Multi-factor dimensionality reduction applied to a large prospective investigation on gene-gene and gene-environment interactions.

It is becoming increasingly evident that single-locus effects cannot explain complex multifactorial human diseases like cancer. We applied the multi-factor dimensionality reduction (MDR) method to a large cohort study on gene-environment and gene-gene interactions. The study (case-control nested in the EPIC cohort) was established to investigate molecular changes and genetic susceptibility in relation to air pollution and environmental tobacco smoke (ETS) in non-smokers. We have analyzed 757 controls and 409 cases with bladder cancer (n=124), lung cancer (n=116) and myeloid leukemia (n=169). Thirty-six gene variants (DNA repair and metabolic genes) and three environmental exposure variables (measures of air pollution and ETS at home and at work) were analyzed. Interactions were assessed by prediction error percentage and cross-validation consistency (CVC) frequency. For lung cancer, the best model was given by a significant gene-environment association between the base excision repair (BER) XRCC1-Arg399Gln polymorphism, the double-strand break repair (DSBR) BRCA2-Asn372His polymorphism and the exposure variable 'distance from heavy traffic road', an indirect and robust indicator of air pollution (mean prediction error of 26%, P<0.001, mean CVC of 6.60, P=0.02). For bladder cancer, we found a significant 4-loci association between the BER APE1-Asp148Glu polymorphism, the DSBR RAD52-3'-untranslated region (3'-UTR) polymorphism and the metabolic gene polymorphisms COMT-Val158Met and MTHFR-677C>T (mean prediction error of 22%, P<0.001, mean CVC consistency of 7.40, P<0.037). For leukemia, a 3-loci model including RAD52-2259C>T, MnSOD-Ala9Val and CYP1A1-Ile462Val had a minimum prediction error of 31% (P<0.001) and a maximum CVC of 4.40 (P=0.086). The MDR method seems promising, because it provides a limited number of statistically stable interactions; however, the biological interpretation remains to be understood.

Intake of specific flavonoid classes and coronary heart disease--a case-control study in Greece.

OBJECTIVE: Dietary intake of flavonoids has been reported to protect against coronary heart disease (CHD) risk, but associations of specific classes of flavonoids with CHD have not been adequately studied. DESIGN: Hospital-based case-control study relying on interviewer administered questionnaires. SETTING: Cardiology Department of the University of Athens Medical School in the Hippokrateion General Hospital (1990-1991). SUBJECTS: Cases were 329 patients with electrocardiographically confirmed first coronary infarct or a first positive coronary arteriogram, or both (participation fraction 93%). Controls were 570 patients admitted to the same hospital for minor conditions unrelated to nutrition (participation fraction 95%). All cases and controls were interviewed in the hospital wards by experienced interviewers, and a 110-item food frequency questionnaire was administered. RESULTS: There was statistically significant evidence (P approximately 0.03) for an inverse association between intake of flavan-3-ols and CHD risk, an increase of about 21 mg per day corresponding to a 24% decrease in CHD risk. The inverse association between flavan-3-ols and CHD risk was largely accounted for by the intake of wine and to a lesser extent tea. For none of the other flavonoid classes was there statistically significant evidence of an association. CONCLUSION: Flavan-3-ols, which are largely found in wine and tea, are inversely associated with, and may be protective against, coronary heart disease.

Daily variations in air pollution and respiratory health in a multicentre study: the PEACE project. Pollution Effects on Asthmatic Children in Europe.

The Pollution Effects on Asthmatic Children in Europe (PEACE) study is a multicentre study of the acute effects of particles with a 50% cut-off aerodynamic diameter of 10 microm (PM10), black smoke (BS), sulphur dioxide (SO2) and nitrogen dioxide (NO2) on the respiratory health of children with chronic respiratory symptoms. The study was conducted in the winter of 1993/1994 by 14 research centres in Europe. A total of 2,010 children, divided over 28 panels in urban and suburban locations, was followed for at least 2 months. Exposure to air pollution was monitored on a daily basis. Health status was monitored by daily peak expiratory flow (PEF) measurements and a symptom diary. The association between respiratory health and air pollution levels was calculated with time series analysis. Combined effect estimates of air pollution on PEF or the daily prevalence of respiratory symptoms and bronchodilator use were calculated from the panel-specific effect estimates. Fixed effect models were used and, in cases of heterogeneity, random effect models. No clear associations between PM10, BS, SO2 or NO2 and morning PEF, evening PEF, prevalence of respiratory symptoms or bronchodilator use could be detected. Only previous day PM10 was negatively associated with evening PEF, but only in locations where BS was high compared to PM10 concentrations. There were no consistent differences in effect estimates between subgroups based on urban versus suburban, geographical location or mean levels of PM10, BS, SO2 and NO2. The lack of association could not be attributed to a lack of statistical power, low levels of exposure or incorrect trend specifications. In conclusion, the PEACE project did not show effects of particles with a 50% cut-off aerodynamic diameter of 10 microm, black smoke, sulphur dioxide or nitrogen dioxide on morning or evening peak expiratory flow or the daily prevalence of respiratory symptoms and bronchodilator use.

Epithelial, possibly precancerous, lesions of the lung in relation to smoking, passive smoking, and socio-demographic variables.

We have undertaken an autopsy-based study to evaluate the etiologic importance of active and passive smoking, as well as socio-demographic variables, in the development of pathologic precursors of lung cancer. Lung specimens were taken at autopsy from 531 persons who had died within four hours from a cause other than respiratory or cancer in Athens (Greece) or the surrounding area. Specimens were examined blindly for basal cell hyperplasia, squamous cell metaplasia, cell atypia and mucous cell metaplasia, i.e., pathological entities considered as epithelial, possibly precancerous, lesions (EPPL). Interviews were conducted with next of kin of the deceased. Suitable autopsy specimens as well as completed interviews were eventually available for 275 subjects. EPPL score was regressed on the available independent variables. EPPL score was higher among active smokers than among nonsmokers, while ex-smokers occupied an intermediate position. Conditional on smoking habits, EPPL score was higher among women than among men and higher among manual than among non-manual workers, in agreement with the corresponding patterns with respect to lung cancer. Nonsmoking women married to ever smokers had significantly higher EPPL score than those married to never smokers. The overall findings of this study suggest that EPPL is a valuable indicator of lung cancer risk and that exposure to environmental tobacco smoke is associated with higher EPPL levels and therefore with higher lung cancer risk.

A case-control study of endometrial cancer in relation to reproductive, somatometric, and life-style variables.

A hospital-based case-control study of cancer of the endometrium was conducted in Athens, Greece, from 1992 to 1994. The cases were 145 women residents of Greater Athens with histologically confirmed incident cancer of the endometrium, operated in the two cancer hospitals of the Greater Athens area or the major University Department of Obstetrics and Gynecology. Controls were 298 women residents of Greater Athens hospitalized for bone fractures or other orthopedic conditions in the accident hospital of Greater Athens. The data were analyzed by modeling through multiple logistic regression. The risk of endometrial cancer decreased with the number of livebirths (p for trend < 0.01), with early age at menopause (p = 0.03), and with later age at menarche (p = 0.11), whereas miscarriages and induced abortions were clearly unrelated. There were nonsignificant relations of disease risk with smoking (inverse), alcohol (inverse), and menopausal estrogens (positive), whereas oral contraceptive use was too uncommon to allow meaningful study. The lower risk of the disease associated with current occupations requiring manual activity (p = 0.03) and the lower, although not significantly so (p = 0.36), energy intake of cases in comparison to controls suggest that physical inactivity could be an important risk factor for endometrial cancer. Women with endometrial cancer were significantly taller than control women (p = 0.02). The latter results indicate that excess energy intake in early life, leading to higher attained stature, and excessive energy intake in later life, on account of physical inactivity and leading to higher body weight, converge in increasing the risk for endometrial cancer.

Dietary factors and the risk of endometrial cancer: a case--control study in Greece.

In a hospital-based case-control study of endometrial cancer undertaken in Athens (1992-94), 145 women residents of Greater Athens with confirmed cancer of the endometrium were compared with 298 control patients with orthopaedic diseases. Personal interviews were conducted in the hospital setting, and diet was assessed using a validated semiquantitative food frequency questionnaire. Nutrient intakes for individuals were calculated by multiplying the nutrient intake of a typical portion size for each specified food item by the frequency at which the food was consumed per month and summing these estimates for all food items. Data were modelled through logistic regression, controlling for demographic, reproductive and somatometric risk factors for endometrial cancer as well as for total energy intake. No macronutrient was significantly associated with endometrial cancer risk, but increasing intake of monounsaturated fat, mostly olive oil, by about one standard deviation was associated with a 26% risk reduction (odds ratio = 0.74; 95% confidence interval 0.54-1.3). Among micronutrients, only calcium intake was significantly inversely associated with endometrial cancer risk, whereas there was evidence against retinol and zinc imparting protection against the disease. With respect to food groups, there was weak and non-significant evidence that vegetables are protective, whereas consumption of pulses was positively associated with disease possibly because they contribute substantially in Greece to energy intake in excess of physical activity-dependent requirements.

Diet and coronary heart disease: a case-control study in Athens, Greece.

We conducted a case-control study in Athens, Greece, between January 1990 and April 1991 to examine the association between diet and coronary heart disease. The case series comprised 329 patients with electrocardiographically confirmed first coronary infarct or a first positive coronary arteriogram, or both, who were admitted to a major teaching hospital during a 16-month period. Controls were 570 patients admitted to the same hospital for minor conditions believed to be unrelated to nutrition. Total energy intake was inversely associated with coronary heart disease risk, a quintile energy increase corresponding to a relative risk of 0.96. After controlling for total energy intake, dietary fat was positively related to coronary heart disease, and total carbohydrates were negatively related to coronary heart disease, the nutrient-specific relative risks for a quintile increase being 1.19 (95% confidence interval = 0.96-1.48) and 0.81 (95% confidence interval = 0.67-0.97), respectively. Major fat components (saturated, monounsaturated, and polyunsaturated fat) did not appear to have differential risk implications for coronary heart disease; however, cooking with margarine was associated with an increased relative risk (1.87; 95% confidence interval = 0.82-4.28). Dietary proteins, cholesterol, and vitamin C were not associated with coronary heart disease.

Induced abortions, miscarriages, and tobacco smoking as risk factors for secondary infertility.

STUDY OBJECTIVE: The aim was to determine whether induced abortions could increase the risk of secondary infertility. DESIGN: This was a case-control study; cases were women with secondary infertility, individually matched to two controls who were currently pregnant. Each participant was interviewed by one of two medical doctors using a questionnaire that sought information on their demographic, socioeconomic, medical, and reproductive status. The data were analysed by conditional logistic regression. SETTING: The study took place in the Alexandra Maternity Hospital in Athens, Greece, in 1987-88. PARTICIPANTS: 84 women consecutively admitted with secondary infertility and 168 pregnant controls took part. MAIN RESULTS: Eight cases and no controls reported a previous ectopic pregnancy, confirming that the occurrence of a pregnancy of this type dramatically increases the risk of secondary infertility. Furthermore, the occurrence of either induced abortions or spontaneous abortions independently and significantly increased the risk of subsequent development of secondary infertility. The logistic regression adjusted relative risks (and 95% confidence intervals) for secondary infertility were 2.1 (1.1-4.0) when there was one previous induced abortion and 2.3 (1.0-5.3) when there were two previous induced abortions. Tobacco smoking significantly increased the risk of secondary infertility, the adjusted relative risk being 3.0 (1.3-6.8). CONCLUSIONS: Legalised induced abortions, as currently practised in Greece, appear to increase slightly the relative risk of secondary infertility.

A case-control study of coronary heart disease in Athens, Greece.

A case-control study of coronary heart disease (CHD) was conducted in Athens, Greece. The case series consisted of 329 patients with electrocardiographically confirmed coronary infarct or a diagnostic coronary arteriogram, or both, who were admitted during a 16-month period to a major teaching hospital. Controls were 570 patients admitted to the same hospital just before or after the CHD cases for minor surgery; eye, ear, nose or minor urological problems; or chest problems definitely shown to be unrelated to CHD. All cases and controls were interviewed in the hospital wards and selected laboratory data were abstracted. The main analysis was done by modelling through multiple logistic regression, controlling for demographic variables as well as for the mutual confounding effects of the investigated risk factors. Obesity, hypertension, diabetes mellitus, elevated blood cholesterol and excessive coffee intake were significant (P < 0.02) independent risk factors with relative risk estimates in the 2- to 3-fold range. Non-significant positive associations were found with respect to tobacco smoking and modest coffee consumption, whereas non-significant negative associations were noted with respect to alcohol intake and regular exercise. A negative association with duration of afternoon siesta was of borderline statistical significance.

Active and passive smoking and pathological indicators of lung cancer risk in an autopsy study.

OBJECTIVE: The association between involuntary smoking and lung cancer has been supported by most epidemiologic studies, but a number of authors and interest groups claim that the possibility of bias has not been excluded. Few autopsy-based studies have explored the role of active smoking and other exposures in lung carcinogenesis, and none has been previously done to examine the role of passive smoking. We have undertaken such an autopsy-based study in Athens, Greece. DESIGN: Lung specimens were taken at autopsy from 400 persons 35 years of age or older, of both genders, who had died within 4 hours from a cause other than respiratory or cancer in Athens or the surrounding area. For each person at least seven tissue blocks were taken from the main and lobar bronchi and at least five blocks from the parenchyma, including an average of about 20 smaller cartilaginous bronchi and membranous bronchioles. The specimens were examined without knowledge of the exposures of the particular subject in Turin, Italy. For 283 (71%) of the subjects the preservation of the bronchial epithelium was satisfactory for pathological examination, and for 206 among them (73%) an interview could be arranged with their next of kin, focusing on smoking habits of the deceased and their spouses, as well as other variables. The interviewers were not aware of the results of the pathological examinations. MAIN OUTCOME MEASURE: Specimens were examined for basal cell hyperplasia, squamous cell metaplasia, cell atypia, and (in membranous bronchioles and bronchiolo-alveolar airways) mucous cell metaplasia, ie, pathological entities that may be lung cancer risk indicators or epithelial, possibly precancerous, lesions (EPPL). The gland and wall thicknesses were also measured and their ratio calculated (Reid Index). RESULTS: In comparison with nonsmokers, EPPL values were significantly higher among current smokers and higher, but not significantly so, among former smokers. Furthermore, EPPL values were significantly higher among deceased nonsmoking women married to smokers rather than to nonsmokers. In this set of data neither occupation nor residence was associated with EPPL, but this could be due to the poor correlation of residential history with exposure to air pollution and the lack of adequate standardization of contemporary Greek occupations. The Reid Index was higher among smokers and former smokers in comparison with nonsmokers, among subjects with mainly urban residence in comparison with those with mainly rural residence, and among nonsmoking women married to smokers in comparison with those married to nonsmokers, but none of these differences was statistically significant. CONCLUSION: These results provide support to the body of evidence linking passive smoking to lung cancer, even though they are based on a study methodologically different from those that have previously examined this association.

A case-control study of air pollution and tobacco smoking in lung cancer among women in Athens.

BACKGROUND: A case-control study exploring the role of smoking and outdoor air pollution in the causation of lung cancer, by histologic type, in nonsmoking women, was undertaken in Athens between 1987 and 1989. METHODS: One hundred one women with lung cancer and 89 comparison women with fractures or other orthopedic conditions, all permanent residents of Greater Athens, were included in the study. Smoking habits were ascertained through interviews, whereas lifetime exposure to air pollution was assessed by linking blindly lifelong residential and employment addresses of all subjects with objectively estimated or presumed air pollution levels. RESULTS: The age-adjusted relative risk and 95% confidence intervals for lung cancer among current smokers compared with nonsmokers was 3.40 (1.75-6.61); it was 7.43 (2.88-19.13) among those smoking for more than 30 years and 7.46 (2.40-23.17) among those smoking more than 20 cigarettes per day. The age-adjusted relative risk was 1.70 (0.75-3.89) for adenocarcinoma and 6.45 (2.73-15.25) for other histologic types of lung cancer; statistically significant dose-response trends were evident for both histologic groups. Air pollution levels were associated with increased risk for lung cancer but the relative risk was small and statistically not significant. However, when both air pollution and duration (or quantity) of tobacco smoking, as well as their interaction, were introduced in a multiple logistic regression model, the interaction term was significant at the suggestive level of 0.10. CONCLUSION: Whereas there is no effect of air pollution among nonsmokers, the relative risk contrasting extreme quartiles of air pollution among smokers of 30 years duration was 2.23. The interaction was almost exclusively accounted for by the nonadenocarcinoma lung tumors.

Induced abortions, contraceptive practices, and tobacco smoking as risk factors for ectopic pregnancy in Athens, Greece.

A case-control study of the role of induced abortion and other factors on the subsequent occurrence of ectopic pregnancy was undertaken in 1986-1987 in Athens, Greece, where a similar study 20 years ago found a tenfold risk of ectopic pregnancy among women with one or more illegal induced abortions. Seventy women residents of Athens, consecutively admitted to the major state maternity hospital with a diagnosis of ectopic pregnancy, were individually matched with women with a newly diagnosed pregnancy of the same order as the ectopic index pregnancy. Two control women were found for each of 63 cases, but only one control for each of the remaining seven cases. All cases and controls were interviewed by the same qualified obstetrician. Statistical analysis was undertaken with stratification of individual matched triplets and pairs, as well as through conditional multiple regression procedures. The relative risk of recurrence of an ectopic pregnancy was 6.39 with 95% confidence interval (CI) 1.96-21.04. Miscarriages did not increase the risk of ectopic pregnancy. The relative risk for subsequent ectopic pregnancy among women with one or more induced abortion, compared to women without such abortions, was 1.87 (CI 0.84-4.16) controlling only for the matching factors, and 1.71 (CI 0.69-4.27) when marital status (a possible selection factor) was also accounted for in the conditional logistic regression. There was no evidence for increasing risk with increasing number of induced abortions. Past use of an intrauterine device (IUCD) was associated with a relative risk of 3.89 (0.72-21.02); the relative risk increased with the duration of use of the IUCD.(ABSTRACT TRUNCATED AT 250 WORDS)

Correlation of early pathological lesions in the bronchial tree with environmental exposures: study objectives and preliminary findings.

Lung specimens were taken at autopsy from 214 subjects aged 35 years and over who had died from nonpulmonary causes in Athens or the surrounding countryside. The samples were coded and examined for entities thought to be linked to environmental exposures, reflecting epithelial, possibly precancerous, lesions, as well as for morphological features, which were summarized using Reid's index. Of the 214 specimens, 142 were suitable for pathological examination. Next-of-kin of 101 of the dead people were identified and asked about the subject's exposure to active smoking, passive smoking, possible occupational hazards, dietary factors and proxy indicators of air pollution (residence). Preliminary analysis, controlling for age and sex, indicates that active smoking is related, although not statistically significantly, to both the Reid index (difference, 0.28, corresponding to a one-tailed p value of 0.07) and epithelial, possibly precancerous lesions (difference, 16.7, corresponding to a one-tailed p value of 0.09). Nonsignificant differences were found in the preliminary analysis of this ongoing study with respect to the other environmental factors examined.

Air pollution and cause specific mortality in Athens.

STUDY OBJECTIVE: The aim was to investigate the reported association between air pollution and cause specific mortality in the city of Athens. DESIGN: Cause specific mortality was contrasted between 199 d with high values of air pollution and 2*199 comparison days with low pollution, matched in a 1:2 ratio on the basis of various confounding factors. Statistical analysis was done, taking matching into account, using analysis of variance for randomised blocks. SETTING: The study was confined to the city of Athens, using data obtained between 1975 and 1982. PARTICIPANTS: Cause of death was assessed in all 25 138 persons dying in the 3*199 d studied. MEASUREMENTS AND MAIN RESULTS: Causes of death were evaluated blindly by two medically qualified investigators on the basis of information in the death certificates. Mortality was generally higher during the high pollution days but the difference was more pronounced and more significant for respiratory conditions, even though the number of deaths in this category was smaller than the corresponding numbers in the other two categories examined (cardiac and "other" deaths). CONCLUSION: The results show that the short term association between air pollution and overall mortality in Athens is likely to be causal, since it is particularly evident with respect to respiratory conditions, for which a biological air pollution link is more plausible.

Passive smoking and diet in the etiology of lung cancer among non-smokers.

A case-control study was undertaken in Athens to explore the role of passive smoking and diet in the causation of lung cancer, by histologic type, in non-smoking women. Among 160 women with lung cancer admitted to one of seven major hospitals in Greater Athens between 1987 and 1989, 154 were interviewed in person; of those interviewed, 91 were life-long non-smokers. Among 160 identified controls with fractures or other orthopedic conditions, 145 were interviewed in person; of those interviewed, 120 were life-long non-smokers. Marriage of a non-smoking woman to a smoker was associated with a relative risk for lung cancer of 2.1 (95% confidence interval [CI] 1.1-4.1); number of cigarettes smoked daily by the husband and years of exposure to husband's smoking were positively, but not significantly, related to lung cancer risk. There was no evidence of any association with exposure to smoking of other household members, and the association with exposure to passive smoking at work was small and not statistically significant. Dietary data collected through a semi-quantitative food-frequency questionnaire indicated that high consumption of fruits was inversely related to the risk of lung cancer (the relative risk between extreme quartiles was 0.27 (CI 0.10-0.74)). Neither vegetables nor any other food group had an additional protective effect; furthermore, the apparent protective effect of vegetables was not due to carotenoid vitamin A content and was only partly explained in terms of vitamin C. The associations of lung cancer risk with passive smoking and reduced fruit intake were independent and did not confound each other.(ABSTRACT TRUNCATED AT 250 WORDS)

The effect of involuntary smoking on the occurrence of chronic obstructive pulmonary disease.

One hundred and three ever-married women with newly diagnosed Chronic Obstructive Pulmonary Disease (COPD), who have never smoked, and 179 ever-married women who were visiting friends or relatives at the same hospital during the same period and have never smoked, were interviewed regarding the smoking habits of their husbands. There was statistically marginally significant difference between the COPD cases and the controls with respect to their husband's smoking habits. The association was irregular with respect to daily number of cigarettes smoked but there was a smooth dose response curve with respect to life long total number of cigarettes smoked, with women whose husband smoked more than 300 thousand cigarettes having a relative risk of 1.8 (90% confidence interval of 0.9-3.6) compared to women whose husband has never smoked. These findings, and converging related evidence, indicate that exposure to environmental tobacco smoke may be a risk factor for the development of COPD.

Time trends of tobacco smoking, air pollution, and lung cancer in Athens.

Athens is a city with a serious air pollution problem which has existed for more than 20 years. To evaluate whether air pollution has affected lung cancer incidence (and hence, mortality) in the population of Athens we have compared standardized lung mortality between Athens and the rest of Greece taking into account the tobacco consumption trends in the respective populations and varying the postulated latency between 0 and 20 years. There is no evidence for an independent or interactive (with tobacco smoking) effect of air pollution on lung cancer mortality; the tobacco-adjusted mortality appears, if anything, lower in Athens than in the rest of Greece and the slopes of lung cancer mortality on tobacco consumption are almost identical in Athens and in the rest of Greece. By contrast, the same data are compatible with a strong effect of tobacco smoking on lung cancer mortality, an effect which appears to involve not only the early carcinogenic stages but also some of the later ones. The results of the present analysis do not support the hypothesis that air pollution, at least in Athens until 1980, has increased the incidence of lung cancer to an extent large enough to be detectable in ecological correlation analyses. Nevertheless the inherent limitations of these methods indicate that their results should be interpreted with caution and only as a step toward the gradual understanding of a complex issue.

Short-term effects of air pollution on mortality in Athens.

Short-term effects of air pollution on mortality in Athens during the years 1975-1982 were studied. Daily values of sulphur dioxide (SO2) and smoke, measured by a five-station network of the National Observatory of Athens, were used as air pollution indicators. Mortality data were abstracted from the Town Registries of Athens and 18 other contiguous towns within the Greater Athens area. It was found that the adjusted daily mortality (estimated by subtracting from the observed value of mortality an 'expected' value, calculated after fitting a sinusoid curve to the empirical mortality data) depends positively and significantly on the level of SO2 (b = +0.0058, p = 0.05). This relation is independent of temperature, relative humidity, secular, seasonal, monthly and weekly variations of mortality as well as of synergistic effects of the above variables with season. No relation was found between smoke and adjusted daily mortality. An analysis for the determination of a possible threshold in the levels of SO2 causing health effects was also undertaken, by studying changes in the SO2 regression coefficients after successive deletion from the regression model of the days with the highest SO2 values. Our study shows that if there is an SO2 threshold it must lie slightly below the level of 150 micrograms/m3 (mean daily value).