Relation of Glycohemoglobin and Adiposity to Atherosclerosis in Youth
Abstract In a cooperative multicenter study (Pathobiological Determinants of Atherosclerosis in Youth, PDAY) of 1532 young persons 15 through 34 years of age who died of external causes and were autopsied in medical examiners’ laboratories, we quantified atherosclerosis of the aorta and the right coronary artery. We analyzed postmortem blood cells for glycohemoglobin and postmortem serum for lipoprotein cholesterol and thiocyanate (as an indicator for smoking). We measured the thickness of the panniculus adiposus and the body mass index (weight per height squared) as indicators of adiposity. Glycohemoglobin levels exceeding 8% were associated with substantially more extensive fatty streaks and raised lesions in the right coronary artery in persons more than 25 years of age and with more extensive raised lesions in the aorta in persons more than 30 years of age. Both thickness of the panniculus adiposus and body mass index were associated with more extensive fatty streaks and raised lesions in the right coronary artery. The associations of atherosclerotic lesions with glycohemoglobin and adiposity were not explained by a less favorable lipoprotein profile or smoking. The results show that atherosclerosis in young adults is associated with the prediabetic or early diabetic state, as indicated by elevated glycohemoglobin levels, and with obesity.
Reprint requests to Jack P. Strong, MD, Department of Pathology, Louisiana State University Medical Center, 1901 Perdido St, New Orleans, LA 70112-1393.
- Received August 28, 1994.
- Accepted January 20, 1995.
The predisposition of persons with diabetes mellitus to atherosclerosis1 and to its clinical sequelae has long been known.2 Diabetics have an excess of dyslipoproteinemia and hypertension, but these do not account for the twofold to fourfold increased risk of atherosclerotic disease among diabetics.3 Subjects with impaired glucose tolerance and hyperglycemia but without clinically manifest diabetes have a higher risk of atherosclerotic disease.4
A new method of detecting elevated blood glucose emerged when electrophoretic variants of hemoglobin5 were found to result from the covalent binding of glucose to reactive amino acid groups,6 forming the complex called glycohemoglobin. The glycohemoglobin concentration is now recognized as a marker for the blood glucose concentration during the 2 to 3 months before it is measured7 and may indicate a preclinical stage of diabetes.8
Obesity is well known to predispose to diabetes.9 Obesity has been linked independently to atherosclerotic disease, but the relation is variable and inconsistent.10 11 12 Clarification of this relation, particularly in young persons, would be helpful in evaluating the health risk associated with obesity.
A multicenter cooperative study of atherosclerosis in 15- to 34-year-old autopsied young persons from whom arterial tissues and postmortem blood samples were collected provided the opportunity to assess the association of hyperglycemia, as indicated by glycohemoglobin levels, and adiposity with the extent and severity of atherosclerosis in its early stages.
Fourteen cooperating centers adopted a Standard Operating Protocol and Manual of Procedures for collecting specimens and information and submitting them to central laboratories for analysis. A statistical coordinating center received all data pertaining to each case from the collection centers and central laboratories.
Study subjects were 15- to 34-year-old individuals who died of external causes (accidents, homicides, or suicides) within 72 hours after injury and were autopsied within 48 hours after death in one of the cooperating medical examiners’ laboratories. Age and race were obtained from death certificates. Persons of race other than black or white and those with congenital heart disease, Down’s syndrome, acquired immunodeficiency syndrome, or hepatitis were excluded.
From a total of 1692 cases collected between June 1, 1987, and August 31, 1990, we excluded 160 cases because they did not meet these inclusion criteria or because of incorrect sampling or incomplete information. The Institutional Review Board of each participating center approved the use of tissue, blood, and data from the human subjects in this study.
Dissection and Preservation of Arteries
An autopsy technician removed the aorta from a point 2 cm proximal to the ligamentum arteriosum to a point 2 cm distal from the iliac bifurcation. Branching arteries were severed close to the aortic wall, and adventitial fat was removed by sharp dissection. The Pathobiological Determinants of Atherosclerosis in Youth (PDAY) technician opened the aorta along a line on the dorsal surface midway between the orifices of the intercostal and lumbar arteries, rinsed the intimal surface with Hanks’ modified balanced salt solution, and flattened it with the adventitial surface downward. The PDAY technician then bisected the aorta longitudinally along a line on the ventral surface and midway between the intercostal and lumbar ostia; prepared the right half for other studies, including histochemical and chemical analyses; and placed the left half on a piece of cardboard with the adventitia downward. This left half was covered with absorbent cotton and fixed in 10% neutral buffered formalin in a flat pan for 48 hours.
The PDAY technician opened the right coronary artery from its origin to the point at which it turned downward along the posterior interventricular sulcus with blunt-point microdissecting scissors, dissected it from the heart, removed the epicardial fat, and fixed it in the same manner as the aorta. The other main branches of the coronary artery system were prepared for other studies.
The collection centers placed each aorta and coronary artery in a plastic bag and shipped accumulated tissues to the central laboratory each month. The central laboratory stained the arteries with Sudan IV and packaged each artery with its identification number in a transparent plastic bag with a slight excess of 10% formalin.
The body was weighed in the units commonly used by the local medical examiner or coroner before organs, liquid, or other material was removed from the body. Weight was recorded to the nearest 0.5 kg or to the nearest 1 lb. A record was made when an amputation or other operative procedure that might alter weight appreciably was present.
Cadaver length, from the vertex of the cranium to the base of the heel, was measured in units commonly used by the local medical examiner or coroner. The measuring instrument was laid parallel to the body, which was in a supine position with the inferior extremities extended. Measurements were recorded to the nearest 1 cm or 0.5 in.
The autopsy technician measured subcutaneous fat, including the subcutaneous tissue from the inner edge of the rectus sheath, to the nearest millimeter at a point halfway between the xyphoid process and the umbilicus. For statistical analysis, cases were classified into categories of panniculus thickness corresponding to quartiles of the entire sample.
Body Mass Index
We computed body mass index (BMI) as weight (in kilograms) divided by height (in meters) squared. Cases were classified into categories of BMI <25, BMI=25 to 30, and BMI >30 kg/m2.13
Grading Arterial Specimens
Pathologists, blinded to demographic, clinical, or pathological observations and collection site, evaluated the right coronary arteries and left halves of the aortas. They visually estimated the extent of intimal surface involved with fatty streaks, fibrous plaques, complicated lesions, and calcified lesions by procedures developed in the International Atherosclerosis Project.14 A fatty streak was a flat or slightly elevated intimal lesion stained by Sudan IV and without other underlying changes. A fibrous plaque was a firm, elevated intimal lesion, sometimes partially or completely covered by sudanophilic deposits. A complicated lesion was a plaque with hemorrhage, thrombosis, or ulceration. A calcified lesion was an area in which calcium was detectable, either visually or by palpation, without overlying hemorrhage, ulceration, or thrombus. The sum of the percentages of surface involved with fibrous plaques, complicated lesions, and calcified lesions by gross visual grading was designated “raised lesions.” Consensus grading of the lesions was the average of independent gradings by the three pathologists. Intraobserver variability was assessed by repeated independent gradings of coded specimens randomly interspersed among new specimens. Agreement among observers was reported previously.15
The PDAY pathologists experienced in evaluating atherosclerosis regarded 5% or more of the intimal surface area involved with raised lesions as biologically significant for this age group. We applied this cut point to both fatty streaks and raised lesions. The prevalence of 5% or more surface area involvement was based on the consensus grading of lesions.
Blood collected at autopsy from the aorta, heart, or vena cava was centrifuged, and frozen serum and cells were shipped to the central laboratory for analyses.
Glycohemoglobin was measured by affinity column chromatography (Helena Laboratories) after a sample of thawed cell hemolysate was mixed with hemolysate reagent to ensure complete lysis. The column was an insoluble cellulose resin bound to dihydroxyboryl groups with an affinity for cis diol groups present in glucose. This method separates all the glycated hemoglobins from the nonglycated hemoglobins and is not interfered with by labile glycated hemoglobin.16 The coefficient of variation for blind duplicate analyses was 7.0%. Of the 1532 cases, postmortem red blood cells were available from 1335 cases. Values of 8% glycated hemoglobin and above were defined as elevated.17
Serum cholesterol and HDL cholesterol (HDL-C), after precipitation of other lipoproteins by heparin/MnCl2, were measured by the cholesterol oxidase method.18 The non–HDL-C concentration or the VLDL plus LDL cholesterol (VLDL+LDL-C) was obtained by subtraction. The coefficient of variation for blind duplicate analyses of serum cholesterol was 1.3%; for HDL cholesterol, 5.2%. Several studies demonstrated that postmortem levels of serum cholesterol and lipoproteins were representative of premortem levels.19 20 21 22 However, because emergency medical technology teams often administer large quantities of intravenous fluids to some individuals immediately before death resulting from violent causes, we excluded all serum values from the statistical analysis when serum cholesterol was less than 100 mg/dL.
We measured the color produced by the thiocyanate–ferric nitrate complex after treatment of trichloroacetic acid filtrates of serum with ferric nitrate.23 The coefficient of variation for blind duplicate analyses was 5.5%. A smoker was defined as having a serum thiocyanate level equal to or greater than 90 μmol/L.
The effects of glycohemoglobin, BMI, thickness of the panniculus adiposus, sex, race, and 5-year age group on percent intimal surface area involved with lesions were analyzed by ANOVA.24 The linear model included the main effects and two-factor interactions. We report only the effects that are related to the variables of interest in this report and do not report findings for the effects of sex, race, and age, which were included in the model. We applied a logit transformation to the proportion of surface area involved with lesions.25 A small constant (0.001) was added to avoid the logarithm of zero. We applied a logarithmic transformation to the serum lipoprotein concentrations. The transformations made the data better satisfy the assumptions underlying the statistical analysis. To determine whether the observed associations of lesions and glycohemoglobin, BMI, or panniculus might be due to other risk factor variables, the covariates VLDL+LDL-C, HDL-C, and smoking status were added to the model. The prevalence of cases having 5% or more of the intimal surface involved with lesions was analyzed with logistic regression.26 Convergence problems in the logistic regression maximum-likelihood procedure necessitated a model simpler than that used for the ANOVA. Preliminary investigation indicated that a main-effects model was adequate. Tests of hypotheses used the likelihood ratio test.
Table 1⇓ shows the mean levels of glycohemoglobin, BMI, and thickness of panniculus adiposus by sex, race, and age. Blacks have higher glycohemoglobin levels than whites, and males tend to have higher glycohemoglobin levels than females. Glycohemoglobin changes little with age, and there are no interactions among sex, race, and age. There is no increase with age in prevalence of glycohemoglobin levels equal to or greater than 8% (results not shown).
White males have greater BMIs than black males, but white females have smaller BMIs than black females, resulting in a sex-by-race interaction. BMI increases with age in whites but not in blacks, resulting in a race-by-age interaction.
White males have a thicker panniculus than black males, but white females and black females have about the same panniculus thickness, producing a sex-by-race interaction. The panniculus increases with age in all sex and race groups except black females, among whom there is little or no increase with age.
Correlations of Glycohemoglobin With Adiposity
Glycohemoglobin is positively but weakly correlated with BMI (partial correlation coefficient adjusted for sex, race, and age, .084; P=.0023) but not with the panniculus (partial correlation coefficient, .029; P=.3025). BMI and panniculus are correlated with each another (partial correlation coefficient, .578; P=.0001).
Association of Glycohemoglobin and Atherosclerosis
Table 2⇓ compares the extent of atherosclerotic lesions in the aorta and right coronary artery for cases with glycohemoglobin less than 8% and those with glycohemoglobin equal to or greater than 8%. Elevated glycohemoglobin is substantially and significantly associated with more extensive raised lesions in both segments of the aorta and in the right coronary artery and is associated with more extensive fatty streaks in the right coronary.
There is a sex-by-glycohemoglobin interaction for raised lesions in the thoracic aorta, with greater involvement among females with glycohemoglobin greater than or equal to 8% compared with males. No other interactions involve sex and race. There is, however, an interaction of glycohemoglobin with age: glycohemoglobin is associated with more extensive raised lesions in the older age groups (particularly the 30- to-34-year-old group) in both aortic segments and with more extensive fatty streaks and raised lesions in the right coronary arteries of persons over 25 years of age. There is no effect of glycohemoglobin at levels less than 8% on any type of lesion in the aorta or right coronary artery (results not shown).
Individual Cases With Elevated Glycohemoglobin
Because the number of cases with glycohemoglobin levels greater than or equal to 8% was small, we matched the 32 subjects with glycohemoglobin greater than or equal to 8% by age, race, sex, and percent surface area involved with lesions to 32 subjects with glycohemoglobin less than 7% to determine whether qualitative differences in the atherosclerotic lesions existed between the two groups. Fifteen variables that might differentiate between the two groups were selected. After three pathologists reviewed the pairs of specimens, 6 variables were selected for further study: intensity and diffuseness of Sudan staining, confluence of lesions, demarcation of lesions, distal distribution of lesions, and diffuse intimal thickening. Each of three pathologists independently evaluated the specimens. None of these 6 gross characteristics identified the high-glycohemoglobin cases.
Association of Adiposity and Atherosclerosis
There is no consistent effect of BMI on either fatty streaks or raised lesions in the aorta except for a weak positive association with fatty streaks in the thoracic aorta (Table 3⇓). BMI shows a strong association with the extent of fatty streaks and raised lesions in the right coronary artery of males but not in females (Table 4⇓).
To ensure that the associations reported in Table 4⇑ were not due to the presence of adolescents and very young adults in our sample, we analyzed the association of BMI and right coronary artery lesions using only individuals 25 years of age and older. Although significance levels were larger because of the smaller number of cases, results similar to those reported in Table 4⇑ were obtained (results not shown).
Panniculus thickness is associated with more extensive fatty streaks and raised lesions in the right coronary artery, but there is no interaction with sex (Table 5⇓).
We used the two measures of adiposity together as an indication of central (abdominal) obesity. Because we found an association of adiposity only with coronary artery lesions and an association of BMI only with coronary artery lesions in males, we analyzed the combined association of BMI and panniculus thickness with lesions only in males. We compared atherosclerosis between two BMI classes within each of two classes of panniculus thickness. We used two classes of BMI because there is little effect of BMI below 30 (Table 4⇑) and two classes of panniculus to simplify the presentation. Table 6⇓ shows that BMI above 30 is associated with increased right coronary artery fatty streaks regardless of panniculus thickness, although the increase is greater with greater panniculus thickness. BMI exceeding 30 is associated with increased right coronary artery raised lesions only within the higher classification of panniculus thickness.
Glycohemoglobin and Adiposity
Beyond the variance accounted for by sex, race, and age, glycohemoglobin explains 5.38% (P=.0001) of the variance in thoracic aorta raised lesions, 1.54% (P=.0004) of the variance in abdominal aorta raised lesions, 1.45% (P=.0022) of the variance in right coronary artery fatty streaks, and 1.70% (P=.0005) of the variance in right coronary artery raised lesions. Adiposity (both BMI and panniculus thickness considered simultaneously) explains 5.60% (P=.0001) of the variance in right coronary artery fatty streaks and 4.21% (P=.0085) of the variance in right coronary artery raised lesions.
The variance explained by glycohemoglobin is similar after adiposity is accounted for (thoracic aorta raised lesions, 4.81% [P=.0001]; abdominal aorta raised lesions, 1.47% [P=.0005]; right coronary fatty streaks, 1.15% [P=.0095]; and right coronary raised lesions, 1.37% [P=.0032]). Likewise, the variance explained by adiposity is similar after glycohemoglobin is accounted for (right coronary fatty streaks, 5.31% [P=.0001]; right coronary raised lesions, 3.88% [P=.0205]). These findings suggest that the association of elevated glycohemoglobin with lesions is not accounted for by adiposity and likewise that the association of adiposity with lesions is not accounted for by elevated glycohemoglobin.
Prevalence of Raised Lesions Greater Than 5%
Elevated glycohemoglobin and both measures of obesity are associated with a higher prevalence of fatty streaks and raised lesions in the right coronary artery; higher glycohemoglobin levels are associated with a higher prevalence of raised lesions in the thoracic aorta (Table 7⇓).
Relation of Glycohemoglobin and Adiposity to Serum Lipoproteins and Smoking
Because of hemodilution and our inability to collect blood from all individuals, the number of cases available for analyses with serum lipoprotein concentrations and smoking is smaller. VLDL+LDL-C is associated directly with glycohemoglobin, BMI, and thickness of panniculus. Although the relation is not statistically significant, HDL-C is inversely associated with the same variables (Table 8⇓). Smokers have a slightly lower BMI than nonsmokers (24.80 versus 24.14; P=.0586) and tend to have a slightly thinner panniculus (20.66 versus 19.30 mm, P=.1369).
Adjustment for Serum Lipoproteins and Smoking
When lesion values are adjusted for VLDL+LDL-C, HDL-C, smoking, sex, race, and age, the significance of the effect of glycohemoglobin on atherosclerosis is diminished, probably because the number of cases is reduced (Table 9⇓). Where the association with glycohemoglobin is significant, the pattern of differences between cases with glycohemoglobin less than 8% and cases with glycohemoglobin greater than or equal to 8% is essentially the same as in Table 2⇑. The same is true for the effects of BMI (Table 10⇓) and panniculus thickness (Table 11⇓) on lesions in the right coronary artery.
Variance Explained by Glycohemoglobin and Adiposity
Table 12⇓ shows the estimated variance as percent of total variance in percent intimal surface area involved with lesions in the right coronary artery explained by glycohemoglobin, BMI, or panniculus thickness after adjustment for sex, race, age, VLDL+LDL-C, HDL-C, and smoking status. The percent variance explained is first given when all available cases are used in the analysis. Next, the percent variance explained is given only for those cases for which we have measurements of VLDL+LDL-C, HDL-C, and thiocyanate; however, no adjustment is made for these additional risk factor variables. Comparison of lesions in the subset with those in all cases indicates the effects of using only a subset of the cases defined by the availability of acceptable serum samples. The significance levels change considerably owing to the reduced number of cases, but the variance explained is similar to that obtained when all cases are used. Finally, the percent variance explained by glycohemoglobin, BMI, or panniculus thickness after adjustment for VLDL+LDL-C, HDL-C, and smoking status is presented. This similarity indicates that the associations observed between lesions and glycohemoglobin, BMI, or panniculus thickness (Tables 2 through 5⇑⇑⇑⇑) are not explained by the association of glycohemoglobin, BMI, or panniculus thickness with a less favorable lipoprotein profile or smoking.
Summary of Results
Elevated glycohemoglobin levels are associated with more extensive and more advanced atherosclerosis in the aorta and right coronary artery of persons between 25 and 34 years of age. The effect is not accounted for by serum lipoprotein cholesterol levels, smoking, or adiposity. Adiposity, as measured by either BMI or thickness of the panniculus adiposus, is associated with more extensive and more advanced atherosclerosis of the right coronary artery in persons 15 to 34 years of age. The association of lesions with adiposity is not explained by serum lipoprotein cholesterol levels or smoking.
Significance of Elevated Glycohemoglobin Levels
Hemodilution resulting from the administration of blood, plasma, or other fluids to injured persons before death is a problem in the interpretation of results of serum cholesterol levels.18 27 28 However, hemodilution (or hemoconcentration) does not interfere with obtaining accurate measurements for glycohemoglobin. A glycohemoglobin concentration equal to or exceeding 8% (by the analytical method used in this study) indicates a mean blood glucose concentration equal to or greater than about 150 mg/dL for the preceding 2 to 3 months17 and may precede or accompany the early preclinical and presymptomatic stages of diabetes.8
Although small, the variance explained by glycohemoglobin beyond that explained by other variables such as sex, race, and age exceeds that expected by chance, as indicated by the small significance levels. The small increment in R2 associated with elevated glycohemoglobin levels (Table 12⇑), which contrasts with the large effect on affected individuals (Table 2⇑), is due to the low frequency of elevated glycohemoglobin in the population.
Several recent articles reviewed the potential mechanisms by which diabetes, hyperinsulinemia, and hyperglycemia augment atherogenesis.29 30 31 Of the many mechanisms suggested, two seem most likely to be involved in the association of glycohemoglobin concentration in these young adults: the effects of dyslipoproteinemia and hyperinsulinemia in the prediabetic state, as suggested by Haffner et al,32 and a direct effect of glycosylation of proteins on atherogenesis.33
Prediabetic Dyslipoproteinemia and Hyperinsulinemia
The glycohemoglobin effect is not accounted for by serum lipoprotein cholesterol levels (VLDL+LDL-C and HDL-C), and there is little correlation between glycohemoglobin levels and serum lipoprotein levels. However, the analyses of glycohemoglobin and serum lipoprotein levels are limited to about half the total number of cases (approximately 800) because serum was not available or the data were excluded because of hemodilution. Furthermore, the limitations of a single determination of serum cholesterol or lipoprotein cholesterol being representative of the level over a longer period are well known.34 Hemodilution, which does not occur in all cases, introduces additional variability. These sources of error in the measurement of serum cholesterol are expected to degrade the association of serum cholesterol with lesions.35 In a nonfasting sample, it was not possible to measure triglycerides and thereby reliably estimate the true LDL-C level, nor was it feasible to measure plasma insulin levels under these conditions. Therefore, the possibility remains that individuals with elevated glycohemoglobin levels may have had abnormal lipoprotein profiles and hyperinsulinemia.
Advanced Glycosylation End Products
A process similar to the glycosylation of hemoglobin occurs in other proteins. Subsequently, the carbohydrate-protein complex undergoes chemical rearrangement to form irreversible advanced glycosylation end products, which have a variety of deleterious effects on cells and tissues.33 Control of hyperglycemia would reduce the intensity and duration of exposure of tissue proteins to blood glucose. There is hope for prevention of the deleterious effects of glycosylation of proteins by agents such as aminoguanidine.29
Obesity and Atherosclerosis
The health effects of obesity have been difficult to study because (1) body weight and composition are influenced by many different conditions (eg, caloric intake, physical activity, smoking, and genetic factors), (2) its definition is not precise, (3) fat distribution may be as important as total fat, (4) duration of exposure and age of the subject influence its effects, and (5) obesity is associated with a variety of health risks (hypertension, coronary heart disease, stroke, non–insulin-dependent diabetes, cholelithiasis, and some forms of cancer).10 11 12 Obesity enhances three other established risk factors for coronary heart disease—hypertension, dyslipoproteinemia, and diabetes mellitus—and is inversely related to another major risk factor, smoking. In many instances, when obesity is associated with coronary heart disease in univariate analyses, multivariate analyses including the other risk factors do not show an independent effect, but several long-term longitudinal studies have found an independent effect after controlling for other risk factors.36 37 38
Few studies have examined the association of obesity with atherosclerosis measured in the arteries of autopsied persons, and the results have been inconsistent or inconclusive. The relation is obscured or confounded in most natural deaths because body mass and adipose tissue are often affected by terminal illnesses and because many deaths are due to atherosclerosis and its complications. A large international survey of atherosclerosis in autopsied persons showed no association, even among accidental deaths in which there would be the least effect of terminal illness.39 However, in a subset of 672 men from New Orleans, La, for which smoking status was ascertained, there was a weak but significant association of thickness of panniculus adiposus with coronary artery but not aortic atherosclerosis after exclusion of individuals dying of diseases related to atherosclerosis and after adjustment for smoking.40 In another 1108 New Orleans males dying of causes unrelated to atherosclerosis but whose smoking status was unknown, there were significant correlations of the ratio of body weight to body length and ponderal index (body length divided by the cube root of body weight) with coronary artery raised lesions.41
The present results, which are derived from accidental deaths and were adjusted (in a subset of cases) for smoking status and lipoprotein cholesterol concentrations, confirm and extend those observations by demonstrating a positive and significant association of obesity with coronary atherosclerosis in young persons. Although direct measurements of body fat distribution were not made, results from the combination of two measurements of adiposity as an indicator of central obesity are consistent with the closer association of central obesity with coronary heart disease that was observed in a number of epidemiological studies of living persons.42 43 44 45
The association of obesity with atherosclerosis among young persons, who were the focus of this study, is particularly significant for primary prevention programs in light of recent reports that obesity has greater predictive power for coronary heart disease after longer follow-up periods37 38 46 and that the prevalence of obesity is increasing in the United States.47
The observations reported here suggest that both elevated glycohemoglobin levels, possibly associated with the prediabetic state, and obesity are associated with accelerated atherogenesis in the third and fourth decades of life. The results provide hope that early detection and control of obesity and hyperglycemia in young persons will reduce the risk of atherosclerotic disease in later life.
PDAY Research Group
The investigators cooperating in the multicenter PDAY study and the grants from the National Heart, Lung and Blood Institute supporting this research are listed below.
Robert W. Wissler (director), PhD, MD, University of Chicago, and Abel L. Robertson, Jr (associate director), MD, PhD, University of Illinois.
J. Fredrick Cornhill, DPhil, Ohio State University; Henry C. McGill, Jr, MD, Southwest Foundation for Biomedical Research; C. Alex McMahan, PhD, University of Texas Health Science Center at San Antonio; Abel L. Robertson, Jr, MD, PhD, University of Illinois; Jack P. Strong, MD, Louisiana State University Medical Center; Robert W. Wissler, PhD, MD, University of Chicago.
Standard Operating Protocol and Manual of Procedures Committee Chair
Margaret C. Oalmann, DrPH, Louisiana State University Medical Center.
University of Alabama, Birmingham, Department of Medicine: Steffen Gay, MD (principal investigator); Renate E. Gay, MD; and Guoquiang Huang, MD (HL-33733). Department of Biochemistry: Edward J. Miller, PhD (principal investigator); Donald K. Furuto, PhD; Margaret S. Vail; and Annie J. Narkates (HL-33728).
Albany Medical College, NY: Assad Daoud, MD (principal investigator); Adriene S. Frank, PhD; Mary A. Hyer; and E. Carol McGovern (HL-33765).
Baylor College of Medicine, Houston, Tex: Louis C. Smith, PhD (principal investigator), and Faith M. Strickland, PhD (HL-33750).
University of Chicago, Ill: Robert W. Wissler, PhD, MD (principal investigator); Dragoslava Vessellinovitch, DVM, MS; Akio Komatsu, MD, PhD; Yoshiaki Kusumi, MD; Gregory M. Culen, DPM; Alyna Chien, BA; Alexis Demopoulos, BA; Gertrud Friedman, BA; R. Timothy Bridenstein, MS; Robert J. Stein, MD; Robert H. Kirschner, MD; Manuela Bekermeier, ASCP; Blanche Berger, ASCP; and Laura Hiltscher, ASCP (HL-33740).
University of Illinois, Chicago: Abel L. Robertson, Jr, MD, PhD (principal investigator); Robert J. Stein, MD; Edmund R. Donoghue, MD; Robert J. Buschmann, PhD; Yoshihisa Katsura, MD; Tae Lyong An, MD; Eupil Choi, MD; Nancy Jones, MD; Mitra S. Kalelkar, MD; Yuksel Konakci, MD; Barry Lifschultz, MD; V. Ramana Gumidyala, MD; Rose M. Harper, BS; and Francis Norris, HTL (ASCP) (HL-33758).
Louisiana State University Medical Center, New Orleans: Jack P. Strong, MD (principal investigator); Gray T. Malcom, PhD; William P. Newman III, MD; Margaret C. Oalmann, DrPH; Paul S. Roheim, MD; Ashim K. Bhattacharyya, PhD; Miguel A. Guzman, PhD; Ali A. Hatem, MD; Conrad A. Hornick, PhD; Carlos D. Restrepo, MD; Richard E. Tracy, MD, PhD; Cecilia C. Breaux, MS; Stephanie E. Hubbard; Cynthia S. Zsembik; DeAnne G. Gibbs; and Dana A. Troxclair (HL-33746).
University of Maryland, Baltimore: Wolfgang Mergner, MD, PhD (principal investigator); James H. Resau, PhD; Robert D. Vigorito, MS, PA; Q-C Yu, MD; and J. Smialek, MD (HL-33752).
Medical College of Georgia, Augusta: A. Bleakley Chandler, MD, and Raghunatha N. Rao, MD (co–principal investigators); D. Greer Falls, MD, Ross G. Gerrity, PhD, and Benjamin O. Spurlock, BA (coinvestigators); Kalish B. Sharma, MD, and Joel S. Sexton, MD (associate investigators); and K.K. Smith, HT(ASCP), and G.W. Forbes (research assistants) (HL-33772).
University of Nebraska Medical Center, Omaha: Bruce M. McManus, MD, PhD (principal investigator); Jerry W. Jones, MD; Todd J. Kendall, MS; Jerrold A. Remmenga, BS; and William C. Rogler, BS (HL-33778).
Ohio State University, Columbus: J. Fredrick Cornhill, DPhil (principal investigator); William R. Adrion, MD; Patrick M. Fardel, MD; Brian Gara, MS; Edward Herderick; John Meimer, MS; and Larry R. Tate, MD (HL-33760).
Southwest Foundation for Biomedical Research, San Antonio, Tex: James E. Hixson, PhD (principal investigator), and Patricia K. Powers (HL-39913).
University of Texas Health Science Center at San Antonio: C. Alex McMahan, PhD (principal investigator); George M. Barnwell, PhD (deceased); Henry C. McGill, Jr, MD; Yolan Marinez, MA; Thomas J. Prihoda, PhD; and Herman S. Wigodsky, MD, PhD (HL-33749).
Vanderbilt University, Nashville, Tenn: Renu Virmani, MD (principal investigator); James B. Atkinson, MD, PhD; Charles W. Hartland, MD; Linda Gleaves, RA; Crystal Gleaves, HT; and Manik Paul, RA (HL-33770).
West Virginia University Health Sciences Center, Morgantown: Singanallur N. Jagannathan, PhD (principal investigator); Bruce Caterson, PhD; James Frost, MD; K. Murali K. Rao, MD; Syamala Jagannathan; Peggy Johnson; and Nathaniel F. Rodman, MD (HL-33748).
Kunkel HG, Wallenius G. New hemoglobin in normal adult blood. Science. 1955;122:288.
Goldstein DE, Little RR, Wiedmeyer H-M, England JD, McKenzie EM. Glycated hemoglobin: methodologies and clinical applications. Clin Chem. 1986;32:B64-B70.
Mann GV. The influence of obesity on health. N Engl J Med. 1974;291:178-185, 226-232.
Barrett-Connor EL. Obesity, atherosclerosis, and coronary artery disease. Ann Intern Med. 1985;103:1010-1019.
Bray GA. An approach to the classification and evaluation of obesity. In: Bjorntorp P, Brodoff BN, eds. Obesity. Philadelphia, Pa: JB Lippincott Co; 1992:294-308.
Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Natural history of aortic and coronary atherosclerotic lesions in youth: findings from the PDAY study. Arterioscler Thromb. 1993;13:1291-1298.
Klenk DC, Hermanson GT, Krohn RI, Fujimoto EK, Mallia AK, Smith PK, England JD, Wiedmeyer H-M, Little RR, Goldstein DE. Determination of glycosylated hemoglobin by affinity chromatography: comparison with colorimetric and ion-exchange methods, and effects of common interferences. Clin Chem. 1982;28:2088-2094.
Santiago JV. Lessons from the Diabetes Control and Complications Trial. Diabetes. 1993;42:1549-1554.
Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Relationship of atherosclerosis in young men to serum lipoprotein cholesterol concentrations and smoking: a preliminary report from the Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. JAMA. 1990;264:3018-3024.
Glanville JN. Post-mortem serum cholesterol levels. Br Med J. 1960;2:1852-1853.
Enticknap JB. Lipids in cadaver sera after fatal heart attacks. J Clin Pathol. 1961;14:496-499.
Bowler RG. The determination of thiocyanate in blood serum. Biochem J. 1944;38:385-388.
Draper NR, Smith H. Applied Regression Analysis. New York, NY: John Wiley & Sons Inc; 1966:58-134.
Carroll RJ, Ruppert D. Transformation and Weighting in Regression. New York, NY: Chapman and Hall; 1988:9-51.
Hosmer DW Jr, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley & Sons Inc; 1989:1-36.
Strong JP, Oalmann MC, Malcom GT, Newman WP II, McMahan CA, for the Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Research Group. Pathobiological Determinants of Atherosclerosis in Youth (PDAY): rationale, methodology, and selected risk factor findings. Cardiovasc Risk Factors. 1992;2:22-30.
Bierman EL. Atherogenesis in diabetes. Arterioscler Thromb. 1992;12:647-656.
Schwartz CJ, Valente AJ, Sprague EA, Kelley JL, Cayatte AJ, Rozek MM. Pathogenesis of the atherosclerotic lesion: implications for diabetes mellitus. Diabetes Care. 1992;15:1156-1167.
McGill HC Jr, McMahan CA, Wene JD. Unresolved problems in the diet-heart issue. Arteriosclerosis. 1981;1:164-176.
Stern MP, Haffner SM. Body fat distribution and hyperinsulinemia as risk factors for diabetes and cardiovascular disease. Arteriosclerosis. 1986;6:123-130.
Thompson CJ, Ryu JE, Craven TE, Kahl FR, Crouse JR III. Central adipose distribution is related to coronary atherosclerosis. Arterioscler Thromb. 1991;11:327-333.
Casassus P, Fontbonne A, Thibult N, Ducimetière P, Richard JL, Claude J-R, Warnet J-M, Rosselin G, Eschwège E. Upper-body fat distribution: a hyperinsulinemia-independent predictor of coronary heart disease mortality: the Paris Prospective Study. Arterioscler Thromb. 1992;12:1387-1392.
Garrison RJ, Castelli WP. Weight and thirty-year mortality of men in the Framingham Study. Ann Intern Med. 1985;103:1006-1009.