Jane Schulman

Congenital Malformations and Birthweight in Areas with Potential Environmental Contamination

Public concern exists about the potential for reproductive damage that may result from exposures to environmental contaminants. Therefore, the authors sought to determine if there was an association between a childs congenital malformation or a childs lowered weight at birth and his or her mothers residence in a census tract where a site of environmental contamination had been documented. Exposure designations were derived from existing sources of information. Except for an elevated risk (odds ratio = 1.5) for infants with malformations of the heart and circulatory system, this investigation did not reveal increased risks for most malformations or for lowered birthweight among babies born to women who lived in these census tracts. Methodologic issues inherent to investigations that rely on existing data sources are discussed.

Sudden infant death syndrome and birth weight

We studied 2962 cases of sudden infant death syndrome (SIDS), derived from linked birth and death records, to specify further the descriptive epidemiologic data on recognized SIDS risk factors and to examine interrelationships among multiple risk factors and SIDS while controlling for the influence of birth weight (using logistic regression). The results generally confirmed those of other studies, with the exception of a higher incidence of SIDS among Chinese and Japanese babies and a lower incidence among Hispanic babies, all in comparison with white, non-Spanish (Anglo) infants. Median age at death was found to vary by birth weight, with very low birth weight babies being about 6 weeks older (postnatal age) than normal birth weight babies at time of death. The association of the risk factors with SIDS remained when birth weight data were statistically controlled; the association of these risk factors with SIDS cannot be explained by their relationship to birth weight. An interaction was found between race and maternal age and between multiparity and type of hospital of birth. There was no interaction between birth weight and the other risk factors.

Distance and risk measures for the analysis of spatial data: a study of childhood cancers.

Three statistical approaches, used to detect spatial clusters of disease associated with a point source exposure, are applied to childhood cancer data for the city of San Francisco (1973-88). The distributions of incident cases of leukemia (51 cases), brain cancer (35 cases), and lymphatic cancer (37 cases) among individuals less than 21 years of age are described using three measures of clustering: distance on a geopolitical map, distance on a density equalized transformed map, and relative risk. The point source of exposure investigated is a large microwave tower located southwest of the center of the city (Sutro Tower). The three analytic approaches indicate that the patterns of the major childhood cancers are essentially random with respect to the point source. These results and a statistical model for spatial clustering are used to explore distance and risk measures in the analysis of spatial data. Both types of measures of spatial clustering are shown to perform similarly when a specific area of exposure can be defined.

Congenital cardiac anomalies relative to selected maternal exposures and conditions during early pregnancy

Recently, several studies have explored possible relations between various exposures to women and congenital cardiac anomalies in their infants (1, 12-17). We also investigated potential relations between a variety of maternal exposures or conditions during pregnancy and congenital cardiac anomalies using data from a study designed to estimate risks between maternal exposure to drinking water and congenital cardiac anomalies (11). Approximately 70,000 children born alive in Santa Clara County, California from January 1981 through December 1983 were eligible. Cases, ascertained by the California Birth Defects Monitoring Program, were children with congenital cardiac disease diagnosed by echocardiogram, catheterization, surgery, or autopsy up to one year of age. Excluded were cases with: chromosomal anomalies, isolated ventricular and atrial septal defects (unless diagnosed by catheterization or surgery or associated with heart failure), isolated patent foramen ovale, patent ductus arteriosus among infants < 38 weeks gestation, isolated pulmonic stenosis, arrhythmia, or functional murmur, unless accompanied by congestive heart failure. Control children were randomly selected from vital statistics files of all livebirths without a cardiac anomaly in the county for the same time period. Of 172 cases that met diagnostic criteria, 141 (82.0%) mothers were interviewed. Of the 210 control infants, 176 mothers (83.80/0) were interviewed. Interviews were conducted with mothers primarily by telephone an average 4.6 (range 2.9 to 6.7 years, cases and controls were the same) years after the birth. The effect of maternal exposures or conditions on the risk of a child being born with a cardiac anomaly was estimated by the odds ratio (OR) and its 95°/0 confidence interval. Analyses were performed with all cases as well as with a subset of cases having defects of the conotruncus (truncus arteriosus communis, including aortopulmonary window; tetralogy of Fallot; dextro-transposition of the great vessels; double outlet fight ventricle; pulmonary atresia with a ventricular septal defect; or interruption of the aortic arch) (3). Approximately two-thirds of the diagnoses for the 141 cases were: ventricular septal defect, atrial septal defect, transposition of the great vessels, coarctation of the aorta, hypoplastic left heart, tetralogy of Fallot, pulmonary valve stenosis, or pulmonary valve atresia. Cases and controls were similar with respect to parental race, maternal age, child sex, and parental education. For all cardiac anomalies as a group, crude ORs greater than 1.5 were observed for epilepsy (OR = 3.8), diabetes (OR = 2.2), arthritis (OR = 1.8), insulin (OR = 5.1), anticonvulsant (OR = 2.5), cigarette use (< 0.5 pack per day) (OR = 1.9), heavy alcohol use (OR = 1.7) (defined as 4 or more drinks per occasion at least once per week, or 2-3 drinks per occasion if drank at least several times a week), and use of oral contraceptives (OR = 2.1) near the time of conception

U.S. Air Force Recruit Injury and Health Study

OBJECTIVES To assess the types, rates, and risks of injury for male and female USAF recruits. DESIGN Outpatient visits for female (5250) and male recruits (8656) were collected and analyzed for rate of injury, types of injuries, and risk of injury throughout a 6-week training period. RESULTS One third of female recruits and one sixth of male recruits were injured at least once during recruit training. The overall rate for injuries in women and men was 63.0 and 27.8 per 1000 person-weeks, respectively. The adjusted relative risk for women versus men for all injuries was 2.22, and was consistent (1.67 to 3.27) across injury sites. Despite declining absolute rates of injury by week (106.1-13.4 for women and 53.7-13.2 for men), relative risk of injury for women versus men remained fairly constant throughout each training week. The relative risk for injury serious enough to result in medical hold was 1.69 for women vis-a-vis men. Approximately half of all medical discharges for women and men were for injuries. CONCLUSIONS Female recruits were injured twice as often as male recruits, and were 1.5 times more likely to be removed from their training cohort for injury. Relative risk for injuries to specific body areas remained fairly consistent, indicating that no gender-specific injuries were occurring. Further efforts to determine the cause of injuries should be undertaken, and interventions aimed at reducing the disparate risk of injuries in women should be developed and evaluated.

Early-stage breast cancer treatment among medically underserved women diagnosed in a national screening program, 1992–1995

AbstractBackground. Little research has been conducted on the breast cancer treatment of low income, underserved women. This study was designed to describe initial treatment of breast cancer among low-income women diagnosed through federally funded screening programs in Detroit, Michigan, and the states of New Mexico and California; and to compare the treatment received by program women with early-stage breast cancer with that of all women diagnosed in those regions. Methods. Data from the three screening programs were linked with cancer registry data from the corresponding geographic areas. All women diagnosed between 1992 and 1995 through the state-based screening programs and all women contemporaneously diagnosed with breast cancer in the three regions were studied. Descriptive analyses were done of the proportion of women with breast cancer receiving treatment; the proportion of early-stage breast cancer (stage I or II) cases treated with breast-conserving surgery, and the proportion treated with mastectomy; and among women with breast-conserving surgery, the proportion receiving radiation therapy. Logistic regression models controlled for age and stage at diagnosis, race or ethnicity and geographic region. Results. Less than 2% of program women diagnosed with breast cancer received no treatment. More than two of five women with early-stage breast cancer underwent breast-conserving surgery, with 72% of these women receiving radiation therapy. Multivariate regression analysis revealed that women with stage IIA or IIB breast cancer had lower odds of undergoing breast-conserving surgery than women with stage I (0.51 [95% CI = 0.30–0.87] and 0.36 [95% CI = 0.19–0.70], respectively). Women over age 65 and those with incompletely staged cancer had the lowest odds for receiving radiation therapy after breast-conserving surgery (0.29 [95% CI = 0.09–0.99] and 0.14 [95% CI = 0.03–0.72], respectively). Women diagnosed through the screening programs had odds of undergoing breast-conserving surgery similar to those of all women in the regions (1.11 [95% CI = 0.89–1.39]). Conclusions. Treatment patterns for women diagnosed with early-stage breast cancer through three state-based screening programs appear to have been similar to those reported in the literature. In addition, their treatment appears to have been similar to that of other women during the same time period.

Exposure Misclassification Due to Residential Mobility during Pregnancy in Epidemiologic Investigations of Congenital Malformations

This study addressed the question of how maternal migration between conception and birth affects estimates of risk in studies of congenital malformations when movement is related to the exposure. For example, in studying the potential association between proximity to a chemical waste site and the occurrence of birth defects, incorrect inferences might be drawn if maternal residence at birth was used as a surrogate for exposure at conception in the case when a significant amount of media attention influenced some women to move away from the site after becoming pregnant. A simple statistical model is proposed that defines the distance to a fixed exposure point measured at birth as a function of the distance to the point measured at conception, the probability of movement, the direction of movement, and the distance moved. Bias is the difference between the expected results when distance is measured at birth versus conception. The amount of bias can be substantial for movement patterns that may be likely to occur. This simplified model was used in an effort to explore and better understand the relationships between maternal migration and risk.

Sudden infant death syndrome among Asians in California

We analyzed California linked birth and infant death records for 1978 to 1985 to determine the ethnic-specific incidence of sudden infant death syndrome in five Asian-American subgroups, and to assess the association of sudden infant death syndrome with cultural assimilation. The overall incidence of sudden infant death syndrome for these groups was 1.1/1000 live births (194 cases). The ethnic-specific incidence was statistically associated with the immigration status of each ethnic group, as measured by the proportion of all live births for which the mother was born in the United States, with a higher incidence for groups that have been in the United States for the longest period. A logistic model was used to examine simultaneously the association of sudden infant death syndrome with maternal ethnicity, ethnic homogeneity or heterogeneity of the parents, and maternal birthplace (United States, or elsewhere). The logistic analysis did not yield statistically significant evidence to support cultural assimilation as a factor in the incidence of sudden infant death syndrome for these groups. This finding may be due to small sample size and inadequate indicators of assimilation. It may also be that other factors relating to immigration and selective migration affect the incidence of sudden infant death syndrome among Asian-Americans.

Interpoint squared distance as a measure of spatial clustering

The expectation and variance for the mean interpoint squared distance are presented. In order to evaluate these expressions it is necessary to calculate the moments of a bivariate uniform distribution defined over an arbitrary polygon. Expressions for these moments are presented, allowing the mean interpoint squared distance to be used as a measure of spatial clustering. The distribution and power of this test statistic is explored on the unit square, and the spatial distribution of 11 cases of non-Hodgkins lymphoma is investigated to illustrate an application of the approach.

Detection of Excess Disease Near an Exposure Point: A Case Study

Many studies have evaluated the likelihood of adverse health effects associated with environmental contamination from point source exposures. Two statistical measures used in these studies are the ratio, R, of the observed to the expected number of cases occurring in the area containing the point and the average distance, D, between the cases and the point. This paper estimates the probability of detecting an association between disease and exposure when one actually exists (power) for R and D for several specific and plausible statistical models. Results are presented in the context of rare diseases such as congenital malformations. The practical implications of using these measures for the evaluation of risk of disease in environmental epidemiologic investigations are discussed.