The burden of smoking-attributable mortality among African Americans — Indiana, 1990

Addictive Behaviors. Vol. 20, No. 5. pp. 563-569, 1995 Copyright 0 1995Elsevier Science Ltd Printed in the USA. All rights reserved 0306-4603/95$9.50 + .OO

0306-4603(95)00016-X

THE BURDEN OF SMOKING-ATTRIBUTABLE MORTALITY AMONG AFRICAN AMERICANS - INDIANA, 1990 SETH L. EMONT,

SUSAN M. DORRELL, and ROGER McCLAIN

KELLY

Indiana State Department of Health, Indianapolis,

BISHOP,

IN

- Recent trends in smoking prevalence in Indiana indicate a slow decline in smoking prevalence among adults aged I8 years of age and under. Despite this decline, the smoking prevalence has remained consistently higher among Blacks compared to Whites. The purpose of this investigation was to evaluate the extent of mortality due to cigarette smoking in 1990 among Blacks and Whites in the state of Indiana. Estimates of smokingattributable mortality (i.e., smoking-attributable mortality [SAM] and years of potential life lost [YPLL]) associated with smoking in 1990 were calculated using SAMMEC, a software program developed by the Centers for Disease Control and Prevention and designed to measure the disease impact associated with cigarette smoking. Although the SAM and YPLL for Whites accounted for the greatest percentage of total SAM and total YPLL, SAM rates and YPLL rates were higher among Blacks compared to Whites, for both men and women. The SAM rate for Black men was 4% higher than that of White men and 9% higher for Black women compared to White women. Similarly, the YPLL to life expectancy rate for Black men was 7% higher than the YPLL rate for White men and 29% higher for Black women compared to White women. The greater burden of tobacco-related mortality among Blacks is evidenced through higher SAM and YPLL rates compared to Whites. It is imperative that high-risk populations be targeted at national, state, and local levels through antitobacco campaigns. Abstract

INTRODUCTION

Smoking is the leading cause of premature mortality in the United States. Estimates since 1985 indicate that hundreds of thousands of Americans have died from smoking-attributable diseases (Centers for Disease Control [CDC], 1991, 1993b; U.S. Department of Health and Human Services [DHHS], 1989). Although prevalence rates continued to decline in the United States at about .5 percentage points each year, over the past 2.5 years smoking prevalence rates have remained consistently higher for males than females and, with the exception of 1990, for Blacks compared to Whites (DHHS, 1989; CDC, 1992a). In addition, national smoking prevalence estimates remained virtually unchanged between 1990 and 1991 (CDC, 1993a). These gender- and race-specific differences in smoking prevalence are reflected in differences in smoking-attributable mortality (SAM) and years of potential life lost (YPLL) rates. In 1988, the SAM rate for men was more than twice the rate for women and the SAM rate for African Americans was 12% higher than that for whites. Additionally, the rate of smoking-attributable YPLL before age 65 for men was about three times that for women and twice as high for African Americans as for whites (CDC, 1991). The results of this evaluation were presented at the Ninth National Conference on Chronic Disease Prevention and Control, Centers for Disease Control and Prevention, Washington, DC, December S-8, 1994. Requests for reprints should be sent to Dr. Seth Emont, Director, Epidemiology and Chronic Diseases, Johnson & Johnson Health Care Systems Inc., Division of Health Management. 410 George Street, New Brunswick, NJ 08901. 563

564

S. L. EMONT et al.

The Smoking-Attributable Morbidity and Mortality and Economic Costs (SAMMEC, version 2.1) software developed by the Centers for Disease Control and Prevention (CDC) is a useful tool designed to measure the disease impact associated with cigarette smoking (CDC, 1992b). This software is available to state health departments to assist them in monitoring the health and economic consequences of smoking. The CDC has used SAMMEC to track the disease burden associated with smoking at the national level (CDC, 1991, 1993b; DHHS, 1990), and similar methodology has been applied to document smoking-attributable mortality in other countries, as well (DHHS, 1992a). National and state-specific prevalence estimates of high-risk behaviors, such as smoking, and their subsequent impact on chronic illness are useful in tracking progress in reducing the disease burden in the United States and in identifying target populations for implementation of specific intervention programs at the state and county level. Federal-, state-, and county-based public health agencies, alike, have developed plans for tracking chronic diseases through the Year 2000 (DHHS, 1991a, 1991b, 1992b). While there is variation in state initiatives addressing chronic illness and high-risk behavior, tobacco use is clearly a cross-cutting, priority issue addressed by the majority of state public health plans (DHHS, 1992b). Recently, the CDC published state-specific estimates of SAM for 1992. However, race-specific state estimates of SAM have not been widely published and are typically limited to sex-specific breakdowns (CDC, 1993d). The current study of race differences in smoking-attributable mortality was evaluated in the state of Indiana. Little documentation is available on smoking-attributable mortality in the state of Indiana. As in other states, recent trends in smoking prevalence in Indiana from Behavioral Risk Factor Surveillance data (BRFSS; see the Method section), indicate a slow decline in smoking prevalence between 1984 and 1992 among adults aged 18 years or older, with a consistently higher prevalence among Blacks compared to Whites (Figure 1) (Indiana State Department of Health, 1993a). This reflects national estimates in smoking prevalence (CDC, 1992a; DHHS, 1989). Although the proportion of Blacks is lower in Indiana compared to national estimates - 7.8% in Indiana (Indiana State Department of Health, 1993b) vs. 12.3% in the general U.S. population (DHHS, 1992~) - mortality rates for some chronic illnesses and prevalence of high-risk behaviors are higher than those of the general population. For example, the Northwestern National Life Insurance Company’s overall ranking of general health status in Indiana decreased from 1992 to 1993. This state experienced one of the largest negative changes in overall health due primarily to an increase in smoking prevalence (Northwestern National Life Insurance Co., 1993). The purpose of this investigation, then, was to compare the burden of tobaccorelated mortality among Blacks and Whites in 1990 in the state of Indiana. Such information underscores the importance of the implementation of targeted, statebased smoking cessation and prevention programs and can serve as a model for continued tobacco use surveillance among high-risk populations in other states. METHODS

Estimates of SAM and YPLL associated with smoking in 1990 were calculated using SAMMEC 2.1 (CDC, 1992b). SAM was calculated using relative risk estimates

Smoking-attributable

mortality -

565

Indiana 1990

50

& 40 (P E i? $j 30 P 20

10

0

L

1984

Fig. 1. Percentage

1988

1988

of adult current smokers by race -

1990

1992

Indiana, 1984-1992.

for 22 adult (aged 35 years and older) smoking-related diseases and relative risk estimates for four perinatal (aged under one year) conditions and is based on the smoking-attributable fraction for these diseases. The smoking-attributable fraction was calculated using the following formula: [PO + PIW~I)

+ P2(RR2)1 -

1

[PO + PIWRI) + P2w32>1

where po, pl, and p2 represent the percentage of never, current, and former smokers, respectively, in the group under study, and RR1and RR2represent the relative risk of death for current and former smokers, respectively, compared to that of never smokers. Relative risk estimates were obtained from the first 4 years of follow-up (19821986) of the American Cancer Society’s Cancer Prevention Study II (Stellman & Garlinkel, 1986). General disease categories include neoplastic, cardiovascular, respiratory, and pediatric diseases. Mortality data for 1990 were obtained from the Public Health Statistics Division of the Indiana State Department of Health. Data on deaths caused by cigarette-ignited fires were estimated by taking 50% of total burn deaths, a conservative estimate of cigarette-related deaths (CDC, 1992b). Current and former smoking prevalence estimates in 1990 for adults aged 35 years or older and for women aged 18-44 years were obtained from the BRFSS. The BRFSS is a statewide survey conducted monthly on a representative sample of the state’s adult population. The survey collects self-reported information on the prevalence of personal health practices and behaviors related to the leading causes of death, including cigarette smoking, physical activity, high blood pressure control, and alcohol use (Remington et al., 1988). Respondents to the BRFSS were asked:

566

S. L. EMONT et al.

Table I. Estimated

smoking-attributable mortality and years of potential life lost (YPLL) by race and sex - Indiana, 1990 Smoking-attributable

mortality

Smoking-attributable YPLL to life expectancy

Race

Men

Women

Pediatric

Total”

Men

Whites Blacks Other Total populationa

6,270 468 IO 6,748

2,908 237 2 3,146

40 I2 0 52

9,218 717 I2 9,946

70,305 5,359 Il.5 75,779

Women 38,083 3,719 33 41,755

Pediatric

Total”

2,968 800 0 3,768

111,276 9,878 I48 121,302

“Sums may not equal total because of missing values for race.

“Have you smoked at least 100 cigarettes in your entire life?” and “Do you smoke cigarettes now?” Current smokers were defined as those who reported smoking at least 100 cigarettes and who were currently smoking, and former smokers were defined as those who reported having smoked at least 100 cigarettes and who were not smoking now. Smoking-attributable YPLL was determined by multiplying the number of deaths by YPLL to life expectancy by the smoking-attributable fraction for each diagnosis. Age-adjusted SAM and YPLL rates were standardized to the 1990 U.S. population. RESULTS

In 1990, an estimated 9,900 deaths were attributable to smoking. Approximately 9,200 deaths occurred among Whites and 700 deaths occurred among Blacks (Table 1). The majority of deaths (63%) occurred among White men. For both Whites and Blacks, the majority of smoking-attributable deaths were due to cardiovascular diseases (43% and 43%, respectively), followed by neoplasms (35% and 42%), respiratory diseases (21% and ll%), and perinatal and other diseases (1% and 4%; Figure 2). 3,263

(35%)

,95

303 (42%)

(1%)

African Americans

Whites Disease Category 0 Neoplasms

q Cardiovascular

Fig. 2. Smoking-attributable

0 Respiratory

q PerinataVOther

mortality by disease category and race -

Indiana, 1990.

Smoking-attributable

Table 2. Age-adjusted

mortality -

smoking-attributable (S-A) mortality ratesa and years of potential life lost to life expectancy (YPLL) rates by race and sex - Indiana, 1990 S-A mortality rate

Whites Blacks Other Total population

561

Indiana 1990

S-A YPLL rate

Males

Females

Both

Males

Females

Both

587.8 613.0 126.3 573.0

229.4 250.1 31.4 223.1

392.4 412.2 75.7 381.9

6400.3 6876.7 1523.0 6322.8

2923.7 3777.8 375.4 2934.2

4521.3 5 166.8 912.0 4490.7

“Per 100,000 persons ~35 years old (adjusted to the 1990 U.S. population).

Smoking-attributable YPLL to life expectancy by race and sex reflected the patterns observed for SAM. Approximately 121,000 YPLL to life expectancy resulted from cigarette smoking in 1990. The majority of YPLL was attributed to Whites (111,276 YPLL), followed by Blacks (9,878 YPLL) and other races (148 YPLL). As with SAM, the majority of YPLL was attributed to White men (58% of total YPLL; Table 1). Although the SAM and YPLL for Whites accounted for the greatest percentage of total SAM and total YPLL, SAM rates and YPLL rates were higher among Blacks compared to Whites, for both men and women (Table 2). The SAM rate for Black men was 4% higher than that of White men (z = 2.43, p < .05, one-tailed), and 9% higher for Black women compared to White women (z = 1.84, p < .07, one-tailed). Similarly, the YPLL to life expectancy rate for Black men was 7% higher than the YPLL rate for White men (z = 9.06, p < .OOl, one-tailed), and 29% higher for African American women compared to White women (z = 16.65, p < .OOl, onetailed). In addition, overall SAM rates were 2.6 times higher among men compared to women, while YPLL rates were 2.2 times greater among men compared to women. DISCUSSION

Although smoking prevalence continues to decline slowly in the state of Indiana, the aftermath of high smoking prevalence rates within the past four decades is evidenced by the relatively high death rates from smoking today. Each year, the economic burden of smoking-related illness is estimated at $1.4 billion in the state of Indiana (Emont, 1993). It is important to note that the SAM and YPLL estimates presented are conservative estimates since they are based on prevalence data for 1990, and observed deaths actually reflect higher smoking prevalence rates among older birth cohorts. Additionally, this estimate does not include lung cancer and cardiovascular disease deaths resulting from passive smoking. Despite the greater percentage of deaths attributed to smoking among Whites compared to Blacks, the greater burden of tobacco-related mortality among Blacks is indicated through higher SAM and YPLL rates compared to Whites. This underscores the need for more targeted smoking-control interventions, since these racial differences can be explained, in part, by higher smoking rates among Blacks compared to Whites. For example, the 1991 National Health Interview Survey estimates of smoking prevalence indicated a significantly higher prevalence among African Americans compared to whites - 29.2%, 95% CI: 27.7 - 30.7 vs. 25.5%, 95% CI: 24.9 - 26.0, respectively) (CDC, 1993a). Additionally, the higher SAM and YPLL

568

S. L. EMONT et al.

rates among African Americans suggests earlier onset of smoking-attributable disease compared to whites (CDC, 1991). The predominately higher YPLL rates experienced among Blacks, particularly among women, may reflect a higher prevalence of preventable risk factors early in life, more advanced disease at diagnosis, delay in treatment, or other factors related to health care access and quality (CDC, 1992~). A number of community-based tobacco-control initiatives that target minority groups have been launched, including the Office on Smoking and Health’s National Tobacco Prevention and Control Program’s Initiatives to Mobilize for the Prevention and Control of Tobacco Use (IMPACT, 1994), the Robert Wood Johnson Foundation’s (1994) SmokeLess States program, and the American Stop Smoking Intervention Study for Cancer Prevention (ASSIST), cosponsored by the National Cancer Institute and the American Cancer Society (Gruman and Lynn, 1993). The Indiana State Department of Health is engaged in ASSIST and has identified a number of key initiatives to reduce the prevalence of smoking in the state. As part of the policy advocacy initiative, ASSIST-Indiana is targeting African American populations to consider alternative sponsorship of community-based activities, such as health fairs, which have typically been sponsored by tobacco and alcohol companies. It is imperative that high-risk groups, such as minority populations, continue to be targeted through antitobacco campaigns. Data from the Center for Disease Control’s (CDC) 1991 National Health Interview Survey indicate that Blacks are less likely than either Hispanics or Whites to maintain abstinence after quitting smoking (CDC, 1993~). Substantial reductions in overall morbidity and mortality due to smoking can be realized only through a multicomponent approach to prevent initiation among youth and promote cessation among adults. Tobacco-control strategies that mobilize statewide coalitions, utilize public education campaigns, and address policy advocacy initiatives hold promise in achieving these objectives. REFERENCES Centers for Disease Control. (1991). Smoking-attributable mortality and years of potential life lost United States, 1988. MMWR, 40, 62-71. Centers for Disease Control. (1992a). Cigarette smoking among adults - United States, 1990. MMWR, 41, 354-362. Centers for Disease Control. (1992b). SAMMEC 2.1. Rockville, MD: Author. Centers for Disease Control. (1992~). Years of potential life lost before age 65, by race, Hispanic origin, and sex - United States, 1986-1988. MMWR, 41(SS-6), 13-23. Centers for Disease Control. (1993a). Cigarette smoking among adults - United States, 1991. MMWR, 42, 230-233. Centers for Disease Control. (1993b). Cigarette smoking-attributable mortality and years of potential life lost - United States, 1990. MMWR, 42, 645-649. Centers for Disease Control. (1993~). Smoking cessation during previous year among adults - United States, 1990 and 1991. MMWR, 42, 504-507. Centers for Disease Control. (1993d). [State-specific estimates of smoking-attributable mortality, 1990.1 Rockville, MD: Office on Smoking and Health. Unpublished raw data. Centers for Disease Control. (1994). National Tobacco Prevention and Control Program’s initiatives to mobilize for the orevention and control of tobacco use (IMPACT). Rockville, MD: Author. Emont, S. L. (1993):Health and economic harden of smoking: The public he&h henejit of increusing tobacco taxes in Indiana. Indianapolis, IN: Indiana State Department of Health, Division of Chronic Diseases. (Unpublished manuscript). Gruman, J., Lynn, B. (1993) Worksite and community intervention for tobacco control. In C. T. Orleans & J. Slade (Eds.), Nicotine addiction: Principles und management (Chapter 21). New York: Oxford University Press, Inc. Indiana State Department of Health, Division of Health Education. (1993a). [Smoking prevalence estimates from the Behavioral Risk Factor Surveillance System.] Unpublished raw data. Indiana State Department of Health, Division of Vital Statistics. (1993b). [Age- and race-specific population estimates, Indiana, 19901. Unpublished raw data.

Smoking-attributable

mortality -

Indiana 1990

569

National Life Insurance Company. (1993). The NWNL SIUIC hculth runkings: Resrrlrs, and discussion. Minneapolis, MN: Author. Remington, P. L., Smith, M. Y., Williamson, D. F., Anda, R. F., Gentry, E. M., Hogelin, G. C. (1988). Design characteristics, and usefulness of state-based behavioral risk factor surveillance: 1981-1987. Northwestern

methodology,

Public Health Reports,

103(4), 366-376.

The Robert Wood Johnson Foundation.

(1994). 19 states receive $10 million to fight tobacco use. NeH,s New Brunswick, NJ: Author. Stellman, S. D., & Garfinkel, L. (1986). Smoking habits and tar levels in a new American Cancer Society prospective study of 1.2 million men and women. Journul of the Nationcll Cancer Institute. 76, 1057-1063. United States Department of Health and Human Services. (1989). Reducing the he&h c’onsequences of smoking: Twenty-jive yeurs of progress. A report of‘ the Surgeon General. (DHHS publication no. [CDC] 89-841 I .) Rockville, MD: Author. United States Department of Health and Human Services. (1990). Smoking und health: A nutionul ~tutus report. (DHHS publication no. [CDC] 87-8396 [Rev. 01/90]). Rockville. MD: Author. United States Department of Health and Human Services. (199la). Herr/thy communities 2000: Model standards (3rd ed.). Washington, DC: Public Health Service. United States Department of Health and Human Services. (1991b). Heulrhy people 2000: Nurionul heulth promotion and disease prevention objecriues. (DHHS publication no. [PHS] 91-50212.) Washington, DC: Public Health Service. United States Department of Health and Human Services. (1992a). Smoking und heulrh in the Americu.5. (DHHS publication no. [CDC] 92-8419). Atlanta. GA: Author. United States Department of Health and Human Services. (1992b). Heulthy prtiple 2000: Sture uc,rion. (USDHHS, PHS. Publication no. 324-482160426). Washington, DC: Author. United States Department of Health and Human Services. (1992~). He&h, United Stures. 1992. Washington, DC: National Center for Health Statistics. United States Department of Health and Human Services. from the Roberr Wood Johnson

Foundufion.