Nocturia in patients with cognitive dysfunction: a systematic review of the literature

The MEDLINE, EMBASE, COCHRANE and CENTRAL databases were used with specific keywords (MeSH or Emtree terms) combined with Boolean operators. The complete search strategy is available as supplementary data (Additional File 1). The search was completed in November 2019 and no date limits were applied. No limit of language was applied. References cited in systematic reviews or meta-analyses were searched. Once key articles were identified, additional searches were performed on PubMed using the “Related Articles” search feature.

A flowchart of the database search and selection process is provided in Fig. 1. Of the 1023 non-duplicate articles identified from the databases, 64 articles were eligible for systematic review. A total of 8 studies were included in the final analysis.

Study characteristics are summarized in Table 1. The 8 studies included all featured a cross-sectional study design and community-dwelling participants. Four studies specifically enrolled older patients [17, 21‐23], 3 studies [18‐20] were conducted in Parkinson’s disease (PD) patients and 1 enrolled participants with probable Alzheimer’s disease (AD) [11]. Cognitive dysfunction was assessed by Mini-Mental State Examination (MMSE) in 6 studies [11, 17, 20‐23], and also the Clinical Dementia Rating (CDR) scale in 1of these studies [11]. The Montreal Cognitive Assessment was used in 1 study [18], and a specific validated scale for impulsivity in PD in 1 study [19]. The prevalence of cognitive dysfunction ranged from 4 to 58% in the studies which reported this parameter [17, 18, 23]. Nocturia was defined as self-reported awakening from sleep to pass urine 2 or more times per night [21‐23], or without any definite threshold [11, 17] or by symptom questionnaires [18‐20]. Overall, nocturia was reported in 19 to 61% of study participants [17, 20].

The association between cognitive dysfunction and nocturia was statistically significant in 6 studies [11, 17, 18, 21‐23], all in the same direction (Table 2). In 1 study [18], nocturia was more frequently identified in PD patients with cognitive dysfunction compared to those without (56% vs. 37% respectively). In 4 other studies [17, 21‐23], a higher MMSE score was a protective factor for nocturia, whereas Jung et al. (2017) did not find any correlation between MMSE score and nocturia in patients with probable AD. Lee et al. (2012) determined that a higher MMSE score was an independent protective factor after adjustment for multiple potential confounders (OR 0.6, 95% CI [0.5–0.9], p = 0.006). In the study by Zhang et al. (2016), this association was independent in multivariable analysis (OR 1.7, 95%CI [1.1–2.8], p = 0.03), but it should be noted that the group studied the entire urinary domain (including nocturia) of the Non-Motor Symptoms Scale (opposed to nocturia as a standalone entity). Minimal clinically important difference in cognitive scores was reached in two of the three studies where assessment was possible [17, 22].

The risk of biases is summarized in Fig. 2. Regarding the selection of participants, the risk of bias was unclear in all studies, since no criteria could be applied to cross sectional studies using the RoBANS instrument. Nevertheless, 3 studies [21‐23] randomly selected their participants, with a response rate ranging from 75 to 98%. Of the 8 studies included, only 3 controlled their analyses for confounding factors [18, 22, 23]. Potential confounding factors considered included age (in all of these studies), ethnicity [23], gender [18], education (in all of these studies), PD history (age of onset, disease severity assessed by the Unified PD Rating Scale, hallucination or gastrointestinal symptoms) in the study by Zhang et al. (2016), urological history (transitional volume zone, history of benign prostate hyperplasia, alpha blockers intake) in the study by Lee et al. (2012) and comorbidities such as sleep or fatigue symptoms [18], lower limbs edema, hypertension, obesity or urinary incontinence [23], depression [18, 22] and anxiety [18]. When gender was not considered, analyses were stratified according to it. All but 1 of the studies [23] were deemed to be at low risk for bias in the measurement of exposure. Except for one study [21], no studies published information on blinding of outcome assessments or incomplete data. Although no studies referenced a published protocol with pre-defined outcomes, expected outcomes were reported in the results section of all 8 studies.

To date, only 8 cross-sectional studies involving older populations, patients with PD, or patients with AD have assessed the association between cognitive dysfunction and nocturia. The overall risk of bias of these studies was unclear. All 6 studies that identified a statistically significant association between nocturnal voiding frequency and cognitive dysfunction did so in the same direction [11, 17, 18, 21‐23]. This association was shown to be independent in 2 studies. Accordingly, current literature suggests that a lower cognitive function is associated with a higher risk of nocturia, although more research is indeed needed to further elucidate the relationship between cognitive dysfunction and nocturia.

Although several studies have found a significant association between nocturia and cognitive dysfunction, the clinical relevance of these associations must be questioned.

This systematic review is the first on the topic of cognitive dysfunction and nocturia. We identified studies which included nearly 4500 older patients in total.

However, the included study populations were heterogeneous, and varied in the means by which potential confounding factors were assessed. Indeed, age, sleep disorders, cardio-metabolic and central nervous system diseases are the main confounding factors in the relationship between cognitive disorders and nocturia (Fig. 3). If age is considered in each study including a multivariable analysis, only Zhang et al. (2016) have considered sleep disorders, cardio-metabolic diseases were partially considered in the study by Burgio et al. (2010) and neurological diseases were ignored if we exclude the study in PD patients [18]. Moreover, the definition and assessment of nocturia (e.g., self-reported voiding frequency vs. symptom questionnaire data) varied across studies, which was consistent with a previous study [25]. At the time of publication, none of the studies included in this review used the exact terminology recommended by the International Continence Society, which states that each void is preceded and followed by sleep or the intention to sleep [2, 26]. Nevertheless, several studies used the clinically significant threshold of two or more nocturnal voids [17, 20‐23]. In addition, none of the studies assessed nocturia on frequency volume charts, but only on validated questionnaires or by history taking. This may have led to a risk of misclassification bias. Indeed, it has been shown that nocturia prevalence is overestimated when using questionnaires [27, 28]. Mis-estimating (over- or under-estimating) nocturnal voiding frequency to a greater extent in patients with cognitive dysfunction compared to those without could lead to this bias. Regarding cognitive dysfunction, the MMSE, was used in most of the included studies. It is a short, patient-friendly instrument, but may be influenced by educational attainment and age. These limitations, coupled with the relatively small number of studies which met the criteria for inclusion, precluded quantitative synthesis of study results.

Although several studies have found a significant association between nocturia and cognitive dysfunction, the clinical relevance of these associations must be questioned. While it is difficult to estimate for some studies, it appears to have been observed in others. For example, in the study by Lee et al. (2012), the observed difference in MMSE between patients with nocturia and those without nocturia was 1.5 points, which is considered clinically significant [29].

The fact that only cross-sectional studies met the criteria for inclusion in this systematic review precludes identification of the precise cause-and-effect relationship between cognitive dysfunction and nocturia. Nevertheless, several non-mutually exclusive hypotheses may explain this association (Fig. 3).

Cognitive dysfunction and nocturia share many of the same risk factors (Part (a) of Fig. 3) including central nervous system pathologies [30], altered circadian rhythms of key hormones [15, 31, 32], reduction in sex hormones [33, 34], sleep disorders [15, 34], comorbidities [13, 34], polypharmacy and modifiable lifestyle factors [13, 34, 35]. Furthermore, cognitive dysfunction could give rise to nocturia (part (b) of Fig. 3), mainly through impaired executive function, which is associated with overactive bladder syndrome [11, 12], and decrease in physical activity [36]. These two manifestations are known causes of nocturia [3, 37]. Conversely, nocturia could give rise to cognitive impairment (part (c) of Fig. 3). Antimuscarinic agents use, a first-line therapy in the management of overactive bladder, may contribute to cognitive dysfunction [38] through an increase of the anticholinergic load. Moreover, nocturia-induced sleep impairment might lead to cognitive decline [13], via factors such as benzodiazepine use [15].