Tobacco Use Among US Racial/Ethnic Minority Groups, Report of the Surgeon General [3]
Images (114)
Showing first 60 of 114 pages.
Document
| id |
id
26413746
|
|---|---|
| contentType |
contentType
document
|
| source |
source
import
|
Source image fields (6)
Extracted text
OCR Page 1 of 114Surgeon General's Report
tobacco). However, this information may not be fully
An analysis of the data from the Coronary
valid, resulting in misclassification of exposure to ciga-
Artery Risk Development in (Young) Adults Study
rette smoking. A previous report of the Surgeon Gen-
(CARDIA) showed that there were higher rates of
eral reviewed the classification of cigarette smoking
misclassification in terms of self-reported nonsmok-
status and the consequences of misclassification
ers who had serum cotinine levels of at least 14 ng/
(USDHHS 1990).
mL among African Americans (5.7 percent) than
Misclassification of smoking information merits
among non-Hispanic whites (2.8 percent) (Wagen-
consideration in investigating tobacco use among
knecht et al. 1992). Alternative explanations for
racial/ethnic populations, because of the potential for
underreporting, such as more efficient smoking and
bias in comparing the effects of smoking across racial/
differences in cotinine metabolism, could not be
ethnic groups. To date, such bias has not been identi-
excluded.
fied, although several studies show that Hispanics may
Two additional studies examined the relation-
underreport cigarette smoking. In a population-based
ship between ancestry of origin and levels of biochemi-
survey in New Mexico, Coultas and colleagues (1988)
cal markers in smokers. In a study of participants in
compared self-reports of smoking against salivary
CARDIA, African American smokers demonstrated
cotinine level (a product of nicotine that has been used
higher cotinine levels than non-Hispanic white smok-
as a measure of exposure to nicotine) and end-tidal car-
ers after controlling for several dimensions of cigarette-
bon monoxide concentration. Based on the question-
smoking behavior (Wagenknecht et al. 1990). Lactose
naire results, the age-standardized prevalence rates of
intolerance, which elevates breath hydrogen concen-
current smoking were 30.9 and 27.1 percent for His-
tration, may increase the apparent level of expired air
panic men and women, respectively. After adjusting
carbon monoxide, a readily measured marker of ac-
for cotinine and carbon monoxide levels, these percent-
tive smoking (McNeill et al. 1990). Lactose intolerance
ages were 39.1 and 33.2. The rate of misclassification
is common in a number of racial/ethnic groups, in-
was greater in self-reported former smokers than in
cluding Asian Americans and African Americans.
(b)
never smokers, but self-reported never smokers also
/
had levels of cotintine and carbon monoxide indica-
tive of active smoking.
Classification of Race/Ethnicity
Using information from the Hispanic Health and
Nutrition Examination Survey (HHANES), Pérez-
The data included in this chapter are derived
Stable and colleagues (1992) documented the
from diverse sources, including vital statistics, cancer
misclassification of smoking status through compari-
registries, and epidemiological studies on smoking.
Race/ethnicity has been classified in these studies us-
sons of self-reports with serum cotinine levels. Among
65 Mexican American former smokers participating in
ing various techniques, including designation on death
the HHANES in 1982 through 1983, 7 (10.8 percent)
certificate, classification according to cancer registry
had a cotinine level indicative of active smoking;
protocols, self-reports, birthplace, language use, and
among 124 reported never smokers, 5 (4 percent) were
surname. The validity of each of these approaches is
probably active smokers based on their cotinine lev-
undoubtedly imperfect; moreover, validity varies
els. In a number of surveys, Hispanics, particularly
across regions and over time. However, comprehen-
Latino groups in the southwestern and western United
sive assessments of the validity of racial/ethnic mi-
States, have been found to smoke about one-half pack
nority classification in various types of health data
have not been reported.
of cigarettes per day, compared with non-Hispanic
The limited information available indicates some
whites who typically report smoking one pack per day
potential for misclassification. For example, Frost
(Coultas et al. 1994). Pérez-Stable and colleagues (1992)
and colleagues (1992) compared the classification of
used data from 547 Mexican American participants in
"Native American," as recorded by the Seattle-Puget
the HHANES to examine underreporting of cigarette
Sound registry of the Surveillance, Epidemiology, and
consumption using the ratio of serum cotinine to
End Results (SEER) Program against an Indian Health
self-reports of the number of cigarettes smoked per
Service (IHS) registry of patients eligible for services.
day as the "gold standard." This study found that
A substantial portion of patients with invasive cancer
among Mexican Americans, 20.4 percent of men
in the IHS registry were not similarly classified
and 24.7 percent of women who were self-reported
by the Seattle-Puget Sound cancer registry. Similarly,
smokers underreported smoking between one and
an injury registry for the state of Oregon under-
nine cigarettes per day. Self-reported Mexican Ameri-
counted those with injuries (Sugarman et al.
can smokers who reported smoking greater numbers
1993). Using data from the National Longitudinal
of cigarettes per day underreported less frequently.
Mortality Study, Sorlie and colleagues (1992) compared
demographic characteristics reported on the CPS of the
186 Chapter 3
Relations
belongs_to