배경 : HbA1c는 혈액투석환자에서는 혈당을 저평가하는 경 향이 있어 근래에는 당화알부민이 혈당조절지표로서 대두되고 있다. 저자들은 당화알부민 시약인 Lucica GA-L (Asahi Kasei Pharma Co., Japan)의 검사수행능을 검증하고 그것의 평균혈당 추정능을 HbA1c와 비교하고자 했다. 방법 : 정밀도와 직선성, 간섭물질에 대한 검증을 시행하고 참 고치를 설정했다. 3개월, 2개월, 1개월 전의 혈당 결과를 가진 ‘비투석군’38명과‘혈액투석군’70명에서 HbA1c와 당화알부 민을 측정하고 두 군에서 혈당 가중평균과 HbA1c 또는 당화알 부민의 상관계수와 단순선형회귀식의 기울기를 비교하였다. 단 계별 변수선택법으로 중회귀분석을 시행하여 혈액투석여부가 HbA1c와 당화알부민에 유의한 영향을 미치는지 보고자 했다. 결과 : 총변이계수는 당화알부민 13.73%에서 2.2%였고 24.62 %에서 2.8%였다. 15.74-62.09%의 농도 범위에서 직선성이 확 인되었다. 참고치는 남성에서 9.98-16.49%, 여성에서 11.35- 17.55%이었다. 혈액투석군에서의 혈당 가중평균과 HbA1c의 상관계수는 0.568-0.625이었고 비투석군에서는 0.735-0.783 이었다. 혈당 가중평균과 당화알부민의 상관계수는 혈액투석군 에서 0.682-0.713이었고 비투석군에서는 0.640-0.677이었 다. 혈액투석군에서 HbA1c와 혈당가중평균의 회귀식 기울기는 0.012-0.014로 비투석군의 0.029-0.032에 비해 약 44.0% 수 준이었고 당화알부민과 혈당가중평균의 회귀식 기울기는 혈액 투석군에서 0.080-0.090으로 비투석군에서의 0.130-0.147에 비해 약 62.0% 수준이었다. HbA1c와 당화알부민 사이의 회귀 식 기울기는 혈액투석군에서 0.119로 비투석군의 0.163에 비해 약 73.0% 수준이었다. 중회귀분석 상, 혈액투석여부가 HbA1c 에는 유의한 영향을 준 반면, 당화알부민에는 그렇지 않았다. 결론 : 당화알부민 시약 Lucica GA-L의 검사수행능은 적절 한 것으로 확인되었다. 혈액투석군에서 측정된 HbA1c는 비투 석군에 비해 혈당가중평균에 둔감하게 반응하고 혈액투석 여부 의 영향을 받으므로 혈액투석 환자에서는 HbA1c의 약점을 보 완하기 위한 당화알부민의 측정이 추천된다.

Background : The HbA1c has been considered to underestimate glucose level in diabetic patients on hemodialysis, therefore, glycated albumin (GA) was recently introduced to assess the glycemic control for those cases. We evaluated the performance of GA assay kit of Lucica GA-L (Asahi Kasei Pharma Co., Japan) and compare it with HbA1c for estimating glucose levels. Methods : Tests for precision, linearity and interference were performed and reference interval was determined. Thirty eight of non-hemodialysis and seventy of hemodialysis patients were recruited, whose glucose levels of three-, two- and one-month prior to this study were available for calculating weighted means of glucose (WMGs). The correlation coefficients and the slopes of regression equation between WMG and HbA1c or GA were compared between two groups. Multiple linear regression analyses were used to determine significant predictor for HbA1c and GA. Results : Total CV was 2.2% at concentration of 13.7% and 2.8% at 24.6%. The dilution curve between 15.7% and 62.1% was linear. Reference intervals were 10.0% to 16.5% for male and 11.4% to 17.6% for female. The correlation coefficients between WMG and GA were 0.682-0.713 in hemodialysis and 0.640-0.677 in non-hemodialysis. Those between WMG and HbA1c were 0.568-0.625 in hemodialysis and 0.735-0.783 in non-hemodialysis. The slopes of regression equation between GA and WMG in hemodialysis were 0.080-0.090 and 0.130-0.147 in non-hemodialysis. Those between HbA1c and WMG in hemodialysis were 0.012-0.014 and 0.029-0.032 in non-hemodialysis. GA was not influenced by hemodialysis status while HbA1c was. Conclusions : The claimed performance characteristic of Lucica GA-L were verified. WMG were better reflected by GA rather than HbA1c in patients on hemodialysis.

Background : The HbA1c has been considered to underestimate glucose level in diabetic patients on hemodialysis, therefore, glycated albumin (GA) was recently introduced to assess the glycemic control for those cases. We evaluated the performance of GA assay kit of Lucica GA-L (Asahi Kasei Pharma Co., Japan) and compare it with HbA1c for estimating glucose levels. Methods : Tests for precision, linearity and interference were performed and reference interval was determined. Thirty eight of non-hemodialysis and seventy of hemodialysis patients were recruited, whose glucose levels of three-, two- and one-month prior to this study were available for calculating weighted means of glucose (WMGs). The correlation coefficients and the slopes of regression equation between WMG and HbA1c or GA were compared between two groups. Multiple linear regression analyses were used to determine significant predictor for HbA1c and GA. Results : Total CV was 2.2% at concentration of 13.7% and 2.8% at 24.6%. The dilution curve between 15.7% and 62.1% was linear. Reference intervals were 10.0% to 16.5% for male and 11.4% to 17.6% for female. The correlation coefficients between WMG and GA were 0.682-0.713 in hemodialysis and 0.640-0.677 in non-hemodialysis. Those between WMG and HbA1c were 0.568-0.625 in hemodialysis and 0.735-0.783 in non-hemodialysis. The slopes of regression equation between GA and WMG in hemodialysis were 0.080-0.090 and 0.130-0.147 in non-hemodialysis. Those between HbA1c and WMG in hemodialysis were 0.012-0.014 and 0.029-0.032 in non-hemodialysis. GA was not influenced by hemodialysis status while HbA1c was. Conclusions : The claimed performance characteristic of Lucica GA-L were verified. WMG were better reflected by GA rather than HbA1c in patients on hemodialysis.