ABSTRACT
Aims/Introduction
Although a low body mass index (BMI) is traditionally associated with high mortality in type 1 diabetes (T1D), the increasing prevalence of obesity and its association with mortality warrant further research.
Materials and Methods
Using a nationwide Korean database from 2002 to 2018, we classified patients with T1D into six categories based on metabolic health and BMI (≥25, 18.5–25, and <18.5 kg/m2): metabolically healthy obesity (MHO), metabolically unhealthy obesity (MUO), metabolically healthy normal weight (MHNW), metabolically unhealthy normal weight (MUNW), metabolically healthy underweight (MHUW), and metabolically unhealthy underweight (MUUW). Mortality data were collected until 2020.
Results
Among 8,838 patients with T1D (mean age 54.7 years, 60.6% male), the prevalence of MUUW, MHUW, MUNW, MHNW, MHO, and MUO was 1.1%, 3.0%, 28.5%, 31.2%, 6.3%, and 29.9%, respectively. Over a median 8.3‐year follow‐up, patients with MHO exhibited the lowest all‐cause mortality risk (adjusted hazard ratio [aHR], 0.70; P = 0.007), and those with MUO also had a lower risk (aHR, 0.77; P < 0.001) compared with the MHNW group. Conversely, the underweight groups had significantly higher all‐cause (MHUW: aHR, 2.08; MUUW: aHR, 3.94) and cardiovascular disease (CVD)‐related mortalities (MHUW: aHR, 2.47; MUUW: aHR, 6.31). Patients with MHO had a lower risk of CVD (aHR, 0.76), whereas those with MUO, MUNW, and MUUW had increased risks (aHRs, 1.18, 1.32, and 2.22).
Conclusions
Metabolic unhealthiness was associated with a higher CVD risk, and the risks of mortality and CVD were particularly high in patients with MUUW, underscoring the importance of maintaining metabolic health in T1D.
Keywords: metabolic syndrome, type 1 diabetes
In type 1 diabetes, mortality was highest in metabolically unhealthy underweight patients, and metabolic unhealthiness was associated with higher cardiovascular disease risk across all BMI categories. Comprehensive metabolic management is needed.
INTRODUCTION
The global burden of type 1 diabetes (T1D) is estimated to significantly increase in the upcoming years, presenting a major public health challenge. 1 Although Korea has historically had a low incidence of T1D, a noticeable increase in the number of cases has been reported recently. 2 T1D is strongly associated with an increased risk of cardiovascular disease (CVD) and early mortality. 3 , 4 , 5 , 6 , 7 , 8 , 9
The Diabetes Control and Complications Trial demonstrated that intensive glucose control reduces the risk of microvascular and macrovascular complications in patients with T1D. 10 However, intensive glucose control often causes weight gain, resulting in a high prevalence of obesity among patients with T1D. 11 , 12 A nationwide study of over one million Korean adults 13 reported the highest mortality in individuals with a BMI < 18.5 kg/m2, underscoring underweight as a clinically important risk indicator in the Korean population. Although underweight patients consistently show an increased risk of CVD and mortality, the relationship between obesity and these outcomes in patients with T1D has been inconsistent across studies. 3 , 4 , 5 , 6 , 14 Further, the heterogeneous mortality risks identified across different body mass index (BMI) categories indicate contributions from additional factors, such as metabolic syndrome.
The incidence of metabolic syndrome, characterized by a cluster of cardiovascular risk factors, 15 has recently increased in patients with T1D. 2 This trend is concerning, as metabolic syndrome may further increase the risk of CVD and early mortality in patients with T1D. 16 , 17 , 18 , 19 The Asian population may be particularly vulnerable to these complications because they often have higher visceral fat content than Caucasians with similar BMIs. 20 , 21 Consequently, the effects of metabolic syndrome on CVD risk and mortality in Asian patients with T1D may be potentiated even at lower BMI thresholds.
In this context, we evaluated the impact of BMI and metabolic health on all‐cause mortality and CVD in patients with T1D and determined their relative importance using the data from the Korean National Health Insurance Service (NHIS) database.
METHODS
Data source and study population
The NHIS is operated by the Korean government and covers approximately 97% of the Korean population. The Korean National Health Information Database (NHID) contains health checkup data, including height, weight, waist circumference, systolic and diastolic blood pressure, fasting plasma glucose, other laboratory variables, medical history based on claims using the disease diagnosis codes of the International Classification of Diseases (ICD‐10), and a hospital utilization database. Further details regarding this database have been described previously. 22 In addition, this database was merged with the death records managed by the Korean National Statistical Office. The study protocol was approved by the Institutional Review Board (IRB) of Korea University and was performed in accordance with the Declaration of Helsinki of the World Medical Association (IRB number: 2020AN0369). The IRB granted an informed consent exemption for this study because only anonymous and de‐identified data from the NHIS were used.
A total of 19,592 patients with T1D were identified between January 2002 and December 2018, and 9,950 patients who underwent national health checkups between January 2009 and June 2020 were selected. The enrollment date was defined as the date of the first health examination conducted at age 20 years or older after the diagnosis of T1D. The following patients were excluded: (i) those who had no health checkup after the diagnosis of T1D (n = 264), (ii) those who had no health checkup after age ≥ 20 years (n = 15), (iii) those with missing data (n = 774), and (iv) those who died within 6 months of enrollment in the study (n = 59) (Figure S1). Consequently, 8,838 participants were included in the final analysis.
Definition of study populations and outcomes
Patients with T1D were defined as those with three or more prescriptions of insulin, as defined by the ICD‐10 code E10. We excluded subjects who met the following criteria to extract accurate data: (i) a primary diagnosis of E11–14 was claimed during hospital admissions or outpatient visits, or if a secondary diagnosis of E11–14 and oral hypoglycemic agents were claimed within 730 days after the first insulin prescription, (ii) no insulin prescription was made between 365 and 730 days after the first insulin prescription, and (iii) a claim for pancreatectomy. These criteria aligned with the operational definition of type 1 diabetes validated in previous Korean NHIS studies. 5 , 23 Patients were divided into six groups according to their metabolic health status and BMI: metabolically healthy obesity (MHO), metabolically unhealthy obesity (MUO), metabolically healthy normal weight (MHNW), metabolically unhealthy normal weight (MUNW), metabolically healthy underweight (MHUW), and metabolically unhealthy underweight (MUUW). Underweight was defined as a BMI <18.5 kg/m2, and obesity was defined as a BMI ≥25 kg/m2 according to the Asia–Pacific BMI criteria. 24 The NCEP‐ATP III criteria state that a metabolically unhealthy status is defined as the presence of two or more of the following factors: (i) waist circumference ≥90 cm in men and ≥85 cm in women based on the International Diabetes Federation criteria for Asians, 24 (ii) systolic blood pressure ≥130/85 mmHg or use of anti‐hypertensive agents under the ICD‐10 codes for hypertension (I10–I15), (iii) triglyceride levels ≥150 mg/dL or current use of lipid‐lowering agents under the ICD‐10 code for dyslipidemia (E78), and (iv) high‐density lipoprotein concentrations <40 mg/dL in men or <50 mg/dL in women or current use of lipid‐lowering agents under the ICD‐10 code for dyslipidemia (E78).
The primary outcome was all‐cause mortality by December 2020. Secondary outcomes included cause‐specific mortality and CVD outcomes. Cause‐specific mortality was defined as death within 30 days of diagnosis using the ICD‐10 codes for CVD, cancer, kidney disease, and diabetes mellitus with coma or ketoacidosis. CVD outcomes included nonfatal myocardial infarction, nonfatal stroke, and heart failure. The detailed definitions of the outcomes are presented in Table S1.
Definitions of covariates
Hypertension was defined as the presence of ≥1 claim per year under the ICD‐10 codes I10 or I11 and ≥1 claim per year for the prescription of anti‐hypertensive agents or systolic/diastolic blood pressure ≥140/90 mmHg. Dyslipidemia was defined as the presence of ≥1 claim per year under ICD‐10 code E78 and ≥1 claim per year for the prescription of a lipid‐lowering agent or total cholesterol level ≥240 mg/dL. Chronic kidney disease (CKD) was defined as an estimated glomerular filtration rate (eGFR) of <60 mL/min/1.73 m2, whereas proteinuria was defined as a positive result on a urine dipstick test. Alcohol consumption was defined as drinking alcohol at least twice per week. Current smoking status was determined using a questionnaire. Regular exercise was defined as moderate‐to‐vigorous activity for >30 min/day or vigorous activity for >20 min/day for at least 3 days/week. Low socioeconomic status (SES) was defined as the lowest tertile of the national health insurance premiums.
Statistical analyses
Continuous data are presented as the mean ± standard deviation (SD), or median with interquartile range (IQR), and categorical data are presented as numbers (percentages). Differences among the six groups were assessed using analysis of variance (anova), Kruskal–Wallis test, or chi‐squared test. The incidence rates of all‐cause mortality and CV events were calculated by dividing the total number of events by the total follow‐up period and are presented as rates per 1,000 person‐years. The adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs) were determined using multivariate Cox proportional hazards regression models. For mortality according to cause of death, cause‐specific hazard ratios were estimated using cause‐specific Cox proportional hazards models, with deaths from causes other than the cause of interest treated as censoring events at the time of death. These models were applied to analyze the effects of metabolic health and obesity, both in combination and separately. The results were visually represented using forest plots. The proportional hazards assumption of the Cox model was evaluated using graphs of the scaled Schoenfeld residuals against time and was confirmed to be satisfied. We adjusted for several factors across different models. Model 1 was adjusted for age and sex; Model 2 was additionally adjusted for smoking status, alcohol consumption, physical activity, SES, eGFR, proteinuria, duration of diabetes, fasting blood glucose levels, family history of diabetes, previous history of CVD, and arrhythmia; and Model 3 was further adjusted for mean fasting blood glucose levels during follow‐up to address the impact of glycemic control. Several sensitivity analyses were performed to ensure the robustness of the findings. First, we examined the association between preexisting conditions (CVD, CKD, and cancer) and mortality outcomes. Subsequently, we conducted an additional sensitivity analysis that excluded patients who had not been followed up for at least 2 years after enrollment or were current smokers. We further redefined metabolic unhealthiness by excluding waist circumference from our criteria. For cardiovascular outcomes, we performed a separate sensitivity analysis that excluded patients with previously diagnosed CVD to focus exclusively on incident cases. Finally, we evaluated the associations of metabolic unhealthiness with mortality and CVD outcomes within each BMI category. Missing data were not imputed, and all analyses were performed using the available data for each variable. Statistical analyses were performed using SAS Enterprise Guide version 7.1 (SAS Institute Inc., Cary, NC, USA), and P‐values <0.05 were considered statistically significant.
RESULTS
Baseline characteristics
The characteristics of the participants according to their metabolic and body phenotypes are presented in Table 1. A total of 8,838 patients with T1D were included, 60.6% of whom were men. The mean age at diagnosis was 48.1 years, and the mean age at examination health checkup was 54.7 years. The prevalences of MUUW, MHUW, MUNW, MHNW, MUO, and MHO were 1.1% (n = 96), 3.0% (n = 264), 28.5% (n = 2,519), 31.2% (n = 2,754), 29.9% (n = 2,647), and 6.3% (n = 558), respectively. The percentage of participants in the metabolically unhealthy group was 59.8%. The percentages of underweight, normal‐weight, and obese individuals were 4.1%, 59.7%, and 36.3%, respectively. Compared to the metabolically healthy group, the metabolically unhealthy group demonstrated a higher mean age at diagnosis and health checkup, lower eGFR, increased prevalence of proteinuria, lower rates of smoking, reduced regular exercise, and a higher prevalence of comorbidities, such as hypertension, dyslipidemia, chronic kidney disease, and CVD. Underweight subjects were characterized by younger age at T1D diagnosis, higher fasting glucose, lower low‐density lipoprotein cholesterol, a higher proportion of current smokers, less regular exercise, and a higher proportion of low SES. However, the mean fasting glucose levels did not differ significantly between the metabolic and body phenotypes (Table S2).
Table 1.
Baseline characteristics
| Mean (SD), n (%) | MUUW | MHUW | MUNW | MHNW | MUO | MHO | P |
|---|---|---|---|---|---|---|---|
| (N = 96) | (N = 264) | (N = 2,519) | (N = 2,754) | (N = 2,647) | (N = 558) | ||
| Age at diagnosis (years) | 48.6 (15.3) | 38.6 (15.6) | 53.0 (12.8) | 41.5 (16.6) | 52.5 (12.9) | 41.6 (16.6) | <0.001 |
| Age at baseline (years) | 56.2 (14.5) | 45.2 (15.4) | 59.8 (12.3) | 48.1 (15.8) | 58.9 (12.7) | 48.0 (15.7) | <0.001 |
| Diabetes duration (years), median (IQR) | 7.9 (5.9, 10.1) | 7.3 (3.0, 9.1) | 7.5 (4.4, 9.0) | 7.3 (3.5, 9.1) | 7.1 (3.5, 8.7) | 7.3 (3.3, 8.9) | <0.001 |
| Men (n [%]) | 56 (58.3) | 143 (54.2) | 1,451 (57.6) | 1746 (63.4) | 1,585 (59.9) | 377 (67.6) | <0.001 |
| Body weight (kg) | 45.4 (6.1) | 47.2 (5.4) | 59.7 (7.9) | 59.6 (7.9) | 74.0 (10.9) | 72.0 (8.9) | <0.001 |
| Body mass index (kg/m2) | 17.4 (0.8) | 17.5 (0.9) | 22.7 (1.6) | 21.8 (1.7) | 27.8 (2.5) | 26.5 (1.7) | <0.001 |
| Waist circumference (cm) | 71.0 (6.6) | 67.7 (6.0) | 82.0 (7.0) | 77.0 (6.5) | 92.9 (7.5) | 85.6 (6.1) | <0.001 |
| Systolic blood pressure (mmHg) | 123.7 (18.6) | 114.7 (13.4) | 130.8 (17.5) | 121.0 (15.5) | 132.4 (16.6) | 122.6 (13.4) | <0.001 |
| Diastolic blood pressure (mmHg) | 71.0 (6.6) | 67.7 (6.0) | 77.4 (10.6) | 73.9 (9.4) | 79.1 (10.5) | 75.6 (8.9) | 0.003 |
| Fasting glucose (mg/dL) | 167.4 (90.6) | 162.4 (95.6) | 153.4 (73.4) | 156.3 (83.1) | 151.6 (65.9) | 146.2 (69.0) | 0.142 |
| Total cholesterol (mg/dL) | 176.3 (56.9) | 180.8 (42.3) | 182.9 (56.5) | 183.0 (37.9) | 183.7 (43.9) | 187.5 (35.8) | <0.001 |
| Triglyceride (mg/dL), median (IQR) | 136 (82.0, 174.5) | 80 (52.0, 106.5) | 141 (94.0, 202.0) | 81 (60.0, 110.0) | 151 (105.0, 212.0) | 94 (69.0, 124.0) | <0.001 |
| High‐density lipoprotein cholesterol (mg/dL) | 50.2 (17.5) | 64.3 (18.8) | 48.0 (14.7) | 60.2 (15.6) | 46.9 (12.8) | 55.9 (12.0) | <0.001 |
| Low‐density lipoprotein cholesterol (mg/dL) | 96.0 (44.8) | 98.5 (33.7) | 101.9 (50.2) | 104.5 (33.3) | 102.1 (38.4) | 111.3 (32.5) | <0.001 |
| Aspartate aminotransferase (U/L) | 25.0 (15.3) | 24.1 (14.8) | 25.8 (16.9) | 24.8 (23.4) | 27.9 (26.4) | 24.9 (15.4) | <0.001 |
| Alanine aminotransferase (U/L) | 22.1 (15.3) | 19.7 (14.8) | 24.8 (19.4) | 22.8 (19.8) | 28.9 (28.0) | 25.8 (21.5) | <0.001 |
| Creatinine | 1.87 (1.92) | 1.17 (1.63) | 1.57 (2.02) | 1.10 (1.18) | 1.36 (1.63) | 1.06 (1.00) | <0.001 |
| eGFR (mL/min per 1.73 m2) | 65.3 (39.7) | 88.6 (38.2) | 68.9 (44.6) | 85.5 (39.6) | 71.7 (41.7) | 84.1 (26.1) | <0.001 |
| Proteinuria (dipstick +) | 19 (22.1) | 35 (13.6) | 557 (22.9) | 273 (10.1) | 581 (22.4) | 42 (7.6) | <0.001 |
| Current smoker (n [%]) | 25 (27.5) | 79 (31.1) | 565 (23.0) | 706 (26.6) | 542 (21.1) | 143 (27.0) | <0.001 |
| Alcohol consumption (≥2 times/week), n (%) | 14 (14.6) | 42 (15.9) | 437 (17.3) | 575 (20.9) | 511 (19.3) | 103 (18.5) | 0.016 |
| Regular exercise (≥3 times/week), n (%) | 7 (7.7) | 38 (14.8) | 531 (21.6) | 616 (23.2) | 503 (19.6) | 130 (24.7) | <0.001 |
| Low socioeconomic status (lowest tertile), n (%) | 37 (38.5) | 93 (35.2) | 753 (29.9) | 815 (29.6) | 730 (27.6) | 155 (27.8) | <0.001 |
| Comorbidities (n [%]) | |||||||
| Hypertension | 75 (78.1) | 49 (18.6) | 2026 (80.4) | 768 (27.9) | 2,123 (80.2) | 138 (24.7) | <0.001 |
| Dyslipidemia | 60 (62.5) | 17 (6.4) | 1,495 (59.3) | 206 (7.5) | 1,457 (55.0) | 45 (8.1) | <0.001 |
| Chronic kidney disease (eGFR<60) | 42 (43.8) | 39 (14.8) | 894 (35.5) | 386 (14.0) | 857 (32.4) | 66 (11.9) | <0.001 |
| Cardiovascular disease | 51 (53.1) | 57 (21.6) | 1,163 (46.2) | 601 (21.8) | 1,221 (46.1) | 121 (21.7) | |
| Ischemic heart disease (I20‐25) | 37 (38.5) | 36 (13.6) | 822 (32.6) | 414 (15.0) | 893 (33.7) | 82 (14.7) | <0.001 |
| Any stroke (I60‐69) | 26 (27.1) | 28 (10.6) | 567 (22.5) | 248 (9.0) | 596 (22.5) | 43 (7.7) | <0.001 |
| Heart failure (I42, I43, I50) | 10 (10.4) | 9 (3.4) | 203 (8.1) | 93 (3.4) | 245 (9.3) | 23 (4.1) | <0.001 |
| Arrhythmia | 3 (3.1) | 14 (5.3) | 195 (7.7) | 122 (4.4) | 221 (8.3) | 24 (4.3) | <0.001 |
| Cancer (C00–99) | 9 (9.4) | 22 (8.3) | 181 (7.2) | 177 (6.4) | 169 (6.4) | 35 (6.3) | 0.538 |
| Medications (n [%]) | |||||||
| Statin | 39 (40.6) | 1,086 (43.1) | 1,201 (45.4) | ||||
| Anti‐platelet | 37 (38.5) | 31 (11.7) | 1,149 (45.6) | 485 (17.6) | 1,232 (46.5) | 112 (20.1) | <0.001 |
Statin and anti‐platelet use were defined as prescriptions for ≥30 consecutive days within 1 year before baseline. Hypertension and dyslipidemia are components of metabolic unhealthiness, causing high prevalence in metabolically unhealthy groups; statin users were also classified as metabolically unhealthy by the study definition. GFR, glomerular filtration rate; MHNW, metabolically healthy normal weight; MHO, metabolically healthy obese; MHUW, metabolically healthy underweight; MUNW, metabolically unhealthy normal weight; MUO, metabolically unhealthy obese; MUUW, metabolically unhealthy underweight.
Risk of all‐cause and cardiovascular mortality according to metabolic and body phenotypes
Over a median (IQR) follow‐up period of 8.3 (5.1–10.6) years, 2,062 deaths from all causes and 644 deaths related to CVD were recorded. For all‐cause mortality using Model 2, subjects in the MUUW group exhibited a significantly increased risk (aHR: 3.94, 95% CI: 2.88–5.41, P < 0.001), as did those in the MHUW group (aHR: 2.08, 95% CI: 1.58–2.73, P < 0.001) (Figure 1 and Table 2). In contrast, the MHO (aHR: 0.70, 95% CI: 0.53–0.90, P = 0.007) and MUO (aHR: 0.77, 95% CI: 0.68–0.88, P < 0.001) groups showed decreased risks compared to the MHNW group. Further, compared to the MHNW group, the MUUW group also demonstrated a significantly higher risk of cardiovascular mortality (aHR: 6.31, 95% CI: 3.76–10.59, P < 0.001). Furthermore, compared with the MHNW group, the MHUW (aHR: 2.47) and MUNW (aHR: 1.36) groups displayed an increased risk of cardiovascular mortality. After adjusting for mean fasting glucose levels during follow‐up, the associations between metabolic and body phenotypes and mortality outcomes remained consistent (Table S3). Similar association patterns were observed after excluding patients who did not undergo follow‐up for at least 2 years after enrollment (Table S4) and those classified as current smokers (Table S5). In the sensitivity analysis considering a previous history of CVD, CKD, and cancer (Table S6), the risks of all‐cause mortality were potentiated in the MUUW (aHR, 6.34) and MHUW (aHR, 2.67) groups. However, no significant differences were observed in the MHO group. When we redefined metabolic unhealthiness by excluding waist circumference from our criteria, the pattern of associations remained consistent; the MUUW group demonstrated the highest risk for both all‐cause mortality (aHR, 3.98) and cardiovascular mortality (aHR, 6.36) compared to the MHNW group (Table S7). In patients with T1D, the risk of all‐cause mortality was 2.59 times higher in the underweight group and 29% lower in the obese group than in the normal weight group (Table S8). Furthermore, metabolically unhealthy patients had a 16% increase in all‐cause mortality rates compared with metabolically healthy patients.
Figure 1.
Forest plots of adjusted hazard ratios for mortality outcomes by metabolic and body phenotype. Adjusted hazard ratios (95% CI) for mortality across different metabolic health and body phenotype categories. P‐values indicate the statistical significance of the difference compared to the reference group (MHNW). Hazard ratios are adjusted for age, sex, smoking, alcohol consumption, physical activity, socioeconomic status, glomerular filtration rate, diabetes duration, fasting glucose level, family history of diabetes, cardiovascular disease (ischemic heart disease, stroke, and heart failure), and arrhythmia. MUUW, metabolically unhealthy underweight; MHUW, metabolically healthy underweight; MUNW, metabolically unhealthy normal weight; MHNW, metabolically healthy normal weight (reference); MUO, metabolically unhealthy obese; MHO, metabolically healthy obese. CI: confidence interval.
Table 2.
Risk of all‐cause and cause‐specific mortality according to metabolic and body phenotypes
| Hazard ratio (95% CI) | |||||||
|---|---|---|---|---|---|---|---|
| No. of events (IR) | Unadjusted | P | Model 1* | P | Model 2 † | P | |
| All‐cause mortality | |||||||
| MUUW (N = 96) | 55 (111.93) | 5.87 (4.43–7.77) | <0.001 | 4.34 (3.26–5.77) | <0.001 | 3.94 (2.88–5.41) | <0.001 |
| MHUW (N = 264) | 63 (34.86) | 1.72 (1.32–2.24) | <0.001 | 2.31 (1.77–3.01) | <0.001 | 2.08 (1.58–2.73) | <0.001 |
| MUNW (N = 2,519) | 799 (42.02) | 1.97 (1.76–2.22) | <0.001 | 1.23 (1.10–1.39) | 0.001 | 1.07 (0.94–1.21) | 0.294 |
| MHNW (N = 2,754) | 449 (21.27) | 1 (Reference) | 1 (Reference) | 1 (Reference) | |||
| MUO (N = 2,647) | 626 (29.98) | 1.39 (1.23–1.57) | <0.001 | 0.87 (0.77–0.98) | 0.026 | 0.77 (0.68–0.88) | <0.001 |
| MHO (N = 558) | 70 (16.09) | 0.75 (0.58–0.96) | 0.023 | 0.71 (0.55–0.91) | 0.007 | 0.70 (0.53–0.90) | 0.007 |
| Cardiovascular mortality | |||||||
| MUUW (N = 96) | 22 (44.77) | 9.52 (6.02–15.05) | <0.001 | 6.72 (4.21–10.73) | <0.001 | 6.31 (3.76–10.59) | <0.001 |
| MHUW (N = 264) | 19 (10.51) | 2.08 (1.28–3.39) | 0.003 | 2.79 (1.71–4.56) | <0.001 | 2.47 (1.48–4.12) | 0.001 |
| MUNW (N = 2,519) | 266 (13.99) | 2.66 (2.13–3.32) | <0.001 | 1.58 (1.26–1.98) | <0.001 | 1.36 (1.07–1.73) | 0.013 |
| MHNW (N = 2,754) | 111 (5.26) | 1 (Reference) | 1 (Reference) | 1 (Reference) | |||
| MUO (N = 2,647) | 207 (9.91) | 1.86 (1.48–2.34) | <0.001 | 1.09 (0.86–1.38) | 0.471 | 0.97 (0.75–1.24) | 0.791 |
| MHO (N = 558) | 19 (4.37) | 0.82 (0.50–1.34) | 0.427 | 0.78 (0.48–1.27) | 0.314 | 0.74 (0.44–1.24) | 0.258 |
| Cancer mortality | |||||||
| MUUW (N = 96) | 5 (10.18) | 3.05 (1.23–7.53) | 0.016 | 2.56 (1.03–6.38) | 0.044 | 2.88 (1.15–7.21) | 0.024 |
| MHUW (N = 264) | 8 (4.43) | 1.25 (0.60–2.59) | 0.548 | 1.66 (0.80–3.45) | 0.173 | 1.61 (0.77–3.37) | 0.209 |
| MUNW (N = 2,519) | 112 (5.89) | 1.62 (1.21–2.16) | 0.001 | 1.11 (0.83–1.50) | 0.485 | 1.08 (0.80–1.47) | 0.61 |
| MHNW (N = 2,754) | 77 (3.65) | 1 (Reference) | 1 (Reference) | 1 (Reference) | |||
| MUO (N = 2,647) | 96 (4.60) | 1.25 (0.93–1.69) | 0.144 | 0.86 (0.63–1.17) | 0.334 | 0.83 (0.61–1.14) | 0.254 |
| MHO (N = 558) | 12 (2.76) | 0.75 (0.41–1.38) | 0.356 | 0.71 (0.39–1.31) | 0.275 | 0.74 (0.40–1.36) | 0.334 |
| Kidney disease‐related mortality | |||||||
| MUUW (N = 96) | 19 (38.67) | 11.30 (6.86–18.62) | <0.001 | 7.95 (4.78–13.24) | <0.001 | 4.97 (2.73–9.04) | <0.001 |
| MHUW (N = 264) | 17 (9.41) | 2.53 (1.50–4.27) | <0.001 | 3.29 (1.94–5.57) | <0.001 | 3.22 (1.82–5.71) | <0.001 |
| MUNW (N = 2,519) | 223 (11.73) | 3.02 (2.34–3.89) | <0.001 | 1.88 (1.45–2.44) | <0.001 | 1.28 (0.96–1.71) | 0.094 |
| MHNW (N = 2,754) | 82 (3.88) | 1 (Reference) | 1 (Reference) | 1 (Reference) | |||
| MUO (N = 2,647) | 167 (8.00) | 2.03 (1.56–2.65) | <0.001 | 1.26 (0.96–1.66) | 0.091 | 0.96 (0.71–1.29) | 0.784 |
| MHO (N = 558) | 15 (3.45) | 0.88 (0.51–1.52) | 0.641 | 0.84 (0.49–1.46) | 0.546 | 0.92 (0.51–1.64) | 0.766 |
| Diabetes with coma or ketoacidosis | |||||||
| MUUW (N = 96) | 9 (18.32) | 10.63 (5.15–21.94) | <0.001 | 7.83 (3.71–16.53) | <0.001 | 8.55 (3.81–19.15) | <0.001 |
| MHUW (N = 264) | 8 (4.43) | 2.41 (1.13–5.16) | 0.023 | 3.13 (1.45–6.72) | 0.004 | 2.91 (1.33–6.38) | 0.008 |
| MUNW (N = 2,519) | 72 (3.79) | 2.00 (1.36–2.94) | <0.001 | 1.31 (0.88–1.95) | 0.184 | 1.19 (0.78–1.83) | 0.417 |
| MHNW (N = 2,754) | 40 (1.89) | 1 (Reference) | 1 (Reference) | 1 (Reference) | |||
| MUO (N = 2,647) | 44 (2.11) | 1.10 (0.72–1.69) | 0.663 | 0.73 (0.47–1.13) | 0.153 | 0.68 (0.43–1.09) | 0.109 |
| MHO (N = 558) | 2 (0.46) | 0.24 (0.06–0.99) | 0.049 | 0.23 (0.06–0.95) | 0.043 | 0.12 (0.02–0.89) | 0.038 |
IR, incidence rate 1,000 person‐years.
MHUW, metabolically healthy underweight; MUUW, metabolically unhealthy underweight; MHNW, metabolically healthy normal weight; MUNW, metabolically unhealthy normal weight; MHO, metabolically healthy obese; MUO, metabolically unhealthy obese.
Model 1 was adjusted for age and sex.
Model 2: adjusted for age, sex, smoking, alcohol consumption, physical activity, socioeconomic status, glomerular filtration rate, proteinuria, diabetes duration, fasting glucose, family history of diabetes, cardiovascular disease (ischemic heart disease, stroke, and heart failure), and arrhythmia.
Regarding the cause‐specific mortality, the MUUW group had the highest risk of cancer‐ (aHR 2.88) and kidney‐related mortality (aHR 4.97). Diabetic coma or ketoacidosis‐related mortality was notably higher in the MUUW (aHR: 8.55, 95% CI: 3.81–19.15) and MHUW groups (aHR: 2.91, 95% CI: 1.33–6.38) than in other groups. In contrast, the MHO group displayed a significantly lower risk (aHR: 0.12, 95% CI: 0.02–0.89) of diabetic coma or ketoacidosis‐related mortality than did the MHNW and other groups.
Risk of cardiovascular disease according to metabolic and body phenotypes
Figure 2 and Table S9 show the aHR for incident cardiovascular composite outcomes. The cardiovascular composite outcome included a U‐shaped pattern according to metabolic and body phenotypes when MHNW were used as a reference. The MUUW group reported the highest risk (aHR: 2.22, 95% CI: 1.39–3.55, P = 0.001), followed by the MHUW (aHR: 1.46), MUNW (aHR: 1.32), MUO (aHR: 1.18), and MHO (aHR: 0.76) groups. These patterns remained consistent even after additional adjustments for mean fasting glucose levels during follow‐up (Table S10), excluding patients with a history of CVD (Table S11), and metabolic unhealthiness redefined without waist circumference (Table S12). Compared to the MHNW group, the MHUW, MUNW, and MUO groups had higher risks of heart failure. The MUNW group had higher risks of nonfatal MI (aHR: 1.40, 95% CI: 1.12–1.76), and MHO had a lower risk of nonfatal ischemic stroke (aHR: 0.40, 95% CI: 0.20–0.78) than did the other groups.
Figure 2.
Forest plot of adjusted hazard ratios for cardiovascular disease by metabolic and body phenotype. Adjusted hazard ratios (95% CI) for cardiovascular disease across different metabolic health and body phenotype categories. P‐values indicate the statistical significance of the difference compared to the reference group (MHNW). Hazard ratios are adjusted for age, sex, smoking, alcohol consumption, physical activity, socioeconomic status, glomerular filtration rate, diabetes duration, fasting glucose level, family history of diabetes, cardiovascular disease (ischemic heart disease, stroke, and heart failure), and arrhythmia. MUUW, metabolically unhealthy underweight; MHUW, metabolically healthy underweight; MUNW, metabolically unhealthy normal weight; MHNW, metabolically healthy normal weight (reference); MUO, metabolically unhealthy obese; MHO, metabolically healthy obese; CI: confidence interval.
Based solely on body phenotype, the underweight group had a higher risk of these outcomes than did the normal‐weight group (Table S13). Conversely, considering only metabolic health, the metabolically unhealthy group had increased risks for cardiovascular outcomes (aHR: 1.38, 95% CI: 1.21–1.58), nonfatal MI (aHR: 1.47, 95% CI: 1.20–1.81), and heart failure (aHR: 1.42, 95% CI: 1.18–1.72) compared to the metabolically healthy group.
Associations of metabolic unhealthiness with mortality and cardiovascular disease according to BMI category
Compared with metabolically healthy status, metabolically unhealthy status was associated with a higher risk of all‐cause and cardiovascular mortality, particularly among individuals with BMI <18.5 kg/m2 (Table S14). In contrast, this association was attenuated and not statistically significant in participants with BMI ≥25 kg/m2. However, the association between metabolic unhealthiness and cardiovascular composite outcomes did not differ significantly across BMI categories (p for interaction = 0.624) (Table S15).
DISCUSSION
This large nationwide population study demonstrated that underweight patients had a higher risk of all‐cause mortality than normal‐weight or obese patients. In particular, the MUUW group had the highest risk of both all‐cause and cardiovascular mortality, whereas the MHO group had the lowest risk. Furthermore, the MUUW group had the highest risk of CVD, followed by the MHUW, MUNW, and MUO groups, with a more pronounced effect of metabolic health than body phenotype. To our knowledge, this is the first study to evaluate the effects of metabolic health according to body phenotype on all‐cause mortality and CVD in T1D.
Previous studies have investigated the relationship between BMI and the risks of early mortality and cardiovascular events in T1D. 3 , 4 , 5 , 6 , 7 , 14 One meta‐analysis including 23,407 patients reported a 3.4‐fold higher mortality risk in the underweight group than in the normal‐weight group, with no difference observed in the overweight/obese group. 4 Similarly, Lee et al. 5 demonstrated that baseline BMI was inversely associated with CVD incidence and all‐cause mortality. Consistent with these findings, low BMI was associated with higher risks of all‐cause mortality and several cause‐specific mortality outcomes in our study. This association may partly reflect underlying conditions such as occult malignancy, renal disease, chronic inflammation, or advanced diabetic complications. 25 Despite the small size of the underweight subgroups, MUUW patients showed a higher risk of all‐cause mortality than their MHUW counterparts. Estimates for some cause‐specific mortality outcomes were based on limited event counts and should therefore be interpreted with caution.
The lower mortality in the MHO group may reflect preserved metabolic health and greater physiological reserve, including better nutritional status, 13 preserved muscle mass, 26 and younger biological age, rather than a protective effect of obesity per se. Although Edqvist et al. 3 reported an association between higher BMI and increased mortality in T1D, their study excluded individuals with BMI <18.5 kg/m2, included a higher proportion of patients with BMI ≥30 kg/m2 (8.3% vs 5.2% in our cohort), and did not stratify participants by metabolic health phenotypes. Asian populations tend to have greater visceral adiposity and metabolic risk at a given BMI than Caucasians, 20 , 21 suggesting that BMI alone may not adequately capture metabolic risk. 27 These differences may explain the divergent findings and suggest that metabolic health provides information beyond BMI.
The association between MHO and lower mortality was less evident for CVD outcomes, possibly because of the transient nature of MHO, in which individuals potentially progress to metabolically unhealthy phenotypes. 28 , 29 In our study, metabolic unhealthiness was associated with a higher risk of cardiovascular composite outcomes, and this association did not differ significantly across BMI categories, underscoring the importance of metabolic health in CVD risk assessment. However, metabolic health status was assessed only at baseline and may change during follow‐up, potentially resulting in exposure misclassification. Further studies using a time‐varying covariate approach are needed to assess the association between changes in metabolic health status and CVD risk.
Beyond metabolic health status, glycemic control itself may influence CVD risk in T1D. However, adjustment for fasting plasma glucose and mean fasting glucose during follow‐up did not substantially alter the associations between metabolic phenotypes and CVD risk, suggesting that nonglycemic metabolic factors such as hypertension and dyslipidemia may also contribute to CVD risk in patients with T1D. 30 Nonetheless, postprandial hyperglycemia or glycemic variability, which may also contribute to CVD risk in T1D, 8 , 31 were not assessed. Further studies incorporating detailed glycemic measures over long‐term follow‐up are needed.
The NCEP‐ATP III criteria have been widely used to assess mortality and CVD risk in the general population 32 and type 2 diabetes 33 and applied to metabolic health in T1D. 2 , 19 , 34 These criteria include multiple metabolic abnormalities, such as dyslipidemia and hypertension. 35 However, their applicability to T1D warrants caution, as intensive insulin therapy may increase adiposity while lowering triglycerides, potentially affecting the metabolic health classification in both directions. The estimated glucose disposal rate (eGDR) has been proposed as a simple surrogate marker of insulin resistance in T1D, 36 and higher eGDR levels have been consistently associated with lower risks of CVD and mortality in T1D. 37 More precise measures of metabolic health and insulin resistance should be evaluated in future studies of T1D. 38
The current study has several limitations. First, T1D was defined based on claims data; we could not incorporate C‐peptide, anti‐glutamic acid decarboxylase antibody, or insulin antibody, and insulin‐treated type 2 diabetes or adult‐onset atypical diabetes may not be fully distinguished from T1D. Although the mean age at diagnosis in our cohort was higher than that reported in Western studies, 3 , 30 it was comparable to that in a previous Korean nationwide study, 2 reflecting the higher proportion of adult‐onset or fulminant T1D in the Korean population. Second, HbA1c is not collected in the NHIS database, which limited the assessment of overall glycemic burden. 31 In addition, insulin dose and delivery method, continuous glucose monitoring use, severe hypoglycemia, nutritional status, muscle mass, and severity of diabetic complications were unavailable. Third, we could not evaluate causal relationships because of the retrospective study design. Fourth, waist circumference, a component of the metabolic health criteria, may partly overlap with the BMI‐based body phenotype; nevertheless, the results remained consistent after its exclusion. Fifth, our study focused on the Korean population, where the proportion of individuals with a BMI of 30–35 kg/m2 was low (4.7%), and only 0.5% had a BMI >35 kg/m2, potentially limiting generalizability to countries with higher obesity rates. Sixth, Fine–Gray subdistribution hazard models were not additionally performed; therefore, the cumulative incidence of cause‐specific mortality accounting for competing causes of death was not directly evaluated.
Despite these limitations, this is the first study to evaluate the associations of metabolic health and body phenotype with all‐cause mortality and CVD in patients with T1D based on nationwide data. Moreover, we improved robustness by incorporating extensive data from health examinations, detailed histories of underlying diseases, current medications, and lifestyle factors (including alcohol consumption, smoking, and SES), which affect mortality outcomes.
Conclusions
The risks of mortality and CVD were particularly high in patients with MUUW, and metabolic unhealthiness was consistently associated with a higher risk of CVD across BMI categories. Comprehensive and tailored management strategies that focus on preserving metabolic health and managing underweight are essential to reduce premature mortality and CVD in patients with T1D.
DISCLOSURE
The authors declare no conflicts of interest.
Approval of the research protocol: N/A.
Informed consent: N/A.
Registry and the registration no. of the study/trial: N/A.
Animal studies: N/A.
Supporting information
Table S1. Codes for comorbidities and outcomes.
Table S2. Changes in fasting blood glucose levels during follow‐up according to metabolic and body phenotypes.
Table S3. Risk of all‐cause and cause‐specific mortality according to metabolic and body phenotypes in patients with values of fasting glucose during follow‐up period of mortality.
Table S4. Risk of all‐cause and cause‐specific mortality according to metabolic and body phenotypes with the exclusion of patients who had not followed‐up for at least 2 years.
Table S5. Risk of all‐cause and cause‐specific mortality according to metabolic and body phenotypes with the exclusion of patients who were current smokers.
Table S6. Risk of all‐cause and cause‐specific mortality according to metabolic and body phenotypes with the exclusion of patients with a history of cardiovascular disease, chronic kidney disease, and cancer.
Table S7. Risk of all‐cause and cause‐specific mortality according to metabolic status and body phenotypes after excluding waist circumference from metabolic criteria.
Table S8. Risk of mortality according to obesity or metabolic unhealthy status.
Table S9. Risk of cardiovascular disease according to metabolic and body phenotypes.
Table S10. Risk of cardiovascular disease according to metabolic and body phenotypes in patients with values of fasting glucose during the follow‐up period.
Table S11. Risk of cardiovascular disease according to metabolic and body phenotypes with the exclusion of patients with a history of cardiovascular disease.
Table S12. Risk of cardiovascular disease according to metabolic and body phenotypes after excluding waist circumference from metabolic criteria.
Table S13. Risk of cardiovascular disease according to obesity or metabolically unhealthy status.
Table S14. Risks of mortality according to metabolically unhealthy status by BMI category.
Table S15. Risks of cardiovascular disease according to metabolically unhealthy status by BMI category.
Figure S1. Flowchart.
ACKNOWLEDGMENT
This study was conducted in cooperation with NHIS. The National Health Information Database constructed by the NHIS was used (NHIS‐2021‐1‐599), and the results do not necessarily represent the opinions of the National Health Insurance Corporation. This research was supported by a grant from Korea University Anam Hospital, Seoul, Republic of Korea (O2412471). The funder had no role in the study design; data collection, analysis, or interpretation; writing of the manuscript; or the decision to submit the manuscript for publication.
DATA AVAILABILITY STATEMENT
This study used de‐identified data from the NHIS‐NHID. Due to data protection policies, raw data are not publicly available. All analyses were conducted using datasets formally approved and extracted through the NHIS process, and the data were destroyed after the permitted usage period. However, access to the data analyzed may be granted upon reasonable request, in accordance with NHIS data access policy.
References
- 1. Gregory GA, Robinson TIG, Linklater SE, et al. Global incidence, prevalence, and mortality of type 1 diabetes in 2021 with projection to 2040: A modelling study. Lancet Diabetes Endocrinol 2022; 10: 741–760. [DOI] [PubMed] [Google Scholar]
- 2. Lee YB, Han K, Kim B, et al. High proportion of adult cases and prevalence of metabolic syndrome in type 1 diabetes mellitus population in Korea: A Nationwide study. Diabetes Metab J 2019; 43: 76–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Edqvist J, Rawshani A, Adiels M, et al. BMI, mortality, and cardiovascular outcomes in type 1 diabetes: Findings against an obesity paradox. Diabetes Care 2019; 42: 1297–1304. [DOI] [PubMed] [Google Scholar]
- 4. Jung HN, Kim S, Jung CH, et al. Association between body mass index and mortality in type 1 diabetes mellitus: A systematic review and meta‐analysis. Journal of Obesity & Metabolic Syndrome 2023; 32: 151–162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Lee YB, Kim B, Park J, et al. Early mortality and cardiovascular disease, varied association with body mass index and its changes in insulin‐treated diabetes: A nationwide study. Int J Obes 2021; 45: 2482–2489. [DOI] [PubMed] [Google Scholar]
- 6. Flegal KM, Kit BK, Orpana H, et al. Association of all‐cause mortality with overweight and obesity using standard body mass index categories: A systematic review and meta‐analysis. JAMA 2013; 309: 71–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Conway B, Miller RG, Costacou T, et al. Adiposity and mortality in type 1 diabetes. Int J Obes 2009; 33: 796–805. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Vergès B. Cardiovascular disease in type 1 diabetes, an underestimated danger: Epidemiological and pathophysiological data. Atherosclerosis 2024; 394: 117158. [DOI] [PubMed] [Google Scholar]
- 9. Lind M, Svensson AM, Kosiborod M, et al. Glycemic control and excess mortality in type 1 diabetes. N Engl J Med 2014; 371: 1972–1982. [DOI] [PubMed] [Google Scholar]
- 10. Nathan DM, Cleary PA, Backlund JY, et al. Intensive diabetes treatment and cardiovascular disease in patients with type 1 diabetes. N Engl J Med 2005; 353: 2643–2653. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Fang M, Jeon Y, Echouffo‐Tcheugui JB, et al. Prevalence and Management of Obesity in U.S. adults with type 1 diabetes. Ann Intern Med 2023; 176: 427–429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Van der Schueren B, Ellis D, Faradji RN, et al. Obesity in people living with type 1 diabetes. Lancet Diabetes Endocrinol 2021; 9: 776–785. [DOI] [PubMed] [Google Scholar]
- 13. Kim NH, Lee J, Kim TJ, et al. Body mass index and mortality in the general population and in subjects with chronic disease in Korea: A Nationwide cohort study (2002‐2010). PLoS One 2015; 10: e0139924. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Price SA, Gorelik A, Fourlanos S, et al. Obesity is associated with retinopathy and macrovascular disease in type 1 diabetes. Obes Res Clin Pract 2014; 8: e178–e182. [DOI] [PubMed] [Google Scholar]
- 15. Grundy SM. Metabolic syndrome: a multiplex cardiovascular risk factor. J Clin Endocrinol Metab 2007; 92: 399–404. [DOI] [PubMed] [Google Scholar]
- 16. Merger SR, Kerner W, Stadler M, et al. Prevalence and comorbidities of double diabetes. Diabetes Res Clin Pract 2016; 119: 48–56. [DOI] [PubMed] [Google Scholar]
- 17. Powell‐Wiley TM, Poirier P, Burke LE, et al. Obesity and cardiovascular disease: A scientific statement from the American Heart Association. Circulation 2021; 143: e984–e1010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Tune JD, Goodwill AG, Sassoon DJ, et al. Cardiovascular consequences of metabolic syndrome. Transl Res 2017; 183: 57–70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Thorn LM, Forsblom C, Wadén J, et al. Metabolic syndrome as a risk factor for cardiovascular disease, mortality, and progression of diabetic nephropathy in type 1 diabetes. Diabetes Care 2009; 32: 950–952. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Lim U, Ernst T, Buchthal SD, et al. Asian women have greater abdominal and visceral adiposity than Caucasian women with similar body mass index. Nutr Diabetes 2011; 1: e6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Wang D, Li Y, Lee SG, et al. Ethnic differences in body composition and obesity related risk factors: Study in Chinese and white males living in China. PLoS One 2011; 6: e19835. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Kim MK, Han K, Lee SH. Current trends of big data research using the Korean National Health Information Database. Diabetes Metab J 2022; 46: 552–563. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Kim KS, Kim B, Han K. Big data research for diabetes‐related diseases using the Korean National Health Information Database. Diabetes Metab J 2025; 49: 13–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. World Health Organization, Regional Office for the Western Pacific . The Asia‐Pacific Perspective: Redefining Obesity and its Treatment. Sydney: Health Communications Australia, 2000. [Google Scholar]
- 25. Hamasaki H. Obesity paradox in individuals with type 1 diabetes. Front Clin Diabetes Healthc 2025; 6: 1670312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Srikanthan P, Horwich TB, Tseng CH. Relation of muscle mass and fat mass to cardiovascular disease mortality. Am J Cardiol 2016; 117: 1355–1360. [DOI] [PubMed] [Google Scholar]
- 27. Re F, Oguntade AS, Bohrmann B, et al. Associations of general and central adiposity with hypertension and cardiovascular disease among South Asian populations: a systematic review and meta‐analysis. BMJ Open 2023; 13: e074050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Kim JA, Kim DH, Kim SM, et al. Impact of the dynamic change of metabolic health status on the incident type 2 diabetes: A Nationwide population‐based cohort study. Endocrinol Metab (Seoul) 2019; 34: 406–414. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Opio J, Croker E, Odongo GS, et al. Metabolically healthy overweight/obesity are associated with increased risk of cardiovascular disease in adults, even in the absence of metabolic risk factors: A systematic review and meta‐analysis of prospective cohort studies. Obes Rev 2020; 21: e13127. [DOI] [PubMed] [Google Scholar]
- 30. Bebu I, Braffett BH, Pop‐Busui R, et al. The relationship of blood glucose with cardiovascular disease is mediated over time by traditional risk factors in type 1 diabetes: The DCCT/EDIC study. Diabetologia 2017; 60: 2084–2091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Hershon KS, Hirsch BR, Odugbesan O. Importance of postprandial glucose in relation to A1C and cardiovascular disease. Clin Diabetes 2019; 37: 250–259. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Hinnouho G‐M, Czernichow S, Dugravot A, et al. Metabolically healthy obesity and the risk of cardiovascular disease and type 2 diabetes: The Whitehall II cohort study. Eur Heart J 2015; 36: 551–559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Guzder RN, Gatling W, Mullee MA, et al. Impact of metabolic syndrome criteria on cardiovascular disease risk in people with newly diagnosed type 2 diabetes. Diabetologia 2006; 49: 49–55. [DOI] [PubMed] [Google Scholar]
- 34. Lavens A, De Block C, Oriot P, et al. Metabolic health in people living with type 1 diabetes in Belgium: A repeated cross‐sectional study. Diabetologia 2024; 67: 2678–2690. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Cleland SJ, Fisher BM, Colhoun HM, et al. Insulin resistance in type 1 diabetes: What is ‘double diabetes’ and what are the risks? Diabetologia 2013; 56: 1462–1470. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Epstein EJ, Osman JL, Cohen HW, et al. Use of the estimated glucose disposal rate as a measure of insulin resistance in an urban multiethnic population with type 1 diabetes. Diabetes Care 2013; 36: 2280–2285. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Guo L, Zhang J, An R, et al. The role of estimated glucose disposal rate in predicting cardiovascular risk among general and diabetes mellitus population: A systematic review and meta‐analysis. BMC Med 2025; 23: 234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38. Bielka W, Przezak A, Molęda P, et al. Double diabetes‐when type 1 diabetes meets type 2 diabetes: Definition, pathogenesis and recognition. Cardiovasc Diabetol 2024; 23: 62. [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
Table S1. Codes for comorbidities and outcomes.
Table S2. Changes in fasting blood glucose levels during follow‐up according to metabolic and body phenotypes.
Table S3. Risk of all‐cause and cause‐specific mortality according to metabolic and body phenotypes in patients with values of fasting glucose during follow‐up period of mortality.
Table S4. Risk of all‐cause and cause‐specific mortality according to metabolic and body phenotypes with the exclusion of patients who had not followed‐up for at least 2 years.
Table S5. Risk of all‐cause and cause‐specific mortality according to metabolic and body phenotypes with the exclusion of patients who were current smokers.
Table S6. Risk of all‐cause and cause‐specific mortality according to metabolic and body phenotypes with the exclusion of patients with a history of cardiovascular disease, chronic kidney disease, and cancer.
Table S7. Risk of all‐cause and cause‐specific mortality according to metabolic status and body phenotypes after excluding waist circumference from metabolic criteria.
Table S8. Risk of mortality according to obesity or metabolic unhealthy status.
Table S9. Risk of cardiovascular disease according to metabolic and body phenotypes.
Table S10. Risk of cardiovascular disease according to metabolic and body phenotypes in patients with values of fasting glucose during the follow‐up period.
Table S11. Risk of cardiovascular disease according to metabolic and body phenotypes with the exclusion of patients with a history of cardiovascular disease.
Table S12. Risk of cardiovascular disease according to metabolic and body phenotypes after excluding waist circumference from metabolic criteria.
Table S13. Risk of cardiovascular disease according to obesity or metabolically unhealthy status.
Table S14. Risks of mortality according to metabolically unhealthy status by BMI category.
Table S15. Risks of cardiovascular disease according to metabolically unhealthy status by BMI category.
Figure S1. Flowchart.
Data Availability Statement
This study used de‐identified data from the NHIS‐NHID. Due to data protection policies, raw data are not publicly available. All analyses were conducted using datasets formally approved and extracted through the NHIS process, and the data were destroyed after the permitted usage period. However, access to the data analyzed may be granted upon reasonable request, in accordance with NHIS data access policy.