Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
JACC Vol. 25, No. 4 March 15, 1995:895 900 895 C O C A I N E USE Lack of Association of Recreational Cocaine and Alcohol Use With Left Ventricular Mass in Young Adults The Coronary Artery Risk Development in Young Adults (CARDIA) Study GEORGEANNE S. H O E G E R M A N , J A M E S M. R A C Z Y N S K I , M D , C O R A E. L E W I S , M D , M S P H , J O H N F L A C K , M D , M P H , * PI-ID,* J E N N I F E R CAVENY, MS, J U L I U S M. G A R D I N , M D , FACC'[" Birmingham, Alabama; Minneapolis. Minnesota; and Orange, California Objectives. This study examined the associations of left ventricular mass with self-reported cocaine and alcohol use prevalent in the young adult population. Background. Increased left ventricular mass has been associated with long-term use of cocaine and alcohol; however, few of the published studies have been population based. Methods. Data from 3,446 black and white participants (mean age 29.9 years) in the Coronary Artery Risk Development in Young Adults (CARDIA) study were used to examine the associations between echocardiographically measured left ventricular mass obtained from 1990 to 1991 and self-reported cocaine and alcohol use. Categories of cocaine use were those who denied ever using cocaine (n = 2,122), experimental users who admitted to cocaine use only 1 to 10 times in their lifetime (n = 755) and recurrent users who admitted to cocaine use >10 times in their lifetime (n = 568). For alcohol consumption, categories were abstainers who consistently denied any alcohol consumption in the year before each of the three CARDIA examinations (n = 275), occasional users who admitted consuming alcohol less than once a week or not at all during the year before the third examination (n = 1,322), moderate users who consumed 1 to 209 ml of alcohol/week during the year before the third examination (n = 1,524) and heavy users who consumed >210 ml of alcohol/week during the year before the third examination (n = 323). Estimated power to detect a 10% difference in left ventricular mass between groups was >80%. Results. For white women, left ventricular mass was significantly higher among those who reported 1 to 10 lifetime uses of cocaine than in never-users (128.5 g [SE 2.0] vs. 122.7 g [SE 1.4], p = 0.002). There were no other significant di~erences in left ventricular mass among categories of cocaine or alcohol use in unadjusted analyses or among analyses controlling for age, body mass index, alcohol or cocaine use, physical activity, cigarette smoking status and systolic blood pressure. Conclusions. At the levels of consumption reported, neither cocaine nor alcohol use was associated with left ventricular mass in this cohort of healthy young adults. Increased left ventricular mass as measured by echocardiography precedes the clinical presentation of some forms of cardiac disease (i.e., congestive heart failure, hypertensive cardiomyopathy) (1-5) and is itself a risk factor for cardiovascular mortality (5,6). Several recent studies (7-11) have reported an association of left ventricular mass with long-term use of either cocaine or alcohol. Increased left ventricular mass has been reported among long-term male cocaine abusers admitted to a Veterans Affairs hospital for treatment of cocaine addiction compared with age-matched male nonusers (7). Studies of alcoholic subjects (8-11) and Framingham study participants (12) have also shown associations of alcohol intake with left ventricular mass. However, interpretation of much of the available data is limited because of small sample sizes, lack of appropriate control subjects and failure to adjust for tobacco consumption and other confounders. Furthermore, few studies have examined these associations in general population samples, where use patterns differ markedly from samples of patients seeking treatment for abuse or for medical consequences of substance use. Finally, there have been few studies examining associations of left ventricular mass with use of cocaine or alcohol in women and blacks. Thus, we examined the associations of left ventricular mass with self-reported cocaine and alcohol use in a large, well described, longitudinal population-based sample to determine the relevance of prevalent cocaine and alcohol use to left ventricular mass in the young adult population. From the Division of Preventive Medicine, Departments of Medicine and Health Behavior, University of Alabama at Birmingham, Birmingham, Alabama; *Division of General Internal Medicine, University of Minnesota, Minneapolis, Minnesota; and tDivision of Cardiology, Department of Medicine, University of California-Irvine Medical Center, Orange, California. This study was supported by Contracts N01-HC-48047, N01-HC-48048, N01-HC-48049, NOI-HC-48050, N01-HC-95095 and N01-HC-95100 from the National Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland. Manuscript received July 14, 1994; revised manuscript received October 27, 1994, accepted November 1, 1994. Address for corremondence: Dr. Cora E. Lewis, 1717 1lth Avenue South, Room 734, Birmingham, Alabama 35205. © 1995 by the American College of Cardiology (J Am Coll Cardiol 1995;25:895-900) 0735-1097/95/$9.50 0735-1097(94)00469-7 896 H O E G E R M A N ET AL. C O C A I N E AND LEFT V E N T R I C U L A R MASS Methods Study population. The Coronary Artery Risk Development in Young Adults (CARDIA) study is a prospective epidemiologic study designed to determine the antecedents of cardiovascular disease risk factors in young adults and to identify health habits and behaviors that predict the levels of, and changes in, these risk factors. During 1985 and 1986, 5,115 young adults 18 to 30 years old were recruited from four cities: Birmingham, Alabama; Chicago, Illinois; Minneapolis, Minnesota; and Oakland, California. The study population was approximately balanced at baseline according to age (18 to 24 and 25 to 30 years), gender, race (black and white) and educational level (high school graduate or less and beyond high school). Details of the study design and participant characteristics at baseline have been described elsewhere (13). The study was approved by the institutional review committee of each center; examination procedures followed established human use guidelines. Participants gave written informed consent at each examination. Ninety-one percent of the surviving participants returned during 1987 to 1988 for a second examination, and 86% returned for a third examination during 1990 to 1991, when echocardiography was performed. Retention rates for the third examination were similar for various categories of alcohol or cocaine use: 89% of those who admitted to consuming >210 ml of alcohol/week at the second examination returned for the third examination, compared with 88% of alcohol abstainers; 91% of those who admitted to >10 lifetime uses of cocaine at the second examination returned for the third examination, compared with 89% of those who denied ever using cocaine. The study sample included 3,446 of the 4,352 participants attending both the second and third examinations. We excluded from analyses 238 participants who had missing drug use data at the second examination, 34 who had missing drug use data at the third examination, 180 who had missing left ventricular mass data and 277 who had inconsistent drug use data (participants reporting any drug use at the second examination who denied ever having used drugs at the third examination). Data collection. Drug and alcohol use. Self-administered questionnaires given at both the second and third examinations were used to determine whether a participant had ever used a variety of illicit drugs (marijuana, cocaine, opiates or amphetamines). Follow-up questions quantified use of each drug over the participant's lifetime and use within the previous 30 days. As described previously (14), alcohol consumption was determined at each examination by an interviewer-administered questionnaire that quantified the number of drinks consumed per week of wine (5 oz [148 ml]), beer (12 oz [355 mlD and liquor (1.5 oz [44 ml]). The total milliliters of absolute alcohol consumed per week was calculated as follows: [(Wine × 17.02) + (Beer × 16.7) + (Hard liquor x 19.09)]. Echocardiography. The echocardiographic protocol used in the CARDIA study was very similar to that previously de- JACC Vol. 25, No. 4 March 15, 1995:895-900 scribed for the Cardiovascular Health Study (15). M-mode and two-dimensional echocardiograms were obtained with patients lying in the left lateral decubitus position using a 2.5-MHz Acuson phased-array sector scanner. Stop-frame M-mode and real-time two-dimensional images were recorded at 30 frames/s on 0.5-in. super-VHS tape and sent to the central Echocardiography Reading Center, University of CaliforniaIrvine. For each participant, the two-dimensional parasternal long-axis view, parasternal short-axis view at the level of the papillary muscle tips and the apical four and two-chamberviews were obtained. M-mode images of the left ventricle were recorded using two-dimensional parasternal images to direct the M-mode cursor perpendicular to the ventricular septum and left ventricular posterior wall at approximately the level of the mitral leaflet tips. All recordings were performed with respiration suspended at midexpiration. Measurements were made from digitized images using a Dextra D-200 off-line image analysis system. Images were played back on a super-VHS videotape recorder and digitized at 30 frames/s in real time. All measurements were traced on a video screen using a digitizer, with primary measurement data going directly into a dBase IV data base (Ashton-Tate). Quality control measures for echocardiography included standardized training of echocardiographic technicians and readers, technician observation by a trained echocardiographer, blind duplicate readings to establish intrareader (4% sample) and interreader (8% sample) measurement variability, periodic review sessions, phantom studies on the ultrasound equipment and quality control audits. M-mode left ventricular mass was calculated from the formula of Devereux et al. (16) with the correction factor for measurements performed using the following American Society of Echocardiography convention (17): Left ventricular mass (g) = 0.08 {1.04[(VSTd + LVIDd + PWTd) 3 (LVIDd)3]} + 0.6, where VSTd = ventricular septal thickness at end-diastole (cm); LVIDd = left ventricular internal dimension at end-diastole (cm); and PWTd = left ventricular posterior wall thickness at end-diastole (cm). Left ventricular internal dimension at end-systole (cm) was also determined from M-mode images. Average values from 3 beats were used in analyses; measurements were made on a stop-frame superVHS image from cycles in which the right and left side of the ventricular septum, and the endocardial and the epicardial interfaces of the left ventricular posterior wall, were clearly defined. Other factors. Potential confounding factors included age, race, gender, height, body mass index, physical activity, cigarette smoking and blood pressure. Body weight was measured in light clothing to the nearest 0.5 lb (0.227 kg) with a calibrated balance-beam scale; height (without shoes) was measured to the nearest 0.5 cm using a vertical ruler. Body mass index was computed as follows: [Weight (lb)/2.2]/[Height (cm)/100] 2. Physical activity was assessed using the reported frequency of participation in each of 13 physical activities during the previous year, weighted by the intensity level of each activity, and then summed to give a total physical activity JACC Vol. 25, No. 4 March 15, 1995:S95-900 score (18). Self-reported measures of cigarette smoking (present, past and never) (19), age at the third examination, race and gender were also used in analyses. Three sitting blood pressure measurements were obtained, after a 5-min rest and at l-rain intervals from the right arm of all participants; the average of the second and third measurements was used in analyses. Trained and certified technicians used a Hawksley random-zero sphygmomanometer (W. A. Baums Co.) and common protocols for blood pressure measurements. Statistical analysis. Three categories were created to evaluate cocaine use at the third examination on the basis of the participants' report of any lifetime use of cocaine. These categories divided the cohort (n = 3,446) into those who denied ever using cocaine (n = 2,122), experimental users who admitted to cocaine use only 1 to 10 times in their lifetime (n = 755) and recurrent users who admitted to cocaine use >10 times in their lifetime (n = 568). The latter category included participants reporting 11 to 99 lifetime uses and ->100 uses, which were combined because of small numbers in the highest category. Because only 94 participants admitted using cocaine within 30 days of their third CARDIA examination, stratification for this use pattern was not done because of small numbers in the race-gender groups. Cocaine use could not be further separated from use of other illicit drugs because only two participants admitted exclusive use of cocaine. The small number of participants (n = 9) who admitted to use of other illicit drugs (i.e., opiates or amphetamines) in the absence of marijuana or cocaine use precluded separate analyses of these groups. We created categories to evaluate alcohol use in participants for whom data were available on alcohol consumption (n = 3,129); categories were based on the findings of an association between alcohol intake and left ventricular mass in Framingham subjects (12). We defined abstainers as participants who consistently denied any alcohol consumption (i.e., those who denied any use in the year before each of their three CARDIA examinations [n = 275]). Definitions of the other categories were based on current use (i.e., that reported during the year before the third examination). Occasional users admitted consuming alcohol less than once a week or not at all during the year before the third examination (n = 1,322); moderate users admitted consuming 1 to 209 ml of alcohol/ week during this period (n = 1,524); and heavy users admitted consuming ->210 ml of alcohol/week (n = 323). To examine the effects of larger intakes of alcohol and the stability of findings at different levels, we also examined associations for those consuming 1 to 350 ml of alcohol/week (n = 1,705) and those consuming >350 ml of alcohol/week (n = 142); however, because the results were similar, these data are not presented. We excluded from analyses 177 participants who were known to be hypertensive (self-reported history of antihypertensive drug use or systolic blood pressure >140 mm Hg or diastolic blood pressure >90 mm Hg at examination). However, data that included participants with hypertensive blood pressure levels who were not taking antihypertensive medication yielded similar results to those reported here. H O E G E R M A N ET AL. COCAINE AND LEFT V E N T R I C U L A R MASS 897 Associations of left ventricular mass and baseline characteristics of potential confounders were examined between race-gender groups using analysis of variance and chi-square tests where appropriate. Pairwise comparisons were tested using the Scheff6 procedure (20). Associations of unadjusted left ventricular mass with cocaine and alcohol use were examined with trend tests within the race-gender groups (21). Analysis of covariance was also used to examine associations of left ventricular mass with cocaine and alcohol use for the race-gender groups, controlling for significantly confounding factors, including systolic blood pressure, age, physical activity, smoking status and alcohol or cocaine consumption. Analyses were repeated, controlling for number of cigarettes smoked per day (0 for nonsmokers and for ex-smokers) instead of smoking status; these findings are not presented because of similar results. Separate models were used to control for body mass index and height. The experimental error rate of alpha 0.05 was adjusted with a Bonferonni correction, depending on the number of comparisons made (20). Analyses were performed with SAS version 6.04 software. Results Because of the initial eligibility criteria, the sample was young (mean age 29.9 years), had higher than a high school education on average and had approximately equal representation from black and white, as well as male and female, groups (Table 1). Left ventricular mass was significantly greater in men than in women and in blacks than in whites. There were also significant race- or gender-related differences, or both, in cocaine and alcohol use and in demographic and other characteristics of the participants. Although >50% of the participants denied ever having smoked tobacco or used cocaine, <10% of the cohort consistently reported that they abstained from alcohol use at each of the three examinations. Left ventricular mass was not consistently associated with either increased cocaine (Table 2) or alcohol use (Table 3) when compared with that of nonusers of these substances. In unadjusted analyses, cocaine was associated with left ventricular mass only among white women; however, a clear doseresponse association was not present in pairwise comparisons (Table 2). No significant associations with alcohol were present (Table 3). When the data were adjusted for height as well as other potential confounders (systolic blood pressure, age, physical activity, smoking status, alcohol or cocaine consumption), there were again no associations of either cocaine or alcohol use with left ventricular mass within race-gender groups (data not shown). In models adjusted for the other potential confounders and body mass index, the only significant difference found was in white women who reported 1 to 10 lifetime uses of cocaine compared with those who reported never having used cocaine (129.5 g [SE 1.9] vs. 121.9 g [SE 1.4], respectively, p = 0.002). Further analyses by type of alcoholic beverage (beer, wine or liquor) within race-gender groups revealed no association between left ventricular mass and alcohol consumption. 898 HOEGERMAN ET AL. COCAINE AND LEFT VENTRICULAR MASS JACC Vol. 25, No. 4 March 15, 1995:895-900 Table 1. Characteristics of 3,446 Participants: C A R D I A Study, 1990 to 1991 Men Age (yr) Education (yr) Systolic BP (ram Hg) Diastolic BP (ram Hg) Body mass index (kg/m:) Physical activity, score LVM (g) Never smoked tobacco (%) Never used cocaine (%) Alcohol abstainers (%) Women Black White Black White 29.3 (3.7)* 13.5 (2.1)* 112 (9.4)*,1, 72 (8.5)*,1, 26.4 (5.0)*,1, 488.7 (347.7)t 174.3 (42.6)*t 56t 63't 10" 30.5 (3.4)* 15.3 (2.6)* 109 (9.5)*t 70 (8.3)*+ 25.3 (3.8)*+ 465.9 (289.9),1, 168.7 (39.5)*,1, 61 50,1,* 5* 29.3 (3.9)* 13.7 (1.9)* 105.4 (9.3)*t 68 (8.7)*t 27.8 (7.2)*t 261.8 (227.5)*i 132.2 (34.9)*t 64",1, 77',1, 13' 30.5 (3.5)* 15.3 (2.4)* 101 (8.6)*t 65 (8.1)*,1, 24.2 (4.8)*,1, 345.0 (247.7)*,1, 125.1 (33.1)*,1, 58* 56*+ 6* *p < 0.0125, race comparison within gender (black men vs. white men, black women vs. white women), tp < 0.0125, gender comparison within race (black men vs. black women, white men vs. white women), Bonferroni correction 0.05/4 0.0125. Data presented are mean value (SD), unless otherwise indicated. BP - blood pressure; LVM = left ventricular mass. Echocardiographic measures of left ventricular internal dimension at end-diastole and end-systole, ventricular septal thickness at end-diastole and left ventricular posterior wall thickness at end-diastole were also not significantly associated with cocaine or alcohol use (data not shown). Discussion Our data do not support associations of cocaine or alcohol use with left ventricular mass at the levels of consumption reported in this healthy population of black and white young men and women. The isolated, statistically significant association of left ventricular mass with cocaine use in white women did not show a monotonic dose-response relation (22) and was small and inconsistent with the findings in the other racegender groups. Thus, this finding is unlikely to be of clinical significance, Previous studies of cocaine and left ventricular mass. Our results differed from those of others, who have reported associations of cocaine with left ventricular mass. Brickner et al. (7) reported that alcohol consumption >210 ml/week and cocaine abuse were independently associated with left ventricular mass index and posterior wall thickness in men undergoing inpatient treatment for cocaine abuse compared with hospital house staff and employees matched for race and age who denied illicit drug use. However, interpretation of these data is limited because of failure to adjust for tobacco consumption, small sample size, use of a control sample that differed from the study sample on many characteristics besides cocaine use and lack of female patients. In addition, the level and duration of use of cocaine or other substances may have been substantially different from that typical of CARDIA Table 3. Unadjusted Left Ventricular Mass (g) by Alcohol Consumption: C A R D I A Study, 1990 to 1991 Table 2. Unadjusted Left Ventricular Mass (g) by Lifetime Cocaine Use: C A R D I A Study, 1990 to 1991 Men Never used cocaine No. Mean (SE) 1-10 uses of cocaine No. Mean (SE) >10 uses of cocaine No. Mean (SE) Overall p value? p value trend Men Women Black (n = 662) White (n = 862) (n Black 923) White (n = 998) 417 173.6 (2.1) 433 167.7(1.9) 712 131.6(1.3) 560 122.7(1.4) 119 174.4 (3.9) 235 168.8(2.6) 132 132.8(3.0) 269 128.5(2.0)* 126 176.5 (3.8) 0.80 0.51 194 170.8(2.8) 0.67 0.22 79 136.8(3.9) 0.45 0.38 169 127.3(2.5) 0.04 0.04 *p < 0.025, comparisons within race-gender groups, using never-used cocaine as reference group; Bonferroni correction 0.05/2 = 0.025. ?Analysis of variance. Alcohol abstainers* No. Mean (SE) Alcohol use more than once a wk No. Mean (SE) 1-209 ml alcohol/wk No. Mean (SE) ->210 ml alcohol/wk No. Mean (SE) Overall p value+ p value trend Women Black (n = 661) White (n - 862) (n Black 924) White (n = 997) 63 173.4 (5.4) 40 172.6(6.2) 117 131.1 (3.2) 55 122.9(4.5) 196 174.4 (3.0) 207 167.8(2.7) 480 131.4(1.6) 439 126.6(1.6) 276 174.8 (2.6) 492 167.3(1.8) 290 132.8(2.0) 466 123.4(1.5) 126 173.1 (3.8) 0.98 0.94 123 174.5(3.6) 0.28 0.48 37 141.0(5.7) 0.42 0.21 37 129.2(5.4) 0.40 0.64 *p < 0.017, comparisons within race-gender groups, using alcohol abstainers as reference group; Bonferroni correction 0.05/3 = 0.017. ?Analysis of variance. JACC Vol. 25, No. 4 March 15, 1995:895 900 participants. In a similar study of 52 male veterans seeking treatment for crack cocaine addiction, Chakko et al. (23) reported that 54% were found to have left ventricular hypertrophy by echocardiography, whereas none of the normal volunteers and psychiatric control patients studied had this finding. Measures of left ventricular mass were also significantly higher in the abusers than in the control groups. Limitations of that study were that race was not balanced across groups (there were many more blacks in the crack users than in the other groups) and although Chakko et al. stated that tobacco and alcohol use was similar between crack users and psychiatric controls, these factors were not controlled for in the analyses. Finally, the groups were somewhat older on average (36 years for crack users to 33 years for normal control subjects) than those in the CARDIA cohort. Numerous other investigators (24-32) have reported cocaine-associated cardiovascular effects in addition to increased left ventricular mass. Interpretation of the associations in these retrospective studies is complicated by many factors, including 1) current and previous polysubstance use (i.e., alcohol, tobacco and other drugs) (24-26,29,30,33,34); 2) pattern of cocaine use (long term, intermittent or short term) (25,26,30,35-37); 3) method of cocaine administration (intranasal, inhalation or injection) (25,26,31,35-37); 4) life-style associated with illicit drug use (crime, violence, trauma and stress) (30,34-36); and 5) sample bias (35-38). Previous studies of alcohol and left ventricular mass. Other recent studies have addressed the relation between alcohol consumption and left ventricular mass (8-12). In 4,491 Framingham Study participants (mean age >50 years) who were free of apparent cardiovascular disease at baseline, left ventricular mass was independently associated with total alcohol intake in men but not in women (12). The gender difference was solely attributed to the inverse association of liquor (as opposed to beer and wine) with left ventricular mass in women. That study did not address the effect of alcohol consumption on left ventricular mass in younger adults, which is the age group with the highest prevalence of heavy drinking (39,40). Several other studies of alcoholic subjects (8-11) have also reported an association of alcohol with increased left ventricular mass; however, it is very difficult to interpret these findings because of the small sample sizes involved and lack of control for confounders. Study limitations. Although we report no association between cocaine or alcohol use and left ventricular mass, extrapolation of our data to other populations should be done cautiously because of several limitations of this study, First, although the CARDIA study is a large population-based sample, the sample sizes of the highest use categories were relatively small. The relatively low prevalence of heavy cocaine or alcohol use in our sample may have limited our power to detect associations of left ventricular mass with use of these substances. However, we estimate that our power to detect a 10% difference in left ventricular mass between use groups was >80% in all groups and was frequently >90%. Furthermore, the level of use in our study may be more reflective of that in H O E G E R M A N ET AL. COCAINE AND LEFT V E N T R I C U L A R MASS 899 the general population and thus may better indicate the risk of increased left ventricular mass attributable to cocaine and alcohol in the general population. Second, the criteria that we selected may be too inclusive for both alcohol and cocaine. Our operational definition of recurrent cocaine use included participants who admitted to use on at least 11 occasions in their lifetime. Most of these recurrent users denied use of cocaine within the 30 days before their CARDIA examination and may not have used cocaine for years. Although we were able to define alcohol consumption in terms of average weekly intake in the year preceding the third examination, we were not able to quantify duration of alcohol consumption at those levels. Undoubtedly, not all of the CARDIA cohort classified as heavy users were long-term alcohol abusers. Finally, cardiovascular sequelae, such as increased left ventricular mass, may be dependent on both quantity of cocaine or alcohol consumed and duration of heavy use. Because our study sample was young, cocaine and alcohol use probably did not extend beyond a decade or two. It is thus possible that cardiovascular effects of continued long-term cocaine or alcohol consumption, or both, in the CARDIA cohort might be demonstrable after more sustained use. Prevalence of substance use. A further limitation of this study is that our categorization of alcohol and cocaine use was based solely on self-report. Furthermore, the generalizability of our findings could be affected by differences in substance use patterns between CARDIA participants and other young urban adults. We therefore compared alcohol and drug use characteristics in the CARDIA study with those of other samples to determine any differences in drug-use history. The self-reported prevalence of abstinence, experimentation and recurrent use of cocaine and alcohol in the CARDIA study is similar to that found in other population-based surveys, such as the Behavioral Risk Factor Surveillance Study (40) and the National Household Survey (39). In specific analyses, the lifetime use of alcohol in CARDIA participants was comparable to that of similarly aged respondents to the National Household Survey on Drug Abuse (39). Lifetime use of cocaine was higher in CARDIA participants; however, the proportion of heavier cocaine use was similar. In the CARDIA study, 43% of participants who admitted to ever using cocaine reported using cocaine on >10 occasions in their lifetime compared with 41% of 26 to 34-year old respondents interviewed in the 1990 National Household Survey on Drug Abuse (39). In contrast to the National Household Survey, CARDIA participants were recruited from large metropolitan areas where a higher prevalence of illicit drug use is found compared with areas with a lower population density (39). The prevalence of illicit drug use found in the longitudinal Harlem Health Study was similar to that in the CARDIA study; in the Harlem study 46% of 26 to 31-year old respondents (n = 411) reported "nonexperimental" lifetime cocaine use, and 96% any lifetime alcohol use (41). Thus, although underreporting of cocaine or alcohol use, or both, by participants could potentially have biased our results toward the null, the generally consistent reporting between the second and third examina- 900 HOEGERMAN ET AL. COCAINE AND LEFT VENTRICULAR MASS tions by participants and the similar prevalence of use between CARDIA and other population-based studies makes this unlikely. Conclusions. We found no consistent associations of left ventricular mass with the levels of cocaine or alcohol use reported by this group of healthy young adults. However, we cannot conclude that alcohol and cocaine are not risk factors for the development of cardiovascular complications if used over longer periods of time or at different levels of consumption than reported by CARDIA participants. Further research is needed to determine whether exposure thresholds exist for these substances. However, the pattern of use prevalent in this sample was not associated with left ventricular mass and implies that cocaine and alcohol use may not be important determinants of left ventricular mass in the young adult population overall. References 1. Casale PN, Devereux RB, Milner M, et al. Value of echocardiographic measurement of left ventricular mass in predicting cardiovascular morbid events in hypertensive men. Ann lntern Med 1986;105:173-8. 2. Devereux RB. Does increased blood pressure cause left vcntricular hypertrophy or vice versa? Ann Intern Med 1990;112:157-9. 3. Devereux RB, Alonso DR, Lutas EM, et al. Echocardiographic assessment of left ventricular hypertrophy: comparison with necropsy findings. Am J Cardiol 1986;57:450-8. 4. Gottdiener JS, Brown J, Zoltick J, Fletcher RD. Left ventricular hypertrophy in men with normal blood pressure: relation to exaggerated blood pressure response to exercise. Ann lntern Med 1990;112:161-6. 5. Le~ D, Garrison RJ, Savage DD, Kannel WB, Castelli WP. Prognostic implications of ecbocardiographically determined left ventricular mass in The Framingham Heart Study. N Engl J Med 1990;322:1561-6. 6. Lee DK, Marantz PR, Devereux RB, Kligfield P, Alderman MH. Left ventricular hypertrophy in black and white hypertensives--standard electrocardiographic criteria overestimate racial differences in prevalence. JAMA 1992;267:3294-9. 7. Brickner ME, Willard JE, Eichhorn EJ, Black J, Grayburn PA. Left ventricular hypertrophy associated with chronic cocaine abuse. Circulation 1991;84:1130-5. 8. Mathews EC Jr, Gardin JM, Henry WL, ct al. Echocardiographic abnormalities in chronic alcoholics with and without overt congestive heart failure. Am J Cardiol 1981:47:570-8. 9. Cregler LL, Worner TM, Mark H. Echocardiographic abnormalities in chronic symptomatic Alcoholics. Clin Cardiol 1989;12:122-8. 10. Kupari M, Koskinen P, Suokas A. Left ventricular size, mass and function in relation to the duration and quantity of heavy' drinking in alcoholics. Am J Cardiol 1991;67:274-9. 11. Urbano-Marquez A, Estruch R, Navarro-Lopez F, Grau JM, Mont L, Rubin E. The effects of alcoholism on skeletal and cardiac muscle. N Engl J Med 1989;320:409-15. 12. Manolio TA, Le~ D, Garrison RJ, Castelli WP, Kannel WB. Relation of alcohol intake to left vcntricular mass: the Framingham Study. J Am Coll Cardiol 1991;17:717-21. 13. Friedman GD, Cutter GR, Donahue RP, et al. CARDIA: study design, recruitment, and some characteristics of the examined subjects. J Clin Epidemiol 1988;41:1105-16. 14. Dyer AR, Cutter GR, Liu K, et al. Alcohol intake and blood pressure in young adults: the CARDIA study. J Clin Epidemiol 1990;43:1-13. 15. Gardin JM, Wong ND, Bommer W, et al. Echocardiographic design of a multi-center investigation of free-living elderly subjects: the Cardiovascular Health study. J Am Soc Echo 1992;5:63-72. JACC Vol. 25, No. 4 March 15, 1995:895-9011 16. Devereux RB, Alonso DR, Lutas EM, et al. Echocardiographic assessment of left ventricular hypertrophy: comparison with necropsy findings. Am J Cardiol 1986;57:450-8. 17. 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;58:1072-83. 18. Jacobs DR, Hahn LP, ttaskell WL, Pirie P, Sidney S. Validity and reliability of short physical activity history: CARDIA and the Minnesota Heart Health Program. J Cardiopulm Rehab 1989;9:448-59. 19. Wagenknecht LE, Perkins LL, Cutter GR, et al. Cigarette smoking behavior is strongly related to educational status: the CARDIA Study. Prey Med 1990;19:158-69. 20. Woolson RF. Statistical Methods for the Analysis of Biomedical Data. New York: Wiley, 1987:327-37. 21. Rosner B. Fundamentals of Biostatistics. 3rd edition. Boston: PWS-Kent, 1990:35%60. 22. Maclure M, Greenland S. Tests for trend and dose response: misinterpretations and alternatives. Am J Epidemiol 1992;135:96-104. 23. Chakko S, Fernandez A, Mellman TA, Milanes FJ, Kessler KM, Myerburg RJ. Cardiac manifestations of cocaine abuse: a cross-sectional study of asymptomatic men with a history of long-term abuse of "crack" cocaine. J Am Coil Cardiol 1992;20:1168-74. 24. Amin M, Gabelman G, Karpel J, Buttrick P. Acute myocardial infarction and chest pain syndromes after cocaine use. Am J Cardiol 1990;66:1434-7. 25. lsner JM and Chokshi SK. Cardiac complications of cocaine abuse. Annu Rev Med 1991;42:133-8. 26. Kloner RA, Hale S, Alker K, Rezkalla S. The effects of acute and chronic cocaine use on the heart. Circulation 1992;85:407-19. 27. Moskowitz RM, Errichetti AJ. Cardiovascular evaluation 'alter withdrawal from chronic alcohol or cocaine-alcohol abuse. J Addict Dis 1991;10:4%65. 28. Pett GW, Brust JC, Tatemichi TK. Embolic stroke after smoking "crack" cocaine. Stroke 1990;21:1632-5. 29. Smart RG. Crack cocaine use: a review of prevalence and adverse effects. Am J Drug Alcohol Abuse 1991;17:13-26. 30. Tazelaar HD, Karch SB, Stephens BG, Billingham ME. Cocaine and the heart. Hum Pathol 1987;18:I95-9. 31. Virmani R, Robinowitz M, Smialek JE, Smyth DF. Cardiovascular effects of cocaine: an autopsy study of 40 patients. Am Heart J 1988;115:1068-6. 32. Perper JA, VanThiel DH. Cardiovascular complications of cocaine abuse. In: Galanter M, editor. Recent Developments in Alcoholism, Vol. 10: Alcohol and Cocaine: Similarities and Differences. New York: Plenum Press, 1992;343-61. 33. Bertolet BD, Freund G, Martin CA, Perchalski DL, Williams CM, Pepine CJ. Unrecognized left ventricular dysfunction in an apparently healthy cocaine abuse population. Clin Cardiol 1990;13:323-8. 34. Gawin FH. Cocaine addiction: psychology and neurophysiology. Science 1991;251:1580-6. 35. Tardiff K, Gross E, Wu J, Stajic M, Millman R. Analysis of cocaine-positive fatalities. J Forensic Sci 1989;34:53-63. 36. Dressler FA, Malekzadeh S, Roberts WC. Quantitative analysis in amounts of coronary, arterial narrowing in cocaine addicts. Am J Cardiol 1990;65: 303-8. 37. Karch SB, Billingham ME. The pathology and etiology,of cocaine-induced heart disease. Arch Pathol Lab Med 1988;112:225-30. 38. VanDette JM, Cornish LA. Medical complications of illicit cocaine use. Clin Pharm 1989;8:401-11. 39. Division of Epidemiology and Prevention Research, NIDA. National Household Survey on Drug Abuse: Main Findings 1990. Rockville (MD): USDHHS, 1991:no.(ADM)91-1788. 40. Anda RF, Waller MN, Wooten KG, Mast EE, Escobedo LG, Sanderson LM. Behavioral risk factor surveillance, 1988. Morbid Mortal Week Rep 1990; 39:1-21. 41. Brunswick AF. Health and substance use behavior: the Longitudinal Harlem Health study. J Addict Dis 1991;11:119-37.