Tobacco Documents Online

 r I I I a I I I I I I ka~tr: Closing the Cap for Cardiovascular Disease Presenter: Dennis D. Tolsma, M.D., M.P.H. Director Center for Health Promotion and Education, CDC Project officer: Craig C. White, M.D. Medical Epidemiologist Behavioral Epidemiology and Evaluation Branch Center for Health Promotion and Education, CDC Consultants: Manning Feinlieb, M.D., M.P.H. Director National Center for Health Statistics Suzanne G. Haynes, Ph.D. Research Associate Professor of Epidemiology School of Public Health University of north Carolina at Chapell Hill Mary Jane Jesse, M.D. Deputy Director of Reserch American Heart Association Dan L. McGee, Ph.D. Senior Statistician Agent Orange Projects, CDc Dan McGee, Jr. Programmer The Carter Center "Closing the Gap" Health Policy Project, CDC Jeffrey M. Newman, M.D., M.P.H. Medical Epidemiologist Behavioral Epidemiology and Evaluation Branch Center for Healtlh Promotion and Education, CDC Leonard Syme, Ph.D. Professor of Epidemiology School of Public Health University of California at Berkeley H.A. Tyroler, M.D. Professor of Epidemiology School of Public Health University of North Carolina at Chapel Hill - B89 - r

Lawrence Watkins, M.D. Cardiologist Section of Cardiology Medical College of Georgia Craig C. White, M.D. Medical Epidemiologist Behavioral Epidemiology and Evaluation Branch Center for Health Promotion and Education, CDC

^ E ^ E ^ E 7 7-1 ^ a 71 7 E I E 71 ^ 71 CARTER CENTER HEALTH POLICY PROJECT Position Paper: Closing the Gap for Cardiovascular Disease Executive Sunmary Authored by Suzanne G. Haynes, Ph.D, Craig White, M.D., Dennis D. Tolsma, M.P.H., Daniel McGee, Jr, and Jeffrey M. Newman, M.D., M.P.H. EXTENT AND IMPACT OF THE CARDIOVASCULAR DISEASE PROBLEM Today, more than half of all deaths in the United States are attributed to diseases of the heart and vascular system. This paper reviews the status and potential reductions of negative consequences for coronary heart disease (CHD), cerebrovascular disease (stroke), and total cardiovascular disease (all forms of circulatory disease, including CHD and stroke.) Although heart disease and stroke have been the leading and third leading causes of death, respectively, over the period 1940-1980, a significant decline has occurred in the rates of these diseases over the last 16 years. Between 1968 and 1979, the noncardiovascular disease mortality rate declined by 12 percent, while CHD dropped by 27 percent and stroke dropped by 40 percent. Nevertheless, cardiovascular diseases (CVD) continue to contribute significantly to the burden of death, illness, disability, and economic costs in the United States. -B91-

On the basis of 30 years of research, a great deal is known about the risk factors of CVD. Of the many risk factors that have been studied, at least six have come to ba considered standard risk factors for CVD: age; male sex; cigarette smoking; serum cholesterol; systolic or diastolic blood pressure; and glucose intolerance. This paper presents specific estimates of the amount of CHD, stroke, and total CVD that is attributable to changes in each of three risk factors: smoking; elevated serum cholesterol (greater than 219 mg/dl); and elevated systolic blood pressure (greater than 139 mmHg). Specific estimates for three other factors, exercise, diabetes, and obesity, will be added later. The tables of data accompanying this paper document a number of important differences in the distribution of CVD in subgroups of the population. Age-adjusted death rates show that males are at higher risk than females, and blacks are at higher risk than whites. Hence, black males are the race/sex group at highest risk of CVD. In general, blacks have about the same death rates from CHD as whites, but an almost two-fold higher death rate from stroke. Death rates rise steadily from age 35 onward; after age 45, the rates rise about 2 1/2-fold from each 10-year age group to the next. Many of these deaths are premature. One way to quantify the prematurity of death is "potential years of life lost" before the age of 65. For example, a death at age 60 represents 5 potential years represents 20. Nearly 2.5 million years of life lost, one at age 45 of life are lost prematurely because of CVD; CHD accounts for 1.4 million years, while stroke adds 0.3 million. Deaths among males contribute 70 percent of these life years lost. - B92 -

I  I I 4  I I I I I There are no national data with which to monitor either incidence (new cases occurring in a year) or prevalence (the amount of disease existing at a point in time) of CVD. This paper presents estimates of period prevalence of CVD for 1980. although CVD mortality rates have been declining, it appears that the prevalence has increased between 1972 and 1980. Approximately 48 million Americans suffered from some form of cardiovascular disease in 1980. Not suprisingly, morbidity of this magnitude is associated with very large expenditures for personal health care. Expenditures for medical care for heart diseases totaled over $14 billion in 1980, along with $5 billion for stroke. The total medical care expenditures for CVD exceeded $33 billion in that year. POTENTIAL IMPACT OF ELIMINATING CARDIOVASCULAR RISK FACTORS The three risk factors for which estimates are presented in this paper make a major contribution to cardiovascular disease rates. In order to compute the number of deaths or cases of CV1) attributable to each risk factor, we calculated the Population Attributable Risk Fraction (PARF) for each risk factor. -B93- 9

Simply stated, this statistic is the percentage of total events (e.g., deaths) in a population that are attributable to a particular risk factor. Hence, PARF can be interpreted from an etiologic point of view--the causal outcome of a risk factor-or from a prevention view point--the events that would not occur if the risk factor were eliminated. The size of the percentage is influenced by two things: The magnitude of the relative risk, and the prevalence of the risk factor in the population. The larger the relative risk, the larger the PARF, all other things being equal. Similarly, the larger the percentage of the population with that risk factor, the larger the PARF. For example, a very powerful risk factor would have a large relative risk. However, if only a few persons have that risk factor, it would only contribute to a small fraction of cardiovascular deaths. Conversely, even if a risk factor has only a moderate relative risk, but many persons have it, the risk factor can contribute to a large fraction of deaths. In determining the attributable risk for smoking, high blood pressure, and elevated serum cholesterol, the following assumptions were made: o high blood pressure: The paper focuses on the risk from defined hypertension (systolic blood pressure over 159 mmHg) as well as borderline hypertension (systolic blood pressure over 139 mmHg.) In the Hypertension Detection and Follow-up Program, substantial reductions in CVD followed treatment of mild hypertension. o elevated serum cholesterol: The paper defines the risk from elevated serum cholesterol as greater than 219 mg/dl. - B94 -

I I I I I I I I I I I I I r I I I I o cigarette smoking: Since the purpose of the analysis is to define the total CVD attributable to a risk factor, we calculated the PARF assuming the elimination of cigarette smoking. The Population Attributable Risk Fractions presented in this report are based on manipulations of logistic regression equations derived from a CDC-sponsored report by Dr. Lester Breslow and colleagues. Basically, equations from major CVD studies conducted during the past 30 years were pooled to develop a series of equations--for men and women, for whites and blacks, for MI morbidity and mortality and for stroke morbidity and mortality. Prevalence estimates for the three risk factors (by age, race, and sex) are inserted in the equation, and CVD mortality or morbidity outcomes are calculated for those distributions of the risk factors. Outcomes are then recalculated separately assuming the elimination of one of the risk factors. The PARF for that risk factor is the difference between the two outcomes, divided by the CVD mortality or morbidity outcome from the first calculation. The PARF is thus a fraction. We computed it separately for white males, black males, white females, and black females for the age groups 25-44, 45-64, and 65 and older. The actual number of deaths attributed to a risk factor is computed by multiplying that fraction and the 1980 deaths in each age/sex/race group. A number of assumptions must be made in using these equations, which are reviewed in the paper. Smoking Attributable Risk. While smoking has declined overall during the past 15-20 years, this decline masks an increase in the number of cigarettes consumed per smoker and an increase in the prevalence of smoking among women. -B95- IN

A significant portion of the public understands that smoking is harmful--for example, that is causes cancer--but fewer seem to understand that the number of cardiovascular deaths due to smoking actually exceeds smoking-related cancer deaths. A significantly greater proportion of CHn deaths can be attributed to smoking in males, particularly black males, than in females. The PARF for smoking is 33 percent for black males and 21 percent for white males; it is 22.2 percent for males as compared to 3.7 percent for females. This relationship is similar for CHD morbidity, but the 2-fold difference between males and females is less pronounced. Overall, 14 percent of CHD deaths, or about 78,000 deaths, are attributable to smoking. Similarly, cigarette smoking accounts for 14 percent of CHD morbidity, more than three-quarters of a million cases. For smoking and stroke, the PARFs for male morbidity and mortality are almost identical to those for CHn. Overall, about 11 percent, or 240,000 cases, of stroke could be prevented if smoking were eliminated. Smoking is responsible for a total of 145,319 cardiovascular disease deaths in 1980. In excess of 7 million cases of CVD can be attributed to smoking. CVn cases attributable to smoking are not substantially different between white men and white women; however, there are about 80 percent more cases among black men than among black women.

 I  I I I I I I I I I I I I I I I Years of life lost due to smoking-related CVD show some striking differences. Overall, smoking accounts for 15 percent of CVD deaths but 28 percent of CVD lifeyears lost--more than 700,000 person years of premature death. the largest percentage differences are among white women and white men, but the highest PARF--36 percent--was for black men. There is little difference in CHD and stroke fractions for men, but there is a fourfold larger PARF for stroke among both white and black females compared to CHD. All four race/sex groups have very similar patterns. Hypertension Attributable Risk. Reduction of systolic blood pressure for all hypertensives and borderline hypertensives to 139 mmHg or less has dramatic impact on cardiovascular disease mortality. ADproximately 29 percent of CHn deaths, 32 percent of stroke deaths, and 30 percent of total CVD deaths are attributable to high blood pressure. There are no substantial differences among any race/sex group in any of these categories.. The number of CVD deaths averted if high blood pressure were eliminated is 292,504. Some 148,988 of these occur among men, and 142,514 among women. More (164,837) are CHD deaths, and 54,642 are stroke deaths. than half Thc population attributable risk fractions for potential years of life lost are virtually identical for CHID, stroke, and total CVD-about one-fifth of all years of life lost are attributable to high blood pressure, a total of 488,233 years of life annually. However, there is a a striking difference between racial groups. Compared to whites, the PARF for blacks is more than 40 percent higher. - B97 - 

With regard to hypertension-attributable morbidity, the fractions for blacks are again higher than for whites, but only modestly so. Stroke morbidity is consistently higher than CHD for all race/sex groups. More than 7 million cases of CVD cam be attribIted to blood pressure greater than 139 mmHg, and with them nearly 10 million hospital days, 155 million disability days, and $6.3 billion in expenditures for personal medical care. Cholesterol-attributable risk. Ten percent of the nearly 1 million CVD deaths that occur each year is attributable to serum cholesterol greater than 219 mg/dl. Only 5 percent of the CHD deaths and 3 percent of the stroke deaths among men are attributable to elevated serum cholesterol. However, among women, the comparable figures are 19 percent of CHD deaths and 8 percent of the stroke deaths. Thus, 80 percent of the deaths attributable to eleveated ~serum cholesterol occur among women. These tend to be among older persons, so cholesterol accounts for a smaller fraction (9 percent) of potential life years lost than for total mortality. This is especially true for stroke, where cholesterol accounts for only 4 percent of potential lifeyears lost. Population attributable risk fractions for CHD morbidity are much higher than for stroke morbidity-CHD accounts for 22 percent, or 1,162,248 cases of CHD, while stroke only accounts for 2 percent, or 51,724 cases. There is little difference by race or sex in the PARFs. Yet, there are very large differences in hospitalization days and disability days for women compared to men, both for CHD and for total CVD. Elevated serum cholesterol accounts for 184 million or 20 percent of all CVD disability days. Women account for 76 - B98 -