EPIDEMIOLOGIC ASSOCIATION BETWEEN PROSTATITIS AND PROSTATE CANCER LESLIE K. DENNIS, CHARLES F. LYNCH, AND JAMES C. TORNER ABSTRACT Objectives. To quantify the relationship between prostatitis and prostate cancer by pooling previous epidemiologic studies of this association. Methods. A comprehensive search for articles published through 2000 was performed, blinded reviews of each study were conducted, data were abstracted, and all such studies were pooled. Results. In this meta-analysis, an increased risk was seen among men with a history of prostatitis (odds ratio = 1.6), particularly with population-based case-control studies (odds ratio = 1.8). Increased relative risk estimates were also seen among men with a history of syphilis and a history of gonorrhea. Conclusions. These associations with prostate cancer suggest that infections may represent one mechanism through which prostate cancer develops. However, causality is unclear, because recall bias and detection bias cannot be ruled out. Future cohort studies of prostate cancer should examine sexually transmitted infections, as well as other infections, as potential risk factors. UROLOGY 60: 78–83, 2002. © 2002, Elsevier Science Inc. V arious carcinomas, including liver, large bowel, bladder, esophageal, and stomach, have been shown to be caused by chronic inflam- matory diseases. Inflammations may both induce and promote cancers through exposure to highly reactive compounds and growth factors and in- creased cell turnover. 1 The current classification system for prostatitis includes four types: acute bacterial, chronic bacterial, chronic nonbacterial (inflammatory), and chronic pelvic pain syn- drome. 2 It is plausible that prostatitis is carcino- genic and may be important in the etiology of pros- tate cancer. A previous meta-analysis evaluating aspects of sexual activity and prostate cancer suggested an association between prostate cancer and sexually transmitted infections. 3 This finding supports the consideration of other infectious diseases as poten- tial risk factors for prostate cancer. The overall aim of this study was to examine both the strength and the consistency of the observed associations be- tween prostatitis and prostate cancer by conduct- ing a detailed meta-analysis. We do not attempt to distinguish between the types of prostatitis, be- cause the studies reviewed did not report such in- formation. MATERIAL AND METHODS LITERATURE SEARCH Medline automated citation files were searched using MESH headings, key words, and text words among studies published from 1966 through the end of December 2000. The MESH headings, key words, and text words searched included pros- tatic neoplasms, prostate cancer, and prostatitis. The reports found through the literature searches were gathered, and the references listed in each study were checked for additional relevant studies. DATA ABSTRACTION The studies considered in this meta-analysis were case-con- trol studies or cohort studies that evaluated prostate cancer and prostatitis. The data were abstracted by two of three inde- pendent reviewers (L.K.D., C.F.L., J.C.T.) and included infor- mation available on the odds ratio (OR), 95% confidence in- tervals (CIs), and cell counts of cases and controls by exposure. Inconsistencies were reviewed again until agree- ment was achieved. Third-party resolution of disagreements was sought when needed. Most studies reported ORs that were adjusted for age, although some adjusted for additional con- founders; when available, covariate-adjusted ORs were used in the meta-analysis. STATISTICAL ANALYSIS Two techniques were used to estimate the pooled relative risk estimates of these dichotomous factors: the standard From the Department of Epidemiology, University of Iowa Col- lege of Public Health, Iowa City, Iowa Reprint requests: Leslie K. Dennis, Ph.D., Department of Epi- demiology, University of Iowa College of Public Health, 200 Hawkins Drive, C21-H GH, Iowa City, IA 52242 Submitted: November 5, 2001, accepted (with revisions): Feb- ruary 12, 2002 ADULT UROLOGY © 2002, ELSEVIER SCIENCE INC. 0090-4295/02/$22.00 78 ALL RIGHTS RESERVED PII S0090-4295(02)01637-0