The prevalence and correlates of urinary tract symptoms in Norwegian men: The HUNT Study
Arnfinn Seim, Department of Public Health and General Practice, Norwegian University of Science and Technology, MTFS, NO-7489 Trondheim, Norway.
To estimate the prevalence of lower urinary tract symptoms (LUTS) by severity (using the International Prostate Symptom Score, IPSS) in a population-based study of men aged ≥ 20 years, and to assess the association between putative risk factors and the presence of moderate to severe LUTS.
SUBJECTS AND METHODS
Between 1995 and 1997, LUTS data were collected from 21 694 male residents aged ≥ 20 years in Nord Trøndelag County in Norway, using the IPSS; from the IPSS (score 0–35) LUTS was defined as a score of ≥ 8, indicating moderate to severe symptoms. We estimated the prevalence of LUTS and used logistic regression analysis to study lifestyle and anthropometric factors, and comorbidity related to LUTS.
The overall prevalence of moderate to severe LUTS was 15.8% (13.2% moderate and 2.6% severe). The prevalence of LUTS increased strongly with age, from ≈ 5% among men aged < 40 years to > 30% when aged ≥ 70 years. Factors positively associated with an increased risk of moderate and severe LUTS were anthropometric (body mass index and waist hip ratio) and lifestyle factors (alcohol consumption and smoking), as well as comorbid conditions, including diabetes, history of stroke, muscle complaints and osteoarthritis.
The findings from this population-based study suggest that the prevalence of LUTS among men aged ≥ 20 years may be lower than previously estimated. Although LUTS may be viewed as an inevitable consequence of ageing, it appears to be exacerbated by lifestyle factors and comorbid conditions.
body mass index
LUTS can cause personal suffering and reduced quality of life for many men [1–5]; in the USA, ≈ 380 000 TURPs are carried out annually to alleviate LUTS, and the procedure is also common in other western societies. However, the reported prevalence of LUTS varies considerably among studies. This can partly be explained by the lack of consensus on a common definition of LUTS; e.g. some authors [6,7] restrict the diagnosis of LUTS to patients with severe symptoms, whereas others have used a broader definition.
Recently, the definition of LUTS was standardized and the IPSS has now been widely adopted [8,9]. Nonetheless, studies that have used the IPSS have also shown substantial variation in the prevalence of LUTS. In one study comparing the community prevalence of LUTS among four countries, the proportion of men who reported moderate to severe symptoms was 14%, 18%, 38% and 56% in France, Scotland, Olmsted County (USA) and Japan, respectively . A recent study showed less variation in prevalence (16–25%) in the Netherlands, France, UK and Korea , but other community studies reported different prevalence estimates among populations [12,13].
Despite the evidence that LUTS represent a burden, especially among elderly men, little is known about their causes . Apart from the positive association with age, high levels of circulating androgens appear to increase the risk of LUTS, possibly via its association with benign prostate enlargement [9,14–16]. LUTS also appear to be negatively associated with socio-economic status and positively associated with obesity [14,17]. Other lifestyle factors, e.g. cigarette smoking, alcohol consumption and consumption of coffee or tea, may also be positively associated with LUTS [14,17,18], whereas physical activity may be inversely related to the symptoms [19,20]. However, the results of studies are conflicting [14,17–21]. LUTS also tend to occur in conjunction with other age-related conditions. Thus, studies have shown that LUTS may be present together with cardiovascular disease, diabetes and metabolic syndrome, neurological disorders, and rheumatic diseases [14,19,22].
The purpose of the present study was to estimate the prevalence of LUTS by severity (using the IPSS) in a population-based study of men aged ≥ 20 years, and to assess the association between putative risk factors and the presence of moderate to severe LUTS.
SUBJECTS AND METHODS
Data were derived from the second Nord-Trøndelag Health Study (HUNT-2), a population-based study of residents in Nord-Trøndelag County in Norway who were aged ≥ 20 years between 1995 and 1997. The HUNT study was conducted as a collaboration between the National Health Screening Service, the National Institute of Public Health, and the Norwegian University of Science and Technology, and has been described in detail elsewhere [23,24]. Briefly, participants were asked to complete a baseline questionnaire, which was mailed together with the invitation to attend a physical examination. The examination included measurements of height, weight and blood pressure. At the examination, all participants received a gender-specific questionnaire to be completed at home and returned by mail. This questionnaire included the seven IPSS questions that aim at detecting the presence and severity of LUTS .
In all, 66 140 men and women were invited to the HUNT-2 Study, and 71.2% of those who were invited attended the examination. Among the 30 860 men who attended and received the second questionnaire, 21 856 (70.8%) completed and returned the IPSS questions. From those who responded, we excluded 162 because they had previously been diagnosed with prostate cancer. Thus, information from 21 694 men aged ≥ 20 years was used in the current analysis.
All participants signed a consent form that included information about the study objectives. The study was approved by the regional committee for ethics in medical research, by the Norwegian Data Inspectorate, and by the Institutional Review Board of Duke University Medical Center.
The IPSS was calculated for all men who responded to the seven symptom questions (items). Each item has response categories of 0 (not at all) to 5 (almost always); the IPSS is the sum of the seven items, with a total score of 0–35. In the analysis, we categorized LUTS according to total score as no symptoms (IPSS 0), mild (1–7), moderate (8–19) or severe symptoms (20–35), according to recommendations made by the developers of the IPSS .
From the general health survey we also used information on age, height, weight, waist and hip circumference, cigarette smoking, consumption of alcohol, coffee or tea, and information on comorbidity, including diabetes, muscular complaints, osteoarthritis, and history of stroke or coronary heart disease.
The overall prevalence of LUTS was calculated by clinical severity in 10-year age groups, and related to factors previously associated with LUTS. We used logistic regression analysis to estimate the association between the independent variables and the presence of LUTS, by comparing men who reported moderate to severe LUTS with men who reported mild or no LUTS. The associations are presented as odds ratios (OR, corrected for age) with 95% CIs.
Among the 21 694 men, 15.8% reported moderate to severe LUTS (13.2% moderate and 2.6% severe;Table 1); the prevalence of LUTS increased rapidly with age, such that ≈ 95% of men aged < 40 years reported no LUTS, whereas of men aged 60–69 years, 26.3% had LUTS. The prevalence increased gradually from hardly any LUTS in the younger groups to more than a third with LUTS among men aged ≥ 70 years.
Table 1. The proportion of LUTS as measured by the IPSS in different age groups
|20–29 (2414)||45.4|| 50.5||3.9||0.2|
|30–39 (3515)||43.6|| 50.4||5.4||0.5|
|40–49 (5067)||36.7|| 53.0||8.9||1.4|
|50–59 (4192)||24.0|| 58.5||14.6||2.9|
|60–69 (3260)||15.7|| 58.0||21.6||4.7|
|70–79 (2570)||12.7|| 57.8||23.6||5.8|
|80–89 (640)||10.0|| 52.0||30.8||7.2|
|≥90 (36)||11.1|| 41.7||33.3||13.9|
|Total (21694)||29.5|| 54.6||13.2||2.6|
Table 2 shows the association between anthropometric variables and the prevalence of LUTS. There was a weak but positive association between body mass index (BMI) and the prevalence of LUTS. Men with a BMI of 35–39 kg/m2 were ≈ 40% more likely to have LUTS than men with a BMI of <25, and there was a positive trend across categories of BMI. For the waist/hip ratio (WHR), there was a slightly higher prevalence of LUTS among those with a higher WHR, but with no clear trend across categories (quintiles) of WHR.
Table 2. ORs (95% CI) for moderate or severe LUTS (IPSS ≥ 8) associated with age and anthropometric measures, lifestyle factors and disease status
|20–29|| 2 316|| 98||1.0 (Reference)|
|30–39|| 3 306|| 209||1.49 (1.16–1.90)|
|40–49|| 4 547|| 520||2.69 (2.16–3.36)|
|50–59|| 3 459|| 733||4.99 (4.01–6.19)|
|60–69|| 2 404|| 856||8.38 (6.75–10.4)|
|70–79|| 1 813|| 757||9.83 (7.89–12.2)|
|80–89|| 397|| 243||14.4 (11.1–18.6)|
|≥90|| 19|| 17||21.1 (10.6–41.8)|
|<25|| 6 526||1033||1.0 (Reference)|
|25–29|| 9 156||1816||1.13 (1.04–1.23)|
|30–34|| 2 206|| 483||1.20 (1.06–1.35)|
|35–39|| 286|| 65||1.39 (1.04–1.85)|
|≥40|| 45|| 11||1.79 (0.90–3.56)|
|<0.851|| 3 940|| 425||1.0 (Reference)|
|0.851–0.881|| 3 683|| 591||1.20 (1.05–1.38)|
|0.882–0.906|| 3 663|| 700||1.22 (1.07–1.39)|
|0.907–0.942|| 3 642|| 731||1.11 (0.97–1.26)|
|≥0.943|| 3 273|| 959||1.32 (1.15–1.50)|
|Height, cm (quintiles)|
|<173|| 4 116||1048||1.0 (Reference)|
|173–176|| 3 973|| 825||1.10 (0.99–1.23)|
|177–179|| 3 070|| 570||1.15 (1.02–1.29)|
|180–183|| 3 641|| 562||1.11 (0.99–1.26)|
|≥184|| 3 419|| 403||1.04 (0.90–1.18)|
|0|| 3 007|| 707||1.0 (Reference)|
|1–5||10 616||1738||1.12 (1.01–1.25)|
|6–10|| 1 515|| 258||1.41 (1.19–1.66)|
|≥11|| 323|| 46||1.23 (0.88–1.72)|
|Smoking, n cigarettes/day|
|Never|| 7 147|| 999||1.0 (Reference)|
|1–5|| 592|| 158||1.08 (0.89–1.32)|
|6–10|| 1 920|| 462||1.10 (0.97–1.25)|
|11–15|| 909|| 213||1.25 (1.06–1.49)|
|16–20|| 839|| 214||1.34 (1.13–1.59)|
|≥20|| 470|| 141||1.52 (1.24–1.87)|
|1–5|| 558|| 103||0.99 (0.79–1.24)|
|6–10|| 1 728|| 273||0.94 (0.81–1.09)|
|11–15|| 1 442|| 176||0.89 (0.75–1.06)|
|16–20|| 731|| 142||1.38 (1.14–1.68)|
|≥20|| 195|| 48||1.72 (1.23–2.39)|
|Coffee/tea consumption, cups/day|
|0–5|| 9 976||1864||1.0 (Reference)|
|≥6|| 8 071||1499||1.09 (1.01–1.17)|
|No||17 753||3228||1.0 (Reference)|
|Yes|| 474|| 185||1.25 (1.04–1.49)|
|No||17 966||3265||1.0 (Reference)|
|Yes|| 251|| 145||1.61 (1.30–2.00)|
|No||16 031||2640||1.0 (Reference)|
|Yes|| 947|| 325||1.68 (1.46–1.93)|
|No||16 307||2592||1.0 (Reference)|
|Yes|| 1 054|| 513||1.69 (1.50–1.91)|
Several lifestyle factors were also associated with the risk of LUTS. Men who reported having 6–10 drinks (of beer, wine or liquor) per week were more likely (OR 1.41, 95% CI 1.19–1.66) to report LUTS than men who abstained from alcohol (Table 2). Comparing current and former smokers to never-smokers, cigarette smoking was associated with a higher prevalence of LUTS. Among former smokers the results indicated a dose-response relationship between the number of cigarettes and prevalence of LUTS. Men who had smoked > 20 cigarettes/day were ≈ 50% more likely to have LUTS than men who had never smoked. Among current smokers the pattern was somewhat different; smoking > 16 cigarettes/day was positively associated with LUTS. Although not significantly, smoking < 16 cigarettes/day seemed to be associated with a lower prevalence of LUTS. Consumption of coffee or tea also showed weak positive associations with LUTS.
Men with comorbid conditions were also more likely to report LUTS (Table 2); men with diabetes were more likely to have LUTS than men without. There was a similar association between muscle complaints and LUTS, and between osteoarthritis and LUTS. Men with a history of stroke also had higher prevalence of LUTS.
In this population-based study of nearly 22 000 Norwegian men aged ≥ 20 years, where the IPSS was used to assess the presence of LUTS, the total prevalence of moderate to severe LUTS was ≈ 16%, but it increased rapidly with age. Before the age of 40 years LUTS was very rare, but one in five men aged ≥ 40 years reported moderate to severe LUTS. Among men aged ≥ 70 years about a third had LUTS.
Compared with the present results the prevalence of LUTS was substantially higher among African-Americans  and among American men in the NHANES study , and among Australian  and European men . However, in a population-based study from the Netherlands, France, UK and Korea the prevalence of LUTS was similar to that in the present study . These differences suggest that the prevalence of LUTS may have geographical or racial variations, but the results also suggest that men in different cultures may have different thresholds for discomfort before reporting the presence of urinary symptoms.
Previously, some investigators reported a negative association between cigarette smoking and LUTS [19,20,28,29], but others found a positive association [7,14,17,30]. In the present study there was a positive association between cigarette smoking and LUTS among former smokers. A possible negative association between light or moderate smoking and LUTS among current smokers has been shown by others . Previous studies also reported discrepant findings for the consumption of coffee or tea and related LUTS [18,32]. Because coffee and tea are ubiquitous exposures, the potential for public health intervention can be sizeable if the association is causal. However, there was no clear association with tea or coffee consumption in the present men.
Men with comorbid conditions were also more likely to report LUTS. Possibly, these conditions may have a general impact on lower urinary tract function, or the medical treatment of the conditions may influence lower urinary tract function or urine excretion. Thus, men with diabetes were more likely to report LUTS than men without diabetes, suggesting that the pathogenesis may either be vascular or neurological. Patients whose diabetes is not optimally treated will tend to have glucosuria, and thereby be more likely to report urinary symptoms. Similarly, men who have muscle complaints may experience pain that could possibly influence smooth muscle function in the urinary system through neurological mechanisms. However, pain-relieving medication per se has been suggested to increase the risk of LUTS . A positive association between osteoarthritis and LUTS was also reported by others .
In conclusion, ≈ 20% of men aged ≥ 40 years reported moderate to severe LUTS; in men aged ≥ 70 years about a third reported these symptoms. Although LUTS may be viewed as an inevitable consequence of ageing, our results also suggest that LUTS are associated with lifestyle factors, and may be exacerbated by comorbid conditions.
The authors thank Jaspreet Chowdhary, MPH, for technical help with the manuscript. The Norwegian Medical Research Council and Merck Co supported the study financially.
CONFLICT OF INTEREST