| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
EDITORIAL |
1 From the Departments of Nutrition (RMvD and FBH) and Epidemiology (FBH), Harvard School of Public Health, Boston MA, and the Channing Laboratory (RMvD and FBH), Department of Medicine, Brigham and Women's Hospital, and Harvard Medical School, Boston MA
See corresponding article on page 832.
2 Reprints not available. Address correspondence to RM van Dam, Department of Nutrition, Harvard School of Public Health, Building 2, 665 Huntington Avenue, Boston MA 02115. E-mail: rvandam{at}hsph.harvard.edu.
Whole grains are a major source of fiber, vitamins, minerals, lignans, and other phenolic compounds and are commonly defined as grains in which the bran and germ components are present in their natural proportions. Epidemiologic evidence for beneficial effects of whole grain consumption on risk of chronic diseases, including cardiovascular diseases (1) and type 2 diabetes (2), has been accumulating. All of these studies are based on self-reported whole grain intake. In this issue of the Journal, Landberg et al (3) evaluated plasma alkylresorcinol (AR) concentrations as potential biomarkers of whole grain intake. ARs are phenolic lipids that are found in substantial amounts in the bran fraction of rye and wheat, but in few other foods (4). About half (45–71%) of AR is absorbed (5), and AR can be measured in plasma.
Several issues contribute to measurement error in the assessment of self-reported whole grain intake. There has been discussion about what constitutes whole grains; it can also be difficult for consumers to recognize whole grains. Some measurement error is inevitable because people may under- or overestimate food intakes, and within-individual variation in intakes can lead to misclassification of long-term intake. An additional consideration is that not all consumed whole-grain foods are listed in food-frequency questionnaires. For these reasons, it would be desirable to develop biomarkers for whole grain intake.
Biomarkers for whole grain intake can be used as the reference method in validation studies of assessments based on self-reports, because the measurement errors affecting the 2 methods can be expected to be independent. This information can be used to correct estimates of associations between whole grain intake and disease risk for measurement error (6). In addition, accurate biomarkers of whole grain intake can be used to assess compliance in whole-grain feeding studies and can directly be studied in relation to disease risk in epidemiologic studies.
For application as biomarkers in epidemiologic and validation studies, evidence is required that plasma AR concentrations reflect variation in whole grain intake in free-living populations. What does the study by Landberg et al tell us about the usefulness of plasma AR for these applications? Consistent with previous feeding trials (7, 8), plasma AR concentrations were substantially greater after the whole-grain intervention than after the refined-grain intervention, which showed that plasma AR concentrations are sensitive to large increases in whole wheat and rye intakes. However, this does not necessarily imply that plasma ARs capture variation in whole grain intake in epidemiologic studies. First, the study included only 30 individuals who were willing to participate in the trial and who therefore probably substantially differed from a more general population. Indeed, the baseline AR intake of participants was twice as high as that estimated in the Swedish population, which in turn has a much higher intake than does the UK population (9). Second, the authors combined data from different intervention periods, which greatly increased the variation in whole grain intake and led to a substantially higher correlation (0.58) between AR intake and plasma concentrations than during periods in which diet is self-selected (0.33 and 0.40). The latter correlations can be expected to better reflect the performance of plasma AR as a biomarker in epidemiologic studies. Third, given that plasma ARs are proposed as biomarkers of whole grain intake rather than biomarkers of AR intake, correlations with habitual whole grain intake would be most relevant. Landberg et al used cereal fiber intake from wheat and rye as a measure of whole grain intake and observed a correlation of 0.47 for the combined intervention periods; they did not separately report this correlation for the periods during which self-selected diets were consumed. Therefore, it remains important to evaluate the agreement between AR concentrations and whole grain intakes in population-based studies. The main sources of ARs include the bran fractions of wheat and rye, whereas barley contains small amounts, and refined grains and rice and oats contain negligible amounts (4). Therefore, plasma AR will be of limited value as a biomarker of whole grain intake in populations in whom intake of other grain species, such as brown rice or oats, is high.
Because the half-life of plasma AR is 
5 h (10), plasma concentrations reflect intake over the last couple of days. For the etiology of chronic diseases, long-term exposure over several years is likely to be most relevant. Plasma AR concentrations only reflect long-term intake in populations with stable whole grain consumption habits, but this condition may well be met given that major grain sources, such as bread and breakfast cereals, are typically eaten daily. Evaluation of long-term reproducibility in repeatedly collected samples in population-based studies can be used to assess whether plasma AR reflects intake over a sufficiently long period. An advantage of food-frequency questionnaire assessment of whole grain intake is that a long-term consumption pattern is assessed.
Whole grains contain lignans, and the plasma or urine concentration of enterolignans (enterodiol and enterolactone) can therefore also reflect whole grain intake. Higher whole grain intake has been associated with higher enterolignan concentrations (8, 11). However, associations of whole grain intake with enterolignans have been weaker than for AR, which is expected because vegetables, tea, and coffee are also important sources of lignans, and enterolignan production by intestinal microflora can be affected by factors such as antibiotic use (11). Therefore, measurement of enterolignans is of interest because they may affect health through their antioxidant or phytoestrogen properties, but their specificity as biomarkers of whole grains is limited.
In summary, measurement of plasma AR concentrations may contribute to information about the accuracy of methods based on self-report and, thus, may be useful in better estimating measurement error and the actual magnitude of effects of whole grain intake on disease risk. Plasma ARs may also be directly studied in relation to disease risk in epidemiologic studies. However, more information on accuracy in population-based samples and long-term reproducibility is needed. Furthermore, it should be noted that studies of biomarkers that reflect dietary intakes can still be confounded by the same risk factors that are associated with dietary intakes. Despite these considerations, the availability of a potentially accurate and specific biomarker for whole grain intake is an exciting opportunity for improving our insight into the health effects of this important dietary factor.
ACKNOWLEDGMENTS
Neither author reported a conflict of interest.
REFERENCES
Related articles in AJCN:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
