- Email: [email protected]
Adult Female Urinary Incontinence And Childhood Bedwetting
-:-:-:-.---. E
- .
• -
'-•:.--~~---:---.• • -
.-4
• - - _"':,.
0022-534 7/94/1521-0085$03.00/0 Vol. 152, 85-88, July 1994
THE JOURNAL OF UROLOGY Copyright © 1994 by AMERICAN UROLOGICAL ASSOCIATION, INC .
Printed in U.S.A.
ADULT FEMALE URINARY INCONTINENCE AND CHILDHOOD BED WETTING ANDERS FOLDSPANG
AND
S0REN MOMMSEN
From the Institute of Epidemiology and Social Medicine, Aarhus University, and Department of Urology, Aarhus University Hospital, Aarhus, and Department of Urology, Odense University Hospital, Odense, Denmark
ABSTRACT
A cross-sectional random population sample of women 30 to 59 years old was sent a questionnaire on urinary incontinence and, among other things, childhood bedwetting. Among 2,613 responders 17 .0% reported prevalent urinary incontinence (14. 7% stress provoked, 8.3% associated with urge, 6.8% stress and urge overlap, 2.2% occurring especially during sleep and 3.9% occurring especially when anxious), and 6.5% reported childhood bedwetting after age 5 years and 3.3% after age 10 years. Childhood bedwetting was associated with prevalent urge urinary incontinence (p <0.01) and incontinence occurring during sleep (p <0.0001) but was less marked with urinary incontinence occurring in situations of anxiety (p <0.025). It is concluded that some individuals suffer a long-lasting disturbance of the balance between micturition and sleep processes. KEY WORDS:
urinary incontinence, enuresis, population, urination disorders
Childhood bedwetting is frequent in Western industrialized populations but less prevalent in Europe than in other areas of the world. 1- 3 The occurrence of childhood bedwetting has been related to genetic, 4 ethnic,5 physical, 6 - 8 mental 6 • 3 - 10 and social5 • 6 • 8 • 11 factors. Some physicians have investigated more specific urological aspects,7 • 12 and the role of the sleep and dream pattern.7 • 11 • 13 Based on a multifactorial etiology, 14 - 19 urinary incontinence also is prevalent in adult female populations. 20 The similarity of the nature of childhood bedwetting and urinary incontinence suggests the possibility of a mutual relationship to be a straightforward object for research interest, which might shed further light on the nature of either condition. Moreover, the number of studies on the topic is extremely low, for example among physicians concerned with the prevalence of urinary incontinence in women only Yarnell et al investigated childhood bedwetting also.21 Conditional on bed wetting after the ages of 5 and 9 years among 990 women 18 years old and older, they found moderately increased relative risks of having urinary incontinence (relative risk 1.3, p -0.05, and relative risk 1.4, p -0.05, respectively). In 500 women with urological conditions Moore found a strong association between present idiopathic detrusor instability and the patient who reported bedwetting in childhood (relative risk 2.7, p <0.0001), 22 We have been unable to identify any childhood cohorts who were observed regarding the occurrence of urinary incontinence in adult life. We presently investigate the possible association between childhood bedwetting and urinary incontinence in women. As contrasted to stress urinary incontinence, we previously found urge incontinence to occur relatively often during sleep and in situations of anxiety.23 Considering this finding, and the urodynamic, psychological and social correlates of childhood bedwetting alluded to, we hypothesized a priori bedwetting to be associated primarily with urge urinary incontinence and, furthermore, with urge incontinence occurring during sleep and in situations of anxiety.
MATERIALS AND METHODS
The applied cross-sectional design and its data have been presented previously.2 0 During January and Febr uary 1988, the members of an age-stratified random sample of 3,114 women 30 to 59 years old and living in the Municipality of Aarhus (total female population of the same age 47,702) were mailed a self-administered postal questionnaire. Nonrespondents were addressed 2 additional times. Finally, 2,631 subjects (84 .5%) responded, with a slight but significant reduction by age (odds ratio 0.93 per 5-year group, p <0.025 [multiple logistic regression estimate]). The present analysis comprises 2,613 women who reported their age at investigation. Among other things, the questionnaire asked about episodes of urinary incontinence during 1987. If present, were urinary incontinence episodes generally provoked by physical stress, for example lifting, coughing, sneezing or laughing (stress incontinence), or accompanied by a feeling of urge to void (urge incontinence) or both (mixed incontinence). If the woman reported urinary incontinence, did it occur especially during sleep or when she felt nervous. Had t he woman experienced childhood bed wetting and, if so, at what age was she able to control micturition during sleep. Additional information was collected on menopausal status, childbirth, surgical experience, cystitis, and body height (cm.) and weight (kg.). Prevalence (%) of reports of childhood bedwetting by age at investigation in 2,613 women 30 to 59 years old, Aarhus, Denmark, 1988
TABLE 1.
Bedwetting Cessation Age(%) Age (yrs.)
30-34 35- 39 40-44 45-49 50- 54 55- 59
Total No.
415 420 508 366 445 459 2,613
6-18 Yrs.
6- 10 Yrs.
11-18 Yrs.
8.9 6.9 8.3 4.6 4.9 4.8 6.5 0.87t
3.9 4.3 4.1 2.5 2.0 2.2 3.2 0.85t
5.1 2.6 4.1 2.2 2.9 2.6 3.3 0.89§
Totals Odds ratio* * Odds ratio indicates the change in prevalence per 5 years of subject age at investigation (logistic regression estimate). t p < 0.005. :j: p > 0.025. § p < 0.10.
Accepted for publication December 3, 1993. Su pported by Grant 12-7621, 1987 from the Danish Medical Research Council, the Health Care Division of the Miilnlycke Co., the KabiVitrum Co., the Danish Urological Society (the Rahbeck Foundation) and the Research Fund of the Aarhus County Division of the Danish Medical Association. 85
~
86
-
ADULT FEMALE URINARY INCONTINENCE AND CHILDHOOD BEDWETTING TABLE 3 . Prevalence of urinary incontinence occurring especially during sleep, by age at cessation of childhood bedwetting and urge incontinence in 2,613 women 30 to 59 years old, Aarhus, Denmark, 1987
Prevalence(%} of bedwetting 10
8
Age at Cessation of Childhood Bedwetting
-All
6
- · Present age 30-39 ···Present age 40-49 4
Urge Urinary Incontinence
- Present age 50-59
7
8
9 10 11 12 13 14 15 16 17 18 19 Age (years)
Fm. 1. Survival analysis of childhood bedwetting in 2,613 women 30 to 59 years old, Aarhus, Denmark, 1988.
Body mass index is defined as the body weight (kg.) divided by the squared body length (m.). In the text the term "prevalence" is used synonymously with "period prevalence 1987". The statistical analysis applies survival analysis, Spearman's correlation coefficient r , Pearson's chi-square test, Fisher's exact test, Mantel-Haenszel's stratified analysis and multiple logistic regression. 24 A p value of 0.05 was used as a general level of significance. Logistic regression was performed by use of a combination of forward selection and backward elimination. Besides an odds ratio regression point estimate and a 95% interval estimate (in parentheses), each predictor was assigned an independent p value based on the chi-square distributed goodness of fit likelihood ratio statistic. The statistical analysis was performed by 1 of us (A. F.). RESULTS
A total of 169 women (6.5%) reported childhood bedwetting after age 5 years (table 1). Of the women 83 (3.2%) and 86 (3.3%) had ceased bedwetting during ages 6 to 10 and 11 to 18 years, respectively. The proportion of women who reported childhood bedwetting decreased significantly by increasing present age. Moreover, this trend was evident regarding childhood bedwetting reported to have ceased before the age of 11 years, whereas bedwetting later in childhood was reported with equal frequency irrespective of age at investigation (fig. 1). Overall, urinary incontinence was reported by 17.2% of the subjects (stress incontinence by 14.8% and urge incontinence by 8.6%). Most women (82.7%) with urge incontinence also had stress incontinence. In 22 women (0.8%) urinary incontinence was neither provoked by stress nor associated with
T ABLE 2 .
Odds Ratio*
25 (11.8) 9.9t 31 (1.3) 11-18 7 (50.0) 17.0t 72 4 (5.6) Totals 2,6 13 67 (2.6) 10.8t,§ * Odds ratio indicates the relative risk of incontinence occurring during sleep conditional on clinical type (urge) of incontinence. t p < 0.00001. t p < 0.0002. § Weight ed Mantel-Haenszel estimate indicating the association of prevalent urge incontinence to incontinence occurring especially during sleep, corrected for age at cessation of childhood bedwetting. 0--10
2
0 6
Total No.
Urinary Incont inence Occurring During Sleep No. (%)
Yes No Yes No
211 2,316 14
urge (unspecific incontinence). Incontinence occurred especially during sleep and 1n situations of a nxiety in 2.6% a nd 4.1% of the women, respectively. Women who reported childhood bedwetting had a high prevalence of urinary incontinence associated with urge, and occurring during sleep and in situations of anxiety (table 2). The symptoms of urge and experiencing incontinence during sleep were associated with childhood bedwetting having ceased late in childhood, while urinary incontinence in anxious situations was correlated with earlier cessation of bedwetting. When corrected for childbirth, curettage, cystitis experience and body mass index, the pattern of association presented in table 2 remained the same, and the risk estimates changed by ±0.2 units at most. The crucial relationship among age at cessation of childhood bedwetting, urge incontinence and incontinence occurring during sleep is illustrated in table 3. Among women reporting childhood bedwetting, only urinary incontinence during sleep was associated with linear age at cessation ofbedwetting (odds ratio 1.2 per year of age, p < 0.05). Unspecific urinary incontinence was associated with incontinence during sleep (p < 0.00001) but unassociated with childhood bedwetting. Former bedwetters did not report a longer mean interval since the first adult incontinence episode than did other women. DISCUSSION
The reliability of our dat a depends on the ability and willingness of study participants to report validly. Because of resource limitations this questionnaire was not validated itself but, using clinical evaluation as a reference, others h ave found questionnaire-based information on prevalent
Prevalence (%) of urinary incontinence by clinical type, situation of occurrence and age at cessation of childhood bedwetting in 2,613 women 30 to 59 years old, Aarhus, Denmark, 1988 Urinary Incontinence
Age at Cessation of Bedwetting (yrs.)
All Pts. %
Odds Ratio*
All Stress %
Odds Ratio*
All Urge %
Odds Ratio*
Mixed %
Odds Ratio*
During Sleep %
Odds Ratio*
When Nervous %
Odds Ratio*
6- 10 19.3 1.2 18.1 1.4 10.8 1.4 9.6 1.5 3.6 1.5 8.4 2.6t 11- 18 19.8 1.1 16.3 12.8 2.0§ 12.8 5.8 1.6 1.2 15.1 2.2t 6.711 6- 18 19.5 1.2 16.6 1.2 13.6 11.2 8.3 7.1 2. l t l.8t l.8t 4.311 Nonbedwetters 14.7 8.3 6.8 2.2 3.9 17.0 * Odds ratio, standardized for age (years) at investigation, indicates th e urinary incontinence prevalence relative risk as compared to that of nonbedwetters and women h aving ceased bedwetting during other ages. t p < 0.025. t p < 0.01. § p < 0.05. 11 p < 0.0001.
ADULT FEMALE URINARY INCONTINENCE AND CHILDHOOD BEDWETTING Prevalence(%) of bedwetting
iO
- - - - -~
8
6
- Literature - · Present study
4 2
06
7
8
9 10 11 12 13 14 15 16 17 18 19
Age (year s)
FIG. 2. Logistic regression based estimates of prevalence of bedwetting by age in previously published studies (total number of children examined 3, 786)7 • 25 · 27 and present study of 2,613women 30 to 59 years old, Aarhus, Denmark 1988. Data found in literature-odds ratio for age 0.83 (range 0.79 to 0.87), p < 0.0001. Present data-odds ratio for age 0.78 (range 0.76 to 0.79), p < 0.0001.
urinary incontinence to be rather reliable. 14• 16 In regard to childhood bedwetting, our cross-sectional study population constitutes a cohort observed until at least age 30 years. To a varying extent, the reliability of the information depends on the memory performance of the respondents and, possibly also, on differences in social and hygienic norms across age strata. Thus, by experience one might expect a certain amount of under-repor ting associated with increasing present patient age, as was obviously the case regarding bedwetting in early childhood (table 1 and fig. 1). Moreover, despite large sample numbers, and differences in definitions and research methodology applied, our results regarding childhood bedwetting prevalence over subject age did not differ significantly from those of 4 identifiable European population-based childhood studies that provided information on the sex and age-distribution of bedwetting sufficient for reanalysis, nor was there any significant prevalence variation within the reports themselves (fig. 2).7 • 25- 27 Despite the current age-dependent variation found in reporting bedwetting in early childhood (fig. 1), the mean childhood bedwetting prevalence levels found in this stu dy do not seem unrealistic. Moreover, age-specific prevalence of childhood bedwetting actually may have changed with time. If so, this was not reflected in the 4 studies used for met a-analysis (fig. 2).
The aim of our analysis was, however, not to study childhood bedwetting levels as such but rather to investigate a possible association between childhood bedwetting after age 5 years and adult urinary incontinence. Considered t h at present-age stratum-specific odds ratios were shown to be homogeneous (p > 0.45), an information bias 28 due to underreporting does not seem probable. Moreover, because of the relative deficiency in childhood bed wetting frequency, underreporting may have introduced a tendency towards the nil hypothesis. In other words, a contemporary study of the same size might be expected to produce even stronger associations. Childhood bedwetting being associated with adult urge urinary incontinence (p < 0.01) in our data is in accordance with the previously cited observations by Yarnell et al 2 1 and Moore.22 In accordance with the preferably ext ernal etiology of stress urinary incontinence, 14 - 1 8 it was not associated with childhood bedwetting. Moreover, we found a strong association (p < 0.0001) between bedwetting lat e in childhood a nd a dult urinary incontinence occurring especially during sleep , indicating the existence of a nocturnal incontinence pattern persistent during several years of life. Since biological, psychological and also social traits tend to
be stable for years of life, persistence itself does not indicate the nature of the condit ion. Moreover, by definition nocturnal voiding, in ch ildhood as well as in adult life, occurs when the individual cannot consciously control t he bladder function, that is it occurs as a consequence of the mutual balance of micturition and sleep processes. Consequently, the balance may be overturned when the bladder function is unstable 12 and when the sleep processes are provocative,11 for example if voiding acts as a strong "dream substitute."13 Further process details cannot be derived from our data. However, it may be noted that we found weak associations (p < 0.025) between childhood bedwetting a nd adult urinary incontinence occurring especially in situations of nervousness, stressing t hat the primary focus of interest is locat ed t o the nocturnal balance rather than conscious mental processes that may, however, impress it. 8 REFERENCES
1. Jonge, G. A.: Epidemiology of enuresis: a survey of the literature. In: Bladder Control and Enuresis. E dited by I. Kalvin, R. C. MacKeith and S. R. Meadow. London: William Heinemann Medical Books, Ltd., chapt. 4, pp. 39 - 46, 1973. 2. Oppel, W. C., Harper, P. A and Rider, R. V.: The age of attaining bladder control. Pediatrics, 42: 614, 1968. 3. McGee, R. , Makinson, T. , Williams, S., Simpson, A and Silva , P. A : A longitudinal study of enuresis from five to nine years. Aust. Paed. J ., 20: 39, 1984. 4. Bakwin, H .: The genetics of enuresis. In: Bladder Control and Enuresis. Edited by I. Koluin, R. C. MacKeith and S. R. Meadow. London: William Heinemann Medical Books, Ltd., chapt. 9, pp. 73-77, 1973. 5. Dodge, W. F., West, E . F., Bridgforth, E. B. and Travis, L. B.: Nocturnal enuresis in 6- to 10-year-old children. Correlation with bacteriuria, proteinuria, and dysuria. Amer. J. Dis. Child., 120: 32, 1970. 6. Hallgren, B.: Enuresis. II. A study with reference to certain physical, mental, and social factors possibly associated with enuresis. Acta Psychiatr. Scand., 31: 405, 1956. 7. Mar tin, C.R. A: A New Approach to Nocturnal Enuresis. Incidence-Causes-Treatment. London: H. K. Lewis & Co. Ltd. , 1966. 8 . Oppel, W. C., Harper, P . A. and Rider, R. V.: Social, psychological, and neurological factors associated with nocturnal enuresis. Pediatrics, 42: 627, 1968. 9. Berg, I., Fielding, D. and Meadow, R. : Psychiatric distur bance, urgency, and bacteriuria in children with day and nigh t wetting. Arch. Dis. Child., 52: 651, 1977. 10. Tapia, F., Jekel, J . and Domke, H. R. : Enuresis: an emotional symptom? J. Nerv. Ment. Dis ., 130: 61, 1960. 11. Stein, Z. and Susser, M.: Social factors in the development of sphincter control. Develop. Med. Child. Neural., 9: 692, 1967. 12. Mayo, M. E. and Burns, M. W. : Urodynamic studies in children who wet. Brit. J . Ural., 65: 641, 1990. 13. Pierce, C. M., Whitman, R. M., Maas, J. W. and Gay, M. L.: Enuresis and dreaming. Experimental studies. Arch. Gen. Psychiatr., 4: 166, 1961. 14. Diokno, A C., Brock, B. M., H er zog, A R. and Bromberg, J.: Medical correlates of urinary incontinence in the elderly. Urology, 36: 129, 1990. 15. Foldspang, A., Mommsen, S., Lam, G. W. and Elving, L.: P arity a s a correlate of a dult female urinary incontinence prevalence. J . Epidemiol. Com. Health, 46: 595, 1992. 16. Milsom, J ., Ekelund, P., Molander, U., Arvidsson, L. and Areskoug, B.: The influence of age, parity, oral contraception, hyster ectomy and menopause on the prevalence of urinary incontinence in women. J . Ural. , 149: 1459, 1993. 17. Mommsen, S., Foldspang, A., Elving, L. and Lam, G. W.: Association between urinary incontinence in women and a previous history of surgery. Brit. J. Ural., 72: 30, 1993. 18. Mommsen, S., Foldspang, A., Elving, L. and Lam, G. W.: Cystitis as a correlate of prevalent female urinar y incontinence. Int. Urogynecol. J., in press. 19. Mommsen, S. and Foldspang, A.: Body mass index and adult female urinary incontinence. World J. Ural., in press. 20. Elving, L., Foldspang, A , Lam, G. W. and Mommsen, S .: Descriptive epidemiology of urinary incontinence in 3,100 women
!1 '
88
ADULT FEMALE URINARY INCONTINENCE AND CHILDHOOD BEDWETTING
age 30-59. Scand. J. Urol. Nephrol., suppl., 125: 37, 1989. 21. Yarnell, J. W., Voyle, G. J., Sweetnam, P. M., Milbank, J., Richards, C. J. and Stephenson, T. P.: Factors associated with urinary incontinence in women. J. Epidemiol. Com. Health, 36: 58, 1982. 22. Moore, K. H., Richmond, D. H . and Parys, B. T.: Sex distribution of adult idiopathic detrusor instability in relation to childhood bedwetting. Brit. J. Urol., 68: 479, 1991. 23. Lam, G. W., Foldspang, A., Elving, L. B. and Mommsen, S.: Social context, social abstention, and problem recognition correlated with adult female urinary incontinence. Danish Med. Bull., 39: 565, 1992.
24. Armitage, P. and Berry, G. : Statistical Methods in Medical Research. Oxford: Blackwell Scientific Publications, 1988. 25. Blomfield, J.M. and Douglas, J. W. B.: Bedwetting; prevalence among children aged 4-7 years. Lancet, 1: 850, 1956. 26. Forrester, R. M., Stein, Z. and Susser, M. W. : A trial of conditioning therapy in nocturnal enuresis. Develop. Med. Child. Neurol. , 6: 158, 1964. 27. Weir, K. : Night and day wetting among a population of threeyear-olds. Develop. Med. Child. Neurol., 24: 479, 1982. 28. Kleinbaum, D. G., Kupper, L. L. and Morgenstern, H. : Epidemiologic Research. Principles and Quantitative Methods. London: Lifetime Learning Publications, 1982.
- .
• -
'-•:.--~~---:---.• • -
.-4
• - - _"':,.
0022-534 7/94/1521-0085$03.00/0 Vol. 152, 85-88, July 1994
THE JOURNAL OF UROLOGY Copyright © 1994 by AMERICAN UROLOGICAL ASSOCIATION, INC .
Printed in U.S.A.
ADULT FEMALE URINARY INCONTINENCE AND CHILDHOOD BED WETTING ANDERS FOLDSPANG
AND
S0REN MOMMSEN
From the Institute of Epidemiology and Social Medicine, Aarhus University, and Department of Urology, Aarhus University Hospital, Aarhus, and Department of Urology, Odense University Hospital, Odense, Denmark
ABSTRACT
A cross-sectional random population sample of women 30 to 59 years old was sent a questionnaire on urinary incontinence and, among other things, childhood bedwetting. Among 2,613 responders 17 .0% reported prevalent urinary incontinence (14. 7% stress provoked, 8.3% associated with urge, 6.8% stress and urge overlap, 2.2% occurring especially during sleep and 3.9% occurring especially when anxious), and 6.5% reported childhood bedwetting after age 5 years and 3.3% after age 10 years. Childhood bedwetting was associated with prevalent urge urinary incontinence (p <0.01) and incontinence occurring during sleep (p <0.0001) but was less marked with urinary incontinence occurring in situations of anxiety (p <0.025). It is concluded that some individuals suffer a long-lasting disturbance of the balance between micturition and sleep processes. KEY WORDS:
urinary incontinence, enuresis, population, urination disorders
Childhood bedwetting is frequent in Western industrialized populations but less prevalent in Europe than in other areas of the world. 1- 3 The occurrence of childhood bedwetting has been related to genetic, 4 ethnic,5 physical, 6 - 8 mental 6 • 3 - 10 and social5 • 6 • 8 • 11 factors. Some physicians have investigated more specific urological aspects,7 • 12 and the role of the sleep and dream pattern.7 • 11 • 13 Based on a multifactorial etiology, 14 - 19 urinary incontinence also is prevalent in adult female populations. 20 The similarity of the nature of childhood bedwetting and urinary incontinence suggests the possibility of a mutual relationship to be a straightforward object for research interest, which might shed further light on the nature of either condition. Moreover, the number of studies on the topic is extremely low, for example among physicians concerned with the prevalence of urinary incontinence in women only Yarnell et al investigated childhood bedwetting also.21 Conditional on bed wetting after the ages of 5 and 9 years among 990 women 18 years old and older, they found moderately increased relative risks of having urinary incontinence (relative risk 1.3, p -0.05, and relative risk 1.4, p -0.05, respectively). In 500 women with urological conditions Moore found a strong association between present idiopathic detrusor instability and the patient who reported bedwetting in childhood (relative risk 2.7, p <0.0001), 22 We have been unable to identify any childhood cohorts who were observed regarding the occurrence of urinary incontinence in adult life. We presently investigate the possible association between childhood bedwetting and urinary incontinence in women. As contrasted to stress urinary incontinence, we previously found urge incontinence to occur relatively often during sleep and in situations of anxiety.23 Considering this finding, and the urodynamic, psychological and social correlates of childhood bedwetting alluded to, we hypothesized a priori bedwetting to be associated primarily with urge urinary incontinence and, furthermore, with urge incontinence occurring during sleep and in situations of anxiety.
MATERIALS AND METHODS
The applied cross-sectional design and its data have been presented previously.2 0 During January and Febr uary 1988, the members of an age-stratified random sample of 3,114 women 30 to 59 years old and living in the Municipality of Aarhus (total female population of the same age 47,702) were mailed a self-administered postal questionnaire. Nonrespondents were addressed 2 additional times. Finally, 2,631 subjects (84 .5%) responded, with a slight but significant reduction by age (odds ratio 0.93 per 5-year group, p <0.025 [multiple logistic regression estimate]). The present analysis comprises 2,613 women who reported their age at investigation. Among other things, the questionnaire asked about episodes of urinary incontinence during 1987. If present, were urinary incontinence episodes generally provoked by physical stress, for example lifting, coughing, sneezing or laughing (stress incontinence), or accompanied by a feeling of urge to void (urge incontinence) or both (mixed incontinence). If the woman reported urinary incontinence, did it occur especially during sleep or when she felt nervous. Had t he woman experienced childhood bed wetting and, if so, at what age was she able to control micturition during sleep. Additional information was collected on menopausal status, childbirth, surgical experience, cystitis, and body height (cm.) and weight (kg.). Prevalence (%) of reports of childhood bedwetting by age at investigation in 2,613 women 30 to 59 years old, Aarhus, Denmark, 1988
TABLE 1.
Bedwetting Cessation Age(%) Age (yrs.)
30-34 35- 39 40-44 45-49 50- 54 55- 59
Total No.
415 420 508 366 445 459 2,613
6-18 Yrs.
6- 10 Yrs.
11-18 Yrs.
8.9 6.9 8.3 4.6 4.9 4.8 6.5 0.87t
3.9 4.3 4.1 2.5 2.0 2.2 3.2 0.85t
5.1 2.6 4.1 2.2 2.9 2.6 3.3 0.89§
Totals Odds ratio* * Odds ratio indicates the change in prevalence per 5 years of subject age at investigation (logistic regression estimate). t p < 0.005. :j: p > 0.025. § p < 0.10.
Accepted for publication December 3, 1993. Su pported by Grant 12-7621, 1987 from the Danish Medical Research Council, the Health Care Division of the Miilnlycke Co., the KabiVitrum Co., the Danish Urological Society (the Rahbeck Foundation) and the Research Fund of the Aarhus County Division of the Danish Medical Association. 85
~
86
-
ADULT FEMALE URINARY INCONTINENCE AND CHILDHOOD BEDWETTING TABLE 3 . Prevalence of urinary incontinence occurring especially during sleep, by age at cessation of childhood bedwetting and urge incontinence in 2,613 women 30 to 59 years old, Aarhus, Denmark, 1987
Prevalence(%} of bedwetting 10
8
Age at Cessation of Childhood Bedwetting
-All
6
- · Present age 30-39 ···Present age 40-49 4
Urge Urinary Incontinence
- Present age 50-59
7
8
9 10 11 12 13 14 15 16 17 18 19 Age (years)
Fm. 1. Survival analysis of childhood bedwetting in 2,613 women 30 to 59 years old, Aarhus, Denmark, 1988.
Body mass index is defined as the body weight (kg.) divided by the squared body length (m.). In the text the term "prevalence" is used synonymously with "period prevalence 1987". The statistical analysis applies survival analysis, Spearman's correlation coefficient r , Pearson's chi-square test, Fisher's exact test, Mantel-Haenszel's stratified analysis and multiple logistic regression. 24 A p value of 0.05 was used as a general level of significance. Logistic regression was performed by use of a combination of forward selection and backward elimination. Besides an odds ratio regression point estimate and a 95% interval estimate (in parentheses), each predictor was assigned an independent p value based on the chi-square distributed goodness of fit likelihood ratio statistic. The statistical analysis was performed by 1 of us (A. F.). RESULTS
A total of 169 women (6.5%) reported childhood bedwetting after age 5 years (table 1). Of the women 83 (3.2%) and 86 (3.3%) had ceased bedwetting during ages 6 to 10 and 11 to 18 years, respectively. The proportion of women who reported childhood bedwetting decreased significantly by increasing present age. Moreover, this trend was evident regarding childhood bedwetting reported to have ceased before the age of 11 years, whereas bedwetting later in childhood was reported with equal frequency irrespective of age at investigation (fig. 1). Overall, urinary incontinence was reported by 17.2% of the subjects (stress incontinence by 14.8% and urge incontinence by 8.6%). Most women (82.7%) with urge incontinence also had stress incontinence. In 22 women (0.8%) urinary incontinence was neither provoked by stress nor associated with
T ABLE 2 .
Odds Ratio*
25 (11.8) 9.9t 31 (1.3) 11-18 7 (50.0) 17.0t 72 4 (5.6) Totals 2,6 13 67 (2.6) 10.8t,§ * Odds ratio indicates the relative risk of incontinence occurring during sleep conditional on clinical type (urge) of incontinence. t p < 0.00001. t p < 0.0002. § Weight ed Mantel-Haenszel estimate indicating the association of prevalent urge incontinence to incontinence occurring especially during sleep, corrected for age at cessation of childhood bedwetting. 0--10
2
0 6
Total No.
Urinary Incont inence Occurring During Sleep No. (%)
Yes No Yes No
211 2,316 14
urge (unspecific incontinence). Incontinence occurred especially during sleep and 1n situations of a nxiety in 2.6% a nd 4.1% of the women, respectively. Women who reported childhood bedwetting had a high prevalence of urinary incontinence associated with urge, and occurring during sleep and in situations of anxiety (table 2). The symptoms of urge and experiencing incontinence during sleep were associated with childhood bedwetting having ceased late in childhood, while urinary incontinence in anxious situations was correlated with earlier cessation of bedwetting. When corrected for childbirth, curettage, cystitis experience and body mass index, the pattern of association presented in table 2 remained the same, and the risk estimates changed by ±0.2 units at most. The crucial relationship among age at cessation of childhood bedwetting, urge incontinence and incontinence occurring during sleep is illustrated in table 3. Among women reporting childhood bedwetting, only urinary incontinence during sleep was associated with linear age at cessation ofbedwetting (odds ratio 1.2 per year of age, p < 0.05). Unspecific urinary incontinence was associated with incontinence during sleep (p < 0.00001) but unassociated with childhood bedwetting. Former bedwetters did not report a longer mean interval since the first adult incontinence episode than did other women. DISCUSSION
The reliability of our dat a depends on the ability and willingness of study participants to report validly. Because of resource limitations this questionnaire was not validated itself but, using clinical evaluation as a reference, others h ave found questionnaire-based information on prevalent
Prevalence (%) of urinary incontinence by clinical type, situation of occurrence and age at cessation of childhood bedwetting in 2,613 women 30 to 59 years old, Aarhus, Denmark, 1988 Urinary Incontinence
Age at Cessation of Bedwetting (yrs.)
All Pts. %
Odds Ratio*
All Stress %
Odds Ratio*
All Urge %
Odds Ratio*
Mixed %
Odds Ratio*
During Sleep %
Odds Ratio*
When Nervous %
Odds Ratio*
6- 10 19.3 1.2 18.1 1.4 10.8 1.4 9.6 1.5 3.6 1.5 8.4 2.6t 11- 18 19.8 1.1 16.3 12.8 2.0§ 12.8 5.8 1.6 1.2 15.1 2.2t 6.711 6- 18 19.5 1.2 16.6 1.2 13.6 11.2 8.3 7.1 2. l t l.8t l.8t 4.311 Nonbedwetters 14.7 8.3 6.8 2.2 3.9 17.0 * Odds ratio, standardized for age (years) at investigation, indicates th e urinary incontinence prevalence relative risk as compared to that of nonbedwetters and women h aving ceased bedwetting during other ages. t p < 0.025. t p < 0.01. § p < 0.05. 11 p < 0.0001.
ADULT FEMALE URINARY INCONTINENCE AND CHILDHOOD BEDWETTING Prevalence(%) of bedwetting
iO
- - - - -~
8
6
- Literature - · Present study
4 2
06
7
8
9 10 11 12 13 14 15 16 17 18 19
Age (year s)
FIG. 2. Logistic regression based estimates of prevalence of bedwetting by age in previously published studies (total number of children examined 3, 786)7 • 25 · 27 and present study of 2,613women 30 to 59 years old, Aarhus, Denmark 1988. Data found in literature-odds ratio for age 0.83 (range 0.79 to 0.87), p < 0.0001. Present data-odds ratio for age 0.78 (range 0.76 to 0.79), p < 0.0001.
urinary incontinence to be rather reliable. 14• 16 In regard to childhood bedwetting, our cross-sectional study population constitutes a cohort observed until at least age 30 years. To a varying extent, the reliability of the information depends on the memory performance of the respondents and, possibly also, on differences in social and hygienic norms across age strata. Thus, by experience one might expect a certain amount of under-repor ting associated with increasing present patient age, as was obviously the case regarding bedwetting in early childhood (table 1 and fig. 1). Moreover, despite large sample numbers, and differences in definitions and research methodology applied, our results regarding childhood bedwetting prevalence over subject age did not differ significantly from those of 4 identifiable European population-based childhood studies that provided information on the sex and age-distribution of bedwetting sufficient for reanalysis, nor was there any significant prevalence variation within the reports themselves (fig. 2).7 • 25- 27 Despite the current age-dependent variation found in reporting bedwetting in early childhood (fig. 1), the mean childhood bedwetting prevalence levels found in this stu dy do not seem unrealistic. Moreover, age-specific prevalence of childhood bedwetting actually may have changed with time. If so, this was not reflected in the 4 studies used for met a-analysis (fig. 2).
The aim of our analysis was, however, not to study childhood bedwetting levels as such but rather to investigate a possible association between childhood bedwetting after age 5 years and adult urinary incontinence. Considered t h at present-age stratum-specific odds ratios were shown to be homogeneous (p > 0.45), an information bias 28 due to underreporting does not seem probable. Moreover, because of the relative deficiency in childhood bed wetting frequency, underreporting may have introduced a tendency towards the nil hypothesis. In other words, a contemporary study of the same size might be expected to produce even stronger associations. Childhood bedwetting being associated with adult urge urinary incontinence (p < 0.01) in our data is in accordance with the previously cited observations by Yarnell et al 2 1 and Moore.22 In accordance with the preferably ext ernal etiology of stress urinary incontinence, 14 - 1 8 it was not associated with childhood bedwetting. Moreover, we found a strong association (p < 0.0001) between bedwetting lat e in childhood a nd a dult urinary incontinence occurring especially during sleep , indicating the existence of a nocturnal incontinence pattern persistent during several years of life. Since biological, psychological and also social traits tend to
be stable for years of life, persistence itself does not indicate the nature of the condit ion. Moreover, by definition nocturnal voiding, in ch ildhood as well as in adult life, occurs when the individual cannot consciously control t he bladder function, that is it occurs as a consequence of the mutual balance of micturition and sleep processes. Consequently, the balance may be overturned when the bladder function is unstable 12 and when the sleep processes are provocative,11 for example if voiding acts as a strong "dream substitute."13 Further process details cannot be derived from our data. However, it may be noted that we found weak associations (p < 0.025) between childhood bedwetting a nd adult urinary incontinence occurring especially in situations of nervousness, stressing t hat the primary focus of interest is locat ed t o the nocturnal balance rather than conscious mental processes that may, however, impress it. 8 REFERENCES
1. Jonge, G. A.: Epidemiology of enuresis: a survey of the literature. In: Bladder Control and Enuresis. E dited by I. Kalvin, R. C. MacKeith and S. R. Meadow. London: William Heinemann Medical Books, Ltd., chapt. 4, pp. 39 - 46, 1973. 2. Oppel, W. C., Harper, P. A and Rider, R. V.: The age of attaining bladder control. Pediatrics, 42: 614, 1968. 3. McGee, R. , Makinson, T. , Williams, S., Simpson, A and Silva , P. A : A longitudinal study of enuresis from five to nine years. Aust. Paed. J ., 20: 39, 1984. 4. Bakwin, H .: The genetics of enuresis. In: Bladder Control and Enuresis. Edited by I. Koluin, R. C. MacKeith and S. R. Meadow. London: William Heinemann Medical Books, Ltd., chapt. 9, pp. 73-77, 1973. 5. Dodge, W. F., West, E . F., Bridgforth, E. B. and Travis, L. B.: Nocturnal enuresis in 6- to 10-year-old children. Correlation with bacteriuria, proteinuria, and dysuria. Amer. J. Dis. Child., 120: 32, 1970. 6. Hallgren, B.: Enuresis. II. A study with reference to certain physical, mental, and social factors possibly associated with enuresis. Acta Psychiatr. Scand., 31: 405, 1956. 7. Mar tin, C.R. A: A New Approach to Nocturnal Enuresis. Incidence-Causes-Treatment. London: H. K. Lewis & Co. Ltd. , 1966. 8 . Oppel, W. C., Harper, P . A. and Rider, R. V.: Social, psychological, and neurological factors associated with nocturnal enuresis. Pediatrics, 42: 627, 1968. 9. Berg, I., Fielding, D. and Meadow, R. : Psychiatric distur bance, urgency, and bacteriuria in children with day and nigh t wetting. Arch. Dis. Child., 52: 651, 1977. 10. Tapia, F., Jekel, J . and Domke, H. R. : Enuresis: an emotional symptom? J. Nerv. Ment. Dis ., 130: 61, 1960. 11. Stein, Z. and Susser, M.: Social factors in the development of sphincter control. Develop. Med. Child. Neural., 9: 692, 1967. 12. Mayo, M. E. and Burns, M. W. : Urodynamic studies in children who wet. Brit. J . Ural., 65: 641, 1990. 13. Pierce, C. M., Whitman, R. M., Maas, J. W. and Gay, M. L.: Enuresis and dreaming. Experimental studies. Arch. Gen. Psychiatr., 4: 166, 1961. 14. Diokno, A C., Brock, B. M., H er zog, A R. and Bromberg, J.: Medical correlates of urinary incontinence in the elderly. Urology, 36: 129, 1990. 15. Foldspang, A., Mommsen, S., Lam, G. W. and Elving, L.: P arity a s a correlate of a dult female urinary incontinence prevalence. J . Epidemiol. Com. Health, 46: 595, 1992. 16. Milsom, J ., Ekelund, P., Molander, U., Arvidsson, L. and Areskoug, B.: The influence of age, parity, oral contraception, hyster ectomy and menopause on the prevalence of urinary incontinence in women. J . Ural. , 149: 1459, 1993. 17. Mommsen, S., Foldspang, A., Elving, L. and Lam, G. W.: Association between urinary incontinence in women and a previous history of surgery. Brit. J. Ural., 72: 30, 1993. 18. Mommsen, S., Foldspang, A., Elving, L. and Lam, G. W.: Cystitis as a correlate of prevalent female urinar y incontinence. Int. Urogynecol. J., in press. 19. Mommsen, S. and Foldspang, A.: Body mass index and adult female urinary incontinence. World J. Ural., in press. 20. Elving, L., Foldspang, A , Lam, G. W. and Mommsen, S .: Descriptive epidemiology of urinary incontinence in 3,100 women
!1 '
88
ADULT FEMALE URINARY INCONTINENCE AND CHILDHOOD BEDWETTING
age 30-59. Scand. J. Urol. Nephrol., suppl., 125: 37, 1989. 21. Yarnell, J. W., Voyle, G. J., Sweetnam, P. M., Milbank, J., Richards, C. J. and Stephenson, T. P.: Factors associated with urinary incontinence in women. J. Epidemiol. Com. Health, 36: 58, 1982. 22. Moore, K. H., Richmond, D. H . and Parys, B. T.: Sex distribution of adult idiopathic detrusor instability in relation to childhood bedwetting. Brit. J. Urol., 68: 479, 1991. 23. Lam, G. W., Foldspang, A., Elving, L. B. and Mommsen, S.: Social context, social abstention, and problem recognition correlated with adult female urinary incontinence. Danish Med. Bull., 39: 565, 1992.
24. Armitage, P. and Berry, G. : Statistical Methods in Medical Research. Oxford: Blackwell Scientific Publications, 1988. 25. Blomfield, J.M. and Douglas, J. W. B.: Bedwetting; prevalence among children aged 4-7 years. Lancet, 1: 850, 1956. 26. Forrester, R. M., Stein, Z. and Susser, M. W. : A trial of conditioning therapy in nocturnal enuresis. Develop. Med. Child. Neurol. , 6: 158, 1964. 27. Weir, K. : Night and day wetting among a population of threeyear-olds. Develop. Med. Child. Neurol., 24: 479, 1982. 28. Kleinbaum, D. G., Kupper, L. L. and Morgenstern, H. : Epidemiologic Research. Principles and Quantitative Methods. London: Lifetime Learning Publications, 1982.