A Cohort Study of Urinary Tract Infections in People Living with Dementia: Epidemiology and Diagnostic Challenges ================================================================================================================= * Raphaella Jackson * Martin Tran * Kirsten Jensen * Michael A. Crone * Alexander J. Webb * Loren P. Cameron * Ramin Nilforooshan * David Wingfield * David J. Sharp * Paul S. Freemont ## Abstract **Background** Urinary tract infections (UTIs) are a leading cause of hospitalization in people living with dementia, making accurate detection and prompt treatment critical in this vulnerable population. **Methods** This retrospective, longitudinal cohort study evaluated the diagnostic accuracy of self-reported symptoms and urine dipstick test results for identifying UTIs in people with dementia. Participants included 78 community-dwelling individuals over the age of 50 with a confirmed dementia diagnosis. These participants were recruited from cohorts established by the Surrey and Borders Partnership NHS Foundation Trust and the Hammersmith & Fulham Partnership Primary Care Network between late 2019 and 2023. **Results** The frequency of UTIs in people living with dementia (PLWD) is highly variable, some individuals having nearly chronic infections and others not acquiring a single UTI during their enrolment in the study. The taxonomy of the microbes associated with the urinary tract infections in this cohort were consistent with those that have been observed in other demographics. When we modelled the relationship between self-reported symptoms, dipstick results, and the presence of UTIs we found that there was no significant relationship between UTI and symptoms. However, there was a highly significant correlation between dipstick test results positive for leucocytes and nitrites with the presence of a UTI. **Conclusions** Urine dipstick tests for nitrites and leucocytes should be incorporated into diagnostic protocols and should also be considered for ongoing surveillance of UTIs in people living with dementia. **Study Registration Numbers** Research Ethics Committee (REC) reference: 19/LO/0102 Integrated Research Application System (IRAS) project ID: 257561 Keywords * Urinary Tract Infection * Dementia * Neurodegenerative Disease * Urine Dipstick Test ## Background At any given time, a quarter of hospital beds in the United Kingdom are occupied by people living with dementia (PLWD). Approximately 20% of these admissions are due to acute conditions, such as infections, where early intervention could prevent more serious progression (1,2). One of the most prevalent of these acute conditions is urinary tract infection (UTI), which accounts for around 9% of the total hospital admissions for people living with dementia (1–3). For urinary tract infections, concomitant to the risk of hospitalization, delayed diagnosis also elevates the risk of increased confusion, episodes of delirium, bloodstream infection, and accelerated clinical decline (4–6), along with potentially worsening chronic neurodegeneration (7,8). It is likely that a major obstacle to timely detection and treatment of UTIs in people living with dementia is presentation of symptoms. UK and European guidelines indicate that individuals should present with localized signs or symptoms of a UTI before diagnosis and treatment (9,10). However, symptoms of infection are often atypical or absent in the elderly, particularly in those with UTIs (11,12). Additionally, the ability of PLWD to recognise or communicate these symptoms may be impaired by the nature of the disease (13). Currently, there are no easily measurable and well-established biomarkers for detecting urinary tract infections in the absence of reported symptoms. The presence of bacteria in the urine is not uncommon in elderly populations (14) and studies on treating low-level persistent asymptomatic bacteriuria have not shown any positive outcomes (15–17). While there has been extensive research on urinary tract infections in many different populations there is limited information on what uropathogens may be affecting people living with dementia. Given the impact of urinary tract infections on both the well-being of this population and the cost to public health it is critical we better understand precisely what is happening in this demographic. In this study, we undertook a longitudinal analysis of urine and urinary pathogens in people living with Alzheimer’s disease, the most common form of dementia. We gathered a variety of molecular, microbiological, and clinical data to evaluate which bacteria are colonising or infecting individuals with Alzheimer’s disease who are still living at home and not in care. We chose to focus on this cohort because studies on this demographic are often restricted to individuals living in care because they are more accessible. ## Methods ### 1.1 Study Cohort This study involves a cohort of participants that have been established by the Surrey and Borders Partnership NHS Foundation Trust clinical care team (SABP) and the Hammersmith & Fulham Partnership Primary Care Network (HFP PCN) between late 2019 and the end of 2023. People with a confirmed diagnosis of dementia (of any type) were recruited through health and social care partners. These included NHS Community Mental Health Teams for older adults (CMHT-OP) and specialist memory services at SABP NHS Trust and within North West London. The study was also advertised on the national ‘Join Dementia Research’ website. Inclusion Criteria for Participants: * Male or female 50 years of age and older at baseline * Confirmed diagnosis of dementia (any type) by specialist assessment * Willing and able to provide informed consent * Standard Mini Mental State Examination (MMSE) score at baseline interview over 12 * Living in the community * Have sufficient functional English to allow completion of the assessment instruments Exclusion Criteria for Participants: * Persons living in a residential care home * Persons with unstable mental state including severe depression, severe psychosis, agitation, anxiety, or active suicidal ideation at screening and baseline. * Persons with severe sensory impairment at screening and baseline. * Persons with an MMSE below 12 * Persons unable to communicate verbally * Persons receiving treatment for terminal illness ### 1.2 Urine collection and symptom reporting Urine samples were collected every 4 to 6 weeks from participants using either sterile urine containers or, for those with fine motor difficulties, a sterile kidney-shaped dish and then transferred to a sterile container. On the day of urine collection, participants completed a questionnaire on UTI-related symptoms (Supplementary Table 1). The urine samples were transported on ice for further analysis. ### 1.3 Dipstick analysis Dipstick analysis on the urine was performed on the Clinitek Status (Siemens Healthcare Diagnostics) using Siemens Multistix 10SG Urinalysis Strips according to the manufacturer’s protocol. ### 1.4 Chromogenic agar plates To numerate the bacterial colonies present in urine samples, 1 µL of urine was cultured on pre-made chromogenic agar plates (Brilliance UTI Clarity Agar, Oxoid Ltd). Plates were incubated for 18-24 hours at 37°C and the resulting colonies were counted and recorded as the number of Colony Forming Units per millilitre (CFU/mL). Counts were only recorded up to 99,000 CFU/mL. All colonies present at over 99,000 CFU/ml were recorded as 100,000+ CFU/mL. ### 1.5 Overnight cultures of bacterial strains Bacterial strains of interest were picked from the chromogenic agar plates and grown overnight in Tryptone Soy Broth (Oxoid Ltd) shaking at 200 rpm at 37°C. 400 µl of overnight culture was mixed gently with 400µl 50% glycerol in 2ml 2D barcoded matrix tubes (Nunc™) and stored at -80°C for future experiments. ### 1.6 16S rRNA gene amplification, sequencing, and classification To confirm the identity of the bacterial strains isolated, 16s rRNA sequencing was undertaken as follows: 1 µL aliquot of the overnight bacterial culture was mixed with 99 µl PBS and incubated at 95°C for 30 minutes to inactivate and lyse the bacteria. The V3 and V4 hypervariable regions of the 16s rRNA gene were amplified by adding 1 µL of the inactivated bacteria samples to a 19 µL reaction with Thermo Scientific DreamTaq PCR Master Mix (2X) using the primers the following primers (final primer concentration of 500 nM): * - Forward: AGGGTTTTCCCAGTCACGACGTTCCTACGGGNGGCWGCA * - Reverse: GACTACHVGGGTATCTAATCC Amplification conditions were as follows: initial denaturation step at 95°C for 5 minutes, followed by 30 cycles of denaturation at 95°C for 30 seconds, annealing at 50°C for 30 seconds and extension at 72°C for 30 seconds, then final extension at 72°C for 5 minutes. A 5 µL aliquot of each amplified product was incubated separately with 2 µL of ExoSAP-IT™ PCR Product Cleanup Reagent at 37°C for 4 minutes to degrade remaining primers and nucleotides, followed by incubation at 80°C for 1 minute to inactivate the ExoSAP-IT ™ reagent. The amplified products were then sent to Eurofins Genomics for Sanger sequencing using the M13 uni (-43) primer. All Sanger sequences were filtered based on QV30 base quality. The sequences were then classified using the silva alignment, classification, and tree service (ACT)(18). Default parameters for the search and classify service were used. ### 1.7 Data Pre-Processing and Classification #### 1.7.1 Participant Symptom Reporting Participant responses to the symptom questionnaire were reviewed manually and re- coded in cases where participants responded with words other than a singular yes or no. Recoding was based on the following principles. 1. Answers which included a yes or no answer with additional exposition or answers synonymous with yes or no were re-coded as yes or no as appropriate. 2. Failure to answer, answers which referred to a medication or unrelated condition as the cause of the symptom, answers indicating the symptom was not a recent development, or responses indicating uncertainty were coded as other. A full table of symptom question text, original answers, and their re-coded responses are available as Supplementary Table 1. #### 1.7.2 Data Analysis The data was imported into and analysed in R. The code used along with a render of the output is included in the supplementary materials. ## 2 Results ### 2.1 Overview of Cohort and UTI Frequency Over a study period of approximately 4 years, 637 routine (approximately every six weeks) urine samples were collected from 91 participants (Figure 1). All participants were still living at home and while some were living alone, the majority (56%) were living with a carer (Figure 1). The gender balance of this cohort was nearly equal, 44% of participants were female, and the most common dementia diagnosis was Alzheimer’s, which accounted for 65% of diagnoses. The average age of participants when they first started submitting urine samples was 82, with an interquartile range from 77 to 88. Because the study has rolling recruitment, the number of samples collected from each participant was not equal. A mean of 7 samples per person was collected, with an interquartile range from 3 to 10 samples. ![Figure 1:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/12/11/2024.12.10.24318775/F1.medium.gif) [Figure 1:](http://medrxiv.org/content/early/2024/12/11/2024.12.10.24318775/F1) Figure 1: Participant Demographics and Samples Collected A Sankey plot showing characteristics of the study cohort including living situation, gender, number of samples collected from each participant, diagnosis, and whether each participant submitted at least one sample that was a UTI sample. The clinicians within our research team decided whether each urine sample was a UTI sample based on urine dipstick results, urine culture results, and participant reported symptoms. Based on these determinations we collected 512 non-UTI samples, 106 UTI samples, and 19 samples where further monitoring or investigation was required to make a determination. Overall, 36% of participants submitted at least one UTI sample (Figure 1). We found that the frequency of UTIs was highly varied among participants (Figure 2). While some participants had chronic UTIs, others did not have a single infection over the course of our study. We grouped participants based on the observed frequency of the UTIs to better understand this phenomenon. For this grouping we considered only participants that had been enrolled in the study 6 months or more. We quantified UTI frequency using 6-week windows, the approximate frequency of sample collection, where a positive window was one where a UTI event had been recorded. Those who had no positive windows were classed as “No UTI”. Those with a ratio of positive to negative windows less than 0.20 were “Low Frequency”, those with a ration between 0.2 and 0.5 “High Frequency”, and those with a ratio greater than 0.50 “Chronic”. We had a total of 8 Chronic, 14 High Frequency, 22 Low Frequency, and 19 no-UTI participants. 28 participants were labelled “Short-Term” and excluded from further classification due to study participation of less than 6 months. ![Figure 2:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/12/11/2024.12.10.24318775/F2.medium.gif) [Figure 2:](http://medrxiv.org/content/early/2024/12/11/2024.12.10.24318775/F2) Figure 2: UTI Frequency in Cohort A plot showing whether a participant had a UTI in each 6-week window following their first sample collection. Each row illustrates the UTI history of one participant. Participants are grouped by their UTI Frequency Type as described in the main text and methods. Participants with less than 6 months since their first sample collection are labelled “Short-Term” and not classified further. ### 2.2 Microbiological Characteristics of Collected Samples Each urine sample collected was characterized in the laboratory using a urine dipstick test and through culture on chromogenic agar. These two steps are standard procedure for the analysis of urine samples if a UTI is suspected. Whilst a urine dipstick test can quantify a range of different biomarkers relevant to different conditions, the two metrics of greatest significance for UTI diagnosis are leucocytes and nitrites (Figure 3). The presence of leucocytes or nitrites in the urine is indicative of a possible UTI, although nitrites are not produced by all bacteria, so the absence of nitrites does not suggest the absence of a UTI. Of the 641 samples collected all but 5 were able to be tested with a dipstick test. The majority, 69%, were negative for both leucocytes and nitrites. This was slightly lower than the proportion of samples which were determined to be non-UTI samples (80%). In only 11 cases (2% of total cases) was the participant believed to have a UTI despite a negative result on the dipstick test for both leucocytes and nitrites. ![Figure 3:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/12/11/2024.12.10.24318775/F3.medium.gif) [Figure 3:](http://medrxiv.org/content/early/2024/12/11/2024.12.10.24318775/F3) Figure 3: Dipstick Test Results A bar plot showing the result of a dipstick test for leucocytes and nitrites for every urine sample that could be analysed. When cultured on chromogenic agar, 392 of the urine samples collected from participants produced bacterial growth. Of these, 1 sample had mixed growth at a level that made counting the individual isolates impossible, but all colony forming units on the remaining 391 plates were counted (Figure 4). The greater the level of observed growth the more likely the sample was to be a UTI sample, however even samples that had pure growth exceeding the standard threshold for UTI (100,000 CFU/mL) were occasionally determined to be non-UTI samples. In many of these cases, the participant had trace or lower levels of leucocytes which may have been indicative that these samples had some fecal contamination. ![Figure 4:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/12/11/2024.12.10.24318775/F4.medium.gif) [Figure 4:](http://medrxiv.org/content/early/2024/12/11/2024.12.10.24318775/F4) Figure 4: Plate Culture Growth A bar plot showing how much total bacterial growth (in CFU/mL) was observed on each chromogenic agar plate after culturing each urine sample. Samples are separated based on whether the growth was pure growth with a single isolate, mixed growth with a dominant isolate accounting for >=70% of total growth, or mixed growth. ### 2.3 Bacteria and Bacterial Pathogens in the Urine of PLWD Across the 392 urine sample cultures that produced bacterial growth there were 771 morphologically distinct bacterial isolates. We used Sanger sequencing of the 16s rRNA gene to determine the taxonomy of approximately half (352) of these isolates to understand what taxa were represented across samples in both UTI and non-UTI contexts (Figure 5A). We found that there was significant variation across taxa in terms of their rate of occurrence in UTI versus non-UTI samples (Fisher Test, p<0.001) and the level of growth they typically had in culture (Fisher Test, p<0.001). Additionally, we identified the most abundant isolate in culture for the UTI samples, with this being the likely cause of the UTI, and compared the rate at which different bacterial taxa appeared (Figure 5B). ![Figure 5:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/12/11/2024.12.10.24318775/F5.medium.gif) [Figure 5:](http://medrxiv.org/content/early/2024/12/11/2024.12.10.24318775/F5) Figure 5: Overview of All Identified Bacterial Taxa Present in Urine Cultures (A) A bar plot showing the taxonomic identity of bacterial isolates resulting from culturing each sample on chromogenic agar. Bars are coloured to reflect the abundance of the isolate in culture. Samples are grouped by whether the isolate occurred in a UTI sample. Please note: this visualization does not identify the causative isolate for UTI samples. Every isolate that was found in a UTI sample, irrespective of whether it was likely to be causative, is grouped under “UTI”. (B) A bar plot showing the taxonomic identity of the most abundant bacterial isolate in UTI samples only. Bars are coloured to reflect the abundance of the isolate in culture. Consistent with research on the primary causes of UTI in other demographics, Escherichia is by far the most commonly associated bacteria with urinary tract infection (19). Other common causes of UTI include Streptococcus, Klebsiella, Enterococcus, Pseudomonas, and Staphylococcus which we also observed. The most common non-UTI associated bacteria were Staphylococcus, which is also the most frequently observed bacterial taxa overall, and Enterococcus. There are two potential explanations for the frequency at which Staphylococcus and Enterococcus are observed, particularly at low levels of growth in culture. These two genera are habitual members of the skin and intestinal microbiome respectively. Their presence could be attributed to low level contamination with urine sample collection, reflective that the urinary tract is more easily colonized by nearby bacteria within this cohort, or a combination of both. It is important to note, however, that both these genera are known to cause urinary tract infections, and we do observe this in several instances in this cohort, so their presence should not be readily disregarded. Finally, we considered what bacteria were causing urinary tract infections in participants with differing UTI frequency (Figure 6). We hypothesized that some species of bacteria might be more likely to cause long-term persistent urinary tract infections that were resistant to treatment. However, we did not observe any clear association between bacterial genera and frequent urinary tract infections. Additionally, we observed that participants affected by chronic urinary tract infections did not appear to be consistently infected with a singular type of bacteria. While some participants, like P061 and P004, were predominantly infected with a singular genus of bacteria, others like P054 had a variety of different bacteria appearing as the dominant isolate in culture across UTI samples. ![Figure 6:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/12/11/2024.12.10.24318775/F6.medium.gif) [Figure 6:](http://medrxiv.org/content/early/2024/12/11/2024.12.10.24318775/F6) Figure 6: Overview of All Identified Bacterial Taxa Present in Urine Cultures A bar plot showing the taxonomic identity of bacterial isolates causing UTI in participants where strains were profiled. Participants are grouped by the overall frequency of UTI they experienced, as outlined in “Overview of Cohort and UTI Frequency”. ### 2.4 Dipstick Superior to Symptoms as an Indicator of UTI With each sample collection we asked participants to respond to a series of seven questions addressing whether they were experiencing symptoms of urinary tract infection. Participants were asked if they were experiencing increased urgency, increased frequency, dysuria, urinary retention, change in urine smell or colour, localized pain, or if they felt generally unwell. We then studied how participant symptom reporting correlated with the type of urine sample given. In the current diagnostic pipeline reported symptoms are treated as the primary indicator of UTI. However, in our cohort symptom reporting was an exceedingly poor indicator of urinary tract infection. Irrespective of UTI status, the majority of the responses were negative for any given symptom (Figure 7). Additionally, symptoms were reported by participants without UTI approximately as frequently as those with UTI. By contrast, dipstick results that were positive for either nitrites or leucocytes appeared to be highly correlated with the presence of urinary tract infection (Figure 7). ![Figure 7:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/12/11/2024.12.10.24318775/F7.medium.gif) [Figure 7:](http://medrxiv.org/content/early/2024/12/11/2024.12.10.24318775/F7) Figure 7: Symptom Reporting and Dipstick by Presence of UTI A) A bar plot showing the proportion of responses to each symptom question based on the type of urine sample submitted(UTI/Non-UTI). B) A bar plot showing the outcome of a dipstick test for leucocytes and nitrites for all samples based on the type of urine sample submitted (UTI/Non-UTI). We modelled the relationship between UTI status, dipstick results, and reported symptoms using generalized estimating equations to explore this relationship further. We found there was no significant association between reported symptoms and the presence of UTI but there were highly significant associations between UTI status and dipstick results (P of Wald Test < 0.001). When we compare the presence of UTI specific symptoms to the presence of leucocytes or nitrites (at levels trace or higher) as predictors of UTI, we found that dipstick results outperform symptoms in predicting UTI in every respect (Figure 8). Rates of true negatives and positives are higher and fewer samples are unquantified, due to reduced ambiguity, in dipstick results as opposed to reported answers. ![Figure 8:](http://medrxiv.org/https://www.medrxiv.org/content/medrxiv/early/2024/12/11/2024.12.10.24318775/F8.medium.gif) [Figure 8:](http://medrxiv.org/content/early/2024/12/11/2024.12.10.24318775/F8) Figure 8: Symptom and Dipstick Results as Predictors of UTI A bar plot showing the number of samples which would be classified correctly or incorrectly as UTI samples based on whether participants were symptomatic or by whether they received a positive dipstick result. Participants were classified as symptomatic if they answered “Yes” to any of the five main symptoms of UTI (Q1-Q5). Participants were considered to have a positive dipstick result if the result for either leucocytes or nitrites was trace or higher. Samples were unquantified for symptoms if the participant’s response to any of the UTI symptom questions was classified as other (and they had not responded yes to any other symptom of UTI) or if they reported a non-UTI specific symptom (Q6-Q7). Samples were unquantified for dipstick if a dipstick result could not be obtained for the sample. ## Discussion ### 2.5 Principal Findings In this study, we characterize the nature and frequency of urinary tract infections in people living with dementia. It is the first study to use a longitudinal routine sampling method to look at how UTIs are affecting individuals in this cohort. It is also unique in how it focuses specifically on people living with dementia who are still living at home, as opposed to living in care facilities. This is noteworthy given how this demographic is less often studied due to increased challenges associated with working with dispersed cohorts in the general population. We found that urinary tract infections are more common in this demographic than in the general population, which is consistent with previous findings (20). However, we additionally found that there is a significant amount of variation in how frequently members of this demographic are acquiring UTIs. While some individuals have nearly perpetual urinary tract infections, despite repeated attempts at treatment, others do not acquire a single infection. Critically, we found that symptoms are not correlated with the presence of urinary tract infections. This is of great relevance given the current pipeline for UTI diagnosis in individuals over 65 years of age in the UK requires the presence of symptoms to diagnose a UTI. Our findings suggest that the current diagnostic criteria will not only fail to detect a significant proportion of urinary tract infections, but also lead to significant expenditure of resources on further testing in individuals without UTI due to high rates of symptom reporting in those without UTI. In contrast, we found that dipstick test results for leucocytes and nitrites are highly correlated with the presence of UTI. Currently Public Health England guidelines discourage the use of dipsticks for individuals over 65 years old, however these results suggest that dipstick results would be of far greater utility then reported symptoms. While both the frequency and detection of UTIs in PLWD may be distinct from the general population, the causes of UTI appear to be consistent across groups. The bacteria most frequently associated with UTIs in our cohort, most notably Escherichia, align well with previous research on the primary causative strains of UTIs (19). ### 2.6 Implications and Future Directions Given these findings, we suggest that our fundamental approach towards detecting urinary tract infections in PLWD needs to change. We cannot continue to rely on participant- reported symptoms to indicate the presence of UTI. People can become extremely ill before there are signs of infection. Indeed, a recent study on bacteremic urinary tract infections caused by E. coli showed elderly individuals frequently presented without urinary symptoms even when sepsis was present by SIRS criteria (11). If reliance on symptoms to distinguish those in need of treatment is flawed, then there is a need to switch to a surveillance model where individuals living with dementia are regularly tested for certain biomarkers. However, it is as yet unclear which biomarkers would serve this purpose. Previous studies have treated asymptomatic bacteriuria based on low-level growth of the same bacterial genera over months (17), but this failed to improve participant mortality. This makes it clear that developing a good methodology and set of biomarkers for the early identification and treatment of urinary tract infections is not trivial. We would, however, propose that we could begin by re-evaluating the use of urine dipsticks for people living with dementia which is currently not recommended by Public Health England guidelines for those over 65 years old (9). We believe that the most critical subject for future study is biomarker discovery for UTIs in PLWD. If we could identify a biomarker, or set of biomarkers, that indicates those who need treatment before they become severely ill, it would reduce patient suffering, greatly lower the burden on the public health system and address issues with inappropriate antibiotic usage. This will undoubtedly take significant resources as potential biomarkers will need to be evaluated using interventional clinical trials. However, given the benefits, the cost is more than warranted. ## Supporting information Supplemental Acknowledgments [[supplements/318775_file02.docx]](pending:yes) Code and Output [[supplements/318775_file03.zip]](pending:yes) Supplementary Table 1 [[supplements/318775_file04.csv]](pending:yes) ## Data Availability All data used in the present work are available upon reasonable request to the authors. ## Declarations ### Ethics approval and consent to participate Consent was obtained from all participants. Participants were not eligible for the study if they were unable to give informed consent. Research Ethics Committee (REC) reference: 19/LO/0102 Integrated Research Application System (IRAS) project ID: 257561 ## Consent for publication Not applicable ## Availability of data and materials The code used to analyse the data is available in the supplementary materials. The dataset is available from the authors upon reasonable request. ## Competing interests The authors declare that they have no competing interests. ## Authors’ contributions RJ contributed to the study design, curated and analysed the data, and wrote the manuscript. (Conceptualization, Data curation, Formal Analysis, Validation, Visualization, Methodology, Writing - original draft, Writing – review & editing) KJ, MAC, and LC developed the methods for sample collection and laboratory analysis (plating, dipstick and 16s PCR method) and collected data for this study. (Conceptualization, Investigation, Methodology) AJW and MT collected data for this study. (Investigation) DW, KJ, MAC, AJW, MT and LC reviewed the manuscript. (Writing-review & editing) RN and DW form part of the clinical team that make decisions on what is considered a UTI (Investigation, Resources, Methodology). DJS supervised the study and reviewed the manuscript. (Conceptualization, Supervision, Funding acquisition, Writing-review & editing) PSF designed the study, supervised and reviewed the manuscript. (Conceptualization, Supervision, Funding acquisition, Writing-review & editing) ## Acknowledgements We thank the patients and study partners who took part in the study. We also thank all members of the UK Dementia Research Institute Care Research & Technology Centre who contributed in some way to this work. A full list of members can be found in supplementary acknowledgment list. Finally, we thank the Surrey and Borders partnership who were sponsors of this study. * Received December 10, 2024. * Revision received December 10, 2024. * Accepted December 11, 2024. * © 2024, Posted by Cold Spring Harbor Laboratory This pre-print is available under a Creative Commons License (Attribution-NonCommercial-NoDerivs 4.0 International), CC BY-NC-ND 4.0, as described at [http://creativecommons.org/licenses/by-nc-nd/4.0/](http://creativecommons.org/licenses/by-nc-nd/4.0/) ## References 1. 1.Alzheimer’s Society. Counting the cost: caring for people with dementia on hospital wards. 2009. 2. 2.Public Health England. Reasons why people with dementia are admitted to a general hospital in an emergency [Internet]. 2015 [cited 2022 Aug 8]. Available from: [https://webarchive.nationalarchives.gov.uk/ukgwa/20170302124526/](https://webarchive.nationalarchives.gov.uk/ukgwa/20170302124526/)[http://www.yhpho.org.uk/default.aspx?RID=207311](http://www.yhpho.org.uk/default.aspx?RID=207311) 3. 3.Corrado O, Swanson B, Hood C, Morris A, Ofili S, Capistrano J, et al. National Audit of Dementia Care in General Hospitals 2018-2019 Round Four Audit Report [Internet]. 2019. Available from: [www.hqip.org.uk/national-programmes](https://www.hqip.org.uk/national-programmes). 4. 4.Eriksson I, Gustafson Y, Fagerström L, Olofsson B. Urinary tract infection in very old women is associated with delirium. Int Psychogeriatr [Internet]. 2011 Apr;23(3):496–502. Available from: [http://www.ncbi.nlm.nih.gov/pubmed/20716391](http://www.ncbi.nlm.nih.gov/pubmed/20716391) 5. 5.Phelan EA, Borson S, Grothaus L, Balch S, Larson EB. Association of incident dementia with hospitalizations. JAMA [Internet]. 2012 Jan 11;307(2):165–72. Available from: [http://www.ncbi.nlm.nih.gov/pubmed/22235087](http://www.ncbi.nlm.nih.gov/pubmed/22235087) 6. 6.Gharbi M, Drysdale JH, Lishman H, Goudie R, Molokhia M, Johnson AP, et al. Antibiotic management of urinary tract infection in elderly patients in primary care and its association with bloodstream infections and all cause mortality: population based cohort study. BMJ [Internet]. 2019;364:l525. Available from: [http://www.ncbi.nlm.nih.gov/pubmed/30814048](http://www.ncbi.nlm.nih.gov/pubmed/30814048) 7. 7.Perry VH, Cunningham C, Holmes C. Systemic infections and inflammation affect chronic neurodegeneration. Nat Rev Immunol [Internet]. 2007 Feb;7(2):161–7. Available from: [http://www.ncbi.nlm.nih.gov/pubmed/17220915](http://www.ncbi.nlm.nih.gov/pubmed/17220915) 8. 8.Dunn N, Mullee M, Perry VH, Holmes C. Association between Dementia and Infectious Disease. Alzheimer Dis Assoc Disord [Internet]. 2005 Apr;19(2):91–4. Available from: [http://journals.lww.com/00002093-200504000-00008](http://journals.lww.com/00002093-200504000-00008) 9. 9.Public Health England. Urinary tract infection: diagnostic tools for primary care [Internet]. 2020 [cited 2022 Jul 22]. Available from: [https://www.gov.uk/government/publications/urinary-tract-infection-diagnosis](https://www.gov.uk/government/publications/urinary-tract-infection-diagnosis) 10. 10.Bonkat G, Bartoletti R, Bruyère F, Cai T, Geerlings SE, Köves B, et al. EAU Guidelines on Urological Infections [Internet]. 2022 [cited 2022 Jul 22]. Available from: [https://uroweb.org/guidelines/urological-infections](https://uroweb.org/guidelines/urological-infections) 11. 11.Bai AD, Bonares MJ, Thrall S, Bell CM, Morris AM. Presence of urinary symptoms in bacteremic urinary tract infection: a retrospective cohort study of Escherichia coli bacteremia. BMC Infect Dis. 2020 Dec 1;20(1). 12. 12.Berman P, Hogan DB, Fox RA. The atypical presentation of infection in old age. Age Ageing. 1987 Jul;16(4):201–7. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1093/ageing/16.4.201&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=3630842&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F12%2F11%2F2024.12.10.24318775.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1987M027900001&link_type=ISI) 13. 13.Cerejeira J, Lagarto L, Mukaetova-Ladinska EB. Behavioral and Psychological Symptoms of Dementia. Front Neurol [Internet]. 2012;3. Available from: [http://journal.frontiersin.org/article/10.3389/fneur.2012.00073/abstract](http://journal.frontiersin.org/article/10.3389/fneur.2012.00073/abstract) 14. 14.Boscia JA, Kobasa WD, Knight RA, Abrutyn E, Levison ME, Kaye D. Epidemiology of bacteriuria in an elderly ambulatory population. Am J Med. 1986 Feb;80(2):208–14. [CrossRef](http://medrxiv.org/lookup/external-ref?access_num=10.1016/0002-9343(86)90011-2&link_type=DOI) [PubMed](http://medrxiv.org/lookup/external-ref?access_num=3946436&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F12%2F11%2F2024.12.10.24318775.atom) [Web of Science](http://medrxiv.org/lookup/external-ref?access_num=A1986A069800015&link_type=ISI) 15. 15.Nicolle LE. Asymptomatic bacteriuria in the elderly. Infect Dis Clin North Am [Internet]. 1997 Sep;11(3):647–62. Available from: [https://linkinghub.elsevier.com/retrieve/pii/S0891552005703780](https://linkinghub.elsevier.com/retrieve/pii/S0891552005703780) 16. 16.Abrutyn E, Mossey J, Berlin JA, Boscia J, Levison M, Pitsakis P, et al. Does asymptomatic bacteriuria predict mortality and does antimicrobial treatment reduce mortality in elderly ambulatory women? Ann Intern Med [Internet]. 1994 May 15;120(10):827–33. Available from: [http://www.ncbi.nlm.nih.gov/pubmed/7818631](http://www.ncbi.nlm.nih.gov/pubmed/7818631) 17. 17.Nicolle LE, Mayhew WJ, Bryan L. Prospective randomized comparison of therapy and no therapy for asymptomatic bacteriuria in institutionalized elderly women. Am J Med [Internet]. 1987 Jul;83(1):27–33. Available from: [http://www.ncbi.nlm.nih.gov/pubmed/3300325](http://www.ncbi.nlm.nih.gov/pubmed/3300325) 18. 18.Pruesse E, Peplies J, Glöckner FO. SINA: accurate high-throughput multiple sequence alignment of ribosomal RNA genes. Bioinformatics [Internet]. 2012 Jul 15;28(14):1823–9. Available from: [http://www.ncbi.nlm.nih.gov/pubmed/22556368](http://www.ncbi.nlm.nih.gov/pubmed/22556368) 19. 19.Flores-Mireles AL, Walker JN, Caparon M, Hultgren SJ. Urinary tract infections: epidemiology, mechanisms of infection and treatment options. Nat Rev Microbiol [Internet]. 2015 May 24;13(5):269–84. Available from: [http://www.ncbi.nlm.nih.gov/pubmed/25853778](http://www.ncbi.nlm.nih.gov/pubmed/25853778) 20. 20.Chae JHJ, Miller BJ. Beyond Urinary Tract Infections (UTIs) and Delirium: A Systematic Review of UTIs and Neuropsychiatric Disorders. J Psychiatr Pract. 2015 Nov;21(6):402–11. [PubMed](http://medrxiv.org/lookup/external-ref?access_num=26554322&link_type=MED&atom=%2Fmedrxiv%2Fearly%2F2024%2F12%2F11%2F2024.12.10.24318775.atom)