- Email: [email protected]
Cigarette Smoking Among Native and Foreign-Born African Americans
Cigarette Smoking Among Native and Foreign-Born African Americans GARY KING, PhD, ANTHONY P. POLEDNAK, PhD, ROBERT BENDEL, PhD, AND DEBORAH HOVEY, MSW
PURPOSE: To examine differences in current smoking status and the number of cigarettes consumed daily between foreign and native-born African Americans, and the impact of demographic and socioeconomic status (SES) factors on smoking behavior. METHODS: Data were obtained from combining the 1990–1994 National Health Interview Surveys and consisted of 16,738 U.S. born and foreign-born African Americans between 18 and 64 years of age. The statistical analysis included cross-tabulations and weighted multiple logistic regression (MLR) using the Statistical Packages for the Social Sciences (SPSS) and the Survey Data Analysis (SUDAAN) computer programs. RESULTS: Adjusted MLR analysis revealed that native-born African Americans were more likely (odds ratio (OR) 5 2.7, p , 0.001) to be current smokers than foreign-born blacks. Within the nativeborn group, smoking prevalence decreased with increasing education and income, but these associations were not found for foreign-born blacks. Women in both groups were less likely than men to be current smokers. Statistically significant differences were not found between the two groups in the number of cigarettes smoked per day. CONCLUSIONS: This analysis of nativity and smoking behavior further demonstrates the social diversity among African Americans and suggests the differential impact of social and cultural factors on smoking behavior within racially classified social groups. In areas where there are substantial numbers of foreign-born blacks, researchers should consider differentiating smoking status by nativity. Though differences in smoking prevalence were apparent for native and foreign-born American Americans, prevention and cessation programs are needed for both groups. Ann Epidemiol 1999;9:236–244. 1999 Elsevier Science Inc. All rights reserved. KEY WORDS:
Cigarette Smoking, African Americans, Blacks, Migrant Health, Minority Health.
INTRODUCTION An important contemporary theme in the study of the health of African Americans is intra-group diversity or social heterogeneity (1–3). Socio-demographic variables such as age, gender, geographic region, and socio-economic status (SES) help to explain variations in health behavior and outcomes (4–8). Nativity or country of birth also has been shown to be associated with disease risks and health status among minority groups such as Hispanics and Asian Americans (9–11). In the 1990 U.S. Census, it was estimated that 5.8% of African Americans were foreign-born (12) and more current
From the Department of Biobehavioral Health (G.K.), Penn State University, University Park, PA; Department of Community Medicine and Health Care (Adjunct) (A.P.P.), University of Connecticut Health Center, Farmington, CT; Center for Environmental Health (R.B.), University of Connecticut, Storrs, CT; and Department of Community Medicine and Health Care (D.H.), University of Connecticut Health Center, Farmington, CT. Address reprint requests to: Gary King, Ph.D., Department of Biobehavioral Health, 315 E. Health and Human Development, Penn State University, University Park, PA 16801. Received May 19, 1998; revised September 08, 1998; accepted September 09, 1998. 1999 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010
data indicates that this figure may exceed 25% in some urban areas (e.g., New York City and Dade County, Florida) (13, 14). Studies of nativity can be useful in examining various social experiences and influences on the health of African Americans including acculturation, social mobility, cultural practices and social networks, and exposure to (or protective strategies against) racism. Moreover, this research could have important implications for health intervention programs and policies and may suggest multiple strategies for studying African American health. Despite numerous studies on cigarette smoking among African Americans, differences in smoking behavior between foreign and native-born blacks have rarely been investigated (15). In a 1984 study of black pregnant women in Boston, foreign-born women were less likely than nativeborn women to be smokers at any time during pregnancy (16). In a study of New York high school students a lower smoking prevalence rate was found among Caribbean-born compared to native-born African-Americans (17). A study of adults living in New York City in 1992 revealed a lower rate of ever smoking among foreign-born relative to nativeborn African Americans (18). Despite the contributions of these studies, the results are based mostly on data from 1047-2797/99/$–see front matter PII S1047-2797(98)00052-0
AEP Vol. 9, No. 4 May 1999: 236–244
Selected Abbreviations and Acronyms SES 5 socio-economic status MLR 5 multiple logistic regression NHIS 5 National Health Interview Survey CCS 5 Cancer Control Supplement CES 5 Cancer Epidemiology Supplement SPSS 5 Statistical Packages for the Social Sciences SUDAAN 5 Survey Data Analysis
small select samples (i.e., Caribbeans, pregnant women, and students) (16, 17) and/or did not examine current smoking or consumption frequency (18). Using data from the 1990–1994 National Health Interview Survey (NHIS), the major objectives of this paper were: 1) to determine whether differences in smoking status and the number of cigarettes consumed daily exist between foreign and native-born African Americans; and 2) to assess the relative importance of demographic and SES factors on smoking behavior between the two groups. This study of a large sample of African Americans is broadly representative of adults, 18–64 years old, and provides previously unreported information about tobacco addiction and nativity among blacks in the U.S.
METHODS Sample Design The NHIS is a national cross-sectional household sample of non-institutionalized civilians. The survey is approximated by a three stage stratified cluster probability sample design. The response rates of African Americans to each of the NHIS surveys between 1990 and 1994 are not known but probably do not differ substantially from other groups (19). The overall annual response rates were about 95% and information about the NHIS has been previously published (20). The annual unweighted sample sizes are displayed in Table 1 with a cumulative total of 16,738 respondents. In this analysis, the Cancer Control Supplement (CCS) and Cancer Epidemiology Supplement (CES) to the 1992 NHIS were merged to increase the representation of African Americans and the weights of the combined sample were halved as suggested by Botman and Jack (21). For all other years, the analysis was conducted using either the Year 1990 or Year 2000 Objectives supplements. Instrument The term foreign-born refers to persons who identified as black or African American and whose country of nativity is outside the United States or its territories. Persons who selected their racial classification as African American or black and who were between 18 and 64 years of age were
237
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
TABLE 1. Socio-demographic characteristics of U.S.-born Blacks and foreign-born Blacks U.S.-born Foreign-born Total Blacks Blacks N 5 16,738a N 5 15,660a N 5 1078a %b %b %b Age of respondents 18–24 years old 25–34 35–44 45–54 55–64 Gender Male Female Education Some high school High school graduate Some college College graduate or more Employment status Employed Unemployed Not in the labor force Region West South Midwest Northeast Family income Less than $15,000 $15,000 to $19,999 $20,000 to $49,999 $50,000 or more Marital status Married Unmarried a b
19.5 29.4 24.1 15.3 11.8
19.8 28.8 23.9 15.4 12.1
15.7 36.4 26.6 13.7 7.5
45.3 54.7
45.0 55.0
48.5 51.5
24.1 41.8 22.1 12.0
24.3 42.6 21.8 11.3
21.7 32.2 25.9 20.2
67.2 6.4 26.3
66.6 6.5 26.9
74.6 5.7 19.6
9.7 53.2 20.8 16.4
9.8 55.2 22.0 13.0
8.1 28.8 6.7 56.5
34.3 12.9 39.2 13.5
35.2 13.1 38.5 13.2
23.6 11.1 48.0 17.3
45.0 55.0
44.2 55.8
54.0 46.0
National Health Interview Survey, 1990–1994 unweighted data. Percentages are the weighted prevalence rates.
included. The NHIS does not include questions specifying the particular country of birth. However, immigration data show that between 1994 and 1996, immigrants from Caribbean and African countries, respectively comprised 72.3% and 27.7% of persons from these geographic regions. In this paper, racial classification is viewed as a social variable rather than a phenotypic indicator of biologic or genetic differences between human populations. Also, because only a relatively small proportion of African American smokers are 65 years or older, the analysis was restricted to ages 18–64. Current smoking status was the primary dependent variable and consisted of persons who were either current smokers or nonsmokers (former and never smokers). In 1990, current smokers were defined as persons who: 1) reported smoking at least 100 cigarettes in their entire life; and 2) were presently smoking cigarettes. In 1991–1994, current smokers included persons who: 1) reported smoking at least
238
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
100 cigarettes in their entire life; and 2) currently smoked either “everyday” or “some days.” Data for 1990 were included despite the change in the definition of current smoking because the impact of the change was small. In a separate analysis it was found that excluding the additional someday smokers reduced the increase in the smoking prevalence rate of African Americans from 3.7% to 3.3% between 1990 and 1991 (5). Furthermore, excluding the 1990 data would have reduced the sample of foreign-born blacks. The second response variable was the number of cigarettes smoked per day by all current smokers in 1990 and by “everyday” smokers in 1991–1994. It was dichotomized as fewer than 15 cigarettes and 15 or more cigarettes per day. Most African Americans (64%) smoked fewer than 15 cigarettes per day so that the 15 or more category defines heavier smokers (5, 11). It was not possible to investigate fully the age of smoking initiation among this sample of NHIS respondents because this variable is only available for the year 1992. An analysis of the 1992 data indicates that the initiation rate among African American adults is greatest among the age group 18–24 (8). The independent variables selected for analysis were age, gender, years of education, annual family income, geographic region, nativity, and length of U.S. residence. To construct the education variable, the completed number of years of education was divided into four categories: 1) less than high school (, 12 years); 2) high school graduate (12 years); 3) some college (1–3 years); and 4) four or more years of college. The annual family income groups were: , $15,000, $15,000–$19,999, $20,000–$49,999, and $50,000 or more. A regional variable based on the U.S. Census definition (i.e., Northeast, Midwest, South, West) was included to assess geographic variations. Nativity refers to persons who were born within the U.S. (native-born) and those born outside of the U.S. (foreign-born). Length of U.S. residence pertains solely to foreign-born persons and includes the categories of , 15 years and 15 years or more. The NHIS response categories for this variable are , 1 year, 1 year–less than 5 years, 5 years–less than 10 years, 10 years–less than 15 years, and 15 years or more. Our analysis did not reveal statistically significant differences between the individual categories of , 15 years and the dependent variables of current smoking and number of cigarettes smoked daily. As a result, we combined the categories of , 1 year to , 15 years of residence to create a variable , 15 years vs. 15 years or more.
AEP Vol. 9, No. 4 May 1999: 236–244
the prevalence rates, logistic regression coefficients, and odds ratios (ORs) reflect the population relationships and are obtained for each subject from the inverse probability of selection with modification for oversampling and missing data. Weights were rescaled (for only SPSS computations) so that the sum of the weights was equal to the sample size instead of the subpopulation size. Rescaling of the weights using SPSS is likely to yield p-values that are too small as it does not take into account the clustering in the NHIS design. Consequently, the SPSS computations were used only as a tool to identify potential associations, i.e., for model building purposes. Indicator coding was used to create single degree of freedom components for categorical variables with more than two categories. Both SPSS and SUDAAN treated all of the single degree of freedom components associated with a main effect or interaction collectively, so that p-values and likelihood ratio tests applied to the “entire” main effect or interaction. In presenting the results, ORs were computed as OR 5 exp (b) where b is the estimate of the logistic regression coefficient. The 95% confidence intervals for the ORs were computed as: exp (b 6 t (.025) * SE) where SE is the (adjusted) standard error of b and t is the upper 2.5 percentile of the ‘t’ distribution with degrees of freedom (df) from SUDAAN (df 5 # of primary sampling units (PSUs) 2 # of fixed strata for the NHIS multistage design where PSUs are sampled without replacement from the fixed strata) (22). Standard errors were adjusted upward by approximately 18% to account for the increased clustering due to combining samples from 5 years. Computation of exact standard errors is not possible because the NHIS typically samples households in the same areas of the same PSUs and this clustering is not identifiable when combining data from 5 years. Botman and Jack, however, note that the design effect (defined as the ratio of the variance of the estimator under a complex survey design to the variance of the estimator under a simple random sample) for estimates of percentages based on a single year of survey data is about 1.3, and based on three years of combining survey data is about 1.6 (21). Hence, for three years the standard errors would be increased 11% (i.e., 100 * sqrt(1.6/1.3)) and for five years, a reasonable (conservative) extrapolation would be 18% (i.e., 11 1 (2/3) * 11).
Statistical Analysis The statistical analysis included crosstabulations and weighted logistic regression modeling using both the Statistical Packages for the Social Sciences (SPSS) and the Survey Data Analysis (SUDAAN) programs. Weights were used so that
RESULTS The age distribution differed slightly for the two subpopulations (Table 1) and the mean age was 36.8 years for nativeborn and 35.8 years for foreign-born blacks. A higher per-
AEP Vol. 9, No. 4 May 1999: 236–244
centage of foreign-born than native-born blacks was found in the highest income and education categories. Most (55.2%) native-born blacks resided in the South compared to foreign-born blacks who mostly (56.5%) were located in the Northeast. A greater proportion of foreign-born (54%) than native-born blacks (44.2%) were married. The prevalence of current cigarette smoking (Table 2) was more than twice as high for native-born compared to foreign-born African Americans (30.4% vs. 14.1%, p , 0.001). This finding was evident in all categories of age, gender, region, education, income, employment, and marital status. Notably, smoking prevalence was more than 4 times higher in native-born than in foreign-born women. Among both the native and foreign-born groups (Table 2), smoking prevalence was highest among 35–44 year-olds and lowest among 18–24 year olds. Within the native-born group, prevalence varied with education (p , 0.001) and income (p , 0.001) and was highest among lower SES persons (e.g., unemployed, lower income, and non-high school graduates). These associations were not found for foreign-born blacks. The ratio of men to women smokers was much higher among the foreign-born (3.4) compared to native-born respondents (1.4). A significant association was not found in the foreign-born population between length of years of residence and smoking, although those who resided 15 or more years in the U.S. had the highest smoking rate (17.5%; Table 2). The unadjusted logistic regression analysis revealed that native-born blacks were 2.6 times (p , 0.001) more likely to be smokers than foreign-born blacks (Results not shown). Using multivariate logistic regression, this association remained after controlling for the effects of other demographic predictors such as age, education, and gender (OR 5 2.73, p , 0.001; Table 3). In the regression model (Table 3) which included only foreign-born blacks, the significant predictors of smoking status were gender (OR 5 0.16, p , 0.001) and to a lesser extent age. None of the SES variables (i.e., education, income, employment status) were statistically significant predictors of smoking. Length of U.S. residence was marginally significant (p , 0.10) and there was no statistically significant interaction between age and duration of U.S. residence (data not shown). The multivariate analysis (Table 3) showed that the impact of education on smoking among foreign-born blacks was less evident at the lowest levels (i.e., substantial differences were not apparent between high school and non-high school graduates). The results were similar for the Northeast region, where a majority of foreignborn blacks resided. Among current smokers, the mean number of cigarettes consumed per day differed only slightly and not statistically significantly between native-born (14.5) and foreign-born (13.8) blacks. Both native (38.6%) and foreign-born African American women (29.8%) consumed 15 or more ciga-
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
239
TABLE 2. Prevalence of current smokersa in U.S.- and foreignborn Blacks by socio-demographic characteristics of 18–64 year oldsb
Age of respondents 18–24 years old 25–34 35–44 45–54 55–64 Gender Male Female Education Some high school High school graduate Some college College graduate or more Employment status Employed Unemployed Not in the labor force Region West South Midwest Northeast Income Less than $15,000 $15,000 to $19,999 $20,000 to $49,999 $50,000 or more Years in US Less than 15 years 15 years or more Marital status Married Unmarried Total
All Blacks %
U.S.-born Blacks %
Foreignborn Blacks %
14.3c 31.1 36.4 34.4 26.6
14.8c 32.8 38.3 36.0 27.5
6.5d 15.9 17.3 14.2 10.5
34.6c 24.6
35.7c 26.0
22.2c 6.6
38.4c 29.9 24.3 17.0
40.2c 30.8 25.3 17.8
14.2 15.4 14.7 11.7
26.9c 39.6 32.1
28.2c 40.9 33.4
14.1 24.3 11.5
25.4c 28.0 34.8 27.7
26.1c 28.6 35.5 32.8
15.5 15.1 10.4 13.9
35.8c 29.7 25.8 21.8
36.8c 30.9 27.1 22.5
16.7 12.2 13.7 16.6
––
––
12.1 17.5
28.1d 30.0 29.1
29.6 31.1 30.4
14.3 14.0 14.1
a
Current smoking status is defined as the proportion of persons who had consumed more than 100 cigarettes in their lifetime and were smoking either every day or some days at the time of the National Health Interview Survey. b Percentages are the weighted prevalence rates. Note: Statistical significance comparing categories within each variable is denoted by d p < 0.05, c p < 0.001.
rettes daily less frequently than native (52.6%) and foreign born (51.8%) men (Table 4). Age, education, and employment status were significantly associated with heavier smoking among native born but not foreign-born blacks (Table 4). These findings may be due to the smaller size for foreignborn blacks. In multiple logistic regression analysis (Table 5), nativity was not significantly related to heavier smoking (15 or 15 or more cigarettes daily). Within the native-born group,
240
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
AEP Vol. 9, No. 4 May 1999: 236–244
TABLE 3. Logistic regression models of smokinga behavior among U.S.- and foreign-born Blacks, 1990–1994 (lower and upper 95% limits) All Blacks
U.S.-born Blacks
Foreign-born Blacks
1.00 3.46 4.88 3.91 2.34
1.00 4.29 4.62 3.47 2.47
b
Age categories (years) 18–24 25–34 35–44 45–54 55–64 Gender Male Female Education Less than high school High school graduate Some college College graduate or more Household income Less than $15,000 $15,000 to $19,999 $20,000 to $49,999 $50,000 or more Number of years in U.S. 15 years or more Less than 15 years Nativity Foreign-born Blacks U.S.-born Blacks Employment Employed Unemployed Not in the labor force Marital status Married/partnered Unmarried/not partnered Region Northeast South Midwest West
1.00 3.50 4.87 3.90 2.37
(2.83, (3.93, (3.08, (1.87,
4.31)e 5.99)e 4.92)e 2.99)e
(2.80, (3.97, (3.81, (2.29,
4.26)e 6.05)e 3.99)e 2.39)e
(1.11, (1.19, (0.81, (0.54,
16.70)c 17.91)c 14.71) 11.23)
1.00 0.53 (0.47, 0.59)e
1.00 0.55 (0.49, 0.62)e
1.00 0.16 (0.08, 0.33)e
1.00 0.71 (0.60, 0.84)e 0.60 (0.50, 0.72)e 0.36 (0.28, 0.47)e
1.00 0.70 (0.59, 0.82)e 0.58 (0.48, 0.70)e 0.35 (0.27, 0.45)e
1.00 1.04 (0.71, 3.15) 0.81 (0.34, 1.90) 0.53 (0.21, 1.32)
1.00 0.76 (0.63, 0.91)d 0.67 (0.58, 0.77)e 0.61 (0.48, 0.77)e
1.00 0.76 (0.63, 0.91)d 0.67 (0.58, 0.77)e 0.59 (0.47, 0.74)e
1.00 0.90 (0.38, 2.15) 0.91 (0.45, 1.84) 1.14 (0.38, 3.41)
— —
1.00 0.63 (0.35, 1.13)
— — 1.00 2.73 (1.98, 3.81)e
— —
— —
1.00 1.72 (1.36, 2.17)e 1.19 (0.82, 1.72)
1.00 1.70 (1.34, 2.15)e 1.18 (1.03, 1.36)c
1.00 1.77 (0.68, 4.61) 1.18 (0.53, 2.60)
1.00 1.19 (0.25, 5.54)
1.00 1.17 (1.03, 1.31)c
1.00 1.62 (0.88, 2.97)
1.00 0.76 (0.63, 0.92)d 1.06 (0.86, 1.31) 0.76 (0.59, 0.99)c
1.00 0.74 (0.61, 0.89)d 1.05 (0.85, 1.30) 0.75 (0.58, 0.97)c
1.00 1.15 (0.56, 2.37) 0.42 (0.11, 1.55) 0.72 (0.22, 2.31)
a
Current smoking status is defined as the proportion of persons who had consumed more than 100 cigarettes in their lifetime and were smoking either every day or some days at the time of the National Health Interview Survey. b The adjusted logistic regression model is based on weighted data and included the effects of age, gender, education, household income, years in the U.S., nativity, employment status, marital status, and region of residence within the U.S. Note: Statistical significance comparing categories to the reference group is denoted by c p < 0.05; d p < 0.01; e p < 0.001.
adjusted ORs for most foreign-born blacks, age categories and gender were statistically significant. Among foreignborn blacks, there were no statistically significant adjusted ORs as the small number of smokers resulted in wide confidence limits. DISCUSSION This study shows large differences in the smoking behavior of native and foreign-born African Americans. Native-born African Americans were far more likely to be smokers than foreign-born blacks even after controlling for differences in
sociodemographic indicators. Though women in both groups were less likely to be smokers, native-born women were 4 times more likely to be smokers than their foreignborn counterparts. Interestingly, among foreign-born blacks, income was inconsistently related to smoking and neither income nor education were statistically significant predictors of smoking status. This particular finding was also supported by investigators examining SES and ever smoking status among foreign-born blacks (18). Caribbean and sub-Saharan African countries generally have low smoking rates, especially among women, and migrants from these countries to the U.S. appear to have
AEP Vol. 9, No. 4 May 1999: 236–244
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
TABLE 4. Daily smoking frequencya in U.S.- and foreign-born Blacks and socio-demographic characteristics of 18–64 year olds
Age of respondents 18–24 years old 25–34 35–44 45–54 55–64 Gender Male Female Education Some high school High school graduate Some college College graduate or more Employment status Employed Unemployed Not in the labor force Region West South Midwest Northeast Income Less than $15,000 $15,000 to $19,999 $20,000 to $49,999 $50,000 or more Years in U.S. Less than 15 years 15 years or more Marital status Married Unmarried Total
U.S.-born smokers who consume >15 cigarettesb %
Foreign-born smokers who consume >15 cigarettesb %
30.7c 39.3 48.9 53.8 52.8
47.8 44.1 49.9 56.0 19.8
52.6c 38.6
51.8d 29.8
49.5c 44.7 43.8 43.2
34.7 41.2 63.3 40.0
44.9d 47.1 47.9
49.4 42.0 29.2
43.0 45.8 47.8 45.1
53.4 34.2 56.9 50.0
45.7 47.0 42.9 48.6
40.7 52.0 46.9 56.8
— —
47.9 44.4
47.9 44.4 46.0
46.6 45.8 46.2
a
Current smoking status is defined as the proportion of persons who had consumed more than 100 cigarettes in their lifetime and were smoking either every day or some days at the time of the National Health Interview Survey. b Proportion of current smokers in the National Health Interview Survey who consumed 15 or more cigarettes per day. Percentages are the weighted prevalence rates. Note: Statistical significance comparing categories within each variable is denoted by d p < 0.05; c p < 0.001.
retained lower smoking rates than native populations (23– 26). Lower smoking prevalence among Caribbean and African migrants compared to native populations has also been observed in Canada, Great Britain and other European countries (27–31). Differences in smoking rates between foreign and native-born blacks in the U.S., however, could decrease in the future (if migration to the U.S. continues) because of the increasing prevalence of smoking among
241
youth (particularly young males) in Caribbean and African countries; contributing factors are greater economic development and the proliferation of cigarette advertisement (23–26). Also, smoking rates may decline more precipitously among native-born African Americans. The possibility that the most “acceptable” (e.g., highly educated) and healthy individuals more often chose or are allowed to immigrate (32), may have only a small effect on the smoking prevalence among the foreign-born in the U.S. because of the currently low rates of smoking in African and Caribbean countries (23–26). Although our study could not measure explicitly the effects of acculturation, foreign-born blacks (especially women) were less likely than native-born African Americans to smoke cigarettes even after a considerable postmigration period. The highest foreign-born smoking rate, however, was found among long-term residents (151 years), albeit it did not equal the prevalence of native-born blacks and the p value exceeded the threshold of statistical significance in the adjusted logistic regression model. Protective factors operative within foreign-born black communities that promote smoking prevention or cessation may involve religious proscriptions, social mores, and pressures against adopting particular American habits (especially for women) (33, 34). Other related factors may include alternative ways of addressing racism and social stress, smaller and insular immigrant communities, stronger regard for health detriments of smoking, costs of cigarettes (especially among newly arrived migrants), and different cultural orientations (35, 36). Key research issues involve identifying the “protective” factors that maintain lower rates of smoking initiation and whether they can be promoted among nativeborn blacks (and conceivably other racially classified social groups). The finding that smoking was not found to vary significantly with income or education in foreign-born blacks points to the need to examine social organization and cultural norms, sanctions, and support systems. This is a major difference between this group and native African Americans and may be suggestive of the stronger role of cultural and environmental factors. For example, a plausible hypothesis based on cultural socialization theory suggests that the differences in smoking prevalence between native-born and foreign-born African Americans may be partially explained by large numbers of adults in the older age groups among the foreign-born who never smoked cigarettes. This “elder” group may serve as standard bearers of the culture or role models for adolescents and young adults to prevent smoking. Study limitations include the possibility of underestimating smoking prevalence due to undocumented residents and non-respondent bias or sampling error related to undercoverage (e.g., institutionalized populations and “hidden” males). Also, as these NHIS data are cross-sectional surveys, age differences may be influenced by cohort effects. Evidence
242
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
AEP Vol. 9, No. 4 May 1999: 236–244
TABLE 5. Logistic regression models predicting daily smoking frequencya among U.S. and foreign-born Black current smokers, 1990–1994
Age categories (years)b 18–24 25–34 35–44 45–54 55–64 Gender Male Female Education Less than high school High school graduate Some college College graduate or more Household income Less than $15,000 $15,000 to $19,999 $20,000 to $49,999 $50,000 or more Years in U.S. 15 years or more Less than 15 years Nativity Foreign-born Blacks U.S.-born Blacks Employment Employed Unemployed Not in the labor force Marital status Married/partnered Unmarried/not partnered Region Northeast South Midwest West
All Blacks
U.S.-born
Foreign-born
Smokers who consumed >15 cigarettes daily
Smokers who consumed >15 cigarettes daily
Smokers who consumed >15 cigarettes daily
1.00 1.40 2.09 2.51 2.13
(0.94, (1.41, (1.61, (1.40,
2.07)c 3.12)c 3.91)c 3.26)c
1.00 1.42 2.15 2.55 2.25
(0.93, (1.44, (1.64, (1.48,
2.16)c 3.18)c 3.99)c 3.42)c
1.00 2.42 4.95 5.38 0.67
(0.24, (0.45, (0.48, (0.03,
23.80) 54.95) 59.53) 12.90)
1.00 0.55 (0.46, 0.67)c
1.00 0.56 (0.46, 0.67)c
1.00 0.44 (0.13, 1.48)
1.00 0.88 (0.70, 1.11) 0.99 (0.75, 1.31) 0.78 (0.52, 1.16)
1.00 0.88 (0.70, 1.11) 0.94 (0.70, 1.23) 0.80 (0.53, 1.18)
1.00 0.74 (0.15, 3.51) 1.86 (0.32, 10.72) 0.47 (0.06, 3.79)
1.00 1.01 (0.73, 1.40) 0.86 (0.68, 1.09) 1.08 (0.73, 1.61)
1.00 1.01 (0.71, 1.43) 0.85 (0.66, 1.10) 1.05 (0.71, 1.56)
1.00 1.64 (0.16, 16.88) 1.53 (0.40, 5.96) 1.82 (0.21, 15.63)
— —
— —
1.00 1.11 (0.34, 3.64)
1.00 0.99 (0.57, 1.75)
— —
— —
1.00 1.25 (0.88, 1.77) 1.22 (0.94, 1.58)
1.00 1.28 (0.90, 1.82) 1.23 (0.95, 1.60)
1.00 0.56 (0.08, 3.89) 0.59 (0.11, 3.06)
1.00 0.98 (0.78, 1.24)
1.00 0.95 (0.77, 1.17)
1.00 3.19 (0.79, 12.96)
1.00 0.95 (0.71, 1.25) 1.05 (0.76, 1.46) 0.86 (0.59, 1.25)
1.00 0.97 (0.73, 1.28) 1.06 (0.77, 1.47) 0.85 (0.56, 1.30)
1.00 0.30 (0.07, 1.24) 3.99 (0.73, 22.10) 0.97 (0.19, 4.98)
a
Proportion of current smokers in the National Health Interview Survey who consumed 15 or more cigarettes per day. The adjusted logistic regression model is based on weighted data and included the effects of age, gender, education, household income, years in the U.S., nativity, employment status, marital status, and region of residence within the U.S. Note: Statistical significance comparing categories to the referent group is denoted by c p < 0.001.
b
on the accuracy of self-report data argue against substantial bias due to underreporting of smoking status despite few studies of tobacco consumption among foreign-born blacks (37, 38). An important implication of this work is that smokingrelated illnesses (and associated social and medical costs) may vary substantially between the two groups. This point is supported by the finding that lung cancer mortality was lower in Caribbean-born residents than in either nativeborn blacks or whites in New York City (39). Incidence rates of lung and other neoplasms were also reported to be
much lower in some Africans (i.e., in Gambia, Mali, and South Africa) compared to African Americans (40). Another implication involves the effect of the foreignborn population on the overall smoking rates of African Americans. Deleting foreign-born blacks from the overall smoking rate of African Americans in the Northeast increased the prevalence rates from 27.7% to 32.8%. In cities (e.g., New York) where the proportion of foreign-born blacks is substantial, use of overall smoking prevalence rates will underestimate prevalence in native-born African Americans. Thus in those geographic areas where the for-
AEP Vol. 9, No. 4 May 1999: 236–244
eign-born population exceeds 20%, it seems important to differentiate smoking by nativity to obtain an accurate profile of smoking risk behaviors and implications for smoking related diseases. This approach would likely enhance targeting smoking prevention and intervention programs to black populations at greatest risk (18). The number of cigarettes smoked per day differed little between foreign and native-born black smokers in this study. Although additional research is needed on cigarette content, smoking styles, and biochemical validation, these findings based on self-reports suggest that, once addicted, foreign-born and native-born African Americans may not differ in smoking behavior. Thus, smoking cessation efforts are needed for both groups. The consistency of our findings with other studies on cigarette smoking among foreign-born African and Caribbean migrants to the U.S. and Europe supports the need to investigate intergenerational smoking behavior. For example, to what extent do smoking rates of first generation blacks carry-over to succeeding generations who become “native-born” and presumably experience greater assimilation? Further, future studies should consider diversity within foreign-born black populations; language differences, cultural and health practices, history of U.S. migration, and social and community ties all may be relevant to smoking behavior (18). In sum, research on native and foreign-born African Americans not only can provide important insights on diversity within this racially classified social group but also help to clarify our understanding of the “race” concept and its association with health status and disease risk factors such as cigarette smoking (15). In this regard, researchers should be encouraged to investigate the role of nativity by examining cross-national data, theories of cultural change and the social diffusion of innovations, tobacco and alcohol consumption, and community organization.
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
243
and cancer incidence among blacks and whites. J Natl Cancer Inst. 1992;83:551–567. 8. King G, Bendel R, Delaronde S. Differences in Smoking among African American Adolescents and Young Adults: A Review of NHIS, 1990– 1993. 124th Annual Meeting of the American Public Health Association. New York, NY; November 1996. 9. Marin G, Perez-Stable EJ, Marin BV. Cigarette smoking among San Francisco Hispanics: The role of acculturation and gender. Am J Public Health. 1989;79:196–198. 10. Klatsky AL, Armstrong MA. Cardiovascular risk factors among Asian Americans living in Northern California. Am J Public Health. 1991;81:1423–1428. 11. U.S. Department of Health and Human Services. Tobacco use among U.S. racial/ethnic minority groups—African Americans, American Indians and Alaska natives, Asian Americans and Pacific Islanders and Hispanics: A report of the Surgeon General. Atlanta, GA: U.S. Dept. of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 1998. 12. U.S. Bureau of the Census, 1992. Public Use Microdata Samples: Technical Documentation. Washington DC: Government Printing Office; 1993. 13. Lobo PA. The Newest New Yorkers, 1990–1994. Department of City Planning. New York: City of New York; 1996. 14. Dunn M. Black Miami in the 20th Century. Gainesville, FL: University Press of Florida; 1997;332–333. 15. King G. The “Race” Concept in smoking: A review of the research on African Americans. Soc Sci Med. 1997;45:1075–1087. 16. Cabral H, Fried LE, Levenson S, Amaro H, Zuckerman B. Foreignborn and US-born black women: Differences in health behaviors and birth outcomes. Am J Public Health. 1990;80:70–72. 17. Welte JW, Barnes GM. Youthful Smoking: Patterns and relationships to alcohol and other drug use. J Adolesc. 1987;10:327–340. 18. Taylor KL, Kerner JF, Gold KF. Mandelblatt JS. Ever -vs.- never smoking among an urban multiethnic sample of Haitian-, Caribbean-, and U.S.-born Blacks. Prev Med. 1997;26:855–865. 19. Botman SL, Thornberry OT. Survey design features correlates of nonresponse. Proceedings of the Section on Survey Research Methods of the 1992 Annual Meeting of the American Statistical Association, Boston, MA; 1992. 20. National Center for Health Statistics. Vital and Health Statistics: Current Estimates from the National Health Interview Survey, 1993. Vital Health Stat 10. 1993; No. 190. 21. Botman SL, Jack SS. Combining national health interview survey datasets: Issues and approaches. Stat Med 1995;14:669–677.
REFERENCES 1. Williams DR. Race and health: Basic questions, emerging directions. Ann Epidemiol. 1997;7:322–333. 2. King G, Williams DR. Race and health: A multidimensional approach to African American health. In: Amick BC, Levine S, Walsh DC, Tarlov AR, eds. Society and Health. New York: Oxford University Press; 1995. 3. Polednak AP. Racial and Ethnic Differences in Disease. New York: Oxford University Press; 1989. 4. Novotny TE, Warner KE, Kendrick JS, Remington PL. Smoking by black and whites: Socioeconomic and demographic differences. Am J Public Health. 1988;78:1187–1189. 5. King G, Bendel R, Delaronde SR. Social heterogeneity in smoking among African Americans. Am J Public Health. 1998;88:1081–1085. 6. Escobedo LG, Zhu BP, Giovino GA, Eriksen MP. Educational attainment and racial differences in cigarette smoking. J Natl Cancer Inst. 1995;87:1552–1553. 7. Baquet, CR, Horm JW, Gibbs T, Greenwald, P. Socioeconomic factors
22. Shah BV, Barnwell BG, Bieler GS. SUDAAN User’s Manual Release 7.0. Research Triangle Park, NC: Research Triangle Institute; 1996. 23. Pan American Health Organization. Tobacco or Health: Status in the Americas; A report of the Pan American Health Organization. Scientific Publication No. 536. Washington: PAHO; 1992. 24. White A. Cigarette marketing in Senegal, West Africa. Tobacco Control. 1997; 6:243–245. 25. Leonard L. Cigarette smoking and perceptions about smoking and health in Chad. East Afr Med J. 1996;73:509–512. 26. Walker ARP. Cancer outlook: An African perspective. J R Soc Med. 1995;88:5–13. 27. Edwards NC, MacMillan K. Tobacco use and ethnicity: The existing data gap. Can J Public Health. 1990;81:32–36. 28. Millar WJ. Place of birth and ethnic status: Factors associated with smoking prevalence among Canadians. Health Reports. 1992;4:7–24. 29. Balarajan R, Yuen P. British smoking and drinking habits: Variations by country of birth. Comm Med. 1986;237–239. 30. Waterson EJ, Murray-Lyon IM. Alcohol, smoking and pregnancy:
244
31.
32. 33. 34. 35.
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
Some observations on ethnic minorities in the United Kingdom. Br J Addiction 1989;84:323–325. Bouchardy C, Parkin DM, Wanner P, Khlat M. Cancer mortality among North African migrants in France. Int J Epidemiol. 1996; 25:5–13. Chen J, Ng E, Wilkins R. The health of Canada’s immigrants in 1994–95. Health Reports. 1996;7:33–45. Kalmijn M. The socioeconomic assimilation of Caribbean American blacks. Social Forces. 1996;74:911–930. Model S. Caribbean Immigrants: A black success story? Int Migr Rev. 1991;25:248–276. Szapocznik J, Kurtines W. Acculturation, biculturalism, and adjustment among Cuban Americans. In: Padilla A. ed. Acculturation: Theory, Models, and New Findings. Boulder, CO: Westview Press; 1980.
AEP Vol. 9, No. 4 May 1999: 236–244
36. Bettes BA, Dunsenbury L, Kerner J, James-Ortiz S, Botvin GJ. Ethnicity and psychosocial factors in alcohol and tobacco use in adolescence. Child Develop. 1990;61:557–565. 37. Patrick DL, Cheadle A, Thompson DC, Diehr P, Koepsell T, Kinne S. The validity of self-reported smoking: A review and meta analysis. Am J Public Health. 1994;84:1086–1093. 38. Clark PI, Gautam SP, Hlaing WM, Gerson LW. Response error in self-reported current smoking frequency by black and white established smokers. Ann Epidemiol. 1996;6:483–489. 39. Fang J, Madhavan S, Alderman MH. Influence of nativity on cancer mortality among black New Yorkers. Cancer. 1997;80:129–135. 40. Walker ARP, Walker BF, Segal I. Cancer patterns in three African populations compared with the United States black population. Eur J Cancer Prev. 1993;2:313–320.
PURPOSE: To examine differences in current smoking status and the number of cigarettes consumed daily between foreign and native-born African Americans, and the impact of demographic and socioeconomic status (SES) factors on smoking behavior. METHODS: Data were obtained from combining the 1990–1994 National Health Interview Surveys and consisted of 16,738 U.S. born and foreign-born African Americans between 18 and 64 years of age. The statistical analysis included cross-tabulations and weighted multiple logistic regression (MLR) using the Statistical Packages for the Social Sciences (SPSS) and the Survey Data Analysis (SUDAAN) computer programs. RESULTS: Adjusted MLR analysis revealed that native-born African Americans were more likely (odds ratio (OR) 5 2.7, p , 0.001) to be current smokers than foreign-born blacks. Within the nativeborn group, smoking prevalence decreased with increasing education and income, but these associations were not found for foreign-born blacks. Women in both groups were less likely than men to be current smokers. Statistically significant differences were not found between the two groups in the number of cigarettes smoked per day. CONCLUSIONS: This analysis of nativity and smoking behavior further demonstrates the social diversity among African Americans and suggests the differential impact of social and cultural factors on smoking behavior within racially classified social groups. In areas where there are substantial numbers of foreign-born blacks, researchers should consider differentiating smoking status by nativity. Though differences in smoking prevalence were apparent for native and foreign-born American Americans, prevention and cessation programs are needed for both groups. Ann Epidemiol 1999;9:236–244. 1999 Elsevier Science Inc. All rights reserved. KEY WORDS:
Cigarette Smoking, African Americans, Blacks, Migrant Health, Minority Health.
INTRODUCTION An important contemporary theme in the study of the health of African Americans is intra-group diversity or social heterogeneity (1–3). Socio-demographic variables such as age, gender, geographic region, and socio-economic status (SES) help to explain variations in health behavior and outcomes (4–8). Nativity or country of birth also has been shown to be associated with disease risks and health status among minority groups such as Hispanics and Asian Americans (9–11). In the 1990 U.S. Census, it was estimated that 5.8% of African Americans were foreign-born (12) and more current
From the Department of Biobehavioral Health (G.K.), Penn State University, University Park, PA; Department of Community Medicine and Health Care (Adjunct) (A.P.P.), University of Connecticut Health Center, Farmington, CT; Center for Environmental Health (R.B.), University of Connecticut, Storrs, CT; and Department of Community Medicine and Health Care (D.H.), University of Connecticut Health Center, Farmington, CT. Address reprint requests to: Gary King, Ph.D., Department of Biobehavioral Health, 315 E. Health and Human Development, Penn State University, University Park, PA 16801. Received May 19, 1998; revised September 08, 1998; accepted September 09, 1998. 1999 Elsevier Science Inc. All rights reserved. 655 Avenue of the Americas, New York, NY 10010
data indicates that this figure may exceed 25% in some urban areas (e.g., New York City and Dade County, Florida) (13, 14). Studies of nativity can be useful in examining various social experiences and influences on the health of African Americans including acculturation, social mobility, cultural practices and social networks, and exposure to (or protective strategies against) racism. Moreover, this research could have important implications for health intervention programs and policies and may suggest multiple strategies for studying African American health. Despite numerous studies on cigarette smoking among African Americans, differences in smoking behavior between foreign and native-born blacks have rarely been investigated (15). In a 1984 study of black pregnant women in Boston, foreign-born women were less likely than nativeborn women to be smokers at any time during pregnancy (16). In a study of New York high school students a lower smoking prevalence rate was found among Caribbean-born compared to native-born African-Americans (17). A study of adults living in New York City in 1992 revealed a lower rate of ever smoking among foreign-born relative to nativeborn African Americans (18). Despite the contributions of these studies, the results are based mostly on data from 1047-2797/99/$–see front matter PII S1047-2797(98)00052-0
AEP Vol. 9, No. 4 May 1999: 236–244
Selected Abbreviations and Acronyms SES 5 socio-economic status MLR 5 multiple logistic regression NHIS 5 National Health Interview Survey CCS 5 Cancer Control Supplement CES 5 Cancer Epidemiology Supplement SPSS 5 Statistical Packages for the Social Sciences SUDAAN 5 Survey Data Analysis
small select samples (i.e., Caribbeans, pregnant women, and students) (16, 17) and/or did not examine current smoking or consumption frequency (18). Using data from the 1990–1994 National Health Interview Survey (NHIS), the major objectives of this paper were: 1) to determine whether differences in smoking status and the number of cigarettes consumed daily exist between foreign and native-born African Americans; and 2) to assess the relative importance of demographic and SES factors on smoking behavior between the two groups. This study of a large sample of African Americans is broadly representative of adults, 18–64 years old, and provides previously unreported information about tobacco addiction and nativity among blacks in the U.S.
METHODS Sample Design The NHIS is a national cross-sectional household sample of non-institutionalized civilians. The survey is approximated by a three stage stratified cluster probability sample design. The response rates of African Americans to each of the NHIS surveys between 1990 and 1994 are not known but probably do not differ substantially from other groups (19). The overall annual response rates were about 95% and information about the NHIS has been previously published (20). The annual unweighted sample sizes are displayed in Table 1 with a cumulative total of 16,738 respondents. In this analysis, the Cancer Control Supplement (CCS) and Cancer Epidemiology Supplement (CES) to the 1992 NHIS were merged to increase the representation of African Americans and the weights of the combined sample were halved as suggested by Botman and Jack (21). For all other years, the analysis was conducted using either the Year 1990 or Year 2000 Objectives supplements. Instrument The term foreign-born refers to persons who identified as black or African American and whose country of nativity is outside the United States or its territories. Persons who selected their racial classification as African American or black and who were between 18 and 64 years of age were
237
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
TABLE 1. Socio-demographic characteristics of U.S.-born Blacks and foreign-born Blacks U.S.-born Foreign-born Total Blacks Blacks N 5 16,738a N 5 15,660a N 5 1078a %b %b %b Age of respondents 18–24 years old 25–34 35–44 45–54 55–64 Gender Male Female Education Some high school High school graduate Some college College graduate or more Employment status Employed Unemployed Not in the labor force Region West South Midwest Northeast Family income Less than $15,000 $15,000 to $19,999 $20,000 to $49,999 $50,000 or more Marital status Married Unmarried a b
19.5 29.4 24.1 15.3 11.8
19.8 28.8 23.9 15.4 12.1
15.7 36.4 26.6 13.7 7.5
45.3 54.7
45.0 55.0
48.5 51.5
24.1 41.8 22.1 12.0
24.3 42.6 21.8 11.3
21.7 32.2 25.9 20.2
67.2 6.4 26.3
66.6 6.5 26.9
74.6 5.7 19.6
9.7 53.2 20.8 16.4
9.8 55.2 22.0 13.0
8.1 28.8 6.7 56.5
34.3 12.9 39.2 13.5
35.2 13.1 38.5 13.2
23.6 11.1 48.0 17.3
45.0 55.0
44.2 55.8
54.0 46.0
National Health Interview Survey, 1990–1994 unweighted data. Percentages are the weighted prevalence rates.
included. The NHIS does not include questions specifying the particular country of birth. However, immigration data show that between 1994 and 1996, immigrants from Caribbean and African countries, respectively comprised 72.3% and 27.7% of persons from these geographic regions. In this paper, racial classification is viewed as a social variable rather than a phenotypic indicator of biologic or genetic differences between human populations. Also, because only a relatively small proportion of African American smokers are 65 years or older, the analysis was restricted to ages 18–64. Current smoking status was the primary dependent variable and consisted of persons who were either current smokers or nonsmokers (former and never smokers). In 1990, current smokers were defined as persons who: 1) reported smoking at least 100 cigarettes in their entire life; and 2) were presently smoking cigarettes. In 1991–1994, current smokers included persons who: 1) reported smoking at least
238
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
100 cigarettes in their entire life; and 2) currently smoked either “everyday” or “some days.” Data for 1990 were included despite the change in the definition of current smoking because the impact of the change was small. In a separate analysis it was found that excluding the additional someday smokers reduced the increase in the smoking prevalence rate of African Americans from 3.7% to 3.3% between 1990 and 1991 (5). Furthermore, excluding the 1990 data would have reduced the sample of foreign-born blacks. The second response variable was the number of cigarettes smoked per day by all current smokers in 1990 and by “everyday” smokers in 1991–1994. It was dichotomized as fewer than 15 cigarettes and 15 or more cigarettes per day. Most African Americans (64%) smoked fewer than 15 cigarettes per day so that the 15 or more category defines heavier smokers (5, 11). It was not possible to investigate fully the age of smoking initiation among this sample of NHIS respondents because this variable is only available for the year 1992. An analysis of the 1992 data indicates that the initiation rate among African American adults is greatest among the age group 18–24 (8). The independent variables selected for analysis were age, gender, years of education, annual family income, geographic region, nativity, and length of U.S. residence. To construct the education variable, the completed number of years of education was divided into four categories: 1) less than high school (, 12 years); 2) high school graduate (12 years); 3) some college (1–3 years); and 4) four or more years of college. The annual family income groups were: , $15,000, $15,000–$19,999, $20,000–$49,999, and $50,000 or more. A regional variable based on the U.S. Census definition (i.e., Northeast, Midwest, South, West) was included to assess geographic variations. Nativity refers to persons who were born within the U.S. (native-born) and those born outside of the U.S. (foreign-born). Length of U.S. residence pertains solely to foreign-born persons and includes the categories of , 15 years and 15 years or more. The NHIS response categories for this variable are , 1 year, 1 year–less than 5 years, 5 years–less than 10 years, 10 years–less than 15 years, and 15 years or more. Our analysis did not reveal statistically significant differences between the individual categories of , 15 years and the dependent variables of current smoking and number of cigarettes smoked daily. As a result, we combined the categories of , 1 year to , 15 years of residence to create a variable , 15 years vs. 15 years or more.
AEP Vol. 9, No. 4 May 1999: 236–244
the prevalence rates, logistic regression coefficients, and odds ratios (ORs) reflect the population relationships and are obtained for each subject from the inverse probability of selection with modification for oversampling and missing data. Weights were rescaled (for only SPSS computations) so that the sum of the weights was equal to the sample size instead of the subpopulation size. Rescaling of the weights using SPSS is likely to yield p-values that are too small as it does not take into account the clustering in the NHIS design. Consequently, the SPSS computations were used only as a tool to identify potential associations, i.e., for model building purposes. Indicator coding was used to create single degree of freedom components for categorical variables with more than two categories. Both SPSS and SUDAAN treated all of the single degree of freedom components associated with a main effect or interaction collectively, so that p-values and likelihood ratio tests applied to the “entire” main effect or interaction. In presenting the results, ORs were computed as OR 5 exp (b) where b is the estimate of the logistic regression coefficient. The 95% confidence intervals for the ORs were computed as: exp (b 6 t (.025) * SE) where SE is the (adjusted) standard error of b and t is the upper 2.5 percentile of the ‘t’ distribution with degrees of freedom (df) from SUDAAN (df 5 # of primary sampling units (PSUs) 2 # of fixed strata for the NHIS multistage design where PSUs are sampled without replacement from the fixed strata) (22). Standard errors were adjusted upward by approximately 18% to account for the increased clustering due to combining samples from 5 years. Computation of exact standard errors is not possible because the NHIS typically samples households in the same areas of the same PSUs and this clustering is not identifiable when combining data from 5 years. Botman and Jack, however, note that the design effect (defined as the ratio of the variance of the estimator under a complex survey design to the variance of the estimator under a simple random sample) for estimates of percentages based on a single year of survey data is about 1.3, and based on three years of combining survey data is about 1.6 (21). Hence, for three years the standard errors would be increased 11% (i.e., 100 * sqrt(1.6/1.3)) and for five years, a reasonable (conservative) extrapolation would be 18% (i.e., 11 1 (2/3) * 11).
Statistical Analysis The statistical analysis included crosstabulations and weighted logistic regression modeling using both the Statistical Packages for the Social Sciences (SPSS) and the Survey Data Analysis (SUDAAN) programs. Weights were used so that
RESULTS The age distribution differed slightly for the two subpopulations (Table 1) and the mean age was 36.8 years for nativeborn and 35.8 years for foreign-born blacks. A higher per-
AEP Vol. 9, No. 4 May 1999: 236–244
centage of foreign-born than native-born blacks was found in the highest income and education categories. Most (55.2%) native-born blacks resided in the South compared to foreign-born blacks who mostly (56.5%) were located in the Northeast. A greater proportion of foreign-born (54%) than native-born blacks (44.2%) were married. The prevalence of current cigarette smoking (Table 2) was more than twice as high for native-born compared to foreign-born African Americans (30.4% vs. 14.1%, p , 0.001). This finding was evident in all categories of age, gender, region, education, income, employment, and marital status. Notably, smoking prevalence was more than 4 times higher in native-born than in foreign-born women. Among both the native and foreign-born groups (Table 2), smoking prevalence was highest among 35–44 year-olds and lowest among 18–24 year olds. Within the native-born group, prevalence varied with education (p , 0.001) and income (p , 0.001) and was highest among lower SES persons (e.g., unemployed, lower income, and non-high school graduates). These associations were not found for foreign-born blacks. The ratio of men to women smokers was much higher among the foreign-born (3.4) compared to native-born respondents (1.4). A significant association was not found in the foreign-born population between length of years of residence and smoking, although those who resided 15 or more years in the U.S. had the highest smoking rate (17.5%; Table 2). The unadjusted logistic regression analysis revealed that native-born blacks were 2.6 times (p , 0.001) more likely to be smokers than foreign-born blacks (Results not shown). Using multivariate logistic regression, this association remained after controlling for the effects of other demographic predictors such as age, education, and gender (OR 5 2.73, p , 0.001; Table 3). In the regression model (Table 3) which included only foreign-born blacks, the significant predictors of smoking status were gender (OR 5 0.16, p , 0.001) and to a lesser extent age. None of the SES variables (i.e., education, income, employment status) were statistically significant predictors of smoking. Length of U.S. residence was marginally significant (p , 0.10) and there was no statistically significant interaction between age and duration of U.S. residence (data not shown). The multivariate analysis (Table 3) showed that the impact of education on smoking among foreign-born blacks was less evident at the lowest levels (i.e., substantial differences were not apparent between high school and non-high school graduates). The results were similar for the Northeast region, where a majority of foreignborn blacks resided. Among current smokers, the mean number of cigarettes consumed per day differed only slightly and not statistically significantly between native-born (14.5) and foreign-born (13.8) blacks. Both native (38.6%) and foreign-born African American women (29.8%) consumed 15 or more ciga-
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
239
TABLE 2. Prevalence of current smokersa in U.S.- and foreignborn Blacks by socio-demographic characteristics of 18–64 year oldsb
Age of respondents 18–24 years old 25–34 35–44 45–54 55–64 Gender Male Female Education Some high school High school graduate Some college College graduate or more Employment status Employed Unemployed Not in the labor force Region West South Midwest Northeast Income Less than $15,000 $15,000 to $19,999 $20,000 to $49,999 $50,000 or more Years in US Less than 15 years 15 years or more Marital status Married Unmarried Total
All Blacks %
U.S.-born Blacks %
Foreignborn Blacks %
14.3c 31.1 36.4 34.4 26.6
14.8c 32.8 38.3 36.0 27.5
6.5d 15.9 17.3 14.2 10.5
34.6c 24.6
35.7c 26.0
22.2c 6.6
38.4c 29.9 24.3 17.0
40.2c 30.8 25.3 17.8
14.2 15.4 14.7 11.7
26.9c 39.6 32.1
28.2c 40.9 33.4
14.1 24.3 11.5
25.4c 28.0 34.8 27.7
26.1c 28.6 35.5 32.8
15.5 15.1 10.4 13.9
35.8c 29.7 25.8 21.8
36.8c 30.9 27.1 22.5
16.7 12.2 13.7 16.6
––
––
12.1 17.5
28.1d 30.0 29.1
29.6 31.1 30.4
14.3 14.0 14.1
a
Current smoking status is defined as the proportion of persons who had consumed more than 100 cigarettes in their lifetime and were smoking either every day or some days at the time of the National Health Interview Survey. b Percentages are the weighted prevalence rates. Note: Statistical significance comparing categories within each variable is denoted by d p < 0.05, c p < 0.001.
rettes daily less frequently than native (52.6%) and foreign born (51.8%) men (Table 4). Age, education, and employment status were significantly associated with heavier smoking among native born but not foreign-born blacks (Table 4). These findings may be due to the smaller size for foreignborn blacks. In multiple logistic regression analysis (Table 5), nativity was not significantly related to heavier smoking (15 or 15 or more cigarettes daily). Within the native-born group,
240
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
AEP Vol. 9, No. 4 May 1999: 236–244
TABLE 3. Logistic regression models of smokinga behavior among U.S.- and foreign-born Blacks, 1990–1994 (lower and upper 95% limits) All Blacks
U.S.-born Blacks
Foreign-born Blacks
1.00 3.46 4.88 3.91 2.34
1.00 4.29 4.62 3.47 2.47
b
Age categories (years) 18–24 25–34 35–44 45–54 55–64 Gender Male Female Education Less than high school High school graduate Some college College graduate or more Household income Less than $15,000 $15,000 to $19,999 $20,000 to $49,999 $50,000 or more Number of years in U.S. 15 years or more Less than 15 years Nativity Foreign-born Blacks U.S.-born Blacks Employment Employed Unemployed Not in the labor force Marital status Married/partnered Unmarried/not partnered Region Northeast South Midwest West
1.00 3.50 4.87 3.90 2.37
(2.83, (3.93, (3.08, (1.87,
4.31)e 5.99)e 4.92)e 2.99)e
(2.80, (3.97, (3.81, (2.29,
4.26)e 6.05)e 3.99)e 2.39)e
(1.11, (1.19, (0.81, (0.54,
16.70)c 17.91)c 14.71) 11.23)
1.00 0.53 (0.47, 0.59)e
1.00 0.55 (0.49, 0.62)e
1.00 0.16 (0.08, 0.33)e
1.00 0.71 (0.60, 0.84)e 0.60 (0.50, 0.72)e 0.36 (0.28, 0.47)e
1.00 0.70 (0.59, 0.82)e 0.58 (0.48, 0.70)e 0.35 (0.27, 0.45)e
1.00 1.04 (0.71, 3.15) 0.81 (0.34, 1.90) 0.53 (0.21, 1.32)
1.00 0.76 (0.63, 0.91)d 0.67 (0.58, 0.77)e 0.61 (0.48, 0.77)e
1.00 0.76 (0.63, 0.91)d 0.67 (0.58, 0.77)e 0.59 (0.47, 0.74)e
1.00 0.90 (0.38, 2.15) 0.91 (0.45, 1.84) 1.14 (0.38, 3.41)
— —
1.00 0.63 (0.35, 1.13)
— — 1.00 2.73 (1.98, 3.81)e
— —
— —
1.00 1.72 (1.36, 2.17)e 1.19 (0.82, 1.72)
1.00 1.70 (1.34, 2.15)e 1.18 (1.03, 1.36)c
1.00 1.77 (0.68, 4.61) 1.18 (0.53, 2.60)
1.00 1.19 (0.25, 5.54)
1.00 1.17 (1.03, 1.31)c
1.00 1.62 (0.88, 2.97)
1.00 0.76 (0.63, 0.92)d 1.06 (0.86, 1.31) 0.76 (0.59, 0.99)c
1.00 0.74 (0.61, 0.89)d 1.05 (0.85, 1.30) 0.75 (0.58, 0.97)c
1.00 1.15 (0.56, 2.37) 0.42 (0.11, 1.55) 0.72 (0.22, 2.31)
a
Current smoking status is defined as the proportion of persons who had consumed more than 100 cigarettes in their lifetime and were smoking either every day or some days at the time of the National Health Interview Survey. b The adjusted logistic regression model is based on weighted data and included the effects of age, gender, education, household income, years in the U.S., nativity, employment status, marital status, and region of residence within the U.S. Note: Statistical significance comparing categories to the reference group is denoted by c p < 0.05; d p < 0.01; e p < 0.001.
adjusted ORs for most foreign-born blacks, age categories and gender were statistically significant. Among foreignborn blacks, there were no statistically significant adjusted ORs as the small number of smokers resulted in wide confidence limits. DISCUSSION This study shows large differences in the smoking behavior of native and foreign-born African Americans. Native-born African Americans were far more likely to be smokers than foreign-born blacks even after controlling for differences in
sociodemographic indicators. Though women in both groups were less likely to be smokers, native-born women were 4 times more likely to be smokers than their foreignborn counterparts. Interestingly, among foreign-born blacks, income was inconsistently related to smoking and neither income nor education were statistically significant predictors of smoking status. This particular finding was also supported by investigators examining SES and ever smoking status among foreign-born blacks (18). Caribbean and sub-Saharan African countries generally have low smoking rates, especially among women, and migrants from these countries to the U.S. appear to have
AEP Vol. 9, No. 4 May 1999: 236–244
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
TABLE 4. Daily smoking frequencya in U.S.- and foreign-born Blacks and socio-demographic characteristics of 18–64 year olds
Age of respondents 18–24 years old 25–34 35–44 45–54 55–64 Gender Male Female Education Some high school High school graduate Some college College graduate or more Employment status Employed Unemployed Not in the labor force Region West South Midwest Northeast Income Less than $15,000 $15,000 to $19,999 $20,000 to $49,999 $50,000 or more Years in U.S. Less than 15 years 15 years or more Marital status Married Unmarried Total
U.S.-born smokers who consume >15 cigarettesb %
Foreign-born smokers who consume >15 cigarettesb %
30.7c 39.3 48.9 53.8 52.8
47.8 44.1 49.9 56.0 19.8
52.6c 38.6
51.8d 29.8
49.5c 44.7 43.8 43.2
34.7 41.2 63.3 40.0
44.9d 47.1 47.9
49.4 42.0 29.2
43.0 45.8 47.8 45.1
53.4 34.2 56.9 50.0
45.7 47.0 42.9 48.6
40.7 52.0 46.9 56.8
— —
47.9 44.4
47.9 44.4 46.0
46.6 45.8 46.2
a
Current smoking status is defined as the proportion of persons who had consumed more than 100 cigarettes in their lifetime and were smoking either every day or some days at the time of the National Health Interview Survey. b Proportion of current smokers in the National Health Interview Survey who consumed 15 or more cigarettes per day. Percentages are the weighted prevalence rates. Note: Statistical significance comparing categories within each variable is denoted by d p < 0.05; c p < 0.001.
retained lower smoking rates than native populations (23– 26). Lower smoking prevalence among Caribbean and African migrants compared to native populations has also been observed in Canada, Great Britain and other European countries (27–31). Differences in smoking rates between foreign and native-born blacks in the U.S., however, could decrease in the future (if migration to the U.S. continues) because of the increasing prevalence of smoking among
241
youth (particularly young males) in Caribbean and African countries; contributing factors are greater economic development and the proliferation of cigarette advertisement (23–26). Also, smoking rates may decline more precipitously among native-born African Americans. The possibility that the most “acceptable” (e.g., highly educated) and healthy individuals more often chose or are allowed to immigrate (32), may have only a small effect on the smoking prevalence among the foreign-born in the U.S. because of the currently low rates of smoking in African and Caribbean countries (23–26). Although our study could not measure explicitly the effects of acculturation, foreign-born blacks (especially women) were less likely than native-born African Americans to smoke cigarettes even after a considerable postmigration period. The highest foreign-born smoking rate, however, was found among long-term residents (151 years), albeit it did not equal the prevalence of native-born blacks and the p value exceeded the threshold of statistical significance in the adjusted logistic regression model. Protective factors operative within foreign-born black communities that promote smoking prevention or cessation may involve religious proscriptions, social mores, and pressures against adopting particular American habits (especially for women) (33, 34). Other related factors may include alternative ways of addressing racism and social stress, smaller and insular immigrant communities, stronger regard for health detriments of smoking, costs of cigarettes (especially among newly arrived migrants), and different cultural orientations (35, 36). Key research issues involve identifying the “protective” factors that maintain lower rates of smoking initiation and whether they can be promoted among nativeborn blacks (and conceivably other racially classified social groups). The finding that smoking was not found to vary significantly with income or education in foreign-born blacks points to the need to examine social organization and cultural norms, sanctions, and support systems. This is a major difference between this group and native African Americans and may be suggestive of the stronger role of cultural and environmental factors. For example, a plausible hypothesis based on cultural socialization theory suggests that the differences in smoking prevalence between native-born and foreign-born African Americans may be partially explained by large numbers of adults in the older age groups among the foreign-born who never smoked cigarettes. This “elder” group may serve as standard bearers of the culture or role models for adolescents and young adults to prevent smoking. Study limitations include the possibility of underestimating smoking prevalence due to undocumented residents and non-respondent bias or sampling error related to undercoverage (e.g., institutionalized populations and “hidden” males). Also, as these NHIS data are cross-sectional surveys, age differences may be influenced by cohort effects. Evidence
242
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
AEP Vol. 9, No. 4 May 1999: 236–244
TABLE 5. Logistic regression models predicting daily smoking frequencya among U.S. and foreign-born Black current smokers, 1990–1994
Age categories (years)b 18–24 25–34 35–44 45–54 55–64 Gender Male Female Education Less than high school High school graduate Some college College graduate or more Household income Less than $15,000 $15,000 to $19,999 $20,000 to $49,999 $50,000 or more Years in U.S. 15 years or more Less than 15 years Nativity Foreign-born Blacks U.S.-born Blacks Employment Employed Unemployed Not in the labor force Marital status Married/partnered Unmarried/not partnered Region Northeast South Midwest West
All Blacks
U.S.-born
Foreign-born
Smokers who consumed >15 cigarettes daily
Smokers who consumed >15 cigarettes daily
Smokers who consumed >15 cigarettes daily
1.00 1.40 2.09 2.51 2.13
(0.94, (1.41, (1.61, (1.40,
2.07)c 3.12)c 3.91)c 3.26)c
1.00 1.42 2.15 2.55 2.25
(0.93, (1.44, (1.64, (1.48,
2.16)c 3.18)c 3.99)c 3.42)c
1.00 2.42 4.95 5.38 0.67
(0.24, (0.45, (0.48, (0.03,
23.80) 54.95) 59.53) 12.90)
1.00 0.55 (0.46, 0.67)c
1.00 0.56 (0.46, 0.67)c
1.00 0.44 (0.13, 1.48)
1.00 0.88 (0.70, 1.11) 0.99 (0.75, 1.31) 0.78 (0.52, 1.16)
1.00 0.88 (0.70, 1.11) 0.94 (0.70, 1.23) 0.80 (0.53, 1.18)
1.00 0.74 (0.15, 3.51) 1.86 (0.32, 10.72) 0.47 (0.06, 3.79)
1.00 1.01 (0.73, 1.40) 0.86 (0.68, 1.09) 1.08 (0.73, 1.61)
1.00 1.01 (0.71, 1.43) 0.85 (0.66, 1.10) 1.05 (0.71, 1.56)
1.00 1.64 (0.16, 16.88) 1.53 (0.40, 5.96) 1.82 (0.21, 15.63)
— —
— —
1.00 1.11 (0.34, 3.64)
1.00 0.99 (0.57, 1.75)
— —
— —
1.00 1.25 (0.88, 1.77) 1.22 (0.94, 1.58)
1.00 1.28 (0.90, 1.82) 1.23 (0.95, 1.60)
1.00 0.56 (0.08, 3.89) 0.59 (0.11, 3.06)
1.00 0.98 (0.78, 1.24)
1.00 0.95 (0.77, 1.17)
1.00 3.19 (0.79, 12.96)
1.00 0.95 (0.71, 1.25) 1.05 (0.76, 1.46) 0.86 (0.59, 1.25)
1.00 0.97 (0.73, 1.28) 1.06 (0.77, 1.47) 0.85 (0.56, 1.30)
1.00 0.30 (0.07, 1.24) 3.99 (0.73, 22.10) 0.97 (0.19, 4.98)
a
Proportion of current smokers in the National Health Interview Survey who consumed 15 or more cigarettes per day. The adjusted logistic regression model is based on weighted data and included the effects of age, gender, education, household income, years in the U.S., nativity, employment status, marital status, and region of residence within the U.S. Note: Statistical significance comparing categories to the referent group is denoted by c p < 0.001.
b
on the accuracy of self-report data argue against substantial bias due to underreporting of smoking status despite few studies of tobacco consumption among foreign-born blacks (37, 38). An important implication of this work is that smokingrelated illnesses (and associated social and medical costs) may vary substantially between the two groups. This point is supported by the finding that lung cancer mortality was lower in Caribbean-born residents than in either nativeborn blacks or whites in New York City (39). Incidence rates of lung and other neoplasms were also reported to be
much lower in some Africans (i.e., in Gambia, Mali, and South Africa) compared to African Americans (40). Another implication involves the effect of the foreignborn population on the overall smoking rates of African Americans. Deleting foreign-born blacks from the overall smoking rate of African Americans in the Northeast increased the prevalence rates from 27.7% to 32.8%. In cities (e.g., New York) where the proportion of foreign-born blacks is substantial, use of overall smoking prevalence rates will underestimate prevalence in native-born African Americans. Thus in those geographic areas where the for-
AEP Vol. 9, No. 4 May 1999: 236–244
eign-born population exceeds 20%, it seems important to differentiate smoking by nativity to obtain an accurate profile of smoking risk behaviors and implications for smoking related diseases. This approach would likely enhance targeting smoking prevention and intervention programs to black populations at greatest risk (18). The number of cigarettes smoked per day differed little between foreign and native-born black smokers in this study. Although additional research is needed on cigarette content, smoking styles, and biochemical validation, these findings based on self-reports suggest that, once addicted, foreign-born and native-born African Americans may not differ in smoking behavior. Thus, smoking cessation efforts are needed for both groups. The consistency of our findings with other studies on cigarette smoking among foreign-born African and Caribbean migrants to the U.S. and Europe supports the need to investigate intergenerational smoking behavior. For example, to what extent do smoking rates of first generation blacks carry-over to succeeding generations who become “native-born” and presumably experience greater assimilation? Further, future studies should consider diversity within foreign-born black populations; language differences, cultural and health practices, history of U.S. migration, and social and community ties all may be relevant to smoking behavior (18). In sum, research on native and foreign-born African Americans not only can provide important insights on diversity within this racially classified social group but also help to clarify our understanding of the “race” concept and its association with health status and disease risk factors such as cigarette smoking (15). In this regard, researchers should be encouraged to investigate the role of nativity by examining cross-national data, theories of cultural change and the social diffusion of innovations, tobacco and alcohol consumption, and community organization.
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
243
and cancer incidence among blacks and whites. J Natl Cancer Inst. 1992;83:551–567. 8. King G, Bendel R, Delaronde S. Differences in Smoking among African American Adolescents and Young Adults: A Review of NHIS, 1990– 1993. 124th Annual Meeting of the American Public Health Association. New York, NY; November 1996. 9. Marin G, Perez-Stable EJ, Marin BV. Cigarette smoking among San Francisco Hispanics: The role of acculturation and gender. Am J Public Health. 1989;79:196–198. 10. Klatsky AL, Armstrong MA. Cardiovascular risk factors among Asian Americans living in Northern California. Am J Public Health. 1991;81:1423–1428. 11. U.S. Department of Health and Human Services. Tobacco use among U.S. racial/ethnic minority groups—African Americans, American Indians and Alaska natives, Asian Americans and Pacific Islanders and Hispanics: A report of the Surgeon General. Atlanta, GA: U.S. Dept. of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, 1998. 12. U.S. Bureau of the Census, 1992. Public Use Microdata Samples: Technical Documentation. Washington DC: Government Printing Office; 1993. 13. Lobo PA. The Newest New Yorkers, 1990–1994. Department of City Planning. New York: City of New York; 1996. 14. Dunn M. Black Miami in the 20th Century. Gainesville, FL: University Press of Florida; 1997;332–333. 15. King G. The “Race” Concept in smoking: A review of the research on African Americans. Soc Sci Med. 1997;45:1075–1087. 16. Cabral H, Fried LE, Levenson S, Amaro H, Zuckerman B. Foreignborn and US-born black women: Differences in health behaviors and birth outcomes. Am J Public Health. 1990;80:70–72. 17. Welte JW, Barnes GM. Youthful Smoking: Patterns and relationships to alcohol and other drug use. J Adolesc. 1987;10:327–340. 18. Taylor KL, Kerner JF, Gold KF. Mandelblatt JS. Ever -vs.- never smoking among an urban multiethnic sample of Haitian-, Caribbean-, and U.S.-born Blacks. Prev Med. 1997;26:855–865. 19. Botman SL, Thornberry OT. Survey design features correlates of nonresponse. Proceedings of the Section on Survey Research Methods of the 1992 Annual Meeting of the American Statistical Association, Boston, MA; 1992. 20. National Center for Health Statistics. Vital and Health Statistics: Current Estimates from the National Health Interview Survey, 1993. Vital Health Stat 10. 1993; No. 190. 21. Botman SL, Jack SS. Combining national health interview survey datasets: Issues and approaches. Stat Med 1995;14:669–677.
REFERENCES 1. Williams DR. Race and health: Basic questions, emerging directions. Ann Epidemiol. 1997;7:322–333. 2. King G, Williams DR. Race and health: A multidimensional approach to African American health. In: Amick BC, Levine S, Walsh DC, Tarlov AR, eds. Society and Health. New York: Oxford University Press; 1995. 3. Polednak AP. Racial and Ethnic Differences in Disease. New York: Oxford University Press; 1989. 4. Novotny TE, Warner KE, Kendrick JS, Remington PL. Smoking by black and whites: Socioeconomic and demographic differences. Am J Public Health. 1988;78:1187–1189. 5. King G, Bendel R, Delaronde SR. Social heterogeneity in smoking among African Americans. Am J Public Health. 1998;88:1081–1085. 6. Escobedo LG, Zhu BP, Giovino GA, Eriksen MP. Educational attainment and racial differences in cigarette smoking. J Natl Cancer Inst. 1995;87:1552–1553. 7. Baquet, CR, Horm JW, Gibbs T, Greenwald, P. Socioeconomic factors
22. Shah BV, Barnwell BG, Bieler GS. SUDAAN User’s Manual Release 7.0. Research Triangle Park, NC: Research Triangle Institute; 1996. 23. Pan American Health Organization. Tobacco or Health: Status in the Americas; A report of the Pan American Health Organization. Scientific Publication No. 536. Washington: PAHO; 1992. 24. White A. Cigarette marketing in Senegal, West Africa. Tobacco Control. 1997; 6:243–245. 25. Leonard L. Cigarette smoking and perceptions about smoking and health in Chad. East Afr Med J. 1996;73:509–512. 26. Walker ARP. Cancer outlook: An African perspective. J R Soc Med. 1995;88:5–13. 27. Edwards NC, MacMillan K. Tobacco use and ethnicity: The existing data gap. Can J Public Health. 1990;81:32–36. 28. Millar WJ. Place of birth and ethnic status: Factors associated with smoking prevalence among Canadians. Health Reports. 1992;4:7–24. 29. Balarajan R, Yuen P. British smoking and drinking habits: Variations by country of birth. Comm Med. 1986;237–239. 30. Waterson EJ, Murray-Lyon IM. Alcohol, smoking and pregnancy:
244
31.
32. 33. 34. 35.
King et al. SMOKING, NATIVITY, AND AFRICAN AMERICANS
Some observations on ethnic minorities in the United Kingdom. Br J Addiction 1989;84:323–325. Bouchardy C, Parkin DM, Wanner P, Khlat M. Cancer mortality among North African migrants in France. Int J Epidemiol. 1996; 25:5–13. Chen J, Ng E, Wilkins R. The health of Canada’s immigrants in 1994–95. Health Reports. 1996;7:33–45. Kalmijn M. The socioeconomic assimilation of Caribbean American blacks. Social Forces. 1996;74:911–930. Model S. Caribbean Immigrants: A black success story? Int Migr Rev. 1991;25:248–276. Szapocznik J, Kurtines W. Acculturation, biculturalism, and adjustment among Cuban Americans. In: Padilla A. ed. Acculturation: Theory, Models, and New Findings. Boulder, CO: Westview Press; 1980.
AEP Vol. 9, No. 4 May 1999: 236–244
36. Bettes BA, Dunsenbury L, Kerner J, James-Ortiz S, Botvin GJ. Ethnicity and psychosocial factors in alcohol and tobacco use in adolescence. Child Develop. 1990;61:557–565. 37. Patrick DL, Cheadle A, Thompson DC, Diehr P, Koepsell T, Kinne S. The validity of self-reported smoking: A review and meta analysis. Am J Public Health. 1994;84:1086–1093. 38. Clark PI, Gautam SP, Hlaing WM, Gerson LW. Response error in self-reported current smoking frequency by black and white established smokers. Ann Epidemiol. 1996;6:483–489. 39. Fang J, Madhavan S, Alderman MH. Influence of nativity on cancer mortality among black New Yorkers. Cancer. 1997;80:129–135. 40. Walker ARP, Walker BF, Segal I. Cancer patterns in three African populations compared with the United States black population. Eur J Cancer Prev. 1993;2:313–320.