| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Folkhälsan Research Center, Institute for Preventive Medicine, Nutrition and Cancer, and Division of Clinical Chemistry, Biomedicum, PB 63, University of Helsinki, FIN-00014 Helsinki, Finland
aauthor for correspondence: fax 358-9-191-25452, e-mail katariina.stumpf{at}helsinki.fi
Enterolactone, a mammalian lignan, is produced by colonic microflora from precursors present in food plants. Because intake of vegetables, fruits, berries, or whole grains is related to the enterolactone concentration in blood (1)(2) and excretion in urine (3)(4), enterolactone may function as a biomarker of fiber-rich foods. Important characteristics for a biomarker include convenient, low-risk collection of samples; specific, reliable laboratory measurement; and a high ratio of between-person to total variability [intraclass correlation (ICC)] (5). In epidemiologic studies, analytes with a low ICC often show weak associations with any disease (6).
The present study describes the short-term variation in enterolactone in serum, 24-h urine, and spot-urine enterolactone:creatinine ratio and the relationship between enterolactone concentrations in serum, 24-h urine, and spot urine.
The study protocol was approved by the Ethics Committee on Epidemiology and Public Health in the hospital district of Helsinki and Uusimaa. Twenty volunteers (13 females and 7 males) were recruited among university students. Exclusion criteria included age <18 years, antibacterial treatment during the preceding 3 months, and any major illness or regular medication, except contraceptive pills. The average age of the participants was 22.2 years [95% confidence interval (CI), 21.4–22.9 years], and the average body mass index was 22.3 (20.9–23.6) kg/m2. Of the female participants, seven took oral contraceptives. The female participant who regularly took an antidepressant reported this only at the end of the study. One participant dropped out because of antibacterial treatment for a urinary tract infection. One spot-urine sample was missing for one female participant, who was thus excluded from that analysis.
The samples were collected on 5 successive days (Monday to Friday) for within-week variation and on the following 3 Mondays for within-month variation. Participants began collecting their urine 24 h before serum and spot-urine samples were collected. No specific timing was demanded for spot-urine collection. The venous samples were drawn after a 4-h fast, starting at 1600 h. The blood samples were collected by venipuncture and centrifuged before separation of the serum. The volumes of the 24-h urine samples were measured before storing. Some participants reported the approximate amount of urine not included in the 24-h collection, and this reported volume was added to the measured urine volume. All samples were stored frozen at -20 °C.
Enterolactone concentrations were analyzed by time-resolved fluoroimmunoassay for plasma and urine (7)(8)(9). For serum samples, 250 µL of serum was hydrolyzed overnight with 250 µL of hydrolysis reagent containing ß-glucuronidase and sulfatase. After hydrolysis, the serum samples were extracted twice with diethyl ether, after which 250 µL of assay buffer was added. [3H]Estradiol glucuronide was used to determine the recovery of the hydrolysis and extraction steps. For urine samples, 50 µL of urine was hydrolyzed overnight with 450 µL of hydrolysis reagent. After hydrolysis, the urine samples were diluted with the assay buffer. The serum, spot-urine, and 24-h urine samples were each hydrolyzed in one batch.
Enterolactone measurements were performed with the AutoDELFIA 1235 Automatic Immunoassay System (Wallac Oy). For the assay, 20 µL of sample, 100 µL of assay buffer containing the antiserum, and 100 µL of assay buffer with europium-labeled enterolactone derivative were pipetted in duplicate on anti-rabbit microtitration strips. The plates were incubated on a shaker for 90 min, followed by washing and shaking with enhancement solution for 5 min before the fluorescence measurement. The measurement was repeated if the results of the measurements deviated more than both 15% and 1.0 nmol/L. Urine samples were diluted further when the concentrations exceeded 100 nmol/L to reach the linear part of the calibration curve. The between-assay CVs determined from the quality-control samples were 7.8%, 9.8%, and 8.8% (at concentrations of 9.4, 20, and 54 nmol/L) for the serum samples and 4.0% and 5.4% (3970 and 4090 nmol/L) for the urine samples.
Urinary creatinine concentrations were determined by a colorimetric method with a modified Benedict-Behre reaction on a DCA 2000 Analyzer (Bayer).
The statistical analyses were performed by SPSS 10.0 for Windows (SPSS Inc.). All calculations were performed on a logarithmic scale.
Estimates of between-subject (
2B), within-subject (2W), and analytical (2A) variances as well as ICCs with 95% CIs were obtained by the SPSS Reliability command using one-way ANOVA. The maximum likelihood estimate was used, assuming equal variances. 2A was obtained by measuring the day 1 samples in two aliquots. Individual variances (2I) were calculated by the following formula: 
. Standard deviations () were estimated as square roots of the respective variances, and CVs were calculated by the formula: 
(10). The number of repeated measurements (in different weeks) needed to estimate the underlying homeostatic set point was calculated to within ± 50% with 80% confidence (11).
Correlations between the logarithms of enterolactone results for serum and urine samples were calculated with one sample from each participant and expressed as Pearson correlation coefficients with 95% CIs (12). The slope of the line was additionally calculated separately for each person from all eight samples. One-way ANOVA was used to test outliers with the Tanhame test (not assuming equal variances) as a post hoc test.
The ranges of the enterolactone measurements for each participant are presented in the Data Supplement that accompanies the online version of this Technical Brief (supplemental file Fig. 1; http://www.clinchem.org/content/vol49/issue1/). The median serum concentration and urinary excretions, the ICCs, and the numbers of samples required to estimate the underlying homeostatic set point are presented in Table 1
. The analytical (CVA), within-week (CVIW), within-month (CVIM), and biological (CVB) variation for enterolactone measurements was 26%, 67%, 68%, and 582%, respectively, for serum; 5.4%, 67%, 68%, and 725% for 24-h urines; and 4.5%, 108%, 125%, and 1041% in spot urines. The scatter plots comparing the logarithms of enterolactone serum concentration, 24-h urinary excretion, and spot-urine enterolactone:creatinine ratio are presented in Fig. 1
.
|
|
Both serum and urine samples can reflect the enterolactone status of the individual, as demonstrated by the linear correlation between enterolactone serum concentration and urinary excretion. The advantages of serum samples include the moderate CVW and a specific immunoassay method (7)(8). The problem with these serum samples was the substantial number of low concentrations; 13% of samples had enterolactone concentrations <3 nmol/L. Because of the high CVAs for two of the lowest samples (108% and 82%), the overall CVA for serum was quite fairly high (26%). After exclusion of these two samples, the CVA for serum was 7.7%.
Enterolactone excretion in 24-h urine demonstrated a low CVA and an ICC comparable to that of serum. The problem with urine samples is that the immunoassay for urine enterolactone gives 30% higher results than those obtained with gas chromatography–mass spectrometry (9). In addition, collecting 24-h urine is more difficult than drawing blood, and it is always questionable whether all urine has been collected.
The spot-urine enterolactone:creatinine ratio demonstrated a lower reliability than did serum and 24-h urine. Limitations in interpreting urine results adjusted to creatinine exist and may explain the wider variation (13):
The ICC for enterolactone measurements may vary between populations, but in an American study in which plasma enterolactone concentrations were measured in two samples collected on subsequent days, the correlation coefficient (r = 0.84) (2) was comparable to ours. Thus, the short-term reliability of serum enterolactone seems equal in American and Finnish populations. The long-term reliability of serum enterolactone measurements over a 2-year period was lower in one American population (ICC = 0.55) (14). Whether this is true in the Finnish population is unknown. The weakness of the present study was that the participants represented a homogeneous group with similar age and lifestyle. Applying these results to the whole Finnish population should be done with caution.
This study may be the first to examine short-term variation in enterolactone in urine. The long-term reproducibility of the overnight urine enterolactone:creatinine ratio has been studied in a Dutch population (15). The reproducibility of four samples collected with a lag time of 1–4.5 years was even poorer than in the present study, with correlation coefficients varying between 0.27 and 0.58.
Enterolactone measurement had a considerable individual variation (CVI) even within a period as short as 1 week. The CVI may be explained by errors in sample collection and storage or by changes in factors explaining differences in enterolactone concentrations between individuals: diet, smoking, intestinal transit time, use of antibacterial drugs, and intake of coffee and alcohol (1)(2)(16). The present study protocol excluded individuals currently or recently undergoing antibacterial treatment. Because the content of plant lignans in food items is known for only two of the several recognized enterolactone precursors (17), we could not test whether fluctuation in plant lignan intake explained the CVI. Alcohol and coffee intake showed no correlation with enterolactone measurements (data not shown).
In conclusion, considerable short-term within-subject variation existed in serum and urine enterolactone measurements. Because estimation of the homeostatic set point within ± 50% with 80% confidence requires three serum or 24-h urine samples, but 10 spot-urine samples, epidemiologic studies should measure enterolactone in several samples rather than only one.
Acknowledgments
We express our warmest thanks to Jouko Verho for helpful statistical advice and to Adile Samaledtin and Inga Wiik for skillful technical help. This study was supported by a Finnish Cultural Foundation Year 2000 grant and a grant from the Finska Läkarsällskapet (Finnish Swedish Medical Society).
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  A. Kilkkinen, I. Erlund, M. J. Virtanen, G. Alfthan, K. Ariniemi, and J. Virtamo Serum Enterolactone Concentration and the Risk of Coronary Heart Disease in a Case-Cohort Study of Finnish Male Smokers Am. J. Epidemiol., April 15, 2006; 163(8): 687 - 693. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Kuijsten, I. C. W. Arts, P. van't Veer, and P. C. H. Hollman The Relative Bioavailability of Enterolignans in Humans Is Enhanced by Milling and Crushing of Flaxseed J. Nutr., December 1, 2005; 135(12): 2812 - 2816. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Stuedal, I. T. Gram, Y. Bremnes, H. Adlercreutz, M. B. Veierod, and G. Ursin Plasma Levels of Enterolactone and Percentage Mammographic Density among Postmenopausal Women Cancer Epidemiol. Biomarkers Prev., September 1, 2005; 14(9): 2154 - 2159. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Bhakta, I. dos Santos Silva, C. Higgins, L. Sevak, T. Kassam-Khamis, P. Mangtani, H. Adlercreutz, and A. McMichael A Semiquantitative Food Frequency Questionnaire Is a Valid Indicator of the Usual Intake of Phytoestrogens by South Asian Women in the UK Relative to Multiple 24-h Dietary Recalls and Multiple Plasma Samples J. Nutr., January 1, 2005; 135(1): 116 - 123. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Hausner, N. F. Johnsen, J. Hallund, and I. Tetens A Single Measurement Is Inadequate to Estimate Enterolactone Levels in Danish Postmenopausal Women Due to Large Intraindividual Variation J. Nutr., May 1, 2004; 134(5): 1197 - 1200. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Kilkkinen, J. Virtamo, M. J. Virtanen, H. Adlercreutz, D. Albanes, and P. Pietinen Serum Enterolactone Concentration Is Not Associated with Prostate Cancer Risk in a Nested Case-Control Study Cancer Epidemiol. Biomarkers Prev., November 1, 2003; 12(11): 1209 - 1212. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
