John E. Vena

Positional Accuracy of Geocoded Addresses in Epidemiologic Research

Background Geographic information systems (GIS) offer powerful techniques for epidemiologists. Geocoding is an important step in the use of GIS in epidemiologic research, and the validity of epidemiologic studies using this methodology depends, in part, on the positional accuracy of the geocoding process. Methods We conducted a study comparing the validity of positions geocoded with a commercially available program to positions determined by Global Positioning System (GPS) satellite receivers. Addresses (N = 200) were randomly selected from a recently completed case–control study in Western New York State. We geocoded addresses using ArcView 3.2 on the GDT Dynamap/2000 U.S. Street database. In addition, we measured the longitude and latitude of these addresses with a GPS receiver. The distance between the locations obtained by these two methods was calculated for all addresses. Results The distance between the geocoded point and the GPS point was within 100 m for the majority of subject addresses (79%), with only a small proportion (3%) having a distance greater than 800 m. The overall median distance between GPS points and geocoded points was 38 m (90% confidence interval [CI] = 34–46). Distances were not different for cases and controls. Urban addresses (median = 32 m; CI = 28–37) were slightly more accurate than nonurban addresses (median = 52 m; CI = 44–61). Conclusions. This study indicates that the suitability of geocoding for epidemiologic research depends on the level of spatial resolution required to assess exposure. Although sources of error in positional accuracy for geocoded addresses exist, geocoding of addresses is, for the most part, very accurate.

Association of a Dietary Inflammatory Index With Inflammatory Indices and Metabolic Syndrome Among Police Officers

Objectives:To determine whether the dietary inflammatory index (DII) is associated with inflammatory or metabolic biomarkers and metabolic syndrome (MetSyn) among police officers. Methods:Cross-sectional data from the Buffalo Cardio-Metabolic Occupational Police Stress study were derived from saliva and fasting blood samples, anthropometric measurements, long-term shiftwork histories, and demographic, stress/depression, and food frequency questionnaires (FFQs). Metabolic syndrome was defined using standard criteria. Results:Officers in DII quartiles 2 to 4 were more likely to exceed a threshold of 3.0 mg/L for C-reactive protein (odds ratio [OR] = 1.88; 95% confidence interval [95% CI] = 1.02 to 3.45; OR = 2.17; 95% CI = 1.19 to 3.95; OR = 1.57; 95% CI = 0.85 to 2.88, respectively) compared with quartile 1. The glucose intolerance component of MetSyn was more prevalent among officers in DII quartile 4 than among those in quartile 1 (OR = 2.03; 95% CI = 1.08 to 3.82). Conclusions:A pro-inflammatory diet was associated with elevated CRP and with the glucose intolerance component of MetSyn.

Menstrual cycle characteristics and the risk of endometriosis.

The present study utilized a case-control design to assess the relation between menstrual cycle factors and risk of endo-metriosis in women age 19–45 years. From a gynecologic specialty clinic in Western New York in 1987, we identified 104 laparoscopically confirmed cases of endometriosis. We used two control groups: (1) 100 friend controls (friends of cases), and (2) 98 medical controls (patients from the same medical practice with conditions other than endometriosis). Controls were frequency matched to cases by age. Because crude odds ratios (ORs), stratified by age (< 30 vs ≥ 30 years), revealed effect modification by age, we performed all analyses by age group. We used logistic regression to calculate ORs and 95% confidence intervals (CIs), controlling for pregnancy, religion, and body mass index. Among women under age 30, using friend controls, we observed elevated risks for menstrual flow ≥6 days per month (OR = 2.5, 95% CI = 1.1–5.9), heavy flow (OR = 2.5, 95% CI = 1.1–6.3), severe cramps (OR = 2.5, 95% CI = 1.2–6.0), increasing symptoms (OR = 6.6, 95% CI = 2.6–16.5), and tampon use ≥14 years (OR = 3.6, 95% CI = 1.04–13.5). Risks were also elevated for women age 30 and over, but not to the same degree as among younger women. Using medical controls, we observed elevated risks for menstrual flow ≥6 days (OR = 2.3, 95% CI = 1.1–5.1), severe cramps (OR = 2.3, 95% CI = 1.1–5.3), and increasing symptoms (OR = 8.3, 95% CI = 3.3–21.0) among women under age 30. These results support theories that associate endometriosis with increased exposure to menstruation.

Breast cancer risk, meat consumption and N‐acetyltransferase (NAT2) genetic polymorphisms

Although inconsistencies exist, some studies have shown that meat consumption is associated with breast cancer risk. Several heterocyclic amines (HAs), formed in the cooking of meats, are mammary carcinogens in laboratory models. HAs are activated by polymorphic N‐acetyltransferase (NAT2) and rapid NAT2 activity may increase risk associated with HAs. We investigated whether ingestion of meat, chicken and fish, as well as particular concentrated sources of HAs, was associated with breast cancer risk, and if NAT2 genotype modified risk. Caucasian women with incident breast cancer (n = 740) and community controls (n = 810) were interviewed and administered a food frequency questionnaire. A subset of these women (n = 793) provided a blood sample. Polymerase chain reaction and restriction fragment length polymorphism analyses were used to determine NAT2 genotype. Consumption of red meats, as well as an index of concentrated sources of HAs, was not associated with increased breast cancer risk, nor did risk vary by NAT2 genotype. In post‐menopausal women, higher fish consumption was inversely associated with risk (odds ratio = 0.7; 95% confidence interval, 0.4–1.0); among pre‐menopausal women, there was the suggestion of inverse associations between risk and pork and chicken intake. Our results suggest that consumption of meats and other concentrated sources of HAs is not associated with increased breast cancer risk. However, due to the strong biologic plausibility for a role of some HAs in mammary carcinogenesis, and the likely measurement error in evaluation of sources of HAs in this study, further studies of these possible relationships are warranted. Int. J. Cancer75:825–830, 1998. Published 1998 Wiley‐Liss, Inc.

Parental consumption of contaminated sport fish from Lake Ontario and predicted fecundability.

Wildlife studies suggest that consumption of contaminated fish from the Great Lakes may expose humans to polychlorinated biphenyls and persistent chlorinated pesticides. To assess whether time to pregnancy or fecundability is affected, we conducted a telephone survey in 1993 with female members of the New York State Angler Cohort Study who were considering pregnancy between 1991 and 1994 (N = 2,445). Among the 1,234 (50%) women who became pregnant, 895 (73%) had a known time to pregnancy. Upon enrollment into the cohort in 1991, both partners reported duration and frequency of Lake Ontario sport fish consumption. We estimated lifetime exposure to polychlorinated biphenyls from recent consumption and used a discrete-time analog of Cox proportional hazards analysis to estimate conditional fecundability ratios and 95% confidence intervals (CIs) for fish consumption among couples with complete exposure data who discontinued birth control to become pregnant (N = 575). Maternal consumption of fish for 3-6 years was associated with reduced fecundability (fecundability ratio = 0.75; 95% CI = 0.59-0.91), as was more than a monthly fish meal in 1991 (fecundability ratio = 0.73; 95% CI = 0.54-0.98). Our findings suggest that maternal but not paternal consumption of contaminated fish may reduce fecundability among couples attempting pregnancy.

Mortality of a police cohort: 1950–1990

This study presents findings from an updated retrospective cohort mortality study of male police officers from January 1, 1950 to December 31, 1990 (n = 2,593; 58,474 person-years; 98% follow-up). Significantly higher than expected mortality rates were found for all cause mortality (Standardized mortality ratio [SMR] = 110; 95% confidence interval [95% CI] = 1.04-1.17), all malignant neoplasms (SMR = 125; 95% CI = 1.10-1.41), cancer of the esophagus (SMR = 213; 95% CI = 1.01-3.91), cancer of the colon (SMR = 187; 95% CI = 1.29-2.59), cancer of the kidney (SMR = 2.08, 95% CI = 100-3.82), Hodgkins disease (SMR = 313; 95% CI = 1.01-7.29), cirrhosis of the liver (SMR = 150; 95% CI = 1.00-2.16), and suicide (SMR = 153; 95% CI = 1.00-2.24). All accidents were significantly lower (SMR = 53; 95% CI = 0.34-0.79). Mortality by years of police service showed higher than expected rates for (1) all malignant neoplasms in the 1- to 9-years-of-service group; (2) all causes, bladder cancer, leukemia, and arteriosclerotic heart disease in the 10 to 19-year group; and (3) colon cancer and cirrhosis of the liver in the over 30 years of service group. Hypotheses for findings are discussed.

Alcohol dehydrogenase 3 genotype modification of the association of alcohol consumption with breast cancer risk

Objectives: Because alcohol dehydrogenase 3 (ADH3) is rate-limiting in alcohol oxidation and is polymorphic, we examined ADH3 genotype in relation to alcohol intake and breast cancer risk.Methods: We conducted a case–control study among Caucasian women aged 40–85 with incident, pathologically confirmed breast cancer and controls, frequency-matched on age and county. Queries included alcohol intake in the past 20 years. Genomic DNA was genotyped for the exon VIII ADH polymorphism by PCR followed by restriction enzyme digestion. Computation of odds ratios (OR) and 95% confidence intervals (CI) was by unconditional logistic regression.Results: We found increased risk among pre- (OR 2.3, 95%, CI 1.2–4.3) but not postmenopausal women (OR 1.1, 95% CI 0.7–1.7) associated with ADH31-1 compared to ADH31-2 and ADH32-2 genotypes. Risk was increased for premenopausal women with the ADH31-1 genotype and alcohol intake above the median (OR 3.6, 95% CI 1.5–8.8) compared to lighter drinkers with the ADH32-2 or ADH31-2 genotypes. ORs were close to null for premenopausal women in other drinking and genotype groups and for postmenopausal women categorized by genotype and alcohol consumption.Conclusion: Among premenopausal women there may be a group more genetically susceptible to an alcohol consumption effect on breast cancer risk.

Body mass index, weight gain, and risk of endometrial cancer

Excess weight near the time of diagnosis is a well-established risk factor for endometrial cancer; less is known about the influence of weight at earlier periods of a womans life or weight gain in adulthood. In a case-control study in western New York State, interviews were conducted with 232 incident endometrial cancer cases, diagnosed between 1986 and 1991, and 631 community controls. Body mass index at 16 years of age and 20, 10, and 2 years before interview and changes in body mass index between these time periods were examined. While being relatively heavy at 16 years of age was associated with slightly increased risk [adjusted odds ratio (OR) = 1.28, 95% confidence interval (CI) = 0.84-1.96], large gains over the entire period from 16 years of age to 2 years ago (OR = 3.45, CI = 2.13-5.57) and high body mass index close to the time of diagnosis (OR = 3.21, CI = 2.01-5.15) were associated with greater risk. Differences in mean body mass index between cases and controls increased over time.

Cardiorespiratory Fitness as a Predictor of Fatal and Nonfatal Stroke in Asymptomatic Women and Men

Background and Purpose— Prospective data on the association between cardiorespiratory fitness (CRF) and stroke are largely limited to studies in men or do not separately examine risks for fatal and nonfatal stroke. This study examined the association between CRF and fatal and nonfatal stroke in a large cohort of asymptomatic women and men. Methods— A total of 46 405 men and 15 282 women without known myocardial infarction or stroke at baseline completed a maximal treadmill exercise test between 1970 and 2001. CRF was grouped as quartiles of the sex-specific distribution of maximal metabolic equivalents achieved. Mortality follow-up was through December 31, 2003, using the National Death Index. Nonfatal stroke, defined as physician-diagnosed stroke, was ascertained from surveys during 1982 to 2004. Cox regression models quantified the pattern and magnitude of association between CRF and stroke. Results— There were 692 strokes during 813 944 man-years of exposure and 171 strokes during 248 902 woman-years of exposure. Significant inverse associations between CRF and age-adjusted fatal, nonfatal, and total stroke rates were observed for women and men (Ptrend≤0.05 each). After adjusting for several cardiovascular disease risk factors, the inverse association between CRF and each stroke outcome remained significant (Ptrend<0.05 each) in men. In women, the multivariable-adjusted relationship between CRF and nonfatal and total stroke remained significant (Ptrend≤0.01 each), but not between CRF and fatal stroke (Ptrend=0.18). A CRF threshold of 7 to 8 maximal metabolic equivalents was associated with a substantially reduced rate of total stroke in both men and women. Conclusions— These findings suggest that CRF is an independent determinant of stroke incidence in initially asymptomatic and cardiovascular disease-free adults, and the strength and pattern of the association is similar for men and women.

Exercise, occupational activity, and risk of endometrial cancer

We conducted a case-control study in western New York state among 232 women with newly-diagnosed endometrial cancer and 631 controls selected from the community. Physical activity was measured by participation in vigorous exercise and walking at four time periods: at age 16, and at 20, 10, and 2 years before the interview and by occupational activity based on a detailed lifetime history. Women who did a moderate amount of vigorous exercise at age 16 and at 20 years before the interview were at reduced risk as compared with those who reported no activity, with odds ratios (OR) (95% confidence intervals) of 0.51 (0.31-0.83) and 0.50 (0.29-0.89), respectively. However, there was no evidence of declining risk with greater amount of activity. At later times, 10 years and 2 years before the interview, being in the highest group with regard to vigorous activity was associated with a slightly but nor significant lower risk as compared with women who reported no activity; the adjusted OR were 0.72 (0.43-1.19) and 0.67 (0.42-1.09), respectively. Being in the highest category of miles walked at age 16 (i.e., > or = 15 miles per week) was associated with a slightly reduced risk as compared with not walking at all (OR 0.64 (0.26-1.16)), whereas the number of miles walked at other times was not related to reduced risk. Occupational physical activity was not related to the risk of endometrial cancer. Overall, these results indicate that physical activity at levels prevalent in this population has at most a modest relationship to reduced risk of endometrial cancer.