Association between smoking and deaths due to colorectal malignant carcinoma: a national population-based case-control study in China.

BACKGROUND: This study explored the association between smoking and colorectal malignant carcinoma (CRC) in the Chinese population at the national level for the first time. METHODS: In the China Nationwide Retrospective Mortality Survey conducted during 1989-1991, 12,942 CRC cases among 1,136,336 all-cause deaths aged ≥30 years were randomly assigned 25,884 control interviews from 325,255 surviving spouses of all-cause deaths across 103 urban and rural areas. RESULTS: Compared with non-smokers, smoking significantly increased the risk of CRC-specific mortality by 9.8% (odds ratio (OR)=1.098, 95% confidence interval (CI)=1.046-1.153) adjusted for sex, age, and residence. There were significant dose-response relationships between smoking and CRC, such as smoking years, cigarettes smoked daily, and age at onset of smoking. Long-term heavy smokers aged ≥50 years with ≥30 smoking years and ≥20 cigarettes daily had an excess risk of CRC deaths of 30.2% (OR=1.302, 95% CI=1.214-1.397). The strongest association between these smoking variables, such as long-term heavy smokers (OR=1.604, 95% CI=1.341-1.919), and CRC was observed among rural men. CONCLUSIONS: Quitting smoking at any time would likely be beneficial to CRC prevention. Long-term heavy smokers and rural men should be viewed as special targets for smoking prevention and cessation programs.

Smoking and risk of death due to pulmonary tuberculosis: a case-control comparison in 103 population centers in China.

OBJECTIVE: To examine the relationship between smoking and the risk of tuberculosis (TB) mortality in a large population-based case-control study in China using an alternative control group selection design. METHODS: During 1989-1991, a nationwide mortality survey was conducted of deaths among adults from 1986 to 1988. Surviving spouses or other informants provided detailed information about their own as well as the deceased person's smoking history. For the present study, all persons who died of TB at age > or =40 were used as cases, whereas all surviving spouses of deceased persons who died from causes other than those attributed to smoking were used as controls. RESULTS: It was estimated that for 22.5% of men and 6.6% of women, smoking was a contributing factor for TB deaths. Although variations in TB death rates by smoking status were not obvious before the age of 60, these differences increased substantially with age thereafter. This trend occurred in both urban and rural areas, although rural TB death rates were double those observed in urban areas. CONCLUSIONS: Tobacco smoking was associated with a large number of deaths from TB in China. The current study confirms results from previous studies about the relationship between smoking and TB mortality.

Birth weight, maternal weight and childhood leukaemia.

There is mounting evidence that childhood leukaemia is associated with high birth weight, but few studies have examined the relationship between leukaemia and other perinatal factors that influence birth weight, such as maternal weight or gestational weight gain. This case-cohort study included 916 acute lymphocytic leukaemia (ALL) and 154 acute myeloid leukaemia (AML) cases diagnosed prior to age 10 years between 1985 and 2001 and born in New York State excluding New York City between 1978 and 2001. Controls (n=9686) were selected from the birth cohorts for the same years. Moderate increased risk of both ALL and AML was associated with birth weight 3500 g or more. For ALL, however, there was evidence of effect modification with birth weight and maternal prepregnancy weight. High birth weight was associated with ALL only when the mother was not overweight while heavier maternal weight was associated with ALL only when the infant was not high birth weight. Increased pregnancy-related weight gain was associated with ALL. For AML, birth weight under 3000 g and higher prepregnancy weight were both associated with increased risk. These findings suggest childhood leukaemia may be related to factors influencing abnormal fetal growth patterns.

Sex of first child and breast cancer survival in young women.

Two studies have reported that young women with breast cancer face increased risk of early mortality if their first child was male rather than female. An immunological mechanism has been suggested. We sought to confirm these results in a larger, historical cohort study of 223 parous women who were aged <45 years at breast cancer diagnosis during 1983-1987. Subjects were identified through the Maine Cancer Registry. Follow-up data were obtained from hospitals, physicians, and death certificates. Reproductive history data were obtained from the next of kin of the deceased women, birth certificates, physicians, hospitals, and lastly, subjects. With a 7-year follow-up, multivariate modeling found a lower mortality risk in women with a male first child (hazard ratio [HR] 0.51, 95% confidence interval [CI] 0.32-0.81, log-rank comparison). The survival advantage remained for at least 13 years in women with a male firstborn. Thus, previous studies were not confirmed. Mortality risk in young women with breast cancer is not increased by having borne a male first child rather than a female first child.

Perinatal factors and risk of neuroblastoma.

Neuroblastoma is a malignancy commonly diagnosed during infancy or early childhood, raising speculation about the role of perinatal factors and risk of disease. Using a case-control design, cases included 155 infants and children aged 0-5 years with histologically confirmed neuroblastoma diagnosed and reported to the New York State Cancer Registry between 1976 and 1987. Controls were randomly selected from the State's Livebirth Registry and were frequency matched to cases on year of birth (n = 310). Medical records of cases were used to verify histology and stage of disease. Data on perinatal factors were ascertained from birth certificates and standardised telephone interviews with mothers. Unconditional logistic regression was used to estimate (un)adjusted odds ratios (OR) and 95% confidence intervals (CI). Both preterm (< 37 weeks) and post-term (> 42 weeks) birth were associated with a reduction in risk (OR = 0.4 [CI = 0.1, 0.9] and OR = 0.3 [CI = 0.1, 0.7] respectively) after controlling for confounders in unconditional logistic regression analysis. Elevated risk factors included: smoking during pregnancy (OR = 1.6; CI = 0.9, 2.8), contracted pelvis (OR = 2.3; CI = 0.6, 9.8), birth injury (OR = 2.9; CI = 0.3, 24.9) and 1-min Apgar Score < or =3 (OR = 6.0; CI = 0.9, 38.6); all confidence intervals included one. These data suggest that extremes in gestation may be associated with a reduced risk, although aetiological mechanisms remain unknown.

Parental occupational exposures and risk of neuroblastoma: a case-control study (United States).

BACKGROUND: A case-control study was conducted with 183 histologically confirmed neuroblastoma cases aged 0-14 years diagnosed among residents of New York State, excluding New York City, between 1976 and 1987. Three hundred seventy-two controls were selected from the New York State live birth certificate registry and were matched to cases on year of birth. METHODS: Parental occupational exposures at the time of each child's birth were obtained from maternal telephone interviews, successfully completed for 85% of cases and 87% of controls. RESULTS: Odds ratios were significantly elevated for maternal occupation in the service (OR = 2.0, 95% CI = 1.0 4.1) and retail (OR = 2.0, 95% CI = 1.1-3.7) industries and paternal occupation in materials handling (OR = 3.8, 95% CI = 1.1-14.6). Odds ratios were also significantly elevated for maternal report of occupational exposure to acetone (OR = 3.1, 95% CI = 1.7-5.6), insecticides (OR = 2.3, 95% CI = 1.4-3.7), lead (OR = 4.7, 95% CI = 1.3-18.2) and petroleum (OR = 3.0, 95% CI = 1.5-6.1) and paternal exposure to creosote (OR = 2.1, 95% CI = 1.1-4.3), dioxin (OR = 6.9, 95% CI = 1.3-68.4), lead (OR = 2.4, 95% CI = 1.2-4.8), and petroleum (OR = 1.8, 95% CI = 1.1-2.8). CONCLUSIONS: Due to the uncertainty of the biologic plausibility of these associations and the possibility of alternative explanations, these results should be interpreted cautiously.

Multiple births and risk of epithelial ovarian cancer.

BACKGROUND AND METHODS: Prevailing hypotheses about the causes of ovarian carcinogenesis predict that women with a history of multiple births (twins, triplets, etc.) should be at increased risk of epithelial ovarian cancer. However, the scant available evidence suggests that they may actually be at lower risk. To resolve this issue, we pooled data from eight studies involving 2859 parous women with epithelial ovarian cancer (case patients) and 7434 parous women without ovarian cancer (control women). In addition to assessing their history of multiple births (and the sex of the children, where available), we obtained information on age, parity, oral contraceptive use, and other reproductive factors for each woman. Details of tumor histology were available for all case patients. We estimated the relative risks of various histologic types of ovarian cancers associated with multiple births by using multivariable logistic regression analysis, adjusting for matching and confounding variables. RESULTS: Among these parous women, 73 case patients (2. 6%) and 257 control women (3.5%) had a history of multiple births. The adjusted summary odds ratio (OR) for developing all types of epithelial ovarian cancer that are associated with multiple births was 0.81 (95% confidence interval [CI] = 0.61-1.08). We found no evidence that risks associated with multiple births differed among women with borderline or invasive tumors and among women with same-sex and opposite-sex offspring from multiple births. The risk reductions appeared specific for nonmucinous tumors (n = 2453; summary adjusted OR = 0.71 [95% CI = 0.52-0.98]); in contrast, associations with mucinous tumors (n = 406) were heterogeneous across studies. CONCLUSIONS: Parous women with nonmucinous ovarian cancer are no more likely to have a history of multiple births than other parous women, counter to the predictions of current hypotheses for causes of ovarian cancer.

Factors associated with repeat mammography in a New York State public health screening program.

Screening mammography is the most effective method for early detection of breast cancer, but repeat mammography rates are not optimal in most populations. Since 1988, New York State has supported a program of breast cancer screening for underserved, uninsured, or underinsured women. The present study was designed to identify sociodemographic and clinical factors associated with failure to return for repeat mammography screening after a negative initial mammogram. Of women initially screened between 1988 and 1991 (N = 9,485), 27 percent obtained repeat mammograms by 1993. The final logistic regression model contained program site, race and ethnicity, family income, and time since last mammogram.

Current problems that are likely to affect the future of epidemiology.

The current discussion focuses on criticism as a positive force for improving epidemiologic practice through periodic reexamination of the basic approach to the discipline and the strategy for meeting the future educational needs of students and practicing epidemiologists. The types of epidemiologic research conducted and the settings within which the research will be conducted are also discussed. Epidemiology can be expected to play a major role in new areas of research that are created by changes in the medical care system and the development of large data systems associated with these approaches to health care delivery. This paper also discusses the growing threat to data access, the problems of communicating epidemiologic research findings to the public through the media, and the expanding interface between epidemiologic research and the legal system. The role of epidemiologic organizations in helping to shape the discipline's response to these issues and the opportunities these issues or problems present for improving epidemiologic research are also discussed.

Interlaboratory agreement among results of human papillomavirus type 16 enzyme-linked immunosorbent assays.

Serological assays for measuring antibodies to human papillomavirus type 16 (HPV-16) virus-like particles (VLPs) have become important epidemiologic tools in recent years. However, the interlaboratory replicability of these assays has not been assessed. In this investigation, three laboratories tested a panel of specimens obtained from two different groups: 265 subjects in a vulvar cancer case-control study and 107 healthy volunteer blood donors. Each laboratory used an enzyme-linked immunosorbent assay (ELISA), but no attempt was made to standardize assay procedures among the three laboratories. The data showed good day-to-day intralaboratory replicability in laboratory 1 (correlation coefficient, > or = 0.88) and good intra-assay variability in laboratory 3 (correlation coefficient, > or = 0.93). Interlaboratory correlations, likewise, ranged between 0.61 and 0.80 in both case-control study subjects and healthy blood donors, indicating that ELISA optical density (OD) values between laboratories were linearly related regardless of the population. Kappa coefficients (kappa), based on each laboratory's categorical interpretation of its results (as positive or negative), showed good agreement (kappa, > 0.6) in case-control study subjects and moderate agreement (kappa, > or = 0.4) in blood donors, a population that had few strongly positive sera. When OD values near seropositive cutoffs were treated as indeterminates, there was little discordance between laboratories in either population. The data suggest that each laboratory measured the same humoral immune response and that their HPV-16 VLP ELISAs performed similarly (Pearson correlations). Interlaboratory differences, however, probably due to reagents and procedures, were considerably greater than intralaboratory day-to-day variability. Interlaboratory agreement in determining seropositivity (kappa) could be improved by sharing positive and negative serum controls and by treating marginal results as indeterminate. As part of continuing cooperation to improve interlaboratory agreement, we are preparing bulk serum control specimens to be shared and made available to interested researchers.

Human papillomavirus-specific serologic response in vulvar neoplasia.

Epidemiological and virological evidence suggests that invasive squamous cell carcinoma (SCC) of the vulva is etiologically heterogeneous and that basaloid or warty SCC (BWSCC) and vulvar intraepithelial neoplasia (VIN) are linked to human papillomavirus (HPV) infections while keratinizing SCC (KSCC) is a non-HPV-associated malignancy. In the present study, HPV-specific antibodies in sera of patients with BWSCC, VIN, and KSCC and of controls were examined by ELISA for antibodies reactive to HPV-16 virus-like particles (VLP) and in radioimmunoprecipitation assays for antibodies to HPV-16 E6 and E7 proteins expressed by in vitro transcription and translation. The prevalences of antibodies to HPV-16 VLPs were significantly higher in HPV-associated VIN (59.1%) and BWSCC (50.0%) than in KSCC (22.2%) and controls (18.2%). Antibodies to E6 and E7 proteins were more prevalent in BWSCC than in any other groups. Prevalence of serum antibodies to any one of the antigen preparations was significantly higher in BWSCC (64.3%) and VIN (59.1%) than in KSCC (27.8%) and controls (22.2%). Also, sera with high antibody titers were found more frequently in BWSCC and VIN cases than in controls. These data provide immunological evidence in support of the observation that VIN and BWSCC, but not KSCC, are associated with HPV infections.

Gravid health status, medication use, and risk of neuroblastoma.

The epidemiology of neuroblastoma suggests that prenatal exposures may be important etiologic factors in this disease. The authors describe the role of maternal health status and prenatal medication usage and risk of neuroblastoma. This retrospective study was based on completed interviews with 183 histologically confirmed neuroblastoma cases aged 0-14 years diagnosed among residents of New York State (excluding New York City) between 1976 and 1987. Controls were matched to cases on year of birth and race and were drawn from the New York State live birth certificate registry. Interviews were satisfactorily completed with 85% of the cases and 87% of controls. Significantly elevated odds ratios were noted for vaginal infections during pregnancy (odds ratio (OR) = 2.2, 95% confidence interval (CI) 1.2-4.0), medical treatments for vaginal infection during pregnancy (OR = 2.4, 95% CI 1.2-4.9), and any reported use of sex hormones during pregnancy (OR = 3.0, 95% CI 1.3-6.9). Point estimates for any hormone use suggested elevated risk among male offspring (OR = 4.4, 95% CI 1.5-13.3). Among the individual exposures comprising any hormone use, only hormone use related to infertility was observed to be significant (OR = 10.4, 95% CI 1.2-89.9). A protective effect was noted for self-reported vitamin use (OR = 0.28, 95% CI 0.03-0.69). Although it is not possible to presume a specific role for prenatal hormone exposure as initiator or promoter, these findings lend support to an association between prenatal hormone exposure and risk of neuroblastoma.

A prospective study of early pregnancy loss.

The New York State Early Pregnancy Detection Study was a prospective study of early pregnancy loss, between implantation and menses, in 217 women attempting to become pregnant during 1989-1992. Women collected urine samples on three consecutive mornings during the late luteal phase of their menstrual cycle, for up to 12 cycles, contributing samples for 1253 menstrual cycles. Urinary human chorionic gonadotrophin (HCG), measured using an immunoradiometric assay, was the biomarker for pregnancy. We observed a range of early pregnancy loss (EPL) rates, from a low estimate of 11.0% to a high estimate of 26.9%, depending on the definition used and the subgroup analysed. Based on a definition of 3 days of HCG concentration > or = 4.00 pmol/l, 2 days > or = 5.33 pmol/l or the last day of HCG > or = 6.67 pmol/l, we identified 115 positive cycles; 95 cycles were clinically confirmed pregnancies and 20 cycles were EPL, giving an EPL rate of 17.4% [95% confidence interval (CI) 11.0-25.6]. In addition, we observed an EPL rate of 19.5% (95% CI 11.3-30.1) for samples collected within a 15 day window around menses, and a rate of 20.3% (95% CI 11.3-32.2) for samples limited to the first three menstrual cycles. Because studies use urine collection schemes other than daily sampling, the definition of pregnancy will be crucial in defining EPL.

Alcohol consumption and breast cancer: estrogen receptor status and histology.

Data from a case-control study conducted in New York State during 1982-1984 were used to evaluate the relation between alcohol consumption and estrogen receptor-positive and estrogen receptor-negative breast cancers and alcohol and various histologic subtypes. The cases were women between 20 and 79 years of age with a diagnosis of primary breast cancer. A total of 794 estrogen receptor-positive and 358 estrogen receptor-negative breast cancer cases were available for study. Controls (n = 1,617) were selected from driver's license files of the New York State Department of Motor Vehicles. Information on estrogen receptor status and histology was obtained from hospital records. The risk of estrogen receptor-positive breast cancer was shown to increase with increasing amounts of alcohol consumption in grams per day (odds ratio (OR) = 1.18 (95% confidence interval (CI) 0.88-1.57) for < 1.5 g/day, 1.28 (95% CI 0.91-1.80) for 1.5-4.9 g/day, 1.28 (95% CI 0.96-1.70) for 5.0-14.9 g/day, and 1.35 (95% CI 0.99-1.85) for > or = 15.0 g/day). There was no relation between alcohol consumption and estrogen receptor-negative tumors (OR = 0.92 (95% CI 0.62-1.36) for < 1.5 g/day, 1.19 (95% CI 0.77-1.83) for 1.5-4.9 g/day, 0.94 (95% CI 0.64-1.35) for 5.0-14.9 g/day, and 1.05 (95% CI 0.70-1.59) for > or = 15.0 g/day). The risk for each of the histologic subtypes studied increased with increasing daily alcohol consumption. These findings suggest that alcohol may only increase a woman's risk of estrogen receptor-positive breast cancers.

Cancer control planning and establishment of priorities for intervention by a state health department.

A number of data sources routinely available to State health departments were analyzed as part of a State health department cancer control planning effort. This planning effort consisted of seven steps; the most challenging one was the establishment of priorities for cancer control interventions. Using data from available sources, however, a framework for prioritizing potential cancer control interventions as well as choosing a geographic area in which to implement selected interventions was developed. Factors considered in this framework for setting intervention priorities included the magnitude of the problem; the existence of scientific consensus regarding the efficacy of intervention techniques; the availability of data needed to plan, implement, and evaluate an intervention; the availability of resources within communities to implement an intervention; and the existence of public demand for the intervention. The development and use of this cancer control planning model and framework for setting cancer control intervention priorities in New York State are described in this paper. In using this planning model and framework for setting priorities, quantitative elements were found to be most necessary to define problems, but qualitative elements were most crucial for decision making.

Relationship of leukemia risk to radiation dose following cancer of the uterine corpus.

BACKGROUND: Radiotherapy has been linked infrequently to secondary leukemia despite extensive exposure of the active bone marrow to ionizing radiation. Few studies include substantial numbers of elderly patients. PURPOSE: We evaluated women with cancer of the uterine corpus, the majority of whom were treated at older ages, to gain additional information on cancer risk following partial-body radiotherapy and to examine differences in risk between external-beam therapy and brachytherapy. METHODS: A cohort of 110,000 women with invasive cancer of the uterine corpus who survived at least 1 year following their initial cancer was assembled from nine population-based cancer registries. Cancer diagnoses occurred from 1935 through 1985, and most patients were diagnosed during the 1960s and 1970s. Radiation doses were computed to 17 sections of the active bone marrow for 218 women who developed leukemia and for 775 matched control subjects. RESULTS: Radiotherapy did not increase the risk of chronic lymphocytic leukemia (CLL) (relative risk [RR] = 0.90; 95% confidence interval [CI] = 0.4-1.9). However, for all leukemias except CLL, a significant risk was identified (RR = 1.92; 95% CI = 1.3-2.9). Overall, the pattern of risk in relation to dose was erratic and was most consistent with a constant increased risk across the entire dose range. The risk following continuous exposures from brachytherapy at comparatively low doses and low dose rates (RR = 1.80; 95% CI = 1.1-2.8; mean dose = 1.72 Gy) was similar to that after fractionated exposures at much higher doses and higher dose rates from external-beam treatment (RR = 2.29; 95% CI = 1.4-3.7; mean dose = 9.88 Gy), indicating a large difference in the estimated risk per unit dose. Risk did not vary by age at first exposure; increased risks were apparent for irradiated patients aged 65 years or older (RR = 1.77; 95% CI = 0.9-3.5). CONCLUSION: The leukemia risk associated with partial-body radiotherapy for uterine corpus cancer was small; about 14 excess leukemia cases were due to radiation per 10,000 women followed for 10 years. Women aged 65 years or older had a radiation risk comparable with that found in younger women. The relationship of leukemia risk to radiation dose was found to be complex due to the competing processes of cell killing, transformation, and repair. At very high doses delivered at high rates, destruction of cells likely dominates, and the risk per unit dose is low. In the low dose range, where dose was protracted and delivered at relatively low dose rates, the leukemia risk appears lower than that projected from risk estimates derived from the instantaneous whole-body exposures of atomic bomb survivors.