Carotid artery measures are strongly associated with left ventricular mass in older adults (A report from the cardiovascular health study)

Carotid Arte Measures Are Strongly Associated Wit ‘x Left Ventricular Mass Older Adults (A Report from the Cardiovascular Health Study) Richard A. Kronmal, Joseph F. Polak,

PhD, Vivienne-Elizabeth M. Gardin,

MD, Julius

MD,

in

Smith, MD, Daniel H. O’Leary, MD, and Teri A. Manolio, MD, MHS

Associations of carotid artery diameter and intimal-medial thickness by ultrasound with echocardiographic left ventricular (LV) structure were examined in 3409 participants in the Cardiovascular Health Study, a population-based study of risk factors for coronary heart disease and stroke in men and women aged 265 years. At baseline, sector-guided M-mode echocardiography and B-mode ultrasound were used to evaluate the I~ ventricle and carotid arteries, respectively. Common carotid artery diameter and intimal-medial thickness were significantly related to LV mass in correlational analysis (r = 0.40 and 0.20, respectively, p
justment for age, gender, weight, systolic and diastolic blood pressure, antihypertensive medication use, prior coronary heart disease, electrocardiographic abnormalities, high-density lipoprotein, and factor VII. We speculate that changes in the arterial wall affect impedance to LV ejection leading to increases in LV mass. Further follow-up of this cohort is in progress and will help to determine whether such carotid artery measures coukf, by exacerbating LV hypertrophy, constitute another important risk factor for adverse cardiovascular outcomes. (Am J Cardiol 1996;77:628-633)

ecent studies confirm that left ventricular (LV) Rh ypertrophy is an important risk factor for adverse cardiovascular outcomes.1*2 LV mass, funda-

ations in LV geometry, particularly in the elderly. We therefore sought to evaluate the association of carotid artery structure with LV mass in a carefully defined population of free-living elderly. The Cardiovascular Health Study is a population-based longitudinal study of 5,201 adults aged ~65 years. Data from the baseline examination for the 3,409 participants whose examinations included an LV mass measurement provided the opportunity for the present study.

mentally determined by physiometric variables (weight, height, and gender), 3m5may increase with the additional hemodynamic burden imposed by elevated arterial pressure, increased vascular stiffness, 6-l’ and valvular heart disease. Pulse contour changes and decreased arterial vessel elasticity in hypertension and aging have been related to increased LV mass.“-‘3 It is further known that large vessel stiffness, pulse-wave velocity, and large vessel wall thickness increase with age.14-16 We hypothesize that the structural changes in conduit arteries associated with such pulse contour alteration might better reflect “hemodynamic history” than serial measurements of the blood pressure alone. In turn, changes in large vessel structure should predict similar alterFrom the Department of Biostatistics, University of Washington, Seattle, Washington; Division of Cardiology, Albany Medical College, Alban , New York. Division of Radiolo y, Geisrnger Medical Center, Brigham and Danviie Pennsylbania. Division 03 Radiology Women’s Hospital, Boston, Massachusetts; Division of Cardiology, Irvine Medical Center, University of California, Irvine, California; Epidemiology and Biometry Pro ram, Division of Epidemiology and Clinical Applications, Nationa 9 Heart, lung, and Blood Institute, Be thesda, Maryland. This study was supported by Contracts NO 1-HC85079, NO1 -HC-85080, NO1 -HC-8508 1, NO1 -HC-85082, NOl-HC-85083, NOTHC-85084, NOl-HC-85085, and NOlHC-85086 from the National Heart, lung, and Blood Institute, Bethesda, Maryland. Manuscript received May 18, 1995; revrsed manuscript received October 24, 1995, and accepted October 29. Address for reprints: Richard A. Kronmal, PhD, Cardiovascular Health Stud Coordinating Center, 1501 Fourth Avenue, Suite 2025, Seatte, Y Washington 98 101.

628

THE

AMERICAN

JOURNAL

OF

CARDIOLOGY@

VOL.

77

METHODS Dota collection:

The 5,201 Cardiovascular Health Study participants were recruited from a random sample of the Health Care Financing Administration Medicare eligibility lists and examined in the period from June 1, 1989, to May 31, 1990 in 4 U.S. communities: Forsyth County, North Carolina; Sacramento County, California; Washington County, Maryland; and Pittsburgh (Allegheny County), Pennsylvania. Committees on human research at each of the institutions involved in the study approved the study protocol. Details of the study design have been published.” Potential participants were excluded if they were institutionalized, wheelchairbound in the home, or currently under treatment for cancer. Eligible participants giving informed consent answered standard questionnaires about socioeconomic status, physical activity, functional status, cognitive function, diet, personal habits, family history, and medical history. Medical history information included recent hospitalizations and prior cardiac diagnoses and procedures. Information on

MARCH

15,

1996

medication use in the preceding 2 weeks was collected from prescription bottles. Sitting blood pressure was measured in the right arm after a 5minute rest using an appropriately sized cuff and a Hawksley random zero sphygmomanometer, model 7076 (Hawksley and Sons Limited, Sussex, England). The average of 2 measurements was used for analysis. Anthropometric measurements included weight, height, and waist and hip circumferences. Blood specimens were drawn and processed by a central laboratory. Among the determinations made from the blood specimens were low-density lipoprotein, high-density lipoprotein, triglycerides, albumin, fibrinogen, factor VII, and factor VIII. A resting electrocardiogram was taken and read in a central laboratory by computer. For those analyses examining participants with versus without coronary heart disease, an aggregate variable was created combining clinical history of (prevalent) coronary disease and/or a major electrocardiographic abnormality compatible with coronary heart disease. Ix The design of the echocardiographic protocol used in the Cardiovascular Health Study has been detailed elsewhere.” For each subject, a baseline echocardiogram was recorded onto super-VHS tape with a Toshiba SSH-160A (Tokyo, Japan) cardiac ultrasonograph using a standardized protocol. LV measurements were obtained from digitized tapes using strict criteria for readability. M-mode measurements were obtained according to conventions established by the American Society of Echocardiography.’ 9 LV mass was derived from the formula described by Devereux et al’“: LV mass (g) = 0.80 x I .04 [(VSTd + LVIDd + PWTd)3 (LVIDd)“] + 0.6, where VSTd = ventricular septal thickness at end-diastole, LVIDd = LV internal dimension at end-diastole, and PWTd = LV posterior wall thickness at end-diastole. The carotid arteries were evaluated with a highresolution linear array transducer (6.7 MHz, -3db point) capable of variable depth focusing (model SSA-270A; Toshiba Corporation) and equipped with a pulsed Doppler spectral analyzer (3.7 MHz, -3db point). A single longitudinal lateral view of the distal 10 mm of the right and left common carotid arteries and 3 longitudinal views in different imaging planes of each internal carotid artery were obtained. The internal carotid artery was defined as including both the carotid bulb, identified by the loss of the parallel wall present in the common carotid artery, and the 10 mm segment of the internal carotid artery distal to the tip of the flow divider that separates the external and the internal carotid arteries. Recorded studies were sent weekly to the Cardiovascular Health Study Ultrasound Reading Center for standardized readings. The high-resolution images of the common and internal carotid arteries were analyzed to calculate combined intimal-medial thickness of the near and far wall and the lumen diameter at each

arterial site. The maximal total diameter of the common carotid artery was determined from the distance between the periadventitiaiadventitia interfaces from the near and far walls, being thereby the external diameter of the artery. For the internal carotid artery, measurement of the maximal intimal-medial wall thickness was obtained at the location of the largest focal plaque, if plaque was present. All measurements were calculated using a specially designed computer program. A detailed description of the scanning and reading protocols, as well as reproducibility results, has been published.” In this report, we use 3 measures taken from the ultrasound examination of the carotid arteries. They are: ( 1) the average of the maximal total diameter of the right and left common carotid arteries, (2) the average of the common carotid near and far wall maximal intimal-medial thickness of the left and right side, and (3) the average of the internal carotid near and far wall maximal intimal-medial thickness of the right and left side for the 3 views. Averaging of dimensions and wall thicknesses was performed both to maximize data yield (allows use of a dataset with any of 4 missing wall thickness values) and to reduce measurement noise. Statistical analyses: Means, SDS, and correlation coefficients for each variable listed in Table I were computed as descriptive measures of the relations between the carotid measurements and LV mass. Partial correlation coefficients were used to estimate linearity of the relation between LV mass and carotid variables after adjustment for the confounding variables. A stepwise multiple regression analysis was initially performed to select the significant predictors of LV mass from among one of the carotid measurements and the variables listed in Table I. Variables that significantly predicted LV mass were then grouped: ( 1) Physiometric: age, gender, weight; (2) Hypertensive: systolic and diastolic blood pressure, and antihypertensive medication use; (3) Laboratory variables that were significant predictors of LV mass. Persons with and without coronary artery disease were analyzed separately because of concern for ischemic or infarction-related distortion of LV geometry. Analysis was then performed in 3 steps by variable group. The purpose of this analysis was to identify the relative contributions of these 3 groups of related variables to the relation between carotid geometry and LV structure. Each of these adjustments was computed using multiple linear regression techniques. The partial correlation coefficients for the prediction of LV mass from one of the carotid measures and the R* (the proportion of the total sum of squares explained by the regression) for the model were reported at each of the 3 steps. First, the LV mass was predicted from each of the carotid measurements separately. Then, these regressions were modified by adjustment for the first group of variables: weight, gender, and age. Then, adjustment was made for the second group of

MlSCELL4NEOUS/CAROTID

ARTERY MEASURES

629

variables: systolic and diastolic blood pressure and use of antihypertensive medication. Finally, an adjustment was made for the significant laboratory predictors: serum plasma high-density lipoprotein and factor VII. Estimates of the relation between LV mass and the carotid measurements allowing for possible nonlinearity were obtained using generalized additive modeling. 22*23This technique fits a nonparametric function for each of the variables in the model, which is then statistically compared with the best-fitting linear function to determine if the use of a nonparametric function provides an improved fit over the linear one. The multiple regression analyses and other descriptive analyseswere performed using SPSSIPC.” All p values should be interpreted cautiously because of the large number of nonindependent tests performed in these analyses; p values were used primarily for descriptive and screening purposes.

cardiogram of sufficient quality for the measurement of LV mass (n = 3,409). Stringent quality assurance was in part responsiblefor the relatively high echo data loss. Those with missing data tended to be older (p
With Left Ventricular

Mass and Carotid

Artery

Dimensions

Correlations

Variables

Mean

Left ventricular moss (g) Maximal total diameter (mm) Common carotid intimol wall thickness [mm) Internal carotid intimol wall thickness [mm) Age (vrl Height (cm) Weight (lb) Waist circumference (cm) Hip circumference (cm) Body moss index (kg/cm2) Kilo-calories expended in physical activity Alcohol consumption (in a week) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Heart rate (30 set) Pack-years of smoking Fibrinogen (mg/dl) Factor VII (%) Factor VIII (%) Triglyceride level (mg/dl) High-density lipoprotein (mg/dl) Low-density lipoprotein (mg/dl) Glucose (mg/dl) Insulin (mg/dl) Albumin (g/dl) Creatinine (mg/dl)

+ SD

LV Mass

Maximal Total Diameter

THE AMERICAN

JOURNAL

0.20' 0.48*

0.17* 0.30' 0.37'

0.40' 0.20'

0.48*

1.57 2 0.92

0.17*

0.30*

0.37'

0.08* 0.34* 0.43* 0.34* 0.24* 0.27' 0.00

0.22* 0.32* 0.28* 0.23' 0.08' 0.11* -0.00

0.22' 0.12" 0.11* 0.10' 0.02 0.04' -0.3

0.20* 0.10* 0.06* 0.08* -0.03 0.00 -0.04'

0.02 0.13* 0.04' -0.10' 0.13* 0.04 -0.13' 0.00 0.07* -0.29* -0.07* 0.16* 0.07* 0.01 0.22*

0.04 0.24* 0.05t -0.04' 0.18' 0.05t -0.14' -0.00 0.02 -0.24* -0.06* 0.13* 0.06' 0.01 0.22*

0.01 0.20' -0.00 -0.01 0.12* 0.04 -0.06' 0.01 0.04 -0.16* 0.07* 0.12' 0.05' 0.03 0.14*

0.05' 0.16* -0.03 0.01 0.21* 0.10* -0.05* 0.02 0.10* -0.18* 0.08' 0.09' 0.03 0.04' 0.17'

t 2 2 ? 2 + +

5.33 9.31 30.26 12.49 9.16 4.33 2,130.71

2.57 135.17 69.89 33.42 16.61 317.96 125.80 121.69 140.52 54.49 133.79 108.74 16.53 4.00 1.05

2 4 2 t zt 2 2 2 -c + + 4 ? 2 +

6.18 21.26 11.30 5.31 25.27 63.97 35.45 37.41 78.94 15.75 35.56 33.07 25.52 0.28 0.36

l p < 0.01;' p < 0.05. LV = left ventricular.

630

Internal Carotid lntimol Wall Thickness

151.62 251.62 9.19 2 0.97 0.99 If: 0.21

72.27 164.60 156.73 92.56 101.16 26.17 900.64

0.40*

Common Carotid lntimal Wall Thickness

OF CARDIOLOGYa

VOL.

77

MARCH

15,

1996

Table I lists the correlations of the 3 carotid artery variables and LV mass with the variables considered in the analysis. It is interesting that maximal total diameter (r = 0.40, p 0.05). The 3-step regression analysis for common carotid total diameter versus LV mass, which was done separately for persons with and without coronary artery disease, yielded results that were similar to those listed in Table II for the entire group after adjusting for the same variables in Table II (partial r = 0.194 and 0.180, both p
Regression

Equations

for left Ventricular

Varioble

DISCUSSION As expected, LV mass showed a strong correlation with weight and gender. However, even after weight and gender adjustment, maximal total common carotid artery diameter remained strongly correlated with LV mass. Our findings confirm those of Roman et al ‘* in a larger, older cohort. We concur that the most plausible explanation for the relation of common carotid total diameter and LV mass is through a shared physiologic response to blood pressure. Unexpectedly however, the addition of blood pressure variables to the regression models had only a trivial effect on this relation. The most likely explanation for this observation is that the current casual blood pressure (taken with the participant seated quietly for 5 minutes) is an imperfect surrogate for average daily blood pressure and lifetime blood pressure history. Such would be particularly relevant in the elderly, in whom systolic pressure has risen gradually over many years Finally, current use of antihypertensive drugs would be expected to both obscure and blunt effects of blood pressure history. In the present study, when systolic and diastolic blood pressure entered the model together, diastolic blood pressure had a slight but negative effect on the relation of LV mass to carotid structural measurements when compared with systolic pressure (partial r = -0.086, p ~0.005, and -0.046, p <0.05 for those with and without coronary artery disease). This finding is in contrast to that of Roman et al.‘* Because both systolic and diastolic blood pressure have been positively correlated with LV mass in younger persons, *’ the latter effect is likely due to the older age of our cohort. Flattening and subsequent decline in diastolic blood pressure, which accompanies age-related increases in systolic blood pressure, have been well described,*7,*8 and may re-

Mass* Partial

Maximal total diameter [mm) Common carotid intimal wall thickness Internal carotid intimal wall thickness

as large. No additional analysis was done for internal intimal-medial thickness because of the nonsignificant correlation after adjustment for the confounders.

r

p Value

Partial

r

p Value

p Value

0.0001

Age (yrl

0.020

0.3

0.032

0.1

0.03 1 0.043

0.08 0.05

Gender (0 = women, 1 = men) Weight (lb] Systolic blood pressure (mm Hg) Diastolic blood pressure [mm Hg) Antihypertensive medication (0 = no, 1 = yes) Prevalent coronary artery disease (0 = no, 1 = yes) KG abnormality (0 = no, 1 = yes) High-density lipoprotein (mg/dl) Factor VII (%)

0.132

0.0001 0.0001 0.0001 0.005 0.010

0.169

0.0001 0.0001 0.0001

0.176 0.318 0.127

0.0001 0.0001 0.0001

0.0005 0.005

-0.077 0.062 0.046 0.203 -0.062 -0.046

0.0001

Each pair of columns

0.303 0.1 16

0.060

0.005

0.185

0.0001

0.193

-0.047 -0.064

0.05 0.346

gives the regression

results using only

1 of the carotid

artery

0.001 0.0001

-0.049 -0.072

0.001

-

0.0001 -

-

-0.070 0.060 0.064

-0.061 0.051

R2 = l

0.073

-

0.282 0.089

-

r

0.166 -

(mm) [mm)

-

Partial

0.05 0.0001

-

-

specified

in the row with a value

MISCELLANEOUS/CAROTID

0.01

0.0001

0.0005 0.01

0.327

0.331 meclsures

0.005

for the partial

ARTERY

r.

MEASURES

631

fleet the course of progressive arterial degeneration or remodeling. Over time, higher pulse pressure would be expected to increase total artery diameter.” Such an increase would incorporate dilation of the vessel as the constraining muscle and connective tissue “deteriorate,” 30*3’as well as a compensatory increase in wall thickness as the system attempts to normalize vessel wall stress. The wall stress would otherwise result in vessel rupture as “the fraying and fragmentation of mural elastin fiber attributed to lifelong cyclic stress” takes its toll.3’ Any adaption of the arterial wall would then be expected to parallel related changes in the heart, in which, as a pump, increases in Wall thickness dominate. We did not enter ‘ ‘pulse pressure’ ’ as a unique variable into the models. Since regression coefficients for systolic and diastolic blood pressure were very similar in magnitude but opposite in sign, we have in effect derived the observation that pulse pressure is the quantity of interest (as opposed to arbitrarily assuming that the optimal relation of interest is systolic minus diastolic blood pressure). The lack of correlation of internal carotid wall thickness to LV mass was not unexpected, because in our study the internal carotid was evaluated near the bifurcation. This region of the carotid is prone to more focal plaque formation and extension3’ in contrast to the common carotid where wall thickness tends to increase more diffusely. Unfortunately, noninvasive imaging techniques, such as carotid ultrasound, cannot reliably distinguish normal from abnormal component contributions to overall wall thickness. Because the distinction between diffuse versus focal thickening ( “plaque” ) is somewhat arbitrary, we have not done analyses related to presence or absence of plaque. The weak negative association with LV mass seen for both high-density lipoprotein and factor VII may reflect their relation to atherosclerosis, 32*33which would be, in turn, related to LV mass through silent coronary artery disease. Against this hypothesis, low-density lipoprotein cholesterol was not a significant predictor of LV mass, and high-density lipoprotein and factor VII had little influence on the strength of the association of common carotid total diameter and common carotid average wall thickness to LV mass. Further, the relation between LV mass and common carotid total diameter is quite similar in participants with and without evidence of coronary artery disease. These 2 results suggest that the relation between LV mass and carotid measurements is probably not due-to atherosclerotic disease. Study limitations: The main limitation of the study is the uncertainty about effects of atherosclerosis (as opposed to arteriosclerosis) on the relation of carotid structure to LV mass as previously noted. Further, although failure to include participants with missing echo data is unlikely to have caused spurious associations, it is possible that bad the entire cohort been

632

THE AMERICAN

JOURNAL

OF CARDIOLOGY@

VOL.

77

included, other relations between the carotid arteries and the heart might have been observed.

1. Levy D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Left ventricular mass and incidence of coronary heart disease in an elderly cohort. The Fmmingham Heart Study. Ann Inrem Med 1989;111:101-107. 2. Levy D, Garrison RJ, Savage DD. Kannel WB, Castelli WP. Prognostic tmphcattons of echocardiographically determined LV mass in the Framingham heart study. N ~~~~ J Med 1990;322: 1561-1566. 3. Devereux RB, Lutas EM, Casale PN, Kligfield P, Eisenberg RR. Hammond lW, Miller DH, Reis G, Alderman MH, Laragh JH. Standardization of M-mode echocardiographic LV anatomic measurements. J Am Co11 Cardiol

Ws4;4:1222-1230.

4. Gardin JM, Savage DD, Ware JH, surface area on echocardiographic LV $iOn1987(~u~~l B):B-36-11-39. 5. de Simone G, Daniels SR, Devereux 0. Alderman MH. LV mass and body assessment of allomettic relations and

Henry WL. Effect of age, sex, and body wall mass in normal subjects. HypenenRB, Meyer RA, Roman MJ, de Divitiis size in normotensive children and adults: impact of overweight. JAm Co11 Cardiol

199WO:r251-1260.

6. Devereux RB, Pickering TG, Harshfield GA, Kleinert HD, Denby L, Clark L, Pregibon D, Jason M, Kleiner B, Borer JS, Laragh JH. LV hypertrophy in patients with hypertension: importance of blood pressure response to regularly recurring stress. Circulation 1983;68:470-476. 7. Nathwani D, Reeves RA, Marquez-Julio A, Leenen FH. LV hypertrophy in mild hypertension: correlation with exercise blood pressure. Am Heart J 1985;104:386-387. 8. Savage DD, Garrison RI, Kannel WB, Levy D, Anderson SJ, Stokes J. Feinlieb M. Castelli WP. The spectrum of LV hypertrophy in a general population sample: The Framingham Study. Circulation 1987;75( suppl I):I-26-I-33. 9. Levy D, Anderson KM, Savage DD, Kannel WB, Christiansen JC, Castelli WP. Echocardiographically detected LV hypertrophy: prevalence and risk factors. The Framingham Heart Study. Ann htem Med 1988;108:7-13. 10. Bouthier JD, DeLuca N, Safar ME, Simon AC. Cardiac hypertrophy and arterial distensibility in essential hypertension. Am He&J 1985;109:345-352. 11. Isnatd R, Pannier B, Laurent S, London G, Diebold B, Safar M. Pulsatile diameter and elastic modulus of the aortic arch in essential hypertension: a noninvasive sNdy. J Am Coff Cardiof 1989;13:900-905. 12. Roman M, Saba P, Pini R, Spitzer M, Pickering TG, Rosen S, Alderman MH, Devereux RB. Parallel cardiac and vascular adaptation in hypertension. Circulation 1992;86:1909-1918. 13. Saba PS. Roman MJ, Pini R, Spitzer M. Ganau A, Devereux RB. Relation of arterial pressure waveform to left ventricular and carotid anatomy in normotensive subjects. J Am Coff Cardiof 1993;22:1873-1880. 14. McVeigh G, Bums D, Finkelstein S, McDonald KM, Mock JE, Feske W, Carlyle PF, Flack 1, Grimm R, Cohn JN. Reduced vascular compliance as a marker for essential hypertension. Am J Hypertens 1991;4:245-251. 15. Avolio A, Deng F, Wei-Chang L, Luo Y-F, Huang Z-D, Xing L-F, O’Rourke MF. Effects of aging on arterial distensibility in populations with high and low prevalence of hypertension: comparison between urban and rural communities in China. Circulation 1985;71:202-210. 16. Bishop L. Heart Complications of arteriosclerosis. In: Arteriosclerosis. Lon-

don:oxfordpreFs,915,

17. Fried L, Borh&i N, Enright P, Furberg C, Gardin I, Kronmai R, Kuller L, Manolio T, Mittelmark M, Newman A, O’Leary D, Psaty B, Rautaharju P, Tracy R, Weiler P. Cardiovascular Health Study: design and rationale. For the CHS Collaborative Research Group. Ann Epidemiof 1991;1:263-276. 18. Gardin JM, Wong ND, Bcmmer W, Klopfenstein HS, Smith V-E, Tabatznik B, Siscovick D, Lobodzinski S, Anton-Culver H, Manolio TA. Echocardiographic design of a multi-center investigation of free-living elderly subjects: n, e c ar d’tovascular Health Study. JAm Sot Echo 1992:5:63-72. 19. Sahn DJ, DeMaria A, Kisslo J, Weyman A. The Committee on M-mode Standardization of the American Society of Echocardiography. Recommendations regarding quantitation in M-mode echocardiography: results of a survey of echocardiographic methods. Circulation 1978;58:10721083. 20. Devereux RB, Alonso DR. Lutas EM. Gottlieb GJ, Campo E, Sachs I, Reichek N. Echocardiographic assessment of LV hypertrophy: comparison with necropsy findings. Am J Cardiof 1986;57:450-458. 21. O’Leary DH, Polak JF, Wolfson SK, Bond MG, Bommer W, Sheth S, Psaty BM. Sharrett AR, Manolio TA. Use of sonogmphy to evaluate carotid atherosclerosis in the elderly: The Cardiovascular Health SNdy. Stroke 1991;22:1155-1163. 22. Hastie T, Tibshirani R. Generalized additive models: Some applications. J Am Stat Assoc 1987;82:371-386. 23. GAIM-Generalized Additive Interactive Modelling. Toronto, S.N. Tibshirani Enterprises Inc., 1991. 24. SPSS/PC Plus V. 4.0. Chicago: SPSS Inc., 1990. 25. Gardin JM, Siscovick D, Anton-Culvar A, Lynch JC, Smith V-E, Klopfen stein S, Bommer WJ, Fried L, O’Leary D. Manolio TA. Sex, age and disease

MARCH

15,

1996

affect echocardiographic left ventricular mass and systolic function in the freeliving elderly. The Cardiovascular Health Study. Circulation 1995;91:17391748. 26. Smith V-E, Tamres LK, Muldoon MF, Williams D, Ding Y, Manuck S, Jacob R. Neither awake nor sleep blood pressures better predict target organ effect. J Cardiovasc Pharnuzcol 1991;17(suppl 2):S125-S128. 27. Master AM, Lasser RP, Jaffe HL. Blood pressure in white people over 65 years of age. Ann Intern Med 1958;48:284-299. 28. Kannel WB. Blood pressure and the development of cardiovascular disease in the aged. In: Chaird FI, Dal1 J, Kennedy RD, eds. Cardiology and Old Age. New York, London: Plenium Press, 1976;143176. 29. Wells HG. The chemistry of arteriosclerosis. In: Cowdry EV, ed. Arteriosclerosis: A Survey of Problems. New York: MacMillan, 1933:328

30. Aschoff L. Introduction. The chemistry of arteriosclerosis. In: Cowdry EV, ed. Arteriosclerosis: A Survey of Problems. New York: MacMillan, 1933:117. 31. O’Rourke M. Vascular impedance in studies of arterial and cardiac function. Physiol Rev 1982:62:570-623 32. Heiss G, Sharrett R, Barnes R, Chambless L, Szklo M, Alzola C, and the Aric Investigators. Carotid atherosclerosis measured by B-mode ultrasound in populations: associations with cardiovascular risk factors in the ARIC study. Am J Epidemiol 1991; 134:250-256. 33. Meade T, Mellows S, Brorovic M, Miller G, Chakrabarti R, North W, Haines A. Stirline Y. Imeson J. Thomuson S. Haemostatic function and ischaemic heart disease: principal results of the Northwick Park Heart Study. Lancer 1986;2:533-537.

MISCELtANEOUS/CAROTID

ARTERY MEASURES

633