Does increased urination frequency protect against bladder cancer?
Debra T. Silverman1*, Juan Alguacil2, Nathaniel Rothman1, Francisco X. Real3,4, Montserrat Garcia-Closas1,
Kenneth P. Cantor1, Nuria Malats3, Adonina Tardon5, Consol Serra4,6, Reina Garcia-Closas7, Alfredo Carrato8,
Josep Lloreta4,9, Claudine Samanic1, Mustafa Dosemeci1and Manolis Kogevinas3,10,11,12
1DivisionofCancer Epidemiology and Genetics, National Cancer Institute,Department ofHealth and Human Services, Bethesda, MD
2Department of Environmental Biology and Public Health, Universidad de Huelva, Huelva, Spain
3Institut Municipal d’Investigacio Medica, Barcelona, Spain
4Unitat de Recerca en Salut Laboral, Universitat Pompeu Fabra, Barcelona, Spain
5Instituto de Oncologia, Departmento de Medicina, Universidad de Oviedo, Oviedo, Spain
6Unitat de Epidemiologia, Corporacio Sanitaria Parc Taul? ı, Sabadell, Spain
7Hospital Universitario de Canarias, La Laguna, Spain
8Hospital General Universitario de Elche, Elche, Spain
9Hospital del Mar, Barcelona, Spain
10Department of Social Medicine, Medical School, University of Crete, Heraklion, Crete, Greece
11Centre for Research in Environmental Epidemiology, Barcelona, Spain
12CIBER Epidemiology and Public Health, Spain
Experimental studies suggest that increased urination frequency
may reduce bladder cancer risk if carcinogens are present in the
urine. Only 2 small studies of the effect of increased urination fre-
quency on bladder cancer risk in humans have been conducted
with conflicting results. Our purpose was to evaluate the effect of
urination frequency on risk of bladder cancer in a large, multicen-
ter case–control study. We analyzed data based on interviews con-
ducted with 884 patients with newly diagnosed, bladder cancer
and 996 controls from 1998 to 2001 in Spain. We observed a con-
sistent, inverse trend in risk with increasing nighttime voiding fre-
quency in both men (p 5 0.0003) and women (p 5 0.07); voiding
at least 2 times per night was associated with a significant, 40–
50% risk reduction. The protective effect of nocturia was appa-
rent among study participants with low, moderate and high water
consumption. The risk associated with cigarette smoking was
reduced by nocturia. Compared with nonsmokers who did not uri-
nate at night, current smokers who did not urinate at night had an
OR of 7.0 (95% CI 5 4.7–10.2), whereas those who voided at least
twice per night had an OR of 3.3 (95% CI 5 1.9–5.8) (p value for
trend 5 0.0005). Our findings suggest a strong protective effect of
nocturia on bladder cancer risk, providing evidence in humans
that bladder cancer risk is related to the contact time of the uro-
thelium with carcinogens in urine. Increased urination frequency,
coupled with possible dilution of the urine from increased water
intake, may diminish the effect of urinary carcinogens on bladder
' 2008 Wiley-Liss, Inc.
Key words: urination frequency; smoking; water intake; bladder
Experimental evidence suggests that if carcinogens are present
in the urine, frequent urination may reduce bladder cancer risk.
Three lines of limited evidence support this hypothesis. First, dogs
exposed to the bladder carcinogen 2-napthylamine who have
undergone surgical diversion of the urine do not develop tumors in
the part of the bladder with no urine contact.1Second, dogs
administered the bladder carcinogen 4-aminobiphenyl and cathe-
terized to regulate urination frequency have decreased levels of
urothelial DNA adducts with increased urination frequency.2
Third, urine appears to be a promoter of bladder carcinogenesis in
rat.3,4A protective effect of increased urinary frequency is further
supported by the results of 2 observational studies. In high-risk
areas for bladder cancer in Israel, infrequent micturition and high
urine concentration were more prevalent than in low-risk areas.5
In Serbia, a significant protective effect was associated with
increased urination frequency based on 130 cases and 130 controls
in a hospital-based case–control study.6This study had a number
of limitations, however, including small sample size and minimal
information on methods used to determine urination frequency
Our purpose was to evaluate the effect of urination frequency
on risk of urothelial (transitional cell) carcinoma of the bladder in
a large, multicenter case–control study of bladder cancer.
Material and methods
We conducted a hospital-based case–control study in 5 areas of
Spain. The case series included all patients with histologically
confirmed carcinoma of the urinary bladder or carcinoma in situ
(International Classification of Diseases 9 5 1880–1889; 2,337)
newly diagnosed in 1998–2001 who were 21–80 years of age at
the time of diagnosis/interview and resided in the catchment areas
of 18 hospitals in 5 areas of Spain (i.e., Barcelona, Valles/Bages,
Asturias, Alicante and Tenerife). Patients with a previous diagno-
sis of cancer of the lower urinary tract (i.e., bladder, renal pelvis,
ureters or urethra) were not eligible for study. Diagnostic slides
for each patient were reviewed by a panel of expert pathologists to
confirm the diagnosis of bladder cancer and ensure uniformity of
classification criteria based on the 1998 World Health Organiza-
tion/International Society of Urological Pathology system7as well
as the histologic subtype. Of the 1,453 patients found to be eligible
for study, we interviewed 1,219 (84%).
For each eligible case, we selected 1 control matched to the
case on age at diagnosis/interview (within 5 years), gender, race/
ethnicity and hospital. Controls were identified from patients
admitted to the same hospital as the case for diseases/conditions
unrelated to the exposures under investigation including urination
frequency. Reasons for hospital admission were: 36% hernias,
12% other abdominal surgery, 24% fractures, 11% hydrocele, 6%
orthopedic conditions, 4% circulatory diseases, 2% dermatological
diseases, 1% ophthalmological diseases and 4% other diagnoses).
Of the 1,442 controls found to be eligible for study, we inter-
viewed 1,271 (88%).
Juan Alguacil: Formerly with the Division of Cancer Epidemiology and
Genetics, National Cancer Institute, Bethesda, MD.
F.X. Real and N. Maltas’s current address is: Programa de Patologia
Molecular, Centro Nacional de Investigaciones Oncologicas, Spain.
Grant sponsor: Intramural Research Program of the NIH, National
Cancer Institute, Division of Cancer Epidemiology and Genetics; Grant
*Correspondence to: National Cancer Institute, 6120 Executive Boule-
vard, Room 8012, Bethesda, MD 20892, USA. Fax: (301) 402-1819.
Received 7 November 2007; Accepted after revision 29 January 2008
Published online 11 July 2008 in Wiley InterScience (www.interscience.
Int. J. Cancer: 123, 1644–1648 (2008)
' 2008 Wiley-Liss, Inc.
Publication of the International Union Against Cancer
Subjects were interviewed in the hospital by trained inter-
viewers using computer-assisted personal interviews (CAPIs).
Before the interview, written informed consent to participate in
the study was obtained from each subject. The CAPI was designed
to elicit detailed information on past usual adult urination fre-
quency during 5-time periods in a typical day: before breakfast,
between breakfast and lunch, between lunch and dinner, between
dinner and bedtime, and the number of times per night that the
subject woke to urinate. Based on this information, we computed
3 exposure metrics: 24-hr urination frequency, daytime urination
frequency and nighttime urination frequency. Among controls, the
median 24 hr and nighttime frequencies were 6 and 0, respec-
tively, for both men and women, which is comparable to urination
frequencies observed in the general population.8,9
The CAPI also obtained information on average adult fluid con-
sumption (instituted after the first year of data collection), smok-
ing habits, dietary factors, medical conditions (e.g., benign pros-
tatic hypertrophy (BPH) and urinary tract infections), occupational
and residential histories, family history of cancer and history of
medication use (i.e., analgesics and nonsteroidal anti-inflamma-
tory drugs). The effects of smoking have been described in detail
elsewhere.10,11This study was approved by the Institutional
Review Board of the National Cancer Institute as well as the
Ethics Committees of all participating hospitals.
The effects of urination frequency on bladder cancer risk were
quantified by the odds ratio (OR). ORs and 95% confidence inter-
vals (CIs) were estimated by unconditional logistic regression.
Final statistical models included terms for exposure (i.e., 24-hr
urination frequency, daytime urination frequency or nighttime uri-
nation frequency), the matching factors (i.e., age at diagnosis/
interview, gender, race/ethnicity and hospital/region), as well as
terms for potential confounding factors (i.e., cigarette smoking,
tap water intake and history of BPH). We selected ‘‘tap water
intake,’’ rather than ‘‘total fluid intake,’’ for inclusion in the final
models because only ‘‘tap water intake’’ was associated with a sig-
nificant reduction in risk in our study; intake of other beverages
appeared unrelated to risk.12Other potential confounders, such as
employment in high-risk occupations for bladder cancer, use of
nonsteroidal anti-inflammatory drugs, consumption of water con-
taining elevated levels of disinfection by-products and consump-
tion of fruits and vegetables, were not included in the final models
because they had little or no impact on estimates of risk. To test
for trend, the median value for each level of the categorical vari-
able among the controls was entered into the model. To test for
interaction between 2 risk factors, we added a cross-product term
to the logistic model and conducted a likelihood ratio test.
Interviewed subjects were excluded from analysis for the fol-
lowing reasons: nonurothelial carcinoma (16 cases) and nonwhite
(5 cases and 1 control). Three hundred fourteen cases and 274 con-
trols received an abridged version of the questionnaire that did not
include detailed information regarding urination frequency (i.e.,
only categorical data on 24-hr urination frequency are available),
water intake and BPH in men. Although these data were not
included in Tables I–III, risks based on subjects with abridged
data were compared to those based on the unabridged question-
naire in the Results section pertaining to Table I. The primary
analysis of urination frequency presented in this work was based
on unabridged interviews with 884 cases (764 men and 120
women) with newly diagnosed, urothelial carcinoma of the blad-
der and 996 controls (850 men and 146 women) with detailed in-
formation on urination frequency.
Table I shows the risk of bladder cancer for 3 urination fre-
quency metrics: nighttime urination frequency, daytime urination
frequency and 24-hr urination frequency. For nighttime urination
frequency, a consistent, inverse trend in risk was apparent with
increasing voiding among both men and women (p-value for trend:
0.0003 in men; 0.07 in women) (Table I). Men who voided at least
twice per night experienced a significant, 40% reduction in risk and
women experienced a marginally significant, 50% reduction in risk.
Among both men and women, no significant trend in risk was
apparent with either increasing daytime urination frequency or
with increasing 24-hr urination frequency, although a nonsignifi-
cant, decreased risk was found among subjects who voided at least
9 times perday duringthe daytime (men:OR 50.6, 95% CI 50.3–
1.1; women: OR 5 0.8, 95% CI 5 0.2–4.3).For 24-hr urination fre-
quency, the inverse trend in risk with increasing urination fre-
tionnaires was not significantly different from that based on partici-
pants withunabridged questionnaires showninTable I (p 50.26).
All risk estimates in Table I were adjusted for potential
confounding from cigarette smoking, age at diagnosis/interview,
TABLE I – ODDS RATIOS AND 95% CONFIDENCE INTERVALS ASSOCIATED WITH FREQUENCY OF URINATION
IN 24 HOURS, DAYTIME AND NIGHTTIME, BY GENDER (884 CASES AND 996 CONTROLS)
95% CICasesControls OR1
p 5 0.00034
p 5 0.074
p 5 0.264
No. voids/24 hr3
p 5 0.294
p 5 0.774
1Adjusted for age, geographic region, cigarette smoking, water intake (24 hr and daytime) and BPH
(men).–3Five controls were excluded because of missing information on no. voids/nighttime.–3No. of voids/
daytime and no. of voids/24 hr could not be computed for 533 cases and 396 controls who did not report
no. of voids for some part of the day or did not have information on water intake.–4p-value for trend.
URINATION FREQUENCY AND BLADDER CANCER
hospital/region and BPH (men). In addition, total daily water
intake was included in the models for 24-hr urination frequency
and daytime urination frequency because it was related to these
exposure metrics as well as to risk. Among men, ORs for water
intake were 1.0 for ?400 ml/day; 0.6 (95% CI 5 0.5–0.9) for
>400–1399 ml/day; 0.5 (95% CI 5 0.3–0.7) for >1399 ml/day
(p value for trend <0.0001). Among women, ORs for comparable
intakes were: 1.0 (referent); 1.6 (95% CI 5 0.7–3.7); 0.6 (95%
CI 5 0.2–1.7) (p-value for trend 5 0.19). Total amount of daily
water intake was, however, unrelated to nighttime urination fre-
quency. Among controls who drank 400 ml per day or less, 60%
did not void at night, 20% voided once per night and 20% voided
at least twice per night; among controls who drank 401–1399 ml
per day, 55% did not void at night, 25% voided once per night and
20% voided at least twice per night; among controls who drank at
least 1,400 ml per day, 53% did not void at night, 24% voided
once per night and 23% voided at least twice per night. Adjust-
ment for total daily water intake had little or no impact on risk
estimates for nighttime urination frequency, and thus, water intake
was excluded from the final models to maximize statistical power.
Among men, further adjustment for potential confounding from
BPH had little impact on estimates of risk. The protective effect of
nocturia was apparent in men with no history of BPH as well as
among women. Among men reporting no history of BPH, ORs
were 1.0 for 0 voids/night; 0.8 (95% CI 5 0.6–1.0) for 1 void/
night; 0.6 (95% CI 5 0.5–0.9) for 2 or more voids/night.
We found no evidence of modification of the effect of nocturia
by age at diagnosis/interview (p-value for interaction 5 0.76 in
men and 0.74 in women). OR for 2 or more voids/night were simi-
lar among participants who were aged 30–64 years and those aged
65–80 years (men: 0.5 (95% CI 5 0.3–0.9) and 0.6 (95% CI 5
0.4–0.9), respectively; women: 0.3 (95% CI 5 0.1–1.5) and 0.6
(95% CI 5 0.3–1.4), respectively.
We examined prevalence of nocturia among controls by reason
for hospitalization to determine if any diagnostic category was
related to nocturia. The prevalence of voiding at least once per
night by main diagnostic category was as follows: 46% for hernia;
36% for other abdominal surgery; 47% for fractures; 46% for hy-
drocele; and 42% for other less common diagnoses. Thus, we
found no evidence of usually high prevalence of nocturia by diag-
nostic category, which would have led to the underestimation of
the effect of nocturia.
The effect of nighttime voiding is stratified by the level of water
intake in Table II. Because questions on water intake were added
after the first year of data collection, Table II was limited to the
379 cases and 647 controls for whom data on water intake and
nighttime voiding were available. The reduced risk associated
with nocturia is apparent in subjects with low, moderate and high
water consumption. Relative to those with low water intake who
did not urinate at night, those with low water intake who urinated
at least twice per night had an OR of 0.6 (95% CI 5 0.3–1.2) and
an OR of 0.2 (95% CI 5 0.1–0.4) for those with high water intake.
Similar protective effects were apparent when nighttime voiding
was stratified by total intake of all fluids made with tapwater (e.g.,
coffee and tea). The protective effect of water intake was also
apparent within each level of nighttime urination frequency. A
strong protective effect was observed when both factors acted to-
gether. Subjects who urinated at least twice per night and drank at
least 1,400 ml of water per day experienced a significant, 80%
reduction in risk compared to risk among those who did not void
at night and drank less than 400 ml of water per day. However,
there was no evidence of statistical interaction between nocturia
and water intake (p 5 0.72).
Table III shows the effect of nighttime voiding cross classified
by smoking status. The reduced risk associated with increased
nighttime voiding frequency was apparent among nonsmokers,
former smokers and current smokers. Moreover, nocturia had a
powerful effect on the bladder cancer risk associated with smoking.
Current smokers who did not urinate at night had an OR of 7.0
(95% CI 5 4.7–10.2), whereas current smokers who voided once
per night had an OR of 4.0 (95% CI 5 2.5–6.5) and those who
voided at least twice per night had an OR of 3.3 (95% CI 5 1.9–
5.8) (p-value for trend 5 0.0005). Former smokers who voided at
least twice per night also experienced a reduced risk (OR 5 2.3,
95% CI 5 1.4–3.7) compared to risk among those who did not void
at night (OR 5 3.4, 95% CI 5 2.3–4.9). Among current smokers,
risk decreased with 1 void per night, whereas the risk reduction in
former smokers and nonsmokers occurred with at least 2 voids per
night. However, no statistical interaction between nighttime void-
ingfrequency and smoking statuswas apparent(p 50.30).
In summary, we observed a consistent, inverse trend in bladder
cancer risk with increasing number of voids per night among both
men and women, which was statistically significant in men but not
in women. Men and women who urinated at least twice per night
experienced a significant 40–50% reduction in risk compared to
those who did not urinate at night. The protective effect of noctu-
ria was apparent among study participants with low, moderate and
high water intake, as well as among nonsmokers, former smokers
and current smokers. It is noteworthy that the risk associated with
cigarette smoking was reduced more than 50% by nocturia.
In the mid-1970s, the urogenous-contact hypothesis was first
introduced, suggesting that ‘‘contact with urine loaded with carci-
nogens is the main important factor in the etiology of carcinoma
of the bladder.’’13–15It is surprising that this observation has been
followed up in only 2 epidemiologic studies. In 1982, Mommsen
et al.16conducted a small case–control study in men in Denmark.
Based on interviews with 165 bladder cancer patients and 165
population controls, they observed that nocturia (defined as noc-
turnal urination twice or more at least 2 years before presenting
symptoms of bladder cancer) was associated with an increased
risk of bladder cancer (relative risk 5 2.05; 95% CI 5 1.27–3.32).
No information on fluid intake was collected, however. In con-
trast, Radosavljevic et al.6reported a significant protective effect
(OR 5 0.27; 95% CI 5 0.18–0.41) for increased urination fre-
TABLE III – ODDS RATIOS1AND 95% CONFIDENCE INTERVALS
BY CIGARETTE SMOKING AND NIGHTTIME VOIDING (865 CASES
AND 950 CONTROLS)2
No. of voids/night
NonsmokersFormer smokersCurrent smokers3
1Adjusted for age, gender, geographic region and BPH (men).–2Ni-
neteen cases and 46 controls were excluded because they were ‘‘occa-
sional smokers.’’–3p-value for trend 5 0.0005.–4Cases/controls.
TABLE II – ODDS RATIOS1AND 95% CONFIDENCE INTERVALS BY WATER
INTAKE AND NIGHTTIME VOIDING (379 CASES AND 647 CONTROLS)2
No. of voids/night
Water intake (ml/day)
Low (?400 ml)
Moderate (401–1,399 ml) High (?1400 ml)
1Adjusted for age, gender, geographic region, cigarette smoking and
BPH (men).–2Five hundred and five cases and 345 controls are excluded
because they did not report information on water intake.–3Cases/con-
SILVERMAN ET AL.
quency after adjustment for fluid intake, smoking and other con-
founding factors, based on interviews with 130 cases and 130 con-
trols in Serbia. Both studies were limited by small numbers and
did not provide detailed information on urination frequency.
The protective effect in our study was primarily associated with
nighttime voiding. As the longest time period in 24 hr when no
voiding typically occurs is nighttime, the protective effect of noctu-
ria may be more pronounced than protection imparted by increased
urination during the waking hours. The lack of a consistent protec-
recall of urination frequency during waking hours compared to that
during sleeping hours. In fact, only 40% of cases and 60% of con-
trols who reported nighttime urination frequency were able to pro-
vide complete information on daytime urination frequency.
The observed protective effect of nocturia appeared unrelated to
the presence of BPH. We observed a reduced risk among those
who voided at least twice per night in men without BPH and
among women. The protective effect of nocturia was also unre-
lated to total daily water intake. We did not, however, obtain in-
formation on timing of water intake so we were unable to evaluate
the effect of evening water intake on nighttime frequency. We
observed a significant, inverse trend in risk with increasing water
intake. Heavy consumers of water (at least 1,400 ml per day)
experienced about 50% reduced risk of bladder cancer,17a finding
consistent with that reported by Michaud et al.17in the Health Pro-
fessionals Follow-up Study. However, the protective effect of noc-
turia in our study was independent of water intake, appearing
among low, moderate and heavy consumers of water. Although no
evidence of interaction between nighttime urination frequency and
water intake was observed, the protective effect of both factors
acting jointly was profound. Subjects who voided at least twice
per night and drank at least 1,400 ml per day experienced a signifi-
cant 80% reduced risk of bladder cancer compared to those who
did not void at night and drank less than 400 ml of water per day.
Increased urination frequency may lessen contact time that any
carcinogens present in the urine have with the urothelium of the
bladder. Because recent data suggest that urine composition is
modified in the lower urinary tract and that the urothelium has a
significant transport function,18the urothelium may have the ability
to reabsorb carcinogens from the urine and decreasing contact time
should result in reduced risk. Increased water intake leads to pro-
duction oflessconcentrated urineaswell asincreased urination fre-
quency, which may also result inreduced bladder cancer risk.
Our study had limitations. First, although our study is the larg-
est and most comprehensive investigation of the effect of urination
frequency on bladder cancer risk, we did not obtain detailed infor-
mation of urination frequency for all study participants. Our analy-
sis was based on interviews with the 884 cases and 996 controls
with detailed information on urination frequency. We did have
limited information (i.e., total number of voids in 24 hr) from par-
ticipants who were asked the abridged questionnaire. We observed
no significant differences in the effect of 24-hr urination frequency
when we compared participants with abridged questionnaires to
those with unbridged questionnaires. In addition, although sub-
jects were queried about their usual adult urination frequency, this
information was obtained after diagnosis of bladder cancer, and it
is not possible to evaluate the possible effect of differential recall
bias. We think it unlikely, however, that bladder cancer patients
would underestimate information on urination frequency.
In conclusion, our findings suggest strong protective effects of
both nocturia and increased water intake on bladder cancer risk. In
particular, the protective effect of nocturia on the risk associated
with current smoking among men in our study was noteworthy.
Current smokers who did not void at night had a 7.0-fold risk that
was more than cut in half to a 3.3-fold risk among current smokers
who voided at least twice per night. This observation suggests that
nocturia may be a powerful factor in reducing bladder cancer risk.
This finding also provides evidence in humans that bladder cancer
risk is related to the contact time of the urothelium with carcino-
gens in urine. If confirmed, innovative approaches will be needed
to translate this finding into meaningful prevention of the occur-
rence of the nearly 357,000 bladder cancer patients newly diag-
nosed worldwide annually and the 145,000 deaths.19,20
Dr. Debra Silverman had full access to all the data in the study
and takes responsibility for the integrity of the data and the accu-
racy of the analysis. Acquisition of data was done by Kogevinas,
Real, Malats, Tardon, Serra, R. Garcia-Closas, Carrato, Lloreta.
Analysis and interpretation of data was done by Silverman, Algua-
cil, Kogevinas, Rothman, Real, M. Garcia-Closas, Cantor, Malats,
Lloreta, Samanic. Drafting, revising, and final approval of manu-
script by Silverman, Alguacil, Rothman, Real, M. Garcia-Closas,
Cantor, Malats, Tardon, Serra, R. Garcia-Closas, Carrato, Lloreta,
Samanic, Dosemeci, Kogevinas.
1.McDonald DF, Lund RR. The role of the urine in vesical neoplasm. I.
Experimental confirmation of the urogenous theory of pathogenesis.
J Urol 1954;71:560–70.
Kadlubar FF, Dooley KL, Teitel CH, Roberts DW, Benson RW,
Butler MA, Bailey JR, Young JF, Skipper PW, Tannenbaum SR.
Frequency of urination and its effects on metabolism, pharmacoki-
netics, blood hemoglobin adduct formation, and liver and urinary
bladder DNA adduct levels in beagle dogs given the carcinogen
4-aminobiphenyl. Cancer Res 1991;51:4371–7.
Oyasu R, Hirao Y, Izumi K. Enhancement by urine of urinary bladder
carcinogenesis. Cancer Res 1981;41:478–81.
Rowland RG, Henneberry MO, Oyasu R, Grayhack JT. Effects of
urine and continued exposure to carcinogen on progression of early
neoplastic urinary bladder lesions. Cancer Res 1980;40:4524–7.
Braver DJ, Modan M, Chetrit A, Lusky A, Braf Z. Drinking, micturi-
tion habits, and urine concentration as potential risk factors in urinary
bladder cancer. J Natl Cancer Inst 1987;78:437–40.
Radosavljevic V, Jankovic S, Marinkovic J, Djokic M. Fluid intake
and bladder cancer. A case control study. Neoplasma 2003;50:
Epstein JI, Amin MB, Reuter VR, Mostofi FK. The World Health Or-
ganization/International Society of Urological Pathology consensus
classification of urothelial (transitional cell) neoplasms of the urinary
bladder. Bladder Consensus Conference Committee. Am J Surg
Larsson G, Victor A. Micturition patterns in a healthy female popula-
tion, studied with a frequency/volume chart. Scand J Urol Nephrol
9. Blanker MH, Bohnen AM, Groeneveld FPMJ, Bernsen RMD, Prins
AD, Ruud Bosch JLH. Normal voiding patterns and determinants of
increased diurnal and nocturnal voiding frequency in elderly men.
J Urol 2000;164:1201–5.
10. Garc? ıa-Closas M, Malats N, Silverman D, Dosemeci M, Kogevinas
M, Hein DW, Tardon A, Serra C, Carrato A, Garcia-Closas R, Lloreta
J, Castano-Vinyals G, et al. NAT2 slow acetylation and GSTM1 null
genotypes increase bladder cancer risk: results from the Spanish Blad-
der Cancer Study and meta-analyses. Lancet 2005;366:649–59.
11. Samanic C, Kogevinas M, Dosemeci M, Malats N, Real FX, Garcia-
Closas R, Sala M, Lloreta J, Tardon A, Rothman N, Silverman D.
Smoking and bladder cancer in Spain: effect of tobacco type, timing,
environmental tobacco smoke, and gender. Cancer Epidemiol Bio-
markers Prev 2006;15:1348–54.
12. Michaud DS, Kogevinas M, Cantor KP, Villanueva CM, Garcia-Clo-
sas M, Rothman N, Malats N, Real FX, Serra C, Garcia-Closas R,
Tardon A, Carrato A, et al. Total fluid and water consumption and the
joint effect of exposure to disinfection by-products on risk of bladder
cancer. Environ Health Perspect 2007;115:1569–72.
13. Oyasu R, Hopp ML. The etiology of cancer of the bladder. Surg.
Gynecol Obstet 1974;138:97–108.
14. Melicow MM. Tumors of the bladder: a multifaceted problem. J Urol
15. Parkash O, Kiesswetter H. The role of urine in the etiology of cancer
of the urinary bladder. Urol Int 1976;31:343–8.
16. Mommsen S, Aagaard J, Sell A. An epidemiological case-control
study of bladder cancer in males from a predominantly rural district.
Eur J Cancer Clin Oncol 1982;18:1205–10.
URINATION FREQUENCY AND BLADDER CANCER
17. Michaud DS, Spiegelman D, Clinton SK, Rimm EB, Curhan GC, Wil-
lett WC, Giovannucci EL. Fluid intake and the risk of bladder cancer
in men. N Engl J Med 1999;340:1390–7.
18. Cahill DJ, Fry CH, Foxall PJ. Variation in urine composition in the
human urinary tract: evidence of urothelial function in situ? J Urol
19. Silverman DT, Devesa SS, Moore LE, Rothman N. Bladder cancer.
In: Schottenfeld D, Fraumeni JF Jr, eds. Cancer epidemiology and
prevention, 3rd edn. New York: Oxford University Press, 2006.
20. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002.
CA Cancer J Clin 2005;55:74–108.
Participating study centers in Spain
Institut Municipal d’Investigaci? o Me `dica, Universitat Pompeu
Fabra, Barcelona-Coordinating Center (M. Kogevinas, N. Malats,
F.X. Real, M. Sala, G. Casta~ no, M. Tora `, D. Puente, C. Villa-
nueva, C. Murta, J. Fortuny, E. L? opez, S. Hern? andez, R. Jara-
millo); Hospital del Mar, Universitat Auto `noma de Barcelona,
Barcelona (J. Lloreta, S. Serrano, L. Ferrer, A. Gelabert, J. Carles,
O. Bielsa, K. Villadiego), Hospital Germans Tries i Pujol, Bada-
lona, Barcelona (L. Cecchini, J.M. Saladi? e, L. Ibarz); Hospital de
Sant Boi, Sant Boi, Barcelona (M. C? espedes); Centre Hospitalari
Parc Taul? ı, Sabadell, Barcelona (C. Serra, D. Garc? ıa, J. Pujadas,
R. Hernando, A. Cabezuelo, C. Abad, A. Prera, J. Prat); Centre
Hospitalari i Cardiolo `gic, Manresa, Barcelona (M. Dome `nech, J.
Badal, J. Malet); Hospital Universitario, La Laguna, Tenerife (R.
Garc? ıa-Closas, J. Rodr? ıguez de Vera, A.I. Mart? ın); Hospital La
Candelaria, Santa Cruz, Tenerife (J. Ta~ no, F. C? aceres); Hospital
General Universitario de Elche, Universidad Miguel Hern? andez,
Elche, Alicante (A. Carrato, F. Garc? ıa-L? opez, M. Ull, A. Teruel,
E. Andrada, A. Bustos, A. Castillejo, J.L. Soto); Universidad de
Oviedo, Oviedo, Asturias (A. Tard? on); Hospital San Agust? ın,
Avil? es, Asturias (J.L. Guate, J.M. Lanzas, J. Velasco); Hospital
Central Covadonga, Oviedo, Asturias (J.M. Fern? andez, J.J.
Rodr? ıguez, A. Herrero), Hospital Central General, Oviedo, Astu-
rias (R. Abascal, C. Manzano, T. Miralles); Hospital de Cabue~ nes,
Gij? on, Asturias (M. Rivas, M. Arguelles); Hospital de Jove, Gij? on,
Asturias (M. D? ıaz, J. S? anchez, O. Gonz? alez); Hospital de Cruz
Roja, Gij? on, Asturias (A. Mateos, V. Frade); Hospital Alvarez-
Buylla (Mieres, Asturias): P. Munta~ nola, C. Pravia; Hospital Jar-
rio, Coa~ na, Asturias (A.M. Huescar, F. Huergo); Hospital Carmen
y Severo Ochoa, Cangas, Asturias (J. Mosquera).
SILVERMAN ET AL.