Lung Cancer
Body mass index and lung cancer : a case-control study of subjects participating in a mass-screening programMaki Kanashiki Study objectives: An inverse relationship between body mass index (BMI) and the risk of lung cancer, suggesting that leanness is a risk factor for lung cancer, has been reported in previous studies. In order to evaluate the risk of lung cancer associated with lower levels of BMI in preclinical patients, we conducted a case-control study based on the results of community mass screening.
Design: The relationship between BMI (at the time of diagnosis, and at 1 to 5 years prior to diagnosis) and lung cancer was investigated in a case-control study of 363 lung cancer cases and 1,089 control subjects conducted between April 1993 and March 2003. Control subjects were selected from mass-screening subjects with no abnormalities on chest radiography and routine laboratory tests.
Results: In men, an inverse association between BMI and lung cancer was observed after adjustment for age and smoking (BMI < 20.8; range of the referent group, [greater than or equal to] 22.9 to < 25.0; odds ratio, 1.9; p = 0.0025; 95% confidence interval, 1.3 to 2.9). In women, however, no association was found between BMI and lung cancer (BMI < 20.8, p = 0.3868; and BMI [less than or equal to] 25.0, p = 0.4603, respectively). In addition, a negative association between BMI at 4 to 5 years prior to diagnosis and lung cancer was not observed in either gender (men, p = 0.2937 to 0.5783; women, p = 0.2042 to 0.9326).
Conclusions: Our present study indicated the possibility that the previously reported association between leanness and the risk of lung cancer in women was not correct, and this apparent association might be influenced by other factors such as smoking and smoking-related respiratory diseases. A larger-scale cohort study combined with mass-screening project will confirm our results. (CHEST 2005; 128:1490-1496)
Key words: body mass index; case control; lung cancer; mass screening
Abbreviations: BMI = body mass index; CI = confidence interval; OR = odds ratio
**********
An elevated risk of lung cancer associated with lower levels of body mass index (BMI) has been reported in case-control and cohort studies. (1-10) However, interpretation of the association between low BMI and lung cancer is complicated by the fact that weight loss may be a sign of the disease. Previous studies (1,2,4,9) were based on subjects with symptomatic lung cancer. Control subjects in a hospital-based study (2) had nonmalignant disease; therefore, the findings may reflect the presence of diseases other than lung cancer in control subjects. In addition, most previous cohort studies. (2,5,7,8) have evaluated BMI using self-reported body size. Some authors (10-12) have noted that self-reported body size was generally close to measured body size; however, small but systematic errors were usually present, with overestimation of height and underestimation of weight.
In order to evaluate the increased risk of lung cancer associated with lower levels of BMI in preclinical patients identified in a community-based mass screening, we conducted a case-control study. In this study, control subjects were selected from the screened population with no abnormalities on chest radiography and normal laboratory results. Information on body size was obtained from measurements obtained by nurses during the mass screening.
METHODS AND MATERIALS
Mass Screening for Lung Cancer
The Japanese government, under the Tuberculosis Control Law, has offered annual chest radiograph screening to all residents aged [greater than or equal to] 16 years since the 1950s. Lung cancer screening has been added to the program since 1987. Local governments in Ibaraki Prefecture and the Ibaraki Health Service Association formulated and managed these screening programs. A total of 2,993,323 people (1,489,724 men and 1,503,599 women) resided in Ibaraki prefecture, and 1,588,931 people were [greater than or equal to] 40 years old. The study area consisted of 71 cities, villages, and towns of 2,434,458 inhabitants (1,283,469 inhabitants aged at least 40 years) in Ibaraki Prefecture where the lung cancer screening was conducted. The participants in our screening numbered 244,704 (April 1997 to March i998), 246,176 (April 1998 to March 1999), 246,678 (April 1999 to March 2000), 250,441 (April 2000 to March 2001), 252,092 (April 2001 to March 2002), and 260,314 (April 2002 to March 2003). The computerized records of subjects who underwent screening were utilized.
An annual chest radiograph examination was conducted for all subjects using 100 x 100-mm miniature photofluorography. At the time of the examination, a self-completed questionnaire supplied information about medical history and general health. The subjects were asked whether they had ever smoked in their lifetime; those who had never smoked were classified as never-smokers. Smokers were classified as former smokers (not presently smoking) and current smokers. Sputum cytology was performed for individuals in the high-risk group, which was defined as individuals [greater than or equal to] 50 years old with a smoking index (number of cigarettes smoked per day multiplied by the number of years of smoking) of at least 600, and individuals [greater than or equal to] 40 years old with a history of bloody sputum. For individuals suspected of having lung cancer on the basis of chest radiography or sputum cytology, further examinations were performed to confirm the diagnosis.
Height and weight were measured in light, indoor clothing without shoes. BMI, calculated as weight in kilograms divided by the square of height in meters, was used as a measure of relative weight. The mean BMI values in Ibaraki Prefecture were 23.7 (SD 1.1) in men and 23.2 (SD 1.2) in women, and these were similar to mean BMI values in Japan: 23.4 (SD 3.1) in men and 22.8 (SD 3.5) in women.
Cases and Control Subjects
The cases analyzed in the present study were the participants in the mass-screening program who received a diagnosis of lung cancer during the study period. Data on the diagnosis of lung cancer were obtained from hospital reports written by chest physicians or thoracic surgeons and/or from the Ibaraki Cancer Registry. Histologic types were classified according to the World Health Organization histologic classification, (13) and TNM stage was classified using the TNM staging system. (14) Information on histologic types and stage was also obtained from hospital reports and/or the Cancer Registry.
Information on diagnosis of lung cancer from the abovementioned hospital reports and/or the Cancer Registry was collected until March 2003. Between April 1997 and March 2003, 363 cases of lung cancer were diagnosed. For each case subject, three control candidates were selected from the screening program lists. They were selected from subjects with no history of malignancy involving any organ and no abnormality on screening, including chest radiography and laboratory tests including blood cell counts and blood chemistry. These control candidates were matched with a case subject by gender and by year of birth. The resulting data set consisted of 363 lung cancer cases and 1,089 control subjects.
Statistical Analysis
This study was carried out on data obtained from April 1993 to March 2003 using BMI at the time of diagnosis and in 1 to 5 years prior to diagnosis. BMI was categorized into four levels on the basis of the distribution in the total study population (BMI < 20.8, leanest; BMI [greater than or equal to] 20.8 to < 22.9, second; BMI [greater than or equal to] 22.9 to < 25.0, third; and BMI [greater than or equal to] 25.0, highest). Relative risks were calculated by using the third category (BMI [greater than or equal to] 22.9 to < 25.0) as the referent group. Conditional logistic regression was used to examine the effect of BMI on lung cancer risk with adjustment for covariates among never-smokers, former smokers, current heavy smokers (more than a pack of cigarettes per day), and other current smokers. The Mann-Whitney U test was applied to elucidate the difference between two independent groups, and these results was compared using the [chi square] test. Results with a p value < 0.05 were regarded as significant.
RESULTS
Table 1 shows the characteristics of the cancer cases. Two thirds of all cases were men. Although the age range for this study was from 50 to 79 years, approximately 60% of the cases were > 70 years old. Smoking history identified 135 never-smokers, 78 former smokers, and 139 current smokers. Of the male case patients, 12 patients (5.4%) were nonsmokers; however, of the female case patients, 123 patients (94.6%) were never-smokers ([chi square] test, p = 0.0001). The smoking index among male case patients (mean, 863) was significantly higher than among male control subjects (mean, 531) [Mann-Whitney U test, p = 0.0001]. There was no difference in smoking index between female case patients and female control subjects (Mann-Whitney U test, p = 0.8125). Histologic subtypes included 215 adenocarcinomas, 94 squamous cell carcinomas, 25 small cell carcinomas, 9 large cell carcinomas, and 20 others. Adenocarcinoma was the most common histologic type in both genders. Of the 133 female case patients, 111 patients (83.5%) had adenocarcinomas. Among the 363 eases, 180 patients (49.6%) had stage 0-I B disease, and 271 patients (74.7%) had operable disease (stage 0-III A), which may differ from the study populations of previous reports. (1-9)
Table 2 shows adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for the association of BMI at time of diagnosis with lung cancer. In men, the leanest group exhibited a twofold-higher rate than the third BMI group after adjustment for age and smoking (BMI < 20.8; OR, 1.9; 95% CI, 1.3 to 2.9; p = 0.0025). The association was nonlinear with a marked increase in the leanest group, a smaller increase in the second, and equally low rates in the third and highest groups. For adenocarcinoma, which was the most common histologic type in both genders, the inverse association was not observed in men. However, women in the second BMI category (BMI [greater than or equal to] 20.8 to < 22.9) had significantly lower ORs for all lung cancer and for adenocarcinoma (OR, 0.5 vs 0.4; 95% CI, 0.3 to 0.9 vs 0.2 to 0.7; p = 0.0154 vs p = 0.0049, respectively). However, no statistical association was observed in the leanest category and highest category of BMI for lung cancers (BMI < 20.8, p = 0.3868; BMI [greater than or equal to] 25.0, p = 0.4603) and adenocarcinomas (BMI < 20.8, p = 0.6410; BMI [greater than or equal to] 25.0, p = 0.6674).
Table 3 shows the association of BMI with lung cancer after stratification by smoking status. In male current smokers, the leanest category of BMI had an elevated OR; however, it was not statistically significant (p = 0.1077). For male former smokers, the leanest category had a significantly elevated OR (2.3; 95% CI, 1.1 to 4.6; p = 0.0226). However, in male never-smokers, no inverse association was found. In female never-smokers, the second category had a significantly low OR (0.5; 95% CI, 0.3 to 0.9; p = 0.0159); however, no statistically significant association was observed in the leanest and highest category of BMI (BMI < 20.8, p = 0.3805; and BMI [greater than or equal to] 25.0, p = 0.5699).
Table 4 shows the association of BMI at 1 to 5 years before diagnosis of lung cancer in men after adjustment for age and smoking status. The leanest and the second BMI categories have elevated ORs at 1 to 2 years before diagnosis. The association was reduced at 3 years before diagnosis, and at 4 years before diagnosis the association was not statistically significant (p = 0.2937 to 0.5783). However, the negative association of BMI at 1 to 5 years before diagnosis was not seen in women (5 years before diagnosis, p = 0.2042 to 0.9326; Table 5). Furthermore, no inverse association was found between lung cancer and BMI based on 1 to 5 years prior to diagnosis in female never-smokers (Table 6).
As already stated, 215 of the 363 cases (59.2%) were adenocarcinomas. There were too-few female case patients to stratify the histologic subtypes, so we examined the association only in adenocarcinoma cases (Table 7). In male former smokers, the leanest category had an elevated OR (3.i; 95% CI, 1.1 to 8.6; p = 0.0319). However, female never-smokers in the second category of BMI had a significantly low OR (0.4; 95% CI, 0.2 to 0.7; p = 0.0073).
We examined the association of BMI before diagnosis and adenocarcinoma adjusted for age and smoking status. Men with BMI in the leanest category (BMI < 20.8) at 3 years before diagnosis and with BMI in the highest category (BMI [greater than or equal to] 25.0) at 3 to 4 years before diagnosis had increased risk for adenocarcinoma; but in 1 year, 2 years, and 5 years before diagnosis, there was no statistically significant association between BMI and adenocarcinoma in men. In women, no inverse association was observed between BMI at 1 to 5 years before diagnosis and adenocarcinoma (data not shown).
DISCUSSION
An inverse gradient between BMI and the incidence of lung cancer has been reported in several case-control and cohort studies. (1-9) However, most of these studies were based on subjects with symptomatic lung cancer. In hospital-based, case-control studies, it has been suggested that the association between leanness and cancer may be due to failure to take into account preexisting and coexisting conditions associated with weight loss. (15) It is also possible that subjects in poor general health are generally leaner. In addition, most previous cohort studies (2,5,7,8) evaluated BMI using self-reported body size. Self-reported body size was generally close to measured body size; however, small but systematic errors were usually present, with overestimation of height and underestimation of weight. (10-12) Moreover, it should be noted that previous study populations in published articles (1,2,4,9) were symptomatic lung cancer patients. No studies have ever attempted to evaluate the association in preclinical subjects. In order to investigate the association between the risk of lung cancer and BMI in preclinical asymptomatic patients, we therefore conducted a case-control study with lung cancer patients identified in a mass-screening program. In this study, control subjects were selected from the same screened population, and had no abnormalities on chest radiography and laboratory studies. Information on body size was obtained from measurements obtained by nurses at the time of screening.
The results of the present study indicate two important points. The first point is an inverse gradient between BMI and risk of lung cancer in men, which was consistent with previous studies. (1-9) The negative association at the time of lung cancer diagnosis and 1 to 5 years before diagnosis between BMI and risk of lung cancer was observed in current and former male smokers. Therefore, we were able to confirm the hypothesis that leanness is an independent risk factor for lung cancer in men who smoke. Leanness was statistically associated with lung cancer after adjustment for smoking status, although this association was not found in never-smokers. More important is the absence of an in verse association between lower BMI and risk of lung cancer in female patients. BMI at 1 to 5 years prior to diagnosis of lung cancer was not significantly associated with cancer risk in women. These data suggest that the absence of an inverse association between BMI and risk of lung cancer in women is inconsistent with the findings of previous studies (2,7,8) of symptomatic lung cancer. In contrast with previous reports, Rauseher et al (16) reported an elevated risk of lung cancer associated with higher levels of BMI, and attributed their results to the large number of nonsmokers included in their analysis. The difference between the findings of previous studies and our own was most probably due to different study populations. As Rauseher et al (16) indicated, a higher proportion of adenocarcinoma and nonsmokers in our study population might also have influenced the results. In addition, the population of the present study was quite different from previous study populations. Namely, all the subjects in this study were preclinical patients even at the time of lung cancer diagnosis. All the control subjects in our study had normal chest radiograph findings as well as routine laboratory test results. But the subjects in previous studies were symptomatic and had a variety of clinical conditions. This, we believe, is the main reason for the different results.
Mass screening for lung cancer in Japan is a unique system (17-21): annual chest radiography and sputum cytology are performed together with blood tests such as peripheral blood tests and biochemical examinations. A study group (22-23) from Boston recently reported that obese white women were less likely to undergo breast, cervical, or colon cancer screening than nonobese women. In mass screening in Japan, no such disparity has been noted. In this study, BMI was correctly measured in all the participants.
Olsen et al (8) examined the association of anthropometric factors stratified by histologic type of lung cancer, and revealed an inverse association between BMI and adenocarcinoma of the lung, even among never-smokers. In our present study, both thin and obese men had an increased risk of adenocarcinoma, and this association was observed 1 to 4 years before the diagnosis of adenocarcinoma. The results of the present study indicate that BMI at 1 to 5 years before diagnosis was not significantly associated, although women with near-average BMI at the time of diagnosis showed a decreased risk for adenocarcinoma. However, there were too few cases of squamous cell carcinoma and small cell carcinoma among female smokers to permit a more precise analysis.
It is well known that smokers tend to be leaner than nonsmokers. (1,24-28) Several previous studies have reported an association between leanness and risk of lung cancer, mainly among smokers, (26,28) and among men with smoking-related disease. (28) In accordance with previous studies, (1-9) we found an increased risk of lung cancer for lower BMI at the time of lung cancer diagnosis in men after stratification by smoking status. The association between leanness and risk of lung cancer was also found at 1 to 2 years before lung cancer diagnosis. It is well known that COPD or reduced pulmonary function is associated with weight loss as well as an increased risk of lung cancer. Adjustment for smoking in this analysis may still have allowed for residual confounding by factors related to impaired pulmonary function caused by smoking. Henley et al (29) indicated that the association of BMI with risk of lung cancer is reliably evaluated only when smokers and subjects with preexisting disease are excluded. In our study the association between lower BMI and risk of lung cancer was found in men, the majority of whom were smokers. It seems reasonable to assume that the inverse association in men was due to the residual effects of smoking and chronic lung disease associated with smoking.
CONCLUSION
The present study tends to confirm the previously reported possibility that an inverse association between leanness and risk of lung cancer in men exists, and this finding is influenced by the effects of smoking and coexisting smoking-related respiratory diseases such as COPD. In addition, our results appear to contradict the previous finding that an inverse association in women exists. A larger cohort study combined with mass-screening project will confirm our results.
REFERENCES (1) Nomura A, Heilbrun LK, Stemmermann GN. Body mass index as a predictor of cancer in men. J Natl Cancer Inst 1985; 74:319-323
(2) Kabat GC, Wynder EL. Body mass index and lung cancer risk. Am J Epidemiol 1992; 135:769-774
(3) Knekt P, Heliovaara M, Rissanen A, et al. Leanness and lung-cancer risk. Int J Cancer 1991; 49:208-213
(4) Chyou PH, Nomura A, Stemmermann GN. A prospective study of weight, body mass index and other anthropometric cancers. Int J Cancer 1994; 57:313-317
(5) Drinkard CR, Sellers TA, Potter JD, et al. Association of body mass index and body fat distribution with risk of lung cancer in older women. Am J Epidemiol 1995; 142:600-607
(6) Kark JD, Yaari S, Rasooly I, et al. Are lean smokers at increased risk of lung cancer? The Israel civil servant cancer study. Arch Intern Med 1995; 155:2409-2416
(7) Goodman MT, Wilkens LR. Relation of body size and risk of lung cancer. Nutr Cancer 1993; 20:179-186
(8) Olsen JE, Yang P, Schmitz K, et al. Differential association of body mass index and tilt distribution with three major histologic type of lung cancer: evidence from a cohort of older women. Am J Epidemiol 2002; 156:606-615
(9) Albanes D, Jones DY, Schatzkin A, et al. Adult stature and risk of cancer. Cancer Res 1988; 15:1658-1662
(10) Stunkard AJ, Albaum JM. The accuracy of self-reported weight. Am J Clin Nutr 1981; 34:1593-1599
(11) Stewart AL. The reliability and validity of self-reported weight and height. J Chron Dis 1982; 35:295-309
(12) Palta M, Prineas RJ, Berman R, et al. Comparison of self-reported and measured height and weight. Am J Epidemiol 1982; 115:223-230
(13) Travis WD, Colby TV, Corrin B, et al. Histological typing of lung and pleural tumors. 3rd ed. Berlin, Germany: Springer, 1999
(14) Mountain CF. Revisions in the international system for staging lung cancer. Chest 1997; 111:1710-1717
(15) Marton KI, Sox HC, Krupp JR. Involuntary weight loss: diagnostic and prognostic significance. Ann Intern Med 1981; 95:568-574
(16) Rauscher GH, Mayne ST, Janerich DT. Relation between body mass index and lung cancer risk in men and women never and former smokers. Am J Epidemiol 2000; 152:506-513
(17) Satoh H, Ishikawa H, Yamashita YT, et al. Outcome of patients with lung cancer detected by mass screening versus presentation with symptoms. Anticancer Res 1997; 17:2293-2296
(18) Kanashiki M, Satoh H, Ishikawa H, et al. Time from finding abnormality on mass-screening to final diagnosis of lung cancer. Oncol Rep 2003; 10:649-652
(19) Okamoto N, Suzuki T, Hasegawa H, et al. Evaluation of a clinic-based screening program for Lung cancer with a case-control design in Kanagawa, Japan. Lung Cancer 1999; 25:77-85
(20) Nishi K, Ueoka H, Kiura K, et al. A case-control study of lung cancer screening in Okayama Prefecture, Japan. Lung Cancer 2001; 34:325-332
(21) Nakayama T, Baba T, Suzuki T, et al. An evaluation of chest X-ray screening for lung cancer in Gunma prefecture, Japan: a population- based case-control study. Eur J Cancer 2002; 38:1380-1387
(22) Wee CC, McCarthy EP, Davis RB, et al. Obesity and breast cancer screening, J Gen Intern Med 2004; 19:324-331
(23) Rosen AB, Schneider EC. Colorectal cancer screening disparities related to obesity and gender. J Gen Intern Med 2004; 19:332-338
(24) Garrison RJ, Feinleib M, Castelli WP, et al. Cigarette smoking as a confounder of the relationship between relative weight and long-term mortality: the Framingham Heart Study. JAMA 1983; 249:2199-2203
(25) Vandenbroucke JP, Mauritz BJ, De Bruin A, et al. Weight, smoking, and mortality. JAMA 1984; 252:2859-2860
(26) Sidney S, Friedman GD, Siegelaub AB. Thinness and mortality. Am J Public Health 1987; 77:317-322
(27) Marti B, Tuomilehto J, Korhonen HJ, et al. Smoking and leanness: evidence for change in Finland. BMJ 1989; 298: 1287-1290
(28) Wannamethee G, Shaper AG. Body weight and mortality in middle-aged British men: impact of smoking. BMJ 1989; 299:1497-1502
(29) Henley SJ, Flanders WD, Manatunga A, et al. Leanness and lung cancer risk: fact or artifact? Epidemiology 2002; 13:268-276
Maki Kanashiki, MD; Toshimi Sairenchi, MSc; Yoko Saito, MD, PhD; Hiroichi Ishikawa, MD, PhD; Hiroaki Satoh, MD, PhD; and Kiyohisa Sekizawa, MD, PhD
* From Ibaraki Health Service Association (Drs. Kanashiki, Sairenchi, and Saito), Mito, Ibaraki, and Division of Respiratory Medicine (Drs. Ishikawa, Satoh, and Sekizawa), Institute of Clinical Medicine, University of Tsukuba, Tsukuba, Ibaraki, Japan.
Manuscript received January 7, 2005; revision accepted March 15, 2005.
Reproduction of this article is prohibited without written permission from the American College of Chest Physicians (www.chestjournal. org/misc/reprints.shtml).
Correspondence w: Hiroaki Satoh, MD, Division of Respiratory Medicine, Institute of Clinical Medicine, University of Tsukuba, Tsukuba, Ibaraki, Japan; e-mail: hirosato@md.tsukuba.ac.jp
Table 1--Characteristics of 363 Lung Cancer Cases *
Characteristics Men Women
Patients. 230 133
Mean age (range), yr 70.9 (50-79) 68.3 (50-79)
Smoking history
Never-smoker 12 123
Former smoker 77 1
Current smoker 133 6
Unknown 8 3
Histologic types
Adenocarcinoma 104 111
Squamous cell carcinoma 87 7
Small cell carcinoma 19 6
Large cell carcinoma 9 0
Other 11 9
* Data are presented as No. unless otherwise indicated.
Table 2--Association of BMI With Lung Cancer and
Adenocarcinoma
Case Control
Patients, Subjects,
BMI No. No.
Men
All lung cancer
< 20.8 77 156
20.8-22.8 52 174
22.9-24.9 48 184
[greater than or equal to] 25 42 176
Adenocarcinoma only
< 20.8 34 77
20.8-22.8 21 78
22.9-24.9 24 81
[greater than or equal to] 25 20 76
Women
All lung cancer
< 20.8 27 84
20.8-22.8 22 112
22.9-24.9 37 90
[greater than or equal to] 25 37 112
Adenocarcincnna only
< 20.8 25 73
20.8-22.8 14 94
22.9-24.9 32 79
[greater than or equal to] 25 30 87
BMI OR 95% CI
Men
All lung cancer
< 20.8 1.9 * 1.3-2.9
20.8-22.8 1.2 0.7-1.8
22.9-24.9 1.0
[greater than or equal to] 25 0.9 0.6-1.4
Adenocarcinoma only
< 20.8 1.5 0.8-2.8
20.8-22.8 0.9 0.5-1.8
22.9-24.9 1.0
[greater than or equal to] 25 0.9 0.5-1.7
Women
All lung cancer
< 20.8 0.8 0.4-1.4
20.8-22.8 0.5 ([dagger]) 0.3-0.9
22.9-24.9 1.0
[greater than or equal to] 25 0.9 0.6-1.4
Adenocarcincnna only
< 20.8 0.9 0.5-1.6
20.8-22.8 0.4 ([dagger]) 0.2-0.7
22.9-24.9 1.0
[greater than or equal to] 25 0.9 0.5-1.6
* P < 0.01.
([dagger]) p < 0.0.5.
Table 3--Association of Smoking Status With OR at
the Time of Diagnosis (Lung Cancers)
Case Control
Patients, Subjects,
BMI No. No.
Men
Current smokers
< 20.8 52 59
20.8-22.8 31 59
22.9-24.9 28 50
[greater than or equal to] 2.5 20 57
Former smokers
< 20.8 24 64
20.8-22.8 17 67
22.9-24.9 17 98
[greater than or equal to] 25 18 77
Never-smokers
< 20.8 1 33
20.8-22.8 4 48
22.9-24.9 3 36
[greater than or equal to] 25 4 42
Women
Never-smokers
< 20.8 25 81
20.8-22.8 20 106
22.9-24.9 35 87
[greater than or equal to] 2.5 36 109
BMI OR 95% CI
Men
Current smokers
< 20.8 1.7 0.9-3.0
20.8-22.8 0.9 0.5-1.9
22.9-24.9 1.0
[greater than or equal to] 2.5 0.6 0.3-1.2
Former smokers
< 20.8 2.3 * 1.1-4.6
20.8-22.8 1.6 0.6-3.5
22.9-24.9 1.0
[greater than or equal to] 25 1.3 0.6-2.7
Never-smokers
< 20.8 0.6 0.1-7.8
20.8-22.8 1.1 0.2-6.3
22.9-24.9 1.0
[greater than or equal to] 25 1.8 0.3-11.1
Women
Never-smokers
< 20.8 0.8 0.4-1.4
20.8-22.8 0.5 * 0.3-0.9
22.9-24.9 1.0
[greater than or equal to] 2.5 0.9 0.5-1.5
* p < 0.05.
Table 4--Association of BMI Based on 1 to 5 Years
Prior to Diagnosis With Lung Cancer in Men, Adjusted
for Smoking Status and Age
Case Control
Patients, Subjects,
BMI No. No.
One year prior to diagnosis
< 20.8 55 160
20.8-22.8 52 161
22.9-24.9 32 194
[greater than or equal to] 25.0 37 175
Two years prior to diagnosis
< 20.8 57 162
20.8-22.8 56 164
22.9-24.9 28 187
[greater than or equal to] 25.0 43 177
Three years prior to diagnosis
< 20.8 48 160
20.8-22.8 58 163
22.9-24.9 31 188
[greater than or equal to] 25.0 36 179
Four years prior to diagnosis
< 20.8 53 157
20.8-22.8 52 169
22.9-24.9 40 180
[greater than or equal to] 25.0 35 184
Five years prior to diagnosis
< 20.8 53 144
20.8-22.8 48 166
22.9-24.9 37 196
[greater than or equal to] 25.0 34 184
BMI OR 95% CI
One year prior to diagnosis
< 20.8 1.8 ([dagger]) 1.1-3.0
20.8-22.8 1.8 ([dagger]) 1.1-2.9
22.9-24.9 1.0
[greater than or equal to] 25.0 1.2 0.7-2.0
Two years prior to diagnosis
< 20.8 2.0 * 1.2-3.4
20.8-22.8 2.0 * 1.2-3.4
22.9-24.9 1.0
[greater than or equal to] 25.0 1.6 0.9-2.7
Three years prior to diagnosis
< 20.8 1.6 0.9-2.7
20.8-22.8 1.9 ([dagger]) 1.1-3.2
22.9-24.9 1.0
[greater than or equal to] 25.0 1.3 0.8-2.3
Four years prior to diagnosis
< 20.8 1.3 0.8-2.1
20.8-22.8 1.3 0.8-2.0
22.9-24.9 1.0
[greater than or equal to] 25.0 0.9 0.5-1.5
Five years prior to diagnosis
< 20.8 1.7 ([dagger]) 1.0-2.8
20.8-22.8 1.4 0.8-2.3
22.9-24.9 1.0
[greater than or equal to] 25.0 1.1 0.7-1.9
* p < 0.01.
([dagger]) p < 0.05.
Table 5--Association of BMI Based on I to 5 Years
Prior to Diagnosis With Lung Cancer in Women,
Adjusted for Smoking Status and Age
Case Control
Patients, Subjects,
BMI No. No.
One year prior to diagnosis
< 20.8 22 88
20.8-22.8 19 99
22.9-24.9 27 103
[greater than or equal to] 25.0 30 109
Two years prior to diagnosis
< 20.8 25 83
20.8-22.8 24 102
22.9-24.9 22 96
[greater than or equal to] 25.0 30 118
Three years prior to diagnosis
< 20.8 28 83
20.8-22.8 24 95
22.9-24.9 19 109
[greater than or equal to] 25.0 36 112
Four years prior to diagnosis
< 20.8 21 80
20.8-22.8 21 105
22.9-24.9 20 105
[greater than or equal to] 25.0 38 112
Five years prior to diagnosis
< 20.8 22 77
20.8-22.8 23 98
22.9-24.9 24 107
[greater than or equal to] 25.0 38 117
BMI OR 95% CI
One year prior to diagnosis
< 20.8 0.9 0.5-1.8
20.8-22.8 0.7 0.2-1.4
22.9-24.9 1.0
[greater than or equal to] 25.0 1.1 0.6-2.1
Two years prior to diagnosis
< 20.8 1.3 0.7-2.6
20.8-22.8 1.0 0.6-2.1
22.9-24.9 1.0
[greater than or equal to] 25.0 1.2 0.6-2.3
Three years prior to diagnosis
< 20.8 1.9 1.0-3.7
20.8-22.8 1.5 0.7-2.8
22.9-24.9 1.0
[greater than or equal to] 25.0 1.9 * 1.0-3.6
Four years prior to diagnosis
< 20.8 1.3 0.7-2.7
20.8-22.8 1.1 0.6-2.2
22.9-24.9 1.0
[greater than or equal to] 25.0 1.9 * 1.0-3.6
Five years prior to diagnosis
< 20.8 1.2 0.6-2.4
20.8-22.8 1.0 0.5-2.0
22.9-24.9 1.0
[greater than or equal to] 25.0 1.5 0.8-2.6
* p < 0.05.
Table 6 Association of BMI Based on 1 to 5 Years
Prior to Diagnosis With Lung Cancer in Female Never-
smokers, Adjusted for Smoking Status and Age
Case Control
Patients, Subjects,
BMI No. No.
One year prior to diagnosis
< 20.8 19 83
20.8-22.8 16 95
22.9-24.9 24 100
[greater than or equal to] 25.0 28 104
Two years prior to diagnosis
< 20.8 22 79
20.8-22.8 19 98
22.9-24.9 21 93
[greater than or equal to] 25.0 29 112
Three years prior to diagnosis
< 20.8 25 79
20.8-22.8 19 90
22.9-24.9 18 108
[greater than or equal to] 25.0 34 105
Four years prior to diagnosis
< 20.8 17 75
20.8-22.8 20 102
22.9-24.9 17 98
[greater than or equal to] 25.0 36 107
Five years prior to diagnosis
< 20.8 19 73
20.8-22.8 20 94
22.9-24.9 21 104
[greater than or equal to] 25.0 36 111
BMI OR 95% CI
One year prior to diagnosis
< 20.8 0.9 0.5-1.8
20.8-22.8 0.7 0.4-1.4
22.9-24.9 1.0
[greater than or equal to] 25.0 1.2 0.6-2.3
Two years prior to diagnosis
< 20.8 1.2 0.6-2.4
20.8-22.8 0.9 0.5-1.9
22.9-24.9 1.0
[greater than or equal to] 25.0 1.3 0.7-2.5
Three years prior to diagnosis
< 20.8 1.9 0.9-3.8
20.8-22.8 1.4 0.7-2.8
22.9-24.9 1.0
[greater than or equal to] 25.0 2.1 * 1.1-4.0
Four years prior to diagnosis
< 20.8 1.3 0.6-2.7
20.8-22.8 1.3 0.6-2.6
22.9-24.9 1.0
[greater than or equal to] 25.0 2.6 * 1.2-4.3
Five years prior to diagnosis
< 20.8 1.0 0.6-2.5
20.8-22.8 1.0 0.5-2.1
22.9-24.9 1.0
[greater than or equal to] 25.0 1.7 0.9-3.1
* p < 0.05.
Table 7--Association of Smoking Status With OR at
the Time of Diagnosis (Adenocarcinomas)
Case Control
Patients, Subjects,
BMI No. No.
Men
Current smokers
< 20.8 20 33
20.8-22.8 13 33
22.9-24.9 13 18
[greater than or equal to] 25.0 6 29
Former smokers
< 20.8 13 32
20.8-22.8 5 22
22.9-24.9 9 46
[greater than or equal to] 25.0 12 29
Never-smokers
< 20.8 1 12
20.8-22.8 3 23
22.9-24.9 2 17
[greater than or equal to] 25.0 2 18
Women
Never-smokers
< 20.8 23 69
20.8-22.8 14 89
22.9-24.9 31 77
[greater than or equal to] 25.0 30 86
BMI OR 95% CI
Men
Current smokers
< 20.8 0.9 0.3-2.4
20.8-22.8 0.6 0.2-1.6
22.9-24.9 1.0
[greater than or equal to] 25.0 0.3 ([dagger]) 0.1-1.0
Former smokers
< 20.8 3.1 ([dagger]) 0.1-1.0
20.8-22.8 1.7 0.5-6.1
22.9-24.9 1.0
[greater than or equal to] 25.0 2.3 0.8-6.3
Never-smokers
< 20.8 1.5 0.1-28
20.8-22.8 2.0 0.2-25
22.9-24.9 1.0
[greater than or equal to] 25.0 1.0 0.1-19
Women
Never-smokers
< 20.8 0.8 0.4-1.6
20.8-22.8 0.4 * 0.2-0.7
22.9-24.9 1.0
[greater than or equal to] 25.0 0.9 0.5-1.6
* p < 0.01.
([dagger]) p < 0.05.
COPYRIGHT 2005 American College of Chest Physicians
COPYRIGHT 2005 Gale Group
|
