- Email: [email protected]
Risk Factors and Atherosclerosis in Youth Autopsy Findings of the Bogalusa Heart Study
Risk Factors and Atherosclerosis in Youth Autopsy Findings of the Bogalusa Heart Study RICHARD E. TRACY, MD,* WILLIAM P. NEWMAN III, MD,* WENDY A. WATTIGNEY, MS,t GERALD S. BERENSON, MOt
ABSTRACT: The Collaborative Pathology Study is one of the most impressive programs of the Bogalusa Heart Study. Attempts are made to obtain complete and uniform necropsy coverage of all deceased young people who may have been examined in the Bogalusa Heart Study. Since 1978, autopsy specimens have been collected from 190 deaths, representing 65% of all known deaths in the study age category. The relation of antemortem risk factors for cardiovascular disease to early atherosclerotic lesions in the aorta and coronary arteries was assessed in those individuals previously examined in the Bogalusa Heart Study (N = 59). Aortic fatty streaks were strongly related to both total and low-density lipoprotein (LDL) cholesterol (r = 0.62< P < 0.0001 for each association), and were inversely correlated with the ratio of high-density lipoprotein (HDL) cholesterol to LDL plus very-Iowdensity lipoprotein (VLDL) cholesterol (r = -0.29, P < 0.01). Coronary artery fatty streaks were associated with elevated total cholesterol, LDL cholesterol, VLDL cholesterol, and systolic blood pressure. Higher levels of LDL and VLDL cholesterol, triglycerides, systolic and diastolic blood pressure, and a lower ratio of HDL to LDL plus VLDL were found in those people with coronary artery fibrous plaques. Microscopy offered additional information about the characteristics of the aortic and coronary arterial intimal disease. Histologic observations have confirmed some of the relationships indicated wi.th gross observations and show the complexity of this disease process. These findings emphasize the importance of an approach to preventive cardiology early in life. KEY INDEXING TERMS: From the *Department of Pathology, Louisiana State University Medical Center, and tTulone National Center for Cardiovasculor Health, Tulane School of Public Health & Tropical Medicine, New Orleans, Louisiana. Supported by funds from the National Heart, Lung, and Blood in· stitute of the U.S. Public Health Service (USPHS), Early Natural History of Arteriosclerosis 5ROI HL38844. Correspondence: Gerald S. Berenson, MD, Tulane Center for Car· diovascular Health, Tulane School of Public Health & Tropical Med· icine, 1501 Canal Street, 14th Floor, New Orleans, LA 70112·2824. THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
Atherosclerosis; Coronary artery disease; Hypertension; Fatty streaks; Fibrous plaques. [Am J Med Sci 1995;310(suppl 1):S37-S41.]
P
rimary prevention of coronary heart disease should begin with children or young adults. Although the crippling and fatal complications of coronary heart disease are usually seen in middle-aged or elderly victims, the course toward these events proceeds by growth of atherosclerotic plaques in coronary arteries beginning in youth. The earliest clear beginning of the process, in the form of raised fibrous plaques, can be found in some adolescents, and these are frequently seen in young adults. 1 The fatty streaks that are often abundant in coronary arteries at very young ages may also represent early stages of plaque growth. The early development of these lesions should be targeted for preventive intervention. 2 The major coronary risk factors, abnormal levels of blood lipoproteins, elevated blood pressure, and smoking, have long been known to correlate with the exte~t of atherosclerosis at ages greater than 40 years; thIS was first shown in the coronary arteries and aorta by autopsy studies,3 and later demonstrated in the carotid arteries of living subjects.4 These correlations, however, have only recently been observed in autopsies of subjects younger than 40 years of age.5-7 One of the three studies providing such data is the Bogalusa Heart Study. Findings from autopsy studies in Bogalusa are reviewed here. Methods Population. The Bogalusa Heart Study is a more than
20-year study of cardiovascular disease risk factors in children and young adults. Risk factor information has been obtained on over 12,000 people from birth through 32 years of age in the biracial (35% black, 65% white) community of Bogalusa, Louisiana (Ward 4 of Washington Parish). Data have also been collected during pilot studies in Franklinton, Louisiana, a nearby community. Since 1978, pathologists associated with the Bogalusa Heart Study have consulted with the coroner of 537
Risk Factors and Atherosclerosis in Autopsy Findings
Washington Parish to obtain consent to autopsy any Washington Parish resident dying between the ages of 3 and 31 years. Autopsies are conducted in local funeral homes and selected tissues (coronary arteries, heart, aorta, kidneys, and blood) are sent to the Department of Pathology at Louisiana State University and the Core Lipid Laboratory of the Bogalusa Heart Study. Most deaths were from vehicular accidents, homicides, or suicides, with only 10% related to natural causes. As of February, 1994, autopsy specimens have been collected from 190 cases, representing over 65% of all eligible deaths. Of these 190 cases, 85 had antemortem screening, many on several occasions. Risk Factor Data. Trained examiners, nurses and technicians, collected data according to specific written protocols. Blood for serum lipids and lipoproteins, blood pressure, and anthropometric measurements were obtained at each examination. Procedures for obtaining risk factor data have been previously described in detail. 8 ,9 Blood pressure was measured in the right arm, with the patient in a sitting relaxed position. An average of six measurements (three replicate readings by each of two examiners) with a mercury sphygmomanometer was used in all analyses. A fourth Korotkoff phase was considered diastolic blood pressure. Gross Evaluation of Lesions. Aortas and coronary arteries were opened longitudinally, stained with Sudan IV, and intermixed with specimens from other research projects. Macrovascular (gross) assessment of the percentage of surface involvement with fatty streaks, fibrous plaques, and complicated lesions was evaluated according to protocols developed in the International Atherosclerosis Project. lO Three pathologists evaluated the vessels independently, and the mean grade was used for all analyses. Histologic Evaluation. After visual grading, photography, and digitization with a Graf/Pen (Science Accesory, South Port, CT), aortas and coronary arteries were sampled for histologic evaluation. Paraffin sections stained with hematoxylin-eosin, and frozen sections stained for fats with Oil red 0, were prepared from the thoracic aorta, abdominal aorta, and the right coronary artery. These were evaluated by previously established methods. ll ,12 Measurements were taken of average intimal and medial thicknesses, extent of specimen affected by foam cell infiltrates, numbers and densities of intimal smooth muscle cells, and average amounts of stainable lipid, both total and extracellular. Results and Discussion Gross Observations of Fatty Streaks. Table 1 shows
the relationship of selected risk factor levels to the extent of fatty streaks in the aorta and coronary arteries. 6 Fatty streaks in the aorta are more extensive in the presence of elevated serum total cholesterol and lowdensity lipoprotein (LDL) cholesterol. Coronary artery fatty streaks were associated with the presence of elevated LDL cholesterol, very-Iow-density lipoprotein (VLDL) cholesterol, and systolic blood pressure. An
538
Table 1. Correlation Coefficients Relating the Percentage of Fatty Streak Involvement to Age-Specific Risk Factor Levels: Bogalusa Heart Study, 1978-1989* Aorta
Coronary Arteries (n = 54)
(n = 59)
Unadjusted Total cholesterol (mg/dL) Triglycerides (mg/dL) Lipoprotein cholesterol (mg/dL) LDL HDL VLDL HDL/(LDL + VLDL) Blood pressure (mm Hg) Systolic Diastolic
AgeAdjusted
Unadjusted
AgeAdjusted
0.62t
0.62t
0.28
0.39:\:
0.03
0
0.30§
0.25
0.63t 0.01 0.28
0.61t 0.06 0.26
0.32§ -0.17 0.37:\:
0.44:\: -0.05 0.36§
-0.29
-0.29
-0.29§
-0.26
0.15 0.01
0.08 -0.04
0.43' 0.27§
0.42t 0.19
* Valves are Spearman correlation coefficients. t P < 0.0001. :\:P<0.01. § P < 0.05. ~ P < 0.001. LDL = low-density lipoprotein; HDL = high-density lipoprotein; VLDL = very low-density lipoprotein. (With permission from Newman et al. Ann NY Acad Sci 1991;623: 16-24.
inverse association of borderline significance was noted between the high-density lipoprotein (HDL) ratio and fatty streaks in the coronaries and aorta. At the present time, it is not clear whether nonelevated fatty streaks are in fact early precursors of advanced atherosclerosis. Nevertheless, the correlations noted with risk factors are in a direction to suggest that lower levels of blood pressure and cholesterol might retard fatty streak development in children and young adults. This result is sufficiently clear to show the importance of understanding the relationship of fatty streaks to progressive atherosclerotic disease. Gross Observations of Raised LeSions. The 16 subjects with fibrous plaques in the coronary arteries at autopsy were significantly different from the 38 subjects without fibrous plaques in several antemortem measures. Higher levels of LDL and VLDL, and a lower ratio, HDL/(LDL + VLDL), are found in those with fibrous plaques; this is consistent with the lack of significant difference in total cholesterol, because the total is the sum of contrary moving constituents. Higher blood pressure, both systolic and diastolic, was found in subjects with fibrous plaques. There is little doubt that the collagen-capped, elevated fibrous plaques are, in fact, early precursors of December 1995 Volume 310 Supplement 1
Tracy at al
progressive atherosclerosis leading to coronary heart disease. The data in Table 2 support the view that young individuals with elevated blood pressure and excessive cholesterol in the LDL and VLDL fractions may be on a faster course toward severe coronary atherosclerosis than are subjects with low levels of these risk factors. These results further strengthen the conclusions suggested by Table 1. Seeing that both fatty streaks and fibrous plaques show similar relationships with risk factors, it is easy to speculate that fatty streaks are, indeed, in some manner or other, indicators of progression toward advanced atherosclerosis. Histologic Observations. Microscopy offers additional information about the characteristics of the aortic and coronary arterial intimal disease. In histologic sections, the intima of coronary arteries is of greatly varied thickness from one specimen to another, and from one place to another within each specimen. At any degree of intimal thickening, Oil red 0 dye applied to frozen sections may show much, little, or no demonstrable lipid. Figure 1 offers some representative examples to illustrate the range of intimal thickness, increasing from below upward in the diagram, and of lipid staining, intensifying from left to right, seen in the youthful coronary arteries ofthis study. Snapshots such as these can be arranged in a variety of ways that give visual portrayal to numerous theories of how plaques may evolve. The arrangement of photographs in Figure 1 depicts a hypothesis of plaque growth that has emerged from recent quantitative and qualitative observations in a number of laboratories. 12- 16 In this theory, "adaptive thickening" of the intima occurs in response to he-
Table 2. Age-Adjusted Mean Values of Risk Factors According to the Presence or Absence of Fibrous Plaques in the Coronary Arteries: Bogalusa Heart Study, 1978-1989 Coronary Artery Fibrous Plaques Risk Factor
Absent (n = 38)
Present (n = 16)
-
148 ± 28 71.7 ± 25 76.8 61.3 8.9 0.81
± ± ± ±
21 18 5 0.3
104.1 ± 6.1 65.9 ± 5
159.5 ± 26 95.2 ± 31 90.4 55.3 13.4 0.57
± ± ± ±
21 18 7 0.3
111.9 ± 9 68.1 ± 6
PValue*
0.17
om
0.04 0.28 0.05 0.03
0.001 0.04
* Based on two·sided Wilcoxon tests. LDL = low-density lipoprotein; HDL = high-density lipoprotein; VLDL = very low-density lipoprotein. (With permission from Newman et at. Ann NY Acad Sci 1991;623: 16-24. THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
, -
~
_.~._.,~__.J •• _
.. lt~•• : . -
I
I~ -
Total cholesterol (mg/ dL) Triglycerides (mg/dL) Lipoprotein cholesterol (mg/dL) LDL HDL VLDL HDL/(LDL + VLDL) Blood pressure (mm Hg) Systolic Diastolic
.:.........- : - . :
----
I~ .
~
I I{t> ."
... '-'"
-
' . ' - '
~
Figure 1. Views of frozen sections stained for lipid with Oil red 0 and counterstained with hematoxylin are taken from coronary arteries of children in the Bogalusa Heart Study to exemplify a systematized variety of typical appearances. Rows 1, 2,3, and 4 represent increasing degrees of "adaptive intimal thickening" (ie, intimal fibroplasia). Column a represents lipid-free structures. Column b represents superficial deposits of mostly intracellular lipid. Column c exemplifies mostly extracellular lipid in deeper locations. Bar in frame 1a is 40 /Lm and applies to all frames.
modynamic conditions at some arterial sites (vertical dimension of Figure 1); lipid then readily deposits in these sites, which are especially "progression prone." Lipid generally appears only in the superficial layers of intima in company with many foam cells at "progression-resistant" sites, where adaptive thickening is
539
Risk Factors and Atherosclerosis In Autopsy Findings
1.5 'C
Q)
1;; ::J
15' oJ
'" '" n; '"
CI
1.0
en "
+
OIl
1.5
.=Male 0= Female Mull r=0.36 It = Necrotic core
It
•
e
• o
ij
~0.26
- - - - - - _e _':ema/es
o
'" :2 ~ oJ
0
•
~
----._
.0 •
UJ
o
2
Serum Total Cholesterol; Z-score
Figure 2. Extracellular lipid in the coronary intima, measured on . a five-point scale of intensity and age-adjusted by regression, is plotted against serum cholesterol age-adjusted by z-scores_ Lines are fitted by multiple regression. (Reproduced with permission from Tracy et al.")
absent (photos Ib,c,2b). At sites of progression-prone intimal thickening, pools of extracellular lipid sometimes appear in levels beneath the zone of foam cells (photos 3c,4c); this appearance is taken to identify the subgroup of fatty streaks that are progressing into fibrous plaques. The images in Figure 1 are representative of some features that were morphometrically quantified in this study for correlation with risk factor measurements. The working hypothesis embodied in the arrangement of the photographs supplies a useful background for interpreting the quantitative data to be reported next. Quantitative Findings. The histologic measures of intimallipid averaged within cases generally reflect the extent of fatty streaks seen by gross inspection. The correlations of Table 1 are therefore often repeated, as expected, in the histologic data. l7 Some additional findings, however; ar~ of further interest. The lack of significant correlation between aortic fatty streaks and blood "pressure, seen in Table 1, is confirmed for microscopic measures of intimal lipid in the thoracic segment. Significant correlations, however, were seen in the abdominal segment, suggesting that blood pressure could be of greater effect in the abdominal segment. The association of coronary fatty streaks with blood cholesterol, seen in Table 1, implies that the microscopic measures of intimal lipid should follow the same pattern. This expectation was confirmed only in the male subjects, being of no significance in the female subjects. This result is shown for extracellular lipid in Figure 2; the pattern was similar for total stainable lipid. The results for extracellular lipid are especially intriguing, because the presence of lipids in the interstitial compartment has been proposed as a marker for progression of fatty streaks into fibrous plaques (Figure 1). The association of blood pressure with coronary intimallipid content, implied by the findings with fatty S40
streaks in Table 1, was confirmed by histologic data, but again only for male subjects.1'his result, shown for extracellular lipid in Figure 3, was also seen with total stainable lipid. This outcome suggests that high blood pressure, just as with elevate(i blood cholesterol, may be associated with a progression-prone type of intimallipid deposit. Some other features of coronary and aortic histology did not relate strongly or consistently with risk factors. These include intimal thickness, smooth muscle cell numbers, and cell densities. This result is disappointing, because these features of the artery wall are all thought to change during the early stages of plaque growth. These results could indicate a linkage of blood lipids and pressure with only the lipid constituents of early plaques, and not with other cellular or connective tissue features in the initial preatheromatous conditions. Such observations show the complexity involved in understanding this disease process. Renal Studies. The histologic examination of renal tissues also offered some interesting results. The hyalinization of arterioles, a feature of the nephrosclerotic kidney that characterizes essential hypertension, varied greatly among subjects. The variation did not parallel blood pressure levels or other risk factors. Associations with coronary intimal thickness and smooth muscle cell numbers were, however, surprisingly strong. Figure 4 shows a linear association between coronary intimal thickness and renal arteriolar hyalinization, with a correlation between the age-adjusted variables of r = 0.53. Males exceeded females in hyalinization of arterioles, and this difference is proportional to the sex difference in coronary intimal thickness. Renal arteriolar hyalinization is associated with those features of the coronary intima that do not correlate with blood pressure or lipids (ie, thickness and cell numbers). Hyalinization of arterioles may therefore offer a clue to the causation of early preatherosclerotic changes that act independently of blood pressure and lipids. ls Alternatively, the hyalinization could be re-
'C
~::J
15'
~
1.0
• = Male 0= Female Mull r = 0.46 It = Necrotic core
CI
'"
n; '"
en "
,j,
It
•
• •
0.5
o
ij
:3' 0.26
It
;;; :;"! a;
0
u
'"
~
UJ
-1
o 1 Systolic Blood Pressure; Z-score
2
3
Figure 3. Extracellular lipid in the coronary intima, as in Figure 2, is plotted against systolic blood pressure age-adjusted by z-scores. Lines are fitted by multiple regression. (Reproduced with permission from Tracyet al_ 17) December 1995 Volume 310 Supplement 1
Tracy et al
..,
200
Ql
iii :::J '0
.=Male 0= Female Mult r= 0.53 It = Necrotic core
'"
References
• It
•
•
•"
Ql Cl ~ :::;J..
iii II)
cu
100
c
-" u
:£
76
"iii
.~
:s 0
·1 0 1 2 Hyalinized arterioles; ,j(N/cm ) Age adjusted
2
Figure 4. Thickness of the coronary intima is plotted against the numbers of hyalinized arterioles in the renal cortex age-adjusted by regression. The line is fitted by regression. (Reproduced with permission from Tracy et alP)
vealing variations in blood pressure that escaped detection by clinical measurement. Medial Changes in Coronary Vessels. The medial thickness of the coronary arteries also showed some interesting properties. (1) A correlation with blood pressure indicates that the coronary media can be used as a postmortem index for blood pressure. (2) Males have greater medial thickness than females. (3) Medial thickness increases with increasing renal arteriolar hyalinization. (4) the sex difference in medial thickness is proportional to the difference in arteriolar hyalinization. (5) Intimal and medial thicknesses are correlated (r = 0.42, age adjusted). As with renal arteriolar hyalinization, coronary medial thickness also appears to be an indicator for the early preatherosclerotic conditions that are, in some degree, independent of the observed risk factors. Conclusions
With the present state of knowledge about atherosclerosis, it remains uncertain whether fatty streaks are precursors of atherosclerotic lesions and to what extent modification of risk factors in children can prevent coronary atherosclerosis from progressing. The findings reviewed here suggest that lower levels of blood pressure and cholesterol are desirable in early life to retard the development of pre atherosclerotic and early atherosclerotic changes in the coronary arteries. The anatomic findings give credibility to the study of cardiovascular risk factors in children and further document the early onset of atherosclerosis and hypertension.
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
1. Strong JP. Coronary atherosclerosis in soldiers: a clue to the natural history of atherosclerosis in the young. JAMA. 1986;256: 2863-6 . 2. Strong JP. Atherosclerotic lesions: natural history, risk factors, and topography. Arch Pathol Lab Med. 1992;116:1268-75. 3. Solberg LA, Strong JP. Risk factors and atherosclerotic lesions: a review of autopsy studies. Arterioscler Tbromb. 1983;3: 187-98. 4. Bonithon-Kopp C, Scarabin RY, Taquet A, Touboul PJ, Malmejac A, Guize L. Risk factors for early carotid atherosclerosis in middle-aged French women. Arterioscler Tbromb. 1991;11:966-72. 5. Strong JP, Oalmann MC, Newman WP ill, et al. Coronary heart disease in young black and white males in New Orleans: community pathology study. Am Heart J. 1984;108:747-59. 6. Newman WP ill, Wattigney W, Berenson GS. Autopsy studies in United States children and adolescents: relationship of risk factors to atherosclerotic lesions. Ann NY Acad Sci. 1991;623:16-24. 7. PDAY (Pathobiological Determinants of Atherosclerosis in Youth) Research Group. Relationship of atherosclerosis in young men to serum lipoprotein cholesterol concentrations and smoking. JAMA. 1990;264:3018-23. 8. Berenson GS, McMahan CA, Voors AW, et al. Cardiovascular risk factors in children. In: Andrews C, Hester HE; eds. The Early Natural History of Atherosclerosis and Essential Hypertension. New York: Oxford University Press; 1980:1-453. 9. Berenson GS, ed. Causation of Cardiovascular Risk Factors in Children: Perspectives on Cardiovascular Risk in Early Life. New York: Raven Press; 1986. 10. Guzman MA, McMahan CA, McGill BC Jr, et al. Selected methodologic aspects of the International Atherosclerosis Project. Lab Invest. 1968;18:479-97. 11. Tracy RE, Kissling GE. Comparisons of human populations for histologic features of atherosclerosis. Arch Pathol Lab Med. 1989;112:1056-65. 12. Tracy RE, Kissling GE, Gandia M, Reynolds C. Spatial dispersion of stainable lipid in frozen sections of human aorta. Virchows Arch [AJ. 1989;415:39-49. 13. Stary BC, Blankenhorn DB, Chandler AB, et al. A definition of the intima of human arteries and of its atherosclerosisprone regions: a report from the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association. AHA Medical/Scientific Statement Special Report. Arterioscler Thromb. 1992;12:120-34. 14. Stary BC, Chandler AB, Glagov S, et al. A definition of initial fatty streak, and intermediate lesions of atherosclerosis: a report from the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association. AHA Medical/ Scientific Statement Special Report. Arterioscler Thromb. 1994;14:840-56. 15. Guyton JR, Klemp KF. Transitional features in human atherosclerosis: intimal thickening, cholesterol clefts, and cell loss in human aortic fatty streaks. Am J Pathol. 1993;143:1444-57. 16. Guyton JR, Klemp KF. Development of the atherosclerotic core region: chemical and ultrastructural analysis of microdissected atherosclerotic lesions from human aorta. Arterioscler Thromb. 1994;14:1305-14. 17. Tracy RE, Newman WP ill, Wattigney WA, Srinivasan SR, Strong JP, Berenson GS. Histologic features of atherosclerosis and hypertension from autopsies of young individuals in a defined geographic population: the Bogalusa Heart Study. Atherosclerosis. 1995;116:163-79. 18. Tracy RE. Blood pressure related separately to parenchymal fibrosis and vasculopathy of the kidney. Am J Kidney Dis. 1992;20:124-31.
541
ABSTRACT: The Collaborative Pathology Study is one of the most impressive programs of the Bogalusa Heart Study. Attempts are made to obtain complete and uniform necropsy coverage of all deceased young people who may have been examined in the Bogalusa Heart Study. Since 1978, autopsy specimens have been collected from 190 deaths, representing 65% of all known deaths in the study age category. The relation of antemortem risk factors for cardiovascular disease to early atherosclerotic lesions in the aorta and coronary arteries was assessed in those individuals previously examined in the Bogalusa Heart Study (N = 59). Aortic fatty streaks were strongly related to both total and low-density lipoprotein (LDL) cholesterol (r = 0.62< P < 0.0001 for each association), and were inversely correlated with the ratio of high-density lipoprotein (HDL) cholesterol to LDL plus very-Iowdensity lipoprotein (VLDL) cholesterol (r = -0.29, P < 0.01). Coronary artery fatty streaks were associated with elevated total cholesterol, LDL cholesterol, VLDL cholesterol, and systolic blood pressure. Higher levels of LDL and VLDL cholesterol, triglycerides, systolic and diastolic blood pressure, and a lower ratio of HDL to LDL plus VLDL were found in those people with coronary artery fibrous plaques. Microscopy offered additional information about the characteristics of the aortic and coronary arterial intimal disease. Histologic observations have confirmed some of the relationships indicated wi.th gross observations and show the complexity of this disease process. These findings emphasize the importance of an approach to preventive cardiology early in life. KEY INDEXING TERMS: From the *Department of Pathology, Louisiana State University Medical Center, and tTulone National Center for Cardiovasculor Health, Tulane School of Public Health & Tropical Medicine, New Orleans, Louisiana. Supported by funds from the National Heart, Lung, and Blood in· stitute of the U.S. Public Health Service (USPHS), Early Natural History of Arteriosclerosis 5ROI HL38844. Correspondence: Gerald S. Berenson, MD, Tulane Center for Car· diovascular Health, Tulane School of Public Health & Tropical Med· icine, 1501 Canal Street, 14th Floor, New Orleans, LA 70112·2824. THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
Atherosclerosis; Coronary artery disease; Hypertension; Fatty streaks; Fibrous plaques. [Am J Med Sci 1995;310(suppl 1):S37-S41.]
P
rimary prevention of coronary heart disease should begin with children or young adults. Although the crippling and fatal complications of coronary heart disease are usually seen in middle-aged or elderly victims, the course toward these events proceeds by growth of atherosclerotic plaques in coronary arteries beginning in youth. The earliest clear beginning of the process, in the form of raised fibrous plaques, can be found in some adolescents, and these are frequently seen in young adults. 1 The fatty streaks that are often abundant in coronary arteries at very young ages may also represent early stages of plaque growth. The early development of these lesions should be targeted for preventive intervention. 2 The major coronary risk factors, abnormal levels of blood lipoproteins, elevated blood pressure, and smoking, have long been known to correlate with the exte~t of atherosclerosis at ages greater than 40 years; thIS was first shown in the coronary arteries and aorta by autopsy studies,3 and later demonstrated in the carotid arteries of living subjects.4 These correlations, however, have only recently been observed in autopsies of subjects younger than 40 years of age.5-7 One of the three studies providing such data is the Bogalusa Heart Study. Findings from autopsy studies in Bogalusa are reviewed here. Methods Population. The Bogalusa Heart Study is a more than
20-year study of cardiovascular disease risk factors in children and young adults. Risk factor information has been obtained on over 12,000 people from birth through 32 years of age in the biracial (35% black, 65% white) community of Bogalusa, Louisiana (Ward 4 of Washington Parish). Data have also been collected during pilot studies in Franklinton, Louisiana, a nearby community. Since 1978, pathologists associated with the Bogalusa Heart Study have consulted with the coroner of 537
Risk Factors and Atherosclerosis in Autopsy Findings
Washington Parish to obtain consent to autopsy any Washington Parish resident dying between the ages of 3 and 31 years. Autopsies are conducted in local funeral homes and selected tissues (coronary arteries, heart, aorta, kidneys, and blood) are sent to the Department of Pathology at Louisiana State University and the Core Lipid Laboratory of the Bogalusa Heart Study. Most deaths were from vehicular accidents, homicides, or suicides, with only 10% related to natural causes. As of February, 1994, autopsy specimens have been collected from 190 cases, representing over 65% of all eligible deaths. Of these 190 cases, 85 had antemortem screening, many on several occasions. Risk Factor Data. Trained examiners, nurses and technicians, collected data according to specific written protocols. Blood for serum lipids and lipoproteins, blood pressure, and anthropometric measurements were obtained at each examination. Procedures for obtaining risk factor data have been previously described in detail. 8 ,9 Blood pressure was measured in the right arm, with the patient in a sitting relaxed position. An average of six measurements (three replicate readings by each of two examiners) with a mercury sphygmomanometer was used in all analyses. A fourth Korotkoff phase was considered diastolic blood pressure. Gross Evaluation of Lesions. Aortas and coronary arteries were opened longitudinally, stained with Sudan IV, and intermixed with specimens from other research projects. Macrovascular (gross) assessment of the percentage of surface involvement with fatty streaks, fibrous plaques, and complicated lesions was evaluated according to protocols developed in the International Atherosclerosis Project. lO Three pathologists evaluated the vessels independently, and the mean grade was used for all analyses. Histologic Evaluation. After visual grading, photography, and digitization with a Graf/Pen (Science Accesory, South Port, CT), aortas and coronary arteries were sampled for histologic evaluation. Paraffin sections stained with hematoxylin-eosin, and frozen sections stained for fats with Oil red 0, were prepared from the thoracic aorta, abdominal aorta, and the right coronary artery. These were evaluated by previously established methods. ll ,12 Measurements were taken of average intimal and medial thicknesses, extent of specimen affected by foam cell infiltrates, numbers and densities of intimal smooth muscle cells, and average amounts of stainable lipid, both total and extracellular. Results and Discussion Gross Observations of Fatty Streaks. Table 1 shows
the relationship of selected risk factor levels to the extent of fatty streaks in the aorta and coronary arteries. 6 Fatty streaks in the aorta are more extensive in the presence of elevated serum total cholesterol and lowdensity lipoprotein (LDL) cholesterol. Coronary artery fatty streaks were associated with the presence of elevated LDL cholesterol, very-Iow-density lipoprotein (VLDL) cholesterol, and systolic blood pressure. An
538
Table 1. Correlation Coefficients Relating the Percentage of Fatty Streak Involvement to Age-Specific Risk Factor Levels: Bogalusa Heart Study, 1978-1989* Aorta
Coronary Arteries (n = 54)
(n = 59)
Unadjusted Total cholesterol (mg/dL) Triglycerides (mg/dL) Lipoprotein cholesterol (mg/dL) LDL HDL VLDL HDL/(LDL + VLDL) Blood pressure (mm Hg) Systolic Diastolic
AgeAdjusted
Unadjusted
AgeAdjusted
0.62t
0.62t
0.28
0.39:\:
0.03
0
0.30§
0.25
0.63t 0.01 0.28
0.61t 0.06 0.26
0.32§ -0.17 0.37:\:
0.44:\: -0.05 0.36§
-0.29
-0.29
-0.29§
-0.26
0.15 0.01
0.08 -0.04
0.43' 0.27§
0.42t 0.19
* Valves are Spearman correlation coefficients. t P < 0.0001. :\:P<0.01. § P < 0.05. ~ P < 0.001. LDL = low-density lipoprotein; HDL = high-density lipoprotein; VLDL = very low-density lipoprotein. (With permission from Newman et al. Ann NY Acad Sci 1991;623: 16-24.
inverse association of borderline significance was noted between the high-density lipoprotein (HDL) ratio and fatty streaks in the coronaries and aorta. At the present time, it is not clear whether nonelevated fatty streaks are in fact early precursors of advanced atherosclerosis. Nevertheless, the correlations noted with risk factors are in a direction to suggest that lower levels of blood pressure and cholesterol might retard fatty streak development in children and young adults. This result is sufficiently clear to show the importance of understanding the relationship of fatty streaks to progressive atherosclerotic disease. Gross Observations of Raised LeSions. The 16 subjects with fibrous plaques in the coronary arteries at autopsy were significantly different from the 38 subjects without fibrous plaques in several antemortem measures. Higher levels of LDL and VLDL, and a lower ratio, HDL/(LDL + VLDL), are found in those with fibrous plaques; this is consistent with the lack of significant difference in total cholesterol, because the total is the sum of contrary moving constituents. Higher blood pressure, both systolic and diastolic, was found in subjects with fibrous plaques. There is little doubt that the collagen-capped, elevated fibrous plaques are, in fact, early precursors of December 1995 Volume 310 Supplement 1
Tracy at al
progressive atherosclerosis leading to coronary heart disease. The data in Table 2 support the view that young individuals with elevated blood pressure and excessive cholesterol in the LDL and VLDL fractions may be on a faster course toward severe coronary atherosclerosis than are subjects with low levels of these risk factors. These results further strengthen the conclusions suggested by Table 1. Seeing that both fatty streaks and fibrous plaques show similar relationships with risk factors, it is easy to speculate that fatty streaks are, indeed, in some manner or other, indicators of progression toward advanced atherosclerosis. Histologic Observations. Microscopy offers additional information about the characteristics of the aortic and coronary arterial intimal disease. In histologic sections, the intima of coronary arteries is of greatly varied thickness from one specimen to another, and from one place to another within each specimen. At any degree of intimal thickening, Oil red 0 dye applied to frozen sections may show much, little, or no demonstrable lipid. Figure 1 offers some representative examples to illustrate the range of intimal thickness, increasing from below upward in the diagram, and of lipid staining, intensifying from left to right, seen in the youthful coronary arteries ofthis study. Snapshots such as these can be arranged in a variety of ways that give visual portrayal to numerous theories of how plaques may evolve. The arrangement of photographs in Figure 1 depicts a hypothesis of plaque growth that has emerged from recent quantitative and qualitative observations in a number of laboratories. 12- 16 In this theory, "adaptive thickening" of the intima occurs in response to he-
Table 2. Age-Adjusted Mean Values of Risk Factors According to the Presence or Absence of Fibrous Plaques in the Coronary Arteries: Bogalusa Heart Study, 1978-1989 Coronary Artery Fibrous Plaques Risk Factor
Absent (n = 38)
Present (n = 16)
-
148 ± 28 71.7 ± 25 76.8 61.3 8.9 0.81
± ± ± ±
21 18 5 0.3
104.1 ± 6.1 65.9 ± 5
159.5 ± 26 95.2 ± 31 90.4 55.3 13.4 0.57
± ± ± ±
21 18 7 0.3
111.9 ± 9 68.1 ± 6
PValue*
0.17
om
0.04 0.28 0.05 0.03
0.001 0.04
* Based on two·sided Wilcoxon tests. LDL = low-density lipoprotein; HDL = high-density lipoprotein; VLDL = very low-density lipoprotein. (With permission from Newman et at. Ann NY Acad Sci 1991;623: 16-24. THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
, -
~
_.~._.,~__.J •• _
.. lt~•• : . -
I
I~ -
Total cholesterol (mg/ dL) Triglycerides (mg/dL) Lipoprotein cholesterol (mg/dL) LDL HDL VLDL HDL/(LDL + VLDL) Blood pressure (mm Hg) Systolic Diastolic
.:.........- : - . :
----
I~ .
~
I I{t> ."
... '-'"
-
' . ' - '
~
Figure 1. Views of frozen sections stained for lipid with Oil red 0 and counterstained with hematoxylin are taken from coronary arteries of children in the Bogalusa Heart Study to exemplify a systematized variety of typical appearances. Rows 1, 2,3, and 4 represent increasing degrees of "adaptive intimal thickening" (ie, intimal fibroplasia). Column a represents lipid-free structures. Column b represents superficial deposits of mostly intracellular lipid. Column c exemplifies mostly extracellular lipid in deeper locations. Bar in frame 1a is 40 /Lm and applies to all frames.
modynamic conditions at some arterial sites (vertical dimension of Figure 1); lipid then readily deposits in these sites, which are especially "progression prone." Lipid generally appears only in the superficial layers of intima in company with many foam cells at "progression-resistant" sites, where adaptive thickening is
539
Risk Factors and Atherosclerosis In Autopsy Findings
1.5 'C
Q)
1;; ::J
15' oJ
'" '" n; '"
CI
1.0
en "
+
OIl
1.5
.=Male 0= Female Mull r=0.36 It = Necrotic core
It
•
e
• o
ij
~0.26
- - - - - - _e _':ema/es
o
'" :2 ~ oJ
0
•
~
----._
.0 •
UJ
o
2
Serum Total Cholesterol; Z-score
Figure 2. Extracellular lipid in the coronary intima, measured on . a five-point scale of intensity and age-adjusted by regression, is plotted against serum cholesterol age-adjusted by z-scores_ Lines are fitted by multiple regression. (Reproduced with permission from Tracy et al.")
absent (photos Ib,c,2b). At sites of progression-prone intimal thickening, pools of extracellular lipid sometimes appear in levels beneath the zone of foam cells (photos 3c,4c); this appearance is taken to identify the subgroup of fatty streaks that are progressing into fibrous plaques. The images in Figure 1 are representative of some features that were morphometrically quantified in this study for correlation with risk factor measurements. The working hypothesis embodied in the arrangement of the photographs supplies a useful background for interpreting the quantitative data to be reported next. Quantitative Findings. The histologic measures of intimallipid averaged within cases generally reflect the extent of fatty streaks seen by gross inspection. The correlations of Table 1 are therefore often repeated, as expected, in the histologic data. l7 Some additional findings, however; ar~ of further interest. The lack of significant correlation between aortic fatty streaks and blood "pressure, seen in Table 1, is confirmed for microscopic measures of intimal lipid in the thoracic segment. Significant correlations, however, were seen in the abdominal segment, suggesting that blood pressure could be of greater effect in the abdominal segment. The association of coronary fatty streaks with blood cholesterol, seen in Table 1, implies that the microscopic measures of intimal lipid should follow the same pattern. This expectation was confirmed only in the male subjects, being of no significance in the female subjects. This result is shown for extracellular lipid in Figure 2; the pattern was similar for total stainable lipid. The results for extracellular lipid are especially intriguing, because the presence of lipids in the interstitial compartment has been proposed as a marker for progression of fatty streaks into fibrous plaques (Figure 1). The association of blood pressure with coronary intimallipid content, implied by the findings with fatty S40
streaks in Table 1, was confirmed by histologic data, but again only for male subjects.1'his result, shown for extracellular lipid in Figure 3, was also seen with total stainable lipid. This outcome suggests that high blood pressure, just as with elevate(i blood cholesterol, may be associated with a progression-prone type of intimallipid deposit. Some other features of coronary and aortic histology did not relate strongly or consistently with risk factors. These include intimal thickness, smooth muscle cell numbers, and cell densities. This result is disappointing, because these features of the artery wall are all thought to change during the early stages of plaque growth. These results could indicate a linkage of blood lipids and pressure with only the lipid constituents of early plaques, and not with other cellular or connective tissue features in the initial preatheromatous conditions. Such observations show the complexity involved in understanding this disease process. Renal Studies. The histologic examination of renal tissues also offered some interesting results. The hyalinization of arterioles, a feature of the nephrosclerotic kidney that characterizes essential hypertension, varied greatly among subjects. The variation did not parallel blood pressure levels or other risk factors. Associations with coronary intimal thickness and smooth muscle cell numbers were, however, surprisingly strong. Figure 4 shows a linear association between coronary intimal thickness and renal arteriolar hyalinization, with a correlation between the age-adjusted variables of r = 0.53. Males exceeded females in hyalinization of arterioles, and this difference is proportional to the sex difference in coronary intimal thickness. Renal arteriolar hyalinization is associated with those features of the coronary intima that do not correlate with blood pressure or lipids (ie, thickness and cell numbers). Hyalinization of arterioles may therefore offer a clue to the causation of early preatherosclerotic changes that act independently of blood pressure and lipids. ls Alternatively, the hyalinization could be re-
'C
~::J
15'
~
1.0
• = Male 0= Female Mull r = 0.46 It = Necrotic core
CI
'"
n; '"
en "
,j,
It
•
• •
0.5
o
ij
:3' 0.26
It
;;; :;"! a;
0
u
'"
~
UJ
-1
o 1 Systolic Blood Pressure; Z-score
2
3
Figure 3. Extracellular lipid in the coronary intima, as in Figure 2, is plotted against systolic blood pressure age-adjusted by z-scores. Lines are fitted by multiple regression. (Reproduced with permission from Tracyet al_ 17) December 1995 Volume 310 Supplement 1
Tracy et al
..,
200
Ql
iii :::J '0
.=Male 0= Female Mult r= 0.53 It = Necrotic core
'"
References
• It
•
•
•"
Ql Cl ~ :::;J..
iii II)
cu
100
c
-" u
:£
76
"iii
.~
:s 0
·1 0 1 2 Hyalinized arterioles; ,j(N/cm ) Age adjusted
2
Figure 4. Thickness of the coronary intima is plotted against the numbers of hyalinized arterioles in the renal cortex age-adjusted by regression. The line is fitted by regression. (Reproduced with permission from Tracy et alP)
vealing variations in blood pressure that escaped detection by clinical measurement. Medial Changes in Coronary Vessels. The medial thickness of the coronary arteries also showed some interesting properties. (1) A correlation with blood pressure indicates that the coronary media can be used as a postmortem index for blood pressure. (2) Males have greater medial thickness than females. (3) Medial thickness increases with increasing renal arteriolar hyalinization. (4) the sex difference in medial thickness is proportional to the difference in arteriolar hyalinization. (5) Intimal and medial thicknesses are correlated (r = 0.42, age adjusted). As with renal arteriolar hyalinization, coronary medial thickness also appears to be an indicator for the early preatherosclerotic conditions that are, in some degree, independent of the observed risk factors. Conclusions
With the present state of knowledge about atherosclerosis, it remains uncertain whether fatty streaks are precursors of atherosclerotic lesions and to what extent modification of risk factors in children can prevent coronary atherosclerosis from progressing. The findings reviewed here suggest that lower levels of blood pressure and cholesterol are desirable in early life to retard the development of pre atherosclerotic and early atherosclerotic changes in the coronary arteries. The anatomic findings give credibility to the study of cardiovascular risk factors in children and further document the early onset of atherosclerosis and hypertension.
THE AMERICAN JOURNAL OF THE MEDICAL SCIENCES
1. Strong JP. Coronary atherosclerosis in soldiers: a clue to the natural history of atherosclerosis in the young. JAMA. 1986;256: 2863-6 . 2. Strong JP. Atherosclerotic lesions: natural history, risk factors, and topography. Arch Pathol Lab Med. 1992;116:1268-75. 3. Solberg LA, Strong JP. Risk factors and atherosclerotic lesions: a review of autopsy studies. Arterioscler Tbromb. 1983;3: 187-98. 4. Bonithon-Kopp C, Scarabin RY, Taquet A, Touboul PJ, Malmejac A, Guize L. Risk factors for early carotid atherosclerosis in middle-aged French women. Arterioscler Tbromb. 1991;11:966-72. 5. Strong JP, Oalmann MC, Newman WP ill, et al. Coronary heart disease in young black and white males in New Orleans: community pathology study. Am Heart J. 1984;108:747-59. 6. Newman WP ill, Wattigney W, Berenson GS. Autopsy studies in United States children and adolescents: relationship of risk factors to atherosclerotic lesions. Ann NY Acad Sci. 1991;623:16-24. 7. PDAY (Pathobiological Determinants of Atherosclerosis in Youth) Research Group. Relationship of atherosclerosis in young men to serum lipoprotein cholesterol concentrations and smoking. JAMA. 1990;264:3018-23. 8. Berenson GS, McMahan CA, Voors AW, et al. Cardiovascular risk factors in children. In: Andrews C, Hester HE; eds. The Early Natural History of Atherosclerosis and Essential Hypertension. New York: Oxford University Press; 1980:1-453. 9. Berenson GS, ed. Causation of Cardiovascular Risk Factors in Children: Perspectives on Cardiovascular Risk in Early Life. New York: Raven Press; 1986. 10. Guzman MA, McMahan CA, McGill BC Jr, et al. Selected methodologic aspects of the International Atherosclerosis Project. Lab Invest. 1968;18:479-97. 11. Tracy RE, Kissling GE. Comparisons of human populations for histologic features of atherosclerosis. Arch Pathol Lab Med. 1989;112:1056-65. 12. Tracy RE, Kissling GE, Gandia M, Reynolds C. Spatial dispersion of stainable lipid in frozen sections of human aorta. Virchows Arch [AJ. 1989;415:39-49. 13. Stary BC, Blankenhorn DB, Chandler AB, et al. A definition of the intima of human arteries and of its atherosclerosisprone regions: a report from the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association. AHA Medical/Scientific Statement Special Report. Arterioscler Thromb. 1992;12:120-34. 14. Stary BC, Chandler AB, Glagov S, et al. A definition of initial fatty streak, and intermediate lesions of atherosclerosis: a report from the Committee on Vascular Lesions of the Council on Arteriosclerosis, American Heart Association. AHA Medical/ Scientific Statement Special Report. Arterioscler Thromb. 1994;14:840-56. 15. Guyton JR, Klemp KF. Transitional features in human atherosclerosis: intimal thickening, cholesterol clefts, and cell loss in human aortic fatty streaks. Am J Pathol. 1993;143:1444-57. 16. Guyton JR, Klemp KF. Development of the atherosclerotic core region: chemical and ultrastructural analysis of microdissected atherosclerotic lesions from human aorta. Arterioscler Thromb. 1994;14:1305-14. 17. Tracy RE, Newman WP ill, Wattigney WA, Srinivasan SR, Strong JP, Berenson GS. Histologic features of atherosclerosis and hypertension from autopsies of young individuals in a defined geographic population: the Bogalusa Heart Study. Atherosclerosis. 1995;116:163-79. 18. Tracy RE. Blood pressure related separately to parenchymal fibrosis and vasculopathy of the kidney. Am J Kidney Dis. 1992;20:124-31.
541