow-up questionnaire administered every five years would have been required; however, this was not conducted.
Fourth, we did not evaluate participants’ nutritional status and therefore could not clarify the association between malnutrition and prediabetes/diabetes. Consequently, the potential influence of conditions such as type 5 diabetes on all-cause mortality could not be assessed.
For future follow-up studies, it will be necessary to further elucidate the associations of MetS, BMI, alcohol consumption, and smoking with life expectancy by incorporating the specific data needed to address these limitations.
Conclusions
For residents of O City, Ehime Prefecture, MetS was not statistically significantly associated with all-cause mortality. However, a positive trend was observed between MetS and all-cause mortality. Our findings suggest that, in both men and women, abstaining from smoking and preventing underweight status and prediabetes/diabetes to maintain good health may contribute to increased life expectancy.
Conflict of interest
The authors declare no conflicts of interest related to this study.
Funding information
This study was supported in part by the Japan Society for the Promotion of Science KAKENHI Grant-in-Aid for Scientific Research (Grant Nos. JP22590587, JP17K09204, and JP22K10557). The funders had no role in the study design, data collection, management, analysis, or interpretation; manuscript preparation; or the decision to submit the article for publication.
Ethics approval and consent to participate
The Research Ethics Committee of the Faculty of Education, Ehime University, approved the study protocol. Written informed consent was obtained from all participants before the use of confidential information derived from medical examinations, health-related questionnaires, and registration data on cerebrocardiovascular events. Furthermore, all cooperating hospitals participating in the cerebrovascular event registry informed participants of their right to opt out.
Consent for publication
Consent for publication was obtained from all co-authors.
Author contributions
MT conducted the data analysis and drafted the manuscript. KM contributed to data analysis and manuscript review. IS collected the data and participated in data analysis. ST, YT, and HO reviewed the manuscript. TK conceived and designed the study and supervised the entire process. All authors have read and approved the final version of this manuscript.
Acknowledgments
We express our sincere appreciation to the participants and staff of the JA Ehime Kouseiren Medical Examination Center and to the physicians of the Kitaishikai Medical Association, Japan, for their valuable cooperation and contributions.
Funding Statement
This study was supported in part by the Japan Society for the Promotion of Science KAKENHI Grant-in-Aid for Scientific Research (Grant Nos. JP22590587, JP17K09204, and JP22K10557).
Data availability statement
The dataset used in this study is not publicly available due to privacy and ethical restrictions.
References
- 1.Ministry of Health, Labour and Welfare. Summary of the 2023 Vital Statistics Monthly Annual Report. 2024; https://www.mhlw.go.jp/toukei/saikin/hw/jinkou/geppo/nengai23/dl/kekka.pdf (in Japanese) (Accessed April 29, 2025).
- 2.Alberti KG, Zimmet P, Shaw J. Metabolic syndrome—a new world-wide definition. A consensus statement from the International Diabetes Federation. Diabet Med 2006; 23: 469–480. doi: 10.1111/j.1464-5491.2006.01858.x [DOI] [PubMed] [Google Scholar]
- 3.Van Hoang D, Inoue Y, Fukunaga A, et al. Metabolic syndrome and the risk of severe cancer events: a longitudinal study in Japanese workers. BMC Cancer 2023; 23: 555. doi: 10.1186/s12885-023-11026-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Saito I, Yamauchi K, Yamaizumi M, et al. [Metabolic syndrome and risk of stroke incidence in a rural community in Japan]. Nihon Koshu Eisei Zasshi 2022; 69: 394–402(in Japanese). [DOI] [PubMed] [Google Scholar]
- 5.World Health Organization. A healthy lifestyle—WHO recommendations. 2010; https://www.who.int/europe/news-room/fact-sheets/item/a-healthy-lifestyle—who-recommendations (Accessed April 29, 2025).
- 6.Fahim O, Naghshi S, Khademi Z, et al. Association of adult height with cardiovascular mortality: a systematic review and meta-analysis of cohort studies. Int J Clin Pract 2022; 2022: 6959359. doi: 10.1155/2022/6959359 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Committee to Evaluate Diagnostic Standards for Metabolic Syndrome.Definition and the diagnostic standard for metabolic syndrome. Nihon Naika Gakkai Zasshi 2005; 94: 794–809(in Japanese). [PubMed] [Google Scholar]
- 8.Japan Atherosclerosis Society. Japan Atherosclerosis Society (JAS) guidelines for prevention of atherosclerotic cardiovascular disease 2022. 2022; https://www.j-athero.org/jp/wp-content/uploads/publications/pdf/GL2022_s/jas_gl2022_3_230210.pdf (in Japanese) (Accessed April 29, 2025). [PubMed]
- 9.Saito I, Iso H, Kokubo Y, et al. Metabolic syndrome and all-cause and cardiovascular disease mortality: Japan Public Health Center-based Prospective (JPHC) study. Circ J 2009; 73: 878–884. doi: 10.1253/circj.CJ-08-1025 [DOI] [PubMed] [Google Scholar]
- 10.Fujiyoshi A, Ohkubo T, Miura K, et al. Observational Cohorts in Japan (EPOCH-JAPAN) Research GroupBlood pressure categories and long-term risk of cardiovascular disease according to age group in Japanese men and women. Hypertens Res 2012; 35: 947–953. doi: 10.1038/hr.2012.87 [DOI] [PubMed] [Google Scholar]
- 11.Kato M, Noda M, Mizoue T, et al. JPHC Study GroupDiagnosed diabetes and premature death among middle-aged Japanese: results from a large-scale population-based cohort study in Japan (JPHC study). BMJ Open 2015; 5: e007736. doi: 10.1136/bmjopen-2015-007736 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Inoue M, Iwasaki M, Otani T, et al. Diabetes mellitus and the risk of cancer: results from a large-scale population-based cohort study in Japan. Arch Intern Med 2006; 166: 1871–1877. doi: 10.1001/archinte.166.17.1871 [DOI] [PubMed] [Google Scholar]
- 13.Hara M, Sobue T, Sasaki S, et al. Smoking and risk of premature death among middle-aged Japanese: ten-year follow-up of the Japan Public Health Center-based prospective study on cancer and cardiovascular diseases (JPHC Study) cohort I. Jpn J Cancer Res 2002; 93: 6–14. doi: 10.1111/j.1349-7006.2002.tb01194.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hirakawa Y, Ninomiya T, Kiyohara Y, et al. Evidence for Cardiovascular Prevention From Observational Cohorts in Japan Research Group (EPOCH-JAPAN)Age-specific impact of diabetes mellitus on the risk of cardiovascular mortality: an overview from the evidence for Cardiovascular Prevention from Observational Cohorts in the Japan Research Group (EPOCH-JAPAN). J Epidemiol 2017; 27: 123–129. doi: 10.1016/j.je.2016.04.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Grundy SM, Cleeman JI, Daniels SR, et al. American Heart AssociationNational Heart, Lung, and Blood InstituteDiagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation 2005; 112: 2735–2752. doi: 10.1161/CIRCULATIONAHA.105.169404 [DOI] [PubMed] [Google Scholar]
- 16.Sasazuki S, Inoue M, Tsuji I, et al. Research Group for the Development and Evaluation of Cancer Prevention Strategies in JapanBody mass index and mortality from all causes and major causes in Japanese: results of a pooled analysis of 7 large-scale cohort studies. J Epidemiol 2011; 21: 417–430. doi: 10.2188/jea.JE20100180 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.City O. Ozu City National Health Insurance Health Service Implementation Plan. 2016; https://www.city.ozu.ehime.jp/uploaded/life/27130_107278_misc.pdf (in Japanese) (Accessed April 29, 2025).
- 18.Fukuma S, Iizuka T, Ikenoue T, et al. Association of the National Health Guidance Intervention for obesity and cardiovascular risks with health outcomes among Japanese men. JAMA Intern Med 2020; 180: 1630–1637. doi: 10.1001/jamainternmed.2020.4334 [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.
Data Availability Statement
The dataset used in this study is not publicly available due to privacy and ethical restrictions.