![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
3
*
Dementia Research Service, Burke Medical Research Institute, White Plains, NY 10605;
ESA, Incorporated, Chelmsford, MA 01824;
**
Antioxidants Research Laboratory, Jean Mayer U.S. Department of Agriculture Human Nutrition Research Center on Aging at Tufts University, Boston, MA 02111; and
Departments of Biochemistry and Neuroscience, Cornell University Medical College, New York, NY 10021
3To whom correspondence should be addressed. E-mail: Bkristal{at}burke.org.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
KEY WORDS: • dietary restriction • serum metabolite • HPLC • multivariate analysis • biomarker • rats
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
–11
). DR extends both mean and maximum life span, reduces age-related morbidity and delays or prevents most age-associated physiologic dysfunction. DR also alters many basic physiologic processes, including metabolism, hormonal balance, and the generation of, detoxification of and resistance to reactive oxygen species (12
). The protective effect of low energy diets in animals appears mirrored in the multiple human epidemiology studies providing evidence for a detrimental effect of obesity (13
).
The multiple, robust effects of DR on physiologic and pathologic indices suggest the feasibility of identifying a metabolic serotype characteristic of this nutritional intervention. Based on this hypothesis, we are attempting to identify serum markers for AL and DR feeding in rats (14
). We are pursuing serotypes of DR (1
–10
,15
–17
) composed of low-molecular-weight molecules that are chemically redox active. Initial studies showed that HPLC separations coupled with coulometric array detectors (18
–27
) could detect 
1200 compounds in rat sera, of which 300 were analytically reliable and 240 were biologically reliable in young female rats (14
). The previous work demonstrated biological and analytical feasibility of the proposed approach by identifying components of the metabolome that could be used to monitor changes in dietary intake. That previous work represents the first stage in identifying of metabolic serotypes, i.e., method validation.
This report addresses the second stage of identifying metabolic serotypes, i.e., determining whether the metabolome retains sufficient information to allow the initial diet groups to be determined based solely on serotypes. In this report, we are concerned only with proof of the principle that the subset of serum metabolome retains sufficient information to determine the dietary group of origin. The companion paper (28
) addresses the next question, i.e., whether such subsets can determine diet groups in independent cohorts. This report demonstrates proof of principle in both female and male rats.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Details of the rats and husbandry conditions used in this study were reported previously (14
). Briefly, male and female Fischer 344 x Brown Norway F1 rats were obtained monthly from the National Institute on Aging colony at Harlan (Indianapolis, IN). The basic animal husbandry for the animals used followed NIA/NIH guidelines as implemented by Harlan. Rats were fed NIH-31 (AL rats only) or vitamin/mineral–fortified NIH-31 (DR rats only) [see (14
) for detailed diet compositions]. All rats were individually housed and DR feeding regimens (40% less food than eaten by AL rats) were implemented at 6 wk of age. Rats were imported from Harlan at 5 mo of age and killed 1 mo later. As noted in the previous report (14
), we found that rats in our colony that consumed food AL ate slightly less food than at Harlan, and restriction was therefore slightly modified. Specifically, once adjusted, food restriction was imposed at a level of 35% less than ad libitum consumption (14
). Sera were collected after killing by decapitation to avoid known differential effects of anesthesia on variables of interest. Collected blood was allowed to clot on ice for 30 min before centrifugation (1000 x g, 10 min). Female cohorts 1 and 2 had 8 AL and 8 DR rats; male cohort 1 had 5 AL and 8 DR rats; male cohort 2 had 7 AL and 8 DR rats. All animal experiments were performed under institutionally approved protocols and complied with NIH guidelines (29
).
HPLC separations and coulometric array detection was conducted essentially as described previously using an ESA CoulArray system (ESA, inc., Chelmsford, MA) (14
,18
,25
,26
). It was noted during validation studies that setting HPLC quantitation algorithms to quantitate on the basis of all matched channels (as opposed to only using those that matched on ratio) yielded a set of analytically valid peaks (metabolites) that partially, but not entirely, overlapped with the set of 297 metabolites previously identified (14
). To maximize the probability of identifying metabolites of interest (i.e., those that help distinguish AL and DR rats), we have validated an additional set of metabolites using the procedures previously defined (14
). This set of 325 metabolites was used in the initial studies of female rats presented here. Merger of the two datasets (i.e., the 297 and 325 datasets) will be conducted when the male and female datasets are combined (unpublished data). A detailed discussion of the analytical issues involved in these distinct quantitation approaches is in press (30
).
Data were analyzed using the programs CEAS 504 (ESA, Chelmsford, MA), Statview 5.0.1 (SAS Institute, Cary, NC) and Pirouette 2.7/3.0 (Infometrix, Woodinville, WA). Data analysis is described in the text.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Males.
We have demonstrated that the HPLC-based approach used could consistently identify 297 peaks in a set of pooled samples (14
). This suggests that we have a valid analytical approach to address the problem in determining serotypes. Biological variability was analyzed in a male cohort (male cohort 1). Visual inspection of the entire sera chromatograms generated from the 13 serum samples showed that the majority of the chromatograms were qualitatively well conserved, whereas some peaks of AL rats differed from those of DR rats. Coulometric array software (CEAS 504) was used to identify the 297 analytically valid metabolites in the 13 HPLC chromatograms from the male cohort. As listed in Table 1
, >87% of the 297 peaks were found in all 13 samples automatically, whereas 98% of the peaks (291 total) were found in at least 5 of 8 DR samples and 3 of 5 AL samples. This suggests that 291 metabolites exist that are potential targets for automated metabolic serotype analysis in male rats.
|
240 metabolites existed that were legitimate targets for automated metabolic serotype analysis in female rats (14
). As described in the methods, this pool of 240 metabolites was identified from a pool of 325 metabolites chosen using slightly different analytical parameters from those used in males. A detailed discussion of the analytical issues involved in these distinct quantitation approaches is in press (30
). Selection of metabolites of interest by statistical comparison.
In theory, it is possible for us to measure all (240/291) metabolites in a (female/male) sample of interest, but there are several practical considerations that argue against this approach. One practical limitation is that, at present, we continue to rely on manual confirmation of computer-driven peak identification and quantitation to improve qualitative and quantitative accuracy. A second consideration is that removing noninformative data can help classification algorithms work more efficiently. For this reason, we continually refine the focus of ongoing work to those metabolites that most effectively distinguish AL and DR rats. We begin to pare the metabolites studied by using a strategy that reduces Type II statistical errors at the expense of making an increased number of Type I statistical errors. Specifically, we chose to avoid Type II errors (believing that the serum concentration of a given metabolite does not differ between AL and DR when in reality it does) because this results in a loss of information that may be helpful in distinguishing AL and DR rats. We are willing to accept an increased frequency of Type I errors (believing that the serum concentration of a given analyte does differ between AL and DR when in reality it does not) for three major reasons. First, even moderately differing peaks can impart information to pattern-recognition algorithms through their interactions with other variables. Second, the classification algorithms available today, including HCA, PCA and pattern recognition–driven classification are all relatively efficient at mathematically identifying the "signal" (i.e., informative variables) amid the noise (i.e., noninformative variables). Third, we expect to (and do) eliminate these errors later by additional studies in future cohorts [see companion paper (28
)].
The strategy was implemented by conducting unpaired t tests (
= P 
0.2) on each of the 240/291 metabolites previously described for young female/male rats to identify metabolites of potential interest. This strategy increases Type I errors (relative to normal statistical conventions for biological research) in two ways. First, no considerations for multiple comparisons were used (e.g., ANOVA, Bonferroni correction). As a result of not adjusting for multiple comparisons, we would predict that 48 variables were determined to differ statistically between groups based solely on statistical artifacts (e.g., sampling). Second, the chosen value of = P 0.2 is much higher than that used conventionally ( = P < 0.05), resulting in a theoretical difference of 15% of the samples being studied. Because we cannot know the true distribution of the set of variables in question in the (statistical) population of AL and DR rats, however, it is impossible to predict accurately the quantity of Type I errors that this manipulation of introduces.
Two-tail, unpaired t tests were conducted on each of the 240/291 metabolites. This t test analysis identified 112 and 94 metabolites for male and female samples, respectively, that differed between AL and DR samples with value smaller than 0.2 after outlier data were deleted from the dataset by visual inspection (Table 2
). Visual inspection of the dataset of female samples used for the t tests led us to include seven additional metabolites that failed the t test for clearly artifactual reasons (e.g., insufficient automated recognition of the peak in one of the groups). Overall, 101 metabolites for female and 112 for male were regarded as the primary dataset to identify dietary groups.
|
; 62 of the 101 female metabolites had P-values < 0.05. The levels of these metabolites in AL and DR male rats were thus significantly different according to more commonly used statistical approaches. HCA with the primary subsets of metabolites distinguishes AL and DR rats.
To determine whether the metabolites identified as potentially interesting retained sufficient information to identify their dietary group of origin, we examined these datasets using HCA. HCA is a method of data analysis that emphasizes the natural groupings of the dataset. In contrast to analytical methods that emphasize distinguishing differences between two groups, HCA uses algorithms that reduce complex datasets to establish these groups without preconceived divisions. In dendrograms, such as those shown in Figure 1
and described in more detail below, relative similarity within the total study population increases as one moves from right (0.0) to left (1.0, biochemical identity) on the horizontal axis. The smaller the distance is from identity (left side) to the point at which two samples (i.e., individual rats in Fig. 1
) are linked by a vertical line, the greater the relatedness of the two samples (i.e., individual rats in Fig. 1
). Alternatively stated, the closer the split between two samples is to the right of the figure, the greater the disparity between two samples or groups of samples.
|
In general, preprocessing is used to help (mathematically) reduce the possibility that choices of arbitrary measurement scales and magnitudes can become a major determinant of the effect of a specific variable on classification. Alternatively stated, and of more direct relevance for the studies conducted, preprocessing attempts to ensure that the relative magnitude of one variable to another is not as significant as are the relative concentration differences within a variable (and between groups). We chose two different preprocessing options. In Autoscale preprocessing, data for each metabolite is mean-centered and then variance scaled. Because of its wide applicability and its effectiveness at standardizing metabolites, we expect that the majority of our future studies will focus on Autoscale data. In Range Scale preprocessing, which is commonly used for graphing, the highest value in the dataset is assigned a value of 1 and the lowest value is assigned a value of 0. Remaining values are determined by proportion. This method is highly sensitive to outliers and has limited utility in subsequent cohorts because the overall range in a subsequent cohort may change; it is useful, however, for initial data exploration. More complete mathematical descriptions of the algorithms chosen are presented in the Pirouette manual.
Clusters can be built using either agglomerative or divisive techniques. Pirouette uses the agglomerative approach, in which samples begin as their own cluster, and the clusters are then progressively grouped on the basis of a predefined distance measure. We chose one representative of each basic approach available to build clusters. Single Link, or nearest neighbor analysis, assigns a sample to the cluster of the single sample which is nearest. Complete Link, or farthest neighbor analysis, assigns a sample to the cluster whose farthest neighbor is closest. Centroid Link, or centroidal linkage analysis, assigns a sample to the cluster whose center is nearest. More complete mathematical descriptions of the algorithms chosen are presented in the Pirouette manual.
As shown in Figure 1
, cluster analysis was 100% successful at distinguishing the diet group of all 16 rats in the female cohort, regardless of the mathematical parameters used.
For male rats, we initially worked with 103 of the 112 metabolites that were identified in either 12 or 13 chromatograms from the 13 rats in the cohort. The results from HCA analysis with the 103-metabolite dataset showed that 5 AL and 8 DR rats were grouped by their dietary origin with 92% accuracy when single, centroid and incremental were used for grouping algorithms, whereas they could be distinguished with only 62% accuracy when complete was used for grouping algorithm (Fig. 2
). The results suggest that "noise" metabolites that were expected on the basis of statistical and analytical considerations in the dataset obtained by automated analysis existed in the dataset and affected the efficiency of HCA to group the samples. To increase the efficiency of the classification algorithm in grouping the rats with the dataset, "noise" metabolites were identified and removed from the dataset (see below).
|
To further evaluate the ability of the metabolites being studied to encode sufficient information to identify their group of origin, we examined this dataset using PCA. PCA, also called eigenvector analysis, is used to determine linear combinations of original metabolites that account for maximal variation. Thus, PCA can be used to reduce the dimensionality of the data by using only some of the eigenvectors. Lower number principal components possess greater ability to explain variation in the dataset, i.e., the ability of principal component 1 to explain variation is greater than that of principal component 2. For our purposes (i.e., classification), the subset of eigenvectors chosen can then be evaluated in terms of their ability to distinguish members of the different groups. In the current report, this is done on the data set from which the original variables were chosen, thereby examining feasibility. In the accompanying report (28
), we test against subsequent, independent cohorts. More important, in the context of our long-range goals, PCA can be used to determine which of the multiple compounds that may differ between AL and DR rats are the most useful for classification purposes.
PCA analysis with female rats.
The first question that we addressed by PCA analysis was, "How many mathematical factors are required to define serotype?" Alternatively stated, "How few mathematical factors, which are themselves linear combinations of the data from the 80 variables (i.e., ‘principal components’ or ‘components’), can replace the majority of the data from the variables?" Figure 3
shows that, for Autoscaled data, 50% of the total variability present can be captured by a single mathematical factor (principal component), and that 5–6 principal components are sufficient to describe 80% of the variability present. For Range-scaled data, the capture is even better, describing 80% with a single mathematical factor. Modeling power analysis (Fig. 4
) reveals that PCA analysis of automated data captures >50% of the variation of over one half the analytes using either Range or Autoscaled data (mean modeling power ± SD = 54 ± 16% [Autoscaled] and 52 ± 16% [Range-scaled]; median modeling power 56% [Autoscale] and 52% [Range-scale]). Overall, the data presented in Figures 3
and 4
show that we are able to capture a good deal of the overall information content of the data using PCA. One aim is to test whether PCA-based approaches offer the potential to distinguish dietary groups of origin. Figure 5
shows that AL and DR samples were completely distinguished by three principal components.
|
|
|
). Figure 6
B shows the individual samples present in the dataset as discriminated by the three primary principal components. Note that the 5 AL and 8 DR rats were distinguished by PCA analysis using these three components. These results indicate that PCA-based approaches can be used to identify the group of origin. The first two principal components captured 50% of the total variability present, and 6 components captured 80% of the variability present (Fig. 6
C). Modeling power analysis (Fig. 6
D) revealed that six components were sufficient to capture 40% (mean 42 ± 25%; median 45%) of the variation of the 103 analytes.
|
PCA and HCA analyses after reducing "noise" metabolites in the dataset by manual inspection.
The studies described above were carried out on metabolite sets that had been evaluated only by automated analysis. To reduce noise in the dataset, we returned to a more detailed study of the female 101 and male 112 peaks initially described above. This secondary analysis addressed two issues: 1) Are the peaks analytically valid? Specifically, are both peak identification and peak quantitation reliable? 2) Is the peak found consistently in other cohorts? This second criterion is important due to the long-range goals of our study to identify serums markers for DR.
The analytical validity of each peak was examined individually. Of the female 101 peaks marked for future study, 18 did not appear sufficiently analytically reliable to warrant continued evaluation. These peaks included, for example, shoulders on other peaks that appeared subject to artifacts in analysis, small peaks in a series of equivalent peaks in which peak identification could not be confirmed, peaks with questionable chromatographic reproducibility and peaks that were misidentified by automated peak identification and analysis software. Each of the 83 peaks that remained in the analysis was then examined qualitatively by inspection in a second cohort. To do this analysis, chromatograms of the 8 female AL rats in the initial cohort were compared with chromatograms of the 8 AL female rats in the next cohort that entered our animal facility. Despite being brought into the colony only 1 mo apart, we found that 20 peaks differed substantially (
5-fold) in the two female cohorts. Many of these 20 peaks were, in fact, apparently absent in the sera from the second cohort. These 20 peaks were removed from the analysis. Note that this analysis was meant to screen out those analytes that displayed major cohort-cohort instability; other analytes also varied by smaller amounts but were retained in the dataset. Thus, 63 female metabolites were defined for further study. Similarly, 76 metabolites were selected from the 112 metabolites for male samples6
. The P-values for the 63 female and 76 male metabolites are shown in Table 3
. In these subsets, 47 of 63 (female) and 44 of 76 (male) metabolites were significantly different between AL and DR rats. HCA and PCA were then conducted on the basis of the new sets of metabolites.
|
Males.
HCA grouped the 13 male samples with grouping algorithm incremental at 92% accuracy, and with single, centroid and complete at 85% accuracy by using the new dataset of 76 metabolites. These results again suggest that the set of metabolites selected for male rats was not as robust as that for female rats in grouping dietary origin. To further improve the metabolite set in male rats, HCA was carried out with metabolite sets that included different numbers of metabolites (from 4 to 76, with an increment of one metabolite at a time) with grouping algorithm single, centroid, complete, and incremental. The metabolites that prevented 100% separation of the dietary groups in males were temporarily excluded from the test set. It was found that 11 metabolites interfered with HCA separations when the incremental grouping algorithm was used. Fifteen metabolites, including these 11, interfered with HCA-based separations using the complete grouping algorithm. Twenty-four metabolites, including the 15 noted above, interfered with HCA separations using the centroid grouping algorithm. The 52 remaining metabolites (76 original metabolites minus 24 interfering metabolites) also enabled HCA to distinguish dietary groups using the single grouping algorithm; 34 of the 52 metabolites were with P < 0.05 between the levels of AL and DR samples (Table 4)
.
|
). PCA results obtained using the 52-metabolite dataset were also improved compared with the 103-metabolite dataset (compare Figs. 8
and 6
). Analysis again showed that there were no outliers in the 13 samples analyzed (Fig. 8
A). Figure 8
B shows that 5 AL and 8 DR rats could be distinguished by the first three principal components. The first principal component captured 48% of the total variability present compared with 42% in the 112-metabolite dataset, and 5 components were able to capture 80% of the variability present, compared with 6 principal components in the 112-metabolite dataset (Fig. 8
C). Analysis of modeling power revealed improvement over the larger dataset (Fig. 8
D, mean 49 ± 19%; median 51%).
|
|
The data presented in this report provide proof of principle, i.e., these analyses indicate that there exists a group of metabolites in rat sera that retain sufficient information to distinguish dietary groups of origin by classification algorithms. As an initial step in validation, the dataset identified will be tested in an independent cohort [see companion paper (28
)]).
Although it has been known for approximately 70 y that the reduction of food intake promotes longevity and reduces age-related morbidity, our understanding of this model and ability to apply this knowledge has remained frustratingly limited. For example, basic questions such as, "What is the mechanism of DR?" remain unanswered, although many theories exist. Identification of metabolic serotypes may help address this question by giving insight into those components of the metabolome that truly distinguish AL and DR. Another open question concerns the applicability of the DR paradigm to higher animals. Although nonhuman primate studies are in progress (e.g., 31
–33
), it is difficult to determine how restricted these animals are compared with the classical rodent model. Once our serotype models are complete, we can address this question directly from the metabolic standpoint.
The serotypes being developed may also contribute to studies in the area of human epidemiology. For example, grouping individuals by body mass index clearly identifies the increasing risk associated with obesity, but is limited by the possibility that two people of equal height and weight may have fundamentally different metabolisms (13
). The serotypes being developed may contribute to a better understanding of the roles of obesity/metabolism in disease by allowing us to categorize individuals biochemically in these studies. Similarly, questionnaire-based epidemiologic studies have other potential problems, including problems with recall and/or fluctuations in intake. Our long-term study aims, again, at allowing metabolic distinctions to be used to reduce the signal-to-noise ratio in such studies. We consider these only a possible list of the applications of these serotypes.
In summary, the major finding of this paper is that specific components of the serum metabolome can retain sufficient information to distinguish AL and DR rats. Note that, in this report, the cohort used to identify the markers of interest was also the cohort that was distinguished by the grouping analysis. Furthermore, because the selection scheme utilized is designed to make type I statistical errors, we know that this dataset includes multiple false positives. Because of this, we have not yet addressed the similarities and differences between the metabolites that enabled us to group male and female rats. In the next report, we begin to validate the datasets by showing that the serotypes developed in this and the previous report (14
) can classify independent cohorts of male and female rats.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Supported by National Institutes of Health National Institute on Aging R01-AG15354 (B.S.K.), ESA, Incorporated and the Winifred Masterson Burke Relief Foundation.
4 Abbreviations used: AL, ad libitum; DR, dietary restricted; HCA, hierarchical cluster analysis; PCA, principal component analysis.
5 This decision was a practical one based on the requirement in Pirouette for all datum cells to have a value. The choice of limiting analysis to those compounds that had been identified in either 15 or 16 chromatograms enabled us to readily use Pirouettes "Fill" function to place a relatively neutral "assigned value" into a single blank cell without creating too much noise. This neutral value can then be replaced later with the true value by manual inspection and manual oversight-based computer quantitation. The choice of using 80 of the 101 initial variables reflected the number of variables in which we had either 15 or 16 values after automated analysis.
6 Although a 103-variable dataset was used for the above HCA and PCA analyses, we inspected all 112 variables to avoid losing useful variables.
Manuscript received 9 June 2001. Initial review completed 25 July 2001. Revision accepted 11 February 2002.
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
1. Kristal, B. S. & Yu, B. P. (1994) Aging and its modulation by dietary restriction. Yu, B. P. eds. Modulation of Aging Processes by Dietary Restriction 1994:1-36 CRC Press Boca Raton, FL. .
2. Weindruch, R. & Walford, R. (1988) The Retardation of Aging and Disease by Dietary Restriction 1988 Charles C. Thomas St. Louis, MO. .
3. McCay, C. M. (1935) Cellulose in the diets of rats and mice. J. Nutr. :435-447.
4.
Maeda, H., Gleister, C. A., Masoro, E. J., Murata, I., McMahan, C. A. & Yu, B. P. (1985) Nutritional influences on aging of Fischer 344 rats: II Pathology. J. Gerontol. 40:671-688.
5. Carlson, A. J. & Hoelzel, F. (1946) Apparent prolongation of the life span of rats by intermittent fasting. J. Nutr. 31:363-375.
6. Goodrick, C. L., Ingram, D. K., Reynolds, M. A., Freeman, J. R. & Cider, N. (1990) Effects of intermittent feeding upon body weight and lifespan in inbred mice: interaction of genotype and age. Mech. Age. Dev. :55.
7. Ross, M. H. & Bras, G. (1970) Food preference and length of life. Science (Washington, DC) 190:165-167.
8. Weindruch, R. H. & Walford, R. L. (2000) Dietary restriction in mice beginning at one year of age: effects on life-span and spontaneous cancer incidence. Science (Washington, DC) 215:1415-1418.
9. Nolen, G. A. (1972) Effect of various restricted dietary regimens on the growth, health, and longevity of albino rats. J.Nutr. 102:1477-1494.
10.
Tannenbaum, A. (1945) The dependence of tumour formation on the composition of the calorie-restricted diet as well as on the degree of restriction. Cancer Res. 5:616-625.
11. Stucklikova, E., Juricova-Horakova, E. M. & Deyl, Z. (1975) New aspects of the dietary effect of life prolongation in rodents: what is the role of obesity in aging. Exp. Gerontol. 10:141-144.[Medline]
12. Yu, B. P. (1996) Aging and oxidative stress: modulation by dietary restriction. Free Radic. Biol. Med. 21:651-668.[Medline]
13.
Willett, W. C., Dietz, W. H. & Colditz, G. A. (1999) Guidelines for healthy weight. N. Engl. J. Med. 341:427-434.
14.
Vigneau-Callahan, K. E., Shestopalov, A. I., Milbury, P. E., Matson, W. R. & Kristal, B. S. (2001) Characterization of diet-dependent metabolic serotypes: analytical and biological variability issues in rats. J. Nutr. 131:924S-932S.
15. Iwasaki, K., Gleister, C. A., Masoro, E. J., McMahan, C. A. & Yu, B. P. (1988) The influence of dietary protein source on longevity and age-related disease processes of Fischer rats. J. Gerontol. 43:B5-B12.[Abstract]
16. Iwasaki, K., Gleister, C. A., Masoro, E. J., McMahan, C. A., Seo, E.-J. & Yu, B. P. (1988) Influence of the restriction of individual dietary components on longevity and age-related disease of Fischer rats: the fat component and the mineral component. J. Gerontol. 43:B13-B21.[Abstract]
17. McCay, C. M., Crowell, M. F. & Maynard, L. A. (1935) The effect of retarded growth upon the length of lifespan and upon the ultimate body size. J. Nutr. 10:63-79.
18. Kristal, B. S., Vigneau-Callahan, K. E. & Matson, W. R. (1998) Simultaneous analysis of the majority of low-molecular-weight, redox-active compounds from mitochondria. Anal. Biochem. 263:18-25.[Medline]
19. Beal, M. F., Matson, W. R., Swartz, K. J., Gamache, P. H. & Bird, E. D. (1990) Kynurenine pathway measurements in Huntington’s disease striatum; evidence for reduced formation of kynurenic acid. J. Neurochem. 55:1327-1339.[Medline]
20. Matson, W. R., Gamache, P. H., Beal, M. F. & Bird, E. D. (1987) EC array sensor concepts and data. Life Sci. 41:905-908.[Medline]
21. Matson, W. R., Bouckoms, A., Svendson, C., Beal, M. F. & Bird, E. D. (1990) Generating and controlling multiparameter databases for biochemical correlates of disorders. Basic, Clinical and Therapeutic Aspects of Alzheimer’s and Parkinson’s Diseases II:513-516 Plenum New York, NY. .
22.
LeWitt, P. A., Galloway, M. P., Matson, W. R., Milbury, P. M., McDermott, M. & Srivastava, D. K. (1992) Markers of dopamine metabolism in Parkinson’s disease: The Parkinson’s Study Group. Neurology 42:2111-2117.
23.
Ogawa, T., Matson, W. R., Beal, M. F., Myers, R. H., Bird, E. D., Milbury, P. E. & Saso, S. (1992) Kynurenine pathway abnormalities in Parkinson’s disease. Neurology 42:1702-1706.
24. Beal, M. F., Matson, W. R., Storey, E., Milbury, P. E., Ryan, E. A., Ogawa, T. & Bird, E. D. (1992) Kynurenic acid concentrations are reduced in Huntington’s disease cerebral cortex. J. Neurol. Sci 108:80-87.[Medline]
25.
Matson, W. R., Langials, P., Volicer, L., Gamache, P. H., Bird, E. D. & Mark, K. A. (1984) n-Electrode three dimensional liquid chromatography with electrochemical detection for determination of neurotransmitters. Clin. Chem. 30:1477-1488.
26. Milbury, P. E. (1997) CEAS generation of large multiparameter databases for determining categorical process involvement of biomolecules. Coulometric Array Detectors for HPLC 1997:125-141 VSP International Science Publication Zeist, The Netherlands. .
27. Acworth, I. N. & Gamache, P. H. (1996) The coulometric electrode array for use in HPLC analysis, part 1: theory. Am. Lab. 5:33-37.
28.
Shi, H., Vigneau-Callahan, K. E., Shestopalov, A. I., Milbury, P. E., Matson, W. R. & Kristal, B. S. (2002) Characterization of diet-dependent metabolic serotypes: III. Primary validation of male and female serotypes in independent cohorts of rats. J. Nutr. 132:1039-1041.
29. National Research Council (1985) Guide to the Care and Use of Laboratory Animals. Publication no. 85–23 (rev.) 1985 National Institutes of Health Bethesda, MD. .
30. Shi, H., Vigneau-Callahan, K., Matson, W. R. & Kristal, B. S. (2002) Attention to relative response across sequential electrodes improves quantitation of coulometric array. Anal. Biochem. 302:239-245.[Medline]
31.
Hansen, B. C., Bodkin, N. L. & Ortmeyer, H. K. (1999) Calorie restriction in nonhuman primates; mechanisms of reduced morbidity and mortality. Toxicol. Sci. 52:56-60.
32. Roth, G. S., Ingram, D. K., Black, A. & Lane, M. A. (2000) Effects of reduced energy intake on the biology of aging: the primate model. Eur. J. Clin. Nutr. 54:S15-S20.
33. Ramsey, J. J., Colman, R. J., Binkley, N. C., Christensen, J. D., Gresl, T. A., Kemnitz, J. W. & Weindruch, R. (2000) Dietary restriction and aging in rhesus monkeys; the University of Wisconsin Study. Exp. Gerontol. 35:1131-1149.[Medline]
This article has been cited by other articles:
|
|
 |
|
  J. A. Milner Incorporating Basic Nutrition Science into Health Interventions for Cancer Prevention J. Nutr., November 1, 2003; 133(11): 3820S - 3826. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. S. Kristal and U. Paolucci Caloric Restriction in trans Sci. Aging Knowl. Environ., July 9, 2003; 2003(27): pe19 - 19. [Abstract] [Full Text]
|
|
|
|
|
|
H. Shi, K. E. Vigneau-Callahan, A. I. Shestopalov, P. E. Milbury, W. R. Matson, and B. S. Kristal Characterization of Diet-Dependent Metabolic Serotypes: Primary Validation of Male and Female Serotypes in Independent Cohorts of Rats J. Nutr., May 1, 2002; 132(5): 1039 - 1046. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
