Cardiovascular severe maternal morbidity in pregnant and postpartum women: development and internal validation of risk prediction models.

OBJECTIVES: To develop and internally validate risk prediction models identifying women at risk for cardiovascular severe maternal morbidity (CSMM). DESIGN: A retrospective cohort study. SETTING: An obstetric teaching hospital between 2007 and 2017. POPULATION: A total of 89 681 delivery hospitalisations. METHODS: We created and evaluated two models, one predicting CSMM at delivery (delivery model) and the other predicting CSMM postpartum following discharge from delivery hospitalisation (postpartum CSMM). We assessed model discrimination and calibration and used bootstrapping for internal validation. MAIN OUTCOME MEASURES: Cardiovascular severe maternal morbidity comprised the following confirmed conditions: pulmonary oedema/acute heart failure, myocardial infarction, aneurysm, cardiac arrest/ventricular fibrillation, heart failure/arrest during surgery or procedure, cerebrovascular disorders, cardiogenic shock, conversion of cardiac rhythm and difficult-to-control severe hypertension. RESULTS: The delivery model contained 11 variables and 3 interaction terms. The strongest predictors were gestational hypertension, chronic hypertension, multiple gestation, cardiac lesions or valvular heart disease, maternal age ≥40 years and history of poor pregnancy outcome. The postpartum model comprised eight variables. The strongest predictors were severe pre-eclampsia, non-Hispanic Black race/ethnicity, chronic hypertension, gestational hypertension, non-severe pre-eclampsia and maternal age ≥40 years at delivery. The delivery and postpartum models had an area under the receiver operating characteristic curve of 0.87 (95% CI 0.85-0.89) and 0.85 (95% CI 0.80-0.90), respectively. Both models were adequately calibrated and performed well on internal validation. CONCLUSIONS: These tools may help providers to identify women at highest risk of CSMM and enable future prevention measures. TWEETABLE ABSTRACT: Risk assessment tools for cardiovascular severe maternal morbidity were developed and internally validated.

Maternal influenza and birth outcomes: systematic review of comparative studies.

BACKGROUND: Although pregnant women are considered at high risk for severe influenza disease, comparative studies of maternal influenza and birth outcomes have not been comprehensively summarised. OBJECTIVE: To review comparative studies evaluating maternal influenza disease and birth outcomes. SEARCH STRATEGY: We searched bibliographic databases from inception to December 2014. SELECTION CRITERIA: Studies of preterm birth, small-for-gestational-age (SGA) birth or fetal death, comparing women with and without clinical influenza illness or laboratory-confirmed influenza infection during pregnancy. DATA COLLECTION AND ANALYSIS: Two reviewers independently abstracted data and assessed study quality. MAIN RESULTS: Heterogeneity across 16 studies reporting preterm birth precluded meta-analysis. In a subgroup of the highest-quality studies, two reported significantly increased preterm birth (risk ratios (RR) from 2.4 to 4.0) following severe 2009 pandemic H1N1 (pH1N1) influenza illness, whereas those assessing mild-to-moderate pH1N1 or seasonal influenza found no association. Five studies of SGA birth showed no discernible patterns with respect to influenza disease severity (pooled odds ratio 1.24; 95% CI 0.96-1.59). Two fetal death studies were of sufficient quality and size to permit meaningful interpretation. Both reported an increased risk of fetal death following maternal pH1N1 disease (RR 1.9 for mild-to-moderate disease and 4.2 for severe disease). CONCLUSIONS: Comparative studies of preterm birth, SGA birth and fetal death following maternal influenza disease are limited in number and quality. An association between severe pH1N1 disease and preterm birth and fetal death was reported by several studies; however, these limited data do not permit firm conclusions on the magnitude of any association. TWEETABLE ABSTRACT: Comparative studies are limited in quality but suggest severe pandemic H1N1 influenza increases preterm birth.

The effect of labour induction on the risk of caesarean delivery: using propensity scores to control confounding by indication.

OBJECTIVE: To use propensity score methods to control for confounding by indication in the association between labour induction and caesarean delivery. DESIGN: Cross-sectional analysis of administrative hospital discharge data supplemented by medical record information. SETTING: Fourteen US member hospitals of the National Perinatal Information Center. SAMPLE: A cohort of 166 559 singleton liveborn deliveries in the period 2007-2012. METHODS: We used propensity scores (PSs) to balance 83 covariates between induced and non-induced women, and compared estimates with traditional covariate adjustment. We estimated PSs for labour induction versus expectant management of pregnancy each week from 34 to 42 weeks of gestation. We estimated risk ratios (RRs) for the association between labour induction and primary caesarean delivery from models with no adjustment, traditional adjustment of five covariates, matched PS, and adjustment for continuous PS. MAIN OUTCOME MEASURE: Caesarean delivery in current or subsequent week of gestation. RESULTS: In crude models labour induction increased the risk of caesarean delivery in all weeks (RR 1.06-1.52), excepting 39 weeks of gestation (RR 0.89). After matching on PS, the analysis showed a significantly decreased risk of caesarean delivery with labour induction during weeks 35-39 (RR 0.77-0.92), and a significantly elevated risk at weeks 40 (RR 1.22) and 41 (RR 1.39). Traditional covariate and PS adjustment resulted in RRs between those from crude and PS-matched models. CONCLUSIONS: There is evidence of considerable confounding by indication in the association of labour induction and caesarean delivery, particularly for preterm deliveries. Using PS methods, we found a reduced risk of caesarean delivery with labour induction before 40 weeks of gestation, and an elevated risk for weeks 40-42. TWEETABLE ABSTRACT: With confounding adjustment, labour induction does not increase the risk of caesarean at 34-39 weeks of gestation.

Disparities in cesarean delivery by ethnicity and nativity in New York city.

Our objective was to examine differences in risk of cesarean delivery among diverse ethnic groups in New York City. Using cross-sectional New York City birth and hospitalization data from 1995 to 2003 (n = 961,381) we estimated risk ratios for ethnic groups relative to non-Hispanic whites and immigrant women relative to US-born women. Adjusting for insurance, pre-pregnancy weight, maternal age, education, parity, birthweight, gestational age, year, medical complications, and pregnancy complications, all ethnic groups except East Asian women were at an increased risk of cesarean delivery, with the highest risk among Hispanic Caribbean women [adjusted risk ratio (aRR) = 1.27, 95 % CI (confidence interval) = 1.24, 1.30] and African American women (aRR = 1.20, 95 % CI = 1.17, 1.23). Among Hispanic groups, immigrant status further increased adjusted risk of cesarean delivery; adjusted risk ratios for foreign-born women compared to US-born women of the same ethnic group were 1.27 for Mexican women (95 % CI = 1.05, 1.53), 1.23 for Hispanic Caribbean women (95 % CI = 1.20, 1.27), and 1.12 for Central/South American women (95 % CI = 1.04, 1.21). Similar patterns were found in subgroup analyses of low-risk women (term delivery and no pregnancy or medical complications) and primiparous women. We found evidence of disparities by ethnicity and nativity in cesarean delivery rates after adjusting for multiple risk factors. Efforts to reduce rates of cesarean delivery should address these disparities. Future research should explore potential explanations including hospital environment, provider bias, and patient preference.

First-trimester bleeding characteristics associate with increased risk of preterm birth: data from a prospective pregnancy cohort.

BACKGROUND: Prior evidence linking first-trimester bleeding with preterm birth (PTB, <37 weeks gestation) risk has been inconsistent and may be biased by subject selection and/or incomplete documentation of bleeding episodes for all participants. Prior studies have not carefully examined the role of bleeding characteristics in PTB risk. In the present study, we estimate the association between first-trimester bleeding and PTB in a non-clinical prospective cohort and test whether bleeding characteristics better predict risk. METHODS: Women were enrolled in Right from the Start (2000-2009), a prospective pregnancy cohort. Data about bleeding and bleeding characteristics were examined with logistic regression to assess association with PTB. RESULTS: Among 3978 pregnancies 344 were PTB and 3634 term. Bleeding was reported by 986 (26%) participants. After screening candidate confounders, only multiple gestations remained in the model. Bleeding associated with PTB [odds ratio (OR)(adjusted) = 1.40, 95% confidence interval (CI) 1.09-1.80]. Risk did not vary by race/ethnicity. Compared with non-bleeders, PTB risk was higher for bleeding with red color (OR(adjusted) = 1.92, 95% CI, 1.32-2.82), for heavy episodes (OR(adjusted) = 2.40, 95% CI 1.18-4.88) and long duration (OR(adjusted) = 1.67, 95% CI 1.17-2.38). CONCLUSIONS: Bleeding associated with PTB was not confounded by common risk factors for bleeding or PTB. PTB risk was greatest for women with heavy bleeding episodes with long duration and red color and would suggest that combining women with different bleeding characteristics may affect the accuracy of risk assessment. These data suggest a candidate etiologic pathway for PTB and warrant further investigation of the biologic mechanisms.

Ethnic differences in the association between gestational diabetes and pregnancy outcome.

The purpose of the study was to examine ethnic variation in the impact of Gestational Diabetes Mellitus (GDM) on birth outcome. The authors examined the association between GDM and pregnancy-induced hypertension, macrosomia, primary Cesarean delivery, and preterm birth, using New York City Birth Certificate data from 2001-2006. Logistic regression was used to evaluate the crude and adjusted odds ratios of GDM with each adverse perinatal event, stratified by ethnicity. GDM was associated with increased risk of adverse perinatal events among all ethnic groups, with modest variation by ethnicity. Across ethnic groups, adjusted odds ratios comparing women with and without GDM ranged from 1.4-2.9 for pregnancy-induced hypertension, 1.0-2.2 for macrosomia, 1.1-1.8 for primary Cesarean delivery, and 1.3-1.8 for preterm birth. Overall, Caribbean, Sub-Saharan African, and African American women tended to show a larger relative impact of GDM, while North African, South Central Asian, and Chinese women showed a comparatively smaller impact of GDM. Although some ethnic variation was seen, differences in effect size were not large enough to support ethnic-specific thresholds for GDM diagnosis and treatment.

Ethnicity and gestational diabetes in New York City, 1995-2003.

OBJECTIVE: To characterise the patterns of occurrence of gestational diabetes among a wide range of ethnic groups that reside in New York City. DESIGN: Birth records and hospital discharge data were linked to more accurately assess the risk of gestational diabetes by ethnicity, compare risk in US-born to foreign-born women, and assess time trends. SETTING: New York City. POPULATION: All singleton live births occurring between 1995 and 2003. METHODS: Multivariable binomial regression analysis of ethnicity and gestational diabetes, yielding adjusted risk ratios with non-Hispanic white women as the referent. MAIN OUTCOME MEASURE: Diagnosis of gestational diabetes on birth certificate or in hospital discharge. RESULTS: Adjusted relative risks (aRRs) were modestly elevated for African-Americans and sub-Saharan Africans and somewhat higher (<2.0) for non-Hispanic Caribbeans, Hispanic Caribbeans, Central Americans, and South Americans. The aRR was 4.7 (95% CI = 4.6-4.9) for South Central Asians (with an absolute gestational diabetes risk of 14.3%), 2.8 (95% CI = 2.7-3.0) among South-East Asian and Pacific Islanders, and 2.3 (95% CI = 2.2-2.4) among East Asians. Among South Central Asians, the greatest risks were found for women from Bangladesh (aRR = 7.1, 95% CI = 6.8-7.3). Foreign-born women consistently had higher risk than US-born women. Risk for gestational diabetes increased over time among South Central Asians, some Hispanic groups, and African-Americans. CONCLUSIONS: Risk of gestational diabetes appears to vary markedly among ethnic groups, subject to potential artefacts associated with screening and diagnosis. These differences would have direct implications for health care and may suggest aetiologic hypotheses.

Stress and placental resistance measured by Doppler ultrasound in early and mid-pregnancy.

OBJECTIVES: To examine the association between self reports and biomarkers of stress and placental resistance (measured by Doppler ultrasound of the uterine and umbilical arteries), to determine if restriction of blood flow to the placenta is a mechanism by which stress might affect health during pregnancy. METHODS: Eight hundred and seventy-two women had ultrasound examinations of the uterine artery at 15-19 weeks' gestation and the uterine and umbilical arteries at 24-29 weeks, and resistance and pulsatility indices were calculated. Psychosocial stress was measured by telephone interview and self-administered questionnaire using several validated tools twice during the pregnancy. Cortisol and corticotropin-releasing hormone (CRH) were measured twice during the pregnancy. Linear and hierarchical models were used to examine the relationships among reported stress, stress hormones and placental Doppler indices. RESULTS: The umbilical artery resistance index was higher in younger women, those with less education, those who were single and those who smoked. The uterine artery pulsatility index was higher in women with pre-eclampsia, those living alone, those with high body mass index, and those who gained the least weight during pregnancy. A higher CRH level was associated with small increases in uterine artery pulsatility and umbilical artery resistance indices. Psychosocial measures of stress were not consistently associated with higher placental resistance. CONCLUSIONS: Increased CRH levels may be associated with increased placental resistance. Otherwise, these findings do not support the hypothesis that restriction of blood flow to the fetus is a major mechanism by which stress affects infant health.

Breast cancer mortality among female electrical workers in the United States.

BACKGROUND: Previous epidemiologic studies have suggested that exposure to electric or magnetic fields in occupational and residential environments may cause cancer. Recent experimental findings provide some support for the hypothesis that exposure to extremely low-frequency electromagnetic fields reduces the pineal gland's nocturnal production of the hormone melatonin, thereby increasing susceptibility to sex hormone-related cancers such as breast cancer. PURPOSE: Our purpose was to assess the evidence that cancer of the female breast might be associated with exposure to extremely low-frequency electromagnetic fields. METHODS: Records of women who had breast cancer as the underlying cause of their death (ICD-9 174) and control subjects (four per case) were selected from computer files of U.S. mortality data for the years 1985-1989. Women 20 years and older at the time of their death were eligible for inclusion if they were residents of and died in one of the 24 states that provided death certification records with occupation and industry codes to the National Center for Health Statistics for at least 1 year during the study interval. Data from death certificates were used to classify the case and control subjects with regard to potential occupational exposure to electric and magnetic fields. Control subjects were a random sample of women who died of any other underlying cause, excluding leukemia and brain cancer. RESULTS: The data analysis contrasted 68 women with breast cancer and 199 controls, all with electrical occupations, with 27,814 women with breast cancer and 110,750 controls, all of whom had other occupations. Electrical workers had excess mortality from breast cancer relative to other employed women [odds ratio (OR) = 1.38; 95% confidence interval (CI) = 1.04-1.82]). Adjusted ORs for specific electrical occupations were 1.73 (95% CI = 0.92-3.25) for electrical engineers, 1.28 (95% CI = 0.79-2.07) for electrical technicians, and 2.17 (95% CI = 1.17-4.02) for telephone installers, repairers, and line workers. There was no excess of breast cancer, however, in seven other occupations held more frequently by women and also involving potentially elevated electrical exposures, including telephone operators, data keyers, and computer operators and programmers. CONCLUSIONS: In light of the limitations inherent in death certification data and the design of this study, any conclusions regarding the hypothesis that exposure to extremely low-frequency electromagnetic fields causes breast cancer among women must be limited. Nevertheless, our findings are broadly consistent with that hypothesis and encourage further investigation with improvements in study design and data quality.

Epidemiology of cervical cancer by cell type.

A multicenter case-control study of 481 invasive cervical cancer patients and 801 population controls enabled comparison of risk factors for squamous cell tumors (n = 418), adenosquamous cancers (n = 23), and adenocarcinomas (n = 40). The epidemiology of the squamous cell tumors resembled that found in other studies, with the major risk factors being absence of Pap smear screening (relative risk = 3.6 to 4.8 for those not screened within 5 yr), multiple sexual partners (relative risk = 2.9 for over ten partners), and history of genital infections or sores (relative risk = 2.3). Although based on small numbers, adenosquamous tumors appeared to share some of these risk factors, notably number of sexual partners, years since last Pap smear, and level of education. Adenocarcinomas were not similarly affected, although sexual practices were marginally predictive. Obesity increased the risk of adenocarcinoma, but no other similarities to endometrial adenocarcinoma were observed. Smoking was a significant predictor of squamous cell tumors but did not affect adenocarcinomas. Extended use of oral contraceptives was a risk factor for all tumor types, especially adenocarcinoma, and a familial tendency to cervical cancer was also observed for all cell types.

Reduced risk of IDDM among breast-fed children. The Colorado IDDM Registry.

The hypothesis that breast-feeding can provide protection against the development of insulin-dependent diabetes mellitus (IDDM) and would, therefore, be less common among subjects with IDDM was tested with a retrospective design. Cases (n = 268) were selected from the Colorado IDDM Registry and the Barbara Davis Center for Childhood Diabetes (Denver, CO). Two control groups were recruited, one from physicians' practices throughout Colorado (n = 291) and the second through random-digit dialing from the Denver area (n = 188). Cases were less likely to have been breast-fed than controls after adjustment for birth year, maternal age, maternal education, family income, race, and sex [adjusted odds ratio (OR) = 0.70; 95% confidence interval (CI) = 0.50-0.97]. This finding was consistent for both control groups and by birth-year intervals. A greater decrease in risk of IDDM was seen among subjects who had been breast-fed to an older age (for breast-feeding duration of greater than or equal to 12 mo, adjusted OR = 0.54, 95% CI = 0.27-1.08). The amount of IDDM that might be explained by breast-feeding habits (population percentage attributable risk) ranged from 2 to 26%, varying according to the breast-feeding prevalence reported in other studies. Replication of this work in different populations, controlled for the strong secular trends in breast-feeding habits, is critical before the hypothesis of protection is accepted.

Glutathione S-transferases M1, T1, and P1 and breast cancer.

We examined associations for glutathione S-transferases M1 (GSTM1), T1 (GSTT1), and P1 (GSTP1) genotypes and breast cancer in the Carolina Breast Cancer Study, a population-based, case-control study in North Carolina. Odds ratios were close to the null value for each GST locus among African-American women (278 cases and 271 controls) and white women (410 cases and 392 controls), as well as pre- and postmenopausal women. For women with a history of breast cancer in one or more first-degree relatives, odds ratios were 2.1 (95% confidence interval, 1.0-4.2) for GSTM1 null and 1.9 (0.8-4.6) for GSTT1 null genotypes. Among women with a family history, age at diagnosis was significantly earlier for those with the GSTM1 null genotype. We did not observe strong evidence for modification of odds ratios for smoking according to GST genotypes. There was no evidence for combined effects of GSTM1, GSTT1, and GSTP1 genotypes, and there were no combined effects for GST genotypes and the catechol O-methyltransferase genotype. We conclude that GSTM1, GSTT1, and GSTP1 genotypes do not play a strong role in susceptibility to breast cancer. However, the role of GST genotypes in age at onset and risk of breast cancer among women with a family history merits further investigation.

Dichlorodiphenyldichloroethene, polychlorinated biphenyls, and breast cancer among African-American and white women in North Carolina.

We examined plasma dichlorodiphenyldichloroethene (DDE) and total polychlorinated biphenyl (PCB) levels in relation to breast cancer in a population-based, case-control study of African-American women (292 cases and 270 controls) and white women (456 cases and 389 controls) in North Carolina. Adjusted odds ratios (ORs) for breast cancer comparing the highest to lowest third of DDE were 1.41 [95% confidence interval (CI), 0.87-2.29] in African-American women and 0.98 (95% CI, 0.67-1.43) in white women. ORs comparing the highest to lowest third of total PCBs were 1.74 (95% CI, 1.00-3.01) in African-American women and 1.03 (95% CI, 0.68-1.56) in white women. Among African-Americans, the OR for total PCBs was highest for obese women (body mass index 234.2; OR, 4.92; 95% CI, 1.63-14.83). In contrast, the OR for DDE was highest for the leanest African-American women (body mass index, <25; OR, 3.84; 95% CI, 0.98-15.08). ORs for DDE were not elevated among women who lived or worked on farms or elevated among farming women who reported exposure to pesticides. Our results suggest absence of a strong effect for DDE or total PCBs in breast cancer but lend support for associations among subgroups of women. In our study, factors such as income, parity, breastfeeding, race/ethnicity, and body mass index influenced the relationship of organochlorines and breast cancer. Differing distributions of such factors may explain some of the inconsistencies across previous studies.

Econometric approaches to epidemiologic data: relating endogeneity and unobserved heterogeneity to confounding.

The concepts of endogeneity and unobserved heterogeneity are well-known among econometricians. However, these issues are rarely addressed in epidemiologic studies. This paper explores these two concepts, their relationship to each other, and the implications for analysis in epidemiologic studies. An endogenous variable is defined as a predictor variable which is partly determined by factors within the model itself, while unobserved heterogeneity is conceptualized as a vector of missing variables acting through the error term. Under certain assumptions, the simultaneous existence of an endogenous variable and unobserved heterogeneity is shown to act in a manner analogous to confounding. Specifically, this occurs due to an association between the error term in the equation and the endogenous predictor variable. The accepted econometric solution to this problem is to replace the endogenous variable with an 'instrumental variable' which is not correlated with the error term and thus not susceptible to confounding. The validity of these concepts and of the proposed solution are discussed.

Effect of a monetary incentive on response to a mail survey.

We assessed the effect of a $1 incentive on response to a two-page questionnaire which was sent to 8356 female cosmetologists between 22 and 36 years old. The study population was randomly assigned to one of three groups in which a $1 incentive was enclosed with either the first or second mailing, or with none of the mailings. Ten percent of questionnaires were returned by the postal service because of an incorrect address or death of the addressee and were omitted from response calculations. Of the remaining questionnaires, 79% were completed and returned after up to three mailings. The cumulative response was highest among cosmetologists who received a $1 incentive with the first mailing, (81%; 95% confidence interval (CI), 80 to 82), intermediate among those who received $1 with the second mailing (78%, 95% CI, 77 to 79), and lowest among cosmetologists who received no incentive (74%; 95% CI, 70 to 78). Characteristics of cosmetologists who responded after having received a "41 incentive were similar to those who responded without having received an incentive. The higher costs per response incurred by the use of an incentive must be weighed against the benefit of higher response.

Exercise during pregnancy among US women.

This report describes the prevalence and pattern of exercise during pregnancy among US women. It is based on 9953 randomly selected women who gave birth to live infants in 1988 and participated in the National Maternal and Infant Health Survey, a cross-sectional survey with a stratified systematic sampling in 48 states, the District of Columbia, and New York City. Self-reported exercise before and during pregnancy, specific activities, and number of months of exercise were examined in relation to maternal demographic characteristics, reproductive history, and prenatal care. Forty-two percent of all women reported exercising during pregnancy, half of whom exercised longer than 6 months. Walking was the leading activity (43% of all activities reported), followed by swimming (12%) and aerobics (12%). Older mothers and women who had multiple gestations, previous children, or an unfavorable reproductive history were less likely to exercise during pregnancy. Given this high prevalence of exercise in pregnancy, more research is warranted on both the beneficial and adverse maternal and fetal effects of exercise in pregnancy.

Male and female occupation in relation to miscarriage and preterm delivery in central North Carolina.

PURPOSE: This study was undertaken to evaluate the role of parental occupation in miscarriage and preterm delivery. Previous studies raise the possibility that both male and female exposures could affect pregnancy. METHODS: Data from a population-based study of miscarriage and preterm delivery in central North Carolina were used to examine potential associations with male and female occupation. Medically treated miscarriage cases (n = 418), preterm delivery cases identified through hospital record review (n = 582), and term, normal birth weight controls (n = 787) were sought for telephone interview. The interview included information on jobs the woman held before and during the pregnancy, reports of her partner's job around the time of pregnancy, and information on potential confounding factors. RESULTS: Female employment overall, or in specific jobs, around the time of conception or early pregnancy was not associated with the risk of miscarriage, whereas working during pregnancy, especially in the seventh month, was inversely associated with risk of preterm delivery. Male employment in several industrial occupations was weakly associated with miscarriage (adjusted odds ratios (OR) of 1.6 to 1.8), and somewhat more strongly associated with preterm delivery, particularly for chemists and sheet metal workers (adjusted OR over 3). Restriction to married men strengthened the associations. CONCLUSIONS: Our results are limited by nonresponse, imprecision, incomplete identification of miscarriages, and lack of detailed occupational exposure information. Nonetheless, we found greater support for further examination of male compared to female jobs in relation to pregnancy outcome.

Population-based case-control study of occupational exposure to electromagnetic fields and breast cancer.

PURPOSE: This population-based case-control study examined occupational exposure to electromagnetic fields in relation to female breast cancer incidence among 843 breast cancer cases and 773 controls. METHODS: Exposure was classified based on work in the two longest-held jobs, and indices of cumulative exposure to magnetic fields based on a measurement survey. RESULTS: Female breast cancer was not associated with employment as an office or industrial worker. For the total study population, cumulative exposure over the entire career, and in the past 0-10 and 10-20 years generally showed odds ratios (ORs) close to the null. Moderately elevated risks were found for intermediate but not high levels of cumulative exposure accumulated 20 or more years ago (OR = 1.5; 95% CI = 1.1-2.0). Associations were stronger for premenopausal women (OR = 1.7; 95% CI = 1.1-2.7) in the past 10-20 years, and those with estrogen-receptor positive (ER+) breast tumors (OR = 2.06; 95% CI = 1.1-4.0). No consistent dose-response patterns were observed. CONCLUSIONS: These findings give little support to the hypothesis that electromagnetic fields cause cancer of the female breast.