||1R03CA226942-01A1 Interpret this number
||Harvard School Of Public Health
||Impact of Screening and Diagnostic Intensity on the Study of Prostate Cancer Epidemiology
Many lifestyle factors have been examined as possible prostate cancer risk factors, but results are
inconsistent across studies. It is possible that failure to adequately account for differences in PSA screening
and diagnostic intensity explains some of these inconsistencies.
The introduction of PSA screening for prostate cancer in the late 1980s has had a major impact on
prostate cancer epidemiology, with a sustained increase in incidence rates, a shift to more localized disease,
and decreased mortality rates. Both screening and willingness to undergo a prostate biopsy, the diagnostic
procedure that follows an elevated PSA, are associated with various demographic and lifestyle factors. These
differences in men’s PSA screening and biopsy behaviors may distort the estimated effects of lifestyle factors
on prostate cancer risk, but few attempts have been made to quantitatively assess this source of possible bias.
Our scientific premise is that appropriately accounting for the impact of screening and biopsy is necessary to
elucidate the underlying associations between lifestyle factors and prostate cancer risk.
The objective of this project is to study the impact of screening and propensity for biopsy on the
association between demographic and lifestyle factors and prostate cancer. This will support more rigorous
interpretation of existing studies of prostate cancer epidemiology and inform data collection and analytic
methods for future studies. We will use data from a large, on-going cohort, the Health Professionals Follow-up
Study (HPFS), a unique resource which has collected information on the use of PSA screening and prostate
biopsy every two years since 1996, and also has updated information on an array of lifestyle factors, and
information on prostate cancer diagnosis and survival. In Aim 1, we will investigate how patterns of prostate
cancer screening and biopsy are associated (a) with demographic and lifestyle factors, including: age, race,
family history, smoking, obesity, physical activity, diet, and medication usage, and (b) with risk of prostate
cancer. In Aim 2, we will use several analytic approaches to study how these lifestyle factors are associated
with prostate cancer risk, independent of screening and biopsy practices.
This study efficiently leverages a unique collection of existing data to provide the first comprehensive
assessment of the impact of prostate cancer screening and diagnosis behaviors on prostate cancer risk
factors. By characterizing and quantifying this detection bias, this project will aid in the interpretation of existing
studies of prostate cancer risk factors and provide direction to future studies on data collection and analytic
methods to better handle these issues. Overall, we hope that this work will shed light on previous
inconsistencies in findings on prostate cancer risk factors and support more rigorous identification of risk
factors for prostate cancer, supporting public health prevention efforts.