Comparison of Population Attributable Fractions of Cancer Incidence and Mortality Linked to Excess Body Weight in Korea from 2015 to 2030
Article information
, Jihye An4, Jeehi Jung5, Hyeon Sook Lee6, Soseul Sung1,3,7, Sungji Moon1,3,8, Inah Kim9, Jung Eun Lee10, Aesun Shin1,2,3,8, Sun Ha Jee11, Sun-Seog Kweon12, Min-Ho Shin12, Sangmin Park2,13, Seung-Ho Ryu14, Sun Young Yang15, Seung Ho Choi15, Jeongseon Kim16, Sang-Wook Yi17, Yoon-Jung Choi16, Sangjun Lee1,2,3, Woojin Lim1,3,7, Kyungsik Kim1,3, Sohee Park11, Jeong-Soo Im18, Hong Gwan Seo16,18, Kwang-Pil Ko,1,19, Sue K. Park,1,2,3
Abstract
Background
The increasing rate of excess body weight (EBW) in the global population has led to growing health concerns, including cancer-related EBW. We aimed to estimate the population attributable fraction (PAF) of cancer incidence and deaths linked to EBW in Korean individuals from 2015 to 2030 and to compare its value with various body mass index cutoffs.
Methods
Levin’s formula was used to calculate the PAF; the prevalence rates were computed using the Korean National Health and Nutrition Examination Survey data, while the relative risks of specific cancers related to EBW were estimated based on the results of Korean cohort studies. To account for the 15-year latency period when estimating the PAF in 2020, the prevalence rates from 2015 and attributable cases or deaths from 2020 were used.
Results
The PAF attributed to EBW was similar for both cancer incidence and deaths using either the World Health Organization (WHO) Asian-Pacific region standard or a modified Asian standard, with the WHO standard yielding the lowest values. In the Korean population, the PAFs of EBW for cancer incidence were 2.96% in men and 3.61% in women, while those for cancer deaths were 0.67% in men and 3.06% in women in 2020. Additionally, PAFs showed a gradual increase in both sexes until 2030.
Conclusion
The EBW continues to have a significant impact on cancer incidence and deaths in Korea. Effective prevention strategies targeting the reduction of this modifiable risk factor can substantially decrease the cancer burden.
INTRODUCTION
Over the past 20 to 30 years, various health issues have emerged due to excess body weight (EBW) in the global population. EBW-related cancers are influenced by a combination of factors, such as deficiency in physical activity, sedentary lifestyle habits, and excessive consumption of fatty and sugary foods, which lead to excess energy [1]. Therefore, reducing EBW, a modifiable risk factor, is crucial for cancer prevention.
According to the International Agency for Research on Cancer (IARC) handbooks published in 2016, the absence of excessive body fat reduces the risk of gastric cardia, liver, gallbladder, pancreatic, ovarian, and thyroid cancer in middle-aged adults and meningioma and multiple myeloma, supported by sufficient evidence from human studies. Moreover, it decreases the risk of fatal prostate cancer, men breast cancer, and large B-cell lymphoma, although evidence for these cancers is limited [2,3]. Furthermore, the 2002 IARC Handbook indicates that excessive body fat increases the risk of esophageal, colorectal, and breast cancers in postmenopausal women, endometrial cancer, and renal cell carcinoma [4]. In addition, the World Cancer Research Fund/American Institute for Cancer Research (WCRF/AICR) reported strong evidence of a convincing level of causality for liver and pancreatic cancers and a probable grade of causality for mouth, pharynx, and larynx (MPL), cardia-gastric, gallbladder, ovarian, and advanced prostate cancers [5].
In Korea, the prevalence of EBW, standardized by the 2005 mid-year population, continued to increase in men (25.1% in 1998 and 42.8% in 2018), while it slightly decreased in women (26.2% in 1998 and 25.5% in 2018) [6]. The highest incidence of EBW was observed in men aged 30–39 years (51.3%) and in women aged ≥70 years (43.0%). Therefore, estimating the contribution of cancer prevention in the absence of excess body fat is necessary, as EBW is a modifiable risk factor. This study aimed to estimate and compare the population attributable fraction (PAF) trends of EBW from 2015 to 2030.
Moreover, the second objective of this study is to compare the attribution of EBW using various definitions, since the definition of overweight/obesity differs between the World Health Organization (WHO) [7] and the WHO Asian-Pacific region for Asian populations, including Korean individuals [8]. This comparison will help determine the most suitable criteria for the Korean population.
METHODS
Definition of trend exposures and estimation of prevalence rates using various methods
The theoretical minimum risk associated with cancer-related EBW is not obesity or overweight, which is defined as a body mass index (BMI) of <23 kg/m2 according to the WHO Asian-Pacific standard. Therefore, a BMI of 23 to 24.9 kg/m2 is defined as overweight, while a BMI of ≥25 kg/m2 is defined as obesity. According to the WHO standards, a BMI of 25 to 29.9 kg/m2 is considered overweight, while a BMI of ≥30 kg/m2 is considered obese in adults with EBW. Additionally, a BMI of 23 to 24.9 kg/m2 is defined as overweight, 25 to 29.9 kg/m2 as obesity Ⅰ, and ≥30 kg/m2 as obesity Ⅱ according to the modified Asian-Pacific and WHO standards.
The past and current prevalence rates of EBW among adults aged ≥20 years were calculated using the Korean National Health and Nutrition Examination Survey (KNHANES) data obtained in 2001 and 2005–2018 [9]. The selected data were standardized using the 2000 mid-year population, and the prevalence rates of EBW were estimated by extrapolation using a linear regression model.
Estimation of relative risks for cancer risk
The relative risk (RR) of developing certain types of cancer was determined based on sufficient evidence from human studies outlined in the IARC’s 2002 and 2016 handbooks [2-4] or strong evidence as recognized by the WCRF/IARC. The types of cancer selected for this study included MPL (oral, pharyngeal, and laryngeal cancers; C00–C14 and C32); esophageal (C15), gastric (C16.0), colorectal (C18–C20), liver (C22), gallbladder (C23), pancreatic (C25), and breast (C50) cancers in postmenopausal women; endometrial and uterine cancers (C54.1); ovarian cancer (C56); prostate cancer (advanced) (C61); and kidney cancer (C64) (Supplemental Table S1).
Thyroid cancer was excluded from the estimation due to the unclearness risk of EBW in the Korean and Asian populations [10,11]. Additionally, meningiomas and multiple myelomas were excluded from this study owing to their low incidence and mortality rates. Men breast cancer and premenopausal large B-cell lymphoma, which were reported to have limited evidence levels in the IARC handbook [3], were also excluded.
Meanwhile, the Korean Cohort Consortium data [12], including the Korean National Health Insurance Service-National Health Screening Cohort [13], the Korean Multicenter Cancer Cohort Study [14], the Namwon/Dong-gu Study [15], the Korea Cancer Prevention Study-phase II [16], the Korea National Cancer Center Screening Cohort [17], and the Kangbuk Samsung Health Study data, were analyzed to estimate the RRs of cancer incidence [18]. The individual RRs of cancer-related deaths were calculated using original data from the Korean Genome and Epidemiology Study [19], the Kangwha Cohort Study [20], and KNHANES [9].
The RRs of cancer incidence and deaths for each cohort study were calculated and then meta-analyzed using a random-effects model.
Estimation of cancer PAF value in 2015 to 2030
Levin’s formula (Equation 1) was used to estimate the trends in EBW PAF in 2015 to 2030, and the 95% confidence interval (CI) of each specific cancer was derived using the Monte Carlo method. As consistent meta-RRs were used when estimating the trend, the prevalence of EBW was adjusted to reflect a 15-year lag period. Considering latency, the prevalence in 2000 was used in the calculation of the 2015 PAF, the prevalence in 2005 in the 2020 PAF, the prevalence in 2010 in the 2025 PAF, and the prevalence in 2015 in the 2030 PAF.
To determine the PAF of EBW per 5 kg/m2 BMI increase, the following formula was used to calculate the PAF (Equation 2): The prevalence rate of BMI ≥23 kg/m2 was computed as “EXP,” while the average BMI of the 2000 population aged 20 years and older was computed as “Exposure.”
For men and women, the attributable cases or deaths (attributable cancers [ACs]) due to EBW were calculated by multiplying the estimated PAF by the number of cases or deaths from each specific cancer, sourced from the Korean Statistical Information Service [21]. In the same year, a PAF was applied to each AC value. On the contrary, the total PAF of EBW-associated cancers was calculated by summing the ACs for men and women and dividing it by the total number of cancer cases or deaths.
Owing to the difficulty in assessing the prevalence rates of various cancer subtypes, the incidence and death rates of the corresponding cancers were additionally considered to determine the AC value. For esophageal cancer, we used the fraction of adenocarcinoma derived from the Asian Cohort Consortium data (1.78% incidence and 2.97% mortality) [22]. Using the proportions of distant and regional tumors (12.8% and 14.6%, respectively) in all types of prostate cancer provided in the Korean Urological Oncology Society, Korean Prostate Cancer Database, approximately 27.4% of all prostate cancer cases were considered advanced and associated with EBW [23].
Ethical statement
This study was approved by the Institutional Review Board of Seoul National University Hospital (IRB number C-1911-188-1084). Written informed consent by the patients was waived due to a retrospective nature of our study.
RESULTS
Prevalence rates of EBW
The prevalence rates of EBW in Korea according to the WHO Health Organization Asia-Pacific region standards are listed in Supplemental Table S2. The trends in 2005, 2010, and 2015 were observed, while projections were made for 2000, 2020, 2025, and 2030. The prevalence rates of overweight (23 to 24.9 kg/m2) slightly decreased from 2000 to 2030 (25.8% to 23.8% in men and 21.6% to 15.7% in women), while those of obesity (≥25 kg/m2) increased dramatically in men but not in women.
Relative risks of cancer
The RRs of cancer incidence and deaths using the different definitions of BMI based on the WHO, WHO Asian-Pacific region, and modified WHO-Asian standards are shown in Supplemental Tables S3, S4. The RRs of cancer incidence in men showed significant cancer risks in prostate (23–24.9 kg/m2 : 1.23; 95% CI, 1.19 to 1.26; ≥25 kg/m2 : 1.26; 95% CI, 1.23 to 1.30) and kidney cancer (23–24.9 kg/m2 : 1.38; 95% CI, 1.31 to 1.45; ≥25 kg/m2 : 1.78; 95% CI, 1.70 to 1.86) based on the WHO Asian-Pacific region standard. Significant cancer incidence risks are observed in liver (25–29.9 kg/m2 : 1.06; 95% CI, 1.04 to 1.08; ≥30 kg/m2 : 1.42; 95% CI, 1.35 to 1.50), kidney cancer (25–29.9 kg/m2 : 1.49; 95% CI, 1.43 to 1.54; ≥30 kg/m2 : 2.06; 95% CI, 1.87 to 2.26) when categorized according to the WHO standard. Moreover, significant cancer incidence risks were seen in colorectal (RRs of 23–24.9, 25–29.9, and ≥30 kg/m2 : 1.10, 1.20, and 1.50, respectively) prostate (1.17, 1.27, and 1.59) and kidney (1.35, 1.73, and 2.40) cancers according to modified WHO-Asian standards. However, for cancer deaths only kidney cancer showed 11% and 53% increased risks (23–24.9 kg/m2 : 1.11; 95% CI, 1.00 to 1.24; ≥25 kg/m2 : 1.53; 95% CI, 1.29 to 1.69) based on the WHO Asian-Pacific region standard.
The risks of liver, gallbladder and uterine endometrium cancer in women increased by approximately more than 10% across all three categories. In addition, kidney cancer showed a significant increase in cancer incidence under the WHO Asian-Pacific region and modified WHO-Asian standards, while the risk of colorectal and postmenopausal breast cancer increased in the WHO standards. Among them, the risk of uterine endometrium cancer has the highest RRs (23–24.9 kg/m2 : 1.38; 95% CI, 1.29 to 1.48; ≥25 kg/m2 : 1.98; 95% CI, 1.86 to 2.11) in the WHO Asian-Pacific region standard; (25–29.9 kg/m2 : 1.57; 95% CI, 1.49 to 1.67; ≥30 kg/m2 : 3.18; 95% CI, 2.88 to 3.51) in the WHO standard (23–24.9 kg/m2 : 1.39; 95% CI, 1.30 to 1.49; 25–29.9 kg/m2 : 1.79; 95% CI, 1.68 to 1.91; ≥30 kg/m2 : 4.93; 95% CI, 2.90 to 8.39). The cancer deaths for women had a similar pattern to cancer incidence.
The cohort study-based RRs caused by EBW per 5 kg/m2 BMI increase in specific cancer incidence and deaths are presented in Supplemental Table S5. The RRs was increased by 13% (RR, 1.13; 95% CI, 1.03 to 1.25) in prostate cancer incidence and 44% (RR, 1.44; 95% CI, 1.10 to 1.88) in kidney cancer deaths in men for every 5 kg/m2 BMI increment, which were the highest. In women, the incidence of kidney cancer (RR, 1.21; 95% CI, 1.03 to 1.42) and deaths of postmenopausal breast cancer (RR, 1.24; 95% CI, 1.18 to 1.31) had the highest RRs. Moreover, the risks of ovarian cancer increased by 7% (RR, 1.07; 95% CI, 1.04 to 1.11) in incidence and 8% (RR, 1.08; 95% CI, 1.03 to 1.14) in deaths, respectively.
Comparison of cancer PAF (%) attributed to EBW
The comparisons of cancer PAF (%) attributed to EBW using different criteria are represented in Table 1 (Supplemental Fig. S1). Except for cancer PAFs calculated using continuous BMI values, the PAF of cancer incidence was relatively higher than that of cancer deaths. When considering all cancer types, a per mean BMI increase in the Asian criteria led to increases in PAFs in men, women, and the total population (4.91%, 8.64%, and 6.67% for cancer incidence and 5.44%, 11.41%, and 7.71% for cancer deaths, respectively). Among the WHO, WHO Asian-Pacific, and modified WHO Asian-Pacific standards, the modified standard yielded the highest values.
Population Attributable Fractions (%) of Cancer Due to Excess Weight, Classified according to Various Methods
A comparison of PAFs (%) for specific cancer types attributed to EBW showed that kidney cancer in men made the largest contribution, both in terms of incidence and deaths. The PAF of kidney cancer incidence using the WHO Asian-Pacific standard reached 24.66%, similar to the value obtained using the modified WHO Asian-Pacific standards (24.22%). Moreover, kidney cancer contributed significantly to the high PAF of deaths, with the highest value obtained using the WHO standard (23.44%), followed by the modified WHO Asian-Pacific standard (21.89%) and the WHO Asian-Pacific standard (15.56%).
For women, the PAFs obtained based on the WHO Asian-Pacific standard and the modified WHO Asian-Pacific standard were similar, with the lowest value observed using the WHO standard. Although the contribution of kidney cancer was equally significant in women as in men, ranking second overall, uterine endometrial cancer made a prominent contribution to cancer incidence (WHO Asian-Pacific standard: 25.59%). For cancer deaths, both uterine endometrial and postmenopausal breast cancers exhibited high contributions (WHO Asian-Pacific standards: 8.57% and 11.44%, respectively) (Fig. 1).
Comparison of cancer population attributable fraction (PAF, %) attributed to excess body weight (EBW) when using different criteria (A: men, B: women, C: total population). MPL, mouth, pharynx, and larynx; WHO, World Health Organization.
Table 2 represents a comparison between the PAF in 2015 and that in 2020. Over the past 5 years, the PAF of cancer incidence and deaths has slightly increased for both men and women, with the increase being more pronounced in men than in women (incidence in men: 2.40% in 2015 to 2.96% in 2020; deaths in men: 0.56% in 2015 to 0.67% in 2020; incidence in women: 3.29% in 2015 to 3.61% in 2020; deaths in women). The number of cancer cases and deaths in men is projected to increase from 2015 to 2030, with colorectal cancer having the highest incidence and kidney cancer having the highest number of deaths based on AC values. In women, cancer cases and deaths due to liver cancer, excluding other cancer types, were expected to decrease within the next 15 years from 2015 (Fig. 2). Also, the proportion of contribution to deaths of liver cancer decreased from 30.4% in 2015 to 25.4% in 2020 (Supplemental Figs. S2-S5).
Population attributable fraction (PAF, %) of cancer cases and deaths of excess body weight (EBW) for cancer cases and deaths from 2015 to 2030. (A) All cancer cases and death. (B) Specific cancer cases in men. (C) Specific cancer death in men. (D) Specific cancer cases in women. (E) Specific cancer death in women.
The PAF of cancer incidence due to EBW increased from 1.49% in 2009 to 2.40% in 2015 for men, and from 2.20% to 3.29% for women. In contrast, the PAF of cancer deaths decreased from 0.94% in 2015 to 0.56% for men, while it elevated from 1.92% to 3.00% for women. Both cancer incidence and deaths in 2009 included fewer types of cancers contributing to EBW compared to 2015. Five cancer types related to EBW, namely, MPL, esophageal cancer, gastric cardia, liver cancer, gallbladder cancer, and ovarian cancer, were included compared with those in 2009 (Supplemental Table S6).
DISCUSSION
Using three different criteria (WHO Asian-Pacific, modified WHO Asian-Pacific, and WHO standards) to calculate the PAF of EBW, varying results were obtained owing to the different BMI cutoff values used for each criterion. Using the WHO Asian-Pacific standard, the PAFs for cancer incidence were 2.40% in men and 3.29% in women. In contrast, the PAFs for cancer-related deaths were lower than those for cancer incidence (0.56% in men and 2.99% in women).
A sensitivity analysis for every 5 kg/m2 increase in BMI was performed to compensate for the limitation of categorical values. This analysis revealed that the PAF was higher using this method than that calculated using the Asian-WHO modified standard. Although the contribution fractions for each cancer type were similar, the PAFs differed due to the variations in BMI trends, which can be linear or U-shaped depending on the cancer type. Specifically, stomach and kidney cancers showed a linear trend, while esophageal, gallbladder, and pancreatic cancers exhibited a U-shaped trend in men, suggesting that the PAFs of these cancer types can be overestimated when estimated linearly. In women, MPL, esophageal cancer, liver cancer, uterine cervix cancer, and endometrial cancer showed U-shaped BMI trends. These findings suggest that using BMI as a continuous variable may not be suitable for the Korean population.
The PAF value obtained according to the WHO standard is the lowest, as the proportion of individuals with a BMI of ≥30 kg/m2 is relatively low in the Korean population, differing significantly from Western populations. Therefore, the modified WHO Asia-Pacific standard is more appropriate for estimating the PAF of cancer incidence in the Korean population. Further refinement of the BMI cutoff values according to the characteristics of specific cancer incidences, in Korea is expected to yield more accurate PAFs.
In women, the high contribution of postmenopausal breast cancer to the PAF of deaths was largely due to the significant effect of obesity on breast cancer mortality. The Women’s Health Initiative clinical trial has demonstrated a more than two-fold increase in the risk of breast cancer deaths among women with grade II and III obesity compared to those with a normal BMI [24]. Many patients with breast cancer undergo hormone replacement therapy (HRT). Although HRT use can obscure the influence of obesity on circulating hormone levels, the breast cancer risk is elevated regardless of HRT use (RR, 1.59; 95% CI, 1.09 to 2.32) [25]. Therefore, breast cancer contributes significantly to cancer-related deaths in women. In men, MPL and esophageal and stomach cancers showed low or negligible contributions to EBW-related deaths. Adipose tissue can cause meta-inflammation and release pro-inflammatory mediators that influence cancer development [26]. However, the link between obesity and MPL deaths is classified as probable, according to the AICR. Smoking and alcohol consumption play significant roles in esophageal cancer development and progression, while Helicobacter pylori infection and dietary habits are more prominent risk factors for stomach cancer. These factors contribute to comparatively lower deaths from these cancers due to EBW.
Numerous studies on the PAFs of EBW have been published (Supplemental Fig. S6). The incidence of PAF in men increased in Western countries from 2010 (Australia: 2.5%) [27] to 2015 (the United Kingdom: 5.2%) [27], marking the highest reported rate. In Asian countries, the PAF of EBW has increased over time; however, the peak value reported in a 2020 Korean study (3.0%) was lower than that observed in Western countries. Additionally, the highest result, estimated at 5.7% in a study in the United States [28], pertained to the PAF of cancer deaths in men. This result is likely driven by prostate cancer, which is closely associated with obesity in Western countries. However, another study conducted in the United States excluded prostate cancer from this estimation. Instead, gallbladder and liver cancers made significant contributions, as did thyroid and multiple myeloma cancers, which were not included in this study. In women, the PAF of cancer incidence mirrored that in men, but the PAFs of deaths were generally higher than those in men. Additionally, the PAFs of cancer-related deaths in Asia have been gradually increasing [27-31].
This study has several limitations. First, calculating the incidence or death RRs of cancer subtypes, such as esophageal adenocarcinoma, gastric cardia cancer, renal cell carcinoma, and endometrial cancer, posed challenges. Recent findings from the Asian Cohort Consortium indicated elevated risks for esophageal squamous cell carcinoma at BMI levels of 18.5 kg/m2 and 25 kg/m2, and for esophageal adenocarcinoma at a BMI of 25 to 29.9 kg/m2 [22]. However, due to the low prevalence of a BMI of ≥25 kg/m2 in the Korean population, estimating these RRs was difficult. Second, the absence of RRs associated with the risk of cancer death necessitated substituting with the meta-analysis of the risk of cancer incidence, potentially leading to an overestimation of PAF related to cancer deaths.
To address these limitations, we conducted estimations by multiplying the incidence and death rates of various cancer subtypes using the AC values, which is a distinct advantage of this study. In response to concerns about EBW and related cancers linked to Westernization and sedentary lifestyles, we calculated the PAFs by including a broad range of cancers in our PAF calculations. Finally, this study is unique as it considers a 15-year latency period for cancer incidence and deaths from 2015 to 2030, at 5-year intervals. This approach contrasts with previous PAF studies related to EBW, which typically focus on estimating the point values for specific years.
In conclusion, the PAF of EBW is projected to increase from 2015 to 2030. The PAF of cancer incidence in men increased by 2.21%, while that in women increased by 0.71%. Moreover, the PAFs of cancer-related deaths increased to 0.54% and 1.09% in men and women, respectively. EBW is a modifiable factor; however, the effective prevention of cancer and chronic diseases requires more than simple weight management. To achieve comprehensive prevention, the vicious cycle of physical inactivity, sedentary behavior, and excessive energy intake must be addressed. Therefore, establishing evidence-based prevention guidelines that consider the interactions between EBW, dietary factors, and physical activity is crucial for effective cancer prevention through nationwide promotion and intervention.
Supplementary Material
Supplemental Table S1.
Cancers Caused by Excess Body Weight
Supplemental Table S2.
Prevalence Ratesa of Exposure to Excess Body Weight in Korea
Supplemental Table S3.
RRa for the Association with Excess Body Weight under Various Categories on Cancer Incidence and Death among Men in Korean Cohort Studies
Supplemental Table S4.
RRa for the Association with Excess Body Weight on Cancer Incidence and Death among Women in Korean Cohort Studies
Supplemental Table S5.
Cohort Study-Based RRsa for the Association with BMI Increase per 5 kg/m2 on Cancer Incidence and Death
Supplemental Table S6.
Comparison between Cancer PAFsa Caused by Excess Body Weightb in 2009 and 2015
Supplemental Fig. S1.
Comparison of cancer population attributable fraction (PAF; %) attributed to excess body weight (EBW) when using different criteria. WHO, World Health Organization; BMI, body mass index.
Supplemental Fig. S2.
The population attributable fraction (PAF; %) of cancer cases attributed to excess body weight and proportion of specific cancers among all cancer cases caused by excess body weight in Korea, 2015. (A) Men, (B) women, and (C) total cancer cases. MPL, mouth, pharynx, and larynx.
Supplemental Fig. S3.
The population attributable fraction (PAF; %) of cancer deaths attributed to excess body weight and proportion of specific cancers among all cancer deaths caused by excess body weight in Korea, 2015. (A) Men, (B) women, and (C) total cancer deaths.
Supplemental Fig. S4.
The population attributable fraction (PAF; %) of cancer cases attributed to excess body weight and proportion of specific cancers among all cancer cases caused by excess body weight in Korea, 2020. (A) Men, (B) women, and (C) total cancer cases. MPL, mouth, pharynx, and larynx.
Supplemental Fig. S5.
The population attributable fraction (PAF; %) of cancer deaths attributed to excess body weight and proportion of specific cancers among all cancer deaths caused by excess body weight in Korea, 2020. (A) Men, (B) women, and (C) total cancer deaths.
Supplemental Fig. S6.
International comparison of population attributable fraction (PAF; %) in excess body weight. (A) Men, (B) women, and (C) total population.
Notes
CONFLICTS OF INTEREST
No potential conflict of interest relevant to this article was reported.
AUTHOR CONTRIBUTIONS
Conception or design: I.K., J.E.L., K.P.K., S.K.P. Acquisition, analysis, or interpretation of data: Y.H., J.A., J.J., H.S.L., I.K., J.E.L., A.S., S.H.J., S.S.K., M.H.S., S.P., S.H.R., S.Y.Y., S.H.C., J.K., S.W.Y., K.P.K., S.K.P. Drafting the work or revising: Y.H., J.A., J.J., S.S., S.M., S.L., W.L., K.K. Final approval of the manuscript: S.P., Y.J.C., J.S.I., H.G.S., K.P.K., S.K.P.
Acknowledgements
This study was funded by the Korean Foundation for Cancer Research (grant no. CB-2017-A-2). Additionally, the study was conducted using a core cohort database provided by the Korean Genome and Epidemiology Study, Korea National Institute of Health, and Korea Disease Control and Prevention Agency; a cohort study based on the Korea National Health and Nutrition Examination Survey and Korea Disease Control and Prevention Agency; and customized cohort databases provided by the National Health Insurance Service (NHIS-2019-1-495 and NHIS-2020-1-164).
The prevalence rates of the risk factors were determined using data provided by the Korea National Institute of Health, Korea Disease Control and Prevention Agency (KDCA), Occupational Safety and Health Research Institute, Korea Occupational Safety and Health Agency, and the Korean Statistical Information Service.
The incidence and mortality rates of cancer were determined using data provided by the Cancer Registration Statistics, Korea National Cancer Center, and Korean Statistical Information Service.