This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kristal, A. R. Articles by Stanford, J. L. Search for Related Content PubMed PubMed Citation Articles by Kristal, A. R. Articles by Stanford, J. L.

Cruciferous Vegetables and Prostate Cancer Risk

Confounding by PSA Screening

Programs in Cancer Prevention, Prostate Cancer Research, and Epidemiology Fred Hutchinson Cancer Research Center Seattle, Washington Department of Epidemiology University of Washington Seattle, Washington

To the Editors: The recent publication by Giovannucci et al. (1) describing associations of cruciferous vegetable consumption with prostate cancer risk is an important contribution to existing literature on diet and cancer. Although cruciferous vegetables were not associated with risk in the primary analyses that included all men in the Health Professionals Follow-up Study, results of analyses among the subgroup of men who reported receiving at least one prostate-specific antigen (PSA) screening test were strikingly different. In these secondary subgroup analyses, there were statistically significant linear decreases in relative risk associated with increasing consumption of cruciferous vegetables. The authors interpret this result as evidence that cruciferous vegetables affect only early stages of prostate carcinogenesis because most cancers detected by PSA screening are organ confined. There is, however, a somewhat more parsimonious interpretation of these results, which has broad implications for current studies of risk factors for prostate cancer.

PSA screening is a strong risk factor for prostate cancer. In a population-based case-control study completed in western Washington State, the age-adjusted odds ratios (95% confidence interval) for prostate cancer were 4.0 (3.1-5.3), 5.5 (3.7-7.3), and 7.1 (4.8-10.5), comparing no PSA test to 1 to 2, 3 to 4, and 5+ PSA tests in the previous 5 years (2). Thus, studies in populations not uniformly screened with PSA are subject to detection bias. In addition, if frequency of PSA testing is positively associated with an hypothesized behavioral risk factor, PSA screening may be act as a special type of confounding variable termed a distorter variable (3), which could mask a true association between behavior and prostate cancer risk (Fig. 1). Indeed, in studies examining healthful dietary behavior and prostate cancer, this is most likely the case. In a population-based survey of Washington State residents, men eating four or more servings of vegetables per day were almost three times more likely to have reported a PSA test in the previous 2 years than men eating fewer than two servings per day (4). Results from the case-control study mentioned above were consistent with this finding, as statistical control for number of PSA tests increased the magnitude of the association between cruciferous vegetable consumption and prostate cancer risk (5). Based on these arguments, we believe that epidemiologic studies of prostate cancer risk need to be controlled for PSA screening, preferably for numbers of recent PSA tests.

View larger version (10K): [in this window] [in a new window]   Figure 1. PSA screening as a distorter variable.

References

1. Giovannucci E, Rimm E, Liu Y, Stampfer M, Willett W. A prospective study of cruciferous vegetables and prostate cancer. Cancer Epidemiol Biomarkers & Prev 2003;12:1403-9.[Abstract/Free Full Text]
2. Stanford JL, Wicklund KG, McKnight B, Daling JR, Brawer MK. Vasectomy and risk of prostate cancer. Cancer Epidemiol Biomarkers & Prev 1999;8:881-6.[Abstract/Free Full Text]
3. Susser M. Causal thinking in the health sciences. New York: Oxford University Press; 1973.
4. Close DR, Kristal AR, Patterson RE, White E. Associations of demographic and health-related characteristics with prostate cancer screening in Washington State. Cancer Epidemiol Biomarkers & Prev 1998;7:627-30.[Abstract]
5. Cohen JH, Kristal AR, Stanford JL. Fruit and vegetable intakes and prostate cancer risk. J Natl Cancer Inst 2000;92:61-8.[Abstract/Free Full Text]

This article has been cited by other articles:

				V. A. Kirsh, U. Peters, S. T. Mayne, A. F. Subar, N. Chatterjee, C. C. Johnson, and R. B. Hayes Prospective Study of Fruit and Vegetable Intake and Risk of Prostate Cancer J Natl Cancer Inst, August 1, 2007; 99(15): 1200 - 1209. [Abstract] [Full Text] [PDF]

				Z. Gong, M. L. Neuhouser, P. J. Goodman, D. Albanes, C. Chi, A. W. Hsing, S. M. Lippman, E. A. Platz, M. N. Pollak, I. M. Thompson, et al. Obesity, diabetes, and risk of prostate cancer: results from the prostate cancer prevention trial. Cancer Epidemiol. Biomarkers Prev., October 1, 2006; 15(10): 1977 - 1983. [Abstract] [Full Text] [PDF]

				A. R. Kristal and J. M. Schenk Directions for Future Epidemiological Research in Lycopene and Prostate Cancer Risk J. Nutr., August 1, 2005; 135(8): 2037S - 2039S. [Full Text] [PDF]

				I. B. King, A. R. Kristal, S. Schaffer, M. Thornquist, and G. E. Goodman Serum Trans-Fatty Acids Are Associated with Risk of Prostate Cancer in {beta}-Carotene and Retinol Efficacy Trial Cancer Epidemiol. Biomarkers Prev., April 1, 2005; 14(4): 988 - 992. [Abstract] [Full Text] [PDF]

This Article Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kristal, A. R. Articles by Stanford, J. L. Search for Related Content PubMed PubMed Citation Articles by Kristal, A. R. Articles by Stanford, J. L.