Division of Clinical Sciences, National Cancer Institute, NIH, Bethesda, MD, USA,
1 CICAMS, Beijing,
2 Yunnan Tin Corporation, Gejiu, Yunnan Province, People's Republic of China,
3 NHANES Laboratory of Biochemical Analyses at the Centers for Disease Prevention and Control, Atlanta, GA,
4 Information Management Services, Inc., Silver Spring MD,
5 Department of Pathology, Johns Hopkins School of Medicine, Baltimore, MD and
6 H. Lee Moffit Cancer Center and Research Institute, Tampa, FL, USA
Received 26 November 1999; in revised form 7 February 2000; accepted 14 February 2000
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The Alpha-Tocopherol Beta-Carotene (ATBC) cancer prevention study (ATBC Cancer Prevention Study Group, 1994) and Beta-Carotene and Retinol Efficacy Trial (CARET) (Omenn et al., 1996
) were randomized double-blind placebo-controlled trials conducted among smokers and/or occupationally exposed men (and women in CARET) to determine the effects of supplemental doses of ß-carotene (and/or
-tocopherol and retinol palmitate) on lung cancer incidence and mortality. Both studies demonstrated an increased incidence of lung cancer among the ß-carotene-supplemented individuals, compared to those taking placebo. Also in both studies, as suggested by Leo and colleagues, alcohol intake potentiated the adverse effect of ß-carotene supplementation on lung cancer risk (Leo and Lieber, 1994
; Albanes et al., 1996
; Omenn et al., 1996
).
In the present study, we examined the association of pre-diagnostic serum carotenoid levels and risk of lung cancer among Chinese tin miners. Based on the finding of the ATBC and CARET studies, we also examined whether the association between pre-diagnostic serum carotenoid levels and subsequent risk of lung cancer was modified by alcohol intake.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
A longitudinal, prospective cohort study of high risk miners of the YTC was established in 1992 with annual follow-up through 1999. Eligible high risk miners were aged 40 years, with
10 years of underground mining and/or smelting experience, and who were free of cancer (except for non-melanoma skin cancer) at baseline in 1992. The baseline and follow-up activities were added to an annual YTC screening of the miners ongoing since 1973. These activities included: an interview about demographic, dietary, residential, occupational, and medical histories; a 24-h food recall, X-rays, physical examinations, and sputum collections. The initial cohort established in 1992 had 6508 miners, and with each annual screen more miners entered the cohort as they reached the eligibility criteria, resulting in 9142 cohort members by 1997. During the annual screenings in 1993 and 1994, miners were asked to provide a fasting blood specimen. The subcohort who gave blood specimens comprised 40% of the screened population and were representative of the larger high risk cohort. Lung cancer cases were ascertained by reports to the Cancer Registry of the Labour Protection Institute of the YTC or from the annual screens. Over two-thirds of the cases were classified as squamous cell carcinomas of the lung.
Definition of exposures
Cumulative radon exposure for each subject was obtained by summing across the estimated working level months (WLM) for each job held at the YTC prior to baseline screening of the high risk cohort. The cumulative individual arsenic exposure for each job was estimated by using an index for arsenic exposure (index of arsenic exposure months or IAEM), which was calculated as a time weighted average of arsenic concentration (mg/m3) times exposure months (mg/m3 x months). Individuals who had smoked cigarettes and/or pipes (water pipes or Chinese long-stem pipes) regularly for 6 months or longer at any time in their life were classified as ever smokers and were asked for information on a variety of smoking-related issues. Pack year equivalents (g/day x years ÷ 20) were used to measure cumulative tobacco consumption, which was calculated separately for cigarettes (1 cigarette = 1 g) water pipe, and long-stem pipe and for total tobacco use (Qiao et al., 1997).
Trained interviewers administered a culture-specific 109-item dietary food frequency questionnaire (FFQ) in the 1992, 1993, and 1995 annual screens. Miners reported the frequency of intake of foods including alcohol from grain liquor, wine, and other spirits during the past year, with specification of month/season of intake when appropriate. Dietary and alcohol intake information was also obtained by a single 24-h food and beverage recall questionnaire, which was the primary source of dietary data presented in the current study (Forman et al., 1999). The data from the single 24-h food and beverage recall questionnaire and the dietary FFQ were quite concordant. In order to further assure the reproducibility of the questionnaire data, a sample of the participants (2%) were re-interviewed daily by the supervisory staff. In a random subsample of the cohort, a diet validation study was conducted in 19921993 and again in 19951996 in order to compare the diet history FFQ to 28 days of food recalls that were collected over seven consecutive days in each of four seasons. De-attenuated Pearson correlation coefficients of the frequency of food intake between the FFQ and food recalls were in the range of 4.0 to 0.72 for both validation studies (Forman et al., 1999
).
Selection of cases and controls
The cases consisted of 106 men and two women, aged 4179 years, diagnosed with primary lung cancer (ICD-9: 162) during the years 19921997 among those who gave blood at either the 1993 or 1994 screening. Using incidence density sampling, the controls were selected from cohort participants who were alive and free of cancer at the time the case was diagnosed. Controls were matched to cases on age (±2 years), sex and date of blood draw (±2 weeks) in a 2:1 ratio for a total of 108 cases and 216 controls. Selection of cases and controls was independent of assessment of exposure status.
Assessment of micronutrient levels
Serum was prospectively collected (on the average 2.5 years prior to case diagnosis) and was analysed in case-control triplet sets. Interspersed throughout the sets were 48 masked quality control (QC) samples (13% of study group) composed of pooled sera arranged also in sets of three to include one set per assay batch. Laboratory personnel were blind to the case-control status and identity of QC samples. The intra-set coefficients of variation (based on masked reference serum assays) were 1.0, 7.1, 4.6, and 2.1 for -carotene, ß-carotene, lutien/zeaxanthin, and ß-cryptoxanthin, respectively. The inter-batch coefficients of variation were 1.1, 8.4, 3.2, and 5.1 for
-carotene, ß-carotene, lutien/zeaxanthin, and ß-cryptoxanthin, respectively. Lycopene was not included in the analyses, because the serum levels were below the detectable limits and had higher coefficients of variation than for the other carotenoids.
Serum levels of carotenoids were measured by isocratic high-performance liquid chromatography with UV/VIS detection at three different wavelengths. Serum was mixed with an ethanol solution of the internal standard nonapreno- ß-carotene. The analytes were extracted into hexane, which was removed under vacuum. The extract was redissolved in ethanol; an equal volume of acetonitrile was then added. The extract was filtered to remove insoluble material. An aliquot of the filtrate was injected onto a C18 reversed-phase column and eluted with a 50% ethanol:50% acetonitrile solution containing 100 µl of diethylamine/l. Chromatograms at 300, 325, and 450 nm were recorded. Quantification was accomplished by comparing the peak height of the analyte in the unknown with the peak height of a known amount of the same analyte in a standard solution. A correction based on the peak height of an internal standard was used. The carotenoids were compared with nonapreno-ß-carotene at 450 nm (Gunter et al., 1996).
Statistical analyses
The Wilcoxon rank sum test was used to test the hypothesis that the distribution of serum carotenoid concentrations was the same for cases and controls. Conditional logistic regression analysis was used to examine the association between serum carotenoids and risk of lung cancer. Modification of the effect of carotenoids by age, season, tobacco, radon, arsenic, and alcohol consumption on lung cancer risk was examined by statistical tests of the first-order interaction term in the logistic regression models. Statistical analyses stratified by alcohol status (drinkers vs non-drinkers, respectively) were conducted by breaking the case-control match to avoid the loss of subjects due to splitting of matched sets that fell into different strata and using unconditional logistic regression adjusted for age, date of blood draw and sex, the original matching criteria, and other potential confounders. Serum carotenoid levels were categorized into tertiles (one lowest and three being the highest) or median splits (one lowest and two being the highest) based on the distribution of serum carotenoids among the controls. In order to conduct linear trend analyses, variables were created using exposure scores based on the median values of each metabolite for the first to third tertiles among the controls and including them in the logistic regression models. Probabilities of differences at the level of P < 0.05 were regarded as statistically significant.
Potential confounding of the association between serum carotenoids and cancer risk by other related risk factors was explored using Spearman correlation analysis and multivariate logistic regression models. Potential confounding variables were entered in stepwise regression models. If the potential confounder had a significant effect on the log likelihood estimate (P < 0.05) and a greater than 10% change in the serum carotenoid ß-coefficient, it was kept in the model for further multivariate analyses. Exclusion of early cases (blood drawn within 1 year of case diagnosis) did not materially alter the risk estimates of any of the serum carotenoids. All analyses were performed using the statistical software package STATA (Texas).
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Significant positive relationships with lung cancer risk were observed for both serum ß-carotene [odds ratio (OR) 2.0, 95% confidence interval (CI) 1.13.8 for individuals in the highest tertile compared to the lowest tertile] and ß-cryptoxanthin (OR 2.9, 95% CI 1.45.8) (Table 2). Excluding cases that were diagnosed with lung cancer within 1 year of blood collection had no effect on the risk estimates for the association between each carotenoid and lung cancer. Adjustment for baseline covariates, such as serum cholesterol, total vegetable, total fruit, meat, and alcohol consumption, arsenic, radon or tobacco exposure, that were correlated with any of the carotenoids and were possibly associated with lung cancer, did not materially change the risk estimates. The risk estimates shown in Table 2
were adjusted for tobacco use (g/day) and radon exposure, because they were significantly associated with lung cancer risk. Furthermore, there was no association between fruit and vegetable intake based on 24-h recalls and lung cancer risk.
|
|
|
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In this population, increased serum ß-carotene and ß-cryptoxanthin levels were apparently associated with increased risk of lung cancer. In our study, there was essentially no association between -carotene or lutein/zeaxanthin levels and lung cancer risk.
Many studies have explored the association between dietary intake or serum levels of carotenoids and lung cancer risk. A study among Finnish men found a moderately strong negative association between dietary -carotene and lung cancer risk (Knekt et al., 1991
). The Finnish study and the Iowa Women's Health Study (Steinmetz et al., 1993
) found no association between dietary lutein and lung cancer risk. Two case-control studies of dietary intake of ß-cryptoxanthin and lung cancer (Candelora et al., 1992
; Le Marchand et al., 1993
), one among Florida women and the other among men and women in Hawaii, described a protective effect from dietary cryptoxanthin intake on the odds of developing lung cancer. Most prospective serologic studies of ß-carotene have shown a consistent protective effect against lung cancer. For example, a serologic study in Washington County, MD, USA (Comstock et al., 1997
) reported an inverse association between
-carotene, ß-carotene, lutein, cryptoxanthin and lung cancer. In contrast, ß-carotene intervention trials among smokers in Finland and the USA showed an increase in lung cancer incidence among the ß-carotene-supplemented individuals, compared to placebo (ATBC Cancer Prevention Study Group, 1994
; Omenn et al., 1996
). Results of these intervention studies among heavy smokers suggest a possible promoting effect of ß-carotene on lung cancer. Both studies also showed an adverse interaction between ß-carotene and alcohol intake that was associated with a further increase in lung cancer incidence in the ß-carotene-supplemented individuals in the highest quartile of alcohol intake (Albanes et al., 1996
; Omenn et al., 1996
).
In the current study, the strong adverse association between serum -carotene, ß-carotene, lutein/zeaxanthin, ß-cryptoxanthin and lung cancer appears among alcohol drinkers only. The consumption of alcohol among the YTC study men is extremely high compared to, for example, drinkers among the ATBC study in Finland (84 vs 20 g/day respectively). Tobacco use among the YTC miners on the other hand was quite similar to the ATBC participants in Finland.
The apparent protective association of carotenoids among non-drinkers and the adverse association among drinkers is intriguing. Alcohol intake has been reported to alter plasma carotenoid concentrations and therefore potentially modify or confound the plasma carotenoidcancer relation (Rimm and Cojditz, 1993). The difficulty in assessing the role of alcohol in the plasma carotenoidcancer relation is further complicated by the strong positive correlation between alcohol intake and smoking (Subar and Harlan, 1993
). In addition, smokers who drink have a different profile of dietary characteristics than non-smokers who drink (Nebeling et al., 1997
). These factors have complicated the alcoholserum carotenoid relation and in turn, its relation to lung cancer.
Alcohol is also known to disrupt lipoproteins, alter liver function and the bioavailability of carotenoids. Ahmed et al., (1994) reported that alcoholics had lower plasma ß-carotene concentrations than control subjects, but heavy drinkers (>200 g/day) had about twice the ß-carotene of those drinking less, with a significant correlation between plasma ß-carotene and alcohol intake. In a controlled diet study, pre-menopausal women had higher plasma - and ß-carotene concentrations and lower lutein concentrations when fed 30 g of alcohol/day for 3 months, compared to 3 months without alcohol intake (Forman et al., 1995
). It has also been shown in a series of studies in baboons that alcohol raises the level of carotenoids in the plasma and delays its clearance (Leo et al., 1992
). In addition, alcohol consumption may also increase the intestinal absorption of carotenoids by improving solubility and aiding in its transport across the mucosal surface. However, these findings do not explain the differential carotenoid levels seen in the current study comparing cases to controls by alcohol status. What we seem to observe is an alcohol- and carotenoid-derived interaction that possibly enhances the mechanisms of lung cancer development. This observation is very similar to the results of the two ß-carotene cancer prevention trials (Albanes et al., 1996
; Omenn et al., 1996
).
One of the strengths of this study is its prospective design. The collection of blood specimens and covariate data before case diagnosis minimized the potential for recall bias and for disease to affect serum carotenoid measurements. However, the average time from blood collection to diagnosis was short (2.5 years). Nonetheless, exclusion of early cases altered neither the risk estimates nor their statistical significance. The biological relevance of examining the relation between a few carotenoids and subsequent lung cancer incidence is a limitation of this and most similar observational studies. The serum carotenoid levels measured are also probably not indicative of lifetime serum carotenoid status. But, serum levels in controls were of a similar range and mean to serum levels observed among a random subcohort of these miners in 1995 and 1996. In addition, the generalizability of these results may be somewhat restricted, because the study only included individuals exposed to mining-associated occupational pollutants.
In summary, this is the first observational study to report a significant adverse association between prospectively collected serum ß-cryptoxanthin and lung cancer. In addition, our data also suggest a modest positive relation between pre-diagnostic serum levels of ß-carotene and lung cancer risk. Furthermore, alcohol seems to potentiate this adverse affect of serum carotenoids with the observed risk increasing as much as sevenfold for serum ß-carotene and ß-cryptoxanthin levels. Paradoxically, there was a suggestion of a protective association for these analytes among the non-drinkers. Future observational studies need to evaluate the effects of alcohol consumption on the carotenoidcancer association.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
FOOTNOTES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Albanes, D., Heinonen, O. P., Taylor, P. R., Virtamo, J., Edwards, B. K., Rautalahti, M., Hartman, A. M., Palmgren, J., Freedman, L. S., Haapakoski, J., Barrett, M. J., Pietinen, P., Malila, N., Tala, E., Liippo, K., Salomaa, E. R., Tangrea, J. A., Teppo, L., Askin, F. B., Taskinen, E., Erozan, Y., Greenwald, P. and Huttunen, J. K. J. (1996) Alpha-tocopherol and beta-carotene supplements and lung cancer incidence in the Alpha-Tocopherol, Beta-Carotene cancer prevention study: effects of base-line characteristics and study compliance. Journal of the National Cancer Institute 88, 15601570.
ATBC Cancer Prevention Study Group (1994) The effect of vitamin E and beta-carotene on the incidence of lung cancer and other cancers in male smokers. New England Journal of Medicine 330, 10291035.
Candelora, E. C., Stockwell, H. G., Armstrong, A. W. and Pinkham, P. A. (1992) Dietary intake and risk of lung cancer in women who never smoked. Nutrition and Cancer 17, 263270.[ISI][Medline]
Comstock, G. W., Alberg, A. J., Huang, H. Y., Wu, K., Burke, A. E., Hoffman, S. C., Norkus, E. P., Gross, M., Cutler, R. G., Morris, J. S., Spate, V. L. and Helzlsouer, K. J. (1997) The risk of developing lung cancer associated with antioxidants in the blood: ascorbic acid, carotenoids, -tocopherol, selenium, and total peroxyl radical absorbing capacity. Cancer Epidemiology Biomarkers and Prevention 6, 907916.[Abstract]
Doll, R. and Peto, R. (1981) The causes of cancer: quantitative estimates of avoidable risks of cancer in the United States today. Journal of the National Cancer Institute 66, 11911308.[ISI][Medline]
Forman, M. R., Yao, S. X., Graubard, B. I., Qiao, Y. L., McAdams, M., Mao, B. L. and Taylor, P. R. (1992) The effect of dietary intake of fruits and vegetables on the odds ratio of lung cancer among Yunnan Tin Miners. International Journal of Epidemiology 21, 437441.[Abstract]
Forman, M. R., Beecher, G. R., Lanza, E., Reichman, M. E., Graubard, B. I., Campbell, W. S., Marr, T., Yong, L. C., Judd, J. T. and Taylor, P. R. (1995) Effect of alcohol consumption on plasma carotenoid concentrations in premenopausal women: a controlled dietary study. American Journal of Clinical Nutrition 62, 131135.[Abstract]
Forman, M. R., Zhang, J., Gunter, E., Yao, S. X., Gross, M., Qiao, Y. L., Graubard, B. I., Taylor, P. R., Keith, S. and Maher, M. (1999) Season specific correlation between dietary intake of fruits and vegetables and levels of serum biomarkers among Chinese tin miners at high risk for lung cancer. Annals of the New York Academy of Sciences 889, 230239.
Fraser, G. E., Beeson, W. I. and Phillips, R. I. (1991) Diet and lung cancer in California Seventh-day Adventists. American Journal of Epidemiology 133, 683693.[Abstract]
Gunter, E. W., Lewis, B. L. and Koncikowski, S. M. (1996) Laboratory methods used for the Third National Health and Nutrition Examination Survey (NHANES III), 19881994. In: NHANES III Reference Manuals and Reports, NCHS CD-ROM. Centers for Disease Control and Prevention, Hyattsville, MD.
Knekt, P., Jarvinen, R., Seppanen, R., Rissanen, A., Aromaa, A., Heinonen, O. P., Albanes, D., Heinonen, M., Pukkala, E. and Teppo, L. (1991) Dietary antioxidants and the risk of lung cancer. American Journal of Epidemiology 134, 471479.[Abstract]
Krinsky, N. I. (1989) Antioxidant functions of carotenoids. Free Radical Biology and Medicine, 7, 617635.[ISI][Medline]
Le Marchand, L., Hankin, J. H., Kolonel, L. N., Beecher, G. R., Wilkens, L. R. and Zhao, L. P. (1993) Intake of specific carotenoids and lung cancer risk. Cancer Epidemiology Biomarkers and Prevention 2, 183187.[Abstract]
Leo, M. A. and Lieber, C. S. (1994) ß-Carotene, vitamin E, and lung cancer. New England Journal of Medicine 331, 612.
Leo, M. A., Kim, C., Lowe, N. and Lieber, C. S. (1992) Interaction of ethanol with ß-carotene: delayed blood clearance and enhanced hepatotoxicity. Hepatology, 15, 883891.[ISI][Medline]
Nebeling, L., Forman, M. R., Graubard, B. I. and Snyder, R. A. (1997) The impact of lifestyle characteristics on carotenoid intake in the United States: The 1987 National Health Interview Survey. American Journal of Public Health 87, 268271.[Abstract]
Omenn, G. S., Goodman, G. E., Thornquist, M. D., Balmes, J., Cullen, M. R., Glass, A., Keogh, J. P., Meyskens, F. L., Valanis, B., Williams, J. H., Barnhart, S., Cherniack, M. G., Brodkin, C. A. and Hammar, S. (1996) Risk factors for lung cancer and for intervention effects in CARET, the Beta-Carotene and Retinol Efficacy Trial. Journal of the National Cancer Institute 88, 15501559.
Paganini-Hill, A., Chao, A., Ross, R. K. and Henderson, B. E. (1987) Vitamin A, ß-carotene, and the risk of cancer: a prospective study. Journal of the National Cancer Institute 79, 443448.[ISI][Medline]
Qiao, Y. L., Taylor, P. R., Yao, S. X., Schatzkin, A., Mao, B. L., Lubin, J., Rao, J. Y., McAdams, M., Xuan, X. Z. and Li, J. Y. (1989) Relation of radon exposure and tobacco use to lung cancer among tin miners in Yunnan Province, China. American Journal of Industrial Medicine 16, 511521.[ISI][Medline]
Qiao, Y. L., Taylor, P. R., Yao, S. X., Erozan, Y. S., Luo, X. C., Barrett, M. J., Yan, Q. Y., Giffen, C. A., Huang, S. Q., Maher, M. M., Forman, M. R. and Tockman, M. S. (1997) Risk factors and early detection of lung cancer in a cohort of Chinese tin miners. Annals of Epidemiology 7, 533541.[ISI][Medline]
Rimm, E. and Cojditz, G. (1993) Smoking, alcohol and plasma levels of carotenes and vitamin E. Annals of the New York Academy of Sciences 686, 323334.[Abstract]
Ross, M. A., Crosley, L. K., Brown, K. M., Duthie, S. J., Collins, A. C., Arthur, J. R. and Duthie, G. G. (1995) Plasma concentrations of carotenoids and antioxidant vitamins in Scottish males: influences of smoking. European Journal of Clinical Nutrition 49, 861865.[ISI][Medline]
SEER Cancer Statistics Review, 19731991 (1994) National Cancer Institute publication No. 942789, Bethesda, MD.
Steinmetz, K. A. and Potter, J. D. (1992) Vegetables, fruit, and cancer. I. Epidemiology. Cancer Causes and Control 2, 325357.[ISI]
Steinmetz, K. A., Potter, J. D. and Folson, A. R. (1993) Vegetables, fruit and lung cancer in the Iowa Women's Health Study. Cancer Research 53, 536543.[Abstract]
Stryker, W. S., Kaplan, L. A., Stein, E. A., Stampfer, M. J., Sober, A. and Willett, W. C. (1988) The relation of diet, cigarette smoking, and alcohol consumption to plasma ß-carotene and -tocopherol levels. American Journal of Epidemiology 127, 283296.[Abstract]
Subar, A. and Harlan, L. (1993) Nutrient and food group intake by tobacco use status: the 1987 Health Interview Survey. Tobacco, smoking and nutrition, Annals of the New York Academy of Sciences 686, 310322.[ISI][Medline]
Yu, S. Y., Mao, B. L., Xiao, P., Yu, W. P., Wang, Y. L., Huang, C. Z., Chen, W. Q. and Xuan, X. Z. (1990) Intervention trial with selenium for the prevention of lung cancer among tin miners in Yunnan, China. A pilot study. Biological Trace Elements Research 24, 105108.
Zheng, S. F., Ershow, A. G., Yang, C. S., Li, G. Y., Li, R. S., Li, H., Zou, X. L., Liu, X. F., Song, L. H. and Qing, Q. S. (1989) Nutritional status in Linxian, China: effects of season and supplementa- tion. International Journal of Vitamins and Nutrition Research 59, 190199.
Ziegler R. G. (1989) A review of epidemiologic evidence that carotenoids reduce the risk of cancer. Journal of Nutrition 119, 116122.[ISI][Medline]
Ziegler, R. G., Mason, T. J., Stemhagen, A., Hoover, R., Schoenberg, J. B., Gridley, G., Virgo, P. W. and Fraumeni, J. F. (1986) Carotenoid intake, vegetables, and the risk of lung cancer among white men in New Jersey. American Journal of Epidemiology 123, 10801093.[Abstract]