Evaluation of Nocturnal Enuresis After Adenotonsillectomy in Children With Obstructive Sleep Apnea: A Secondary Analysis of a Randomized Clinical Trial

Key Points

Question  Do children with nonsevere obstructive sleep apnea experience resolution of nocturnal enuresis after adenotonsillectomy?

Findings  In this econdary analysis of a randomized clinical trial including 393 of the 453 totally enrolled children, those with nonsevere obstructive sleep apnea, those who underwent adenotonsillectomy were more likely to experience resolution of nocturnal enuresis compared with children who did not receive surgery. Nocturnal enuresis was less frequent in girls; other clinical factors, such as age, obesity, and severity of sleep apnea, were not associated with improvement of nocturnal enuresis.

Meaning  The findings of this study suggest that it may be useful for clinicians to inquire about nocturnal enuresis in children with obstructive sleep apnea and counsel caregivers on the potential benefit associated with adenotonsillectomy.

Importance  Children with obstructive sleep apnea (OSA) are at increased risk for nocturnal enuresis (NE). However, randomized clinical trials assessing NE outcomes in children randomized to adenotonsillectomy (AT) vs watchful waiting are lacking.

Objective  To assess the outcomes of AT vs watchful waiting in children with nonsevere OSA who experience NE.

Design, Setting, and Participants  Secondary analysis of data from a multicenter randomized clinical trial conducted at tertiary children’s hospitals was performed. Participants included 453 children aged 5.0 to 9.9 years with nonsevere OSA who were randomized to either watchful waiting or AT as part of the multicenter Childhood Adenotonsillectomy Trial. Caregivers completed the Pediatric Sleep Questionnaire, which includes a binary item on bed-wetting, at baseline and 7-month follow-up. The trial was conducted between October 2007 and June 2012. Evaluation in this secondary analysis involving NE occurred from October 2019 to March 2021.

Interventions  Adenotonsillectomy vs watchful waiting in children with NE.

Main Outcomes and Measures  Prevalence of NE as defined by parental response to the Pediatric Sleep Questionnaire bed-wetting item at baseline and 7-month follow-up.

Results  Of the 453 children enrolled, 393 were included in analysis; of these, 201 were girls (51.1%). Mean (SD) age at baseline was 6.54 (1.40) years. At baseline, the number of children with NE was similar (2.6%; 95% CI, −0.12% to 0.07%) between the AT (59 [30.7%]) and watchful-waiting (67 [33.3%]) groups. The odds of NE in the watchful-waiting group were approximately 2 times higher than the AT group after 7 months (odds ratio, 2.0; 95% CI, 1.3 to 3.1). Following AT, there was a decrease (−11.0%; 95% CI, −16.3% to −5.7%) in the number of children with NE (n = 38). The prevalence of NE did not change significantly (−0.5%; 95% CI, −5.4% to 6.4%) in the watchful-waiting group (n = 66) at follow-up. Although NE was less frequent in girls (adjusted odds ratio, 0.53; 95% CI, 0.33-0.85), other clinical factors, such as age, race and ethnicity, obesity, and apnea-hypopnea index, were not associated with improvement of NE.

Conclusions and Relevance  In this secondary analysis of a randomized clinical trial, AT for the treatment of pediatric OSA appears to result in improvement in NE. Further research is needed to assess whether AT is associated with long-term benefits for NE compared with watchful waiting.

Trial Registration  ClinicalTrials.gov Identifier: NCT00560859

Nocturnal enuresis (NE), also called bed-wetting, involves involuntary urination during sleep. Nocturnal enuresis is the second most common chronic medical condition affecting children and is present in approximately 5% to 10% of otherwise healthy children aged 7 years.¹ Nocturnal enuresis is defined as discrete episodes of urinary incontinence that occur during sleep in children at least aged 5 years.² Nocturnal enuresis is associated with diminished self-esteem and the emergence of childhood psychological problems.^3,4 Although NE has a spontaneous resolution rate of 15% per year, approximately 3% of NE cases will persist into adulthood.⁵

The source of NE is multifactorial and features increased nocturnal urine production, failure to awaken from sleep, and bladder hyperactivity.⁶ One known risk factor for NE is obstructive sleep apnea (OSA), a disease characterized by upper airway collapse during sleep.^5,7 Obstructive sleep apnea has been linked to decreased quality of life, neurocognitive and behavioral impairment, and the development of cardiovascular disease.^8-10 Studies have reported that children with OSA have a higher prevalence of NE than their healthy peers.^10-12 The mechanism linking OSA to NE has not been fully elucidated; however, it is proposed that airway obstruction leads to changes in intrathoracic pressure that cause activation of cardiopulmonary and renal neuroendocrine reflexes.¹³

First-line therapy of NE is behavioral modification in conjunction with either bed alarm or desmopressin therapy. However, adherence to these management strategies is often poor, resulting in refractory NE.³ The primary treatment for children with OSA is adenotonsillectomy (AT). Research has pointed to a reduction in NE in children with OSA following AT.^2,14,15 However, given that the natural history of NE is characterized by spontaneous resolution, the lack of randomized studies on this topic makes it difficult to determine whether AT confers additional benefit. We sought to better understand the interplay between bed-wetting and pediatric OSA by analyzing data from the Childhood Adenotonsillectomy Trial (CHAT).¹⁶ The primary objective of our project was to assess the outcomes associated with AT in children with NE and nonsevere OSA. We hypothesized that children undergoing AT would be more likely to experience improvement in NE than children managed with watchful waiting. A secondary aim of the present study was to examine which factors may be associated with resolution of NE among children with nonsevere OSA.

We analyzed data from CHAT, a large multicenter randomized clinical trial assessing the effects of AT in children with nonsevere OSA on various neurocognitive, behavioral, and polysomnogram (PSG) outcomes conducted from October 2007 to June 2012, and participants were randomized between January 2008 and September 2011.¹⁶ Evaluation in this secondary analysis involving NE occurred from October 2019 to March 2021. Children aged 5 to 9.9 years with parental report of habitual snoring and nonsevere OSA (ie, apnea-hypopnea index [AHI] ranging from 2 to 30 or obstructive apnea index ranging from 1 to 20) confirmed by nocturnal laboratory-based PSG were eligible for inclusion. To be enrolled in the study, patients also had to have tonsillar hypertrophy and be deemed a surgical candidate for AT. Children with attention-deficit/hyperactivity disorder, psychiatric or behavioral disorders, and substantial medical comorbidities, such as craniofacial anomalies and hypotonia, were excluded from CHAT. Children were randomized to AT or watchful waiting and completed a variety of assessments at baseline and 7-month follow-up, including surveys to assess symptom burden, behavior, and PSG. Randomization, stratified by site, age (5-7 or 8-10 years), race and ethnicity (Black or other [no specification]), and weight (≤95th or >95th percentile of body mass index), was performed using a web-based procedure.

This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline. Written informed consent was provided by caregivers for CHAT participants. In addition, children aged 7 years or older provided written assent. The trial protocol is available in Supplement 1. A detailed summary of the CHAT methods has been published, and the study data are publicly accessible for scientific research.¹⁷

The Pediatric Sleep Questionnaire (PSQ), a validated survey assessing sleep-disordered breathing symptom burden,¹⁸ was administered at enrollment and 7 months. The instrument consists of 22 items for which the caregiver indicates either yes or no. Children with a score greater than 0.33 on the PSQ are considered to have a high symptom burden. One of the PSQ questions specifically pertains to NE: “Does your child occasionally wet the bed?” Children were considered to have NE if the caregiver responded yes to this item at baseline. At 7-month follow-up, resolution of bed-wetting was documented if the caregiver responded no to the bed-wetting item. Children were excluded if a response for the bed-wetting item was not available at baseline or follow-up. Demographic and clinical factors were examined to assess indicators of NE resolution.

The initial sample size for CHAT was based on the original primary outcome of change in NEPSY attention/executive function between baseline and the month 7 evaluations.¹⁷ Continuous variables are presented as mean (SD), median, and interquartile range. Categorical variables are presented as frequency and percentage. The Mann-Whitney test was used to compare age at measurement, body mass index, AHI, and oxygen nadir between the AT and watchful-waiting groups. The χ² test was used to compare sex and PSQ score greater than 0.33 between the AT and watchful waiting groups. A generalized linear model was conducted to analyze the outcomes associated with several variables, including sex, race and ethnicity, age, weight (z score), baseline AHI, baseline AHI greater than the study median, follow-up AHI greater than 2, and baseline and follow-up PSQ scores on the likelihood of bed-wetting at follow-up. The odds ratios (ORs) and 95% CIs were computed using the generalized linear model. The reported ORs were adjusted (aOR) for the effect of treatment and sex. All statistical analyses were implemented using the SPSS, version 26 (SPSS Inc). All statistical tests were 2-sided, and P < .05 was considered as statistically significant.

Four hundred fifty-three children underwent randomization in CHAT between January 2008 and September 2011. Three hundred ninety-three children (192 patients in the AT arm and 201 in the watchful-waiting arm) had baseline and follow-up responses for the PSQ bed-wetting item (Figure 1). Of the 393 children, 201 were girls (51.1%) and 192 were boys (48.9%); mean (SD) age at baseline was 6.54 (1.40) years. Other demographic data are displayed in Table 1; there were no significant baseline differences between the AT and watchful-waiting groups in terms of sex (7.9%; 95% CI, −0.02% to 0.18%), age (median, 0; 95% CI, −0.25 to 0.25 years), race and ethnicity (Black: −1.6%; 95% CI, −0.12% to 0.09%; other, 2.5%; 95% CI, −0.04% to 0.09%; White: −0.9%; 95% CI, −0.11% to 0.09%), or body mass index (median z score, 0.10; interquartile range, −0.06 to 0.35). Most CHAT participants were of Black race (210 [53.4%]), and 132 (33.6%) children were obese. Most (249 [63.3%]) children in the study had nonsevere OSA, with a median AHI of 4.75 (interquartile range, 2.72-8.75) in the AT group and 4.50 (interquartile range, 2.57-8.84) in the watchful-waiting group.

At baseline, the rate of NE was similar between the 2 groups, with caregivers reporting NE in 59 children (30.7%) in the AT group and 67 children (33.3%) in the watchful-waiting group (2.6%; 95% CI, −0.12% to 0.07%). The prevalence of NE in the AT group after 7 months improved to 19.8% (n = 38), which represents an 11.0% decrease (95% CI, −16.3% to −5.7%). There was no significant change (0.5%; 95% CI, −5.4% to 6.4%) in the prevalence of NE (66 [33%]) in patients in the watchful-waiting group (Figure 2). The odds of persistent NE in the watchful-waiting group were approximately 2 times higher than that of the AT group at follow-up (aOR, 1.98; 95% CI, 1.25 to 3.14). There was a significant difference in the percentage of children in whom the caregivers reported new bed-wetting (ie, no NE at baseline with NE developing during the study) (Table 2) at 7-month follow-up between the AT and groups (95% CI, 3% to 18%). Compared with the AT group, children in the watchful-waiting group were 5 times more likely to develop new bed-wetting (aOR, 4.98; 95% CI, 1.58 to 20.80).

Children who underwent AT had better outcomes than those who were observed with no intervention in terms of PSG parameters and symptom burden after 7 months. In addition, children in the AT group were more likely to experience resolution of NE (aOR, 1.75; 95% CI, 1.08-2.81). Nocturnal enuresis was reported less frequently among girls compared with boys (aOR, 0.53; 95% CI, 0.33-0.85). Additional demographic factors and PSG variables, including body mass index, age, race and ethnicity, and baseline AHI, were not associated with persistent NE at follow-up. Children in the AT group who continued to have NE after surgery were more likely to have a PSQ score greater than 0.33 compared with those who experienced NE resolution (aOR, 2.43; 95% CI, 1.13-5.25). There was not an association between follow-up AHI greater than 2 and persistent NE. There were no adverse events associated with this analysis.

To our knowledge, this is the only study to examine bed-wetting outcomes between children randomized to AT vs watchful waiting for the management of nonsevere OSA. Our finding that children with OSA who undergo AT appear to be nearly 2 times more likely to experience resolution of their NE compared with those who do not have surgery suggests a modest effect size and is clinically important. Clinicians can use this information when counseling parents about the benefits vs risk for surgical intervention for children with nonsevere OSA. In pediatric patients with nonsevere OSA between ages 5 and 12 years in whom bed-wetting has been a chronic, frustrating problem, AT may be considered.

The pathophysiologic link between OSA and NE is still being investigated. It has been postulated that airway obstruction reduces intrathoracic pressures leading to release of atrial natriuretic peptide and brain natriuretic peptide.^13,19 Elevated levels of these hormones can inhibit antidiuretic hormone and lead to increased urine production during sleep.¹³ Antidiuretic hormone levels have been shown to increase following AT.¹⁹

Previous studies have suggested increased rates of NE in children with snoring and OSA. A large, community-based survey of 20 000 school-aged Japanese children found that habitual snoring was associated with a higher prevalence of NE in an age-adjusted analysis.⁷ Further supporting the link between OSA and NE, a review reported a prevalence of NE of 33% in children with sleep-disordered breathing.¹² In this analysis of the CHAT database, the percentage of children with OSA reporting NE was similar (AT, 30.7%; watchful waiting, 33.3%). This prevalence is substantially higher than the approximate 14% prevalence of NE in otherwise healthy school-aged children. In conjunction with previously published data, the findings of the present study highlight that NE is a common problem in elementary school-aged children with OSA. It may be useful for clinicians to inquire about bed-wetting in children presenting for OSA evaluation.

The association between AT and NE outcomes in children with OSA is still being investigated. A review article described a significant decrease in the prevalence of NE in children following AT from 31% (426/1360) to 16% (95/587).¹² However, only 1 of the studies included in that analysis included a control group. The study with the control group did not show that AT conferred benefit for NE resolution; rather, the rates of NE were similar between the control and AT groups at 6-month follow-up.¹⁹ However, this study was limited by inclusion of children aged 3 years and lack of PSG use to diagnose OSA. The paucity of randomized clinical trials on this topic makes it difficult to assess the role of AT in children with NE because spontaneous resolution can occur in these patients.

In line with Jeyakumar et al’s¹² review on the association between AT and NE, our analysis suggests that AT is associated with decreased bed-wetting 7 months following AT. Earlier studies have identified several risk factors for NE in children, including male sex and family history of NE.¹¹ Data regarding risk factors for persistent NE are lacking. In the present analysis, we found that girls were more likely than boys to experience resolution of NE at 7-month follow-up. This finding is in line with previous research that demonstrated an association between NE and sleep-disordered breathing in boys but not girls.⁷ Several factors have been proposed to account for these differences, including varying renal sensitivity to vasopressin and sex hormone effects.^20,21 Kovacevic et al¹³ reported that children with OSA who had higher arousal index and obstructive AHI on baseline PSG were more likely to experience NE resolution following AT than those with lower index scores. In their study, sex, race and ethnicity, body mass index, history of constipation, and preoperative number of wet nights weekly did not estimate the probability of NE outcomes. In our present study, neither baseline OSA nor persistent OSA at follow-up was associated with NE. We found that children with NE at follow-up were more likely to have PSQ scores greater than 0.33. This finding suggests that persistent NE may be associated with an ongoing high symptom burden in children.

Strengths of this study include the large number of participants in both the experimental and control arms from multiple institutions and the use of the PSG in all subjects to diagnose OSA. In addition, the prospective nature of the study limits recall bias. Children younger than 5 years in whom bladder maturity may not yet have been reached were excluded.

The study has limitations. CHAT was a nonblinded trial, so parental bias in reporting presents a limitation. In addition, the study lacks information regarding the severity of NE. The PSQ does not provide information on the amount of bed-wetting that occurs (ie, nights per month). Thus, we were not able to examine whether additional children may have experienced improvement in bed-wetting even if resolution was not obtained. Furthermore, the PSQ fails to distinguish between primary and secondary NE. Although most literature has focused on primary NE, Weider et al²² reported resolution of secondary NE (bed-wetting that started following a dry period of at least 6 months) in children with sleep-disordered breathing who underwent AT. In our study, children in the watchful-waiting group were more likely to develop new-onset NE at 7-month follow-up compared with those in the AT group. Further studies are needed to assess whether outcomes following AT differ for children with primary vs secondary NE. Long-term data on this topic are lacking; it is unclear whether the differences between the surgery and watchful-waiting groups will remain evident in long-term follow-up. In addition, information is lacking regarding other factors that could influence NE outcomes, such as comorbid medical conditions and whether medical treatment of NE was pursued during the study. Further study is needed to better clarify which patient factors are the best in estimating NE resolution following AT. A prospective study that compares AT with alternative NE treatments, including behavioral modification and bed-wetting alarm, is needed to help further guide management protocols.

This secondary analysis of CHAT noted that, over a 7-month follow-up period, children with nonsevere OSA who underwent AT were more likely to experience resolution of NE than those who were observed but received no intervention. It may be useful for clinicians to inquire about bed-wetting in children with OSA and counsel caregivers regarding the benefit surgery may confer. Further research is needed to compare NE outcomes in children managed with AT vs behavioral modification and bed-wetting alarm.

Accepted for Publication: July 22, 2021.

Published Online: September 9, 2021. doi:10.1001/jamaoto.2021.2303

Corresponding Author: Cristina M. Baldassari, MD, Department of Otolaryngology–Head and Neck Surgery, Eastern Virginia Medical School, 600 Gresham Dr, Ste 1100, Norfolk, VA 23507 (baldassc@gmail.com; cristina.baldassari@chkd.org).

Author Contributions: Dr Baldassari had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Snow, Baldassari.

Acquisition, analysis, or interpretation of data: Snow, Vazifedan.

Drafting of the manuscript: All authors.

Critical revision of the manuscript for important intellectual content: Snow, Vazifedan.

Statistical analysis: Snow, Vazifedan.

Administrative, technical, or material support: Baldassari.

Supervision: Baldassari.

Conflict of Interest Disclosures: None reported.

Data Sharing Statement: See Supplement 2.