이중구간중도절단된 생존자료의 생존함수 비교를 위한 검정: 한국인 암 예방연구 중 당뇨병에의 응용

지선하,남정모,김진흠,Jee, Sun-Ha,Nam, Chung-Mo,Kim, Jin-Heum 한국통계학회 2009 응용통계연구 Vol.22 No.3

이중구간중도절단된 자료의 생존함수를 비교하기 위한 두 검정법을 소개하고 한국인 암 예방연구 (Jee 등, 2005) 자료에 적용하여 당뇨병 잠복시간의 분포를 성별과 연령에 따라 비교하였다. Kim 등 (2006)의 검정법을 이중구간중도절단된 자료로 확장한 검정법은 위험집합의 크기에만 의존하는 가중값을 사용하기 때문에 대용량 자료의 분석에서 Sun (2006)의 검정법보다 계산 시간을 대폭 줄일 수 있으며, 이산형 생존자료뿐만 아니라 연속형 생존자료에도 적용가능한 장점이 있다. 당뇨병의 잠복시간이 성별에 따라 매우 유의하게 달랐으며 여자의 잠복시간이 남자보다 긴 것으로 나타났다. 4개 연령그룹 간 당뇨병의 잠복시간도 성별에 관계 없이 매우 유의하게 달랐으며 여자의 경우가 남자의 경우보다 그 차이가 훨씬 더 유의했다. 한편, 소표본 모의실험을 통해 제안한 검정법과 Sun (2006)의 검정법의 검정력을 비교하였으며 제안한 검정법의 검정력이 Sun (2006)의 검정법보다 더 좋은 것으로 나타났다. Two tests were introduced for comparing several survival functions with doubly interval-censored data and illustrated with data surveyed by Korean Cancer Prevention Study (Jee et al., 2005). The test which extended Kim et al. (2006)'s test to the doubly interval-censored data has an advantage over Sun (2006)'s test in terms of saving computation time because the proposed test only depends on the size of risk set, and also the proposed test is applicable to continuous failure time data as well as discrete failure time data unlike Sun's test. Comparing male with female groups on the incubation time of diabetes was highly different and the survival of female group was longer than that of male one. Regardless of gender, the difference in survival functions of four age groups was highly significant with p-value of less than 0.001. This trend was more remarkable for female group than for male one. Simulation results showed that the significance level of both tests was well controlled and the proposed test was better than Sun's test in terms of power.

허혈성심질환 예측모형을 이용한 사회경제적 위치와 허혈성심질환 위험도의 연관성

고동희,한순실,지선하,김형렬,Koh, Dong-Hee,Han, Sun-Shil,Jee, Sun-Ha,Kim, Hyoung-Ryoul 대한예방의학회 2006 예방의학회지 Vol.39 No.4

Objectives: The object of this study is to assess the relationship between socioeconomic factors and the predicted 10-year risk of cardiovascular disease by using health risk appraisal of ischemic heart disease. Methods: The study population was taken from The 2001 Korea National Health and Nutrition Survey, and it consisted of 1,566 men and 1,984 women aged 30-59. We calculated 10-year risk using the risk function of ischemic heart disease as developed by Jee. The educational level and equivalized household income were dichotomized by a 12 years education period and the median income level. Occupation was dichotomized into manual/non-manual work. We stratified the population by age(10 years) and sex, and then we rated the risk differences according to socioeconomic factors by performing t-tests for each strata. Results: There were gradients of the predicted 10-year risk of ischemic heart disease with the educational level and the equivalized household income, and thet was an increasing tendency of risk differences with age. Manual workers didn't show significant risk difference from non-manual workers. Conclusions: There was definite relationship between low socioeconomic position and the predicted risk of ischemic heart disease in the future.

개인별 여성 유방암 발생예측모형(health risk appraisal) 개발 연구

지선하(Sun Ha Jee),송지원(Ji Won Song),남정모(Chung mo Nam) 한국역학회 2004 Epidemiology and Health Vol.26 No.1

Objectives: The objectives of this study were to develop a health risk appraisal (HRA) model of breast cancer and to evaluate the model's ability to accurately predict individual breast cancer risk Method: The sample data were from a prospective cohort study in Korea with a follow-up period of 8 years in Korea (1995-2002) The cohort was composed of 108,708 Korean women, aged 30 to 64 years, who received health insurance from the Korea Medical Insurance Corporation and who underwent a medical evaluation in 1994 Between 1995 and 2002 (1,138,349 person-year), there were 772 incident cases of breast cancer (67.8/100,000 person-years) First, we developed the HRA based on the Cox propertional hazard models using a prospective cohort data With half-split data (50% random sample) Second, using Cox proportional hazard models, we compared the probability of breast cancer predicted by the model to the actual number of cases observed m the other 50% of the random sample of the study population Results: The HRA Cox proportional hazard model of breast cancer developed in the study included age, age², age at menarche, and lactation. Multivariate Cox proportional hazards models were used to test the HRA model validation Compared with women in the lowest 10 percentile of subjects, women in the highest 10 percentile had an increased fisk of breast cancer (relative fisk, 3.701, 95% confidence interval, 2.554 to 5.364) Conclusion: The model's performance was satisfactory for estimating breast cancer fisk for individual women aged 30years.

허혈성심질환 발생예측모형 (health risk appraisal) 개발 연구

지선하 ( Sun Ha Jee ),송지원 ( Ji Won Song ),조홍근 ( Hong Keun Cho ),김상연 ( Sang Yon Kim ),장양수 ( Yang Soo Jang ),김정희 ( Jung Hee Kim ) 한국지질동맥경화학회(구 한국지질학회) 2004 韓國脂質學會誌 Vol.14 No.2

2012년 흡연으로 인한 건강보험 진료비 추정 연구

지선하 ( Sun Ha Jee ),정금지 ( Keum Ji Jung ),전티나 ( Christina Jeon ),김희진 ( Hee Jin Kim ),윤영덕 ( Young Duk Yun ),김일순 ( Il Soon Kim ) 한국보건정보통계학회(구 한국보건통계학회) 2014 보건정보통계학회지 Vol.39 No.1

Objectives: The purpose of the study was to estimate relative risk and attributable risk of 35 tobacco-related diseases and to compute total medical expenses on smoking by providing a cohort study with 20 years follow-up period. Methods: Smoking-attributable medical costs were calculated by applying the percentages of population attributable risks (PARs) to the estimated medical costs by the tobacco related diseases in 2012. In this study, PARs were obtained by using relative risks from the Korean Cancer Prevention Study and the previous studies, and population smoking prevalence surveyed in 1990 conducted by Korean Institute of Tuberculosis. Results: As a result, the medical expenses from tobacco use were 1,846,562,350,000 won (about 3.86% of total medical expenses). The top 5 medical expenses on tobacco-related diseases were ischemic heart diseases, cerebrovascular diseases, lung cancer, diabetes, and chronic obstructive pulmonary disease, respectively. More than a half percent of total medical expenses (about one billion dollars) were spent from these five, tobacco-related diseases. Conclusions: While the harmful effect of smoking is expected to have a steady increase for a while, antismoking policy should be reinforced to reduce the risk of disease incidence, and the medical expenses for treating the tobacco-related diseases.

한국인의 심혈관질환 위험 예측보고

지선하 ( Sun Ha Jee ) 한국지질동맥경화학회(구 한국지질학회) 2008 한국지질동맥경화학회 추계학술발표논문집 Vol.2008 No.-

혈청 아디포넥틴 측정 키트(AdipoMark)의 개발과 심혈관질환 위험요인과의 관련성 연구

지선하(Sun Ha Jee),이승복(Seungbok Lee),민성식(Sungsik Min),박정용(Jungyong Park),김현숙(Hyon-Suk Kim),김상연(Sang Yeun Kim),윤지은(Ji Eun Yun),이선주(Sun Ju Lee),지은정(Eun Jung Jee),이희연(Hee Yeon Lee),송혜윤(Hye Yun Song) 한국역학회 2007 Epidemiology and Health Vol.29 No.2

  Purposes: To develop a rapid, sensitive, qualitative ELISA-kit for serum adiponectin and examine correlation with adiponectin and cardiovascular risk factors.   Methods: On the base of monoclonal antibodies against adiponectin, apply indirect ELISA to study the performance parameter of the kit. The correlation was examined between adiponectin and cardiovascular risk factors including waist circumference, body mass index, triglyceride, and HDL cholesterol.   Results: The limited concentration of detection of the ELISA-kit was 1㎍/㎖. Linearity with R&D system and AdipoGen with this ELISA-kit was acceptable: the linear equation with R&D system was y=1.0116x+0.4629 (R2=0.97) and linear equation with AdipoGen was y=0.9562x+1.1961 (R2=0.93), respectively. The average recovery rate of the ELISA-kit ranged 92 to 104%. The correlation coefficient of waist circumference with adiponectin was -0.2276 (p<0.0001) among men and -0.2328 (p<0.0001) among women.   Conclusion: This ELISA-kit was quick, sensitive, and stable and can be used to determine adiponectin in serum.

한국인의 혈청 아디포넥틴과 당뇨병 진단기준 설정에 관한 연구

지선하(Sun Ha Jee),이희연(Hee Yeon Lee),이선주(Sun Ju Lee),윤지은(Ji Eun Yun),지은정(Eun Jung Jee),송혜윤(Hye Yun Song),김상연(Sang Yeun Kim),박정용(Jungyong Park),김현숙(Hyon-Suk Kim) 한국역학회 2007 Epidemiology and Health Vol.29 No.2

  Background: As indicators of obesity, waist circumference (WC), body mass index (BMI), and adiponectin are well known risk factor for diabetes mellitus. The objectives of this study were to measure the independent association between these obesity indicators and diabetes and to examine the combined effect of these indicators on diabetes in a Korean population.   Methods: The WC, BMI, and serum adiponectin were measured in 6,505 healthy Koreans and were classified into tertile groups for men and women. The independent and combined associations of the obesity indicators with diabetes were measured using logistic regression analyses. Diabetes was defined as fasting serum glucose greater than 126㎎/dL or taking medication.   Results: Levels of adiponectin were inversely associated with BMI and WC and directly associated with age and high density lipoprotein cholesterol (HDL) cholesterol (P<0.001). After adjusting for age, WC, and other lifestyle factors, low levels of adiponectin were associated with an increased prevalence of diabetes. Further adjustment for HDL cholesterol and triglyceride attenuated this association in both men and women. The best cut-off value of adiponectin in terms of identifying the presence of diabetes was 5.5㎍/㎖ with a sensitivity and specificity of 46.7% and 63.9% for men and 9.5㎍/㎖ with a sensitivity and specificity of 68.2% and 55.2 for women.   Conclusions: These results suggest that adiponectin was associated with diabetes. The association was independent of WC and was partly modified by HDL and triglyceride. There were no effect modifications of adiponectin with WC on diabetes.

한국인 흡연과 사망위험에 대한 11년 추적연구

지선하(Sun-Ha Jee),조인호(In-Ho Jo),윤지은(Ji-Eun Yun),박정용(Jung-Yong Park),설재웅(Jae-Woong Sull),오희철(Hee-Choul Ohrr),이상이(Sang-Vi Lee),윤유식(Yoo-Sik Yoon),Jonathan M Samet, 김일순(II-Soon Kim) 한국역학회 2005 Epidemiology and Health Vol.27 No.1

Objective : In Korea, male smoking prevalence is among the world's highest and mortality rates from smoking-caused cancers, particularly lung cancer, are escalating. This cohort study examined the effects of cigarette smoking on the risk of mortality from all causes, cancers and cardiovascular diseases(CVD), and characterized the relationship of the risk with the amount and duration of cigarette smoking. Method : A eleven-year prospective cohort study was carried out of on 1,207,592 Koreans, 30 to 95 years of age. The study population includes participants in a national insurance program, who completed a questionnaire on smoking and other risk factors. The main outcome measures were death from all causes, cancer and CVD, obtained through record linkage. At baseline, 482,997 men(60.0%) and 19,755(5.3%) women were current cigarette smokers. Results : In multivariate Cox proportional hazards models, controlling for age, alcohol drinking, exercise, and obesity, current smoking among men increased the risks of mortality from all cause death (relative risk[RR], 1.56; 95% confidence interval[CI], 1.52~1.59), all cancer (1.75, 1.68~1.82), and CVD(1.46, 1.38~1.55). Similar results were found for mortality among women. Smoking also increased the risks of mortality for cancer of the lung(4.60, 4.09~5.33) and other cancers, including larynx, bile duct, esophagus, liver, stomach, pancreas, bladder, and also leukemia. Current smoking among women increased the risk of lung cancer mortality(RR=2.83, 95% CI 2.38~3.36). Conclusion : In Korea, smoking is an independent risk factor for death from all causes, CVD and a number of major cancers. The findings affirm the need for aggressive tobacco control in Korea in order to minimize the epidemic of smoking-caused disease.

메타분석 방법을 적용한 서울시 대기오염과 조기사망의 상관성 연구 (1991년$\sim$1995년)

김춘배,지선하,정용,이종태,Kim, Chun-Bae,Jee, Sun-Ha,Chung, Yong,Lee, Jong-Tae 대한예방의학회 1999 예방의학회지 Vol.32 No.2

유럽과 미국을 비롯한 선진 산업국가에서 1930년에서 1950대 사이에 발생하였던 일련의 대기오염사건을 경험한 이후, 대기오염과 인체건강영향과의 상관성을 평가하려는 체계적인 연구수행이 있어왔다. 국내에서도 최근 2-3년 동안 단면적 연구 설계에서 벗어나 시계열적 분석방법을 적용한 연구결과들이 발표되었으며, 외국에서의 결과와 유사하게 현재 수준의 대기오염도에서도 인체건강영향 특히 조기사망발생과 유의한 상관성이 있음이 제시되었다. 특히 서울시를 대상으로 한 일련의 연구결과가 대기오염도와 일별사망과에 유의한 상관관계가 있음을 보이고 있어 이에 대한 보다 정밀한 연구수행 필요성이 제기되었다. 따라서 본 연구는 메타분석적 방법론을 적용하여 서울시내의 구별분석 결과를 통합하여 기존의 연구결과와 비교함으로써 대기오염과 인체건강영향의 원인적 상관성을 규명하는 보조적 자료를 제공하기 위하여 수행되었다. 대기오염 자동측정기가 설치된 서울시 내 총 18개 구를 대상으로 대기오염도와 사망과의 상관관계를 추정하는 구별 예측치를 산출하였다. 이렇게 산출된 각 개별 예측치는 전체로서 총괄되는 가중평균 예측치를 계산하기 위하여 사용되었으며 이 때 모수효과 모형 또는 랜덤효과 모형을 적용하여 가중평균 예측치를 산출하였다. 지역별 또는 구별 예측치간의 변이차를 검정하는 동질성 검정의 결과 세 오염물질의 경우 모두 지역간 변이가 큰 것으로 평가되었으며 따라서 랜덤효과 모형의 결과를 최종결과로 선정하여 제시하였다. 랜덤효과 모형의 결과를 보면 총부유분진과 아황산가스의 경우 일별사망과 유의한 상관관계가 있는 것으로 평가되었으나 오존의 경우는 그렇지 않았다. 또한 메타분석 결과가 기존의 연구결과와 비교할 때 유사하게 평가된 점으로 미루어 현재 수준의 대기오염도와 일별사망 또는 인체건강영향과 밀접한 상관관계가 있음을 알 수 있다. 메타분석 방법을 적용하였어도 직접적 개인노출평가가 이루어지지 않았다는 점에서 정보편견의 가능성을 제외할 수 없다. 그러나 이러한 메타분석 방법이 기존의 연구(서울시 전체를 대상으로 하는)에 비하여 정보편견을 어느정도 줄여줄 수 있을 것으로 기대되었으며, 분석결과 기존 연구에서 제시하는 것과 매우 유사함을 알 수 있다. 결론적으로 대기오염 역학 연구에 있어서 이와 같은 메타분석적 방법이 유용하며, 본 연구결과도 기존의 연구에서와 같이 현재 수준의 대기오염도가 인체 건강에 위해한 영향을 미칠 수 있음을 보여 준다는 점에서 향후 보다 강화된 대기오염관리 방안 마련의 필요성을 제기한다고 하겠다. 이를 위하여 대기 오염기준치의 강화뿐 아니라 대기오염에 민감한 인구집단, 즉 어린이나 노약자 등에 대한 각별한 감시와 연구수행이 필요시 된다. Objectives: To reexamine the association between air pollution and daily mortality in Seoul, Korea using a method of meta-analysis with the data filed for 1991 through 1995. Methods: A separate Poisson regression analysis on each district within the metropolitan area of Seoul was conducted to regress daily death counts on levels of each ambient air pollutant, such as total suspended particulates (TSP), sulfur dioxide $(SO_2)$, and ozone $(O_3)$, controlling for variability in the weather condition. We calculated a weighted mean as a meta-analysis summary of the estimates and its standard error. Results: We found that the p value from each pollutant model to test the homogeneity assumption was small (p<0.01) because of the large disparity among district-specific estimates. Therefore, all results reported here were estimated from the random effect model. Using the weighted mean that we calculated, the mortality at a $100{\mu}g/m^3$ increment in a 3-day moving average of TSP levels was 1.034 (95% Cl 1.009-1.059). The mortality was estimated to increase 6% (95% Cl 3-10%) and 3% (95% Cl 0-6%) with each 50 ppb increase for 9-day moving average of SO2 and 1-hr maximum O3, respectively. Conclusions: Like most of air pollution epidemiologic studies, this meta-analysis cannot avoid fleeing from measurement misclassification since no personal measurement was taken. However, we can expect that a measurement bias be reduced in a district-specific estimate since a monitoring station is hefter representative cf air quality of the matched district. The similar results to those from the previous studios indicated existence of health effect of air pollution at current levels in many industrialized countries, including Korea.