Clinical and Preclinical Obesity in Korean Adults from 2014 to 2023
Article information
Abstract
Background
A new diagnostic framework for clinical obesity, proposed by the Lancet Commission, defines obesity as excess adiposity with organ dysfunction, offering a more functional assessment than traditional body mass index (BMI)-based classification. However, it has not been applied to Asian populations, where obesity-related complications arise at lower BMI thresholds.
Methods
We analyzed 57,863 Korean adults aged ≥20 years from the 2014 to 2023 Korea National Health and Nutrition Examination Survey (KNHANES), a nationally representative, cross-sectional dataset. Clinical obesity was defined as excess adiposity with ≥1 obesity-related complication or functional limitation; preclinical obesity was defined as excess adiposity without complications. Age, sex, and temporal trends were examined and compared with BMI-based classifications using complex sampling weights and age-standardization.
Results
The prevalence of clinical obesity and preclinical obesity was 31.2% and 8.1%, respectively. Among those with BMI-defined obesity (≥25.0 kg/m2), 20.1% had no complications, while 19.4% of overweight individuals (BMI 23.0–24.9 kg/m2) met clinical obesity criteria. Clinical obesity increased with age despite stable BMI, driven by metabolic and functional decline. Cancer prevalence was highest among individuals with clinical obesity (1.77%). Applying Western waist circumference cutoffs drastically reduced the estimated prevalence of clinical obesity to 13.1%. Longitudinal analyses showed a rising trend in clinical obesity over ime.
Conclusion
This is the first large-scale study to apply the clinical obesity framework in an Asian population. The framework identifies high-risk individuals missed by BMI alone and supports a shift toward more functional, ethnically tailored obesity definitions for risk stratification and intervention.
Highlights
• BMI and Lancet clinical obesity showed discordance in 57,863 Korean adults.
• 19.4% of overweight adults met criteria for clinical obesity.
• Metabolic impairment characterized 83.6% of clinical obesity cases.
• Clinical obesity increased with age, despite stable BMI across decades.
• Long-term validation of the clinical obesity framework is warranted.
INTRODUCTION
According to the World Obesity Federation, 24% of the global population is projected to be living with obesity by 2035 [1]. In the Asia-Pacific region, which comprises 60% of the world’s population, an estimated 40% of adults are currently classified as overweight or having obesity, based on body mass index (BMI) categories [1]. Early identification of overweight and obesity, or detection of progression toward these states, is crucial, as they are associated with reduced quality of life and an increased risk of adverse health outcomes, including cardiovascular disease, certain cancers, and mortality [2,3].
Obesity is conventionally defined as excessive fat accumulation that poses health risks. BMI has long served as the primary diagnostic tool in both clinical and public health contexts. However, BMI has several well-recognized limitations: it does not differentiate between fat and lean mass, fails to capture regional body fat distribution, and does not reflect metabolic health [4]. These limitations are particularly pronounced in Asian populations, where individuals with BMI values below traditional obesity thresholds may still exhibit excess adiposity and elevated cardiometabolic risk [5]. A Korean study found that 32% of individuals with a BMI of 18.5–22.9 kg/m2 had body fat levels exceeding 26% in men or 36% in women, conferring a higher risk of metabolic syndrome [6].
To address these limitations, the ‘Lancet Diabetes & Endocrinology Commission’ recently proposed a redefinition of obesity as a spectrum of disease states [7]. This novel framework introduces the categories of preclinical and clinical obesity, representing a shift away from weight-based thresholds towards functional and metabolic assessments. It emphasizes that individuals with increased risk—due to metabolic dysfunction, unfavorable body fat distribution, or genetic predisposition—may benefit from early intervention, whereas others without increased health risk may only require regular monitoring.
In this study, we aim to investigate the prevalence of clinical and preclinical obesity in a nationally representative South Korean adult population using the newly proposed definition [7]. We also explore how this framework improves classification accuracy and highlight important considerations and potential pitfalls when interpreting obesity status through this lens.
METHODS
Study population
We used data from the Korea National Health and Nutrition Examination Survey (KNHANES), a nationally representative, cross-sectional survey conducted by the Korea Disease Control and Prevention Agency (KDCA) [8]. KNHANES employs a stratified, multistage, clustered probability sampling design to collect information on health status, nutritional habits, and sociodemographic factors among the non-institutionalized civilian population of South Korea. Adults aged ≥20 years who participated in the sixth (2014–2015), seventh (2016–2018), eighth (2019–2021), and ninth (2022–2023) waves of KNHANES were included. Individuals with missing data on BMI were excluded.
A total of 57,863 adults met the inclusion criteria and were analyzed to assess the prevalence and characteristics of clinical obesity (Supplementary Fig. 1). Definitions for clinical obesity, BMI-based obesity, and obesity-related diseases/disorders are provided in the Supplementary Fig. 2 and Supplementary Methods. All participants gave written informed consent.
Statistical analysis
We examined trends in the prevalence of clinical and preclinical obesity across age, sex, and survey waves (2014–2023), along with the burden of obesity-related complications and obesity-related cancers. To account for the confounding effect of aging on prevalence estimates, direct age-standardization was applied. For the overall analyses, age-standardized prevalence of clinical obesity and related complications was calculated using the 2023 Korean population as the reference. For the survey wave-specific analyses, age-standardization was performed using the population structure of the final year within each survey period (2015, 2018, 2021, and 2023, respectively). Estimated numbers of individuals in each category (no obesity, preclinical obesity, and clinical obesity) were derived by applying sampling weights to age-standardized proportions, thereby reflecting the corresponding population size.
All analyses accounted for the complex survey design of KNHANES, incorporating sampling weights, strata, and clusters to ensure national representativeness. Descriptive statistics were calculated according to clinical obesity status. Continuous variables were presented as means with standard deviations, and categorical variables as frequencies with proportions. Group comparisons across clinical obesity categories were performed using one-way analysis of variance (ANOVA) for continuous variables and the chi-square test for categorical variables. The Jonckheere–Terpstra test was used to assess linear trends across ordered groups (P for trend), where appropriate.
All statistical analyses were performed using SPSS version 28 (IBM Corp., Armonk, NY, USA) and R version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria). Two-sided P values <0.05 were considered statistically significant.
Ethics approval
This study was based on publicly available data from the KNHANES, which is conducted by the KDCA. All KNHANES participants provide written informed consent. Since this study involved secondary analysis of de-identified, anonymized data, additional institutional review board (IRB) approval was not required under current Korean regulations.
Patient and public involvement
Patients and members of the public were not involved in the design, conduct, reporting, or dissemination plans of this research. The study was a secondary analysis of anonymized, publicly available data from the KNHANES, which does not include direct patient or public involvement in its analytical processes.
RESULTS
Prevalence and clinical manifestations of clinical obesity
Among the 57,863 individuals included in this nationwide cohort, 31.2% (n=18,074) met the criteria for clinical obesity, while 8.1% (n=4,681) were categorized as having preclinical obesity by the newly proposed functional definition of clinical obesity (Table 1, Supplementary Table 1). These findings were consistent in the age-standardized analysis, confirming robustness across age distributions.
Prevalence and characteristics of individuals meeting the new definition of clinical obesity in the Korean National Health and Nutrition Examination Survey 2014 to 2023
Among the diagnostic components for clinical obesity, metabolic impairment—including hyperglycemia and dyslipidemia—was the most prevalent, affecting 82.6% of individuals with clinical obesity (Table 1, Fig. 1). Hypertension was the second most common manifestation (54.8%), substantially higher than among those without obesity (21.0%). Steatotic liver disease (SLD) with fibrosis was present in 17.1% of participants with clinical obesity, compared to 1.2% among individuals without clinical or preclinical obesity. A full ranked list of obesity-related complications by crude and age-standardized prevalence is provided in Supplementary Table 2.
Distribution of diagnostic components among individuals with and without clinical obesity. The proportion of individuals with metabolic impairment, obesity-related disorders (including hypertension, steatotic liver disease [SLD] with fibrosis, chronic kidney disease [CKD], asthma, and arthritis, obstructive sleep apnea [OSA], and activity limitation) among those with clinical obesity compared to those without, based on the new clinical obesity definition. aP<0.05.
Additional complications were more frequent in the clinical obesity group: chronic kidney disease (CKD; 10.0% vs. 4.2%), asthma (4.4% vs. 2.6%), obstructive sleep apnea (0.5% vs. 0.2%), and arthritis (18.0% vs. 7.6%) (all P<0.001). Limitations in daily activities were also higher in individuals with clinical obesity (10.8% vs. 6.0%, P<0.001). Among individuals with clinical obesity, 39.2% met one diagnostic criterion, 33.9% met two, 18.3% met three, 8.4% met four or more.
Based on the conventional BMI classification for Asian populations, 23.0% were overweight (BMI 23.0–24.9 kg/m2), and 35.3% had obesity (BMI ≥25.0 kg/m2), with subclassification into class I (25.0–29.9 kg/m2, 29.6%), class II (BMI 30.0–34.9 kg/m2, 4.9%), and class III obesity (BMI ≥35.0 kg/m2, 0.8%) (Supplementary Table 3). The proportion of individuals reclassified as having clinical obesity increased with BMI category: 71.9% in class I obesity, 81.4% in class II obesity, and 84.4% in class III obesity met criteria for clinical obesity.
However, substantial discordance was observed (Fig. 2). Among individuals with BMI ≥25.0 kg/m2, only 73.5% met the new criteria for clinical obesity, whereas 20.1% were classified as having preclinical obesity and 6.4% as having no obesity (Supplementary Table 4). Even in the overweight range (BMI 23.0–24.9 kg/m2), 19.4% were reclassified as having clinical obesity, and 3.8% as preclinical obesity. Conversely, among those categorized as having no obesity by the new clinical obesity definition, 25.9% had a BMI of 23.0–24.9 kg/m2, and even 12.6% had a BMI ≥25.0 kg/m2.
Relationship between body mass index (BMI) categories and clinical obesity classification. (A) Distribution of individuals classified as non-obesity, preclinical obesity, and clinical obesity according to BMI categories. (B) Distribution of BMI categories (<23, 23.0–24.9, and ≥25 kg/m2) within each group defined by the new clinical obesity classification.
Age-specific trends in clinical obesity
The prevalence of clinical obesity increased progressively with age, from 10.3% among individuals in their 20s to 46.9% among those in their 70s (Table 2, Supplementary Table 1). In contrast, the prevalence of preclinical obesity showed an inverse pattern, declining from 12.0% in the third decade of life to 2.3% in the 70s. Across all age groups, the burden of obesity-related organ dysfunction and physical activity limitation increased with age, and was consistently higher in the clinical obesity group compared to those without obesity (Fig. 3, Supplementary Fig. 3).
Age-specific trends in obesity prevalence and anthropometric measures based on BMI alone and the new definition of clinical obesity in Korean National Health and Nutrition Examination Survey 2014 to 2023
Age-standardized prevalence of major diagnostic components of clinical obesity. Comparison of the age-standardized prevalence of key components defining clinical obesity—(A) metabolic impairment, (B) hypertension, (C) steatotic liver disease (SLD) with fibrosis, and (D) chronic kidney disease (CKD)—between individuals with no obesity and those with clinical obesity. Prevalence estimates were age-standardized using the 2023 Korean population structure to account for demographic differences across age groups. Other obesity-related conditions, including asthma, arthritis, obstructive sleep apnea, and activity limitation, are presented separately in Supplementary Fig. 3.
Sex-specific differences in clinical obesity
Clinical obesity was more common in men than in women in both crude and age-standardized analyses (Supplementary Table 5). Among individuals with clinical obesity, metabolic impairment was more common in men than in women (86.8% vs. 79.6%, P<0.001, men and women, respectively). Men were more likely to have low high-density lipoprotein cholesterol (HDL-C; 55.6% vs. 49.9%, P<0.001) and high hypertriglyceridemia (52.8% vs. 50.4%, P=0.001). Conversely, crude analyses suggested a higher prevalence of hypertension in women compared to men (59.0% vs. 50.6%, P<0.001), a pattern likely influenced by the older average age of women in the sample. After applying age-standardization to account for this imbalance, the trend was reversed, with hypertension more prevalent in men (40.5% vs. 38.9%, P<0.001). Women also had a higher prevalence of SLD with fibrosis (14.4% vs. 19.7%, P<0.001), arthritis (6.9% vs. 29.0, P<0.001), and asthma (3.4% vs. 5.4%, P<0.001), and these trends persisted after age-standardization. In contrast, men showed a higher prevalence of CKD (11.0% vs. 9.0%) and obstructive sleep apnea (0.8% vs. 0.1%). Limitations in daily activities were more common among women than men with clinical obesity (8.1% vs. 13.5%, P<0.001), indicating a greater burden of functional impairment in women.
Temporal trends in clinical obesity
The prevalence of clinical obesity increased from 28.6% in 2014–2015 to 34.5% in 2019–2021, coinciding with a rise in metabolic impairment from 44.1% to 49.5% (P for trend <0.001) (Supplementary Table 6). This upward trend was also observed in age-standardized estimates, which rose from 24.3% to 30.5% over the same period, indicating that the increase was not solely attributable to population aging.
However, this upward trend reversed in 2022–2023, with crude prevalence declining to 31.3% and age-standardized prevalence to 27.5% (both P<0.001). This recent decline was accompanied by improvements in health behaviors and metabolic markers. In 2022–2023, physical activity levels increased (P<0.001), and the proportion of individuals with activity limitation decreased from 7.7% in 2019–2021 to 4.4% in 2022–2023 (P<0.001).
In parallel, metabolic markers showed improvement. HDL-C levels increased from 51±12 mg/dL in 2014–2015 to 57±16 mg/dL (P<0.001). Similarly, the proportion of individuals with metabolic impairment decreased from 49.5% in 2019–2021 to 45.8% in 2022–2023 (P<0.001), alongside reductions glycosylated hemoglobin, lipid profiles, aspartate aminotransferase and alanine aminotransferase levels, contributing to the modest decline in clinical obesity prevalence.
Cancer prevalence and clinical obesity
Obesity-related cancers were most prevalent among those with clinical obesity (1.77%), particularly colorectal and breast cancers (P<0.001), whereas gastric cancer was more common in those without obesity (Supplementary Table 7).
Application of Western anthropometric criteria
To assess the implications of using Western anthropometric cutoffs, we applied waist circumference thresholds of at least 102 cm in men and 88 cm in women to identify excess adiposity. Based on these criteria, 84.4% of participants were classified as having no obesity, 2.5% as having preclinical obesity, and 13.1% as having clinical obesity (Supplementary Table 8).
Individuals with clinical obesity defined by Western cutoffs were substantially older (mean age 60.0 years) than those without obesity (51.6 years) or with preclinical obesity (43.3 years), and were predominantly female (77.5%).
DISCUSSION
The ‘Lancet Commission’ has provided a novel framework for conceptualizing and diagnosing obesity. This study enables comparison between that framework and the traditional, BMI-based classification in a nationally representative sample of 57,863 Korean adults. Using the Commission’s diagnostic approach, we found a lower prevalence of clinical obesity (31.2%) than by traditional Asian BMI-based criteria (35.3%). However, notably, 19.4% of individuals categorized as overweight (BMI 23.0–24.9 kg/m2) met the diagnostic criteria for clinical obesity, suggesting that the new definition may offer greater clinical sensitivity than BMI alone. The framework appears more effective in identifying individuals with medical or functional impairments who may benefit from more intensive treatment approaches. Conversely, 26.5% of individuals with BMI ≥25.0 kg/m2—the cutoff for obesity in Asians—were not classified as having clinical obesity: 6.4% were categorized as non-obese and 20.1% as having preclinical obesity. This reinforces the limitations of BMI alone and the value of incorporating metabolic and functional health indicators.
Several influential frameworks have also aimed to move beyond anthropometric thresholds toward clinically meaningful classification. The Edmonton Obesity Staging System (EOSS) pioneered severity staging based on medical, functional, and psychosocial burden, offering clinicians a practical tool to complement BMI when prioritizing treatment intensity [9]. The European Association for the Study of Obesity (EASO) consensus framework further refines this approach by combining an anthropometric gate (ethnicity-appropriate BMI and waist measures) with a structured clinical component encompassing medical, functional, and mental domains [10]. It provides a graded assessment of disease severity according to the degree of health impact, thereby linking diagnostic evaluation to management intensity and individualized treatment targets. In contrast, the ‘Lancet Diabetes & Endocrinology Commission’ distinguishes preclinical from clinical obesity, defining the latter as a chronic systemic disease caused by excess adiposity, evidenced by organ or tissue dysfunction or age-adjusted limitation of daily activities, and recommending that BMI be used primarily for screening with confirmatory adiposity measures at the individual level [7]. These frameworks are complementary: EOSS provides a pragmatic method to stage disease severity once obesity is established, while EASO offers a structured diagnostic and management framework that helps translate the ‘Lancet Commission’s’ conceptual definition into practical use in both clinical and public health settings (Supplementary Table 9). Collectively, these frameworks underscore a shared movement in obesity science toward pathophysiologic and functional definitions, reinforcing the limitations of BMI as a sole diagnostic criterion.
BMI is a measure of body size, not adiposity distribution or related complications. It offers no information on medical or functional comorbidities of obesity. In our analysis, despite nearly identical mean BMI values (27 kg/m2), individuals with clinical obesity showed markedly worse cardiometabolic profiles than those with preclinical obesity, underscoring the limitations of BMI-based classifications.
This analysis reveals the health burden associated with obesity in an aging population. In age-stratified analyses (Table 2), the prevalence of clinical obesity increased with age: 17.8% in those aged 30–39 years (mean BMI 23.9±4.1 kg/m2) vs. 46.9% in those aged 70–80 years (mean BMI 23.9±3.2 kg/m2), despite similar BMI levels. This age-related difference was mainly due to the higher prevalence of metabolic dysfunction, obesity-related disorders, and functional limitations in older adults. To minimize potential age-related confounding, all prevalence estimates for obesity-related diseases were age-standardized using the 2023 Korean population structure, allowing for fairer comparison across age groups and over time.
Glucose regulation deteriorates with age due to progressive β-cell exhaustion [11], while lipid metabolism—particularly HDL-C—is adversely affected by age-related reduction in muscle mass and physical activity [12]. Furthermore, 76.3% of individuals aged 70–80 years with clinical obesity had hypertension, compared to only 10.2% in the 20–29 years group (Fig. 3), although hypertension in older adults is influenced by multiple factors beyond obesity [13,14].
SLD with fibrosis was also more prevalent in elderly individuals with clinical obesity (Fig. 3). SLD is now classified by etiology into metabolic dysfunction-associated SLD, alcohol-related SLD, and other forms (e.g., drug-induced, monogenic, or mixed etiologies). Insulin resistance accompanied by obesity is strongly associated with SLD progression to fibrosis [15,16]. Genetic susceptibility (e.g., patatin-like phospholipase domain-containing 3 [PNPLA3], transmembrane 6 superfamily member 2 [TM6SF2], membrane-bound O-acyltransferase domain-containing 7 [MBOAT7], and glucokinase regulatory protein [GCKR]), dyslipidemia, and gut dysbiosis further modulate hepatic fat accumulation and fibrosis progression [16–18].
CKD was significantly more prevalent in older individuals with clinical obesity. While obesity contributes to CKD through hemodynamic and inflammatory pathways [19], its development also reflects the influence of systemic conditions, environmental exposures, and individual risk factors. Diabetic kidney disease is the leading cause globally, accounting for 30% to 50% of cases [20], and long-standing hypertension also contributes through glomerulosclerosis and arteriolar hyalinosis [21].
BMI remains the most practical and cost-effective initial tool for identifying individuals at risk of obesity-related complications. Waist circumference offers a better estimate of visceral adiposity, although its reliability depends on proper measurement technique and adequate training of health professionals. Both measures are widely available and simple to implement, making them suitable for population-level screening. More precise assessments such as dual-energy X-ray absorptiometry (DXA) or bioelectrical impedance analysis can quantify body composition more accurately but are not feasible for routine clinical use [22,23]. A pragmatic clinical approach may involve using BMI for initial screening, with waist circumference or other indices of fat distribution added in referral settings or for individuals with ambiguous risk profiles. Notably, very few individuals with BMI ≥30.0 kg/m2 (class 2 or 3 obesity) were classified as having neither preclinical nor clinical obesity, suggesting strong alignment between the diagnostic framework and higher BMI categories. This concordance may reflect the greater burden of obesity-related complications at more extreme levels of adiposity. This tiered strategy could help guide further evaluations by clinicians specializing in obesity, such as endocrinologists.
Nevertheless, a key caveat to relying solely on BMI is that a substantial proportion of individuals with BMI <25.0 kg/m2 met the diagnostic criteria for clinical obesity. Conversely, 34.4% of people classified as having no obesity under the new definition had one or more metabolic impairments, and 21.0% had hypertension (Table 1). Additionally, no individual classified as having preclinical obesity had metabolic impairment or an obesity-related disorder, due to the operational definition of clinical obesity. This distinction may lead to underestimation of risk in younger individuals with high BMI and overestimation in older adults due to age-related deterioration in organ function, metabolism, and physical function—for example, a higher prevalence of hypertension, SLD with fibrosis, CKD, arthritis, and activity limitation.
We also observed important sex-specific differences (Supplementary Table 5). While men exhibited higher rates of metabolic impairment, hypertension, and CKD, women with clinical obesity had higher rates of SLD with fibrosis, arthritis, and limitations in daily activities. In the temporal trend analysis, clinical obesity rose over the past decade, peaking in 2019–2021 during the coronavirus disease 2019 (COVID-19) pandemic, likely reflecting the reduced physical activity and sedentary behavior (Supplementary Table 6) [24]. Encouragingly, a modest decline in clinical obesity in 2022–2023 coincided with improved physical activity, HDL-C, liver enzymes, and waist circumference.
Obesity is associated with an increased risk of several cancers. In this study, colorectal, kidney, breast, and cervical cancers were more prevalent among individuals classified with clinical obesity (Supplementary Table 7). Although gastric cancer is common, it is not typically categorized as an obesity-related malignancy because its epidemiologic associations in Korea are stronger with Helicobacter pylori infection and dietary factors than with obesity itself. However, obesity-related cancers were not included as criteria in the definition of clinical obesity, as the temporal relationship between cancer onset and the development of obesity could not be clearly established.
The new clinical obesity definition may be particularly relevant to Asians, where metabolic disorders such as type 2 diabetes mellitus and metabolic syndrome emerge at lower BMI levels compared to Caucasians [25–27]. Asian individuals tend to have greater visceral fat area associated with higher insulin resistance, lower β-cell function, and poorer lipid profiles, even at lower BMI levels [28–30]. In a multiethnic computed tomography (CT)-based study, East Asians exhibited the highest visceral adipose tissue accumulation and the lowest deep subcutaneous fat with increasing adiposity, resulting in the greatest visceral-to-subcutaneous fat ratio after BMI adjustment [31]. Given that abdominal obesity is a stronger predictor of metabolic risk in Asians, waist circumference and visceral adiposity should be prioritized over BMI in obesity diagnosis and risk assessment [26,32].
When Western thresholds for BMI and waist circumference (≥102 cm for male, ≥88 cm for female) were applied, only 11.1% and 2.0% of the population were classified as having clinical obesity and preclinical obesity, respectively, with a notable female predominance (clinical obesity: 8.8% in female vs. 2.3% in men; preclinical obesity: 1.8% in female vs. 0.3% in male) (Supplementary Table 8). These findings underscore the importance of adopting ethnic-specific criteria for defining clinical obesity to ensure accurate risk stratification and intervention planning.
Our study has important clinical implications as the first large-scale application of the new obesity definition in an Asian population. It addresses the key limitations of the BMI-based approach, demonstrates the impact of age-related metabolic and functional factors essential to the diagnosis of clinical obesity, and highlights broader public health perspectives. A recent analysis from the ‘All of Us’ Research Program in the United States estimated the prevalence of clinical obesity at 39.1% and preclinical obesity at 28.9% [33]. The higher prevalence in the United States may reflect both differences in diagnostic thresholds and underlying population characteristics (Supplementary Table 10), underscoring the need for further studies across diverse ethnic groups.
Importantly, preclinical obesity should not be equated with the metabolically healthy obesity (MHO) phenotype. While preclinical obesity reflects the absence of evident complications, MHO has often been defined using heterogeneous criteria and may include individuals with subtle but clinically meaningful cardiometabolic abnormalities [34]. For instance, a population-based study from China reported that the prevalence of MHO ranged from 4.2% to 13.6% depending on the diagnostic criteria used [35]. Similarly, a Japanese study found that only one-fifth of individuals with obesity met the criteria for MHO [36]. These findings reinforce recent evidence suggesting that truly metabolically healthy individuals—those meeting normal thresholds across all six metabolic markers—are relatively uncommon [34]. While this strictly defined MHO group may represent a stable, low-risk phenotype, the long-term trajectory of preclinical obesity remains unclear [37,38]. Future comparative studies across Asian populations employing unified definitions of clinical obesity are warranted to evaluate the generalizability of this framework and its utility in risk stratification.
Some limitations of the present analyses should be acknowledged. The characterization of individuals in the KNHANES cohort could not be mapped with complete precision to the ‘Lancet Commission’s’ diagnostic framework. The ‘Commission’ defines clinical obesity strictly in terms of obesity-induced organ/tissue dysfunction. For example, type 2 diabetes mellitus was not included as obesity-induced, but rather as obesity-related, acknowledging that factors other than excess adiposity may contribute to its development. In the present study, we adopted a pragmatic approach to identifying obesity-related complications, as detailed in the Supplementary Fig. 2 and Supplementary Methods. Certain conditions included in the original framework—such as heart failure, atrial fibrillation, and pulmonary hypertension—were not assessed in the KNHANES dataset and thus excluded from this analysis, potentially leading to an underestimation of clinical obesity. Additionally, obstructive sleep apnea was assessed using the STOP-BANG questionnaire, which may have underestimated its true prevalence [39]. Lastly, direct quantification of excess fat mass using DXA was not available, and clinical assessments of excess or abnormal adiposity were limited to BMI and waist circumference. Other anthropometric parameters, such as body fat percentage or fat distribution (e.g., visceral vs. subcutaneous adiposity), could not be assessed.
In conclusion, adopting a clinically meaningful definition of obesity provides superior risk stratification compared with traditional BMI-based categories, although important caveats remain. This approach reveals the considerable burden of clinical obesity in Asians (at least Koreans) and underscores the need for multidimensional strategies to identify, monitor, and manage individuals at risk. Implementing such frameworks in clinical and public health settings may enable more targeted interventions and improve obesity-related outcomes across diverse populations. Future research should focus on the long-term refinement and validation of the clinical obesity definition and its impact on healthcare access, patient adherence, and disease prevention strategies in Asian as well as European descendants.
SUPPLEMENTARY MATERIALS
Supplementary materials related to this article can be found online at https://doi.org/10.4093/dmj.2025.0697.
Cross table of new definition of clinical obesity with BMI-based obesity in Korean National Health and Nutrition Examination Survey 2014 to 2023
Top obesity-related complications among adults with clinical obesity
Clinical and metabolic characteristics of individuals by BMI category and corresponding classification under the new definition of clinical obesity in KNHANES 2014 to 2023
Cross table of BMI-based obesity with new definition of clinical obesity in KNHANES 2024 to 2023
Comparison of metabolic impairment, comorbidities, and activity limitation between men and women with clinical obesity in KNHANES 2014 to 2023 (age-standardized using the 2023 Korean population as the reference)
Trends in obesity prevalence by BMI category and clinical obesity criteria in Korean adults (KNHANES 2014 to 2023)
Informed cancers in people with no obesity, preclinical obesity, and clinical obesity in KNHANES 2014 to 2023
Distribution of adults by clinical and preclinical obesity status using Western anthropometric criteria in KNHANES 2014 to 2023
Comparison of key frameworks defining and staging obesity
Comparison of major complications for clinical obesity between Koreans and Americans
Participant flow diagram. BMI, body mass index.
Flowchart of diagnosing preclinical and clinical obesity. BMI, body mass index; HbA1c, glycosylated hemoglobin; HDL-C, high-density lipoprotein cholesterol; SBP, systolic blood pressure. a STOP-BANG is a validated screening tool for obstructive sleep apnea based on clinical and anthropometric risk factors (Snoring, Tiredness, Observed apnea, high blood Pressure, BMI, Age, Neck circumference, and Gender).
Age-specific prevalence of additional diagnostic components among individuals with and without clinical obesity. Comparison of the age-standardized prevalence of additional obesity-related conditions—(A) asthma, (B) arthritis, (C) obstructive sleep apnea (OSA), and (D) activity limitation—between individuals with no obesity and those with clinical obesity. These conditions represent functional or organ-specific consequences of excess adiposity that were not included in the main figure. All estimates were age-standardized using the 2023 Korean population structure to minimize demographic confounding and allow consistent comparison across age groups.
Notes
CONFLICTS OF INTEREST
Hun Jee Choe received a research grant from the Korean Diabetes Association (2024F-7). Jean-Pierre Després has no conflict of interest to report. John P. H. Wilding declares consultancy or advisory board work for the pharmaceutical industry contracted via the University of Liverpool (no personal payment) for Altimmune, Amgen, AstraZeneca, Boehringer Ingelheim, Eli Lilly, Napp, Novo Nordisk, Menarini, Pfizer, Regeneron, Rhythm Pharmaceuticals, Sanofi, Saniona, Tern, Shionogi, and YSOPIA Bioscience; research grants for clinical trials from AstraZeneca and Novo Nordisk; personal honoraria or lecture fees from AstraZeneca, Boehringer Ingelheim, Medscape, Napp, Novo Nordisk, and Rhythm Pharmaceuticals. Donna H. Ryan declares having received consulting honoraria from Abbvie, Altimmune, Amgen, AstraZeneca, Boehringer Ingelheim, Biohaven, Calibrate, Carmot/Roche, CINRx, Currax, eMedd, Epitomee, Fractyl, Gila, Lilly, Nestle, Novo Nordisk, Pfizer, Regeneron, Scientific Intake, Source Bio, Structure Therapeutics, Tenvie, Wondr Health, Zealand; she declares she received stock options from Calibrate, Epitomee, Scientific Intake and Xeno Bioscience. Soo Lim received research grants from Merck Sharp & Dohme, Novo Nordisk, and LG Chem; and honoraria as a consultant or speaker for AstraZeneca, Boehringer Ingelheim, Abbott, LG Chem, Daewoong Pharmaceutical, Chong Kun Dang Pharmaceutical, and Novo Nordisk.
AUTHOR CONTRIBUTIONS
Conception or design: H.J.C., S.L.
Acquisition, analysis, or interpretation of data: H.J.C., S.L.
Drafting the work or revising: all authors.
Final approval of the manuscript: all authors.
FUNDING
Hun Jee Choe received a research grant from the Korean Diabetes Association (2024F-7).
ACKNOWLEDGMENTS
None
DATA AVAILABILITY
The data used in this study are publicly available through the Korea National Health and Nutrition Examination Survey (KNHANES) database, which can be accessed at https://knhanes.kdca.go.kr. Access requires approval by the Korea Disease Control and Prevention Agency (KDCA). No individual participant data are available. The statistical code used for this analysis is available from the corresponding author upon reasonable request.