Key Features.
The Korean Genome and Epidemiology Study (KoGES) has undergone a strategic transformation towards ageing research with its 245 000 participants, now in their sixties and seventies, providing an invaluable long-term resource for studying healthy ageing and longevity.
Since 2017, the cohort has expanded beyond genomics to integrate multi-omics, brain magnetic resonance imaging (MRI), gut microbiome analyses, and geocoded air-pollution data, offering a comprehensive systems biology platform.
More than 20 years of prospective follow-up and robust linkage to national health databases ensure near-complete disease ascertainment and minimal loss to follow-up.
KoGES data and biosamples have been utilized for >2000 publications and these resources are accessible to researchers through the sharing and access platforms of the National Biobank of Korea as well as the Clinical and Omics Data Archive.
The original cohort
Initiated in 2001, the Korean Genome and Epidemiology Study (KoGES) is a large-scale, population-based cohort study to investigate the genetic and environmental determinants of health and disease in the Korean population. The original KoGES consortium profile paper, published in 2017, described the baseline survey, recruitment process, and participant characteristics at the inception of the study [1].
Since 2001, >245 000 participants have been enrolled, with data collected on demographic characteristics, lifestyle behaviours, dietary habits, medical history, and environmental exposure. Biological samples were systematically analysed for genomic markers.
Building on this foundation, the KoGES has expanded its scope to address South Korea’s transition to a super-aged society. With ageing participants, the cohort now serves as a platform for ageing research and precision medicine. This evolution, integrating advanced omics technologies, neuroimaging, and environmental health data, positions the KoGES to inform disease-prevention strategies addressing demographic challenges facing Korea and other rapidly ageing Asian populations (Fig. 1).
Figure 1.
Flowchart of participation in the KoGES. The few individuals who withdrew their consent to participate in the KoGES were excluded; therefore, the numbers deviated slightly from those reported in previous KoGES cohort profiles. aThe follow-up survey was conducted in 17 of the 39 hospitals and health institutions (baseline number of the 17 hospitals and health institutions = 137 448). bThe follow-up survey was conducted in 6 out of 11 counties (baseline number from the 6 counties = 21 714). cThe follow-up recruitment included only immigrant mothers and children. dThe follow-up survey was continued only in Japan because of difficulties in the follow-up of Japanese participants.
What is the reason for the new focus?
South Korea is experiencing the world’s most rapid demographic shift, having transitioned into a ‘super-aged society’ in 2025 as individuals aged ≥65 years exceeded 20% of the population [2]. This unprecedented pace of population ageing presents profound challenges for healthcare delivery, social welfare infrastructure, and economic sustainability [3–5]. Without targeted research to understand age-related health changes and develop evidence-based interventions, healthcare systems risk being overwhelmed by the rising burden of age-related chronic diseases and functional decline [6, 7].
In response, the KoGES has undergone a strategic transformation to position ageing research at its centre. Original cohort participants, now in their sixties and seventies (mean age 70.6 years), with >20 years of prospective data, represent a valuable resource for investigating healthy ageing, age-related disease progression, and modifiable factors promoting successful ageing in Asian populations. This transforms the KoGES into a comprehensive platform for gerontology and precision medicine, integrating five complementary data resources:
Expanded ageing phenotype data: standardized instruments capture multidimensional ageing aspects, including frailty assessment (Fried phenotype) [8], loneliness (UCLA Loneliness Scale) [9], psychological resilience (Connor–Davidson Resilience Scale) [10], and health literacy [11]. These measures enable the investigation of ageing phenotypes beyond traditional biomedical markers.
Integrated multi-omics data: integration of metabolomics, proteomics, and epigenomics enables deep molecular phenotyping of ageing processes [12], facilitating the investigation of how biological pathways interact across multiple molecular levels to identify novel therapeutic targets and biomarkers.
Neuroimaging and cognitive data: brain magnetic resonance imaging (MRI) protocols in KoGES_Ansan and KoGES_CAVAS enable the investigation of structural brain changes, white-matter lesions, and vascular alterations [13]. Validated cognitive batteries complement neuroimaging for comprehensive brain-ageing assessment.
Longitudinal environmental-exposure data: geocoded residential histories, integrated with air-pollution exposure estimates derived from the Community Multiscale Air Quality modelling system and national monitoring networks, enable the investigation of how environmental exposures throughout life influence ageing trajectories [14]. This environmental gerontology component is relevant for understanding respiratory and cardiovascular ageing in populations exposed to significant air pollution during industrialization.
Microbiome data: faecal-sample collection in selected cohorts (particularly the KoGES_Ansung study) enables investigation of the role of the gut microbiome in ageing and age-related diseases. This places the KoGES at the forefront of an emerging research frontier linking the microbiome to immune ageing, metabolic health, and cognitive function [15].
Together, these integrated domains have transformed the KoGES into one of the most comprehensively phenotyped ageing cohorts in Korea, uniquely positioned to address population ageing.
What will be the new areas of research?
The enhanced KoGES data enabled the study of several novel research domains not feasible with the original study design. Pilot analyses and early findings demonstrate the capacity of the data to generate actionable insights into healthy ageing in Asian populations.
Healthy ageing and longevity research
The combination of >20 years of prospective data with comprehensive biobanking enables the investigation of factors distinguishing exceptional longevity from typical ageing. By comparing participants who reached advanced age without major morbidity against those with typical disease trajectories, protective genetic variants, lifestyle patterns, and environmental exposures that promote the compression of morbidity can be identified.
Cognitive ageing and dementia
The integration of longitudinal cognitive assessments, structural and functional brain imaging, genetic data, and environmental exposure enables KoGES to unravel the complex, multifactorial aetiology of cognitive decline and dementia in Asian populations. This is particularly urgent given that dementia prevalence in Asia is projected to triple by 2050 [16].
Frailty and functional decline
Standardized frailty assessments, combined with multi-omics profiling and environmental data, enable investigation of the biological underpinnings of frailty—arguably the most clinically relevant ageing phenotype. Understanding frailty trajectories in the Korean population is essential, as cultural factors and differences in body composition may influence the manifestation and progression of frailty.
Environmental gerontology
The integration of decades of residential history with high-resolution environmental-exposure data and longitudinal health outcomes enables life-course investigation of the environmental determinants of ageing—a critically understudied area. This is particularly relevant for Korean populations. Research can quantify how cumulative exposure to air pollution accelerates cognitive decline [17] and cardiovascular ageing [18], informing both clinical risk stratification and environmental policy interventions.
Precision medicine in ageing
Multi-omics integration with clinical phenotypes enables the development of Korean-population-specific risk-prediction algorithms and the identification of therapeutic targets. Given that most pharmacogenomic and risk-prediction models are derived from European populations, this research is essential for equitable precision medicine that accounts for the genetic and environmental contexts specific to Korean ageing. KoGES data enable the identification of molecular signatures that predict responses to interventions.
Social determinants of healthy ageing
Comprehensive socio-economic data, social-network assessments, and psychosocial measures linked with biological ageing markers enable the investigation of how social factors ‘get under the skin’ to influence ageing trajectories. This is particularly relevant in rapidly changing Korean societies.
Collectively, these six research domains position the KoGES as a unique data source for understanding the multidimensional nature of ageing in Korean populations, generating evidence to inform clinical practice, public health policy, and social interventions for the world’s most rapidly ageing region.
Who is in the cohort?
The baseline characteristics of all participants and those retained at the first follow-up across the KoGES are summarized in Table 1 and Supplementary Table S3. The consortium comprises six distinct cohorts, each investigating specific health aspects. All cohorts contribute longitudinal data via linkage with National Health Insurance data, National Cancer Centre, and mortality records.
Table 1.
Baseline characteristics of all participants and those retained at first follow-up in the KoGES population-based studies.
| Variables | Total |
KoGES_Ansan and Ansung study |
KoGES_HEXA |
KoGES_CAVAS |
||||
|---|---|---|---|---|---|---|---|---|
| Baseline (N = 211 562) | Retained at first follow-up (N = 86 674) | Baseline (N = 10 030) | Retained at first follow-up (N = 8603) | Baseline (N = 173 195) | Retained at first follow-up (N = 65 608) | Baseline (N = 28 337) | Retained at first follow-up (N = 12 463) | |
| Age (years)a | 53.79 ± 8.75 | 54.36 ± 8.54 | 52.29 ± 8.93 | 52.42 ± 8.91 | 53.10 ± 8.37 | 53.72 ± 8.03 | 58.58 ± 9.38 | 59.10 ± 9.26 |
| Age group (years) [n (%)] | ||||||||
| 40–49 | 74 017 (34.99) | 27 308 (31.51) | 4711 (46.97) | 3982 (46.29) | 63 975 (36.76) | 21 062 (32.10) | 5631 (19.87) | 2264 (18.17) |
| 50–59 | 78 462 (37.09) | 33 499 (38.65) | 2617 (26.09) | 2285 (26.56) | 66 704 (38.51) | 27 253 (41.54) | 9141 (32.26) | 3961 (31.78) |
| 60–69 | 52 112 (24.63) | 23 212 (26.78) | 2702 (26.94) | 2336 (27.15) | 39 189 (22.63) | 16 310 (24.86) | 10 221 (36.07) | 4566 (36.64) |
| 70–79 | 6713 (3.17) | 2562 (2.96) | 0 (0.00) | 0 (0.00) | 3627 (2.09) | 983 (1.50) | 3086 (10.89) | 1579 (12.67) |
| ≥80 | 258 (0.12) | 93 (0.11) | 0 (0.00) | 0 (0.00) | 111 (0.17) | 0 (0.00) | 258 (0.91) | 93 (0.75) |
| Sex [n (%)] | ||||||||
| Male | 74 837 (38.19) | 31 237 (36.04) | 4758 (47.44) | 4105 (47.72) | 59 258 (34.21) | 22 290 (33.97) | 10 821 (38.19) | 4842 (38.85) |
| Female | 136 725 (61.81) | 55 437 (63.96) | 5272 (52.56) | 4498 (52.28) | 113 937 (65.79) | 43 318 (66.03) | 17 516 (61.81) | 7621 (61.15) |
| Questionnaires | ||||||||
| Smoking status [n (%)] | ||||||||
| Non-smoker | 150 726 (71.86) | 61 922 (71.95) | 5808 (58.69) | 5016 (59.03) | 125 026 (72.87) | 48 176 (73.98) | 19 892 (70.32) | 8730 (70.14) |
| Ex-smoker | 30 998 (14.78) | 13 542 (15.74) | 1539 (15.55) | 1335 (15.71) | 25 205 (14.69) | 10 211 (15.68) | 4254 (15.04) | 1996 (16.04) |
| Current smoker | 28 032 (13.36) | 10 598 (12.31) | 2549 (25.76) | 2146 (25.26) | 21 340 (12.44) | 6731 (10.34) | 4143 (14.65) | 1721 (13.83) |
| Drinking status [n (%)] | ||||||||
| Non-drinker | 106 223 (50.60) | 44 542 (51.74) | 4596 (46.24) | 3933 (46.12) | 87 152 (50.75) | 34 143 (52.43) | 14 475 (51.23) | 6466 (51.95) |
| Ex-drinker | 9586 (4.57) | 3824 (4.44) | 625 (6.56) | 545 (6.39) | 6908 (4.02) | 2536 (3.89) | 2026 (7.17) | 743 (5.97) |
| Current drinker | 94 116 (44.83) | 37 724 (43.82) | 4691 (47.20) | 4049 (47.48) | 77 672 (45.23) | 28 437 (43.67) | 11 753 (41.60) | 5238 (42.08) |
| Regular exercise [n (%)]b | 102 549 (49.14) | 43 882 (50.88) | NA | NA | 90 307 (52.56) | 36 041 (55.82) | 8860 (31.34) | 4099 (32.93) |
| Education duration (years) [n (%)] | ||||||||
| ≤6 | 52 179 (25.06) | 21 547 (25.16) | 3355 (33.73) | 4845 (56.77) | 30 262 (17.80) | 9746 (15.07) | 16 279 (57.65) | 6956 (55.97) |
| 7–12 | 102 165 (49.06) | 41 999 (49.04) | 5287 (53.15) | 2558 (29.97) | 89 332 (52.53) | 35 016 (54.14) | 9829 (34.81) | 4425 (35.60) |
| ≤13 | 53 900 (25.88) | 22 099 (25.80) | 1305 (13.12) | 1132 (13.26) | 50 463 (29.67) | 19 919 (30.80) | 2132 (7.55) | 1048 (8.43) |
| Total energy intake (kcal) | 1754.87 ± 598.72 | 1758.18 ± 582.76 | 1957.02 ± 714.10 | 1965.50 ± 723.18 | 1759.22 ± 592.03 | 1751.85 ± 560.85 | 1658.89 ± 575.38 | 1651.82 ± 553.74 |
| Protein intake (g)a,c | 59.18 ± 27.46 | 58.97 ± 26.63 | 66.60 ± 30.35 | 66.66 ± 30.57 | 59.90 ± 27.29 | 59.40 ± 26.12 | 52.28 ± 26.15 | 51.54 ± 24.49 |
| Fat intake (g)a,c | 27.61 ± 18.97 | 27.37 ± 18.50 | 32.56 ± 21.49 | 32.49 ± 21.49 | 28.21 ± 18.76 | 27.76 ± 18.02 | 22.28 ± 18.28 | 21.88 ± 17.47 |
| Carbohydrate intake (g)a,c | 313.01 ± 97.75 | 314.58 ± 95.62 | 344.53 ± 119.75 | 346.74 ± 121.74 | 312.27 ± 96.71 | 311.98 ± 91.69 | 306.61 ± 93.54 | 306.55 ± 91.45 |
| Anthropometric measurements | ||||||||
| Height (cm)a | 160.11 ± 8.18 | 160.13 ± 8.08 | 159.98 ± 8.66 | 159.96 ± 8.91 | 160.49 ± 8.03 | 160.58 ± 7.86 | 157.83 ± 8.50 | 157.88 ± 8.43 |
| Weight (kg)a | 61.76 ± 9.96 | 61.83 ± 9.89 | 63.01 ± 10.10 | 63.04 ± 10.11 | 61.83 ± 9.92 | 61.79 ± 9.80 | 60.92 ± 10.07 | 61.20 ± 10.10 |
| Body mass index (kg/m2)a | 24.02 ± 2.96 | 24.04 ± 2.93 | 24.57 ± 3.15 | 24.59 ± 3.07 | 23.94 ± 2.91 | 23.90 ± 2.86 | 24.40 ± 3.15 | 24.49 ± 3.12 |
| Waist circumference (cm)a | 81.57 ± 8.75 | 81.57 ± 8.78 | 82.70 ± 8.82 | 82.86 ± 8.81 | 81.09 ± 8.67 | 80.82 ± 8.63 | 84.10 ± 8.74 | 84.64 ± 8.80 |
| Hip circumference (cm)a | 94.35 ± 5.93 | 94.18 ± 5.87 | 93.53 ± 5.97 | 93.44 ± 5.97 | 94.47 ± 5.83 | 94.31 ± 5.73 | 93.93 ± 6.46 | 93.99 ± 6.46 |
| Waist–hip ratioa | 0.86 ± 0.07 | 0.87 ± 0.07 | 0.88 ± 0.08 | 0.89 ± 0.08 | 0.86 ± 0.06 | 0.86 ± 0.07 | 0.90 ± 0.07 | 0.90 ± 0.07 |
| Systolic blood pressure (mmHg)a | 123.14 ± 16.04 | 122.67 ± 15.69 | 121.69 ± 18.48 | 121.74 ± 18.30 | 122.71 ± 15.49 | 122.32 ± 14.90 | 126.28 ± 17.93 | 121.17 ± 17.45 |
| Diastolic blood pressure (mmHg)a | 76.86 ± 10.35 | 76.84 ± 10.16 | 80.34 ± 11.44 | 80.44 ± 11.34 | 76.24 ± 10.4 | 75.95 ± 9.69 | 79.40 ± 11.13 | 79.02 ± 10.82 |
| Clinical measurements | ||||||||
| Fasting blood glucose (mg/dl)a | 95.29 ± 22.06 | 94.55 ± 20.21 | 87.31 ± 21.41 | 86.981 ± 20.42 | 95.18 ± 21.60 | 94.66 ± 19.33 | 98.67 ± 24.16 | 99.06 ± 22.84 |
| Haemoglobin (g/dL)a | 13.82 ± 1.51 | 13.81 ± 1.47 | 13.58 ± 1.58 | 13.59 ± 1.57 | 13.84 ± 1.51 | 13.84 ± 1.46 | 13.81 ± 1.47 | 13.82 ± 1.44 |
| Total cholesterol (mg/dl)a | 197.45 ± 35.91 | 198.61 ± 35.59 | 191.17 ± 35.83 | 190.98 ± 35.37 | 197.44 ± 35.63 | 196.83 ± 35.33 | 199.75 ± 37.33 | 199.35 ± 36.69 |
| HDL- cholesterol (mg/dl)a | 52.42 ± 13.04 | 51.44 ± 12.87 | 44.64 ± 10.07 | 44.58 ± 9.99 | 54.02 ± 12.94 | 53.60 ± 12.83 | 45.36 ± 11.12 | 44.81 ± 10.79 |
| Triglyceride (mg/dl)a | 131.65 ± 92.71 | 131.79 ± 90.14 | 162.47 ± 104.96 | 162.58 ± 105.06 | 126.88 ± 89.90 | 124.61 ± 85.72 | 149.88 ± 100.32 | 148.28 ± 94.91 |
| AST (IU/L)a | 24.67 ± 18.55 | 24.72 ± 14.90 | 29.83 ± 18.55 | 28.29 ± 27.80 | 23.89 ± 18.03 | 23.63 ± 12.66 | 27.45 ± 20.92 | 26.81 ± 20.29 |
| ALT (IU/L)a | 23.10 ± 21.64 | 23.25 ± 18.88 | 28.13 ± 26.55 | 28.29 ± 27.80 | 22.50 ± 21.84 | 22.30 ± 17.33 | 24.86 ± 17.80 | 24.59 ± 18.21 |
| Creatinine (mg/dL)a | 0.84 ± 0.23 | 0.84 ± 0.21 | 0.84 ± 0.23 | 0.84 ± 0.19 | 0.82 ± 0.23 | 0.81 ± 0.21 | 0.95 ± 0.20 | 0.95 ± 0.19 |
NA, not assessed in this cohort or at this time point. Data are presented as means ± SD.
Defined from the questionnaire regarding whether or not a subject routinely exercises.
The nutrient intake was calculated by using food frequency questionnaires (FFQs) (the Japanese Public Health Centre-based prospective study (JPHC) FFQ was employed in the KoGES_emigrant study).
AST, aspartate aminotransferase; ALT, alanine aminotransferase.
KoGES_Ansan and Ansung study: a community-based cohort of 10 030 participants at baseline, of whom 8603 were retained at the first follow-up (85.8% follow-up; baseline mean age, 52.4 ± 8.9 years; women, 52.3%).
KoGES_HEXA study: the largest cohort, enrolling 173 195 participants at baseline, with 65 608 retained at the first follow-up (37.9% follow-up; baseline mean age, 53.7 ± 8.0 years; 66.0% women).
KoGES_CAVAS study: a cardiovascular-focused cohort of 28 337 participants at baseline, with 12 463 retained at the first follow-up (44.0% follow-up; baseline mean age, 59.1 ± 9.3 years; 61.2% women).
Gene–environment model studies: three cohorts no longer actively recruiting but continuing to contribute epidemiological and genetic data through secondary data linkage.
Baseline characteristics were highly comparable between all participants and those retained, with a consistent female preponderance. The prevalence of hypertension, type 2 diabetes, and dyslipidaemia increased by follow-up, consistently with age-related progression (Table 2).
Table 2.
Prevalence of selected chronic disorders at baseline and first follow-up in the KoGES population-based studies.
| Type of chronic disorders | Number (percentage) of people with chronic disorder |
|||||
|---|---|---|---|---|---|---|
| KoGES_Ansan and Ansung study |
KoGES_HEXA |
KoGES_CAVAS |
||||
| Baseline (N = 10 030) | First follow-up (N = 8603) | Baseline (N = 173 195) | First follow-up (N = 65 608) | Baseline (N = 28 337) | First follow-up (N = 12 463) | |
| Hypertensiona | 3273 (32.6) | 3463 (40.3) | 51 099 (29.5) | 13 641 (20.8) | 12 610 (44.5) | 6340 (50.9) |
| Type 2 diabetesb | 1222 (12.2) | 1241 (14.4) | 16 396 (9.5) | 6077 (9.3) | 3547 (12.5) | 1862 (14.9) |
| Dyslipidaemiac | 4898 (48.8) | 5310 (61.7) | 61 886 (35.7) | 23 582 (35.9) | 14 886 (52.5) | 8369 (67.2) |
| Obesityd | 3404 (33.9) | 3744 (43.5) | 57 011 (32.9) | 20 948 (31.9) | 11 312 (39.9) | 5272 (42.3) |
Hypertension was defined as either a systolic or diastolic blood pressure of ≥140 or ≥90 mmHg, respectively, or when participants self-reported the diagnosed disease.
Type 2 diabetes was defined as either a fasting blood glucose level of ≥126 mg/dl or when participants self-reported the diagnosed disease.
Dyslipidaemia was defined as either one of the NCET-ATPIII criteria (total cholesterol ≥ 240 mg/dl, low-density lipoprotein (LDL) cholesterol ≥ 160 mg/dl, high-density lipoprotein (HDL) cholesterol < 40 mg/dl, triglyceride ≥ 200 mg/dl) or when participants self-reported the diagnosed disease.
Obesity was defined as a body mass index ≥ 25 kg/m2.
What has been measured?
The KoGES has evolved from chronic-disease surveillance to a comprehensive ageing-research platform, generating one of Korea’s most extensive longitudinal datasets (Table 3).
Table 3.
Comprehensive list of data collection by KoGES cohorts.
| Contents and measurements | Baseline | Follow-up |
|---|---|---|
| Social and economic measurements | ||
| Education | √ | √ |
| Household income | √ | √ |
| Marital status | √ | √ |
| Occupation | √ | √ |
| Lifestyle-habit measurements | ||
| Past disease history | √ | √ |
| Medication history | √ | √ |
| Family disease history | √ | √ |
| Smoking | √ | √ |
| Alcohol habits | √ | √ |
| Physical activity | √ | √ |
| Sleep disorder | √ | √ |
| Nutrition | √ | √ (subset) |
| Anthropometric measurements | ||
| Height and weight | √ | √ |
| Waist and hip circumference | √ | √ |
| Body composition | √ | √ |
| Blood pressure and pulse rate | √ | √ |
| Clinical examination | ||
| Physical function | √ | √ |
| Cardiac function | √ | √ |
| Pulmonary function | √ | √ |
| Blood and urine laboratory test | √ | √ |
| Genome sequencing | ||
| Affy. 5.0 | √ | |
| Exome chip | √ | √ (subset) |
| Korea Biobank Array | √ | |
| Biobank | ||
| DNA | √ | √ |
| Blood (serum, plasma, buffy cast) | √ | √ |
| Urine | √ | √ |
| Additional data (see Supplementary Table 4) | ||
| Ageing | √ | |
| Omics data | √ | |
| Brain MRI | √ (subset) | |
| Microbiome | √ (subset) | |
| Air pollution | √ | |
| Link with health data | ||
| Statistics Korea: cause-of-death statistics, date of death, cause of death (ICD codes) | √ | |
| National Health Insurance Service: cancer incidence, primary site (ICD-O-3), SEER stage, first course of treatment | √ | |
| National Cancer Center: qualifications, medical treatments, prescription details, health-screening results | √ |
ICD, International Classification of Diseases; SEER, Surveillance, Epidemiology, and End Results.
Socio-demographic and lifestyle factors: questionnaires assessed education, occupation, income, marital status, smoking, alcohol, physical activity, sleep, and diet [19], enabling the investigation of socio-economic transitions on health.
Anthropometric and clinical measurements: standard protocols measuring height, weight, waist/hip circumference, blood pressure, and body composition.
Biospecimen banking: the National Biobank of Korea maintains extensive DNA, plasma, serum, and urine collections. Recent expansions include faecal samples for microbiome research and additional blood samples for metabolomics, proteomics, and epigenomics [20], establishing the KoGES as a premier molecular ageing resource.
Advanced geriatric assessments: standardized instruments measuring frailty, cognitive function, depression, loneliness, resilience, and health literacy.
Brain imaging: MRI protocols in KoGES_Ansan and KoGES_CAVAS investigate brain ageing and neurodegeneration.
Environmental-exposure assessment: this assessment enables investigation of the effects of air-pollution exposure on cardiovascular and respiratory ageing in the cohort participants.
National database linkage: Statistics Korea’s cause-of-death data indicated ∼15 000 deaths as of December 2023. Linkage with National Health Insurance and cancer registry data enables near-complete outcome ascertainment (Table 4).
Table 4.
Main causes of death in the KoGES population-based studies.
| Cause of death | Number (percentage) of people with cause of death |
|||||
|---|---|---|---|---|---|---|
| KoGES_Ansan and Ansung study |
KoGES_HEXA |
KoGES_CAVAS |
||||
| Men (N = 4752) | Women (N = 5256) | Men (N = 59 233) | Women (N = 113 874) | Men (N = 10 814) | Women (N = 17 499) | |
| All-cause | 1135 (23.9) | 787 (15.0) | 6396 (10.8) | 4856 (4.3) | 2951 (27.3) | 2315 (13.2) |
| Cancera | 411 (8.6) | 225 (4.3) | 2646 (4.5) | 2163 (1.9) | 1037 (9.6) | 626 (3.6) |
| Cardiovascular diseaseb | 99 (2.1) | 102 (1.9) | 555 (0.9) | 390 (0.3) | 279 (2.6) | 238 (1.4) |
| Cerebrovascular diseasec | 96 (2.0) | 76 (1.4) | 381 (0.6) | 370 (0.3) | 206 (1.9) | 259 (1.5) |
| Pneumoniad | 64 (1.3) | 44 (0.8) | 273 (0.5) | 168 (0.1) | 169 (1.6) | 155 (0.9) |
The number of individuals eligible for linkage to the National Statistics Office mortality data may differ from the number of KoGES baselines. This exclusion was due to errors in personal identification information and the omission of individuals who did not consent to the linkage of secondary data. International Classification of Diseases (ICD) codes were used to identify diseases:
cancer (C00–C97),
cardiovascular disease (I20–I25, I30–I52),
cerebrovascular disease (I60–I69), and
pneumonia (J12–J18).
What has it found? Key findings and publications
Over two decades, the KoGES has generated >2000 peer-reviewed papers, advancing the understanding of chronic diseases and ageing in Korea.
Discovery of population-specific variants: large-scale genome-wide association studies (GWAS) identified genetic variants associated with 76 traits [21], including novel loci for serum uric acid [22], lung function [23], and gene–nutrient interactions affecting obesity [24].
Translation to clinical prediction: polygenic risk scores from KoGES data effectively predict atherosclerotic cardiovascular disease [25], demonstrating that Asian-specific genetic data improve risk stratification. These tools are being extended to predict frailty and cognitive decline.
Novel biomarkers for Asian populations: the KoGES established the triglyceride-to-high-density lipoprotein (HDL) cholesterol ratio as a superior predictor of type 2 diabetes over 12 years [26] and demonstrated that non-HDL cholesterol and apolipoprotein B outperform low-density lipoprotein (LDL) cholesterol in cardiovascular risk prediction in Koreans [27]. Critically, these metabolic signatures are now being investigated as biomarkers of accelerated ageing.
Mechanistic insights into cognitive decline: long-term hyperglycaemia is directly associated with accelerated brain atrophy and cognitive decline over 4 years [28], while elevated HbA1c levels predict cognitive impairment in older adults without dementia [29]. This domain is now prioritized through expanded neuroimaging and cognitive assessments. With brain MRI now integrated into the KoGES, ongoing research has investigated whether intensive glycaemic control in midlife can slow brain ageing.
Air pollution and accelerated cardiovascular ageing: integration of geocoded pollution data revealed associations between ambient particulate matter and hypertension in >200 000 participants [14]. Current analyses investigate whether lifelong exposure to particulate Matter 2.5 (PM2.5) accelerates biological ageing, increases frailty risk, and shortens the health span, addressing whether economic development paradoxically undermines healthy ageing in Asia.
Social determinants of health: analyses revealed significant effects of marital status and socio-economic position on hypertension risk [30], demonstrating that social factors shape biological ageing trajectories. With expanded assessments of loneliness, resilience, and social support, the KoGES is now investigating which social interventions can modify biological ageing and reduce morbidity in rapidly changing societies.
These findings establish the KoGES as a premier resource for Asian-specific precision medicine and understanding how rapid modernization shapes ageing trajectories. As the KoGES participants enter their seventies and eighties, this foundation prepares the cohort to answer a defining question of twenty-first-century gerontology: Can populations that age rapidly also age successfully?
What are the main strengths and weaknesses?
The greatest strength of the KoGES lies in its unique historical value: it was launched in 2001 when participants were in their forties and fifties, so the cohort documented the entire transition from midlife to old age during Korea’s transformation into a super-aged society. This 22-year follow-up period enables the investigation of long-term disease trajectories, life-course epidemiology, and ageing processes that cannot be studied in shorter-term cohorts or cross-sectional studies. The KoGES represents a natural experiment in compressed modernization and rapid ageing, offering insights directly relevant to populations in Asia, Africa, and Latin America, who will experience similar transitions in the coming decades. Robust linkage with the National Health Insurance data, National Cancer Centre Cancer registry data, and mortality records via the Korean Personal Identification Number system enables the near-complete ascertainment of disease outcomes, regardless of active follow-up participation.
This infrastructure advantage is rare globally and virtually eliminates loss-to-follow-up bias for major health outcomes. The KoGES uniquely combines 22 years of epidemiological data with cutting-edge molecular phenotyping and environmental-exposure assessment, as well as comprehensive data specialized for ageing research. Consequently, this integration enables the investigation of complex gene–environment–ageing interactions that is not possible with single-domain datasets, establishing the KoGES as being at the forefront of precision gerontology research in Korea. The KoGES fills a critical gap in global health research, in which most genetic and epidemiological data are derived from European-ancestry populations. Asia-specific genetic architecture, disease risk profiles, and environmental exposures differ substantially from those of Western populations. The KoGES enables the development of precision-medicine tools tailored to non-European populations, contributing to health equity and ensuring that ageing research reflects global demographic realities.
Follow-up rates varied across the cohorts, with the HEXA cohort showing a 37.9% follow-up rate at active follow-up. This attrition can introduce selection bias into analyses that require repeated exposure measurements. However, this limitation is substantially mitigated by (i) national database linkage ensuring complete outcome ascertainment for all enrolled participants regardless of active participation, (ii) systematic investigation of attrition patterns enabling bias adjustment in analyses, and (iii) supplementary telephone surveys conducted between 2017 and 2018 (n = 33 037) to ascertain health outcomes and update participant contact status. The KoGES employed regional sampling strategies rather than national probability sampling, which potentially limits its generalizability to the broader Korean population. However, this design choice has critical strengths: detailed characterization of specific communities over two decades, integration with local environmental monitoring, and feasibility of intensive phenotyping, including brain imaging. Moreover, for genetic discovery and mechanistic research, the primary contribution of the KoGES—biological generalizability—often matters more than strict population representativeness. The national database linkage also enables comparison of KoGES findings with population-level disease rates, thereby allowing the assessment of generalizability for specific outcomes. The initial focus on middle-aged adults meant that early-life exposures were captured retrospectively, potentially introducing recall bias. This limitation is inherent to any cohort not initiated at birth and the KoGES addresses it through (i) the use of validated instruments for retrospective assessment, (ii) linkage with historical environmental and socio-economic data where available, and (iii) the focus on research questions where midlife-to-late-life trajectories are most relevant—precisely the questions motivating the ageing-research pivot. Future integration with birth cohorts or multigenerational studies could address life-course questions. The strengths of investigating healthy ageing in rapidly transitioning populations—the strategic focus of the KoGES—are decisive. The 22-year follow-up, which captures the demographic transition, complete outcome ascertainment, multidimensional data integration, and Asian-population specificity, enables research questions that cannot be addressed elsewhere to be answered. Although the attrition and sampling design pose challenges, these are common to long-running cohorts and are mitigated by the unique infrastructure and methodological advantages of the KoGES.
Can I get hold of the data? Where can I find out more?
Epidemiological data and biosamples from the KoGES are valuable resources for the research community. After completing the quality-control procedures, epidemiological and genotypic data will be made available to researchers. Korean investigators with institutional review board (IRB) approval may apply for data and biosamples. Upon review by the Biobank Distribution Committee of the National Biobank of Korea, approval for data and biosample distribution is granted to eligible applicants who meet the required criteria. For further details regarding the application process and data-access conditions, please refer to https://biobank.nih.go.kr. If only epidemiological data are required, then researchers can access the clinical and omics data archives (CODA). Information on the application process is available at https://coda.nih.go.kr. International researchers may contact us (koges@korea.kr) regarding collaboration or data access.
Ethics approval
This study was conducted by following the principles of the Declaration of Helsinki. This study was approved by the Korea Disease Control and Prevention Agency (IRB no. KDCA-2025–07-02). Informed consent was obtained from all participants.
Supplementary Material
Acknowledgements
We acknowledge all the participants and staff of the KoGES.
Contributor Information
Jieun Lyu, Division of Population Health Research, Department of Precision Medicine, National Institute of Health, Cheongju, 28159, Republic of Korea.
Joong-Yeon Lim, Division of Population Health Research, Department of Precision Medicine, National Institute of Health, Cheongju, 28159, Republic of Korea.
Joong Won Lee, Division of Population Health Research, Department of Precision Medicine, National Institute of Health, Cheongju, 28159, Republic of Korea.
Hye-Kyoung Kwak, Division of Population Health Research, Department of Precision Medicine, National Institute of Health, Cheongju, 28159, Republic of Korea.
Hyun-Young Park, National Institute of Health, Cheongju, 28159, Republic of Korea; Department of Internal Medicine, Eulji University College of Medicine, Seoul, 01830, Republic of Korea.
Jae-Pil Jeon, Department of Precision Medicine, National Institute of Health, Cheongju, 28159, Republic of Korea.
the KoGES group:
Chol Shin, Nan Hee Kim, Nam H Cho, Daehee Kang, Dong Hyun Kim, Sue K Park, Ji-Yeob Choi, Aesun Shin, Bo-Youl Choi, Mi Kyung Kim, Yu-Mi Kim, Min-Ho Shin, Insung Chung, Sang-Baek Koh, Hyeon Chang Kim, Joohon Sung, Myung-Hee Shin, and Hye Won Chung
Author contributions
J.L.: conceptualization, formal analysis, visualization, writing—original draft, writing—review and editing. J.-Y.L.: conceptualization, project administration, funding acquisition, writing, review, and editing. J.W.L., H.-k.K., H.-Y.P.: writing, review, and editing. J.-P.J.: conceptualization, funding acquisition, writing—review and editing, supervision.
Supplementary material
Supplementary material is available at IJE online.
Conflicts of interest
None declared.
Funding
This study was funded by the National Institute of Health, Republic of Korea (grant number 6635–302-260–01).
Data availability
KoGES data are available to qualified researchers through the National Biobank of Korea (https://biobank.nih.go.kr) and the Clinical and Omics Data Archive (https://coda.nih.go.kr) upon approval of the research proposals and IRB clearance.
Use of artificial intelligence (AI) tools
The authors have used Gemini to check English grammar during the preparation of this manuscript. All scientific content, data, analyses, and conclusions were generated and verified by the authors.
References
- 1. Kim Y, Han B-G, KoGES Group. Cohort profile: the Korean genome and epidemiology study (KoGES) consortium. Int J Epidemiol 2017; 46:e20-e. 10.1093/ije/dyv316 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Korean Statistical Information Service (KOSIS). Population Projections for South Korea: Daejon: Statistics Korea; https://kosis.kr (15 December 2025, date last accessed). [Google Scholar]
- 3. Jane Osareme O, Muonde M, Maduka CP et al. Demographic shifts and healthcare: a review of aging populations and systemic challenges. Int J Sci Res Arch 2024;11:383–95. [Google Scholar]
- 4. Hong C, Sun L, Liu G et al. Response of global health towards the challenges presented by population aging. China CDC Wkly 2023;5:884–7. 10.46234/ccdcw2023.168 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Iuga IC, Nerişanu RA, Iuga H. The impact of healthcare system quality and economic factors on the older adult population: a health economics perspective. Front Public Health 2024;12:1454699. 10.3389/fpubh.2024.1454699 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Jin K, Simpkins JW, Ji X et al. The critical need to promote research of aging and aging-related diseases to improve health and longevity of the elderly population. Aging Dis 2015;6:1–5. 10.14336/AD.2014.1210 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Guo J, Huang X, Dou L et al. Aging and aging-related diseases: from molecular mechanisms to interventions and treatments. Signal Transduct Target Ther 2022;7:391. 10.1038/s41392-022-01251-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Fried LP, Tangen CM, Walston J et al. Frailty in older adults: evidence for a phenotype. J Gerontol A Biol Sci Med Sci Cardiovascular Health Study Collaborative Research Group 2001; 56: M146–56. 10.1093/gerona/56.3.m146 [DOI] [PubMed] [Google Scholar]
- 9. Russell D, Peplau LA, Cutrona CE. The revised UCLA Loneliness Scale: concurrent and discriminant validity evidence. J Pers Soc Psychol 1980;39:472–80. 10.1037//0022-3514.39.3.472 [DOI] [PubMed] [Google Scholar]
- 10. Connor KM, Davidson JRT. Development of a new resilience scale: the Connor‐Davidson resilience scale (CD‐RISC). Depress Anxiety 2003;18:76–82. 10.1002/da.10113 [DOI] [PubMed] [Google Scholar]
- 11. HeeRan C, Cho S-I, Kim I-H. Validation of the measure of health literacy for the elderly. Korean Public Health Res 2018;44:99–109. [Google Scholar]
- 12. Lorusso JS, Sviderskiy OA, Labunskyy VM. Emerging omics approaches in aging research. Antioxid Redox Signal 2018;29:985–1002. 10.1089/ars.2017.7163 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Fjell AM, Walhovd KB. Structural brain changes in aging: courses, causes and cognitive consequences. Rev Neurosci 2010;21:187–221. 10.1515/revneuro.2010.21.3.187 [DOI] [PubMed] [Google Scholar]
- 14. Woo HD, Song DS, Choi SH et al. Integrated dataset of the Korean Genome and Epidemiology Study cohort with estimated air pollution data. Epidemiol Health 2022; 44: e2022071. 10.4178/epih.e2022071 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Badal VD, Vaccariello ED, Murray ER et al. The gut microbiome, aging, and longevity: a systematic review. Nutrients 2020;12:3759. 10.3390/nu12123759 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Prince M, Bryce R, Albanese E et al. The global prevalence of dementia: a systematic review and metaanalysis. Alzheimers Dement 2013;9:63–75.e2. 10.1016/j.jalz.2012.11.007 [DOI] [PubMed] [Google Scholar]
- 17. O’Piela DR, Durisek GR, Escobar YH et al. Particulate matter and Alzheimer’s disease: an intimate connection. Trends Mol Med 2022;28:770–80. 10.1016/j.molmed.2022.06.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Brook RD, Rajagopalan S, Pope ICA, American Heart Association Council on Epidemiology and Prevention, Council on the Kidney in Cardiovascular Disease, and Council on Nutrition, Physical Activity and Metabolism et al. Particulate matter air pollution and cardiovascular disease: an update to the scientific statement from the American Heart Association. Circulation 2010;121:2331–78. 10.1161/CIR.0b013e3181dbece1 [DOI] [PubMed] [Google Scholar]
- 19. Park S-J, Lyu J, Lee K et al. Nutrition survey methods and food composition database update of the Korean Genome and Epidemiology Study. Epidemiol Health 2024; 46: e2024042. 10.4178/epih.e2024042 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Cho SY, Hong EJ, Nam JM et al. Opening of the national biobank of Korea as the infrastructure of future biomedical science in Korea. Osong Public Health Res Perspect 2012;3:177–84. 10.1016/j.phrp.2012.07.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Nam K, Kim J, Lee S. Genome-wide study on 72,298 individuals in Korean biobank data for 76 traits. Cell Genom 2022;2:100189. 10.1016/j.xgen.2022.100189 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Park JS, Kim Y, Kang J. Genome-wide meta-analysis revealed several genetic loci associated with serum uric acid levels in Korean population: an analysis of Korea Biobank data. J Hum Genet 2022;67:231–7. 10.1038/s10038-021-00991-1 [DOI] [PubMed] [Google Scholar]
- 23. Lee K-S, Kim KH, Oh Y-M et al. A genome wide association study for lung function in the Korean population using an exome array. Korean J Intern Med 2021;36:S142–S150. 10.3904/kjim.2019.204 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Kwon Y-J, Park DH, Choi J-E et al. Identification of the interactions between specific genetic polymorphisms and nutrient intake associated with general and abdominal obesity in middle-aged adults. Clin Nutr 2022;41:543–51. 10.1016/j.clnu.2021.12.040 [DOI] [PubMed] [Google Scholar]
- 25. Song H, Koh Y, Rhee T-M et al. Prediction of incident atherosclerotic cardiovascular disease with polygenic risk of metabolic disease: analysis of 3 prospective cohort studies in Korea. Atherosclerosis 2022;348:16–24. 10.1016/j.atherosclerosis.2022.03.021 [DOI] [PubMed] [Google Scholar]
- 26. Lim T-K, Lee HS, Lee Y-J. Triglyceride to HDL-cholesterol ratio and the incidence risk of type 2 diabetes in community dwelling adults: a longitudinal 12-year analysis of the Korean Genome and Epidemiology Study. Diabetes Res Clin Pract 2020;163:108150. 10.1016/j.diabres.2020.108150 [DOI] [PubMed] [Google Scholar]
- 27. Yun SY, Rim JH, Kang H et al. Associations of LDL cholesterol, non-HDL cholesterol, and apolipoprotein B with cardiovascular disease occurrence in adults: Korean genome and epidemiology study. Ann Lab Med 2023;43:237–43. 10.3343/alm.2023.43.3.237 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Yu JH, Kim REY, Park SY et al. Association of long‐term hyperglycaemia and insulin resistance with brain atrophy and cognitive decline: a longitudinal cohort study. Diabetes Obes Metab 2023;25:1091–100. 10.1111/dom.14958 [DOI] [PubMed] [Google Scholar]
- 29. Kim JS, Chun BC, Lee K. Association between elevated glycosylated hemoglobin and cognitive impairment in older Korean adults: 2009–2010 Ansan cohort of the Korean genome and epidemiology study. Front Public Health 2024;12:1417236. 10.3389/fpubh.2024.1417236 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Son M, Heo YJ, Hyun H-J et al. Effects of marital status and income on hypertension: the Korean Genome and Epidemiology Study (KoGES). J Prev Med Public Health 2022;55:506–19. 10.3961/jpmph.22.264 [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
Data Availability Statement
KoGES data are available to qualified researchers through the National Biobank of Korea (https://biobank.nih.go.kr) and the Clinical and Omics Data Archive (https://coda.nih.go.kr) upon approval of the research proposals and IRB clearance.