Circadian Patterns in Postvoid Residual and Voided Percentage Among Older Women with Urinary Incontinence.

BACKGROUND
Women with urinary incontinence incur an increased risk of elevated postvoid residual (PVR) volume and impaired voiding efficiency (i.e., voided percentage (Void%)), but the clinical significance of these parameters remains poorly described. Further characterization of PVR and voiding efficiency may thus be useful in refining the evaluation and management of urinary incontinence. This study aims to explore possible circadian variations in PVR and Void% in older women with stress (SUI), urge (UUI) and mixed urinary incontinence (MUI).


METHODS
A single center prospective study which enrolled a convenience sample of 90 older women who consulted a tertiary referral hospital for urinary incontinence. Participants underwent an extensive medical interview and were hospitalized to complete a 24-h frequency-volume chart (FVC) with PVR measurement after each void (FVCPVR).


RESULTS
FVCPVR analysis demonstrated no differences in mean PVR and Void% between patients with SUI, UUI and MUI. Likewise, no daytime or nighttime differences were observed in mean PVR or Void% within or between groups.


CONCLUSIONS
No evidence of circadian variation in PVR or Void% was observed in older women with SUI, UUI or MUI.


Introduction
Postvoid residual (PVR) is the volume of urine that remains in the bladder after voluntary micturition [1]. An increased PVR (i.e., incomplete bladder emptying) may be a cause of lower urinary tract symptoms (LUTS) such as urgency, frequency, incontinence and nocturia and, in some cases, may also contribute to upper urinary tract dysfunction [2]. The voided percentage (Void%), defined by the International Continence Society (ICS) as the proportion of bladder content emptied [3], is even more strongly associated with peak flow rate (Qmax) [4], and may thus also be a clinically relevant variable in the evaluation of LUTS mediated by inefficient voiding.
Although there is no consensus as to what volume constitutes an "elevated" PVR [5], and the specific cutoffs employed may vary across different LUTS, values of 50-100 mL are commonly recognized as the floor threshold for abnormal [6], as PVR values in excess of 100 mL are exceedingly rare in middle-aged and older women [7]. Despite some heterogeneity across study populations, PVR values greater than 100 mL may be relatively more prevalent among women with stress urinary incontinence [8] and urge urinary incontinence [9].
Problematically, across all LUTS, significant intra-individual variability in PVR volumes and Void% generally exists, which hinders their potential diagnostic utility in LUTS management [2]. Recent research of uroflowmetry parameters in women with different subtypes of urinary incontinence indeed demonstrated high intra-individual variability in PVR volumes, but did not account for the potential influence of time-of-day on study results [10], which has been recognized as a potential confounding variable in the association between LUTS severity and PVR [11].
Further characterization of the interplay between time-of-day, PVR and voiding efficiency may thus be useful in refining the evaluation and management of urinary incontinence. Although frequency-volume chart (FVC) data has, to our knowledge, never been recorded with PVR across a 24-h period, previous small-scale research involving geriatric subjects with incontinence reported significantly higher intrasubject PVR values in the early morning compared to the afternoon or evening [11]. Accordingly, we aimed to test the hypothesis that PVR is greater in the nighttime vs. daytime in women with urinary incontinence. Secondly, we aimed to characterize circadian variations within and between subgroups of women with stress (SUI), urge (UUI) and mixed (MUI) urinary incontinence.

Study Design and Participants
This study undertook a post hoc analysis of unpublished data from the Think Dry cross-sectional descriptive study of urinary incontinence in older adults. The study population consisted of a convenience sample who consulted a tertiary referral hospital for urinary incontinence from December 2013 to December 2018. Included were patients aged ≥ 65 years with a chief complaint of urinary incontinence. Exclusion criteria were a positive screen for cognitive impairment on the Mini-Cog [12], symptomatic urinary tract infection and incontinence associated with recurrent infection, pain, hematuria, pelvic irradiation, radical pelvic surgery, suspected fistula or urinary retention. Local ethics committee approval (2013/950) was obtained, the Declaration of Helsinki was followed, and all participants provided written informed consent.

Study Protocol
All Think Dry participants completed a 24-h frequency-volume (FVC) chart and PVR (FVC PVR ), a renal function profile (RFP) [13] and urodynamic testing. The data amassed were used by a functional urology specialist (K.E.) to establish specific diagnoses. Participants also completed several validated questionnaires, which were used to characterize their functional status (Belgian-modified Katz Index of Independence in Activities of Daily Living (ADL) [14]), frailty (Tilburg Frailty Indicator (TFI) [15]), fall risk (St. Thomas's Risk Assessment Tool in Falling Elderly Inpatients (STRATIFY) [16]) and the impact of LUTS on their quality of life (International Consultation on Incontinence modular Questionnaire for LUTS in women (ICIQ-fLUTS) [17]).
For the present analysis, female patients with a diagnosis of urodynamic stress incontinence and/or detrusor overactivity who completed a FVC PVR were included, such that 11 patients were excluded because urinary incontinence was not reproduced during urodynamic testing.

Materials
The Katz Index dichotomizes independence/dependence for six ADLs (bathing, dressing, transferring, toileting, continence and feeding). Scores are summated, such that a score of six reflects fully preserved function, four indicates moderate impairment, and ≤2 indicates severe functional impairment [18].
The TFI is a 15-item questionnaire assessing physical, psychological and social frailty, with increasing scores reflecting increased impairment in these domains. For this instrument, the maximum total score is 15, and patients with a score ≥ 5 may be deemed frail [15].
The STRATIFY instrument consists of five items that evaluate risk factors for falling. Scores range from zero to five, with a score ≥ 2 indicating high fall risk [16].
All FVC PVR measurements were obtained by nurses of the inpatient urology department. Nurses recorded the voided volume (VV) (mL), urinary incontinence (UI) weight (g) and PVR (mL) accompanying each micturition, as well as patients' time to bed with the intention of sleeping and time of awakening. PVR was analyzed noninvasively using a validated portable ultrasound device (BladderScan®BVI 9400, Verathon, Inc., Bothell, WA, USA) within 15 min after each micturition [19]. Patients were permitted to consume fluids and solids ad libitum in an effort to best simulate real-world conditions.
Measurements derived from the FVC PVR were defined in accordance with reports from the Standardization subcommittee of the ICS [1,20]. Namely, global polyuria was defined as a 24-h urine output > 40 mL/kg; and nocturnal polyuria (NP) was defined as a nocturnal urine volume > 33% of the total 24-h urine volume in the absence of global polyuria. Nocturia was defined as at least 2 nocturnal voids, because this is the threshold at which most patients are more likely to report clinically significant nocturia-related bother and experience impaired health-related quality of life [21]. Voided percentage (Void%) was calculated as {(VV/[VV + PVR]) × 100} [3]. Missing values were not estimated or replaced, except in the case of one missing daytimevalue, for which values were replaced with corresponding mean daytime values for that patient (n = 9).
Urodynamic studies were conducted according to local protocols, which were guided by current ICS recommendations [22].

Statistical Analysis
Between incontinence subgroups, continuous and categorical variables were compared using the Kruskal-Wallis and Fisher's exact tests, respectively. A pairwise comparison using the Mann-Whitney U and Fisher's exact tests for continuous and categorical variables, respectively, was used to establish partial order between pairs when significant differences were identified on three-way analysis, with a Bonferroni correction (p < 0.017) applied for multiple comparisons. The Wilcoxon Signed-rank test was used to assess intragroup differences in PVR and Void%. Continuous variables are reported as median (confidence interval), and categorical variables are reported as frequency (percent). Data were analyzed using Statistical Package for Social Sciences (Version 24, IBM Corp., Armonk, NY, USA).

Study Population
A total of 90 patients were included, for which the median age was 76 (72-80) years and parity was 2 (1-3) children ( Table 1). Most of the participants (98%) lived in the community, while two (2%) were nursing home residents. The distribution of hysterectomy (p = 0.031), depression (p = 0.048), neurological lesions (p = 0.023) and utilization of selective serotonin reuptake inhibitors (SSRIs) (p = 0.032) were recorded differed between groups. History of hysterectomy was more prevalent in patients with SUI compared to UUI and MUI (57% vs. 28% and 27%, p = 0.019 and p = 0.026, respectively). Patients with UUI were more likely to be affected by depression compared to SUI (25% vs. 4%, p = 0.033). Neurological disorders were more prevalent in patients with UUI compared to SUI (31% vs. 4%, p = 0.008). SSRI utilization was greater among patients with UUI vs. SUI (28% vs. 4%, p = 0.017). No further significant pairwise differences were observed for these parameters.  For questionnaire data, the distribution of functional status (Katz Index) and frailty (TFI) significantly differed between subgroups. (Table 2). A greater proportion of patients with SUI reported a well-preserved ADL function compared to patients with UUI (93% vs. 47%, p < 0.001). Consistently, fewer patients with SUI reported severe ADL impairment compared to those with UUI (0% vs. 31%, p < 0.001). Frailty on the TFI was more prevalent among the UUI subgroup compared to patients with SUI (85% vs. 52%, p = 0.010). No further three-way or pairwise differences were observed in questionnaire analysis.

Discussion
To our knowledge, this is the first study to utilize the 24-h FVC with PVR volume to characterize the circadian pattern of PVR and Void% in older women with SUI, UUI and MUI. In the present analysis, no significant circadian variations were observed in mean PVR or Void% within subgroups.
Abnormal residual volume is a common phenomenon in the setting of urinary incontinence. Although there is no consensus as to what exactly constitutes an elevated PVR, many experts have proposed values between 50-100 mL as the lower threshold for abnormal residual urine volume [24]. In their analysis of PVR in middle-aged women with urodynamic stress incontinence, Tseng et al. recognized that more than one-third of participants had a PVR > 50 mL, and nearly one in five experienced a PVR > 100 mL [8]. Consistently, in the present analysis, a maximum daytime PVR >50 mL was identified in more than half of all participants, and a maximum daytime PVR > 100 mL was recorded in one-third of all subjects.
Importantly, however, while more than one-fourth of all participants in the present study demonstrated mean 24-h PVR > 50 mL, only 7% of all patients experienced a PVR > 100 mL. Thus, single point-of-care PVR measurements may overestimate the true prevalence of an abnormal residual urine volume, particularly when higher cutoffs (e.g., PVR ≥ 100 mL) are applied. Differences in maximum daytime PVR and mean 24-h PVR are most likely a function of the number of measurements, as significant interindividual in PVR has been previously reported [20]. Indeed, this is consistent with data from Saaby and colleagues on the repeatability of PVR ≥ 100 mL among women with uro-gynecological complaints, wherein the prevalence of PVR ≥ 100 was 14%, but declined to 1.3% on repeated measurements [25].
The present post hoc study design is limited by the absence of a prospective power analysis, lack of a concomitant control group and small sample size. Moreover, the convenience sample of patients recruited from a tertiary referral hospital might introduce selection biases in this study and limits the generalizability of this study beyond older women with urinary incontinence. Data about urinary tract infection and stage of prolapse are missing and may be confounding factors that may influence the PVR. Measurement errors cannot be excluded, given that a team of nurses was responsible for the PVR measurements, and interrater reliability was not examined.
In addition, this study relied on 24-h FVCs, which are considered to be less reliable than FVCs of longer duration [26]. Given that there are currently no patient-centered equipment options for home measurement of PVR, the present study design necessitated hospitalization, which is not a viable option in the real-world evaluation of urinary incontinence amongst community-dwelling adults. Future studies should aim to validate our results using three-day FVC PVR instruments and establish precise cut-off values for clinically significant PVR elevations.
Overall, the present study suggests that bladder emptying is comparable in efficacy during the nighttime versus daytime. Moreover, there is no evidence to suggest that measurements of PVR may differentiate the clinical diagnoses of SUI, UUI and MUI in older women. Multiple PVR measurements may be needed to increase intrasubject reliability

Conclusions
The present analysis did not identify significant circadian variation in mean PVR or Void% among older women with stress, urge or mixed urinary incontinence. This study is the first to utilize the 24-h FVC with PVR volume to characterize the circadian pattern of PVR and Void% in older women with SUI, UUI and MUI, which may have bearing on the rational use and interpretation of one-time point-of-care PVR testing