Clinical Chemistry
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Clinical Chemistry 54: 356-365, 2008. First published December 18, 2007; 10.1373/clinchem.2007.091843
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
clinchem.2007.091843v1
54/2/356    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ichihara, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ichihara, K.
Related Collections
Right arrow General Clinical Chemistry
(Clinical Chemistry. 2008;54:356-365.)
© 2008 American Association for Clinical Chemistry, Inc.

General Clinical Chemistry

Sources of Variation of Commonly Measured Serum Analytes in 6 Asian Cities and Consideration of Common Reference Intervals

Kiyoshi Ichihara1,a, Yoshihisa Itoh2, Christopher W. K. Lam3, Priscilla M. K. Poon3, Jeong-Ho Kim4, Hiroshi Kyono5, Naning Chandrawening6, Dewi Muliaty6 Science Committee for the Asian-Pacific Federation of Clinical Biochemistry

1 Faculty of Health Sciences, Yamaguchi University, Ube, Japan; 2 Department of Laboratory Medicine, Asahikawa Medical College, Asahikawa, Japan; 3 Department of Chemical Pathology, Chinese University of Hong Kong, Hong Kong; 4 Department of Laboratory Medicine, Yonsei University, Seoul, Korea; 5 Yuan Ching Clinical Laboratory, Taipei, Taiwan; 6 Prodia Clinical Laboratory, Jakarta, Indonesia.

aAddress correspondence to this author at: Department of Laboratory Sciences, Faculty of Health Sciences, Yamaguchi University Graduate School of Medicine, Minami-Kogushi 1–1-1, Ube, 755-8505 Japan. Fax +81-836-35-5213; e-mail ichihara{at}yamaguchi-u.ac.jp.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: In a previous study to determine the feasibility of common reference intervals in Asia, we found significant differences among populations from 6 cities. In this study, we attempted to define the sources of these differences.

Methods: We enrolled 580 healthy volunteers (279 men, 301 women, 20–62 years old), after a selection process that was based on the Clinical and Laboratory Standards Institute guidelines, and used a lifestyle questionnaire. All sera were obtained at a basal state and frozen at –80 °C until the collective assay was done. We measured 21 basic chemical analytes and 10 serum proteins.

Results: We used 3-level nested ANOVA to separate the variation (SD) into between-city (SD-city), between-sex (SD-sex), between-age (SD-age), and between-individual (SD-indiv) components. SD-indiv corresponds to one-quarter of the "pure" reference interval obtained after removing variations due to city, sex, and age. The SD-sex to SD-indiv ratio was >0.8 for creatinine, urate, retinol-binding protein, and transthyretin. We observed high SD-city to SD-indiv ratios, ranging from 0.4 to 0.7, for 11 analytes including lactate dehydrogenase (LDH), electrolytes, IgG, and complement components and SD-age to SD-indiv ratios >0.4 for LDH, alkaline phosphatase, and total cholesterol. Multiple regression analysis demonstrated several other relevant sources of variation, including body mass index, alcohol consumption, and cigarette smoking, although their contributions were generally smaller than those for sex, region, or age.

Conclusion: We observed unacceptably large regional differences in measured values of some analytes even after adjustment for age, sex, and lifestyle variables. Genetic and environmental factors may account for the residual differences.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Two of the stated goals for the preparation of the international serum protein reference material CRM4701 were to decrease between-manufacturer variance in assays and to determine the feasibility of common reference materials (1)(2)(3)(4). In the previous Asian Reference Interval Project, we used a protocol developed by the IFCC Committee on Plasma Proteins (5) to explore the feasibility of using common reference intervals (RIs) for these proteins in different geographical regions in Asia.

Contrary to our expectation, we found large differences in test results for individuals from 6 Asian cities, not only for major serum proteins but also for other commonly measured biochemical analytes. The analytical results were comparable because all specimens were freshly collected and measured collectively in one laboratory. Unfortunately, we did not obtain detailed demographic information from the donors and were unable to analyze factors possibly accounting for the differences.

In this second study, we sought to confirm our earlier findings, using a lifestyle questionnaire to explore possible causes for the differences and evaluating possible geographical differences in other analytes. Because we did not seek to determine the RI for each analyte, the number of study participants was set to approximately 60–100 per city, adequate for analyzing between-city differences adjusted for various demographic factors.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
study participants
The objectives of the project and the protocol were sent to the member nations of the Asian Pacific Federation of Clinical Biochemistry (APFCB). Six laboratories expressed a desire to participate in the study: the central laboratories of Prince of Wales Hospital (Shatin, Hong Kong), Yonsei University Hospital (Seoul, South Korea), Yamaguchi University Hospital (Ube, Yamaguchi, Japan), Asahikawa Medical College Hospital (Asahikawa, Japan), Yuan Ching Clinical Laboratory (Taipei, Taiwan), and Prodia Clinical Laboratory (Jakarta, Indonesia).

The target population was healthy individuals 20–60 years old. To avoid possible bias attributable to differences in physical activity and/or climatic influences depending on working environment, we limited the participants primarily to those who were working for the clinical laboratory or its affiliated hospital. However, those healthy individuals who were family members of the laboratory staff and regarded as having similar levels of daily physical activities and working indoors were included. Their proportion was <15% in all cities except Taipei, where the size of the laboratory was small. Therefore, white-collar workers from a company affiliated with the laboratory were also enrolled as equivalent to the laboratory workers. The planned number of participants was 60–100 per city, as shown in Table 1 .


View this table:
[in this window]
[in a new window]

  		Table 1. Planned composition of healthy participants.

The following exclusion criteria were applied during recruitment: anyone who reported having any chronic disorder requiring regular medication, who had just recovered (≤14 days) from surgery or any acute illness, or who was pregnant, a heavy smoker (≥25 cigarettes/day), or overweight [body mass index (BMI) >30]. The numbers of individuals who agreed to participate are shown in Table 2 .


View this table:
[in this window]
[in a new window]

  		Table 2. Actual composition (age and sex) of the healthy participants.

lifestyle questionnaire
The questionnaire survey, conducted at the time of blood collection, was designed to obtain demographic information recommended in the Clinical and Laboratory Standards Institute (CLSI) guideline (6) as well as additional information to explore other factors possibly associated with regional differences. These other factors included regular exercise (at least once per week for ≥1 years: yes or no), average h/day of standing, cigarette smoking, approximate alcohol consumption (quantified as g ethanol/day), and amount of various food types consumed per week in 5 grades for meat, fish, vegetables, beans/tofu, milk/dairy, eggs, fruits, sweets, fried food, and salty food.

specimens and assays
Blood was sampled between 0800 and 1000 after an overnight fast (≥10 h). The volunteers were asked to remain seated for at least 15 min before sampling. Blood (10 mL) was drawn from each volunteer into evacuated serum-separator blood collection tubes. The collected samples were centrifuged at room temperature for 10 min except in Taiwan, where the sample tubes were centrifuged for 5 min at 4 °C. The sera were poured into 2 separate containers and then deep frozen at –80 °C.

The measured analytes consisted of 2 groups. Group 1 included aspartate transaminase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), lactate dehydrogenase (LDH), {gamma}-glutamyltransferase (GGT), creatine kinase (CK), amylase, triglyceride (TG), total cholesterol (TCho), HDL cholesterol, sodium, potassium, chloride, calcium adjusted for albumin, inorganic phosphate (IP), total protein (TP), albumin, globulin (Glb, computed as TP – albumin), creatinine (CRE), urea nitrogen (UN), and urate (UA). Group 2 included 10 serum proteins: IgG, IgA, IgM, complement components (C3, C4), transferrin (Tf), transthyretin (TTR), retinol binding protein (RBP), cystatin C (CysC), and C-reactive protein (CRP).

All the specimens were sent by air freight in deep-frozen state to the Special Reference Laboratory, Inc. (Hachiohji, Tokyo) by July 2006. Measurements for the group 1 analytes were performed using a Hitachi 7170 autoanalyzer. The analytical methods employed were as follows: AST, ALT, LDH, ALP, GGT, CK, and amylase by a Japanese Society of Clinical Chemistry (JSCC) recommended method, using the Japanese certified enzyme reference material (7); TP, biuret method; albumin, bromcresol green method; TCho, TG, CRE, UN, and UA, enzymatic methods; HDL, direct enzymatic method with specific inhibition of cholesterol in other fractions; Na, K, and Cl, ion-selective electrode method; Ca, o-cresolphthalein complexone method; and IP, molybdate method. The group 2 analytes were assayed by timed nephelometry using a Behring Nephelometer-II Analyzer (Dade Behring Ltd).

statistical analyses
On the basis of a preliminary analysis, we performed logarithmic transformation of test results for AST, ALT, GGT, CK, TG, and CRP to normalize the data distributions. Values outside the mean and 4SD were deleted as outliers.

The magnitudes of the between-city (btw-city), between-sex (btw-sex), and between-age (btw-age) group components of variation were computed by use of 3-level nested ANOVA (8) after coding each record by city, sex (0 = male, 1 = female), and age group (1 = 20–29, 2 = 30–39, 3 = 40–49, 4 = 50+ years). Each component of variation was derived as a variance and then expressed as SD by taking the square root. These components were designated as between-city SD (SDcity), between-sex SD (SDsex), and between-age SD (SDage). The SD for the residual variance provided by the 3-level nested ANOVA is regarded as the crude between-individual SD (SDindiv). The SDindiv corresponds to one-quarter of "pure" RI derived after removing variations due to age group, sex, and city. Thus the magnitude of SDindiv represents the sum of (a) between-individual variation (not attributable to age, sex, or city, but to other factors, such as smoking, level of exercise, alcohol consumption, or preferences for foods), (b) within-individual variation (mainly dependent on sampling conditions), and (c) analytical variation. For those analytes that were logarithmically transformed before performing the 3-level nested ANOVA, the back-transformed mean (M) and SD were computed as follows from the logarithmic mean and SD (MT and SDT):

Formula

Formula

Multiple regression analysis was performed to explore sources of variation other than city, sex, and age. For this purpose, the above information obtained from the lifestyle questionnaire was used for the explanatory variables. To adjust for regional differences, dummy variables representing each city were introduced by setting Yamaguchi as the reference category. A dummy variable for sex was set as male = 0, female = 1. A best-fit regression model was sought by use of a stepwise selection method; however, the dummy variables for sex and city were always included irrespective of their significance to adjust for their possible confounding influences.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
sex and regional differences
Table 3 shows the results of the 3-level nested ANOVA. The magnitudes of 4 components of variance, btw-city, btw-sex, btw-age, and btw-indiv, were expressed as the relative percentage of the overall variance and listed in columns 4–7 of Table 3 . Contributions >10% are marked by superscript a and b. The btw-indiv variance is regarded as the residual variance obtained after extracting the variation from city, sex, and age.


View this table:
[in this window]
[in a new window]

  		Table 3. Results of the 3-level nested ANOVA.

Each component of variance was transformed to the SD and listed in columns 8–11, denoted btw-city (SDcity), btw-sex (SDsex), btw-age (SDage), and btw-indiv (SDindiv). Because SDindiv is regarded as equivalent to one-quarter of the RI, the ratio of SDcity, SDsex, or SDage to SDindiv (SD ratio) was computed to compare the extent of influence of each SD component in deriving the RI. The ratios are shown within parentheses next to the values for SDs. We empirically set a critical value for the SD ratio as 0.4. Above this value, we usually consider partitioning reference values (see "Discussion" for reasoning).

The SD ratios for btw-sex (SDsex/SDindiv) >0.4 were noted for ALT, GGT, CK, HDL, TG, UN, CRE, UA, IgM, TTR, RBP, and CysC. The SD ratios for btw-city (SDcity/SDindiv) were >0.4 for LDH, TP, Glb, UN, Na, K, Cl, IP, IgG, C3, and C4. SD ratios for btw-age (SDage/SDindiv) were >0.4 for LDH, ALP, TCho, and IP.

When the SD ratio is between 0.25 and 0.4, partitioning reference values may not be required, but the factors have to be considered in interpreting the test results.

Fig. 1 shows the distributions of actual test results for representative analytes, grouped by city and sex. The cities were aligned by geographical latitude from north to south. The box within each scattergram denotes the central 50% range (25th to 75th centiles), and the vertical line in the middle of the box indicates the median. The magnitudes of btw-city and btw-sex differences are shown at the top of each panel.


Figure 1
Figure 1
View larger version (61K):
[in this window]
[in a new window]

  		Figure 1. Distribution of test results for representative analytes subgrouped by city and sex.

Distribution (1-dimensional scattergram) of test results are shown grouped by sex and city. The cities were aligned by geographical latitude from north to south. The box within each scattergram denotes a central 50% range (25th to 75th centile), and the vertical line in the middle of the box indicates the median or 50th centile. The heading at the top of each graph depicts ratios of between-city or between-sex SDs to between-individual SDs. The smaller the SD ratio, the more likely that the reference interval can be computed without subgrouping by city or sex. continued on page 363

other sources of differences
Table 4 shows the results of multiple regression analysis. The contributions of sex- and city-related differences were expressed in terms of overall regression model fit as the multiple correlation coefficient (R). The R for btw-sex difference shown in column 3 was obtained using a regression model that included only sex as an explanatory variable. The R in column 4 was obtained using another model including both dummy variables for sex and city together. Thus the difference of R between columns 3 and 4 represents the contribution of btw-city differences.


View this table:
[in this window]
[in a new window]

  		Table 4. Components of variations of test results revealed by multiple regression analysis

The R in column 5 was derived from a full model containing the relevant explanatory variables selected by the stepwise method. In this model, variables for sex and city were always included as control variables to adjust for their possible confounding influence.

The explanatory variables that exceed a significance level of P <0.01 are listed in the righthand columns. The relative importance of the variables representing age, BMI, cigarette-smoking, alcohol (g ethanol/day), and regular exercise was evaluated via the statistical significance of the t values in the related columns. t Values with minus signs indicate that the test values are inversely related to the level of the factor concerned. Miscellaneous variables are listed in the last 2 columns with the variable names and associated t values.

Comparing the R values in columns 3–5, essentially the same information was obtained as in Table 3Up regarding contribution of sex, city, and age. The contributions of the other variables are relatively small in most cases, as judged from the minor increment of R value by their inclusion.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The international laboratory medicine community is actively pursuing the possibility of global standardization of laboratory tests, at least for commonly measured analytes. Accompanying this trend, there is growing momentum for deriving common RIs by the collaborative effort of clinical laboratories. Such an effort has already been made by the Scandinavian Society of Clinical Chemistry, which published common RIs to be used in the allied countries (9)(10)(11)(12). As a whole, the Scandinavian Society believed there were no practically significant differences among their countries in test results for any analyte. However, as discussed below, this judgment was based on where to set acceptable limits for implementing common RI.

We launched a project in 1999 (5) to explore the feasibility of deriving common RIs in 6 Asian cities (Hong Kong, Shanghai, Seoul, Kuala Lumpur, Taipei, and Tokyo) for 13 serum proteins (group 2 analytes plus 3) standardized by CRM470. We also measured 7 clinical chemical analytes (AST, ALT, GGT, CRE, TCho, HDL, and TG) whose results were intended to be used in excluding individuals with latent abnormalities. Contrary to our expectation, we observed large between-city variations for many analytes, although we assayed all specimens collectively at a single laboratory in Japan.

We compared the results obtained in the 2 studies for 10 proteins measured in common. Moderate to large between-city differences (SDcity/SDindiv ratios) observed in test results for IgG (0.38), C3 (0.78), C4 (0.59), and CRP (0.39) in the first study were also observed in the second study (0.44, 0.63, 0.54, and 0.26, respectively). The lack of noticeable differences for transferrin (0.12), RBP (0.0), and CysC (0.00) in the first study was also seen in the second (0.00, 0.00, and 0.12). As a whole, we confirmed the presence of moderate regional differences in the serum proteins. However, the regional differences observed in the first study for IgA (0.33), IgM (0.33), and TTR (0.50) were not reproduced in the second (0.15, 0.00, and 0.00).

Among the group 2 analytes, IgG, C3, and C4 showed the most prominent between-city differences. These proteins are markers for chronic inflammation. Because the contributions of other factors revealed by multiple regression analysis were much less than the between-city differences, the observed differences are likely to represent a difference in the level of exposure to infectious agents. If the cities are sorted by their latitudes as in Fig. 1Up , progressively higher values of IgG and C3 are seen as the location approaches the tropics, suggesting higher exposure to infectious agents in cities closer to the tropics. A similar regional trend was also observed for TP and Glb.

Racial differences in serum concentrations of proteins cannot be neglected; Johnson et al. (13) reported that whites and Asian Indians living in the same community in England showed different concentrations of serum IgG and {alpha}1-acid glycoprotein, and that serum concentrations of {alpha}1-antitrypsin varied according to genetic phenotype. Therefore, genetic factors may contribute to between-city differences, although we could not distinguish them from environmental factors with the information available in this study.

The 7 common analytes measured in the first study (AST, ALT, GGT, TCho, TG, HDL, and CRE), were also included in this study. In the first, we observed moderate between-city differences in test results for all but AST, but we found no apparent differences in this study. This discrepancy may be due to the stricter criteria used for selecting reference individuals. All of these analytes except CRE are related to the metabolic syndrome, and thus between-city differences in its prevalence may have led to differences in test results for those analytes. The SDcity/SDindiv for BMI in this study was zero, suggesting there was no difference in the prevalence of obesity in the 6 cities.

For the other analytes measured only in the second study, we observed unexpectedly large between-city differences in test results, especially for LDH, TP, Glb, UN, Na, K, Cl, and IP. Lifestyle-related factors may have resulted in spurious between-city differences. However, multiple regression analysis (Table 4Up ) revealed that between-city differences remain highly significant, with only slight changes in the magnitude of R, by inclusion of the lifestyle factors.

As for LDH, no report of regional difference has been published, but plausible factors include (a) hemolysis due to inappropriate sampling procedure in some laboratories, (b) muscular exertion, and (c) genetic factors. Hemolysis-induced LDH increase was unlikely, with no observed correlation between serum concentrations of LDH and K in any city. Furthermore, hemolysis was not reported by the autoanalyzer for any specimen. A different level of exercise could be a factor, because a positive correlation of LDH was observed with both AST and CK (Spearman correlation coefficient rS = 0.446, 0.329, n = 571). However, the questionnaire-based level of exercise showed no association with LDH. Therefore, genetic factors are more likely. Bathum et al. (14) reported, after twin-based analysis of test results for LDH, ALT, GGT, and total bilirubin in an elderly group, that the analytes showed high heritability even after adjustment for alcohol consumption and BMI. The large btw-city variation observed for TP and Glb appears to be related to the clear-cut between-city differences in test results for serum proteins, especially for IgG, C3, and C4. Although concentrations of TP and Glb were lower in smokers, multiple regression analysis showed that between-city differences remain highly significant after adjusting for the level of smoking.

Although large regional differences were observed for Na, K, Cl, IP, and UN, the concentrations of these analytes are less dependent on genetic factors. Because their concentrations are under tight physiologic control by homeostatic mechanisms, the observed between-city differences for these analytes are more likely a function of diet and/or the local climate. Geographical differences in potassium concentration appear to be determined by the balance between oral intake and excretion in urine and sweat (15)(16). Because our study population was limited primarily to reference individuals who worked in the hospital laboratory, differences in excretion by perspiration should be negligible. Therefore, variation in dietary intake is likely a more important determinant in our population than differences in potassium excretion. Unfortunately, our questionnaire was not detailed enough to identify dietary variation that could account for these differences in potassium or those seen with Na, Cl, and UN.

Our current results indicate there are apparent between-city differences in many analytes. Several criteria have been proposed to aid in deciding whether observed differences in analyte concentrations across groups are large enough to justify using separate RIs rather than common RIs. In the Nordic Project cited above, 3 criteria were considered. The Harris and Boyd (17) criterion cited in the CLSI guideline (6) uses a formula based on differences in 2 subgroup means adjusted for the sample sizes. The second criterion, proposed by Lahti (12), is based on the percentages of reference values in each subgroup lying outside the upper and lower limits of the common RI derived without partitioning. If the proportion at either end exceeds defined limits, partitioning is permitted. The third criterion, proposed by Fraser et al. (18), is based on the magnitude of biological variation expressed as the standard deviation (sbio), which includes both between- and within-individual components. If the SD accounting for between-subgroup variation is >0.375sbio, separate RIs are to be considered.

The first 2 criteria are directly applicable when there are only 2 subgroups. In our case with multiple subgroups, it is difficult to make any judgment by all pairwise comparisons. Recently Gellerstedt (19) reported a method for determining the requirement for partitioning multiple subgroups as an extension of Lahti’s approach (12). The method focuses on the presence of subgroups with high frequencies (≥4%) of participants outside the common RI. We applied the method to evaluate between-city differences. Acknowledging the limitation of small sample sizes, we found many subgroups exceeding the critical frequencies in all 31 analytes. Therefore we selected Fraser’s criterion as more appropriate to aid in deciding whether to use a common RI. Our criterion to justify partition, SDcity/SDindiv ≥0.4, is essentially very similar to Fraser’s criterion; however, our SDindiv was obtained by removing btw-age, btw-city, and btw-sex components, whereas Fraser’s sbio corresponds to gross biological variations without subgrouping by age. Therefore, our SDindiv is generally narrower than sbio. If a large between-city difference exists, the RI without partitioning is always wider than the average RI derived after partitioning. As a result, Fraser’s critical SD (0.375sbio) also gets wider proportionately, leading to a more conservative judgment not to allow partition. Therefore, compared to Fraser’s ratio of 0.375, our critical ratio of 0.4 seems to be either comparable to or more permissive of partitioning. We could set the critical ratio to a higher level in compensation for a narrower measure of between-individual variation; however, SD ratios for between-sex variation or SDsex/SDindiv that exceeded 0.4 were observed for ALT, GGT, CK, HDL, TG, UN, CRE, and UA among the group 1 analytes. Separate RIs for each sex are almost always used for all of the analytes except UN. On the other hand, the analytes with SDsex/SDindiv <0.4 have not traditionally had sex-specific RIs in the 20- to 60-year age interval. This empirical practice began long before discussions of partitioning criteria. In that sense, our critical number of 0.4 seems to be in a practical range.

In summary, this study was not an attempt to define RIs for any population, but rather to demonstrate the presence or absence of differences in analyte concentrations in populations that were as similar as possible except in the city of residence. We observed large between-city differences in some commonly measured analytes among 6 Asian cities, but our sample size and target population was not large enough to generalize the finding to a larger population. Our results suggest that the feasibility of setting common RIs should be confirmed by direct samplings and measurements from the population concerned. Our measure of practical difference, the SD ratio derived from the 3-level nested ANOVA, may be useful in judging the feasibility of adopting a common RI.


   Acknowledgments
 
Grant/funding Support: In carrying out this study, we did not use any specific grant or fund, but used personal research funds of general use from various sources.

Financial Disclosures: None declared.

Acknowlegments: This study was made possible by the generous support of reagents for the measurements of serum proteins provided by Dade Behring Marburg GmbH, and by the generous support offered by Special Reference Laboratory (SRL), Tokyo, in carrying out the assays for all the analytes. We are very grateful to the eager support and donation of blood by the staffs and associated members of the participating laboratories. We are also grateful to Myron Johnson for his careful editing of this manuscript to improve not only English expressions but also scientific reasoning.


   Footnotes
 
1 Nonstandard abbreviations: CRM470, certified reference material 470; RI, reference interval; BMI, body mass index; CLSI, Clinical and Laboratory Standards Institute; AST, aspartate transaminase; ALT, alanine aminotransferase; ALP, alkaline phosphatase; LDH, lactate dehydrogenase; GGT, {gamma}-glutamyltransferase; CK, creatine kinase; TG, triglyceride; TCho, total cholesterol; IP, inorganic phosphate; TP, total protein; Glb, globulin; CRE, creatinine; UN, urea nitrogen; UA, urate; Tf, transferrin; TTR, transthyretin; RBP, retinol binding protein; CysC, cystatin C; CRP, C-reactive protein.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Ichihara K, Kawai T. Determination of reference intervals for 13 plasma proteins based on IFCC international reference preparation (CRM470) and NCCLS proposed guideline (C28-P, 1992): a strategy for partitioning reference individuals with validation based on multivariate analysis. J Clin Lab Anal 1997;11:117-124.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  2. Itoh Y, Ichihara K, Kanno T, Sugawara T, Ohkubo A, Hirabayashi Y, et al. Serum protein standardization project in Japan: evaluation of an IFCC reference material (RPPHS/CRM470) and establishment of reference intervals. J Clin Lab Anal 1997;11:39-44.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  3. Itoh Y, Ichihara K. Standardization of immunoassay for CRM 470-related proteins in Japan: from evaluating CRM 470 to setting reference intervals. Clin Chem Lab Med 2001;39:1054-1061.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  4. Johnson AM, Whicher JT. Effect of certified reference material 470 (CRM 470) on national quality assurance programs for serum proteins in Europe. Clin Chem Lab Med 2001;39:1123-1128.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  5. Ichihara K, Itoh Y, Min WK, Sook FY, Lam CWK, Kong XT, et al. Diagnostic and epidemiological implications of regional differences in serum concentrations of proteins observed in six Asian cities. Clin Chem Lab Med 2004;42:800-809.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  6. NCCLS C28–A2: How to Define and Determine Reference Intervals in the Clinical Laboratory. 2nd edition 2000 National Committee for Clinical Laboratory Standards Villanova, PA. Approved Guideline.
  7. Kanno T. Improvement of inter-laboratory imprecision in enzyme measurements. Rinsho Byori 2001;(Suppl 116):16-25[in Japanese].
  8. Sokal RR, Rohlf FJ. Biometry. 3rd edition 1995:272-320 W.H. Frieman New York. .
  9. Rustad P, Felding P, Lahti A, Hyltoft Petersen P. Descriptive analytical data and consequences for calculation of common reference intervals in the Nordic Reference Interval Project 2000. Scand J Clin Lab Invest 2004;64:343-370.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  10. Rustad P, Felding P, Franzson L, Kairisto V, Lahti A, Martensson A, et al. The Nordic Reference Interval Project 2000: recommended reference intervals for 25 common biochemical properties. Scand J Clin Lab Invest 2004;64:271-284.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  11. Stromme JH, Rustad P, Steensland H, Theodorsen L, Urdal P. Reference intervals for eight enzymes in blood of adult females and males measured in accordance with the International Federation of Clinical Chemistry reference system at 37 degrees C: part of the Nordic Reference Interval Project. Scand J Clin Lab Invest 2004;64:371-384.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  12. Lahti A. Are the common reference intervals truly common? Case studies on stratifying biochemical reference data by countries using two partitioning methods. Scand J Clin Lab Invest 2004;64:407-430.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  13. Johnson AM, Hyltoft Petersen P, Whicher JT, Carlstrom A, MacLennan S, . International Federation of Clinical Chemistry and Laboratory Medicine, Committee on Plasma Proteins. Reference intervals for serum proteins: similarities and differences between adult Caucasian and Asian Indian males in Yorkshire, UK. Clin Chem Lab Med 2004;42:792-799.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]
  14. Bathum L, Petersen HC, Rosholm JU, Hyltoft Petersen P, Vaupel J, Christensen K. Evidence for a substantial genetic influence on biochemical liver function tests: results from a population-based Danish twin study. Clin Chem 2001;47:81-87.[Abstract/Free Full Text]
  15. Reidenberg MM, Gu ZP, Lorenzo B, Coutinho E, Athayde C, Frick J, et al. Differences in serum potassium concentrations in normal men in different geographic locations. Clin Chem 1993;39:72-75.[Abstract]
  16. Sriboonlue P, Prasongwatana V, Suwantrai S, Bovornpadungkitti S, Tungsanga K, Tosukhowong P. Nutritional potassium status of healthy adult males residing in the rural northeast Thailand. J Med Assoc Thai 1998;81:223-232.[Medline] [Order article via Infotrieve]
  17. Harris EK, Boyd JC. On dividing reference data into subgroups to produce separate reference ranges. Clin Chem 1990;36:265-270.[Abstract/Free Full Text]
  18. Fraser CG, Hyltoft Petersen P, Libeer JC, Ricos C. Proposals for setting generally applicable quality goals solely based on biology. Ann Clin Biochem 1997;24:8-12.
  19. Gellerstedt M. Partitioning reference values of several Gaussian subpopulations with unequal prevalence: a procedure with computer program support. Clin Chem Lab Med 2006;44:1258-1263.[CrossRef][Web of Science][Medline] [Order article via Infotrieve]



The following articles in journals at HighWire Press have cited this article:


Home page
 Ann Clin BiochemHome page
J. H Barth
Reference ranges still need further clarity
Ann Clin Biochem, January 1, 2009; 46(1): 1 - 2.
[Full Text] [PDF]

Home page
 Ann Clin BiochemHome page
F. Ceriotti, R. Hinzmann, and M. Panteghini
Reference intervals: the way forward
Ann Clin Biochem, January 1, 2009; 46(1): 8 - 17.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Chem.Home page
J. C. Boyd
Cautions in the Adoption of Common Reference Intervals
Clin. Chem., February 1, 2008; 54(2): 238 - 239.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
clinchem.2007.091843v1
54/2/356    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (1)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ichihara, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ichihara, K.
Related Collections
Right arrow General Clinical Chemistry

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS