Smoking and colorectal cancer: different effects by type of cigarettes?

Although smoking is suggested to be a risk factor for colorectal cancer, the evidence to date is conflicting and may be confounded. Moreover, the effect of tobacco smoke may vary by time since initiation, type of tobacco product, anatomic subsites, and among ethnic groups. Data were derived from two consecutive population-based case-control studies conducted among Caucasians, Japanese, Native Hawaiians, Filipinos, and Chinese in Hawaii, including 1,959 ethnicity-, sex-, and age-matched case-control pairs. A lifetime history of smoking for different tobacco products and information on other risk factors were obtained by in-person interviews. Odds ratios (OR) and corresponding 95% confidence intervals (95% CI) were estimated using conditional logistic regression models with adjustment for potential confounders. Subjects who ever smoked were at an increased risk of colorectal cancer compared with never smokers (OR, 1.23; 95% CI, 0.99-1.52 for men and OR, 1.27; 95% CI, 1.01-1.59 for women). Increasing quartiles of pack-years over all tobacco products showed a clear dose-dependent association in men [for the highest quartile, Q4 (>40 pack-years) versus never smokers: OR, 1.48; 95% CI, 1.12-1.96; P(trend) = 0.002]. The dose-response trend was also present in women [for the highest quartile, Q4 (>30 pack-years) versus never smokers: OR, 1.38; 95% CI, 0.91-1.95; P(trend) = 0.04] and each ethnic group. There was a suggestion of a difference in risk with type of tobacco product. Non-filtered cigarettes increased risk of both colon and rectal cancer [for Q4 versus never smokers: OR, 1.59; 95% CI, 1.15-2.21; P(trend) = 0.001 and OR, 1.84; 95% CI, 1.18-2.86; P(trend) = 0.02, respectively], whereas filtered cigarettes seemed to increase risk of rectal but not colon cancer (OR, 1.37; 95% CI, 0.88-2.13; P(trend) = 0.06 and OR, 1.05; 95% CI, 0.79-1.39; P(trend) = 0.98, respectively). The effect of smoking was not limited to the distant past, and accumulated pack-years of smoking seemed to be more important than the time in which smoking occurred. The data from this large study corroborate previous reports of a positive association between smoking and colorectal cancer and suggest that the association may vary by type of cigarette.

Simple measures of dietary variety are associated with improved dietary quality.

The objective of this study was to identify a measure of dietary variety that was associated with improved dietary quality and easily understood by consumers. Dietary quality was measured by nutrient adequacy and intakes of added sugars, saturated fat, cholesterol, and sodium. We developed four definitions of dietary variety: (a) a count of basic commodities consumed; (b) a count of food codes reported; (c) a count of five Food Guide Pyramid (FGP) food groups consumed; and (d) a count of 22 FGP subgroups consumed. The analysis sample included 4,964 men and 4,797 women aged 19 years and older who participated in the Continuing Survey of Food Intakes by Individuals 1994-96. For each day of dietary data, we examined associations of each type of dietary variety with several measures of dietary quality using Spearman's correlations and multivariate linear regression models. After adjusting for energy intake and the number of FGP food group servings, all types of dietary variety were positively associated with mean nutrient adequacy across 15 nutrients, but associations were strongest for commodity-based variety and for 22 FGP subgroup consumption variety. Likewise, all variety measures were inversely associated with intakes of added sugars and saturated fat, with commodity-based variety and 22 FGP subgroup variety the strongest. We conclude that variety measured using 22 FGP subgroups is preferable because it is a good predictor of dietary quality, is relatively simple to calculate, and is easy to explain to consumers.