Prognosis of Male Lower Urinary Tract Symptoms and Associated Mortality

Lower urinary tract symptoms (LUTS) are common among men and their prevalence increases with age. The primary reason for seeking care is the bother associated with such symptoms. While the symptoms of LUTS tend to progress with age, spontaneous resolution is also common due to their multifactorial nature. LUTS is associated with various health conditions such as cardiovascular diseases, metabolic syndrome, and diabetes. Through its association with various comorbidities, LUTS may serve as a potential prognostic factor for premature death. However, due to very uncertain cause-and-effect relationships and the confounding effect of ageing and associated health problems, the relevance of LUTS as a prognostic factor for mortality remains unclear.

The primary aim of this thesis was to evaluate the long-term patterns of some of the most common male LUTS, their impact on bother, and associated mortality in a population-based cohort of middle-aged and older men. The analyses were based on repeated assessments during a follow-up spanning over two decades. The secondary aim was to test the usefulness of male LUTS as a predictor for premature death using one of the machine learning techniques, random forest (RF).

The thesis is part of the Tampere Ageing Male Urologic Study (TAMUS). This study started in 1994 including a cohort of 50-, 60-, and 70-year-old men from Pirkanmaa Region (Finland). The cohort was followed-up with repeated questionnaires in 1994, 1999, 2004, 2009, and 2015. A younger cohort of 30-, 40- and 50-year-old men was added in 2004. The questionnaire included assessments of LUTS using the Danish Prostatic Symptom Score (DAN-PSS-1), as well as sociodemographic, anthropometric, and behavioral factors, and medical conditions. Data on medications were obtained from the comprehensive nationwide prescription database (Social Insurance Institution of Finland). Information on surgical procedures was retrieved from the Care Register for Social Welfare and Health Care (HILMO). Mortality data were obtained from Statistics Finland.

To evaluate the course of one of the most common male LUTS – urinary urgency – and associated bother, we included both younger and older cohorts and utilized assessments from 2004 and 2015. Logistic regression was used to explore the risk factors of increased bother at follow-up. The older cohort was included to assess the association between various male LUTS and mortality utilizing repeated questionnaires between 1994 and 2015. Time-dependent Cox regression was utilized in mortality analysis. RF algorithms were moreover developed to predict mortality using LUTS, demographic, medical, and behavioral factors alone and in various combinations.

A total of 2,480 men were included in the analysis concerning urgency and associated bother, assessing their possible adaptation. Of those men, 1,056 (43%) had persistent mild urgency and 132 (5%) persistent moderate or severe urgency at follow-up. The proportions of those experiencing moderate or severe bother due to persistent mild urgency was only 6%. However, 79% of men with persistent moderate or severe urgency reported at least moderate bother at follow-up. In multivariable-adjusted logistic regression, cardiac disease, pulmonary disease, and a history of medically treated symptoms were associated with bother.

A total of 1,332 men were included in the analyses of the association between mortality and urinary urgency, daytime frequency, and nocturia in 21-year follow-up. In multivariable-adjusted analyses, daytime urinary frequency and nocturia were associated with mortality with hazard ratios (HR) of 1.42 (95% CI 1.11-1.83) and 1.38 (1.07-1.79) respectively. The associations were significant only in analyses where the symptoms and comorbidities were updated every five years while the sensitivity analyses with fixed baseline characteristics showed no significant associations.

A total of 1,167 men were included in the analysis assessing the possible modifying effect of symptom severity and bother on mortality in a 24-year follow-up. In multivariable-adjusted analyses disregarding symptom-specific bother, overall voiding and storage LUTS, daytime frequency, and urgency incontinence were associated with increased mortality: the respective HRs were 1.19 (1.00-1.40), 1.35 (1.13-1.62), 1.31 (1.09-1.58) and 2.19 (1.42-3.37). In the analyses disregarding both symptom severity and bother, daytime frequency and nocturia were associated with increased mortality: the respective HRs were 1.31 (1.09-1.58) and 1.52 (1.21-1.91).

A total of 2,663 men were included in the analysis using RF algorithms to predict death. An RF algorithm including information on LUTS only showed an area under the curve (AUC) of 0.60 (0.52-0.69) for five-year mortality. An expanded RF algorithm, including LUTS, medical history, and behavioral and sociodemographic factors, yielded an AUC 0.73 (0.65-0.81), while an algorithm excluding LUTS yielded an AUC 0.71 (0.62-0.78). The respective sensitivities were 0.74 and 0.58 and specificities 0.60 and 0.74.

In conclusion, persistent male LUTS is common and typically mild and well-tolerated. Moderate and severe male LUTS is potential markers of ill health and risk factors of mortality. Furthermore, daytime urinary frequency and nocturia are
potential mortality risk factors even as mild symptoms. Although more severe symptoms are associated with low trends of symptom adaptation, the bother caused by male LUTS is not a relevant predictor of mortality. In clinical practice, repeated assessments of LUTS are needed to identify clinically relevant and persistent symptoms, often associated with poor health. While it is unlikely that LUTS will improve the accuracy for predicting death if the patient’s background is well known, the management of male LUTS should include assessments of men’s general health and comorbidities.