원저 Lab Med Online Vol. 3, No. 4: , October 진단면역학 아디포넥틴 ELISA 검사성능평가와한국인의혈청아디포넥틴참고구간설정 Evaluatio

Transcription

1 원저 Lab Med Online Vol. 3, No. 4: , October 2013 진단면역학 아디포넥틴 ELISA 검사성능평가와한국인의혈청아디포넥틴참고구간설정 Evaluation of the Performance of an Adiponectin ELISA-based Test and Establishing Serum Adiponectin Reference Intervals for Korean Population 박용정 김영란 김현숙 Yongjung Park, M.D., Young Ran Kim, M.T., Hyon-Suk Kim, M.D. 연세대학교의과대학세브란스병원진단검사의학과 Department of Laboratory Medicine, Severance Hospital, Yonsei University College of Medicine, Seoul, Korea Background: Adiponectin is a plasma protein secreted by adipose tissues and low serum adiponectin concentration has been reported to be associated with insulin resistance and metabolic syndrome (MS). We evaluated the performance of an ELISA-based assay for measuring serum adiponectin levels and established reference intervals of adiponectin for Korean population. Methods: Laboratory performance, including precision and linearity, of the AdipoMark Human Adiponectin ELISA kit (Mesdia Co., Korea) was assessed. Reference intervals of adiponectin concentration were determined after evaluation of 1200 subjects with no history of MS. Adiponectin was also measured in 100 patients with MS. Results: The mean concentrations of serum samples tested for precision evaluation were 6.66, 12.61, and μg/ml: the ELISA showed total imprecision of 13.6%, 9.3%, and 10.5% CV for the respective concentrations. The assay demonstrated linear responses in the range of μg/ml serum adiponectin levels. The 95% reference intervals for Korean population were μg/ml for men and μg/ml for women. ROC-area under the curve values of adiponectin for the diagnosis of MS were 0.85 for men and 0.83 for women. Low adiponectin level was independently associated with MS in the multivariate analysis. Conclusions: The adiponectin quantitation assay evaluated in this study showed acceptable laboratory and clinical performances in an ELISA platform. To meet the ever-increasing demand for a reliable assay for measuring adiponectin levels in the study of various metabolic diseases, this assay could be further improved by the automation of the platform. Key Words: Adiponectin, Laboratory performance, Precision, Reference interval 서론 Adiponectin 은지방세포에특이적으로분포하는 mrna 인 apm1 의 translation 에의해합성되는단백질로 [1], 정상적으로인간 Corresponding author: Hyon-Suk Kim Department of Laboratory Medicine, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemun-gu, Seoul , Korea Tel: ; Fax: kimhs54@yuhs.ac Received: September 11, 2012 Revision received: December 14, 2012 Accepted: December 20, 2012 This article is available from , Laboratory Medicine Online This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ( which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. 의혈장에분포하는용해성기질단백이다 [2]. 1990년대말에사람의혈장에서 adiponectin이발견된후, 한연구진은사람 adiponectin에대한단클론및다클론항체를제조하여 enzyme-linked immunosorbent assay (ELISA) 를개발하였고, 혈장 adiponectin 농도를측정한결과 adiponectin 농도가체질량지수 (body mass index) 와음의상관관계를보임을보고하였다 [2]. 이후여러연구들에서 adiponectin이당뇨병이나심혈관질환등대사질환의발생에있어중요한역할을하는것으로보고하였다. 한연구에서는 adiponectin이 tumor necrosis factor-alpha (TNF-α) 에의한혈관내피세포와단핵구의결합을막고혈관내피세포의 vascular cell adhesion molecule-1 (VCAM-1), endothelialleukocyte adhesion molecule-1 (E-selectin) 및 intracellular adhesion molecule (ICAM-1) 의발현을억제하며, 관상동맥질환환자에서혈장 adiponectin 농도가정상인에비하여유의하게낮다고보고하였고 [3], 혈장 adiponectin 농도측정이심혈관질환의위험도 eissn

2 평가에도움이될것이라고하였다 [3]. 다른연구에서는내피세포의경계가손상되는경우정상적으로혈장에존재하는 adiponectin이혈관벽에축적된다고보고하여 [4], adiponectin의농도감소가심혈관질환의발생과관련되어있음이알려졌다. Adiponectin은대식세포가포말세포 (foam cell) 로변형되는것을막아체내에지방이축적되는것을억제하는것으로알려져있다 [5]. 여러연구들에서체질량지수, 혈청중성지방농도, 당뇨병이나심혈관질환의유무가혈장 adiponectin과유의한상관관계가있고, 당뇨환자및정상인에서체중을감량하는경우혈장 adiponectin 농도가유의하게증가함을보고하였다 [6, 7]. 또한, 인종에관계없이 2형당뇨및비만이혈장 adiponectin 농도감소와관련되어있음이알려졌다 [8]. rhesus 원숭이를이용한동물실험에서도인슐린저항성이증가할수록혈장 adiponectin 농도가감소하였다 [9]. 근래의한연구에서 adiponectin 농도와체질량지수의연관성정도가백인, 흑인, 동양인등인종에따라차이가있으나, 인슐린저항성의지표중하나인 homeostasis model assessment of insulin resistance (HOMA-IR) 과 adiponectin 농도사이의상관관계는인종에따른차이가없어, adiponectin과같은 adipokine의정상적인인종간차이가인종에따른평균적인인슐린저항성차이를설명할수있을것으로보고하였다 [10]. 따라서인종에관계없이대사증후군을평가하기위한지표로서 adiponectin의유용성이보고된바있다. 앞서언급한연구들에서 adiponectin이비만, 인슐린저항성, 당뇨, 심혈관질환등의발생과연관되어있음이보고되면서, 이들질환을포괄하는대사증후군의발생에있어 adiponectin의중요한역할이알려지고, 관련연구들도활발하게이루어졌다. 나아가 adiponectin이악성종양, 자가면역질환, 간에서의탄수화물과지질대사, 염증과패혈증등여러가지질병의발생과관련되어있다고보고되어왔다 [11-13]. 하지만, 현재까지상용화된검사시약이많이보급되지않아임상검사실에서정규검사로 adiponectin 정량측정을많이시행하지않고있는실정이다. 본연구에서는국내에서개발된 adiponectin 측정용 ELISA의기본적인검사성능을평가하고대사증후군이없는한국인의혈청 adiponectin 농도를측정하여한국인의 adiponectin 농도참고구간을제시하고자하였다. 또한, 대사증후군에이환된대상자에서혈청 adiponectin을측정하여대사증후군진단을위한 adiponectin의임상적유용성을평가하였다. 대상및방법 1. Adiponectin 측정 혈청 adiponectin 농도측정을위하여 AdipoMark Human Adiponectin ELISA kit (Mesdia Co., Seoul, Korea) 를사용하였다. 이시약은샌드위치형효소면역측정법원리를활용하여혈청또는혈장 adiponectin 농도를측정한다. Adiponectin 특이마우스단클론항체가부착된 96-well microplate에 5가지농도의표준물질, blank, 양성및음성대조물질과함께 100배로희석한혈청검체를첨가하여 2시간동안반응시키고세척후, biotin이부착된 2차항체를첨가후다시 2시간배양하였다. 세척후, streptavidin과함께 horse-radish peroxidase가부착된 conjugate 용액을첨가하여 30 분간반응시키고세척한후, 발색시약을첨가하고 20분후 450 nm 에서흡광도를측정하였다. 다섯가지농도의표준물질및 blank 물질에서측정한흡광도를이용하여표준곡선을도출한뒤각검체에서측정한흡광도를 adiponectin 농도 (μg/ml) 로환산하였다. 2. 정밀도평가 CLSI document EP4-A2 guideline을참고하여정밀도를평가하였다. 3가지 adiponectin 농도의혈청검체를각각 1일 1회검사당 8번반복하여 10일간측정하고, 각농도당 80회측정한후총비정밀도 (total imprecision), 검사내비정밀도 (intra-assay imprecision, repeatability) 및검사간비정밀도 (inter-assay imprecision) 를계산하였다. 3. 직선성평가 CLSI document EP6-A에따라직선성을평가하였다. Adiponectin 고농도및저농도혈청검체를 1:0, 3:1, 1:1, 1:3, 0:1의비율로각각혼합하여 5가지농도의혈청을제조한후, 각검체를 4회반복측정하고예상농도와측정된농도를선형회귀분석을통해비교하여직선성이유지되는구간을평가하였다. 직선성의정도는회귀식의결정계수 (R 2 ) 로제시하였다. 4. 참고구간설정혈청 adiponectin의참고구간설정및건강한정상인과대사증후군환자와의비교를위하여세브란스병원연구심의위원회 (Institutional Review Board) 의승인을받은후, 건강한한국인과대사증후군에이환된한국인을대상으로하였다. 2006년 9월부터 2011년 12월까지세브란스건강검진센터에내원하여건강검진을받은건강한한국인으로부터건강검진완료후잔여혈청검체를수집하였다. 수집된검체는개인신상정보를폐기하고바코드로검체의식별정보를부여한후 -70 이하로냉동보관하였으며, 이중대사증후군이없는대상자로부터수집한혈청을 20, 30, 40, 50, 60대남, 여각각 120개씩총 1,200개를선정하여 adiponectin 농도를측정하였다. 검사시약의생산 lot 번호에따른검사결과의차이를방지하기위하여전체측정과정동안동 243

3 일 lot 번호의시약만을사용하였으며, ELISA 진행과정중검사 batch에따른차이를최소화하기위하여각 batch마다성별및연령대별대상검체를무작위로배정하여각연령대및성별에따른대상군검체가하나의검사 batch에편중되지않도록하였다. 개정된 National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP-III) 의기준을다음과같이한국인에맞게변형후적용하여대사증후군이있는대상자는배제하였다 : 복부비만 : 남 90 cm, 여 85 cm; 중성지방 150 mg/dl; 고밀도콜레스테롤남자 <40 mg/dl, 여자 <50 mg/dl; 혈압 130/85 mmhg 또는항고혈압약물복용 ; 공복혈당 100 mg/dl 또는당뇨병약물복용의 5가지항목중 3가지이상에해당하는경우대사증후군으로진단하고참고구간설정을위한검체에서제외하였다. 검체를대상자의연령구간및성별에따라나누어각대상군의 adiponectin 농도분포를 Kolmogorov Smirnov 시험을이용하여분석한후, adiponectin 농도가정규분포에따르지않음을확인하고, 2.5 및 97.5 백분위수를산출하여 95% 참고구간을설정하였다. 건강검진수검시측정한혈압, 허리둘레, 체질량지수, 공복혈당, 중성지방및고밀도콜레스테롤농도를함께분석하였다. 5. 대사증후군에서의 adiponectin 농도참고구간설정을위한검체수집과동일하게같은기간동안세브란스건강검진센터에내원하여건강검진을받은대상자중상기진단기준에따라대사증후군으로진단된남녀각각 50명으로부터혈청을수집하여동일한방법으로 adiponectin 농도를측정하였다. 건강검진결과중혈압, 허리둘레, 체질량지수, 공복혈당, 중성지방및고밀도콜레스테롤농도를함께수집하였다. 6. 통계분석통계분석에는 Analyse-it Method Evaluation Edition version 2.27 software (Analyse-it Software Ltd., City West Business Park, Leeds, UK) 와 PASW statistics 18.0 (SPSS Inc., Chicago, IL, USA) 을사용하였다. 대상군간연속변수의차이를비교하기위하여두군간의비교에는 Mann Whitney U 시험을, 세군이상간의비교에는 Kruskal Wallis 시험을사용하였다. Kruskal Wallis 시험에는다중비교시발생하는제1종통계오류를보정하기위하여 Bonferroni 교정법을적용하였다. 참고구간설정을위한대상군의 adiponectin 농도분포는 Kolmogorov Smirnov 시험을이용하여분석하였다. 변수들간의상관성을분석하기위하여 Spearman s rank 시험을이용하여상관계수 (r) 를계산하였다. 대사증후군감별을위한혈청 adiponectin의진단성능을평가하기위하여상대수행능곡선 (receiver operating characteristics curve, ROC) 분석으로곡선하면적 (area under the curve, AUC) 을구하였다. 다변량분석을위하여대사증후군의유무를종속변수로, 혈압, 허리둘레, 체질량지수, 공복혈당, 중성지방, 고밀도콜레스테롤및 adiponectin 농도를독립변수로하여이분형로지스틱회귀분석을시행하였으며, 유의한독립변수의선별을위하여각변수의 likelihood ratio 순서에따른전방단계별분석을적용하였다. 모든통계분석결과는 P-value 0.05 미만인경우유의한것으로간주하였다. 결과 1. AdipoMark ELISA 검사의정밀도 3종류검체 ( 저농도, 중간농도, 고농도 ) 의평균 adiponectin 농도는각각 6.66, 및 μg/ml 이었고검사내비정밀도는변이계수 (coefficient of variation, CV) 가각각 9.6%, 5.5% 및 6.8% 이었다 (Table 1). 해당농도에대한 AdipoMark ELISA 검사의검사간비정밀도는각각변이계수 9.5%, 7.5% 및 8.1% 였으며, 총비정밀도는각각변이계수 13.6%, 9.3%, 10.5% 이었다 (Table 1). 2. AdipoMark ELISA 검사의직선성 5가지 adiponectin 농도의혈청을 4회반복측정한결과를예상농도와비교하여선형회귀분석을하였다. Adiponectin 농도가 μg/ml인구간에서직선성이유지되었으며, y를측정된농도로, x를예상농도로하여분석한선형회귀식은 y = 0.972x 였고이때결정계수 R 2 은 이었다 (Fig. 1). 3. 혈청 adiponectin 참고구간참고구간설정을위한대상군의특성을 Table 2에요약하였다. 체질량지수, 허리둘레, 혈압, 공복혈당, 중성지방및고밀도콜레스테롤농도, adiponectin 농도등모든항목이남녀간에차이가있 Table 1. Precision performance of the AdipoMark ELISA for measuring serum adiponectin levels during a 10-day evaluation period Level N Mean (μg/ml) Intra-assay imprecision Inter-assay imprecision Total imprecision SD (μg/ml) CV (%) SD (μg/ml) CV (%) SD (μg/ml) CV (%) Low Middle High Abbreviations: ELISA, enzyme-linked immunosorbent assay; SD, standard deviation; CV, coefficient of variation

4 Measured concentration (μg/ml) Linearity Plot Expected concentration (μg/ml) Fig. 1. Determination of the linear relationship between the serum adiponectin levels by using AdipoMark Human Adiponectin ELISA kit. Dashed line indicates polynomial fit. 었으며, 남자에서는허리둘레및공복혈당, 여자에서는모든항목 이연령대별대상군에따라차이가있었다. 연령및성별에따른 adiponectin 농도의분포및차이는 Fig. 2 에나타내었다. 혈청 adiponectin 농도는여자에서더높았으며, 남 자의경우연령에따라유의한차이가없었으나, 여자에서는다른 연령에비하여 40 대에서 adiponectin 농도가낮았다. 연령및성별에따른 adiponectin 의참고구간은 Table 3 에정리 하였다. 정규분포와의차이를의미하는 Kolmogorov Smirnov D 값이 20 대여성을제외한모든연령대및성별별대상군에서통계 적으로유의하여정상인의혈청 adiponectin 농도분포가정규분 포에따르지않음을확인하였고, 왜도 (skewness) 및첨도 (kurtosis) 가양의값으로정상인의 adiponectin 농도는정규분포에비하 여낮은농도에편중되어분포하였으므로, 이에따라 95% 참고구 간을 백분위수로계산하였다. 남자는전체연령대대상군 에서 95% 참고구간이 μg/ml 였으며, 여자의경우전체연 령대에서 μg/ml 로남자에비하여높았다. 한편, 참고구간 설정을위한 14 회의 ELISA 검사간비정밀도는평균 adiponectin 농도 1.94 μg/ml 인음성대조물질에대하여변이계수 12.4%, 표준 편차 0.24 μg/ml 였고, 평균농도 μg/ml 인양성대조물질에 대하여변이계수 4.0%, 표준편차 6.16 μg/ml 이었다. 4. Adiponectin 농도와다른변수들간의상관성 Linear fit ( x) R 2 = 전체 1,200 명의건강한한국인에서혈청 adiponectin 농도는체 질량지수 (r = ) 및허리둘레 (r = ) 와중등도의음의상 관관계를나타내었다 (Table 4). 또한, 혈압, 공복혈당및중성지방 농도와도약한음의상관관계를보였다. 반면, adiponectin 농도와고밀도콜레스테롤수치사이에는중등도의양의상관관계가관찰되었다 (r = ). 성별및연령별로 adiponectin 농도와다른지표간의상관성을분석한경우에도, 대부분의대상군에서 adiponectin 농도가체질량지수및허리둘레와유의한음의상관관계를보였으나, 수축기및이완기혈압과공복혈당은 adiponectin 농도와유의한상관관계가없었다. 또한, 중성지방과고밀도콜레스테롤수치는남자의경우 20-40대대상군에서 adiponectin 농도와유의한상관관계를보인반면, 여자에서는 40대이상에서상관관계를보였다. 5. 대사증후군진단을위한 adiponectin의임상적유용성남녀각각 50명의대사증후군환자군의나이는남자군중위수 43.5세 (1-3사분위수 세 ) 및여자군중위수 54.5세 (1-3사분위수 세 ) 로남자환자군의경우정상인남자군의나이 (N = 600, 중위수 43.5세, 1-3사분위수 세 ) 와차이가없었으나 (P = ), 여자환자군은정상인여자군 (N = 600, 중위수 45.0세, 1-3사분위수 세 ) 에비하여고령이었다 (P < ). 혈청 adiponectin은남녀모두대사증후군환자군 ( 남, N= 50, 중위수 4.56 μg/ml, 1-3사분위수 μg/ml; 여, N=50, 중위수 6.29 μg/ml, 1-3사분위수 μg/ml; 전체, N=100, 중위수 5.31 μg/ml, 1-3사분위수 μg/ml) 에서정상인 ( 남, N= 600, 중위수 8.21 μg/ml, 1-3사분위수 μg/ml; 여, N= 600, 중위수 μg/ml, 1-3사분위수 μg/ml; 전체, N=1200, 중위수 μg/ml, 1-3사분위수 μg/ml) 에비하여유의하게낮았다 (P <0.0001). 남녀각각 50명의대사증후군환자와남녀각각 600명의정상인을비교분석한결과, 대사증후군진단을위한 adiponectin의 ROC- AUC는남자의경우 0.85 (95% 신뢰구간 , P <0.0001), 여자의경우 0.83 (95% 신뢰구간 , P <0.0001) 이었으며, 두 AUC 값간에는유의한차이가없었다 (P = ). 남녀전체를대상으로분석하는경우의 AUC는 0.82 (95% 신뢰구간 , P <0.0001) 였다. Youden index에따라대사증후군진단에대한민감도와특이도의합이최대인절단값 adiponectin 농도를산출한결과, 남자의경우최적의절단값이 6.59 μg/ml, 여자는 9.27 μ g/ml였으며, 이때의민감도와특이도는각각남자 86.0% (95% 신뢰구간 73.3%-94.2%) 및 70.5% (95% 신뢰구간 66.7%-74.1%), 여자 78.0% (95% 신뢰구간 64.0%-88.5%) 및 75.0% (95% 신뢰구간 71.3%-78.4%) 였다. 추가로, 정상인남자의 adiponectin 2.5백분위수참고치인 3.6 μg/ml를남자대상군에대한절단값으로적용하는경우, 대사증후군진단에대한민감도는 28.0% (95% 신뢰구간 16.2%-42.5%) 로낮았으며, 이때특이도는참고구간산출에사용한 245

5 Table 2. Characteristics of Korean subjects (N=1,200) for the establishment of reference intervals Characteristics Age group (yr)* Male (N=600) Female (N=600) P-value All (N=1,200) BMI (kg/m 2 ) 20 to (21.3 to 25.7) 20.0 (18.7 to 21.4) < (19.6 to 24.2) 30 to (22.3 to 26.1) 20.8 (19.3 to 22.2) < (20.4 to 24.5) 40 to (22.2 to 25.2) 21.8 (20.3 to 24.1) < (21.0 to 24.9) 50 to (22.6 to 25.9) 22.8 (20.4 to 24.3) < (21.6 to 25.5) 60 to (22.5 to 25.3) 23.2 (22.1 to 24.9) (22.2 to 25.1) All 23.9 (22.2 to 25.6) 21.6 (19.9 to 23.9) < (20.8 to 24.9) P-value < < WC (cm) 20 to (75.0 to 88.6) 69.0 (66.0 to 74.0) < (69.0 to 83.6) 30 to (78.0 to 87.6) 70.0 (67.0 to 76.0) < (70.0 to 85.0) 40 to (81.0 to 88.0) 74.0 (70.0 to 78.0) < (73.0 to 85.0) 50 to (82.0 to 90.0) 77.0 (72.0 to 83.0) < (75.4 to 87.0) 60 to (82.0 to 90.0) 80.0 (75.0 to 84.0) < (78.0 to 88.0) All 84.0 (80.0 to 89.0) 74.0 (69.0 to 80.0) < (73.0 to 86.0) P-value < < SBP (mmhg) 20 to (115.0 to 124.0) (102.4 to 112.0) < (106.4 to 120.0) 30 to (113.0 to 126.0) (102.0 to 111.6) < (105.0 to 121.6) 40 to (111.0 to 123.6) (103.0 to 117.6) < (106.0 to 121.6) 50 to (112.0 to 124.6) (107.4 to 123.0) (110.0 to 124.0) 60 to (113.4 to 131.0) (113.0 to 130.0) (113.0 to 130.6) All (113.0 to 126.0) (104.0 to 120.0) < (108.0 to 123.0) P-value < < DBP (mmhg) 20 to (73.4 to 83.6) 73.0 (67.0 to 77.6) < (70.0 to 80.0) 30 to (73.4 to 85.0) 69.0 (65.0 to 75.6) < (67.4 to 81.0) 40 to (71.4 to 85.6) 74.0 (68.0 to 80.0) < (70.0 to 82.0) 50 to (74.4 to 85.6) 76.0 (69.0 to 82.0) (70.0 to 84.0) 60 to (73.0 to 87.6) 77.0 (72.0 to 84.0) (72.4 to 85.0) All 79.0 (73.0 to 85.0) 74.0 (68.0 to 80.0) < (70.0 to 83.0) P-value < < FBG (mg/dl) 20 to (80.4 to 88.0) 82.0 (78.0 to 86.0) (79.0 to 87.0) 30 to (81.0 to 92.6) 83.0 (78.0 to 90.0) (80.0 to 91.0) 40 to (83.4 to 97.0) 85.0 (80.0 to 90.0) < (82.0 to 95.0) 50 to (85.4 to 98.0) 87.0 (82.0 to 92.6) < (83.0 to 95.0) 60 to (86.4 to 100.6) 90.0 (84.0 to 94.0) (85.0 to 98.0) All 88.5 (83.0 to 96.0) 85.0 (80.0 to 91.0) < (82.0 to 94.0) P-value < < < TG (mg/dl) 20 to (74.0 to 128.6) 72.0 (55.0 to 88.2) < (63.4 to 113.0) 30 to (79.8 to 145.0) 69.0 (58.0 to 92.6) < (63.0 to 124.0) 40 to (81.4 to 144.6) 79.0 (57.4 to 101.6) < (71.0 to 125.6) 50 to (83.0 to 143.6) 95.0 (68.0 to 128.0) (74.0 to 136.0) 60 to (73.0 to 138.8) (75.4 to 133.6) (75.0 to 135.0) All (78.0 to 141.0) 80.0 (60.4 to 109.6) < (68.0 to 127.0) P-value < < HDL (mg/dl) 20 to (46.0 to 59.6) 63.5 (57.0 to 75.6) < (49.0 to 66.0) 30 to (44.0 to 55.0) 61.0 (54.4 to 71.0) < (47.0 to 64.0) 40 to (44.0 to 58.6) 58.0 (49.4 to 68.0) < (46.0 to 64.0) 50 to (43.0 to 57.6) 58.0 (52.0 to 66.0) < (47.0 to 64.0) 60 to (43.0 to 56.6) 55.0 (49.0 to 63.6) (46.0 to 59.0) All 50.0 (44.0 to 57.6) 59.0 (52.0 to 69.0) < (47.0 to 64.0) P-value < < Adiponectin (μg/ml) 20 to (6.0 to 11.3) 14.2 (9.1 to 18.4) < (7.4 to 14.8) 30 to (6.5 to 11.1) 16.7 (11.9 to 22.5) < (7.6 to 17.1) 40 to (6.6 to 13.1) 9.6 (7.4 to 14.8) (6.9 to 13.9) 50 to (5.9 to 9.7) 13.7 (10.1 to 19.3) < (6.9 to 14.6) 60 to (6.4 to 10.6) 13.3 (10.0 to 18.1) < (7.5 to 15.3) All 8.2 (6.2 to 11.0) 13.4 (9.3 to 18.9) < (7.3 to 15.0) P-value < *N=120 for the respective age and sex groups; Men vs. women (P-values were calculated by Mann Whitney U test.); P-value for the difference among the age groups was calculated by Kruskal-Wallis test with Bonferroni correction to compensate alpha-statistical error from multiple comparisons. Abbreviations: BMI, body mass index; WC, waist circumference; SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose; TG, triglyceride; HDL, high-density lipoprotein cholesterol

6 P < P = P < P < P < P = P = P >0.05 Adiponectin level (μg/ml) Age (yr) Sex 20 to to to to to to to to to to 69 Male Female Fig. 2. Distribution of serum adiponectin levels according to the age and sex of the subjects without metabolic syndrome. Serum adiponectin levels were higher in women than in men. Age-related variation in adiponectin levels was only observed in female subjects. P-values were calculated by Mann Whitney U test or Kruskal-Wallis test. The upper and lower ends of the boxes and box inner lines correspond to the upper and lower quartiles and median values, respectively. Whiskers denote the minimum and maximum values, and the circles indicate individual values. Table 3. Reference intervals of serum adiponectin levels in the groups classified according to age and sex Sex Age (yr) N 95% reference interval (μg/ml) Distribution of adiponectin levels 2.5 percentile (CI)* 97.5 percentile (CI)* Skewness Kurtosis Kolmogorov-Smirnov D P-value Male 20 to (2.5 to 4.4) 17.9 (15.8 to 21.1) to (2.5 to 4.5) 15.8 (13.9 to 22.8) to (0.0 to 4.3) 20.3 (17.8 to 21.0) to (1.2 to 4.2) 19.5 (13.6 to 21.0) to (1.4 to 3.7) 23.4 (18.5 to 49.0) < All (2.5 to 4.0) 19.2 (17.8 to 21.0) < Female 20 to (1.3 to 6.2) 30.9 (25.8 to 39.4) > to (0.4 to 7.1) 52.5 (35.4 to 60.6) to (3.4 to 4.9) 25.8 (23.4 to 26.8) < to (3.8 to 5.8) 36.4 (30.2 to 54.5) to (2.9 to 6.6) 34.2 (29.5 to 35.5) < All (3.8 to 4.9) 34.2 (30.2 to 39.4) < All 1, (3.5 to 4.3) 29.2 (26.3 to 32.3) < * 90% CI for each age group and 95% CI for the total age groups for men and women. Abbreviation: CI, confidence interval

7 Table 4. Correlation between serum adiponectin levels (μg/ml) and other variables in normal Korean subjects (N=1,200) with no history of metabolic syndrome* Variable Sex Age N r 95% CI P-value BMI (kg/m 2 ) Male 20 to ( to ) to ( to ) to ( to ) < to ( to ) to ( to ) All ( to ) < Female 20 to ( to ) to ( to ) to ( to ) < to ( to ) < to ( to ) All ( to ) < All All 1, ( to ) < WC (cm) Male 20 to ( to ) < to ( to ) to ( to ) < to ( to ) to ( to ) All ( to ) < Female 20 to ( to ) to ( to ) to ( to ) < to ( to ) < to ( to ) All ( to ) < All All 1, ( to ) < SBP (mmhg) Male 20 to ( to ) to ( to ) to ( to ) to ( to ) to ( to ) All ( to ) Female 20 to ( to ) to ( to ) to ( to ) to ( to ) to ( to ) All ( to ) All All 1, ( to ) < DBP (mmhg) Male 20 to ( to ) to ( to ) to ( to ) to ( to ) to ( to ) Female 20 to ( to ) to ( to ) to ( to ) to ( to ) to ( to ) (Continued to the next page) 248

8 Table 4. (Continued from the previous page) Correlation between serum adiponectin levels (μg/ml) and other variables in normal Korean subjects (N=1,200) with no history of metabolic syndrome* Variable Sex Age N r 95% CI P-value All ( to ) All All 1, ( to ) < FBG (mg/dl) Male 20 to ( to ) to ( to ) to ( to ) to ( to ) to ( to ) All ( to ) Female 20 to ( to ) to ( to ) to ( to ) to ( to ) < to ( to ) All ( to ) All All 1, ( to ) < TG (mg/dl) Male 20 to ( to ) to ( to ) < to ( to ) to ( to ) to ( to ) All ( to ) < Female 20 to ( to ) to ( to ) to ( to ) < to ( to ) to ( to ) All ( to ) < All All 1, ( to ) < HDL (mg/dl) Male 20 to ( to ) < to ( to ) < to ( to ) to ( to ) to ( to ) All ( to ) < Female 20 to ( to ) to ( to ) to ( to ) < to ( to ) to ( to ) All ( to ) < All All 1, ( to ) < * Bold letters imply statistically significant results. Abbreviations: CI, confidence interval; BMI, body mass index; WC, waist circumference; SBP, systolic blood pressure; DBP, diastolic blood pressure; FBG, fasting blood glucose; TG, triglyceride; HDL, high-density lipoprotein cholesterol. 정상군을 ROC 분석의대조군으로사용하였으므로 97.5% (95% 신뢰구간 95.9%-98.6%) 였다. 마찬가지로, 정상인여자의 2.5백분위수참고치인 4.5 μg/ml를여자대상군에대한절단값으로적용하는경우, 대사증후군진단에대한민감도는 30.0% (95% 신뢰구간 17.9%-44.6%) 로낮았고, 특이도는 97.5% (95% 신뢰구간 95.9%- 98.6%) 로남자와동일하였다. 나이등의변수가대사증후군환자군과정상군사이에차이가있었으므로, 다변량분석을시행하여유의한독립변수를선별하 249

9 Table 5. Results of the multivariate analysis using binary logistic regression with metabolic syndrome as a dependent variable and other factors as predictors in patients with metabolic syndrome (N=100) and healthy subjects (N=1,200) Factors* Odds ratio 95% CI P-value Female to < BMI (Kg/m2) to < SBP (mmhg) to < FBG (mg/dl) to TG (mg/dl) to < HDL (mg/dl) to < Adiponectin (μg/ml) to < *Age, diastolic blood pressure, and waist circumference did not show a statistically significant association with the presence of metabolic syndrome in the multivariate analysis. Abbreviations: CI, confidence interval; BMI, body mass index; SBP, systolic blood pressure; FBG, fasting blood glucose; TG, triglyceride; HDL, high-density lipoprotein cholesterol. 였으며, 그결과를 Table 5 에요약하였다. 본연구의대상군에서는 성별, 체질량지수, 수축기혈압, 공복혈당, 중성지방및고밀도콜레 스테롤농도가대사증후군과관련된유의한독립변수였으며, 혈청 adiponectin 농도의경우대사증후군에대한 odds ratio 가 으로, 낮은 adiponectin 농도와대사증후군사이에독립적인관련 성이있었다 (P <0.0001). 고찰 Adiponectin 이 1990 년대말처음발견된이후, 당뇨, 인슐린저 항성, 비만등과관련된복합적대사이상을평가하기위한지표로 서의유용성이주로연구되어왔으나, 근래에는암, 감염질환및자 가면역질환등다양한질병과의연관성이밝혀지고있다. 이러한 임상적중요성에도불구하고임상검사실에서정규검사로서 adiponectin 정량검사를시행하기위한선결과제로 adiponectin 검사 의검사성능을평가하는연구가많이시행되지못하였고, 임상검 사실에서사용하기위한 adiponectin 측정정량검사시약의개발 도활발하게이루어지지않았다. 이연구에서는국내에서개발된 adiponectin 정량검사용 ELISA 의검사성능을평가하고자하였으며, 향후연구를위한기초자료 로사용할수있도록대사증후군이없는한국인 1,200 명에서혈청 adiponectin 농도를측정하여한국인의 adiponectin 참고구간을 설정하고자하였다. 이연구에서평가한 AdipoMark ELISA 검사는 adiponectin 농도 가 6.66 에서 μg/ml 범위의혈청검체에대하여변이계수 9.3% 에서 13.6% 사이의총비정밀도를보였고, 검사내비정밀도는 변이계수 5.5% 에서 9.6% 사이였으며, 검사간비정밀도는변이계수 7.5% 에서 9.5% 범위였다 (Table 1). Adipokine 동시정량검사의검사성능을평가한이전의한연구에서 multiplex bead-based immunoassay 및방사면역측정법 (radioimmunoassay) 원리를사용한 adiponectin 측정검사의검사내비정밀도는각각검사에대하여변이계수 3.8% 및 3.9%, 검사간비정밀도는각각변이계수 13.0% 및 8.5% 였고 [14], 이와유사한다른연구에서도 multiplex beadbased immunoassay와 ELISA의검사내비정밀도를각각변이계수 5.0% 와 6.9% 로, 검사간비정밀도를대상농도에따라 multiplex bead-based immunoassay는변이계수 8.9% 에서 12.0%, ELISA는 4.1% 에서 8.2% 로보고한바있어 [15], 본연구의결과와유사하였으나, 두연구모두총비정밀도를보고하지않아본연구의결과와정확한비교는어려웠다. 상용화된 adiponectin 측정검사가많이보급되지않았고, adiponectin 측정검사의검사성능에대한보고도드물어여러종류의 adiponectin 측정검사의보급을위하여검사성능에대한평가가선행되어야할것으로사료된다. 한편, 최근의한연구에따르면소아의경우출생이후부터 14세경까지혈장 adiponectin 농도가지속적으로감소하므로나이에따른참고구간을적용해야한다고하였으나 [16], 본연구에서는대사증후군에이환되지않은성인만을대상으로혈청 adiponectin 농도를측정한결과, 남자의경우연령별대상군간에 adiponectin 농도의차이가없었고, 여자의경우남자보다높은 adiponectin 농도를보이고연령별대상군간에도차이를보였으며, 40대여성의경우에서가장낮은 adiponectin 농도를보였다. 이전의한연구에서개인내 adiponectin의일중변화가 cortisol의변화에영향을받음을보고한바있으나 [17], 다른연구에서는비만인대상군에비하여정상인에서 30일간개인내 adiponectin 농도변화는매우작음을보고하였고 [18], 다른연구에서도체질량지수의변화를보정한경우 1년간개인내변화가극히적어역학조사를위해 adiponectin을측정하는경우개인당한번만측정해도충분함을제시한바있다 [19]. 본연구에서는특히 40대여성의경우 adiponectin 농도가다른연령대의여성대상군에비하여낮았으나대상군중 40대여성의체질량지수나허리둘레가오히려 50대및 60대여성군에비해서작았으므로, 체질량지수나허리둘레등본연구에서고려한변수이외의요소가여성에서 adiponectin 농도의연령대별차이에영향을준것으로생각된다. 최근의한연구에서중년여성의경우낮은성호르몬결합글로블린 (sex hormone binding globulin) 및높은테스토스테론 (testosterone) 농도가낮은 adiponectin 농도와관련이있음을보고하였는데 [20], 본연구의결과에서 40대이상여성의 adiponectin 농도가감소하는이유는아마도중년이후여성에서발생하는호르몬불균형및남성화때문일것으로추정된다. 따라서 adiponectin의연령에따른생리적변동에대한추가적연구가필요할것으로생각된다. 본연구에서남자의 250

10 경우연령별로 adiponectin 농도에차이가없었으나, 여자의경우연령에따라차이가있었고, 남녀간에도 adiponectin 농도에차이가있었으므로, 여자의경우연령대에따라각각의참고구간을적용하는것이바람직할것으로사료되고, 남자의경우에는연령별로참고구간을구분할필요는없을것으로생각된다. 이전의여러연구들에서보고한바와같이 [2, 6, 8, 10, 21-23], 본연구에서도대부분의연령및성별대상군에서혈청 adiponectin 농도가체질량지수및허리둘레와음의상관관계를보였고, 남자의경우 50세미만, 여자의경우 40세이상에서 adiponectin 농도가고밀도콜레스테롤수치와양의상관관계를보여낮은 adiponectin 농도가당뇨, 심혈관질환및대사증후군의위험도와관련이있을것으로생각되었다. 남녀및연령대별로 adiponectin 농도와다른대사지표간의상관성정도가달랐고, 이는연령및성별에따른지방대사의차이에따른것으로생각된다. 이처럼건강인사이에도연령및성별에따라여러대사지표들에차이가있으므로본연구에서는검체수집시조사된변수들을포함하여다변량분석을시행하였으며, 그결과낮은 adiponectin 농도가대사증후군과독립적으로연관되어있었으므로, adiponectin 농도가대사이상을반영하는중요한지표임을확인하였다. 남녀각각 50명씩 100명의대사증후군환자에서혈청 adiponectin을측정한결과, 남녀모두정상인에비하여대사증후군환자가낮은 adiponectin 농도를보였다. 또한, ROC 분석에서남녀모두 AUC 값이 0.82 이상으로 adiponectin 농도가대사증후군진단에유용한지표임을알수있었다. 그러나, ROC 분석을위한대조군과참고구간산출을위한정상인대상군이같았으므로이연구에서산출한 2.5백분위수참고치를대사증후군진단을위한절단값으로적용하는경우 97.5% 의특이도를보인반면, 민감도는남녀모두 30% 내외로낮아서참고구간을대사증후군진단을위한절단값으로사용하는것은바람직하지않을것으로생각되며, 특정질환의진단을목적으로 adiponectin을측정하는경우질환별로최적의절단값을산출할필요가있을것으로판단된다. 본연구결과에서대사증후군진단을위한최적의 adiponectin 절단값은남자 6.59 μg/ml, 여자 9.27 μg/ml였으며, 이를적용하는경우대사증후군진단에대한민감도와특이도는남자의경우 86.0% 및 70.5%, 여자의경우 78.0% 및 75.0% 였다. 한편, 본연구에서산출한혈청 adiponectin 농도의상위참고구간값이시약의측정가능범위를벗어났으나, adiponectin 농도의감소가질환과관련되어있으므로산출된상위참고구간값의정확성에대한추가적인검증은시행하지않았다. 본연구에서평가한국내제조사의 adiponectin 정량검사시약은 ELISA 원리를사용한검사임을감안할때, 이전의보고들과유사한정밀도를보이는등, 허용할만한수준의검사성능을보였다. Adiponectin과다양한질병의관련성에대한연구가지속적으로이루어지고있는점을고려하여볼때, 자동화된측정원리를적용한검사시약의개발을통하여신속하고용이한검사가가능함은물론, 검사성능을더욱개선함으로써향후임상검사실에서 adiponectin 측정을정규검사로도입할수있을것으로기대된다. 본연구에서설정한한국인의성별및연령별혈청 adiponectin 참고구간은한국인을대상으로다양한질환에서 adiponectin의역할을밝히기위한기본자료로서향후연구에도움을줄수있을것으로기대된다. 요약 배경 : Adiponectin은지방세포에서분비되는혈장단백으로, adiponectin의감소가인슐린저항성및대사증후군과관련이있음이보고되어왔다. 본연구에서는혈청내 adiponectin 농도측정을위한 ELISA 시약의검사및임상성능을평가하고한국인의참고구간을설정하고자하였다. 방법 : AdipoMark Human Adiponectin ELISA 시약 (Mesdia Co., Korea) 의정밀도와직선성을평가하였다. 대사증후군이없는대상군 1,200명에서 adiponectin을측정하여참고구간을설정하였다. 대사증후군환자 100명에서 adiponectin을측정하였다. 결과 : 정밀도평가를위한혈청검체의평균 adiponectin 농도는각각 6.66, 및 μg/ml였고, AdipoMark ELISA 검사는해당농도에대하여각각표준편차 0.90, 1.18 및 2.47 μg/ml, 변이계수 13.6%, 9.3% 및 10.5% 의총비정밀도를보였다. AdipoMark ELISA 검사는혈청 adiponectin 농도가 μg/ml인범위에서직선성을보였다. 한국인의 95% 참고구간은남자 μg/ml, 여자 μg/ml이었다. 대사증후군진단에대한 adiponectin 의상대수행능곡선-곡선하면적은남자 0.85, 여자 0.83이었다. 다변량분석에서낮은 adiponectin 농도가대사증후군과독립적으로연관되어있었다. 결론 : 본연구에서평가한 adiponectin 측정시약은 ELISA 원리를활용한검사임을감안할때허용할만한수준의검사성능을보였다. Adiponectin과관련된것으로알려져있는다양한대사질환에대한연구를위하여신뢰할수있는 adiponectin 측정방법에대한요구가꾸준히증가하고있어, 향후자동화된검사방법을개발하여 adiponectin 측정시약을더개선할수있을것으로기대된다. 감사의글 Adiponectin ELISA 시약을제공해주신메스디아 ( 주 ) 및본연구에사용된검체를제공해주신연세대학교보건대학원지선하교 251

11 수연구팀께감사드립니다. REFERENCE 1. Maeda K, Okubo K, Shimomura I, Funahashi T, Matsuzawa Y, Matsubara K. cdna cloning and expression of a novel adipose specific collagen-like factor, apm1 (AdiPose Most abundant Gene transcript 1). Biochem Biophys Res Commun 1996;221: Arita Y, Kihara S, Ouchi N, Takahashi M, Maeda K, Miyagawa J, et al. Paradoxical decrease of an adipose-specific protein, adiponectin, in obesity. Biochem Biophys Res Commun 1999;257: Ouchi N, Kihara S, Arita Y, Maeda K, Kuriyama H, Okamoto Y, et al. Novel modulator for endothelial adhesion molecules: adipocyte-derived plasma protein adiponectin. Circulation 1999;100: Okamoto Y, Arita Y, Nishida M, Muraguchi M, Ouchi N, Takahashi M, et al. An adipocyte-derived plasma protein, adiponectin, adheres to injured vascular walls. Horm Metab Res 2000;32: Ouchi N, Kihara S, Arita Y, Nishida M, Matsuyama A, Okamoto Y, et al. Adipocyte-derived plasma protein, adiponectin, suppresses lipid accumulation and class A scavenger receptor expression in human monocyte-derived macrophages. Circulation 2001;103: Hotta K, Funahashi T, Arita Y, Takahashi M, Matsuda M, Okamoto Y, et al. Plasma concentrations of a novel, adipose-specific protein, adiponectin, in type 2 diabetic patients. Arterioscler Thromb Vasc Biol 2000;20: Yang WS, Lee WJ, Funahashi T, Tanaka S, Matsuzawa Y, Chao CL, et al. Weight reduction increases plasma levels of an adipose-derived anti-inflammatory protein, adiponectin. J Clin Endocrinol Metab 2001; 86: Weyer C, Funahashi T, Tanaka S, Hotta K, Matsuzawa Y, Pratley RE, et al. Hypoadiponectinemia in obesity and type 2 diabetes: close association with insulin resistance and hyperinsulinemia. J Clin Endocrinol Metab 2001;86: Hotta K, Funahashi T, Bodkin NL, Ortmeyer HK, Arita Y, Hansen BC, et al. Circulating concentrations of the adipocyte protein adiponectin are decreased in parallel with reduced insulin sensitivity during the progression to type 2 diabetes in rhesus monkeys. Diabetes 2001; 50: Rasmussen-Torvik LJ, Wassel CL, Ding J, Carr J, Cushman M, Jenny N, et al. Associations of body mass index and insulin resistance with leptin, adiponectin, and the leptin-to-adiponectin ratio across ethnic groups: the Multi-Ethnic Study of Atherosclerosis (MESA). Ann Epidemiol 2012;22: Liu Q, Yuan B, Lo KA, Patterson HC, Sun Y, Lodish HF. Adiponectin regulates expression of hepatic genes critical for glucose and lipid metabolism. Proc Natl Acad Sci U S A 2012;109: Behnes M, Brueckmann M, Lang S, Putensen C, Saur J, Borggrefe M, et al. Alterations of adiponectin in the course of inflammation and severe sepsis. Shock 2012;38: Dalamaga M, Diakopoulos KN, Mantzoros CS. The role of adiponectin in cancer: a review of current evidence. Endocr Rev 2012;33: Martos-Moreno GÁ, Burgos-Ramos E, Canelles S, Argente J, Barrios V. Evaluation of a multiplex assay for adipokine concentrations in obese children. Clin Chem Lab Med 2010;48: Loo BM, Marniemi J, Jula A. Evaluation of multiplex immunoassays, used for determination of adiponectin, resistin, leptin, and ghrelin from human blood samples, in comparison to ELISA assays. Scand J Clin Lab Invest 2011;71: Cangemi G, Di Iorgi N, Barco S, Reggiardo G, Maghnie M, Melioli G. Plasma total adiponectin levels in pediatrics: reference intervals calculated as a continuous variable of age. Clin Biochem 2012;45: Gavrila A, Peng CK, Chan JL, Mietus JE, Goldberger AL, Mantzoros CS. Diurnal and ultradian dynamics of serum adiponectin in healthy men: comparison with leptin, circulating soluble leptin receptor, and cortisol patterns. J Clin Endocrinol Metab 2003;88: Shand B, Elder P, Scott R, Frampton C, Willis J. Biovariability of plasma adiponectin. Clin Chem Lab Med 2006;44: Pischon T, Hotamisligil GS, Rimm EB. Adiponectin: stability in plasma over 36 hours and within-person variation over 1 year. Clin Chem 2003;49: Wildman RP, Wang D, Fernandez I, Mancuso P, Santoro N, Scherer PE, et al. Associations of testosterone and sex hormone binding globulin with adipose tissue hormones in midlife women. Obesity (Silver Spring) Riestra P, García-Anguita A, Lasunción MA, Cano B, de Oya M, Garcés C. Relationship of adiponectin with metabolic syndrome components in pubertal children. Atherosclerosis 2011;216: Yoon SJ, Lee HS, Lee SW, Yun JE, Kim SY, Cho ER, et al. The association between adiponectin and diabetes in the Korean population. Metabolism 2008;57: Kotani K and Sakane N. Leptin:adiponectin ratio and metabolic syndrome in the general Japanese population. Korean J Lab Med 2011;31: