Transition of Metabolic Health Status and the Risk of Cardiovascular Diseases in a Japanese Population: the Hisayama Study

Abstract

Aims: To investigate the association between metabolic health status, defined by the combination of metabolic syndrome (MetS) and obesity, and cardiovascular disease (CVD) in a Japanese community.

Methods: A total of 2,842 participants without prior CVD, aged 40 years or older, were followed up from 2007 until 2017. Participants were classified into 4 metabolic health statuses based on the presence of obesity (body mass index ≥ 25 kg/m²) and MetS: metabolically healthy normal weight (MHN) (obesity [-] and MetS [-]), metabolically unhealthy normal weight (MUN) (obesity [-] and MetS [+]), metabolically healthy obesity (MHO) (obesity [+] and MetS [-]), and metabolically unhealthy obesity (MUO) (obesity [+] and MetS [+]). The risk estimates were computed by using a Cox hazard regression analysis.

Results: During the follow-up period, 190 participants developed CVD. The MUO group had a 1.94-times greater risk of developing CVD than the MHN group after adjusting for confounders, but no excess risk was observed in the MHO group. Moreover, in 1,595 participants who had undergone a health checkup in 2002, 5 years before baseline, the risk of developing CVD was 2.18-times greater in the group that transitioned from MHO to MUO and 1.75-times higher in the stable MUO group than in the stable MHN group, but was not higher in the stable MHO group.

Conclusions: The present findings suggest that cardiovascular risk increases when metabolic abnormalities are present simultaneously with obesity. In individuals with obesity, it may be important to maintain metabolic health and/or lose weight to prevent CVD.

Introduction

The number of individuals with obesity has globally increased over the past 50 years¹⁾. Obesity is known as a risk factor for type 2 diabetes, hypertension, dyslipidemia, cardiovascular diseases (CVD), osteoarthritis, and cancer, contributing to a decrease in quality of life and survival rates^2-5). Metabolically healthy obesity (MHO), often defined as obesity in the absence of high blood pressure or metabolic abnormalities, has been proposed as an obesity subphenotype. A meta-analysis reported that MHO was associated with a greater risk of CVD than metabolically healthy normal weight (MHN) in Asian populations, but the findings were heterogenous across the study due to the different definitions of MHO and the setting of control groups⁶⁾. A large study in Japan that prospectively investigated the association between MHO and CVD incidence using the Japan Medical Data Center database found that MHO was not associated with an increased risk of cardiovascular events, such as myocardial infarction or stroke⁷⁾. Recently, the prognostic value of MHO has been debated, mainly because MHO likely shifts gradually towards metabolically unhealthy obesity (MUO)⁸⁾. However, the influence of transition from MHO to other metabolic health conditions, such as MUO, on the development of CVD has not been previously addressed in Japan.

The Hisayama Study is a population-based prospective cohort study of CVD and its risk factors that has been ongoing in the town of Hisayama, a suburb of the Fukuoka metropolitan area in southwestern Japan⁹⁾. This study has established multiple cohorts by conducting, every 5 years, a comprehensive survey involving approximately 70% or more of the town’s residents aged 40 years and older, in addition to annual health checkups. The aim of the present study was to investigate the association of MHO with the development of CVD in a community-dwelling Japanese population, with consideration for the influence of the transition from MHO to other metabolic health statuses.

Methods

Study Population

In the present study, we used the prospective longitudinal cohort data from the Hisayama Study. Details of the Hisayama Study have been described elsewhere⁹⁾. Briefly, a total of 3,384 residents aged 40 years or older participated in the health check-up in 2007–2008 (the 2007 survey) and consented to participate in the study. After excluding 8 individuals who did not consent to participate in the study, 224 individuals who had a history of CVD, 255 individuals with body mass index (BMI) of less than 18.5 kg/m², and 55 individuals who could not be diagnosed with metabolic syndrome (MetS) due to missing data on waist circumference (WC) or fasting blood test, the remaining 2,842 participants were eligible for the analysis investigating the association of metabolic health status at baseline with the risk of developing of CVD during the 10-year follow-up periods (median 10.3 years from the 2007 survey to November 30, 2017).

Of these 2,842 participants, moreover, 2,433 participants had undergone a health checkup in 2002–2003 (the 2002 survey). We therefore were able to additionally investigate the association of the transition pattern of metabolic health status over the 5 years prior to baseline (from the 2002 survey to the 2007 survey) with the risk of developing CVD during the subsequent 10-year follow-up periods (from the 2007 survey to November 30, 2017) (Supplementary Fig.1). Among the 2,433 potential participants in this additional analysis, we excluded 147 individuals with BMI of less than 18.5 kg/m², and 24 individuals for whom a diagnosis of MetS was not made due to missing data on WC or fasting blood test, and further selected only those participants who exhibited one of 6 transition patterns of metabolic health status that were focused on the influence of MHO and MUO (Supplementary Fig.2). After these exclusions, a total of 1,595 participants remained. Further details regarding the selection of the 6 transition patterns will be provided later.

Supplementary Fig.1.

Study design for the analysis of transitions of metabolic health status and incidence of cardiovascular disease

Supplementary Fig.2.

Flow chart of the study sample selection

The study protocol was approved by the Kyushu University Institutional Review Boards for Clinical Research (Approval No. 23061-03). All participants provided written informed consent.

Definitions of Metabolic Health Status

Body height and weight were measured in light clothing without shoes at both the 2002 and 2007 surveys. BMI was calculated as weight in kilograms divided by squared height in meters (kg/m²). Obesity was defined as BMI ≥ 25 kg/m². MetS was defined using criteria in the joint statement from the International Diabetes Federation, the National Heart, Lung, and Blood Institute, the American Heart Association, the World Heart Federation, the International Atherosclerosis Society, and the International Association for the Study of Obesity with a modification¹⁰⁾. WC was measured at the umbilical level in a standing position by a trained staff member. Abdominal obesity was defined as a WC of 90 cm or greater in men and 80 cm or greater in women, according to the International Obesity Task Force central obesity criteria for Asians¹¹⁾. The blood pressure was measured three times in a sitting position using an automated sphygmomanometer (BP-203 RVIIIB; Omron Healthcare), and the mean of three measurements was used for the analysis. Elevated blood pressure was defined as average systolic/diastolic blood pressures of 130/85 mmHg or greater and/or current use of antihypertensive medicine. Hypertriglyceridemia was defined as a serum triglyceride level of 150 mg/dL (1.69 mmol/L) or greater. Low serum HDL cholesterol level was defined as less than 40 mg/dL (1.03 mmol/L) in men and less than 50 mg/dL (1.29 mmol/L) in women. Elevated blood glucose level was defined as a fasting blood glucose level of 100 mg/dL (5.6 mmol/L) or greater and/or use of glucose-lowering agents. MetS was defined as the presence of three or more of these five components¹⁰⁾.

For the analysis of the relationship between metabolic health status and CVD, participants were classified into 4 metabolic health status groups based on obesity and MetS: metabolically healthy normal weight (MHN) (obesity [-] and MetS [-]), metabolically unhealthy normal weight (MUN) (obesity [-] and MetS [+]), MHO (obesity [+] and MetS [-]), and MUO (obesity [+] and MetS [+]). For the analysis according to the transition patterns of metabolic health status, participants were classified into 16 categories based on the transition patterns of metabolic health status from the 2002 survey to the 2007 survey as shown in Fig.1. Among these transition patterns, we restricted the analysis to 6 patterns (shown as colored boxes in Fig.1)to specifically focus on the transition between MHO and MUO.

Fig.1. Transition patterns of metabolic health status from the 2002 survey to the 2007 survey

Abbreviations: MHN, metabolically healthy normal weight; MUN, metabolically unhealthy normal weight; MHO, metabolically healthy obesity; MUO, metabolically unhealthy obesity.

Definitions of Outcome

The participants were followed up prospectively until November 2017 (median 10.3 years from the 2007 survey). The primary outcome of the present analysis was CVD that was defined as first-ever development of coronary heart disease or stroke. The criteria for a diagnosis of coronary heart disease included first-ever fatal and non-fatal acute myocardial infarction, silent myocardial infarction, sudden cardiac death within 1 hour after the onset of acute illness, or coronary artery disease followed by coronary artery bypass surgery or percutaneous coronary angioplasty. Acute myocardial infarction was diagnosed when a participant met at least two of the following criteria: (1) typical symptoms, including prolonged severe anterior chest pain; (2) abnormal cardiac enzymes more than twice the upper limit of the normal range; (3) evolving diagnostic electrocardiographic changes; and (4) morphological changes, including local asynergy of cardiac wall motion on electrocardiography, persistent perfusion defect on cardiac scintigraphy, or myocardial necrosis or scars 1 cm long accompanied by coronary atherosclerosis at autopsy. Silent myocardial infarction was defined as myocardial scarring without any historical indication of clinical symptoms or abnormal cardiac enzyme changes. Stroke was defined as a first-ever sudden onset of nonconvulsive and focal neurological deficit persisting for 24 hours. The diagnosis of stroke and the determination of its pathological type were based on the clinical history, neurological examination, and all available clinical data, including brain CT/MRI and autopsy findings. Stroke was classified as either ischemic or hemorrhagic (intracerebral hemorrhage and subarachnoid hemorrhage)¹²⁾.

Clinical Evaluation and Laboratory Measurements

Information on medical history, medications for hypertension, diabetes, smoking habits, alcohol intake, and regular exercise was obtained by a self-administered questionnaire. Trained interviewers checked the questionnaire at the examination. Smoking and drinking habits were categorized by whether the participant currently smoked or drank alcohol. Regular exercise was defined as engaging in sports or other physical exercise including recreational walking at least 3 times a week during leisure time. Electrocardiogram abnormalities were defined as left ventricular hypertrophy (Minnesota code, 3-1), ST depression (4-1, 2, 3), or atrial fibrillation/flutter (8-3). The blood samples were collected from an antecubital vein. Plasma glucose levels were measured by the hexokinase method, and serum insulin levels were determined by using an electro-chemiluminescence immunoassay in 2007 and chemiluminescence enzyme immunoassay in 2002. Insulin resistance was estimated by the homeostasis model assessment of insulin resistance (HOMA-IR) values, which were calculated as follows: HOMA-IR = fasting plasma glucose (mg/dL) × fasting serum insulin (IU/mL)/405 ¹³⁾. Serum low-density lipoprotein (LDL) cholesterol, high-density lipoprotein (HDL) cholesterol, triglycerides (TG), and creatinine levels were measured enzymatically. Serum gamma-glutamyl transferase (GGT) was enzymatically measured in accordance with the consensus method of the Japan Society of Clinical Chemistry. The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation with a Japanese coefficient of 0.81328 ¹⁴⁾. Chronic kidney disease was defined as proteinuria (≥ 1 + on dipstick) or eGFR ＜60 mL/min/1.73 m² according to the Kidney Disease: Improving Global Outcomes (KDIGO) guideline¹⁵⁾. Serum high-sensitivity C-reactive protein (hs-CRP) concentrations were measured by a modified version of the Behring Latex-Enhanced CRP assay on a Behring Nephelometer BN-100 (Behring Diagnostics, Westwood, MA). The fatty liver index was calculated using an algorithm based on TG, GGT, BMI, and WC as follows: Fatty liver index = (e ^{[0.953 × ln(TG) + 0.139 × BMI + 0.718 × ln(GGT) + 0.053 × WC - 15.745]})/(1+ e ^{[0.953 × ln(TG) + 0.139 × BMI + 0.718 × ln(GGT) + 0.053 × WC - 15.745]}) x 100 ¹⁶⁾. The serum small dense low-density lipoprotein (sdLDL) cholesterol concentrations were directly measured using a homogeneous assay (Denka, Tokyo). The serum remnant cholesterol concentrations were measured using an automated assay (Denka, Tokyo).

Statistical Analysis

The characteristics of participants at baseline were summarized according to the metabolic health status in the 2007 survey, and according to the transition of metabolic health status from the 2002 survey to the 2007 survey. Dunnet’s test was used for pairwise comparison. The serum levels of TG, hs-CRP, HOMA-IR, and fatty liver index levels were shown as medians and interquartile ranges in the baseline characteristics of the population and were log-transformed in the analyses due to their skewed distributions. Serum small dense LDL cholesterol and serum remnant cholesterol levels were also log-transformed in the analyses due to their skewed distributions. The cumulative incidence for CVD was demonstrated according to metabolic health status by the Kaplan-Meier method. The incidence rate of CVD was calculated using the person-year method after adjusting for age and sex by means of the direct method. The Cox proportional hazards model was used to estimate the hazard ratios (HRs) and their 95% confidence intervals (CIs) for the development of CVD according to the categories of metabolic health status. In the multivariable analysis, the risk estimates were adjusted for the following potential confounding factors at baseline: age, sex, LDL cholesterol, chronic kidney disease, echocardiogram abnormality, current smoking, current drinking, and regular exercise. In investigating the association between the transition of metabolic health status and the development of CVD, the risk estimates were adjusted for age, sex, chronic kidney disease, echocardiogram abnormality, current smoking, current drinking, and regular exercise. In addition to these variables, we additionally adjusted for serum LDL cholesterol, serum small dense LDL cholesterol, serum remnant cholesterol, and fatty liver index, respectively. We used the analysis of covariance method in calculating the geometric mean values of fatty liver index in the 2002 survey and the 2007 survey. Paired t-test was used in analyzing the difference of fatty liver index between the 2002 survey and the 2007 survey in each transition pattern. We also used analysis by the covariance method in calculating the mean values of serum LDL cholesterol and the geometric mean values of serum sdLDL cholesterol and serum remnant cholesterol according to the transitional patterns. A two-sided value of p＜0.05 was considered statistically significant in all analyses. Statistical analyses were conducted using Statistical Analysis Software (SAS) version 9.4 (SAS Institute, Cary, NC).

Results

Prevalence and Characteristics of Each Metabolic Health Status at Baseline

Among 2,842 participants, 1,604 participants were MHN, 449 participants were MUN, 282 participants were MHO, and 507 participants were MUO. Table 1 exhibits the clinical characteristics of the study population at the 2007 survey according to the metabolic health status. The mean values of systolic blood pressure, diastolic blood pressure, WC, and BMI; the median values of serum hs-CRP and HOMA-IR; and the frequency of use of antihypertensive agents were all significantly higher in participants with MHO compared to those with MHN. However, the values were even higher in participants with MUO, and additionally, those with MUO had significantly higher mean values of age, fasting plasma glucose, and LDL cholesterol, significantly higher median value of TG, and significantly greater frequency of chronic kidney disease compared to those with MHN. Conversely, the mean value of serum HDL cholesterol was significantly lower in participants with MHO compared to those with MHN, and even lower in participants with MUO.

Table 1.Clinical characteristics according to metabolic health status at the 2007 survey

	Metabolic health status
	MHN (n = 1604)	MUN (n = 449)	MHO (n = 282)	MUO (n = 507)
Age (years)	62 (13)	67 (11)＊	61 (12)	63 (11)＊
Male sex (%)	43.3	35.2＊	46.8	47.1
Systolic blood pressure (mmHg)	127 (19)	141 (17)＊	130 (16)＊	143 (16)＊
Diastolic blood pressure (mmHg)	77 (10)	84 (10)＊	79 (9)＊	86 (10)＊
Use of antihypertensive agents (%)	20.3	47.0＊	27.0＊	50.1＊
Fasting plasma glucose (mg/dL)	99 (17)	112 (21)＊	100 (19)	119 (30)＊
Use of glucose-lowering agents (%)	3.0	10.9＊	4.6	16.6＊
Serum LDL cholesterol (mg/dL)	122 (30)	129 (32)＊	126 (31)	131 (31)＊
Serum HDL cholesterol (mg/dL)	71 (17)	60 (17)＊	64 (14)＊	58 (15)＊
Serum triglycerides (mg/dL)	88 (66-120)	159 (105-207)＊	95 (72-118)	147 (100-204)＊
Waist circumference (cm)	81.3 (6.5)	86.6 (5.4)＊	92.5 (6.6)＊	96.6 (7.3)＊
BMI (kg/m2)	21.7 (1.7)	22.8 (1.5)＊	26.7 (1.8)＊	27.9 (2.7)＊
Chronic kidney disease (%)	10.7	18.7＊	13.2	19.1＊
Electrocardiogram abnormalities (%)	14.5	19.8＊	11.7	15.6
Serum hs-CRP (mg/L)	0.32 (0.15-0.65)	0.48 (0.24-1.08)＊	0.56 (0.31-0.97)＊	0.71 (0.43-1.29)＊
HOMA-IR	1.09 (0.79-1.48)	1.70 (1.22-2.33)＊	1.64 (1.18-2.24)＊	2.64 (1.86-3.91)＊
Fatty liver index	14 (7-24)	33 (19-47)	40 (29-53)	62 (47-78)
Current smoking (%)	20.0	20.9	18.8	20.1
Current drinking (%)	50.2	42.8	51.1	48.7
Regular exercise (%)	27.9	33.6＊	28.7	32.0

Abbreviations: BMI, body mass index; LDL, low density lipoprotein; HDL, high density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance; hs-CRP, high-sensitivity C-reactive protein; MHN, metabolically healthy normal weight; MHO, metabolically healthy obesity; MUN, metabolically unhealthy normal weight; MUO, metabolically unhealthy obesity.

Data are presented as the mean values (standard deviation), percentages, or median (interquartile range).

^＊p＜0.05 vs. the MHN group.

Incidence of CVD According to the Metabolic Health Status

During The follow-up period, a total of 190 participants had a CVD event. Among them, 90 participants experienced coronary heart disease and 104 participants experienced stroke (82 cases of ischemic stroke and 22 cases of hemorrhagic stroke). Fig.2 shows that the age- and sex-adjusted cumulative incidence of CVD was significantly higher in participants with MUO than those with MHN. However, there was no apparent increased risk for CVD in participants with MHO compared to those with MHN. Table 2 demonstrates the age- and sex-adjusted incidence rates and multivariable-adjusted HRs for the development of CVD according to the metabolic health status. The age- and sex-adjusted incidence rate of CVD for each category was as follows: MHN, 6.2; MUN, 8.8; MHO, 5.0; and MUO, 12.8 per 1000 person-years (Table 2). The HRs for the development of CVD increased significantly in participants with MUO compared to those with MHN after adjusting for age, sex, LDL cholesterol, chronic kidney disease, echocardiogram abnormality, current smoking, current drinking, and regular exercise: participants with MUO had a 1.94-fold (95% CI, 1.37–2.75) higher risk of CVD than those with MHN. However, no increased risk of CVD was observed in participants with MHO. Similar results were obtained for other outcomes, such as coronary heart disease and ischemic stroke, though these associations were not observed in hemorrhagic stroke.

Fig.2. Age- and sex-adjusted cumulative incidence of cardiovascular disease according to metabolic health status, 2007–2017

^＊p＜0.05 vs the MHN group

Abbreviations: MHN, metabolically healthy normal weight; MUN, metabolically unhealthy normal weight; MHO, metabolically healthy obesity; MUO, metabolically unhealthy obesity.

Table 2.Hazard ratios for the development of cardiovascular disease and its subtypes according to metabolic health status during the follow-up period (2007–2017)

	Metabolic health status
	MHN	MUN	MHO	MUO
Cardiovascular disease
Number of events/participants	84/1604	40/449	12/282	54/507
Age- and sex-adjusted incidence rate (per 103 PYs)	6.2	8.8	5.0	12.8
Age- and sex-adjusted HR (95% CI)	1.00 (reference)	1.41 (0.97-2.07)	0.88 (0.48-1.61)	2.05 (1.46-2.89)＊
Multivariable-adjusted HR (95% CI) a)	1.00 (reference)	1.33 (0.90-1.95)	0.91 (0.50-1.67)	1.94 (1.37-2.75)＊
Coronary heart disease
Number of events/participants	31/1604	25/449	6/282	28/507
Age- and sex-adjusted incidence rate (per 103 PYs)	2.2	5.7	2.4	7.6
Age- and sex-adjusted HR (95% CI)	1.00 (reference)	2.68 (1.57-4.56)＊	1.17 (0.49-2.82)	2.85 (1.71-4.75)＊
Multivariable-adjusted HR (95% CI) a)	1.00 (reference)	2.56 (1.50-4.37)＊	1.22 (0.51-2.93)	2.62 (1.55-4.41)＊
Stroke
Number of events/participants	55/1604	15/449	6/282	28/507
Age- and sex-adjusted incidence rate (per 103 PYs)	4.1	2.9	2.5	5.4
Age- and sex-adjusted HR (95% CI)	1.00 (reference)	0.73 (0.41-1.30)	0.68 (0.29-1.58)	1.60 (1.01-2.52)＊
Multivariable-adjusted HR (95% CI) a)	1.00 (reference)	0.67 (0.38-1.21)	0.70 (0.30-1.62)	1.57 (0.99-2.49)
Ischemic stroke
Number of events/participants	40/1604	14/449	5/282	23/507
Age- and sex-adjusted incidence rate (per 103 PYs)	2.9	2.8	2.0	4.4
Age- and sex-adjusted HR (95% CI)	1.00 (reference)	0.95 (0.52-1.76)	0.78 (0.31-1.99)	1.81 (1.08-3.04)＊
Multivariable-adjusted HR (95% CI) a)	1.00 (reference)	0.89 (0.48-1.65)	0.78 (0.31-1.99)	1.74 (1.03-2.93)＊
Hemorrhagic stroke
Number of events/participants	15/1604	1/449	1/282	5/507
Age- and sex-adjusted incidence rate (per 103 PYs)	1.1	0.1	0.5	0.9
Age- and sex-adjusted HR (95% CI)	1.00 (reference)	0.17 (0.02-1.32)	0.42 (0.06-3.15)	1.01 (0.37-2.80)
Multivariable-adjusted HR (95% CI) a)	1.00 (reference)	0.16 (0.02-1.24)	0.47 (0.06-3.58)	1.10 (0.39-3.05)

Abbreviations: MHN, metabolically healthy normal weight; MHO, metabolically healthy obesity; MUN, metabolically unhealthy normal weight; MUO, metabolically unhealthy obesity; LDL, low density lipoprotein; PYs, person years; HR, hazard ratio; CI, confidence interval.

^＊p＜0.05 vs. the MHN group.

^a) Adjusted for age, sex, serum LDL cholesterol, chronic kidney disease, echocardiogram abnormality, current smoking, current drinking, and regular exercise.

Transition Patterns of Metabolic Health Status from the 2002 Survey to the 2007 Survey

Based on the transition patterns of metabolic health status from the 2002 survey to the 2007 survey, 2,262 participants were classified into the 16 groups shown in Fig.1. Among 1,213 participants with MHN in the 2002 survey, 993 (81.9%) participants maintained MHN status (stable MHN group) in the 2007 survey. Of the 248 participants with MHO in the 2002 survey, 46 (18.5%) transitioned to MHN status, 121 (48.8%) maintained MHO status (stable MHO group), and 71 (28.6%) transitioned to MUO status in the 2007 survey. Among 434 participants with MUO in the 2002 survey, 308 (71.0%) participants maintained MUO status (stable MUO group) and 56 (12.9%) transitioned to MHO in the 2007 survey.

For the analysis regarding CVD risk according to the transition patterns of metabolic health status, the following 6 groups (n = 1,595 participants) were selected from the original 16 in order to focus on the influence of MHO and MUO on the development of CVD as well as due to the relatively large number of participants in each group:

a) Stable MHN (MHN at both examinations in the 2002 and 2007 survey)

b) MHO to MHN (MHO in the 2002 survey and MHN in the 2007 survey)

c) Stable MHO (MHO at both examinations in the 2002 survey and the 2007 survey)

d) MHO to MUO (MHO in the 2002 survey and MUO in the 2007 survey)

e) Stable MUO (MUO at both examinations in the 2002 survey and the 2007 survey)

f) MUO to MHO (MUO in the 2002 survey and MHO in the 2007 survey)

Clinical Characteristics at the 2007 Survey according to the Transition Patterns of Metabolic Health Status from the 2002 Survey to the 2007 Survey

Table 3 exhibits the clinical characteristics at the 2007 survey according to the transition patterns in metabolic health status among 1,595 participants selected as described above. The mean values of systolic blood pressure, diastolic blood pressure, fasting plasma glucose, serum LDL cholesterol, WC, and BMI; the median values of serum TG, serum hs-CRP and HOMA-IR; and the frequency of use of glucose-lowering agents were significantly higher in participants in the MHO to MUO group than those in the stable MHN group. Similar differences were observed in participants in the stable MUO group and in participants in the MUO to MHO group, except that there was no significant difference in serum LDL cholesterol in either of these groups compared to the MHN group, and the mean age and frequency of chronic kidney disease were higher in participants in the stable MUO group than in those in the stable MHN group. In addition, the geometric mean values of serum sdLDL cholesterol and remnant cholesterol in the participants in the MHO to MUO groups and in the stable MUO group were significantly higher than those in the stable MHN group (Supplementary Fig.3).

Table 3.Clinical characteristics at the 2007 survey according to the transition patterns of metabolic health status from the 2002 survey to the 2007 survey

	Transition pattern in metabolic health status from the 2002 survey to the 2007 survey
	Stable MHN (n = 993)	MHO to MHN (n = 46)	Stable MHO (n = 121)	MHO to MUO (n = 71)	Stable MUO (n = 308)	MUO to MHO (n = 56)
Age (years)	63 (11)	67 (12)	63 (10)	63 (10)	65 (10)＊	66 (10)
Male sex (%)	42.1	45.7	43	45.1	42.5	41.1
Systolic blood pressure (mmHg)	126 (18)	126 (19)	129 (16)	139 (15)＊	144 (16)＊	131(14)
Diastolic blood pressure (mmHg)	77 (10)	76 (10)	79 (9)	84 (9)＊	86 (10)＊	80 (7)＊
Use of antihypertensive agents (%)	20.0	32.6＊	25.6	25.4	59.1＊	44.6＊
Fasting plasma glucose (mg/dL)	99 (19)	100 (12)	99 (14)	109 (13)＊	123 (32)＊	104 (34)
Use of glucose-lowering agents (%)	3.1	0.0	5.8	9.9＊	18.5＊	5.4
Serum LDL cholesterol (mg/dL)	124 (29)	120 (27)	124 (29)	137 (32)＊	129 (30)	125 (37)
Serum HDL cholesterol (mg/dL)	73 (17)	67 (17)	68 (15)＊	64 (17)＊	58 (15)＊	62 (14)＊
Serum triglycerides (mg/dL)	87 (64-114)	82 (63-118)	87 (66-105)	126 (97-180)＊	145 (99-195)＊	102 (80-128)
Waist circumference (cm)	81.3 (6.2)	86.2 (5.1)＊	93.0 (7.0)＊	95.9 (6.6)＊	97.5 (7.7)＊	94.5 (6.2)＊
BMI (kg/m2)	21.6 (1.6)	23.9 (0.8)＊	27.0 (1.9)＊	27.2 (1.8)＊	28.4 (3.0)＊	27.0 (1.6)＊
Chronic kidney disease (%)	10.0	15.2	13.2	11.3	21.6＊	16.1
Electrocardiogram abnormalities (%)	14.0	15.2	11.6	14.1	14.3	16.1
Serum hs-CRP (mg/L)	0.30 (0.15-0.62)	0.52 (0.18-0.96)	0.57 (0.31-1.08)＊	0.68 (0.34-1.04)＊	0.75 (0.45-1.37)＊	0.59 (0.35-0.97)＊
HOMA-IR	1.06 (0.77-1.42)	1.11 (0.87-1.61)	1.58 (1.14-2.14)＊	2.23 (1.61-3.33)＊	2.69 (1.95-4.25)＊	1.75 (1.38-2.44)＊
Fatty liver index	13 (7-22)	18 (14-30)＊	39 (26-52)＊	59 (42-75)＊	63 (49-80)＊	43 (37-58)＊
Current smoking (%)	17.9	13.0	17.4	15.5	16.9	8.9
Current drinking (%)	49.1	45.7	51.2	46.5	48.7	35.7
Regular exercise (%)	28.8	39.1＊	27.3	31.0	35.7	37.5

Abbreviations: BMI, body mass index; LDL, low density lipoprotein; HDL, high-density lipoprotein; HOMA-IR, homeostasis model assessment of insulin resistance; hs-CRP, high-sensitivity C-reactive protein; MHN, metabolically healthy normal weight; MHO, metabolically healthy obesity; MUN, metabolially unhealthy normal weight; MUO, metabolically unhealthy obesity.

Data are presented as the mean values (standard deviation), percentages, or median (interquartile range).

^＊p＜0.05 vs. the Stable MHN group.

Supplementary Fig.3. Geometric means of serum lipid levels at the 2007 survey according to groups of transition pattern in metabolic health status

^＊p＜0.05 vs Stable MHN group

Abbreviations: MHN, metabolically healthy normal weight; MHO, metabolically healthy obesity; MUO, metabolically unhealthy obesity; sdLDL, small dense low-density lipoprotein.

Incidence of CVD According to the Transition Patterns of Metabolic Health Status

During the follow-up period, a total of 104 participants developed a CVD event. Fig.3 shows that the age- and sex-adjusted cumulative incidence of CVD was significantly higher in participants in the stable MUO group and in those who transitioned from MHO to MUO than those in the stable MHN group. Table 4 demonstrates the age- and sex-adjusted incidence rates and multivariable-adjusted HRs for the development of CVD by the transitional categories in metabolic health status. The incidence rate of CVD (unit: per 1000 person-years) for each group of the transition pattern of metabolic health status from the 2002 survey to the 2007 survey was as follows: 6.9 for the stable MHN group; 5.6 for the MHO to MHN group; 6.4 for the stable MHO group; 12.3 for the MHO to MUO group; 11.1 for the stable MUO group; and 12.5 for the MUO to MHO group. The risk of developing CVD was significantly higher at 2.18-times in the MHO to MUO group and 1.75-times higher in the stable MUO group than in the stable MHN group after adjusting for age, sex, chronic kidney disease, echocardiogram abnormality, current smoking, current drinking, and regular exercise (Model 2). However, there was no evidence of the significant increased risk for CVD in the other groups, including the MHO to MHN group, the stable MHO group, and the MUO to MHO group. These risks of developing CVD were reduced by further adjustment of serum LDL cholesterol, especially serum sdLDL cholesterol. Furthermore, we adjusted with serum remnant cholesterol and fatty liver index in addition to Model 2 (Supplementary Table 1). As a result, the association between the transition of metabolic health status and the risk of cardiovascular disease were attenuated as well.

Fig.3. Age- and sex-adjusted cumulative incidence of cardiovascular disease according to the transition patterns in metabolic health status, 2007–2017

^＊p＜0.05 vs the Stable MHN group.

Abbreviations: MHN, metabolically healthy normal weight; MHO, metabolically healthy obesity; MUO, metabolically unhealthy obesity.

Table 4.Hazard ratios for the development of cardiovascular disease according to the transition patterns in metabolic health status from the 2002 survey to the 2007 survey during the follow-up period (2007–2017)

	Transition pattern in metabolic health status from the 2002 survey to the 2007 survey
	Stable MHN (n = 993)	MHO to MHN (n = 46)	Stable MHO (n = 121)	MHO to MUO (n = 71)	Stable MUO (n = 308)	MUO to MHO (n = 56)
Cardiovascular disease
Number of events/participants	57/993	2/46	6/121	8/71	31/308	5/56
Age- and sex-adjusted incidence rate (per 103 PYs)	6.9	5.6	6.4	12.3	11.1	12.5
Hazard ratio (95% CI)
Model 1: Age- and sex-adjusted	1.00 (reference)	0.54 (0.13-2.21)	0.91 (0.39-2.12)	2.16 (1.03-4.53)＊	1.74 (1.12-2.70)＊	1.44 (0.58-3.59)
Model 2: Multivariable-adjusted	1.00 (reference)	0.52 (0.13-2.15)	0.93 (0.40-2.15)	2.18 (1.04-4.59)＊	1.75 (1.12-2.72)＊	1.37 (0.55-3.43)
Model 2 + serum LDL cholesterol	1.00 (reference)	0.52 (0.13-2.16)	0.93 (0.40-2.17)	2.00 (0.95-4.21)	1.67 (1.07-2.60)＊	1.41 (0.57-3.54)
Model 2 + log-transformed serum sdLDL cholesterol	1.00 (reference)	0.53 (0.13-2.19)	0.94 (0.40-2.17)	1.75 (0.82-3.73)	1.42 (0.89-2.26)	1.34 (0.54-3.34)

Abbreviations: MHN, metabolically healthy normal weight; MHO, metabolically healthy obesity; MUN, metabolically unhealthy normal weight; MUO, metabolically unhealthy obesity; LDL, low density lipoprotein; sdLDL, small dense low-density lipoprotein; PYs, patient years; CI, confidence interval.

^＊p＜0.05 vs. the Stable MHN group.

Model 2: Adjusted for age, sex, chronic kidney disease, echocardiogram abnormality, current smoking, current drinking, and regular exercise.

Supplementary Table 1.Hazard ratios for the development of cardiovascular disease according to the transition patterns in metabolic health status from the 2002 survey to the 2007 survey during the follow-up period (2007–2017) after additional adjustment for fatty liver index or serum remnant cholesterol

	Transition in metabolic health status from 2002 to 2007
	Stable MHN (n = 993)	MHO to MHN (n = 46)	Stable MHO (n = 121)	MHO to MUO (n = 71)	Stable MUO (n = 308)	MUO to MHO (n = 56)
Cardiovascular disease
Number of events/participants	57/993	2/46	6/121	8/71	31/308	5/56
Age- and sex-adjusted incidence rate (per 103 PYs)	6.9	5.6	6.4	12.3	11.1	12.5
Hazard ratio (95% CI)
Model 2 + log-transformed serum remnant cholesterol	1.00 (reference)	0.52 (0.13-2.14)	0.92 (0.40-2.12)	2.11 (0.995-4.48)	1.70 (1.07-2.71)＊	1.35 (0.54-3.39)
Model 2 + log-transformed fatty liver index	1.00 (reference)	0.47 (0.11-1.96)	0.75 (0.31-1.81)	1.64 (0.72-3.73)	1.27 (0.71-2.30)	1.05 (0.40-2.79)

Abbreviations: MHN, metabolically healthy normal weight; MHO, metabolically healthy obesity; MUN, metabolically unhealthy normal weight; MUO, metabolically unhealthy obesity; LDL, low density lipoprotein; sdLDL, small dense low-density lipoprotein; PYs, patient years; CI, confidence interval.

^＊p＜0.05 vs. the MHN group.

Model 2: Adjusted for age, sex, chronic kidney disease, echocardiogram abnormality, current smoking, current drinking, and regular exercise.

Changes of the MetS Component and the Fatty Liver Index, and the Values of Lipids at Baseline according to Transition of Metabolic Health Status

The changes in the frequencies of participants having each MetS component according to transition patterns of metabolic health status are shown in Supplementary Table 2. The frequencies of participants with abdominal obesity increased during 5 years in the stable MHN group, the stable MHO group, the MHO to MUO group, and the stable MUO group. As compared to the 2002 survey, the frequencies of participants with elevated blood pressure, elevated fasting plasma glucose, and elevated serum TG in the 2007 survey were significantly higher in the MHO to MUO groups and significantly lower in the MUO to MHO group. The geometric means of the fatty liver index increased steeply from the 2002 survey to the 2007 survey in the MHO to MUO group (Supplementary Fig.4).

Supplementary Table 2.Changes in the frequencies of participants having each metabolic syndrome component according to transition patterns of metabolic health status, 2002–2007

Metabolic syndrome components	Transition pattern of metabolic health status from the 2002 survey to the 2007 survey
	Stable MHN (n = 993)	MHO to MHN (n = 46)	Stable MHO (n = 121)	MHO to MUO (n = 71)	Stable MUO (n = 308)	MUO to MHO (n = 56)
Abdominal obesity (%)	18.2 to 34.4＊	50.0 to 54.4	66.9 to 82.6＊	67.6 to 97.2＊	92.9 to 97.4＊	89.3 to 87.5
Elevated blood pressure (%)	34.0 to 43.9＊	41.3 to 47.8	38.8 to 44.6	28.2 to 84.5＊	92.9 to 95.8＊	85.7 to 60.7＊
Elevated fasting plasma glucose (%)	53.9 to 30.8＊	63.0 to 32.6＊	47.1 to 27.3＊	73.2 to 88.7＊	92.5 to 88.0＊	85.7 to 21.4＊
Elevated serum triglycerides (%)	8.5 to 7.9	4.3 to 4.3	7.4 to 5.8	7.0 to 42.3＊	48.4 to 47.7	44.6 to 7.1＊
Low serum HDL cholesterol (%)	3.5 to 2.1＊	6.5 to 0.0†	0.8 to 0.8	4.2 to 7.0	22.4 to 22.1	23.2 to 0.0†

Abbreviations: HDL, high-density lipoprotein; MHN, metabolically healthy normal weight; MHO, metabolically healthy obesity; MUO, metabolically unhealthy obesity.

Data are presented as proportion of the participants in the 2002 survey to the 2007 survey.

^＊p＜0.05 vs. the 2002 survey (for each transition status).

^†The test of difference could not be conducted due to an empty category.

Supplementary Fig.4. Geometric mean of fatty liver index at the 2002 and 2007 survey and their changes according to groups of transition pattern in metabolic health status from the 2002 survey to the 2007 survey

^＊p＜0.05 vs fatty liver indexat the 2002 survey

Abbreviations: MHN, metabolically healthy normal weight; MHO, metabolically healthy obesity; MUO, metabolically unhealthy obesity.

Discussion

The present study demonstrated that individuals with MUO were at higher risk for developing CVD than those with MHN, while no significant excess risk of developing CVD was observed in those with MHO. On the other hand, approximately 30% of the individuals with MHO transitioned to MUO over a 5-year period, and these individuals who transitioned from MHO to MUO (the MHO to MUO group) had a significantly higher risk of CVD than either those who maintained MHN (the stable MHN group) or those who remained MHO (the stable MHO group). However, individuals who maintained MHO (the stable MHO group) or those who transitioned from MUO to MHO did not show the increased risk for developing CVD. Furthermore, the individuals who transitioned from MHO to MUO had higher serum sdLDL cholesterol levels and exhibited an increase in the fatty liver index over the 5 years, and the magnitude of excess risk of developing CVD in individuals who transitioned from MHO to MUO was attenuated by adjusting for serum sdLDL cholesterol levels. These findings suggest that obesity combined with metabolic disorders including elevated blood pressure, elevated blood glucose, and lipid abnormalities increases the risk of developing CVD, possibly due to increased serum sdLDL cholesterol, and the exacerbation of fatty liver may be involved in the combination of obesity and metabolic disorders.

Several epidemiologic studies have reported that individuals with MHO combined with metabolic disorders have a higher risk of CVD than those in the stable MHO group¹⁷⁾. The Multi-Ethnic Study of Atherosclerosis¹⁸⁾ and the UK Biobank¹⁹⁾ found that no increased risk of CVD was observed in the stable MHO group, but there was an increased risk of CVD in those who transitioned from MHO to MUO. A study from China also reported an increased risk of CVD in the group who transitioned from MHO to MUO, although there was no such increase in the stable MHO group²⁰⁾. These results were largely consistent with our findings. In a study using data from the Korean Genome and Epidemiology Study (KoGES) cohort, on the other hand, the risk for developing CVD increased in the stable MHO group²¹⁾. This result differs from ours, but this difference may be due to differences in the study design: in KoGES, the time period for assessing change in metabolic health status and the onset of CVD was the same, whereas the present study assessed the change in the metabolic health status in the 5 years prior to the start of the follow-up period for developing CVD.

The exact mechanism underlying the difference in the effects of MHO and MUO on the risk of CVD has not been fully elucidated. In a genetically modified mouse model of MHO, a large accumulation of fat in sites other than visceral tissues was observed, along with increased expression of adipokines that improve metabolism^{22, 23)}. In humans with MHO, a reduced infiltration of macrophages into adipose tissue has been observed²⁴⁾. Individuals with MHO have higher levels of adipokines that promote metabolism. From these findings, it is suggested that tissues other than visceral fat, such as muscles and subcutaneous fat, increase and cytokines and adipokines work towards maintaining a healthy metabolic state in the MHO group. This would be one of the mechanisms accounting for the increased risk for the development of CVD in participants with MUO, although it was not observed in those with MHO.

In the present study, serum sdLDL cholesterol and remnant cholesterol levels were notably high at the 2007 survey among participants in the MHO to MUO group and in the stable MUO group. Additionally, when we adjusted with serum sdLDL cholesterol and serum remnant cholesterol in addition to conventional risk factors, the association between the transition of metabolic health status and the risk of cardiovascular disease were attenuated. Serum sdLDL cholesterol and remnant cholesterol have highly atherogenic potential and are known as biomarkers of atherosclerotic cardiovascular disease^{25, 26)}. Furthermore, the fatty liver index, which is a useful indicator for fatty liver²⁷⁾, decreased significantly from the 2002 survey to the 2007 survey among participants who transitioned from MHO to MHN, while those who transitioned from MHO to MUO exhibited an apparent increase in the fatty liver index. In fact, there is a population-based cohort study showing that the presence of nonalcoholic fatty liver disease (NAFLD) at baseline influences the transition of MHO to MUO²⁸⁾. These findings suggest that fatty liver is implicated in the complication of metabolic abnormalities such as elevated blood pressure, elevated blood glucose, and lipid abnormalities to obesity, and that prevention and management of fatty liver is important for reducing CVD risk in individuals with obesity.

The strengths of this study include its longitudinal population-based design, high participation rate and follow-up rate, and accurate diagnoses of CVD. Moreover, we could evaluate the transition of metabolic health status between two points using the same definition due to longitudinal measures of the components of MetS. However, some potential limitations of this study should be noted. First, the transition patterns of metabolic health status were determined on the basis of two health check-ups before the start of the follow-up survey, without accounting for the health status during the follow-up period. This limitation could lead to misclassification of the transition pattern of metabolic health status, which may weaken the association found in the present study. Second, we did not have available information on the causes of changes in metabolic health status. For example, weight or body fat loss due to malignancy is not considered as healthy weight loss. Therefore, the present study could not consider the influence of the causes of changing metabolic health status on the CVD risk. Third, approximately 15% of the study population was excluded from the analysis of metabolic health status transitions due to the absence of data from the 2002 survey. The participants included in the analyses were elder and had higher systolic blood pressure and waist circumference, factors that could predispose them to developing MetS during the 5-year period. Furthermore, 373 individuals, who had no history of CVD as of the 2002 survey, participated in the 2002 survey without history of cardiovascular disease but did not participate in the 2007 survey. Of these, the numbers of deaths, cardiovascular events, and cases of lost to follow-up over the five years from 2002 to 2007 were 130, 17, and 226, respectively, over the five-year period from 2002 to 2007. These selection biases may have affected the magnitude of the association between the transition of metabolic health status and the risks of cardiovascular disease to some extent. However, we believe that they would not have substantially altered our findings. Finally, the generalizability of our findings may be limited, since these analyses were conducted in only one cohort of Japanese.

Conclusions

The present study demonstrated that individuals with MHO itself were unlikely to have excess risk of developing CVD, but approximately 30% of the individuals with MHO transitioned to MUO over a 5-year period, and these individuals who transitioned from MHO to MUO were subsequently at greater risk of CVD. Therefore, it may be important to prevent complications of metabolic abnormalities when attempting to reduce the risk of developing CVD among individuals with obesity, and the prevention and close management of fatty liver would be valuable for this purpose.

Acknowledgements and Notice of Grant Support

The authors thank the residents of the town of Hisayama for their participation in the survey and the staff of the Division of Health of Hisayama for their cooperation with this study. The statistical analyses were carried out using the computer resources offered under the category of General Projects by the Research Institute for Information Technology, Kyushu University.

This study was supported in part by the Ministry of Education, Culture, Sports, Science and Technology of Japan (JSPS KAKENHI Grant Number JP22K07421, JP22K17396, JP23K09692, JP23K09717, JP23K16330, JP23K06787, and JP23K09060); by the Health and Labour Sciences Research Grants of the Ministry of Health, Labour and Welfare of Japan (JPMH23FA1006, JPMH23FA1022, and JPMH24GB1002); by the Japan Agency for Medical Research and Development (JP24dk0207053, JP24km0405209, and JP24tm0524003); by Japan Science and Technology (JPMJPF2210); by DENKA Co., Ltd. (Tokyo, Japan); and by Eli Lilly Japan K.K. (Kobe, Japan).

Conflicts of Interest

The authors declare that Toshiharu Ninomiya received research funding from Eli Lilly & Co. and DENKA Co., Ltd, and that the other authors have no conflicts of interest.

References

• 1)  Blüher M: Obesity: global epidemiology and pathogenesis. Nat Rev Endocrinol, 2019; 15: 288-298
• 2)  Guh DP, Zhang W, Bansback N, Amarsi Z, Birmingham CL, Anis AH: The incidence of co-morbidities related to obesity and overweight: a systematic review and meta-analysis. BMC Public Health, 2009; 9: 88
• 3)  Honda T, Ishida Y, Oda M, Noguchi K, Chen S, Sakata S, Oishi E, Furuta Y, Yoshida D, HirakawaY, Hata J, Kitazono T, Ninomiya T: Changes in body weight and concurrent changes in cardiovascular risk profiles in community residents in Japan: the Hisayama Study. J Atheroscler Thromb, 2022; 29: 252-267
• 4)  Riaz H, Khan MS, Siddiqi TJ, Usman MS, Shah N, Goyal A, Khan S, Mookadam F, Krasuski RA, Ahmed H: Association Between Obesity and Cardiovascular Outcomes: A Systematic Review and Meta-analysis of Mendelian Randomization Studies. JAMA Netw Open, 2018; 1: e183788
• 5)  Flegal KM, Kit BK, Orpana H, Graubard BI: 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
• 6)  Huang MY, Wang MY, Lin YS, Lin CJ, Lo K, Chang IJ, Cheng TY, Tsai SY, Chen HH, Lin CY, Liu SJ, Chien KL, Yeh TL: The association between metabolically healthy obesity, cardiovascular disease, and all-cause mortality risk in asia: a systematic review and meta-analysis. Int J Environ Res Public Health, 2020; 17: 1320
• 7)  Itoh H, Kaneko H, Kiriyama H, Kamon T, Fujiu K, Morita K, Michihata N, Jo T, Takeda N, Morita H, Yasunaga H, Komuro I: Metabolically healthy obesity and the risk of cardiovascular disease in the general population analysis if a nationwide epidemiological database. Circ J, 2021; 85: 914-920
• 8)  Iacobini C, Pugliese G, Fantauzzi CB, Federici M, Menini S: Metabolically healthy versus metabolically unhealthy obesity. Metabolism, 2019; 92: 51-60
• 9)  Ninomiya T: Japanese legacy cohort studies; the Hisayama Study. J Epidemiol, 2018; 28: 444-451
• 10)  Alberti KG, Eckel RH, Grundy SM, Zimmet PZ, Cleeman JI, Donato KA, Fruchart JC, James WPT, Loria CM, SmithJr SC: Harmonizing the metabolic syndrome: a joint interim statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity. Circulation, 2009; 120: 1640-1645
• 11)  AlbertiKG, ZimmetP, ShawJ: Metabolic syndrome-a new worldwide definition. A consensus statement from the International Diabetes Federation; DiabetMed, 2006; 23: 469-480
• 12)  Hata J, Ninomiya T, Hirakawa Y, Nagata M, Mukai N, Gotoh S, Fukuhara M, Ikeda F, Shikata K, Yoshida D, Yonemoto K, Kamouchi M, Kitazono T, Kiyohara Y: Secular trends in cardiovascular disease and its risk factors in Japanese: half-century data from the Hisayama Study (1961-2009). Circulation, 2013; 128: 1198-1205
• 13)  Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC: Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia, 1985; 28: 412-419
• 14)  Horio M, Imai E, Yasuda Y, Watanabe T, Matsuo S: Modification of the CKD epidemiology collaboration (CKD-EPI) equation for Japanese: accuracy and use for population estimates. Am J Kidney Dis, 2010; 56: 32-38
• 15)  Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group. KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int Suppl, 2013; 3: 1-150
• 16)  Bedogni G, Bellentani S, Miglioli L, Masutti F, Passalacqua M, Castigilone A, Tiribelli C: The fatty liver index: a simple and accurate predictor of hepatic steatosis in the general population. BMC Gastroenterol, 2006; 6: 33
• 17)  Abiri B, Koohi F, Ebadinejad A, Valizadeh M, Hosseinpanah F: Transition from metabolically healthy to unhealthy overweight/obesity and risk of cardiovascular disease incidence: A systematic review and meta-analysis. Nutr Metab Cardiovasc Dis, 2022; 32: 2041-2051
• 18)  Mongraw-Chaffin M, Foster MC, Anderson CAM, Burke GL, Haq N, Kalyani RR, Ouyang P, Sibley CT, Tracy R, Woodward M, Vaidya D: Metabolically healthy obesity, transition to metabolic syndrome, and cardiovascular risk. J Am Coll Cardiol, 2018; 71: 1857-1865
• 19)  Zhou Z, Macpherson J, Gray SR, Gill JMR, Welsh P, Celis-Morales C, Sattar N, Pell JP, Ho FK: Are people with metabolically healthy obesity really healthy? A prospective cohort study of 381,363 UK Biobank participants. Diabetologia, 2021; 64: 1963-1972
• 20)  Gao M, Lv J, Yu C, Guo Y, Bian Z, Yang R, Du H, Yang L, Chen Y, Li Z, Zhang X, Chen J, Qi L, Chen Z, Huang T, Li L; China Kadoorie Biobank (CKB) Collaborative Group: Metabolically healthy obesity, transition to unhealthy metabolic status, and vascular disease in Chinese adults: a cohort study. PLoS Med, 2020; 17: e1003351
• 21)  Kim NH, Seo JA, Cho H, Seo JH, Yu JH, Yoo HJ, Kim SG, Choi KM, Baik SH, Choi DS, Shin C, Cho NH: Risk of the development of diabetes and cardiovascular disease in metabolically healthy obese people: the Korean Genome and Epidemiology Study. Medicine, 2016; 95: e3384
• 22)  Kim JY, van de Wall E, Laplante M, Azzara A, Trujillo ME, Hofmann SM, Schraw T, Durand JL, Li H, Li G, Jelicks LA, Mehler MF, Hui DY, Deshaies Y, Shulman GI, Schwartz GJ, Scherer PE: Obesity-associated improvements in metabolic profile through expansion of adipose tissue. J Clin Invest, 2007; 117: 2621-2637
• 23)  Kusminski CM, Holland WL, Sun K, Park J, Spurgin SP, Lin Y, Askew GR, Simcox JA, McClain DA, Scherer CL, Scherer PE: MitoNEET-driven alterations in adipocyte mitochondrial activity reveal a crucial adaptive process that preserves insulin sensitivity in obesity. Nat Med, 2012; 18: 1539-1549
• 24)  Klöting N, Fasshauer M, Dietrich A, Kovacs P, Schön MR, Kern M, Stumvoll M, Blüheret M: Insulin-sensitive obesity. Am J Physiol Endocrinol Metab, 2010; 299: E506-E515
• 25)  Liou L, Kaptoge S: Association of small, dense LDL-cholesterol concentration and lipoprotein particle characteristics with coronary heart disease: A systematic review and meta-analysis. PLoS One, 2020; 15: e0241993
• 26)  Delialis D, Georgiopoulos G, Aivalioti E, Konstantaki C, Oikonomou E, Bampatsias D, Mavraganis G, Vardavas C, Liberopoulos E, Stellos K, Stamatelopoulos K: Remnant cholesterol in atherosclerotic cardiovascular disease: A systematic review and meta-analysis. Hellenic J Cardiol, 2023; 74: 48-57
• 27)  Takahashi S, Tanaka M, Higashiura Y, Mori K, Hanawa N, Ohnishi H, Furuhashi M: Prediction and validation of nonalcoholic fatty liver disease by fatty liver index in a Japanese population. Endocr J, 2022; 69: 463-471
• 28)  Kouvari M, Chrysohoou C, Skoumas J, Pitsavos C, Panagiotakos DB, Mantzoros CS; ATTICA study Investigator: The presence of NAFLD influences the transition of metabolically healthy to metabolically unhealthy obesity and the ten-year cardiovascular disease risk: A population-based cohort study. Metabolism, 2022; 128: 154893