Neurourology and Urodynamics 26:218^221 (2007)

Urinary Flow Disturbance as an Early Sign of AutonomicNeuropathy in Diabetic Children and Adolescents,

Laszlo Szabo,1,4,5* Laszlo Barkai,2,4,5 and Bela Lombay3,51Departments of Pediatric Nephrology, Hypertension and Urodynamics, Child Health Center,

Borsod County Teaching Hospital, Miskolc, Hungary2Department of Diabetes, Child Health Center, Borsod County Teaching Hospital, Miskolc, Hungary

3Department of Radiological Imaging, Child Health Center, Borsod County Teaching Hospital, Miskolc, Hungary4Postgraduate Institute of Pediatrics, University of Debrecen, Miskolc, Hungary

5Institute of Health Science, University of Miskolc, Miskolc, Hungary

Aims: Urinary bladder dysfunction is a frequent complication of diabetes mellitus in adults. Theaim of this study is to determine the early disturbances of the detrusor contractility in children, andadolescents with type 1 diabetes.Materials and Methods: The bladder urine £ow parameters wereinvestigated in 37 children with type 1 diabetes, 10/37 with cardiovascular autonomic dysfunction(CAD) voided 482 � 128 mils (Group A), 27/37 without CAD voided 258 � 52 mils (Group B). Froman earlier uro£ow screening study 20 healthy children formed group control A, who had voided488� 94 mils, and another 30 healthy children formed group control B, who had voided 260�50 mils. The compared groups were matched for voided volume, age, weight, and height. There wasno evidence of urinary tract abnormalities in any of the children. Results: Time to maximum £owwas longer in both diabetic groups as compared with controls (P < 0.01), and the acceleration(the ratio of maximum £ow and the time to maximum £ow) of diabetics was signi¢cantly lower(P < 0.01). Bladder emptying was complete in each subject. Bladder wall disturbances were not seenby ultrasound. Conclusions: The decreased acceleration of detrusor muscle contraction maybe interpreted as an early sign of autonomic neuropathy in children and adolescent with type 1 dia-betes, even in patients without CAD. Clinicians may be able to determine autonomic neuropathy byusing uro£owmetry that is easy to perform, sensitive, reproducible, and needs only a child’s mini-mal cooperation. Neurourol. Urodynam. 26:218 ^221, 2007. � 2006 Wiley-Liss, Inc.

Key words: diabetesmellitus; diabetic cystopathy;urine£owacceleration;uro£owmetry

INTRODUCTION

The number of patients with diabetes mellitus is increasing.Diabetic autonomic neuropathy is associated with a markedlyreduced quality of life and poor prognosis.The manifestationsof diabetic autonomic neuropathy cause multiple symptomsand involve the cardiovascular system, respiratory system,gastrointestinal tract, and genitourinary tract. Urinary blad-der dysfunction is a frequent complication of diabetes melli-tus in adults. Diabetic cystopathy is characterized by animpaired sensation of bladder fullness, increased bladdercapacity, reduced bladder contractility, and elevated residualurine volume [Frimodt-Moller, 1980; Massin et al., 1999].However, adult diabetic patients have concomitant lesions,such as benign prostate hyperplasia, stress incontinence, blad-der or prostate cancer and infection, which may coexist withor mimic the symptoms of diabetic cystopathy [Kaplan et al.,1995]. Clinical manifestations of voiding dysfunction in chil-dren with diabetes are fortunately very rare, but only few datahave been published [Barkai and Szabo, 1993; Wahl et al.,2001]. In our earlier study we compared bladder sensation ofdiabetic children with cardiovascular autonomic dysfunction

(CAD) and without CAD.These ¢ndings suggest that diabeticchildren may have diminished sensation of bladder ¢llingindependent of impaired cardiovascular re£exes, however,the degree of bladder dysfunction parallels with CAD, bothdepending on diabetes duration and long-term glycometaboliccontrol [Barkai and Szabo, 1993]. The aim of present study isto determine the early disturbances of the detrusor contracti-lity in children, and adolescents with type 1 diabetes.

MATERIALS AND METHODS

Prospective study includes 37 children with type 1 (insulindependent) diabetes. They were divided into two groups with

No con£ict of interest reported by the author(s).The study was presented at the ICCS-ESPU 2004 Congress in Gent.*Correspondence to: Laszlo Szabo, MD, PhD, Szentpeteri kapu 72, MiskolcPOBox: 188. H-3501Miskolc, Hungary. E-mail: lszabo52@t-online.huReceived 17 December 2005; Accepted 14 August 2006Published online 31October 2006 inWiley InterScience(www.interscience.wiley.com)DOI 10.1002/nau.20349

�2006Wiley-Liss, Inc.

respect to CAD. Ten patients with CAD de¢ned as two ormore abnormal cardiovascular tests were included in GroupA. They had a longer duration of diabetes and bad metaboliccontrol (mean fructosamin level was 4.6 � 0.4 mmol/L,Table I), and their mean voided volume was 482 � 128 mils at¢rst sensation of bladder fullness. Twenty-seven patients with-out CAD were included in Group B. They had a shorter dura-tion of diabetes and good metabolic control (meanfructosamin level was 2.9 � 0.6 mmol/L, Table I), and theirmean voided volume was 258 � 52 mils at ¢rst sensation ofbladder fullness.The controls for GroupAwere obtained froma previous uro£ow screening study, where 231 healthy chil-dren voided at ¢rst and at maximal sensation of bladder full-ness [Szabo and Fegyverneki, 1995]. Twenty healthy childrenvoided 488 � 94 mils and formed control for Group A, andanother 30 healthy children voided 260 � 50 mils and formedcontrol for Group B. The uro£ow parameters highly dependon the voided volumes. The di¡erent uro£ow parameters canbe compared if the voided volumes are same, so we establishedtwo controls groups because of the measured and calculatedurine £ow parameters depend on the voided volume basically.The compared groups were matched for voided volumes, age,weight, and height, too (Table I).The controls did not have anylower urinary tract problem.We have not found dysfunctionalvoiders among them. Neither the healthy, nor the diabeticchildren had been treated with any medicine, except for insu-lin, during the study and the preceding 2 weeks. There was noevidence of vascular complications, urinary tract abnormal-ities or manifest neuropathy in any of the children.

Five cardiovascular tests, originally described for adults[Ewing and Clarke, 1982], and for children [Barkai andMadacsy, 1995], were used to determine whether CAD waspresent. The tests employed included mean resting heart ratefor a period of 1 min, heart rate variation to deep breathing,heart rate response to standing from lying position, fall in sys-tolic blood pressure on standing, and rise in diastolic bloodpressure during sustained handgrip. Heart rates were deter-mined with a routine electrocardiography device (MedicorER 31-A, Budapest) by R-R intervals, for blood pressuremeasurements a digital blood pressure device (Omron HEM-400 C,Tokyo) was used. All diabetic patients had all ¢ve tests.

Urinary bladder function was assessed by ultrasound anduro£owmetry. We measured the voiding volume, the time to

maximum urinary £ow (TQmax), the maximum urinary £ow(Qmax) and the acceleration (ratio of Qmax and TQmax). Blad-der function was assessed within 2 weeks after the cardiovas-cular tests, and both investigations were carried out at least2 hr after the morning insulin dose.Two £ows were performedon each child. Ultrasound was accomplished before and aftermicturition. Scans were obtained with a real time ultrasoundscanner (Hitachi EUB 40, 5 MHz transducer) using a directscanning technique, and the images were recorded on videotape. Uro£ow was performed using an Uro£ow-Cystometer(X002 Metripond, Hungary). The urodynamic methods, de¢-nitions and units we employed conform to the standardsrecommended by the International Children ContinenceSociety, except where specially noted [Norgaard et al., 1998].Data and results are expressed as mean � SD. Data weretested by the Student t-test or by the Mann^Whitney ranksum test when data were not normally distributed. A P-value<0.05 was considered to be statistically signi¢cant.The resultsof patients with diabetes with CAD and Control Group Awere separately compared, and the result of patients with dia-betes without CAD andControlGroupB were also separatelycompared. The results of diabetic patients with and withoutCAD have not been compared.The study was approved by the Local Ethics Committee.

RESULTS

Time to maximum £ow was longer in both diabetic groupsas compared with controls (P < 0.01), and the acceleration (theratio of maximum £ow and the time to maximum £ow) of dia-betics were signi¢cantly lower (P < 0.01) (Table II). Bladderemptying was complete in each subject. There was no measur-able post-micturition residual urine by ultrasound. Bladderwall disturbances were not seen by ultrasound. They did nothave dysfunctional voiding. The di¡erences of uro£ow para-meters were not compared between two diabetic groupsbecause of the voided volumes were di¡erent.

DISCUSSION

Studies in adult patients have shown that autonomic neuro-pathy is a frequent complication of diabetes that contributessigni¢cantly to morbidity and mortality. Lower genito-urin-

Neurourology and Urodynamics DOI 10.1002/nau

TABLE I. Baseline Characteristics of Patients With and Without CAD and Controls

DataDiabetes with

CAD Control ADiabetes without

CAD Control B

n (male/female) 10 (5/5) 20 (10/10) 27 (15/12) 30 (15/15)Age (years) 13.8 � 2.2 13.5 � 2.0 13.6 � 2.0 13.7 � 2.1Weight (kg) 42.3 � 4.2 42.0 � 4.1 39.8 � 6.1 40.2 � 5.2Height (cm) 157 � 6.2 155 � 5.2 156 � 8.0 157� 6.3Diabetes duration (years) 6.3 � 2.4 � 3.4 � 2.0 �Mean fructosamin (mmol/l) 4.6 � 0.4 2.0 � 0.5 2.9 � 0.6 2.0 � 0.5

Diabetic Flow 219

ary problems are part of the polyneuropathy of diabetes.Cystopathy a¡ects 40^85% of adult diabetic patients,although less than half of them are symptomatic. The classicsymptoms are not always observed in patients with diabetes,who often have a varied symptom presentation.

Diabetic cystopathy is induced by polyneuropathythat predominantly a¡ects sensory and autonomic ¢bers[Frimodt-Moller, 1980; Ueda et al., 1997; Ingberg et al., 1998].

Ueda et al. [1997] investigated the possible relationshipbetween bladder dysfunction and autonomic neuropathy inunselected diabetic patients. Cystometrograms exhibited sig-ni¢cant increases in bladder volume at ¢rst desire to void andmaximal bladder capacity, and a decrease in detrusor contrac-tility in diabetic patients compared to those in controls. Wefound increased bladder capacity in our earlier study [Barkaiand Szabo, 1993]. Children with diabetes mellitus who weretested with uro£ow had a signi¢cantly increased voidingvolume, increased average urinary £ow, and increased delayuntil the ¢rst sensation of the need to void, compared withhealthy children. These abnormal urodynamic features areconsistent with damage to the a¡erent sensory pathways.Ishigooka [Ishigooka et al., 1996] evaluated 42 patientswith diabetes mellitus by uro£owmetry. After urodynamicinvestigation 12 patients were diagnosed as having diabeticcystopathy. In patients with diabetic cystopathy, most of theparameters in uro£owmetry were signi¢cantly di¡erent fromthose of normal subjects. Moreover, to our knowledge nostudies have examined detrusor muscle acceleration indiabetic children. In the literature [Susset, 1983; Cucchi,1990; Chancellor et al., 1991; Jorgensen et al., 1993], theacceleration in young male was 3.21ml/sec2.We found higherlevels in normal children, but the acceleration levels signi¢-cantly decreased not only in diabetic children with CAD, butalso in diabetic patient without CAD, but the result was worsein patient with CAD and their diabetes duration was longer(Table I). The standard deviations were high, especially inpatients, and this situation reduces the value of our results.Unfortunately we could not identify patients who were moreat risk of a more rapid progression of their diabetes based onuro£ow parameters.We have showed that patients with mod-

erately abnormal glycolmetabolic control (Group B) alreadyhad abnormal uro£ow parameters (TQmax and Qmax/TQmax)without any other sign of autonomic neuropathy.These resultshave not caused any impact on bladder emptying in ourpatients, but we think that diabetic autonomic neuropathy isa process, and at the end of it, most of the adult diabeticpatients have serious voiding problems with large residualurine. So we hypothesize, that these abnormal uro£ow para-meters can be interpreted as an early sign of autonomic neuro-pathy in diabetic children. We suspect that there is acorrelation between these uro£ow parameters and bloodsugar levels, and di⁄culties of glycometabolic control. Wethink that uro£ow changes can appear early on in the diseaseand uro£ow has good potential to monitor its progress. Werecommend the ¢rst uro£ow at onset of the diagnosis ofdiabetes, because of the patient may have disturbances of thelower urinary tract function originally. We also recommendthe repeated uro£ow test in every 6 months. We shouldcompare the repeated uro£ow results with the initial test.Clinicians should be able to determine autonomic neuropa-

thy by using an uro£owmetry that is easy to perform, sensi-tive, reproducible, and need only a small cooperation of thechildren.The autonomous neuropathic situations occur quite fre-

quently, especially in older subjects who have su¡ered fromdiabetes for more than 10 years. Studies of the few therapeuticmeasures practiced have not proved very conclusive. Only adiligent examination of the signs and symptoms with the aimof early diagnosis and maintenance of good glycometabolicbalance are considered to be e¡ective as preventive measures.Rapidi et al. [2005] also looked for early sign of diabetic

cystopathy and found abnormal tibial somatosensory evokedpotentials even in diabetic patients without lower urinarytract symptoms.Early detection of urinary autonomic dysfunction may be

important to further motivate patients to improve their dia-betes control and hopefully to delay the development of com-plications. Therefore, these diabetic patients need morefrequent control at outpatient clinic. They need more strictdiet and special attention from their parents.

Neurourology and Urodynamics DOI 10.1002/nau

TABLE II. Uroflow Parameters of Patients and Controls

Data Diabetes with CAD Control A Diabetes without CAD Control B

Voided volume (ml) 482 � 128 488 � 94 258 � 152 260 � 50Voiding time (s) 17.0 � 8.1 15.0 � 6.0 17.4 � 6.2 14.5 � 4.2Qmax (ml/sec) 33.5 � 10.2 35.0 � 5.8 23.8 � 11.0 24.5 � 5.8TQmax (s) 11.3 � 6.3* 5.8 � 4.7 8.3 � 5.4** 2.4 � 1.2Qmax/TQmax (ml/sec

2) 2.96 � 1.68* 6.03 � 1.23 4.87 � 2.04** 10.21 � 4.83

*P < 0.01 vs. control group A.**P < 0.01 vs. control group B.Qmax�maximal urine £ow in ml/sec.TQmax�Time to maximal urine £ow in sec.Qmax/TQmax�maximal urine £ow divided by time to maximal urine £ow ¼ acceleration (ml/sec2).

220 Szabo et al.

CONCLUSION

Urine £ow measurement may be a useful technique for rea-lizing early signs of autonomic neuropathy in diabetic chil-dren. This hypothesis will need validation over time withexamination on durability and progression.

REFERENCES

Barkai L, Madacsy L. 1995. Cardiovascular autonomic dysfunction in dia-betes mellitus. Arch Dis Child 73:515^8.

Barkai L, Szabo L. 1993. Urinary bladder dysfunction in diabetic childrenwith and without subclinical cardiovascular autonomic neuropathy. EurJ Pediatr 152:190^2.

ChancellorMB, Blaivas JG, Kaplan SA, et al. 1991. Bladder outlet obstructionversus impaired detrusor contractility: The role of uro£ow. J Urol145:810^2.

Cucchi A. 1990. Acceleration of £ow rate in obstructive detrusor instability.Br J Urol 66:26^9.

Ewing DJ, Clarke BF. 1982. Diagnosis and management of diabetic auto-nomic neuropathy. BMJ 285:916^8.

Frimodt-Moller C. 1980. Diabetic cystopathy: Epidemiology and related dis-orders. Ann Intern Med 92:318^22.

Ingberg CM, Palmer M, Schwarcz E, et al. 1998. Prevalence of urinary tractsymptoms in long-standing type 1 diabetes mellitus. Diabetes Metabolism24:343^6.

Ishigooka M, Suzuki Y, Hayami S, et al. 1996. Role of symptom scoring anduro£owmetry in patients with diabetic cystopathy. Int Urol Nephrol28:761^6.

Jorgensen JB, Jensen KM-E, Mogensen P. 1993. Longitudinal observations onnormal and abnormal voiding in men over the age of 50 years. Brit J Urol72:413^20.

Kaplan AS, Te AE, Blaivas JG. 1995. Urodynamic ¢ndings in patients withdiabetic cystopathy. J Urol 153:342^4.

Massin MM, Ernould C, Derkenne B, et al. 1999. Cardiac autonomic dys-function in diabetic children. Diabetes Care 22:1845^50.

Norgaard JP, van Gool JD, Hjalmas K, et al. 1998. Standardization andde¢nitions in lower urinary tract dysfunction in children. Br J Urol 81:1^16.

Rapidi CA, KarandreasN, Katsifotis C, et al. 2005. A combined urodynamicand electrophysiological study of diabetic cystopathy. Neurourol UrodynNov 2; [Epub]

Susset JG. 1983. Development of nomograms for application of uro£owme-try. In: Hinman F Jr, editor. Benign Prostatic Hyperplasia. New York:SpringerVerlag, 528 p.

Szabo L, Fegyverneki S. Maximum and average urine £ow rates in normalchildren, the Miskolc nomograms. Br J Urol 1995;76:16^20.

UedaT,Yoshimura N,Yoshida O. 1997. Diabetic cystopathy: Relationship toautonomic neuropathy detected by sympathetic skin response. J Urol157:580^4.

Wahl EF, Lahdes-VasamaaTT, Lerman SE, et al. 2001. Prototype system forenhancing cystometric analysis with special emphasis on the pediatricpopulation. J Endourol 15:873^80.

Neurourology and Urodynamics DOI 10.1002/nau

Diabetic Flow 221