Original Article J Health Info Stat 2016;41(4):403-410 https://doi.org/10.21032/jhis.2016.41.4.403 pissn 2465-8014 eissn 2465-8022 공복혈당궤적에따른심장병발생위험 : 국민건강보험공단표본코호트연구 전주은, 조어린, 정금지, 지선하 연세대학교보건대학원국민건강증진연구소 Fasting Blood Glucose Levels Trajectory and Risk of Cardiovascular Disease in Korean Population: National Health Insurance Service-National Sample Cohort (NHIS-NSC) Jooeun Jeon, Eo Rin Cho, Keum Ji Jung, Sun Ha Jee Department of Epidemiology and Health Promotion, Institute for Health Promotion, Yonsei University, Seoul, Korea Objectives: To classify trajectories of fasting blood glucose (FBS) levels and examine each trajectory s associations with risk of cardiovascular disease (CVD). Methods: The National Health Insurance Service-National Sample Cohort (NHIS-NSC) sampled in the 2002 NHIS database was followed until 2010, and 13,829 participants aged 20 years and above had conducted nationwide health examinations annually. We used Cox proportional hazards models to examine the association of trajectories to risk of CVD. Four distinct trajectory groups were identified for FBS: low-stable, moderate-stable, elevated-upward, and High-upward. Results: During 88947.9 person-years of follow-up (mean follow-up, 6.4 years), we documented 2,778 incident case of CVD. Age-standardized incidence rate were increased with FBS levels (5,296.2 in low-stable group, 6,292.6 in moderate-stable group, and 8047.9 in elevated-upward group), but not in High-upward group. In multivariate models adjusted for age and sex, FBS was a significant predictor of CVD in elevatedupward group (hazard ratio (HR) = 1.6, 95% confidence interval (CI):1.4-1.8) and High-upward group (HR = 1.6, 95% CI:1.3-2.1). However, further adjustment for clinical covariates, only elevated-upward group was significantly associated with CVD (HR = 1.2, 95% CI:1.1-1.4). Conclusions: Using the trajectory approach, we found that elevated-upward and High-upward FBS trajectories were associated with greater risk of CVD. These findings indicate the importance of FBS management across the lifespan, prognostic assessments and the targeting of prevention strategies to high-risk individuals. Key words: Fasting blood sugar, Cardiovascular disease, Diabetes, Hypertension, Stroke 서론 제2형당뇨병은심장병에대한위험증가와관련있으며, 이는주로임상적당뇨병발생이전에심장병이발생한다 [1-4]. 최근의한인구집단을대상으로한코호트연구에의하면공복혈당장애또한당뇨병과동일한수준으로사망위험성을증가시킨다고보고하였다 [5]. 이에대 한기전은명확하지않지만, 당뇨병환자에서조절되지않는혈당상태는심혈관질환으로인한사망위험률및이로인한합병증과관련된다는보고가널리알려져왔으며 [6-8], 또한이전의많은연구들을통해공복혈당장애와심장병으로인한사망과의관련성이밝혀졌다 [9-11]. 전세계적으로당뇨유병률은매년급증하고있으며, 당뇨병으로인한사망또한매년증가하고있다 [12,13]. 더욱이당뇨병발병연령은점 Corresponding author: Sun Ha Jee 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, Korea Tel: +82-2-2228-1523, E-mail: jsunha@yuhs.ac Received: October 22, 2016 Revised: November 28, 2016 Accepted: November 28, 2016 *This study was funded by a grant of the Korean Health Technology R&D Project, Ministry of Health & Welfare, Korea (HI13C0715). No potential conflict of interest relevant to this article was reported. How to cite this article: Jeon J, Cho ER, Jung KJ, Jee SH. Fasting Blood Glucose Levels Trajectory and Risk of Cardiovascular Disease in Korean Population: National Health Insurance Service-National Sample Cohort (NHIS-NSC). J Health Info Stat 2016;41(4):403-410. Doi: https://doi.org/10.21032/jhis.2016.41.4.403 It is identical to the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) whichpermit sunrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. 2016 http://www.e-jhis.org 403
Jooeun Jeon, et al. 차낮아지고있으며, 성인초기당뇨병발생위험은증가하고있는가운데, 성인초기에공복혈당장애로정의되는당뇨병전단계유병률또한꾸준히증가하고있다 [14-16]. 최근미국에서는당뇨전단계인사람들이급증하는추세에있고 [16], 국내에서도당뇨병유병률이최근몇년사이소폭증가하였으며, 공복혈당장애유병률또한 30세이상의성인에서 25% 에이르는것으로나타났다 [17]. 당뇨전단계는추후당뇨병및심장병으로발전할가능성이높은잠재적인당뇨병이므로, 국가적인당뇨병관리체계에있어매우중요한예방및관리대상이다. 이를위해당뇨병및심장병발생이전의공복혈당에대하여지속적으로추적관리하는것이무엇보다중요하다 [18]. 이에집단중심추세모형방법론 (group based trajectory modeling methods, GBTM) 을활용하면공복혈당추적관리기간동안시간에따른특유의변화양상을파악하고각기다른양상에따라실제심장병발생양상또한어떠한차이를보이는지파악가능하다. 따라서본연구에서는국민건강보험공단표본코호트자료를통하여공복혈당과심장병발생위험과의관련성을집단중심추세모형방법론을이용한잠재계층분석 (latent class growth analysis, LCGA) 을통해알아보고자하였다. 연구방법 연구대상본연구대상자는전국민을대상으로한국민건강보험공단표본코호트자료에서 2002년부터 2010년까지 9년간매년건강검진을받은수검자를기준으로하였다. 2002년이후, 9년간의검진정보를지닌수검자 15,130명중에서 20 세미만인 566명과당뇨병, 고혈압, 심장병및뇌졸중과거력이있는 412명, 검진이후심혈관질환진단전에이미진단받은자 299명및공복혈당, 수축기혈압, 이완기혈압, 총콜레스테롤변수에결측값을지닌 24명을제외한 13,829 명을최종분석대상으로하였다. 국민건강보험공단표본코호트자료국민건강보험공단표본코호트자료의근간은국민건강보험공단에서 2011 년에구축한 2002년부터 2010년까지건강보험및의료급여권자전체에대한진료명세서와진료내역, 상병내역, 처방전내역등을기존청구일중심에서진료개시일중심으로자료구조를조정한 국민건강정보 DB 이다. 이는자격 DB, 진료 DB 및검진 DB 등으로구성되어있으며, 사실상장기간의전국민전향적코호트연구또는중재의효과를검증할수있는대규모데이터베이스이다 [19]. 본연구는국민건강정보 DB의자격 DB로부터추출된표본중건강검진을받은경우에 만해당되는건강검진 DB를기준으로하였다. 검진정보는 2002년에서 2010년까지국민건강보험공단에서국민들을대상으로실시한일반검진 1차및 2차, 생애전환기건강진단 1차및 2차, 암검진및영유아검진자료등으로구성되어있다. 이러한국가건강검진의수검률은본연구2010년기준, 국민건강보험대상자의약 70% 정도이며 [19], 2014 년현재기준, 약 75% 에달한다 [20]. 연구에사용된변수현재국가검진관련하여당뇨병진단에쓰이는대표적인방법은공복혈당수치로, 식후 8-14시간가량지난뒤혈장속에포함된포도당의양을측정한다 [10]. 당뇨병진단에는공복혈당수치뿐아니라식후혈당및당화혈색소 (HbA1c) 측정치를사용하기도하지만, 국민건강보험공단표본코호트자료에서활용가능한당뇨병진단관련변수로서, 보다보편적지표인공복혈당을본분석에사용하였다 [21]. 이외에연구에포함된변수는일반적특성 ( 성, 연령 ), 생활습관 ( 흡연, 음주, 운동 ), 과거력 ( 결핵, 간염, 간장질환, 암, 기타질환 ), 가족력 ( 당뇨병, 고혈압, 심장병, 뇌졸중 ) 에관한정보를수집하였으며, 건강검진결과로서키, 체중, 수축기혈압, 이완기혈압, 총콜레스테롤수치를포함하였다. 연구대상자의심장병진단은본연구자료의근간인국민건강보험공단표본코호트자료의진료DB에서상병정보를담고있는상병내역자료및명세서자료를연계하여확인하였으며, 상병코드는국제질병분류 ICD-10 기준으로하였다. 표본코호트자료의특성상상병내역자료와명세서자료는청구일련번호로연결되는데, 자료의충실도측면에서청구를목적으로한자료라는자료의제한점이있다. 따라서보다정확한진단기준을적용하기위해해당질병에대해입원이력이있는자들을대상으로하였다. 이를위해 2002년이후명세서자료내에포함된 최초입원일 날짜기준으로하여주상병및부상병코드가심장병인경우, 심장병첫진단일변수를생성하였고, 해당변수를지닌대상자는심장병발생자로분류하였다. 본연구의심장병진단코드로는이전연구를참고하여, 고혈압성심장질환 (I10 I15), 허혈성심장질환 (I20 I25), 죽상경화증관련심장질환 (I44 I52), 동맥성심장질환 (I70 I74), 출혈성뇌졸중 (I60 I62) 및허혈성뇌졸중 (I63 I66) 을포함하였다 [4,22]. 집단중심추세모형방법론집단중심추세모형방법론은연구대상자들각각의개인적특성을규명하고, 특정변수에대해시간또는연령에따른유사한진행패턴을파악하고적합하게분류해주는방법이다. 본연구에서는집단중심추세모형방법론중한가지인잠재계층분석기법을이용하여분석하였다. 잠재계층분석기법은 Nagin et al. [23] 에의해그이론이발전하여 404 http://www.e-jhis.org
Fasting Blood Glucose Levels Trajectory and Risk of Cardiovascular Disease in Korean Population Figure 1. Basic assumptions of group based trajectory modeling methods. 왔으며, 분석도구로는 SAS 소프트웨어의 proc traj나 Stata 소프트웨어의 Traj이주로쓰인다 [23,24]. 본연구에서는 SAS 소프트웨어의 proc traj을분석에활용하였다. 집단중심추세모형방법론에서는데이터의형식에따라분석시각기다른모형을적용하는데, 이분형자료에서는주로로짓기반모형 (logit-based model) 을적용하며, 심리측정식계층구조형자료 (psychometric scale data) 의경우중도절단모형 (censored normal (tobit) model, CNORM) 을적용한다. 집단중심추세모형방법론의주요한가정으로는시간에독립적인공변량은각그룹에속한개인을통해서만관찰되는추세와연관되며, 시간에의존적인공변량은관찰되는추세와직접적으로연관된다는것이다 (Figure 1). 이러한가정에따라, 분석시행시시간에따라변화하는변수와그렇지않은변수들의효과크기를고려하여각개인별그룹에속할확률값을계산해주게된다. 그룹개수에대한모형적합도비교는베이지안정보척도 (Bayesian information criterion) 를통해시행되며, 그값이작을수록적합하다고판정한다 [23]. 잠재계층분석에따른공복혈당그룹본연구에서는 2002년부터 2010년까지의 9년간매년건강검진을받은사람을기준으로하여, 공복혈당수치에로그를취한값으로중도절단모형을통한잠재계층분석을실시하였고, 각각의상이한궤적을통해적합한그룹화를시행한결과 BIC값이가장작은 4개의그룹으로대상자를분류하였다. 4개로분류된그룹별로 2002년에서 2010년까지의공복혈당궤적의일반적특성을파악한후각각의그룹명을부여하였다. 2002년 83.8 mg/dl에서 2010년 88.7 mg/dl로 80-90 mg/dl 수준에머물러있던그룹은 Low-stable 로, 2002년 95.4 mg/dl에서 2010년 99.3 mg/dl로 91-100 mg/dl 수준에머물러있던그룹을 Moderate-stable 로, 2002년 109.4 mg/dl 에서 2010 년 125.9 mg/dl 로이상혈당증범위였으며, 9년동안약간증가 (16.5 mg/dl) 한그룹을 Elevated-upward, 2002년 150.3 mg/dl에서 2010년 179.9 mg/dl로당뇨병범위였으며, 9년동안크게증가 (29.6 mg/dl) 한그룹은 High-upward 이라명명하였다. 통계분석방법잠재계층분석시행결과로그룹화된각공복혈당그룹별일반적특성을파악하였으며, 9년간의공복혈당궤적추세특성에따른십만인년당심장병발생률과심장병발생에대한비교위험도분석을하였다. 또한혼란변수를통제한상태에서콕스비례위험모형 (Cox proportional hazard model) 을통해심장병발생위험도를추정하였으며, 연구대상집단표본크기 (sample size) 에대한콕스비례위험모형가정이만족하는지검토하였다. 분석도구로는 SAS 9.2 프로그램 (SAS Institute, Cary, NC, USA) 을사용하였으며, 통계적유의성은 p-value 0.05미만을기준으로하였다. 연구결과 13,829명의최종분석대상중남자가여자보다월등히많았는데, 9 년간매년검진을받은자를기준으로하여자격 DB의연령그룹변수와성별변수를취한결과, 이와같이매년검진을받은검진 DB의대상자대부분은남자였다. 전체연구대상자의연령대에서 20대와 30대가각각 39.6%, 39.5% 로가장많았으며, 남자에서는 20대 (40.2%) 가, 여자에서는 30대 (37.2%) 가가장많았다. 연구대상자의평균공복혈당수치는남자가여자에비해높았다 (Table 1). 중도절단모형을통하여연구대상자들에대하여 2002년부터 2010년까지의 9년간공복혈당수치에로그를취한값에대한잠재계층분석을실시하였다. 잠재계층분석결과각그룹에대한확률은모두 p-value 0.0001 미만으로통계적인적합수준을만족하였으며, 그룹개수에대한 BIC최소값은 63,348.8로최종4개그룹으로대상자를그룹화하였다. 전체연구대상자에대한각그룹별비율은 Low-stable 47.9% (n = 6,670), Moderate-stable 44.1% (n = 6,087), Elevated upward 6.7% (n=189), High-upward 14% (n= 883) 이었다 (Figure 2). Table 2는잠재계층분석을통한공복혈당그룹별일반적특성을보여주고있다. 4개그룹중 Low-stable 그룹에서여성의비율이가장높았고가장젊은그룹이었으며, 심장병과고혈압가족력을지닌대상자비율은가장높았다. 반면에 High-upward 그룹에서는남성의비율이월등히높았고 30대연령의대상자가가장많았으며, 체질량지수 25 이상의비율이약 58% 에달하였으나, 심장병과고혈압가족력을지닌대상자비율은가장낮았다. 본연구대상자전체평균추적기간은 6.4년이었고, 2010년인구를표준인구로하였을때, 각그룹별심장병에대한 10만명당발생률은 Elevated-upward 그룹이가장높았다 (Table 3, Figure 3). Table 3은공복혈당궤적에따른그룹별심장병발생위험도분석결과를보여주고있다. Low-stable 그룹을기준으로하였을때, 모형1 에서는 Moderate- http://www.e-jhis.org 405
Jooeun Jeon, et al. Table 1.General characteristics of study participants by sex (n=13,829) stable 그룹이 1.3 배, Elevated upward 그룹은 2.1 배, High-upward 그 룹에서는 2.2 배로심장병발생위험이유의하게높았다 (p for trend < 0.0001). 연령과성을보정한모형 2 에서는 Low-stable 그룹에비해 Moderate-stable 그룹은 1.1 배, Elevated upward 그룹은 1.6 배, Highupward 그룹에서는 1.6 배심장병발생위험이유의하게높았으며, 흡 연상태, 체질량지수, 수축기혈압, 총콜레스테롤변수와심장질환, 뇌 졸중, 고혈압및당뇨병에대한가족력을추가보정한모형 3 에서는 Low-stable 그룹에비해 Elevated upward 그룹만심장병에대한발 생위험도가 1.2 배유의하게높았다. Men (n = 11,830) n (%) Women (n = 1,999) p-value Age (y) 20-29 4760 (40.2) 723 (36.2) <0.000 30-39 4717 (40.1) 743 (37.2) 40-49 2129 (18.0) 503 (25.2) 50 200 (1.7) 30 (1.5) Body mass index (BMI, kg/m 2 ) BMI 18.5 293 (2.5) 144 (7.2) <0.000 18.5<BMI<23 4840 (40.9) 1123 (56.2) 23 BMI<25 3271 (27.7) 371 (18.6) 25 BMI<30 3209 (27.1) 341 (17.1) 32 BMI 217 (1.8) 20 (1.0) Family history Chronic heart disease 355 (3.0) 70 (3.5) 0.469 (n = 12,138) Cerebrovascular disease 567 (4.8) 92 (4.6) 0.923 (n = 12,165) Hypertension (n=12,166) 735 (6.2) 153 (7.7) 0.049 Type 2 diabetes (n=12,172) 782 (6.6) 133 (6.7) 0.981 Mean ± SD Fasting blood sugar (mg/dl) 92.1±20.5 87.9±15.0 <0.000 Systolic blood pressure (mmhg) 123.8±13.8 116.6±13.8 <0.000 Diastolic blood pressure (mmhg) 78.9±10.0 74.2±10.0 <0.000 Total cholesterol (mg/dl) 192.0±35.1 184.1±33.2 <0.000 SD, standard deviation. FBS (mg/dl) Group percents 47.9 44.1 6.7 14 181.3 164.0 High-upward 148.4 134.3 121.5 Elevated-upward 109.9 99.5 90.0 Moderate-stable Low-stable 81.4 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 Year Figure 2. Change in log FBS levels during 9 years by trajectory groups. FBS, fasting blood sugar. 들의연령대자체가다소젊은편인것을감안할때, 만성질환발생에 대한위험도를산출하기에는젊은대상자들에대한 9 년의단기추적 으로는충분하지않은결과로반영되어나타나게된것으로생각된다. 또한, 2002 년부터당뇨병수준의공복혈당수치를지닌대상자들은그 만큼심장병및당뇨병과같은만성질환에대한내재적위험이상당할 것이고, 따라서꾸준한건강관리를하였을가능성이높다. 당뇨전단 계에서중요한예방책으로는생활습관의변화, 특히체중감소와운동 량을늘리는것이라고알려져있는데, 따라서그러한효과의반영으로 인해 High-upward 그룹의결과에서는관련성이나타나지않았다고 볼수도있다 [25]. 이와더불어, 공복혈당과심장병으로인한사망관련 된이전의한연구에서당뇨병유병상태인대상자들보다는당뇨병 유병상태는아니나비교적높은수준의공복혈당수준을보인대상 자들에서심장병관련사망에대한비교위험도가더높게나타났는데 [26], 본연구에서도이와유사한결과를보인점에서, 대표성을지닌 비교적젊은코호트자료를이용하여, 잠재계층분석을통한 9 년의짧 은추적기간동안의공복혈당변화에따른심장병발생위험추정을 하였다는점에서연구자체에의미를부여할수있다고본다. 익히알려진당뇨병과심장병의관련성이외에당뇨전단계와심장 병과의관련성에대해서도지금껏많은연구들이이루어져왔다. 2010 년에시행된한메타분석연구에따르면 20 개의연구에대해고정효과 로결합된결과를추정하였을때유의한관련성을보인바있다 [27]. 이 고찰 2002년 109.4 mg/dl 에서 2013년 125.9 mg/dl로이상혈당증범위였으며, 9년동안약간의증가를보였던 Elevated upward 그룹에서심장병발생위험이가장높았다. 2002년부터공복혈당이당뇨병수준이었다가 9년동안계속해서크게증가하였던 High-upward 그룹에서는심장병발생위험과관련성이보이지않았는데, 이는국민건강보험공단표본코호트자료대상자 러한이전연구결과들에비추어, 당뇨전단계에서의심장병위험에대한중요성을시사하는데본연구의결과또한도움이될것을기대하는바이다. 우리나라의당뇨병유병률추이를살펴보면지난 2005년부터꾸준히약 9% 를유지하다가지난 2013년 11.0%, 2014년 10.2% 로약 1-2% 증가하였다. 만 30세이상당뇨병유병률은전체 11.1%, 공복혈당장애유병률에서는전체 25.0% 였으며, 성별로구분해보았을때남자 30.0%, 여자 20.1% 로남자가더높았다 [17,28,29]. 이처럼세계적인당뇨병및당뇨전단계유병률증가추세와더불어우리나라또한증가 406 http://www.e-jhis.org
Fasting Blood Glucose Levels Trajectory and Risk of Cardiovascular Disease in Korean Population Table 2. General characteristics of study participants according to fasting blood sugar trajectories at baseline and 9 years (n=13,829) Low-stable Moderate-stable Elevated-upward High-upward Age (y) 20-29 3,207 (48.1) 2,380 (34.4) 151 (17.1) 31 (16.4) <0.000 30-39 2,380 (35.7) 2,591 (42.6) 422 (47.8) 91 (48.2) 40-49 994 (14.9) 1,294 (21.3) 285 (32.3) 59 (31.2) 50 89 (1.3) 108 (1.8) 25 (2.8) 8 (4.2) Sex (women) 1,367 (20.5) 566 (9.3) 56 (6.3) 10 (5.3) <0.000 Body mass index (BMI, kg/m 2 ) At baseline BMI 18.5 291 (4.4) 140 (2.3) 5 (0.6) 1 (0.5) <0.000 18.5<BMI<23 3,249 (48.7) 2,443 (40.1) 231 (26.2) 40 (21.2) 23 BMI<25 1,626 (24.4) 1,734 (28.5) 243 (27.5) 39 (20.6) 25 BMI<30 1,429 (21.4) 1,664 (27.3) 361 (40.9) 96 (50.8) 30 BMI 75 (1.1) 106 (1.7) 43 (4.9) 13 (6.9) At endline 25 BMI 1,864 (27.9) 2,142 (35.2) 396 (44.8) 80 (42.3) <0.000 Family history Chronic heart disease (n=12,138) 213 (3.2) 186 (3.1) 22 (2.5) 4 (2.1) <0.000 Cerebrovascular disease (n=12,165) 303 (4.5) 308 (5.1) 40 (4.5) 8 (4.2) <0.000 Hypertension (n=12,166) 445 (6.7) 377 (6.2) 56 (6.3) 10 (5.3) <0.000 Type 2 diabetes (n=12,172) 418 (6.3) 388 (6.4) 93 (10.5) 16 (8.5) <0.000 Smoking status (n=12,029) Never smoker 2,844 (42.6) 2,138 (35.1) 309 (35.0) 56 (29.6) <0.000 Current smoker 2,637 (39.5) 2,722 (44.7) 401 (45.4) 92 (48.7) Former smoker 366 (5.5) 400 (6.6) 55 (6.2) 9 (4.8) Drinking status ( per once) Non drinker 2,076 (31.1) 1,442 (23.7) 206 (23.3) 45 (23.8) <0.000 Less than half a bottle of Soju 1,655 (24.8) 1,334 (21.9) 177 (20.1) 32 (16.9) Half a bottle - a bottle of Soju 2,105 (31.6) 2,314 (38.0) 328 (37.2) 76 (40.2) A bottle - 1.5 bottle of Soju 624 (9.4) 753 (12.4) 130 (14.7) 23 (12.2) 1.5 bottle - 2 bottle of Soju or more 210 (3.2) 244 (4.0) 42 (4.8) 13 (6.9) Regular exercise (per a wk) (n=13,395) No exercise 3,088 (46.3) 2,623 (43.1) 369 (41.8) 81 (42.9) <0.000 1-2 days 2,270 (34.0) 2,215 (36.4) 337 (38.2) 57 (30.2) 3-4 days 713 (10.7) 684 (11.2) 102 (11.6) 21 (11.1) 5-6 days 146 (2.2) 124 (2.0) 16 (1.8) 3 (1.6) Everyday 253 (3.8) 228 (3.8) 41 (4.6) 17 (9.0) Mean ± SD Fasting blood sugar (mg/dl) At baseline 83.8±11.0 95.4±16.9 109.4±24.54 150.3±69.4 <0.000 At endline 88.7±10.3 99.3±12.2 125.9±30.0 179.9±62.4 <0.000 Systolic blood pressure (mmhg) At baseline 120.6±13.5 124.0±13.8 129.0±15.6 131.1±15.5 <0.000 At endline 120.9±12.6 124.6±13.0 127.6±12.9 128.7±14.9 <0.000 Diastolic blood pressure (mmhg) At baseline 76.8±9.9 79.0±9.9 82.7±11.0 84.7±11.0 <0.000 At endline 76.6±9.1 78.7±9.3 80.7±9.1 80.8±9.6 <0.000 Total cholesterol (mg/dl) At baseline 186.8±34.1 192.8±34.6 203.0±37.2 213.2±36.8 <0.000 At endline 195.1±33.5 198.5±34.6 198.9±39.1 195.2±40.8 <0.000 SD, standard deviation. n (%) p-value http://www.e-jhis.org 407
Jooeun Jeon, et al. Table 3. Hazard ratios (95% CI) for risk of CVD according to FBS group by trajectory (n=13,829) No. of persons No. of CVD incidences Person years, follow-up Age-adjusted rate FBS (mmhg) Model 1 Model 2 Model 3 At baseline At endline HR (95% CI) HR (95% CI) HR (95% CI) Low-stable 6,670 1,086 37,297.2 5,296.2 83.8 ± 11.0 88.7 ± 10.3 1.0 1.0 1.0 Moderate-stable 6,087 1,326 43,401.1 6,292.6 95.4±16.9 99.3±12.2 1.3 (1.2-1.4) 1.1 (1.0-1.2) 1.0 (0.9-1.1) Elevated-upward 883 309 1,765.1 8,047.9 109.4±24.5 125.9±30.0 2.1 (1.8-2.3) 1.6 (1.4-1.8) 1.2 (1.1-1.4) High-upward 189 67 409.9 5,326.6 150.3±69.4 179.9±62.4 2.2 (1.7-2.8) 1.6 (1.3-2.1) 1.0 (0.8-1.3) CI, confidence interval; CVD, cardiovascular disease; FBS, fasting blood sugar; HR, hazard ratio. Model 1: Uunadjusted, Model 2: Adjusted for age and sex, Model 3: Adjusted for age, sex, current smoker, body mass index, systolic blood pressure, total cholesterol and family history for cardiovascular disease. 1,400 9,000 Age standadized rate of CVD incidence (per 100,000 person-yr) 1,200 1,000 800 600 400 200 8,000 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 Low-stable Moderate-stable Elevated-upward High-upward 0 # of each group of CVD incidence Age-standadized rate of each group 1,086 1,326 309 67 5,296.2 6,292.6 8,047.9 5,326.6 Figure 3. Age standardized rate (per 100,000 person-year) and number of CVD incidence by trajectory groups. CVD, cardiovascular disease. Age-adjusted death rate per 100,000 person-years using the 2010 standard population (Statistics Korea). 추세이다. 최근한관련자료에의하면 2015년기준국내당뇨병및당뇨전단계유병자수는 1천 2백 30만명에달하는것으로나타났으며, 이러한증가추세는노인인구증가와성인초기공복혈당장애의유병자수증가로인해계속될것으로생각된다. 그러나 30-40대의연령대에서당뇨병에대해매우낮은인지율과치료율을보이고있어, 초기성인기에제때치료를하지못한공복혈당장애가중년기에접어들어당뇨병및각종만성질환으로계속진행될가능성이있어이에대한적절한대책이필요하다고본다 [30]. 당뇨병과심혈관질환과의관련성을연구한기존연구의, 병태생리학적관점에서보면, 당뇨로인한만성적인고혈당, 과도한유리지방산의분비, 그리고인슐린저항성등의비정상적인대사상태는혈관내피세포내의대사이상을초래하게되어, 내피세포의기능저하, 혈관의과도한수축, 염증의증가및지속으로인해이상지질혈증을일으킨다. 이는결국혈관내혈전이생기기쉬운상태로진행하게되어심혈관 질환과는매우밀접한연관성을가질수밖에없게된다 [31]. 당뇨병유병률은남녀모두연령이증가할수록증가하는양상을띄는데, 최근에는전세계적으로식습관및생활습관과밀접한관련이있는만성질환발생의패턴변화로인해젊은연령에서의당뇨병발생률이증가하고있다 [15-17]. 이와관련하여, 본연구는우리나라국민을대표하는국민건강보험공단표본코호트자료를이용하여, 비교적젊은연령의대상자들에대해이루어진연구로서갖는의미가있다. 본연구의장점으로는먼저, 장기간의전국민전향적코호트연구또는중재의효과를검증할수있는대규모데이터베이스인공단표본코호트 DB를활용하였다는점이다. 우리나라의국민건강보험제도는세계적으로도인정받고있으며, 전국민이곧가입대상자가되기에국민건강보험가입자의진료내역, 검진결과, 거주지및보험료, 요양기관정보등을바탕으로층화계통추출방법을통해구축된국민건강보험공단표본코호트자료는전국민건강정보를대표하는자료라할수 408 http://www.e-jhis.org
Fasting Blood Glucose Levels Trajectory and Risk of Cardiovascular Disease in Korean Population 있다. 이러한자료를활용한연구라는측면에서볼때, 전국민건강정보의대표성을지닌자료를활용하여잠재계층분석을통한연구모형을적용한연구라는점을장점으로우선꼽을수있다. 또한, 데이터특유의개별적또는분리적인구조에따라자격 DB로부터추출된표본중 2002년건강검진수검자를기준으로하여 9년간의코호트를새롭게구축하여연구하였다는점이있다. 마지막으로, 변수의시간에따른추세와연구대상인구집단의크기를고려하여통계적으로적절히그룹화시켜주는집단중심추세모형방법론을적용하여, 9년간의공복혈당변화에따른심장병발생위험도를추정하는새로운시도를한것또한장점으로꼽을수있다. 집단중심추세모형방법론을이용하여, 장기간의공복혈당추적관리기간동안의특유의변화양상을파악하여질병발생이전에예방적대책을수립하고, 그에따라실제심장병발생양상또한어떠한차이를보이는지를파악하는단계로까지나아간다면, 질병예측측면에서도좋은방향을제시할수있는방법이라사료된다. 집단중심추세모형방법론을역학연구에적용한이전의사례들에서는주로일부연령대만을대상으로하거나한시점만을기준으로하여일정기간의궤적에따른그룹화를시행하고, 각그룹별질병관련지표와의관련성을분석하였다면 [32-34], 본연구에서는 20 대이상전연령의연구대상자에서코호트전체기간동안의공복혈당궤적에따른그룹별심장병발생위험도를추정하였다는특징이있다. 그러나연구에있어몇가지제한점또한있는데, 첫째, 코호트전체기간이모두잠재계층분석에이용되어추적기간동안중도탈락한자, 사망한자없이 9년간매년건강검진을받은생존자에한하여진행된연구라는가장큰취약점이있다. 둘째, 연구를위한코호트구축시처방전데이터까지수렴하는데에한계가있어약복용력을고려하지못하였다는연구의제한점또한있을수있다. 따라서이같이공복혈당값에영향을미칠가능성이있는당뇨병치료제복용여부를고려하지못한이유로당뇨병수준의공복혈당수치를지니고있던대상자들이시간이지남에따라지속적인건강관리의목적으로치료나약복용을하였을가능성이있는데, 그러한점이본연구의 High-ward 그룹의결과에서드러나꾸준히공복혈당수치가증가하였음에도불구하고심장병발생위험과의관련성이나타나지않았을가능성에대해조심스럽게유추해볼수있을것으로사료된다. 셋째, 심장병을기전별로세분화하였을때, 대상자및발생자수의부족으로인해세분화된위험도추정이곤란하였다. 이를위해서는심장병기전별위험도분석이반영되고다양한연령대를아우르는더욱세밀하고보다확장된연구를위해대규모코호트자료를통한공복혈당변화와심장병발생과의관련성연구가더필요할것으로보인다. 결론적으로, 국민건강보험공단표본코호트자료를활용한본연구에서, 공복혈당이 9년동안이상혈당증수준이었으며, 지속적으로약 간의증가를보인 20 세이상성인에서의심장병발생위험이높았다. REFERENCES 1. Hu FB, Stampfer MJ, Haffner SM, Solomon CG, Willett WC, Manson JE. Elevated risk of cardiovascular disease prior to clinical diagnosis of type 2 diabetes. Diabetes Care 2002;25(7):1129-1134. Doi: 10.2337/diacare.25.7.1129. 2. Elizabeth LM, Paul ZZ, Timothy AW, Damien J, Dianna JM, David WD, et al. Risk of cardiovascular and all-cause mortality in individuals with diabetes mellitus, impaired fasting glucose, and impaired glucose tolerance The Australian Diabetes, Obesity, and Lifestyle Study (Aus- Diab). Circulation 2007;116:151-157. Doi: 10.1161/Circulationaha. 106.685628. 3. Kim HK, Park SW, Kim CH, Yun YD, Kim EH, Baek SJ, et al. Impaired fasting glucose and risk of cardiovascular disease in Korean men and women. Diabetes Care 2013;36:328 335. Doi: 10.2337/ dc12-0587. 4. Park CS, Han EJ, Guallar E, Baek SJ, Linton JA, Yun YD, et al. Fasting glucose level and the risk of incident atherosclerotic cardiovascular diseases. Diabetes Care 2013;36:1988 1993. Doi: 10.2337/dc12-1577. 5. Shi Z, Zhen S, Zimmet PZ, Zhou Y, Zhou Y, Magliano DJ, et al. Association of impaired fasting glucose, diabetes and dietary patterns with mortality: a 10-year follow-up cohort in Eastern China. Acta Diabetologica 2016;53(5):799 806. Doi: 10.1007/s00592-016-0875-8. 6. Held C, Björkander I, Forslund L, Rehnqvist N, Hjemdahl P. The impact of diabetes or elevated fasting blood glucose on cardiovascular prognosis in patients with stable angina pectoris. Diabet Med 2005; 22:1326 1333. 7. Yang SW, Zhou YJ, Nie XM, Liu YY, Du J, Hu DY, et al. Effect of abnormal fasting plasma glucose level on all-cause mortality in older patients with acute myocardial infarction: results from the Beijing Elderly Acute Myocardial Infarction Study (BEAMIS). Mayo Clin Proc 2011;86(2):94-104. Doi: 10.4065/mcp.2010.0473. 8. Norhammar A, Tenerz A, Nilsson G, Hamsten A, Efendíc S, Rydén L, et al. Glucose metabolism in patients with acute myocardial infarction and no previous diagnosis of diabetes mellitus: a prospective study. Lancet 2002;359:2140 2144. 9. Wei M, Gibbons LW, Mitchell TL, Kampert JB, Stern MP, Blair SN. Low fasting plasma glucose level as a predictor of cardiovascular dis- http://www.e-jhis.org 409
Jooeun Jeon, et al. ease and all-cause mortality. Circulation 2000;101:2047-2052. 10. The Emerging Risk Factors Collaboration. Diabetes mellitus, fasting glucose, and risk of cause-specific death. N Engl J Med 2011;364:9. Doi: 10.1056/NEJMoa1008862. 11. Williams ED, Renwick C, Rawal L, Shaw JE, Oldenburg BF, Tapp RJ. Risk of cardiovascular and all-cause mortality: impact of impaired health-related functioning and diabetes. Diabetes Care 2012;35:1067 1073. Doi: 10.2337/dc11-1288. 12. World Health Organization. 2015 Global report on diabetes. 2016. 13. Centers for Disease Control and Prevention. Long-term trends in diabetes. GA, USA: Centers for Disease Control and Prevention; 2016. 14. Dabelea D, DeGroat J, Sorrelman C, Glass M, Percy CA, Avery C, et al. Diabetes in Navajo youth: prevalence, incidence, and clinical characteristics: the SEARCH for Diabetes in Youth Study. Diabetes Care 2009;Suppl 2:S141-7. Doi: 10.2337/dc09-S206. 15. Wilmot E, Idris I. Early onset type 2 diabetes: risk factors, clinical impact and management. Ther Adv Chronic Dis 2014;5(6):234 244. Doi: 10.1177/2040622314548679. 16. Menke A, Casagrande S, Geiss L, Cowie CC. Prevalence of and trends in diabetes among adults in the United States, 1988-2012. JAMA 2015;314(10):1021-1029. Doi: 10.1001/jama.2015.10029. 17. Korea Centers for Disease Control and Prevention. The Korea National Health and Nutrition Examination Survey (KNHANES): The Korea health statistics (2014). Cheongju: Korea Centers for Disease Control and Prevention; 2015 (Korean). 18. Chun KH. Evidence-based management and treatment of high risk individuals with pre-diabetes. J Korean Med Assoc 2011;54(10):1020-1027 (Korean). 19. Lee JY, Lee JS, Park SH, Soon Shin A, Kim KW. Cohort profile: The National Health Insurance Service National Sample Cohort (NHIS- NSC), South Korea. Int J Epidemiol 2015;1 8. Doi: 10.1093/ije/ dyv319. 20. National Health Insurance Service. National health screening statistically earbook. Wonju: National Health Insurance Service ; 2015 (Korean). 21. Park TS. Medical opinion for application criteria of Health insurance service for type 2 diabetes medication. Health Poilcy Forum 2011; 9(2):21-26 (Korean). 22. Park JK, Kim GS, Kim CB, Lee TY, Lee KS, Lee DH, et al. The accuracy of ICD codes for cerebrovascular diseases in medical insurance claims. Korean J Prev Med 2000;33(1):76-82 (Korean). 23. Jones BJ, Nagin DS, Roeder K. A SAS procedure based on mixture models for estimating developmental trajectories. Sociological Methods & Research 2001;29(3):374-393. 24. Wen CP, Cheng DYD, Tsai SP, Hsu HL, Wang SL. Increased mortality risks of pre-dabetes (impaired fasting glucose) in Taiwan. Diabetes Care 2005;28:2756 2761. 25. Grundy SM. Pre-diabetes, metabolic syndrome, and cardiovascular risk. J Am Coll Cardiol 2012;59(7):635-643. Doi: 10.1016/j.jacc.2011. 08.080. 26. Nagin DS, Odgers CL. Group-based trajectory modeling in clinical research. Annu Rev Clin Psychol 2010;6:109 138. 27. Ford ES, Zhao G, Li C. Pre-diabetes and the risk for cardiovascular disease: a systematic review of the evidence. J Am Coll Cardiol 2010; 55(13):1310 1317. 28. National Health Insurance Service. Korean diabetes fact sheet. Wonju: National Health Insurance Service; 2015 (Korean). 29. Korea Centers for Disease Control and Prevention. The Korea National Health and Nutrition Examination Survey (KNHANES): main report, Phase 6, 2nd (2014). Cheongju: Korea Centers for Disease Control and Prevention; 2015 (Korean). 30. Korea Centers for Disease Control and Prevention. The Korea National Health and Nutrition Examination Survey (KNHANES): the Korea health statistics (2015). Cheongju: Korea Centers for Disease Control and Prevention; 2016 (Korean). 31. Beckman JA, Creager MA, Libby P. Diabetes and atherosclerosis: epidemiology, pathophysiology, and management. JAMA 2002;287(19): 2570-2581. 32. Allen NB, Siddique J, Wilkins JT, Shay C, Lewis CE, Goff DC, et al. Blood pressure trajectories in early adulthood and subclinical atherosclerosis in middle age. JAMA 2014;311(5):490-497. 33. Theodore RF, Broadbent J, Nagin D, Ambler A, Hogan S, Ramrakha S, et al. Childhood to early-midlife systolic blood pressure trajectories. Hypertension 2015;66:1108-1115. 34. Munthali RJ, Kagura J, Lombard Z, Norris SA. Childhood adiposity trajectories are associated with late adolescent blood pressure: birth to twenty cohort. BMC Public Health 2016;16:665. 410 http://www.e-jhis.org