Abstract P350: Lower Urinary Tract Symptoms (LUTS) Predict Incident Cardiovascular Disease (CVD) in a Cohort of Middle-aged to Elderly Men

Abstract

Introduction: Both lower urinary tract symptoms (LUTS) and cardiovascular disease (CVD) share multiple risk factors. Whether LUTS predict the development of CVD, independent of these factors, remains unclear. Objectives: To examine the cross-sectional and longitudinal relationship between LUTS (total, storage, and voiding LUTS) and first-event CVD. Methods: Participants were drawn from a cohort of randomly-selected, community-dwelling men aged 35-80 years at recruitment (2002-5;n = 1195;sample response rate: 67.8%). LUTS were assessed using the International Prostate Symptom Scale (IPSS). Men with an overall IPSS score >7 were classified as having total LUTS, while men with IPSS scores of >3 for items 2, 4, & 7 and >4 for items 1,3, 5 & 6 were classified with storage and voiding LUTS, respectively. CVD status was obtained by either self-reported physician diagnosis and/or data linkage with a state-wide registry of in-hospital events. The cross-sectional sample (n=1108) consisted of men who had uncomplicated LUTS and complete LUTS measures at baseline. Binomial logistic regression models were used to fit exposure (LUTS) against outcome (CVD) along with qualifying lifestyle, and health-related factors. The longitudinal sample consisted of all CVD-free men at T1 who attended follow-up visits (2007-10;n=701). Cox proportional models were used to determine to hazard ratios (HR) for first-event CVD or date of data-linkage (June 2015). Results: For the cross-sectional sample, 9.4% (n=104) of men examined were found to have CVD. Higher levels of CVD were observed in men with total (35.6% (n=37) vs. 18.2% (n=198) without CVD, p =0.009), storage (40.4% (n=42) vs. 28.3% (n=308), p =0.009) and voiding (31.7% (n=33) vs. 20.0% (n=218), p =0.005) LUTS. In unadjusted binomial models, CVD was associated with total LUTS (OR: 2.38; 95%CI: 1.23-4.31), storage LUTS (1.55;1.08-2.52), and voiding LUTS (1.55;1.08-3.33). In the multi-adjusted model, voiding (1.65; 1.06-3.89) and total (2.18; 1.17-4.11) LUTS was independently associated with CVD, with no significant associations observed for storage LUTS. In the longitudinal analysis (median follow-up duration: 67 months), n=103 CVD cases were detected (14.7% of sample; 20 MI,38 IHD,18 sudden cardiac arrests, and 27 heart failure cases). In unadjusted Cox proportional models, a significant association was detected for voiding (1.28;0.87-1.55) and total (HR: 1.22; 95%CI 1.05-1.83) LUTS, but not storage LUTS (1.18;0.78-2.01). In multi-adjusted models, total LUTS only (2.89;1.37-6.23) were associated with incident CVD. Conclusions: We demonstrate in a group of broadly-representative, middle aged men an independent association between total LUTS at baseline and the development of CVD, suggesting its utility as a prognostic marker for CVD.