Anna H. Wu-Williams

The relationship between dietary fat intake and risk of colorectal cancer: evidence from the combined analysis of 13 case-control studies

The objective of this study was to examine the effects of the intakeof dietary fat upon colorectal cancer risk in a combined analysis of datafrom 13 case-control studies previously conducted in populations withdiffering colorectal cancer rates and dietary practices. Original datarecords for 5,287 cases of colorectal cancer and 10,470 controls werecombined. Logistic regression analysis was used to estimate odds ratios (OR)for intakes of total energy, total fat and its components, and cholesterol.Positive associations with energy intake were observed for 11 of the 13studies. However, there was little, if any, evidence of anyenergy-independent effect of either total fat with ORs of 1.00, 0.95, 1.01,1.02, and 0.92 for quintiles of residuals of total fat intake (P trend =0.67) or for saturated fat with ORs of 1.00, 1.08, 1.06, 1.21, and 1.06 (Ptrend = 0.39). The analysis suggests that, among these case-control studies,there is no energy-independent association between dietary fat intake andrisk of colorectal cancer. It also suggests that simple substitution of fatby other sources of calories is unlikely to reduce meaningfully the risk ofcolorectal cancer.

Exposure of nonsmoking women to environmental tobacco smoke: a 10-country collaborative study.

The interpretation and interpretability of epidemiologic studies of environmental tobacco smoke (ETS) depend largely on the validity of self-reported exposure. To investigate to what extent questionnaires can indicate exposure levels to ETS, an international study was conducted in 13 centers located in 10 countries, and 1,369 nonsmoking women were interviewed. The present paper describes the results of the analysis of self-reported recent exposure to ETS from any source in relation to urinary concentrations of cotinine. Of the total, 19.7 percent of the subjects had nondetectable cotinine levels, the median value was 6 ng/mg, and the cut-point of the highest decile was 24 ng/mg. The proportion of subjects misreporting their active smoking habit was estimated at between 1.9 and 3.4 percent, depending on whether cut-points of 50 or 100 ng/mg creatinine were used. Large and statistically significant differences were observed between centers, with the lowest values in Honolulu, Shanghai, and Chandigarh, and the highest in Trieste, Los Angeles, and Athens. Mean cotinine/creatinine levels showed a clear linear increase from the group of women not exposed either at home or at work, to the group of those exposed both at home and at work. Values were significantly higher for women exposed to ETS from the husband but not at work, than for those exposed at work but not from the husband. The results of linear regression analysis indicated that duration of exposure and number of cigarettes to which the subject reported being exposed were strongly related to urinary cotinine. ETS exposure from the husband was best measured by the number of cigarettes, while exposure at work was more strongly related to duration of exposure. After adjustment of number of cigarettes for volume of indoor places, a similar increase in cotinine (5 ng/mg) was predicted by the exposure to 7.2 cigarettes/8 h/40 m3 from the husband and 17.9 cigarettes/8 h/40 m3 at work. The results indicate that, when appropriately questioned, nonsmoking women can provide a reasonably accurate description of ETS exposure. Assessment of individual exposure to ETS should focus on daily duration and volume of indoor places where exposure occurred.