Introduction
Functional dyspepsia (FD) is one of the most common gut-brain interaction disorders worldwide. According to the Rome IV criteria, it is characterized by persistent or recurrent symptoms lasting at least six months, with symptoms occurring within the past three months. These symptoms include one or more of the following: epigastric pain, epigastric burning, postprandial fullness, and early satiety, excluding organic disease1.
The pathophysiology of FD remains unclear but is believed to involve disturbances in the bidirectional communication between the gastrointestinal tract and the brain. This communication occurs primarily through neural, endocrine, immune, and humoral pathways, affecting brain and gut neural activity. A disturbed gut-brain axis can lead to motility disorders, visceral hypersensitivity, gut microbiota perturbations, and changes in mucosal and immune function2,3.
FD is associated with Helicobacter pylori infection, stress, mood and mental disorders, anxiety, and depression4,5. Additionally, duodenal barrier dysfunction and microbiota imbalance, causing subtle duodenal inflammation, are considered contributing factors to its pathophysiology6,7.
Some female patients with FD exhibited abnormal urinalysis findings, which could be mistaken for UTI. However, unlike true UTI, these abnormalities resolved or improved with antidepressants/anxiolytics without antibiotic intervention8. Unfortunately, there is limited research on whether FD’s pathogenesis is related to urinary system inflammation.
The study by Wuestenberghs et al. found that bladder pain syndrome is present in 26.9% of functional dyspepsia patients9. However, most patients do not have accompanying urinary irritation symptoms, and their urinalysis indicators remain abnormal and do not improve for several months, which is significantly different from the characteristic of asymptomatic bacteriuria spontaneously improving in 7–10 days10.
A recent study proposed a conceptual model linking mental disorders to urinary irritation symptoms in women with overactive bladder (OAB), involving interactions within the brain-gut-bladder axis (BGBA). This model highlights the interplay among the brain-gut axis, brain-bladder axis, and gut-bladder axis11. These interactions involve altered visceral function, cross-sensitization between the gastrointestinal tract and the bladder, neuroendocrine unbalance, autonomic nervous system unbalance, low-grade inflammation, and altered genetic and/or epigenetic susceptibility mechanisms12.
The gut microbiota, along with its metabolic products and immune factors, further affects the brain and bladder, resulting in bladder dysbiosis. Additionally, an animal model of bladder outlet obstruction (BOO) demonstrated that local bladder inflammation mediated by NLRP3 inflammasomes can induce inflammation in the hippocampus, leading to depressive behaviors13. This supports the “outside to inside” mechanism, where peripheral bladder inflammation causes neuroinflammation and induces mood disorders14.
The study by Persson et al. also indicates complex and significant associations between OAB and FD15. The above evidences indicate that there is close signal communication between the gastrointestinal tract, brain, and bladder.
Given these findings, it is hypothesized that there may be a potential link between FD and bladder inflammation. Dysregulated communication in the brain-gut-bladder network facilitates cross-organ interactions between gastrointestinal and bladder diseases, creating a mutually reinforcing cycle that drives shared comorbidities. If abnormal urinalysis findings in patients with FD are not properly recognized and are misdiagnosed as a UTI, these patients may face recurrent symptoms and receive inappropriate antibiotic treatment for a presumed chronic UTI.
Furthermore, such unnecessary antibiotic use can induce gut microbiota dysbiosis and related side effects. Thus, accurately distinguishing between FD and UTI holds both theoretical and practical clinical significance. This study aims to investigate whether urinalysis abnormalities in female FD patients differ from those in UTI patients, to identify the optimal combination of diagnostic indicators for differentiating between the two conditions, and to provide new insights into the pathogenesis of FD.
Methods
Study population
The study was retrospective. FD patients who met the inclusion and exclusion criteria were selected from the outpatient Department of Gastroenterology at the Affiliated Hospital of North Sichuan Medical College between 2022 and 2024. All selected patients were female.
Data collection was conducted through face-to-face and telephone interviews by internists, gathering basic patient information, including sociodemographic data (age, education level), as well as medical histories and urinalysis indicators for both groups. We have de-identified all patient details.
Sample size
In our preliminary research, we found that 18% of FD patients with abnormal urinalysis showed red blood cells without white blood cells, compared to 0% in patients with UTI. Assuming an α error of 0.05 and a β error of 0.20, the calculated sample size was 53 patients per group [The sample size was calculated by the following formula: n = (Zα/2+ Zβ)2× [p1(1-p1) + p2(1-p2)]/(p1-p2)2)]. Consequently, for this study, we selected 60 FD patients and 60 patients diagnosed with UTI, confirmed by positive urine cultures, from the same hospital during the same period.
Inclusion and exclusion criteria
The reporting of this study conforms to STROBE guidelines16. See Fig. 1 for the study flow chart.
Fig. 1Study flow chart.
FD group inclusion criteria1: (a) Meet the Rome IV criteria: experiencing one or more of the following symptoms—epigastric pain, epigastric burning, postprandial fullness, early satiety—with no evidence of organic disease to explain these symptoms; symptoms must have been present for at least six months, with the last three months meeting the diagnostic criteria. Patients must have completed examinations such as gastroscopy, barium meal, complete blood count, biochemistry, thyroid function, glycosylated hemoglobin, tumor marker screening, and abdominal ultrasound. (b) No prior history of renal/kidney disease.
(c) Patients whose abnormal urinalysis indicators improved upon the second re-examination after treatment with PPIs and anxiolytic/antidepressant medications. (d) Abnormal urinalysis was defined as one or more parameters (urinary protein, occult blood, leukocytes, erythrocytes, leukocyte esterase, or epithelial cells) exceeding the upper limit of normal reference values.
FD group exclusion criteria: (a) Systemic diseases, metabolic diseases, and malignancies found through examinations like gastroscopy. (b) Patients with irritable bowel syndrome (IBS). (c) Patients with gastroesophageal reflux disease (GERD) (d) Patients with a history of abdominal surgery.
(e) Pregnant and lactating women. (f) Patients whose abnormal urinalysis indicators worsened upon re-examination after PPI and combined anxiolytic/antidepressant treatment. (g) Patients receiving antibiotic therapy.
(h) Patients who refuse follow-up or refuse to repeat urinalysis.
UTI group inclusion criteria: (a) Urinary tract irritation symptoms such as frequent urination, urgency, and pain during urination, with or without fever. (b) Quantitative culture of bacteria in clean midstream urine with a colony count of ≥ 10 per high-power field, and urine white blood cells with a colony count of ≥ 5 per high-power field. (c) Clean mid-stream urine culture with G+ cocci count ≥ 104 CFU/ml or G− bacilli count ≥ 105 CFU/ml.
(d) For at least 6 months, with no FD symptoms (as defined above) in the past 3 months.
UTI group exclusion criteria: (a) Patients who had received antibiotic treatment prior to the initial urinalysis. (b) Patients with other systemic infections. (c) Patients with catheterization (indwelling or intermittent).
(d) History of pelvic, gastrointestinal, or urinary system surgery or trauma. (e) Patients with neurological or psychiatric disorders. (f) Patients with major chronic diseases or malignancies, such as coronary artery disease, myocardial infarction, stroke, etc.
(g) Incomplete data. (h) Patients who did not return for follow-up after treatment.
Sample collection
Urine samples were collected correctly, avoiding menstruation, and were promptly tested after being placed in sterile containers. Urinalysis was conducted using the FUS-2000 automatic urine chemistry analyzer and chemical strips from Dirui Industrial Co., Ltd. The normal reference ranges for the indicators were: urinary erythrocytes (0–10/µL), urinary leukocytes (0–10/µL), urinary epithelial cells (0–15/µL), urinary leukocyte esterase (0–25 Leu/µL), urinary protein (negative), and urinary occult blood (negative).
Data analysis
Statistical analysis was performed using SPSS version 27. The normality of continuous data was assessed with the Shapiro-Wilk test. For normally distributed age data, independent samples t-tests were used, and results were expressed as mean ± standard deviation.
Non-normally distributed quantitative data, including urinary leukocyte esterase, leukocytes, erythrocytes, and epithelial cells, were expressed as median (25th, 75th percentiles) and compared using the Mann-Whitney U test. Categorical data, including urinary occult blood and protein, were expressed as frequencies (percentages) and compared using Chi-square tests or Fisher’s exact tests. ROC curve analysis was conducted to evaluate the diagnostic efficiency of individual and combined urinalysis indicators for UTI or FD, calculating AUC, sensitivity, specificity, PPV, NPV, and the Youden index.
Logistic regression analysis was performed to identify independent predictive factors, and a nomogram prediction model was created.
Results
Comparison of population characteristics
We included 60 female patients with UTI and 60 female patients with FD. The analysis focused on age, gender, and urinary irritation symptoms. All participants were female, and the age distribution was similar between the two groups (P > 0.05), with both groups primarily consisting of middle-aged women over 50.
The proportion of patients with urinary irritation symptoms was significantly higher in the UTI group (43/60 patients, 71.7%) compared to the FD group (4/60 patients, 6.7%), showing a statistically significant difference between the two groups. The significant difference in urinary irritation symptoms was observed between FD and UTI patients (P < 0.05).
Comparisons the levels of urinary leukocyte esterase, leukocytes, erythrocytes, epithelial cells, occult blood, and protein levels between the two groups
Urinary protein (P < 0.01), occult blood (P < 0.01), leukocyte esterase (P < 0.01), leukocyte count (P < 0.01), and erythrocyte count (P < 0.01) were significantly higher in the UTI group compared to the FD group, where protein levels were almost negative. There was no significant difference in epithelial cell counts (P = 0.66) (Table 1).
Table 1Comparison of Urinalysis indicator levels between UTI and FD. *“0–4” represented “-, +-, 1+, 2+, 3+”.IndicatorsFD(n = 60)Median (IQR)UTI(n = 60)Median (IQR)Z valueP valueProtein*0(0;0)1(0;2)−5.67< 0.01OB*0(0;3)2(0;4)−3.45< 0.01Leukocyte esterase (Leu/µL)75(25;500)500(75;500)−3.38< 0.01WBC (cells/µL)8(4;32.25)80.5(36.2;569.75)−7.17< 0.01RBC (cells/µL)10(6;22.5)26(8;435.75)−3.96< 0.01Epithelial cells (cells/µL)4(0.25;18.5)5(1;23.5)−0.460.66UTI Urinary tract infection, FD Functional dyspepsia, IQR Interquartile range.
Comparisons the abnormal frequencies of urinary leukocyte esterase, leukocytes, erythrocytes, epithelial cells, occult blood, and protein between UTI and FD groups
Chi-square test results showed significant differences between the UTI and FD groups for several indicators. Significant differences were observed for leukocyte esterase levels at 0–25 Leu/µL and ≥ 500 Leu/µL. Lower levels (0–25 Leu/µL) were more indicative of FD (OR < 1), while higher levels (≥ 500 Leu/µL) were more indicative of UTI (OR > 1).
Negative protein levels were predominantly observed in FD patients (95.0%), supporting an FD diagnosis (OR < 1), whereas positive protein levels (1+) were significantly more common in UTI patients (OR > 1). Negative occult blood levels were more supportive of FD (OR < 1), while higher levels (3+) were more indicative of UTI (OR > 1). Lower erythrocyte levels (0–10/µL) supported an FD diagnosis (OR < 1), whereas higher levels (≥ 31/µL) were more indicative of UTI (OR > 1).
Lower leukocyte counts (0–10/µL) supported an FD diagnosis (OR < 1), while higher counts (≥ 71/µL) were strongly indicative of UTI (OR > 1). No significant differences were found in epithelial cell counts between the two groups (P > 0.05). These results suggest that specific urinalysis indicators, such as high leukocyte esterase, positive protein, high occult blood, high erythrocytes, and high leukocytes, were more indicative of UTI, while negative or lower levels of these indicators supported an FD diagnosis (Table 2).
Additionally, patients with abnormal erythrocytes but normal leukocytes were 0/60 (0%) in the UTI group, compared to 14/60 (23.3%) in the FD group (χ² = 15.85, P < 0.01).
Table 2Comparison of abnormal frequencies of urinary leukocyte esterase, leukocytes, erythrocytes, epithelial cells, occult blood, and protein between UTI and FD groups. *“0–4” represented “-, +-, 1+, 2+, 3+”.IndicatorsUTI(n = 60)FD(n = 60)χ2ORP-valueTotalχ2P valueLeukocyte esterase (Leu/µL)[0, 25)2(3.3%)12(20.0%)8.090.14< 0.0113.120.01[25, 50)7(11.7%)9(15.0%)0.290.750.79[50, 250)11(18.3%)13(21.7%)0.210.810.82[250, 500)8(13.3%)10(16.7%)0.260.950.08>=50032(53.3%)16(26.7%)8.893.14< 0.01ProteinNegative27(45.0%)57(95.0%)35.710.04< 0.0137.270.00+-13(21.7%)0(0.0%)14.5819.61< 0.011+13(21.7%)1(1.7%)11.6413.57< 0.012+5(8.3%)2(3.3%)1.372.640.443+2(3.4%)0(0.0%)2.032.810.50OB*Negative16(26.7%)32(53.3%)8.890.32< 0.0121.830.00+-5(8.3%)1(0.0%)2.815.360.091+10(16.7%)11(18.3%)0.060.890.812+7(11.7%)12(20.0%)1.560.530.213+22(36.7%)4(6.7%)15.918.11< 0.01RBC (cells/µL)0–1016(26.7%)32(53.3%)8.890.32< 0.0121.360.0011–208(13.3%)11(18.3%)0.560.690.6221–309(15.0%)12(20.0%)0.520.710.63>=3127(45.0%)5(8.3%)20.639.00< 0.01WBC (cells/µL)0–101(1.7%)32(53.3%)40.160.02< 0.0164.670.0011–3012(20.0%)13(21.7%)0.050.901.0031–507(11.7%)8(13.3%)0.080.861.0051–705(8.3%)2(3.3%)0.541.370.44>=7135(58.3%)1(1.7%)45.8782.60< 0.01Epithelial cells (cells/µL)0–1542430.040.921.000.041.00> 1518170.043.251.00
Diagnostic efficiency analysis of single tests for urinary protein, occult blood, leukocyte esterase, leukocytes, erythrocytes, and epithelial cells in predicting FD or UTI
ROC curve analysis revealed the following AUC values for diagnosing urinary tract infection: The highest AUC was observed for urine white blood cells (0.88, 0.82–0.94), with sensitivity of 75.0%, specificity of 85.0%, positive predictive value of 83.3%, and negative predictive value (NPV) of 77.3%. The Youden index was highest, with the optimal cutoff value of 36.5/µL, indicating that urine white blood cell counts greater than 36.5/µL were most indicative of UTI. Urine protein demonstrated the highest specificity (95.0%) and PPV (91.2%), with the best cutoff value of 0.5, suggesting that positive urine protein levels were more likely to indicate a UTI (Table 3; Fig.
2a).
Table 3Diagnostic efficiency of urinary protein, occult blood, leukocyte esterase, leukocytes, erythrocytes, and epithelial cells. *“0–4” represented “-, +-, 1+, 2+, 3+”.IndicatorsCut-pointAUROC(95%CI)P valueSe(%)Sp(%)YIPPV(%)NPV(%)LRProtein*0.50.74(0.65 ~ 0.83)0.0055.095.00.5091.267.911.00OB*3.50.68(0.58 ~ 0.77)0.0036.793.30.3084.659.65.50Leu esterase3750.67(0.58 ~ 0.77)0.0053.373.30.2766.761.12.00WBC36.50.88(0.82 ~ 0.94)0.0075.085.00.6083.377.35.00RBC27.50.71(0.62 ~ 0.80)0.0050.088.30.3881.163.94.29Epithelial cells0.50.52(0.42 ~ 0.63)0.6485.025.00.1049.448.61.13AUROC Area under the receiver operating curve, CI Confidence interval, Se Sensitivity, Sp Specificity, YI Youden index, PPV Positive predictive value, NPV Negative predictive value, LR Likelihood ratio, WBC White blood cell, RBC Red blood cell, OB Occult blood, Leu esterase Leukocyte esterase, Pro Protein.
Fig. 2(a) ROC curves to diagnose UTI for leukocyte esterase, protein, erythrocytes, leukocytes, occult blood, epithelial cells in urine; (b) ROC curves to diagnose UTI for combined urinalysis indicators.
Combined diagnostic value analysis of urinary protein, occult blood, leukocyte esterase, leukocytes, and erythrocytes in predicting FD or UTI
Excluding epithelial cells, combinations of urinary protein, occult blood, leukocyte esterase, leukocytes, and erythrocytes were analyzed for diagnostic efficiency. The combination of leukocytes and protein had the highest AUC (0.90), with sensitivity of 90.0%, specificity of 78.3%, PPV of 90.0%, and NPV of 63.3%. The combination of leukocytes and occult blood showed the highest specificity (93.3%), PPV (95.0%), and accuracy (83.3%).
The combination of leukocyte esterase, protein, and occult blood exhibited high specificity, PPV, and accuracy, with the highest Youden index (0.73) (Table 4; Fig. 2b).
Table 4Combined diagnostic analysis of urinary protein, occult blood, leukocyte esterase, leukocytes, and erythrocytes.IndicatorAUROC(95%CI)P valueSe(%)Sp(%)YIPPV(%)NPV(%)Diagnostic accuracy (%)WBC + OB0.85(0.78 ~ 0.92)0.0076.693.30.7095.071.783.3WBC + Pro0.90(0.85 ~ 0.91)0.0090.078.30.6890.063.376.7Leu esterase + OB0.78(0.69 ~ 0.87)0.0065.091.70.5763.380.071.7Leu esterase + Pro0.81(0.73 ~ 0.89)0.0080.068.30.4868.378.373.3Leu esterase + RBC0.79(0.71 ~ 0.87)0.0081.765.00.4770.076.773.3Leu esterase + Pro + OB0.81(0.72 ~ 0.89)0.0073.390.00.7390.071.780.8Leu esterase + Pro + RBC0.82(0.74 ~ 0.90)0.0068.385.00.5370.073.371.7AUROC Area under the receiver operating curve, CI Confidence interval, Se Sensitivity, Sp Specificity, YI Youden index, PPV Positive predictive value, NPV Negative predictive value, LR Likelihood ratio, WBC White blood cell, RBC Red blood cell, OB Occult blood, Leu esterase Leukocyte esterase, Pro Protein.
Logistic regression analysis for predicting to diagnose FD or UTI using age, urinary protein, occult blood, leukocyte esterase, leukocytes, erythrocytes, and epithelial cells
Univariate logistic regression analysis assessed the significance of age, urinary protein, occult blood, leukocyte esterase, leukocytes, erythrocytes, and epithelial cells in differentiating between UTI and FD (Table 5). Variables with significant results (P < 0.05) were included in a multivariate logistic regression analysis (Table 5). The multivariate analysis identified urinary protein, occult blood, and leukocyte esterase as significant predictors for distinguishing UTI from FD (P < 0.05).
Although leukocytes were significant in the univariate analysis, their significance diminished in the multivariate model. Erythrocytes were not significant in the multivariate analysis. These findings underscore the diagnostic importance of urinary protein, occult blood, and leukocyte esterase in differentiating between UTI and FD.
Table 5Univariate logistic regression analysis and multivariate logistic regression analysis for predicting FD or UTI diagnosis.VariableβSEOR (95%CI)P valueUnivariate logistic regression analysisAge0.010.011.01(0.98–1.03)0.64Protein1.340.343.81(1.97–7.36)0.00OB0.400.121.48(1.17–1.88)0.00Leu esterase0.000.001.00(1.00–1.01.00.01)0.002WBC0.020.011.02(1.01–1.03)0.00RBC0.020.011.02(1.00–1.03.00.03)0.03Epithelial cells0.000.001.00(1.00–1.01.00.01)0.24Multivariate logistic regression analysisProtein1.1340.343.11(1.60–6.04)0.00OB0.3310.161.39(1.02–1.90)0.04Leu esterase0.0040.001.00(1.00–1.01.00.01)0.00Constant−2.0390.500.130.000,1 represented FD, UTI; β Standardized regression coefficient, SE Standard error, OR Odds ratio, CI Confidence interval, WBC White blood cell, RBC Red blood cell, OB Occult blood, Leu esterase Leukocyte esterase.
Nomogram prediction model for FD or UTI based on abnormal urinalysis
A nomogram prediction model was developed to estimate the risk of FD or UTI based on abnormal urinalysis results, incorporating significant predictors from the multivariate logistic regression analysis: urinary protein, occult blood, and leukocyte esterase. This model provides a visual tool to estimate the probability of UTI or FD in patients (Fig. 3).
It assists clinicians in evaluating the likelihood of FD or UTI based on abnormal urinalysis results, thereby supporting more accurate diagnosis and treatment decisions.
Fig. 3Nomogram prediction model for FD or UTI based on abnormal urinalysis. Diagnostic threshold was calculated as 0.6523.
If the predicted probability is below the diagnostic threshold, it tends to be classified as FD; otherwise, it is predicted as UTI.
Discussion
UTI refers to an infection occurring in the urethra, bladder, or kidneys and is more common in women, which commonly present with urinary irritation symptoms like frequency, urgency, and dysuria17,18. Leukocyte urine is the basis for diagnosing urinary tract infections. Some studies have proposed that the current leukocyte threshold for diagnosing urinary tract infections is too low, which can lead to misdiagnosis of UTI.
It is suggested to use a corrected leukocyte threshold to classify UTI. In this study, it is more likely to diagnose UTI while the leukocyte count is greater than 71/µL, and FD patients often do not have high white blood cells, but only show abnormalities in occult blood or red blood cells. Abnormal erythrocytes with normal leukocytes do not support UTI and suggest a FD diagnosis.
While normal urinary protein is highly specific, it has low sensitivity for excluding UTI. This inflammation was distinct from UTI, with most FD patients not exhibiting urinary irritation symptoms and not requiring antibiotic treatment. Instead, improvements were observed with treatment using PPIs combined antianxiety/antidepressant medications.
Regarding the pathogenesis of abnormal urinalysis in female FD patients, a population-based cohort study in Taiwan indicated a bidirectional relationship between anxiety, depression, and lower urinary tract symptoms (LUTS)19. Peripheral inflammation, such as bladder inflammation, can “migrate” to the central nervous system (CNS) through neural, endocrine, and immune pathways, causing changes in central microglial cells and triggering neuroinflammation, which affects the CNS and subsequently influences bladder activity and symptom occurrence20. Existing studies have demonstrated the detrimental role of neuroinflammation in Helicobacter pyloriinfection and chronic gastritis.
Furthermore, neuroprotection has been proposed as a potential therapeutic strategy for FD, suggesting a link between local organ inflammation and neuroinflammation. These findings suggest that inflammatory mechanisms play a role in brain-organ communication, potentially through an “inside” mechanism within the CNS or an “outside to inside” pathway14.
Urinary abnormalities in FD patients are associated with anti-inflammatory effects of anti-anxiety/depression drugs8. In the present study, FD patients with poor responses to PPI treatment and abnormal urinalysis showed normalization or significant improvement in urinalysis indicators after two weeks of combined treatment with PPI and flupentixol-melitracen8. Some anti-inflammatory agents, including cytokine inhibitors, NSAIDs, omega-3 fatty acids, and statins, have been found to exhibit certain antidepressant effects22–24.
Additionally, some selective serotonin reuptake inhibitors (SSRIs), such as fluoxetine, also possess anti-inflammatory properties25. This supports the potential anti-inflammatory action of these treatments in alleviating bladder inflammation. Other related studies have demonstrated a connection between inflammation and the clinical progression of neuropsychiatric disorders, further substantiating this finding26. –27.
The gut microbiota may also be involved in the mechanism of FD-related abnormal urinalysis. A study investigating the genome of Escherichia coli in the bladder revealed that urine was not sterile and contained colonizing bacteria, including genera such as Lactobacillus, Corynebacterium, Gardnerella, Staphylococcus, and Streptococcus28. In urine samples from patients with bladder diseases, specific bacterial groups exhibited altered concentrations29.
When gut bacteria, such as Escherichia coli, migrate to the bladder via flagellum-mediated motility28,30, they can disrupt the stability of the bladder microbiome, leading to bacterial dysbiosis31,32. This suggests a possible connection between the gut and bladder through the gut microbiota, supporting the notion of interactions between the gastrointestinal tract and the bladder via the brain-gut-bladder axis or the brain-bladder axis. Such interactions could lead to an imbalance in the bladder microbiota, resulting in bladder inflammation.
This could explain the abnormal urinalysis observed in FD patients in this study.
For the differential diagnosis of abnormal urinalysis in female FD patients versus UTI patients, our results indicated that UTI patients often exhibit urinary irritation symptoms, whereas such symptoms were rare in FD patients. In the UTI group, the levels of urinary protein, occult blood, leukocyte esterase, leukocyte count, and erythrocyte count were significantly higher than those in the FD group. ROC curve analysis showed that among the single indicators, urinary leukocytes had the greatest diagnostic significance.
When leukocytes exceed 71/µL, a diagnosis of UTI was more likely. The combined diagnostic efficiency results revealed that the combination of urinary protein, occult blood, and leukocyte esterase had the highest Youden index, indicating the greatest diagnostic value with high specificity and accuracy. However, despite a specificity of 90%, there remains a false positive rate, and the sensitivity was insufficient, leading to potential false negatives.
Therefore, the diagnosis of UTI or FD should be combined with clinical treatment outcomes for confirmation. When the false positive rate was set at 10%, the diagnostic threshold was calculated as 0.6523 based on the combination of occult blood, urinary protein, and leukocyte esterase33. When p > 0.6523, the diagnosis tended towards UTI; otherwise, it tended towards FD.
For example, if a female FD patient had the following urinalysis results: occult blood (3+), protein negative, and leukocyte esterase 250. Using the nomogram for visual calculation: Occult blood score: 21 points, Leukocyte esterase score: 27 points, Protein score: 0 points, Total score: 48 points, corresponds to a diagnostic probability of 0.53, suggesting FD. Subsequent outpatient treatment confirmed this diagnosis.
The patient’s urinalysis significantly improved after the treatment with a combination of PPI and flupentixol-melitracen, without the need for antibiotics8.
Limitations
This study has several limitations. Firstly, as a retrospective study with patients sourced from a tertiary medical center, there is selection bias and the generalizability of the findings requires further investigation34. And there is the danger of overfitting due to the small sample size.
The unique urogenital anatomy of women might contribute to abnormal urinalysis. Secondly, specific confirmatory data such as bladder imaging or histology were not provided. Thirdly, uncontrolled confounding factors: Such as sexual activity, hygiene habits, drinking habits, daily activities, psychosocial stress, vaginal microecology and perimenopausal period, might still influence urinalysis results.
Fourthly, specificity and sensitivity in regression analysis: The regression analysis targeted a specificity of 90%, but a 10% false positive rate and a sensitivity of 73.3% remained. Diagnosis should be confirmed by clinical treatment outcomes. Fifth, studies have shown that the culture positivity rate in patients suspected of having UTI is only around 30%35.
Since FD patients in gastroenterology clinics rarely exhibit urinary irritation symptoms and most patients are reluctant to undergo urine culture tests, coupled with the low culture positivity rate, it remains challenging to distinguish and differentiate abnormal urinalysis indicators in FD from those in UTI. Using post-treatment improvement in urinalysis indicators as an inclusion criterion for the FD group may still lead to misclassification. Sixth, data and equipment specificity: The study utilized the FUS-2000 automatic urine chemistry analyzer and chemical strips from Dirui Industrial Co., Ltd.
Consistency of results with other equipment requires further research. Thus, the generalizability of the findings requires further investigations by large-scale, multicenter data with external validation.
High leukocytes and high occult blood were rare in FD patients with abnormal urinalysis, whereas these indicators were common in UTI patients. Conversely, UTI patients with abnormal erythrocytes but normal leukocytes were extremely rare (0%), while this pattern was observed in 23.3% of FD patients. These differences were statistically significant.
The findings supported the need to avoid misdiagnosing long-term FD as chronic UTI, which could lead to repeated and unnecessary antibiotic treatments. Further multicenter data collection for external validation and prospective animal experiments to investigate underlying mechanisms are required.
Conclusions
This study demonstrated that abnormal urinalysis indicators in female FD patients differ from UTI. The optimal combination for the rapid clinical prediction of UTI or FD is urinary occult blood in combination with protein and leukocyte esterase. Local bladder inflammation may underlie the abnormal urinary indicators observed in patients with FD.
Further prospective observational research is warranted to confirm these results.