Tobacco Documents Online

A MATCHED CASE-CONTROL STUDY OF THE RELATIONSHIP BETWEEN BETA- CAROTENE INTAKE AND LUNG CANCER Tan, A.J., He, S.P., Huang, M.X., Li, G.G., and Shi, L.Y. Department of Epidemiology, Tongji Medical University, Wuhan, China In order to evaluate the relationship between dietary intake and serum level of beta-carotene and lung cancer risk, a 1:1 matched case-control study of 156 newly histologically diagnosed primary lung cancer patients and 156 general patients (without tumors, respiratory, or other related diseases) was conducted in Wuhan. Three ml of venous blood was collected, and all cases and controls were asked to participate in a nutritional assessment and to answer a food frequency questionnaire containing 64 items on 11 kinds of foods. Subjects were instructed to recall dietary habits during the year prior to their first recognition of any symptoms associated with their diseases. Nutrient intake from foods was calculated from the annual intake of various food items by reference to our food composition database. The findings showed that there was a statistically significant difference between dietary intake of beta-carotene in cases and controls (2877.13 ± 393.43 vs 3445 ± 430.98 ug/day), and serum level as well (25.69 ± 5.57 vs 32.26 ± 5.02 ug/dl). By means of several statistical methods (such as conditional logistic regression), we found an inverse association between dietary and serum beta-carotene and lung cancer risk. When controlling for the

potential confounding effects of cigarette smoking and alcohol drinking, a significant linear trend of lower dietary carotene intake and higher lung cancer risk was observed.

Table 1. OR linear tread test for lung cancer by histologic subtypes (controlling for cigarette smoking) according to levels of dietary carotene intake smoke dietary index1 intake2 Squamous cell Adenocarc i noma Other types Control No. OR No. OR No. OR No. <104 >3282 13 1.00 11 1.00 3 1.00 62 2960-3282 15 2.04 9 1.45 3 1.86 35 <2960 20 4.54 8 2.15 2 2.06 21 (P trend=0.00) (P trend=0.14) (P trend=0.30) 104 >3282 4 1.00 8 1.00 8 1.00 20 2960-3282 9 5.00 6 1.67 9 2.50 9 <2960 (P 16 12.33 4 trend=0.00) (P 1.67 trend=0.43) 8 (P 3.33 trend=0.04) 6 77 46 33 156 Adjusted OR(95XC1) 1.00 1.00 1.00 2.61(1.81, 3.78) 1.52(0.92,2.51) 2.21(1.34,3.63) 6.14(3.05,12.35) 1.98(0.88,4.47) 2.82(1.47,5.39) (p trend=0.00) (P trend=0.10) (P trend=0.04) 1. smoking frequency *smoking years 2. unit, µg/day

Table 2. OR linear tread test for lung cancer by histologic subtypes (controlling for alcohol smoking) according to levels of dietary carotene intake --------------------------------------------------------------------------- Alcohol Dietary Squamous cell Adenocarcinoma Other types Control index1 intake2 No. OR No. OR No. OR No. <25 >3282 9 1.00 9 1.00 5 1.00 46 2960-3282 12 1.92 3 0.48 5 1.44 32 <2960 16 2.92 3 0.55 9 2.96 28 (P trend=0.02) (P trend=0.31) (P trend=0.06) >25 >3282 8 1.00 18 1.00 5 1.00 30 2960-3282 14 3.75 11 1.31 2 0.86 14 <2960 18 11.25 2 0.56 7 7.00 6 (P trend=0.00) (P trend=0.81) (P trend=0.01) 77 46 33 156 Adjusted OR(95XCI) 1.00 1.00 1.00 2.59(1.71, 3.93) 0.91(0.76,1.87) 1.19(1.06,1.32) 4.86(2.43, 9.72) 0.55(0.17,1.78) 4.10(1.69,9.92) (P trend=0.00) (P trend=0.10) (P trend=0.04) 1. drinking frequency *drinking years 2. unit, µg/day