Abstract
Objectives
To understand the association between Lower urinary tract symptoms (LUTS) and risk of bladder cancer in a large men's cohort.
Methods
Using data from the Health Professionals Follow-up Study, we examined risk of bladder cancer in relation to severity of LUTS among 30,183 men. During the follow-up period from 1996 until 2010, 476 newly diagnosed cases of bladder cancer occurred. Cox proportional hazards regression was used to adjust for potential confounders.
Results
Among men with severe LUTS, risk of bladder cancer was 64% higher (relative risk (RR): 1.64, 95% confidence interval (CI): 0.87, 3.08) compared with men who reported no LUTS. Subjects with both voiding and storage dysfunction had a significantly higher risk of bladder cancer (RR: 1.60, 95% confidence interval: 1.00, 2.56). Among individual urinary symptoms, urinary hesitancy was strongly associated with bladder cancer; those who experienced urinary hesitancy at least 50% of the time had more than twice the risk of bladder cancer (RR: 2.21, 95% CI: 1.29, 3.78).
Conclusions
Our findings suggest that LUTS, especially urinary hesitancy, are associated with the development of bladder cancer in men.
Keywords: Lower Urinary Tract Symptoms, Bladder cancer, Epidemiology, Hesitancy, Cohort study
Introduction
Bladder cancer is the fourth leading cause of cancer among men in the United States.1 Approximately 72,570 new cases of bladder cancer were diagnosed in the United States in 2013, among which about 54,610 were men.1 Because of the lengthy and resource-intensive course, health care expenditure for bladder cancer treatment and management is estimated to be among the highest of all malignancies.2 Cigarette smoking is the most well-established risk factor for bladder cancer, which has been estimated to account for about 50% of all cases.3 Occupational exposure to chemical carcinogens increases bladder cancer for workers in industries of aromatic amine manufacture, rubber and cable manufacture, and dyestuff manufacture and use.4
Lower urinary tract symptoms (LUTS), a group of common medical conditions among males, can be defined as abnormal voiding sensations that occur with a frequency or severity that affects quality of life.5,6 About 15–60% of men aged older than 40 years reported having LUTS in different studies, and both the prevalence and severity increase substantially with age.7-10 When LUTS are present, there is substantial amount of post-void residual urine retained in the bladder after urination, increasing the contact time of potential carcinogens in urine and the bladder urothelium.11,12 Bladder wall thickening and bladder mass increase have been noticed among men with LUTS in multiple studies.13-15
To our knowledge, there has not been any large cohort study that evaluated the association between LUTS and bladder cancer risk among men. We conducted this study to understand whether the development of bladder cancer can be influenced by existence and severity of LUTS among men in the Health Professionals Follow-up Study (HPFS).
Methods
Study population
The HPFS was initiated in 1986, when 51,529 male health professionals residing in all 50 states of United States, aged 40-75 years, were recruited. Participants completed a baseline questionnaire which provided information on age, body size measures, smoking and drinking history, medical history, medication, lifestyle, and physical activity in 1986. Participants' information was updated biennially.16 Deaths of participants were often reported by family members or by the postal service. The National Death Index was also searched to determine deaths of participants in the cohort. This study was approved by the Institutional Review Board of the Harvard School of Public Health, Boston, MA.
Identification of bladder cancer cases
On each biennial questionnaire, participants were asked to report if they had received a diagnosis of bladder cancer. Medical records for all reported cases were requested from hospitals after receiving permission from the participants. With permission from cases or next-of-kin for the deceased, the medical records and pathology reports were reviewed. 87% of the cases were confirmed by medical record review or cancer registry, and the diagnoses of the remaining cases are confirmed by obtaining additional data from the cohort member or a surviving family member through phone interviews. Cancer registries (listed in acknowledgment section) are mostly used to confirm cancer diagnoses for participants who indicated a cancer diagnosis during follow-up and subsequently died. Name, date of birth, date of death and social security are sent to the cancer registries; all linkages are conducted at the cancer registries.. Bladder cancer cases were included in the analysis only if a medical record or other sources of confirmation were ascertained successfully. Tumors that had invaded subepithelial connective tissue, muscle, perivesical tissues, or pelvic wall or abdominal wall, or metastasis (T1-T4) were categorized as invasive cancer cases. Non-invasive papillary carcinomas (stage Ta) were considered as non-invasive cancer cases, while Carcinoma in situ (CIS) tumors were categorized in the invasive group due to high risk of progression.17
Ascertainment of Lower Urinary Tract Symptoms
LUTS condition was assessed from all participants in 1992, 1994, 1998, and 2000.18 The American Urological Association Symptom Score Index (AUA-7) was slightly modified to fit the constraints of the mailed questionnaires to assess LUTS information in the HPFS.19 The AUA-7 consisted of seven questions assessing two broad categories of male LUTS: storage and voiding dysfunction. The AUA-7 was extensively validated and found to have good construct and criterion validity, and high test-retest reliability.20,21 Each participant was asked to report the frequency of experiencing the six symptoms over the past month, including sensation of incomplete bladder emptying, frequent urination, stopping and urinary intermittence, urinary urgency, weak urinary stream, and urinary hesitancy. For each of the symptoms, a score of 0 to 5 was assigned corresponding to 0, 10, 25, 50, 75, and almost 100 percent of the time that the symptom was experienced. Further, they were also asked about frequency of night urination. A score of 0 to 4 was assigned corresponding to 0 to 4 times of urination per night, and a score of 5 was assigned for 5 or more times per night. If part of the seven questions were responded, values of other items were replaced with a score of 0. The scores for each of the seven questions were summed up. To evaluate the impacts of voiding dysfunction and storage dysfunction, the scores for the four voiding symptoms (sensation, intermittence, weak stream, and hesitancy) and the scores for the two storage symptoms (urgency and frequent urination) and frequency of night urination were summed up separately.
Assessment of smoking history and dietary factors
Information on cigarette smoking was collected on the baseline questionnaire and updated biennially. A validated, 131-item semiquantitative food-frequency questionnaire was used to assess the baseline dietary intake and every 4 years thereafter.22,23
Statistical analysis
We investigated the risk of bladder cancer in relation to total LUTS condition, voiding dysfunction and storage dysfunction, individual urinary symptoms, and frequency of night urination. The average total LUTS scores from 1992 and 1994 assessments, as well as the average total LUTS scores from 1998 and 2000 assessments, were obtained separately. Participants who did not provide valid LUTS information on either 1992 or 1994 questionnaire were excluded (n=249). If LUTS information was missing on both 1998 and 2000 questionnaires, the average of 1992 and 1994 score was carried forward. All participants with valid LUTS information were classified into four overall LUTS severity groups as having: 1) no LUTS (AUA-7≤ 1 of 35 points), 2) mild LUTS (AUA-7>1 but ≤7 of 35 points), 3) moderate LUTS (AUA-7>7 but ≤19 of 35 points), 4) severe LUTS (AUA-7>19 of 35 points). Participants were also grouped into voiding dysfunction group (≥8 of 20 points for sensation, intermittence, weak stream, and hesitancy), or storage dysfunction group (≥6 of 15 points for frequent urination, urgency, night urination).
As some bladder cancer cases may present urinary symptoms that were similar to LUTS which can bias the underlying association away from the null, the 1994 and 1995 follow-up were excluded to minimize the possibility of reverse causation. We calculated the person-time of follow-up for each participant from the return data of the questionnaire in 1996 to the date of diagnosis of bladder cancer, the date of death from any cause, or January 31, 2010, whichever came first. With the 2-year lag, person-time occurred during 1996-2002 was related to the average of 1992 and 1994 LUTS assessments, and person-time occurred after 2002 was related to the average of 1998 and 2000 LUTS assessments. Additional analyses excluded the first 4 years of follow-up to further rule out the possibility of reverse causation.
Relative risks (RRs) for each of the exposed categories were computed by dividing the incidence rates in these categories by the rate in the reference category. Cox proportional hazard models were used to obtain RRs and 95% confidence intervals (CI) adjusting for potential confounders, which were included in the models if they were found to be risk factors of bladder cancer in previous HPFS studies.24-27 All final models were adjusted for age (continuous in years), pack-years of smoking (6 categories), updated smoking status (current, past, never), geographic region (4 regions), cumulative intake of total fluid (in quartiles), cumulative consumption of bacon (3 categories), cumulative intake of fruit and vegetables (4 categories), and diagnosis of diabetes (updated status with 2 year lag). A new data record was created for every questionnaire cycle at which a subject was at risk by use of covariate values at the time that the questionnaire was returned (i.e., time-varying covariates were used for variables that could potentially change during follow-up, including smoking status, pack-years of smoking, and diagnosis of diabetes). For intake of total fluid, bacon, and fruit and vegetables, the cumulative average values were computed using preceding measurements from all available questionnaires for each participant. Statistical analyses in this study were conducted using SAS 9.3 statistical package (SAS Institute Inc.) and all reported p-values are two-sided.
Results
Valid LUTS information was obtained from 37,591, 37,587, 35,633, and 33,186 participants in the cohort in 1992, 1994, 1998, and 2000, respectively. Percentage of men with self-reported total LUTS score (AUA-7) equal to or higher than 8 were noticed to increase over time (17.31% in 1992, 21.36% in 1994, 26.19% in 1998, and 28.64% in 2000). The percentages of men reporting having voiding dysfunction, storage dysfunction, and mixed dysfunction increased by 59.6%, 54.3%, and 51.4% from 1992 to 2000, respectively.
After excluding participants who reported implausibly high or low values for total energy intake (<800 or >4200 kcal/day) or left more than 70 items blank on the baseline dietary questionnaire, who died between 1986 and 1996, and who reported a diagnosis of cancer (not including nonmelanoma skin cancer) before 1996, the remaining 30,183 men were eligible for the analysis. A total of 476 bladder cancer cases and 373,442 person-years occurred in the cohort during 1996 - 2010. Age-standardized characteristics of the remaining cohort in 1996 by categories of LUTS severity status according to the average AUA-7 from 1992 and 1994 assessments are shown in Table 1.
Table 1.
Age-standardized characteristics of the study population who were free of cancer in 1996 according to total LUTS score categories.1
| LUTS status2 | ||||
|---|---|---|---|---|
| AUA-7≤1 | 1<AUA-7≤7 | 7<AUA-7≤19 | 19<AUA-7≤35 | |
|
|
||||
| n | 7,597 | 17,075 | 5,161 | 350 |
| Age, years | 58.8 ± 8.13 | 61.7 ± 8.9 | 65.5 ± 8.6 | 66.3 ± 8.5 |
| Caucasian race, percentage | 92% | 92% | 92% | 92% |
| Current smoker, percentage | 6% | 5% | 5% | 5% |
| Weight (kg) | 82.9 ± 12.2 | 83.2 ± 12.8 | 83.2 ± 13.3 | 84.4 ± 13.9 |
| Body mass index (kg/m2) | 25.9 ± 3.3 | 26.0 ± 3.5 | 26.2 ± 3.8 | 26.5 ± 3.8 |
| Pack-years of smoking | 10.6 ± 16.9 | 11.6 ± 17.3 | 12.7 ± 18.1 | 15.5 ± 21.2 |
| Intake of4 | ||||
| Total fluid (ml/d) | 1,871 ± 772 | 1,987 ± 794 | 2,007 ± 817 | 2,102 ± 877 |
| Bacon (servings/week) | 0.5 ± 1.1 | 0.5 ± 1.2 | 0.5 ± 1.2 | 0.7 ± 1.7 |
| Diabetes, percentage | 3% | 4% | 5% | 6% |
| Hypertension, percentage | 19% | 24% | 27% | 31% |
All variables, except for age, were standardized according to the age distribution of the entire cohort.
LUTS status was determined by the average of 1992 and 1994 total AUA-7 scores.
Mean ± SD (all such values).
Dietary information from the 1994 questionnaire.
Results of the age-adjusted analysis excluding person-years of the first 2 years' follow-up after exposure assessments are shown in Table 2. Among men with severe LUTS status (AUA-7 > 19), bladder cancer risk was 1.64 times as likely (95% CI: 0.87, 3.08) as among men who reported no LUTS. The multivariable RR of bladder cancer for men reporting mixed voiding and storage symptoms were 1.60 (95% CI: 1.00, 2.56), comparing to men reporting no LUTS symptoms. Voiding dysfunction was more strongly related to bladder cancer risk than storage dysfunction (RR voiding: 1.41, 95% CI: 0.95, 2.11; RR storage: 1.28, 95% CI: 0.92, 1.80). RRs showed minimal changes before and after controlling for multiple potential risk factors. Every one-unit increase in total LUTS score, voiding subgroup score, or storage subgroup score was all associated with about 2%-4% increase in risk of bladder cancer (Table 3). When further excluding the first 4 years' follow-up after exposure assessments to further minimize the possibility of reverse causation (Table 2), the RR for mixed voiding and storage symptoms were 1.87 (95% CI: 1.16, 3.00), comparing to men reporting no LUTS symptoms.
Table 2.
Relative risks of bladder cancer associated with updated LUTS status from latency analysis with 2-year lag and 4-year lag.
| LUTS status | Case No. | Person-Years | Age-adjusted RR (95% CI) | Multivariate RR (95% CI)3 |
|---|---|---|---|---|
| 2-year lag1 | ||||
|
|
||||
| No LUTS (ref)5 | 73 | 82,570 | 1.0 | 1.0 |
| Mild LUTS6 | 258 | 207,242 | 1.14 (0.87, 1.49)4 | 1.14 (0.87, 1.49) |
| Moderate LUTS7 | 133 | 78,669 | 1.18 (0.87, 1.58) | 1.18 (0.87, 1.59) |
| Severe LUTS8 | 12 | 4,961 | 1.62 (0.86, 3.03) | 1.64 (0.87, 3.08) |
| p for trend | 0.11 | 0.12 | ||
| Isolated voiding9 | 15 | 9,037 | 1.25 (0.72, 2.18) | 1.18 (0.66, 2.08) |
| Isolated storage10 | 55 | 30,135 | 1.20 (0.68, 1.70) | 1.17 (0.81, 1.69) |
| Mixed V+S11 | 26 | 11,509 | 1.59 (1.00, 2.52) | 1.60 (1.00, 2.56) |
| All voiding12 | 41 | 20,546 | 1.48 (1.01, 2.16) | 1.41 (0.95, 2.11) |
| All storage13 | 81 | 41,644 | 1.32 (0.97, 1.80) | 1.28 (0.92, 1.80) |
| 4-year lag2 | ||||
|
|
||||
| No LUTS (ref)5 | 66 | 71,607 | 1.0 | 1.0 |
| Mild LUTS 6 | 226 | 174,585 | 1.12 (0.84, 1.48) | 1.13 (0.85, 1.50) |
| Moderate LUTS 7 | 118 | 63,304 | 1.24 (0.90, 1.70) | 1.24 (0.90, 1.71) |
| Severe LUTS 8 | 10 | 3,988 | 1.65 (0.83, 3.26) | 1.68 (0.84, 3.33) |
| p for trend | 0.04 | 0.05 | ||
| Isolated voiding9 | 13 | 7,322 | 1.26 (0.69, 2.31) | 1.21 (0.66, 2.23) |
| Isolated storage10 | 46 | 24,055 | 1.15 (0.78, 1.71) | 1.17 (0.79, 1.74) |
| Mixed V+S11 | 26 | 9,236 | 1.82 (1.14, 2.92) | 1.87 (1.16, 3.00) |
| All voiding12 | 39 | 16,558 | 1.58 (1.05, 2.38) | 1.58 (1.04, 2.38) |
| All storage13 | 72 | 33,291 | 1.33 (0.94, 1.89) | 1.35 (0.95, 1.93) |
Person-time during 1996-2002 was related to the average of 1992 and 1994 LUTS scores; Person-time after 2002 was related to the average of 1998 and 2000 LUTS scores.
Person-time during 1998-2004 was related to the average of 1992 and 1994 LUTS scores; Person-time after 2004 was related to the average of 1998 and 2000 LUTS scores.
Adjusted for age, cumulative total fluid intake (in quartiles), cumulative intake of fruits and vegetables (4 categories), intake of bacon (3 categories), updated smoking status (current smoker vs nonsmoker), cumulative pack-years of smoking (6 categories), and diabetes status two years before.
RR; 95% CI in parentheses (all such values).
No LUTS: total AUA-7 score less than or equal to 1.
Mild LUTS: total AUA-7 score greater than 1 but less than or equal to 7.
Moderate LUTS: total AUA-7 score greater than 7 but less than or equal to 19.
Severe LUTS: total AUA-7 score greater than 19.
Isolated voiding: voiding subgroup score equal to or higher than 8 and storage subgroup score less than 6.
Isolated storage: storage subgroup score equal to or higher than 6 and voiding subgroup score less than 8.
Mixed V+S: voiding subgroup score equal to or higher than 8 and storage subgroup score equal to or higher than 6.
All voiding: voiding subgroup score equal to or higher than 8.
All storage: storage subgroup score equal to or higher than 6.
Table 3.
Association of frequency of individual lower urination tract symptoms and night urination with risk of bladder cancer.1
| Frequency (% of time experienced symptoms) | ||||
|---|---|---|---|---|
| 0 | <25 | ≥25 but <50 | ≥50 | |
| Sense of incomplete emptying | ||||
| No. of cases | 258 | 166 | 33 | 19 |
| Person-years | 232,299 | 111,603 | 19,034 | 10,506 |
| Multivariable RR2 | 1.0 | 1.13 (0.93, 1.38) | 1.13 (0.78, 1.64) | 1.16 (0.72, 1.87) |
| Urinary intermittence | ||||
| No. of cases | 270 | 142 | 38 | 26 |
| Person-years | 235,234 | 102,307 | 20,794 | 15,107 |
| Multivariable RR2 | 1.0 | 1.00 (0.81, 1.23) | 1.10 (0.78, 1.56) | 1.09 (0.72, 1.65) |
| Weak stream | ||||
| No. of cases | 188 | 170 | 63 | 55 |
| Person-years | 187,501 | 117,884 | 33,888 | 34,169 |
| Multivariable RR2 | 1.0 | 1.19 (0.96, 1.48) | 1.38 (1.03, 1.85) | 1.16 (0.85, 1.58) |
| Urinary hesitancy | ||||
| No. of cases | 343 | 100 | 18 | 15 |
| Person-years | 284,494 | 76,132 | 8,452 | 4,364 |
| Multivariable RR2 | 1.0 | 1.01 (0.80, 1.27) | 1.42 (0.87, 2.32) | 2.21 (1.29, 3.78) |
| Frequent urination | ||||
| No. of cases | 129 | 247 | 61 | 39 |
| Person-years | 126,028 | 180,591 | 42746 | 24,077 |
| Multivariable RR2 | 1.0 | 1.25 (1.00, 1.55) | 1.17 (0.85, 1.59) | 1.29 (0.89, 1.87) |
| Urinary urgency | ||||
| No. of cases | 182 | 195 | 57 | 42 |
| Person-years | 180,912 | 138,277 | 30,814 | 23,439 |
| Multivariable RR2 | 1.0 | 1.15 (0.93, 1.41) | 1.22 (0.89, 1.66) | 1.09 (0.77, 1.55) |
| Night urination | ||||
| Times per night | 0 | <2 | ≥2 but <3 | ≥3 |
| No. of cases | 61 | 277 | 105 | 33 |
| Person-years | 68,016 | 226,959 | 58,572 | 19,894 |
| Multivariable RR2 | 1.0 | 1.01 (0.76, 1.34) | 1.07 (0.77, 1.49) | 0.82 (0.53, 1.29) |
Person-time occurred during 1996-2002 was related to the average of 1992 and 1994 symptom scores; Person-time occurred after 2002 was related to the average of 1998 and 2000 symptom scores. If both 1998 and 2000 LUTS information was missing, the average of 1992 and 1994 symptom scores was brought forward.
Adjusted for age, cumulative total fluid intake (in quartiles), cumulative intake of fruits and vegetables (4 categories), intake of bacon (3 categories), updated current smoking status (current smoker vs nonsmoker), cumulative pack-years of cigarette smoking (6 categories), and updated diabetes status two years before.
RR; 95% CI in parentheses (all such values).
We evaluated the associations of each of the six urinary symptoms and frequency of night urination with risk of bladder cancer individually in the 2-year-lag multivariable models (Table 3). Results showed that urinary hesitancy was most strongly associated with bladder cancer risk. Men experiencing urinary hesitancy at least 50% of the time had more than twice the risk of bladder cancer compared to men who never experienced this symptom (RR: 2.21, 95% CI: 1.29, 3.78). Having to urinate equal to or more than 3 times per night showed a nonsignificant decreased risk of bladder cancer comparing to those who did not have night urination among the study population (RR: 0.82, 95% CI: 0.53, 1.29). In a separate multivariable model, all categorical variables of the seven urinary items were added in simultaneously, the RR was 2.69 (95% CI: 1.41, 5.12) for experiencing urinary hesitancy at least 50% of the time as compared with never experiencing urinary hesitancy. When the seven urinary items were added in the multivariable model simultaneously as continuous variables, every one-unit increase in the score of urinary hesitancy elevated the risk of bladder cancer by 21% (RR: 1.21, 95% CI: 1.03, 1.42).
To study whether LUTS had different effects on different stages of bladder cancer, we restricted bladder cancer cases to stage of bladder cancer at diagnosis (non-invasive bladder cancer vs invasive bladder cancer) (Table 4). Results excluding the first 2 years of follow-up after exposure assessments showed that there is a stronger association between LUTS severity and non-invasive bladder cancer. The multivariable RRs for the moderate group (7<AUA-7≤19) and severe group (AUA-7>19) compared with no symptoms were 1.87 (95% CI: 1.05, 3.34) and 2.62 (95% CI: 0.84, 8.19) respectively for non-invasive cancer. Participants with mixed storage and voiding dysfunction had 3.36 times of bladder cancer risk compared with no LUTS group (95% CI: 1.52, 7.44). The magnitudes of the above associations increased in analyses further excluding the first 4 or 6 years of follow-up after exposure assessments.
Table 4.
Relative risks of bladder cancer associated with LUTS, according to stage at diagnosis.
| 2-year lag1 | ||||||
|---|---|---|---|---|---|---|
| Noninvasive | Invasive | |||||
|
|
|
|||||
| LUTS status | Case No. | PYs | Multivariate RR7 | Case No. | PYs | Multivariate RR7 |
| No LUTS2 | 19 | 82,517 | 1.0 | 35 | 82,523 | 1.0 |
| Mild LUTS3 | 100 | 239,965 | 1.60 (0.96, 2.69)8 | 129 | 240,003 | 0.97 (0.66, 1.42) |
| Isolated voiding4 | 3 | 9,025 | 1.28 (0.37, 4.46) | 9 | 9,030 | 1.43 (0.67, 3.04) |
| Isolated storage5 | 16 | 30,088 | 1.82 (0.90, 3.70) | 24 | 30,103 | 0.96 (0.56, 1.66) |
| Mixed V+S6 | 11 | 11,491 | 3.36 (1.52, 7.44) | 10 | 11,489 | 1.09 (0.53, 2.26) |
| 4-year lag9 | ||||||
| Noninvasive | Invasive | |||||
|
|
|
|||||
| LUTS status | Case No. | PYs | Multivariate RR7 | Cases No. | PYs | Multivariate RR7 |
| No LUTS2 | 10 | 70,083 | 1.0 | 33 | 70,099 | 1.0 |
| Mild LUTS3 | 84 | 196,162 | 2.69 (1.38, 5.25) | 101 | 196,181 | 0.82 (0.55, 1.24) |
| Isolated voiding4 | 2 | 7,117 | 1.88 (0.40, 8.80) | 8 | 7,123 | 1.34 (0.60, 2.99) |
| Isolated storage5 | 12 | 23,081 | 2.99 (1.24, 7.17) | 20 | 23,094 | 0.94 (0.53, 1.69) |
| Mixed V+S6 | 9 | 8,832 | 6.08 (2.37, 15.64) | 12 | 8,838 | 1.42 (0.70, 2.86) |
Person-time occurred during 1996-2002 was related to the average of 1992 and 1994 LUTS assessments; Person-time occurred after 2002 was related to the average of 1998 and 2000 LUTS assessments.
No LUTS group was defined as the total AUA-7 score less than or equal to 1; it was used as reference group in the regression models.
Mild LUTS group was defined as anyone who did not have high voiding or storage symptoms, or no LUTS.
Isolated voiding group was defined as voiding subgroup score equal to or higher than 8, and storage subgroup score less than 6.
Isolated storage group was defined as storage subgroup score equal to or higher than 6, and voiding subgroup score less than 8.
Mixed V+S group was defined as voiding subgroup score equal to or higher than 8, and storage subgroup score equal to or higher than 6.
Adjusted for age, cumulative intake of fruits and vegetables (4 categories), intake of bacon (3 categories), updated current smoking status (current smoker vs nonsmoker), cumulative pack-years of cigarette smoking (6 categories), and updated diabetes status two years before.
RR; 95% CI in parentheses (all such values).
Person-time occurred during 1998-2004 was related to the average of 1992 and 1994 LUTS assessments; Person-time occurred after 2004 was related to the average of 1998 and 2000 LUTS assessments.
Discussion
In this prospective study with updated LUTS measurements, we observed a consistent, positive association between voiding dysfunction or mixed dysfunction with risk of bladder cancer. There was a stronger association between bladder cancer risk with voiding dysfunction symptoms comparing with storage dysfunction ones. Total AUA-7 score, storage dysfunction score, or specific storage symptoms, were not found to be significantly related to bladder cancer risk in our study.
Voiding dysfunction is classically attributed to the direct effect of mechanical obstruction of the lower urinary tract, inducing extended contact time of any potential carcinogens present in the urine and the bladder urothelium. Among all specific urinary voiding symptoms, the frequency of urinary hesitancy was found to be strongly associated with elevated bladder cancer risk in our study. Urinary hesitancy, indicating a delay between trying to urinate and the urine flow actually beginning, causes high intravesical pressure. We investigated the relation between hesitancy and other urinary symptoms in this study population. At baseline, 1,892 men reported experiencing urinary hesitancy for at least 50% of the time, among which 1,069 (56.5%) had mixed storage and voiding dysfunctions and 642 (33.93%) had isolated voiding dysfunction. Participants reporting experiencing hesitancy for at least 50% of the time had the highest overall total AUA-7 score (18.84) comparing to experiencing the other urinary symptoms for at least 50% of the time. These results suggest that the presence of severe urinary hesitancy might indicate higher volume of post-void residual urine, which can increase the contact time of potential carcinogens in urine and the bladder urothelium and the risk of bladder cancer.
We found that men who urinated equal to or more than 3 times per night might have a decreased risk of bladder cancer. A recent case-control study found that both men and women who urinated at least twice per night had 40-50% reduction in bladder cancer risk comparing to those who did not urinate at night.28 In a separate analysis which was restricted to men with no or mild LUTS, increased frequency of night urination was not related to reduced bladder cancer risk controlling for total fluid intake and other covariates. In our study, because the night urination question was designed as part of the LUTS severity assessment, it was placed under the six urinary symptom questions. Participants who had the habit of night urination but also believed they did not suffer from urinary problems might unintentionally underreport on the night urination frequency question, which may bias the result toward the null. Future studies directly measuring the volume of post-void residual urine are warranted.
Major strengths of this study include its large sample size, long follow-up period, and repeated measurements of LUTS scores, with participants from all geographic regions of the United States. Because all the participants in this study were health professionals, the self-reported urinary symptoms, as well as other medical conditions, were considered to be relatively accurate. Among participants who responded to at least part of the seven LUTS questions, the missing rates of each of the seven symptoms were very small, ranging from 0.38% to 6.17% in 1992 and 0.23% to 2.09% in 1994. When the missing items were replaced with a score of 0, the results would only be attenuated slightly if the corresponding symptoms were present. The prospective design and application of lag analysis allowed us to minimize the possibility of bias due to reverse causation, meaning the cancer should not occur before the urinary symptoms. In addition, urinary symptoms related to undiagnosed bladder cancer occur mostly on patients with later stages of bladder cancer or carcinoma in situ. However, we found a stronger association between the presence of LUTS and early stages of bladder cancer instead of later stages, which further excludes the possibility that the observed positive associations in this study were all due to reverse causation.
Detection bias refers to a potential artifact caused by the use of a particular diagnostic technique or type of equipment, meaning LUTS can lead to a higher chance of bladder cancer diagnosis instead of bladder cancer cases. In order to minimize the possibility of detection bias, sensitivity analyses were conducted. First, we examined the association of the six specific urinary symptoms and night urination frequency with bladder cancer risk limited to invasive or noninvasive stages at diagnosis (with 2-year lag and updated LUTS information). If the observed positive association of LUTS and early stage bladder cancer risk was all due to detection bias, we would expect to see relatively constant RRs across different symptoms and night urination frequency, because all urinary symptoms could lead to higher chance of visiting urologists or getting urinalysis. In the multivariable model where all six urinary symptoms and night urination frequency variables were added simultaneously, we observed that urinary hesitancy was strongly related to diagnosis of bladder cancer while the other symptoms and night urination frequency were not.
In conclusion, our findings suggest that urinary voiding symptoms may increase the risk of bladder cancer. If this association is confirmed, LUTS might be used as simple test to identify men who are at higher risk of bladder cancer; however, despite statistically significant associations with a 4-year lag period, we cannot rule out the possibility that bladder cancer in these men was diagnosed as a result of urinary symptoms.
Acknowledgments
This work was supported by a grant from the National Institutes of Health (UM1 CA167552).
The authors thank the staff of the Health Professionals Follow-up Study for their valuable contributions. They also thank the following state cancer registries for their help: Alabama, Arizona, Arkansas, California, Colorado, Connecticut, Delaware, Florida, Georgia, Idaho, Illinois, Indiana, Iowa, Kentucky, Louisiana, Maine, Maryland, Massachusetts, Minnesota, Nebraska, New Hampshire, New Jersey, New York, North Carolina, North Dakota, Ohio, Oklahoma, Oregon, Pennsylvania, Rhode Island, South Carolina, Tennessee, Texas, Virginia, Washington, Wyoming.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Society AC. Cancer Facts & Figures 2012. Atlanta: American Cancer Society; 2012. [Google Scholar]
- 2.Avritscher EB, Cooksley CD, Grossman HB, et al. Clinical model of lifetime cost of treating bladder cancer and associated complications. Urology. 2006;68:549–553. doi: 10.1016/j.urology.2006.03.062. [DOI] [PubMed] [Google Scholar]
- 3.Freedman ND, Silverman DT, Hollenbeck AR, et al. Association between smoking and risk of bladder cancer among men and women. JAMA : the journal of the American Medical Association. 2011;306:737–745. doi: 10.1001/jama.2011.1142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Olfert SM, Felknor SA, Delclos GL. An updated review of the literature: risk factors for bladder cancer with focus on occupational exposures. South Med J. 2006;99:1256–1263. doi: 10.1097/01.smj.0000247266.10393.72. [DOI] [PubMed] [Google Scholar]
- 5.Roehrborn CG. Male Lower Urinary Tract Symptoms (LUTS) and Benign Prostatic Hyperplasia (BPH) Med Clin N Am. 2011;95:87–100. doi: 10.1016/j.mcna.2010.08.013. [DOI] [PubMed] [Google Scholar]
- 6.Gammack JK. Lower urinary tract symptoms. Clin Geriatr Med. 2010;26:249–260. doi: 10.1016/j.cger.2010.02.006. [DOI] [PubMed] [Google Scholar]
- 7.Marszalek M, Wehrberger C, Temml C, et al. Chronic pelvic pain and lower urinary tract symptoms in both sexes: analysis of 2749 participants of an urban health screening project. Eur Urol. 2009;55:499–507. doi: 10.1016/j.eururo.2008.03.073. [DOI] [PubMed] [Google Scholar]
- 8.Andersson SO, Rashidkhani B, Karlberg L, et al. Prevalence of lower urinary tract symptoms in men aged 45-79 years: a population-based study of 40 000 Swedish men. BJU Int. 2004;94:327–331. doi: 10.1111/j.1464-410X.2004.04930.x. [DOI] [PubMed] [Google Scholar]
- 9.Kupelian V, Wei JT, O'Leary MP, et al. Prevalence of lower urinary tract symptoms and effect on quality of life in a racially and ethnically diverse random sample: the Boston Area Community Health (BACH) Survey. Arch Intern Med. 2006;166:2381–2387. doi: 10.1001/archinte.166.21.2381. [DOI] [PubMed] [Google Scholar]
- 10.Taylor BC, Wilt TJ, Fink HA, et al. Prevalence, severity, and health correlates of lower urinary tract symptoms among older men: The MrOS study. Urology. 2006;68:804–809. doi: 10.1016/j.urology.2006.04.019. [DOI] [PubMed] [Google Scholar]
- 11.Tokgoz O, Tokgoz H, Unal I, et al. Diagnostic values of detrusor wall thickness, postvoid residual urine, and prostate volume to evaluate lower urinary tract symptoms in men. Diagn Interv Radiol. 2012;18:277–281. doi: 10.4261/1305-3825.DIR.4847-11.1. [DOI] [PubMed] [Google Scholar]
- 12.McNeill SA, Hargreave TB, Geffriaud-Ricouard C, et al. Postvoid residual urine in patients with lower urinary tract symptoms suggestive of benign prostatic hyperplasia: pooled analysis of eleven controlled studies with alfuzosin. Urology. 2001;57:459–465. doi: 10.1016/s0090-4295(00)01021-9. [DOI] [PubMed] [Google Scholar]
- 13.Isikay L, Turgay Akgul K, Nuhoglu B, et al. Lower urinary tract symptoms, prostate volume, uroflowmetry, residual urine volume and bladder wall thickness in Turkish men: a comparative analysis. Int Urol Nephrol. 2007;39:1131–1135. doi: 10.1007/s11255-006-9161-y. [DOI] [PubMed] [Google Scholar]
- 14.Manieri C, Carter SS, Romano G, et al. The diagnosis of bladder outlet obstruction in men by ultrasound measurement of bladder wall thickness. J Urol. 1998;159:761–765. [PubMed] [Google Scholar]
- 15.Kojima M, Inui E, Ochiai A, et al. Ultrasonic estimation of bladder weight as a measure of bladder hypertrophy in men with infravesical obstruction: a preliminary report. Urology. 1996;47:942–947. doi: 10.1016/S0090-4295(96)00059-3. [DOI] [PubMed] [Google Scholar]
- 16.Rimm EB, Stampfer MJ, Colditz GA, et al. Effectiveness of various mailing strategies among nonrespondents in a prospective cohort study. American journal of epidemiology. 1990;131:1068–1071. doi: 10.1093/oxfordjournals.aje.a115598. [DOI] [PubMed] [Google Scholar]
- 17.Dorkin TJ, Robson CN, Neal DE. The molecular pathology of urological malignancies. J Pathol. 1997;183:380–387. doi: 10.1002/(SICI)1096-9896(199712)183:4<380::AID-PATH959>3.0.CO;2-7. [DOI] [PubMed] [Google Scholar]
- 18.Rohrmann S, Giovannucci E, Willett WC, et al. Fruit and vegetable consumption, intake of micronutrients, and benign prostatic hyperplasia in US men. The American journal of clinical nutrition. 2007;85:523–529. doi: 10.1093/ajcn/85.2.523. [DOI] [PubMed] [Google Scholar]
- 19.Barry MJ, Fowler FJ, Oleary MP, et al. The American-Urological-Association Symptom Index for Benign Prostatic Hyperplasia. J Urology. 1992;148:1549–1557. doi: 10.1016/s0022-5347(17)36966-5. [DOI] [PubMed] [Google Scholar]
- 20.Welch G, Kawachi I, Barry MJ, et al. Distinction between symptoms of voiding and filling in benign prostatic hyperplasia: findings from the Health Professionals Follow-up Study. Urology. 1998;51:422–427. doi: 10.1016/s0090-4295(97)00626-2. [DOI] [PubMed] [Google Scholar]
- 21.Barry MJ, Cockett AT, Holtgrewe HL, et al. Relationship of symptoms of prostatism to commonly used physiological and anatomical measures of the severity of benign prostatic hyperplasia. J Urol. 1993;150:351–358. doi: 10.1016/s0022-5347(17)35482-4. [DOI] [PubMed] [Google Scholar]
- 22.Rimm EB, Giovannucci EL, Stampfer MJ, et al. Reproducibility and validity of an expanded self-administered semiquantitative food frequency questionnaire among male health professionals. American journal of epidemiology. 1992;135:1114–1126. doi: 10.1093/oxfordjournals.aje.a116211. discussion 1127-1136. [DOI] [PubMed] [Google Scholar]
- 23.Willett WC, Sampson L, Stampfer MJ, et al. Reproducibility and validity of a semiquantitative food frequency questionnaire. American journal of epidemiology. 1985;122:51–65. doi: 10.1093/oxfordjournals.aje.a114086. [DOI] [PubMed] [Google Scholar]
- 24.Michaud DS, Holick CN, Giovannucci E, et al. Meat intake and bladder cancer risk in 2 prospective cohort studies. Am J Clin Nutr. 2006;84:1177–1183. doi: 10.1093/ajcn/84.5.1177. [DOI] [PubMed] [Google Scholar]
- 25.Michaud DS, Spiegelman D, Clinton SK, et al. Fruit and vegetable intake and incidence of bladder cancer in a male prospective cohort. J Natl Cancer Inst. 1999;91:605–613. doi: 10.1093/jnci/91.7.605. [DOI] [PubMed] [Google Scholar]
- 26.Zhou J, Smith S, Giovannucci E, et al. Reexamination of total fluid intake and bladder cancer in the Health Professionals Follow-up Study Cohort. American journal of epidemiology. 2012;175:696–705. doi: 10.1093/aje/kwr359. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Michaud DS, Spiegelman D, Clinton SK, et al. Fluid intake and the risk of bladder cancer in men. N Engl J Med. 1999;340:1390–1397. doi: 10.1056/NEJM199905063401803. [DOI] [PubMed] [Google Scholar]
- 28.Silverman DT, Alguacil J, Rothman N, et al. Does increased urination frequency protect against bladder cancer? Int J Cancer. 2008;123:1644–1648. doi: 10.1002/ijc.23572. [DOI] [PMC free article] [PubMed] [Google Scholar]