Reports on Selected Racial/Ethnic Groups Special Focus: Maternal and Child Health Racial/Ethnic Differences in Smoking, Other Risk Factors, and Low Birth Weight Among Low-Income Pregnant Women, 1978-1988
Ronald R. Fichtner, Ph.D.
Because of the adverse effects of low birth weight (LBW) on infant morbidity and mortality, one of the 1990 health objectives for the nation has been to reduce the incidence of LBW to 5% among all live births in the United States. Public health surveillance of cigarette smoking during pregnancy has demonstrated an association between smoking and an increased risk of LBW, defined as birth weight of less than 2,500 g. For the period 1978-1988, information on nearly 248,000 women from CDC's Pregnancy Nutrition Surveillance System showed an LBW rate of 6.9%, a high prevalence of smoking during pregnancy (29.7%); and a strong association between smoking during pregnancy and the likelihood of delivering an LBW infant in all age, racial/ethnic, and prepregnancy weight groups. The risk of LBW was greater for smokers than for nonsmokers (9.9% versus 5.7%), creating an excess LBW risk of 4.2% associated with smoking. Overall, the average birth-weight deficit related to smoking was 178 g. Among both smokers and nonsmokers, black women had a higher percentage of LBW infants than did white women, and the risk of LBW related to smoking was greater among black women. That risk tended to increase with age, especially among women with low pregravid weight. Major reductions in LBW might be achieved if smoking were eliminated among pregnant women.INTRODUCTION
Birth weight has been found to be the primary predictor of infant survival (1). The incidence of low birth weight (LBW), defined as less than 2,500 g ( less than 5 lb, 8 oz), remains a major public health concern in the United States. Because two-thirds of all infants who die had LBW, the 1990 health objectives for the nation state "that low birthweight babies should constitute no more than five percent of all live births" (2). Blacks are disproportionately burdened by LBW. Of births in 1975, 13.1% of black infants and 6.3% of white infants had LBW (a relative risk of 2.1); in 1987, 12.7% of black infants and 5.7% of white infants had LBW (a relative risk of 2.2 for blacks) (3). Thus, although there has been a general decline in LBW for both blacks and whites, the approximate 2:1 black-to-white ratio has remained relatively stable. LBW data for other racial/ethnic groups are less reliable, but the LBW risks for most of these groups are more comparable with the risks for whites than those for blacks. Accordingly, the 1990 health objectives set a second birth-weight goal, that "no racial or ethnic group of the population should have a rate of low birth-weight infants that exceeds nine percent of all live births" (2).
Evidence clearly supports the negative effects of cigarette smoking during pregnancy on birth weight and shows a dose-response relationship between number of cigarettes smoked during pregnancy and a corresponding decline in birth weight (4). Intrauterine hypoxia, not decreased caloric consumption during pregnancy, is the likely mechanism that links cigarette smoking to LBW (5).
The 1990 health objectives, accordingly, set a goal that "the proportion of women who smoke during pregnancy should be no greater than one-half the proportion of women overall who smoke" (2). National estimates for 1987 indicate that 34% of women between 18 and 24 years of age smoked, whereas, for 1988, an estimated 11% of women between 12 and 17 years of age smoked (6). According to 1985-1986 data from 26 states participating in CDC's Behavioral Risk Factor Surveillance System, 21% of women ages 18-45 years who self-reported that they were pregnant at the time they were interviewed by telephone also reported that they were current smokers, compared with 30% of those not pregnant (7). These data also showed that white pregnant women were more likely to smoke (22%) than black pregnant women (16%) and that among those who smoked, white pregnant women smoked more cigarettes per day than black women.
Through the cooperation of participating state health departments, CDC's Pregnancy Nutrition Surveillance System (PNSS) monitors the prevalence of various risk factors for adverse pregnancy outcomes among low-income pregnant women who participate in publicly funded programs (8). Depending on local requirements, many low-income pregnant women are eligible to participate in public-sector maternal- and child-health clinic programs, such as the U.S. Department of Agriculture's Supplemental Food Program for Women, Infants, and Children (WIC). Clinic visits associated with these programs generate data records that are submitted to the PNSS. Begun in 1978, the PNSS is coordinated by the Division of Nutrition, Center for Chronic Disease Prevention and Health Promotion, CDC. In 1988, the 14 states that participated in the PNSS contributed data records concerning approximately 87,000 pregnancies.
The PNSS is part of CDC's public health surveillance effort to increase national and state capacity to quantify preventable risk behaviors during pregnancy, analyze the relationships of nutritional and behavioral status during pregnancy to birth outcomes, and intervene appropriately. This report examines the prevalence of smoking in the PNSS population and assesses the risks of delivering an LBW infant by racial/ethnic group, smoking status, pregravid weight status, and other risk factors included in the PNSS.METHODS
To study the relationship that smoking and other risk factors have to LBW among women attending WIC clinics, the Division of Nutrition analyzed nearly 248,000 records from the PNSS. These records corresponded to pregnancies that resulted in births between January 1, 1978, and December 31, 1988, among women enrolled in WIC. Only records from states that provided data on smoking were used. Data were used from pregnancies that resulted in live, singleton births, for which the mother had at least one prenatal WIC clinic visit. Adequate data on pregnancy outcome, including infant birth weight, were also required. In order to minimize inclusion of erroneous birth weight data, records with birth weights less than 1,000 g and greater than 6,000 g were excluded. Data were restricted to records from women who had either a hematocrit or hemoglobin level recorded and who, at their first WIC clinic visit, gave their ages as between 15 and 44. All data not related to birth outcome were obtained from the first clinic visit, regardless of the pregnancy trimester in which that visit occurred. Plausibility criteria were established for gestational age, number of previous live births, height, pregravid weight, hemoglobin, and hematocrit. Unlikely values were changed to missing. Smoking status at the time of the first prenatal clinic visit was ascertained by the question, "Are you currently smoking cigarettes?" Women were classified, on the basis of their reported pregravid weight and height, into one of three pregravid body mass index (BMI) groups: underweight, normal weight, and overweight. BMI was defined as weight(kg)/(height(m))2. Underweight was defined as a pregravid BMI less than the 15th percentile ( less than 19.2 BMI units) of the National Health and Nutrition Examination Survey II (NHANES II) reference population of women between the ages 20 and 29 (9). Overweight was defined as a pregravid weight greater than the 85th percentile ( greater than 27.2 BMI units). Anemia status was defined according to CDC criteria (10). Accordingly, adjustments were made to hematocrit and hemoglobin levels among pregnant women who reported smoking. Gestational age was calculated in units of weeks from the mother's self-reported estimated date of confinement and the infant's birth date, but if the estimated date of confinement was not reported, the gestational age was estimated from the self-reported date of last menstrual period.RESULTS
The PNSS population used for this analysis was approximately 51% white, 33% black, and 12% Hispanic (Table 1). More than half of the women were less than 25 years of age, and 27% were between 15 and 19 years of age. Other characteristics of the population include the prevalences of anemia (17%) and cigarette smoking (30%). Large overall differences in LBW were found between preterm and full-term infants, among racial/ethnic groups, between smokers and nonsmokers, and among pregravid BMI classifications.
The prevalence of smoking varied by age, racial/ethnic, pregravid BMI, and anemia categories (Table 2). The most striking difference in smoking prevalence was by race/ethnicity. Whites had the highest prevalence of smoking (approximately 40%) in every age group (Figure 1). However, blacks, who had only a 10% smoking prevalence in the 15- to 19-year age group, had smoking prevalences of 22% and 28% in the 20- to 24-year and 25- to 29-year age groups, respectively. Hispanics, American Indians, and Asians had the lowest smoking prevalences, which showed little variation by age.
The relationship between smoking and LBW by other factors shows that smokers had a higher risk of having an LBW child in every category (Table 2). The difference in risk for having an LBW infant between smokers and nonsmokers was found to be greatest among blacks, to increase with age, to be greater when infants were premature, and to be greatest among pregravid, underweight women.
In general, the percentage of LBW infants born to black and white women, among both smokers and nonsmokers, increased as pregravid BMI decreased (Figures 2-4). However, in every age and pregravid BMI group, for both blacks and whites, smokers had a greater risk of having an LBW infant than nonsmokers. The excess risk of LBW associated with smoking was greater for blacks than for whites and was more extreme for normal weight and underweight pregravid categories.
Because the frequency of preterm infants is not equal among the age, race, and pregravid BMI groups, an additional analysis was restricted to pregnancies that led to full-term deliveries. That analysis yielded the same relations as described previously, except for a lower proportion of LBW and a slightly diminished difference in LBW between smokers and nonsmokers.DISCUSSION
The prevalence of smoking is higher among PNSS participants than it is in the general population (6,7). This higher prevalence may be related to the low-income status of participants in public programs. Black women had lower LBW rates in the PNSS than in the general population, whereas the LBW rates for white women in the PNSS were comparable with those in the general population (3). Nevertheless, among participants in this study, clear risk factors for LBW are evident. This analysis indicates that among blacks and whites, for every age and pregravid BMI group, smokers had a greater risk of having an LBW infant than nonsmokers. The difference between smokers and nonsmokers in LBW risk was greater for blacks than for whites. It increased with age, and it was greatest in the pregravid underweight categories and elevated in normal weight categories when compared with overweight categories. Even among young women ages 15-19, the overall prevalence of smoking (28%) was substantial, and the LBW risk difference associated with smoking (2.7%) was elevated (11). Data in the surveillance summary must be interpreted with caution because of the following limitations. The data were program-based and represented women attending WIC clinics from a nonrandom selection of 13 participating states. In many states, smoking was a selection criterion for program eligibility, which may have inflated the smoking prevalences above that of low-income women in general. Prepregnancy weight, smoking status, estimated date of confinement, date of last menstrual period, age, and number of previous live births were based on information provided by the mother at the first prenatal WIC clinic visit and were not verified. Information on birth outcomes--such as birth weight of the child--was, in most instances, based on a history provided by the mother at a postpartum clinic visit. These histories were not verified.
During the period of study from 1978 to 1988, the PNSS excluded records on pregnancies for which there was no postpartum information available. Thus, the effect on birth outcomes, especially birth weight, among women who had prenatal WIC participation but who either made no postpartum visit or for whom, for other reasons, postpartum information was unavailable, is not calculable.
Information about smoking status was self-reported at a single point during pregnancy. No information about a woman's smoking history before pregnancy or about numbers of cigarettes smoked per day during pregnancy was collected. Furthermore, other medical, social, behavioral, and dietary factors affect LBW, but data on these factors were not available (12). However, recent changes to the PNSS will expand the collection of data on smoking as well as on other risk factors, such as alcohol use, and will also increase the number of participating states (8). These improvements will enhance future analyses.
In addition to the expansion of the PNSS, CDC is assisting states in the area of maternal and child health through the Pregnancy Risk Assessment Monitoring System (PRAMS), coordinated by the Division of Reproductive Health, Center for Chronic Disease Prevention and Health Promotion. PRAMS, a population-based surveillance system now in place in seven states, collects information on maternal behaviors (e.g., smoking) during pregnancy, problems with access to prenatal care, and maternal and infant health complications. These data will also strengthen future analyses.
Despite limitations in these PNSS data, blacks, including smokers and nonsmokers, are found to be at greater risk of delivering LBW infants than are whites. Moreover, the LBW risk difference associated with smoking was greater among blacks, and it increased with age, especially among women with low pregravid weight. These results argue powerfully for smoking prevention and intervention measures directed toward low-income, pregnant women, particularly those who are at high risk for giving birth to LBW infants. CDC has initiated efforts toward this end through the Smoking Cessation in Pregnancy (SCIP) program, coordinated by the Division of Reproductive Health, which provides technical assistance to states in developing smoking-cessation programs for pregnant women.
All MMWR HTML documents published before January 1993 are electronic conversions from ASCII text into HTML. This conversion may have resulted in character translation or format errors in the HTML version. Users should not rely on this HTML document, but are referred to the original MMWR paper copy for the official text, figures, and tables. An original paper copy of this issue can be obtained from the Superintendent of Documents, U.S. Government Printing Office (GPO), Washington, DC 20402-9371; telephone: (202) 512-1800. Contact GPO for current prices.**Questions or messages regarding errors in formatting should be addressed to email@example.com.
Page converted: 08/05/98
This page last reviewed 5/2/01