Abstract
This study aims to investigate the relationships among a novel inflammatory biomarker, the red blood cell distribution width-to-albumin ratio (RAR), and three different types of urinary incontinence (UI) while evaluating the potential clinical significance of the biomarker. The National Health and Nutrition Examination Survey (NHANES)—which spanned 2007–2018—was used in this investigation. The relationship between the RAR and the prevalence of UI was assessed by logistic regression modeling, restricted cubic spline curve (RCS) analysis, and subgroup analysis. Fully adjusted models revealed a significant positive correlation between the RAR and all three types of UI (SUI: OR 1.23, 95% CI 1.12–1.35; UUI: OR 1.43, 95% CI 1.31–1.56; MUI: OR 1.44, 95% CI 1.27–1.62; all p < 0.05). Dividing the RAR into quartiles illustrated that an increased RAR was positively correlated with UI compared with a decreased RAR (SUI: OR 1.40, 95% CI 1.22–1.60; UUI: OR 1.85, 95% CI 1.60–2.14; MUI: OR 1.81, 95% CI 1.47–2.21; all p < 0.05). A nonlinear, inverted “U”-shaped relationship between the RAR and UI was demonstrated using restricted cubic spline analysis, suggesting that the RAR impacted UI only until reaching a threshold. Additionally, subgroup analysis revealed a stronger link between the RAR and SUI in women than in men (p < 0.05). An increased RAR was correlated with a heightened risk of UI, and RAR was more strongly associated with stress urinary incontinence (SUI) in women than in men. Systematic epidemiologic evidence suggests that the novel inflammatory marker RAR is associated with UI, but more research is needed to determine causality.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-025-02574-y.
Keywords: RAR, UI, Inflammation, NHANES
Subject terms: Kidney, Kidney diseases, Renal replacement therapy, Nephrology, Urology
Introduction
Urinary incontinence (UI) is recognized by the World Health Organization (WHO) as an important worldwide health issue1,2. The International Continence Society defines stress urinary incontinence (SUI) as the involuntary expulsion of urine provoked by physical exertion, whereas urgency urinary incontinence (UUI) is characterized by involuntary urine leakage associated with an abrupt, intense urge to urinate3. Mixed urinary incontinence (MUI) denotes the simultaneous presence of both SUI and UUI. Epidemiological data indicate that approximately 30% to 40% of elderly women report experiencing UI and that the incidence of UI in men increases with age. For example, the prevalence of UI in men aged 60–64 years is 11%, increasing to 31% in men aged 85 years. Furthermore, up to 32% of the population reports lower urinary tract symptoms (LUTS)2,3. The EPINCONT study, a longitudinal investigation of Norwegian women conducted from 1995 to 1997 and again from 2006 to 2008, demonstrated a 16% increase in the prevalence of UI across these intervals, with an incidence rate of 18.7%4. The number of women affected by UI in the United States is projected to increase from 18.3 million in 2010 to 28.4 million by 20505. UI in women is often linked to dysfunction of the bladder or pelvic floor muscles, commonly arising during pregnancy, childbirth, or menopause. The incidences of nocturia, urgency, diminished urine flow, hesitation, incomplete bladder emptying, postvoid dribbling, and high daytime frequency have also increased, with males reporting a poorer health-related quality of life (HRQoL) while experiencing UI or overactive bladder (OAB)6. While some studies have demonstrated an increased incidence of UI among white women, additional research has indicated that prevalence rates are similar across various racial and ethnic groups7. Other factors linked to UI include parity, obesity, history of uterine or pelvic surgeries, lung diseases, diabetes, hospitalization, and dementia8. Emerging evidence implicates systemic inflammation as a key mechanism underlying these factors and urinary dysfunction pathogenesis.
Urinary tract infections (UTIs) and inflammation affect the structure and function of the urinary tract epithelium and underlie the pathogenesis of UI9. Some research has revealed a possible correlation between inflammatory indicators and UI. RDW is a prevalent hematological parameter that indicates variability in red blood cell volume and is associated with several clinical pathological disorders. It has been introduced as a novel predictor of inflammation for various diseases. Studies indicate that chronic inflammation may increase RDW levels by affecting the hematopoietic microenvironment and red blood cell lifespan10. Additionally, serum albumin levels serve as indicators of the body’s inflammatory state. A Swedish prospective study revealed that low albumin levels are negatively correlated with patient mortality and can independently predict mortality11. Specifically, both acute and chronic inflammation alter hepatic protein metabolism and induce capillary leakage, thus influencing serum albumin levels12. The RDW-to-albumin ratio (RAR, reported in %/g/dL), serves as a recently established composite measure of inflammation. The RAR is calculated as follows: red blood cell distribution width (%) divided by serum albumin (g/dL). The RAR is usually less than 3. It integrates RDW (which reflects red blood cell volume heterogeneity and is associated with chronic inflammation) and serum albumin (which reflects nutritional status and the ability to suppress inflammation), allowing for a multidimensional assessment of systemic inflammation. However, compared with traditional inflammatory markers such as C-reactive protein, the RAR lacks universal applicability. A recent study suggested that the RAR may function as a potential biomarker for adverse outcomes in multiple disorders, including cardiovascular diseases, sepsis, and chronic kidney disease13–19. Current studies have not systematically elucidated the relevance of this inflammatory-nutritional biomarker to different UI subtypes, and there is a lack of validation based on cross-ethnic, large-sample populations.
The current investigation aims to explore the possible correlation between the RAR and UI among individuals in the United States. Although the RAR has been recognized as a predictive marker for risk in other health disorders, its association with UI remains poorly understood. To fill this gap, information collected by the NHANES between 2007 and 2018 will be analyzed to evaluate the correlation between RDW, serum albumin levels, and the prevalence of UI, offering new insights into the risk assessment of this condition.
Materials and methods
Study description and population
This investigation utilized data from the National Health and Nutrition Examination Survey. As shown in Fig. 1, the initial sample consisted of 59,842 participants. After excluding those under 18 years of age (n = 23,262) and those lacking data from KIQ042 and KIQ044 (n = 6431), the following variables with high proportions of missing values were then deleted based on the proportion of missing values: triglyceride and LDL levels. Random forest was used to interpolate missing values for variables with < 20% missing proportions (n = 30,149), and a small number of questionnaires with invalid responses were excluded (n = 955), leaving complete data for inclusion in the study (n = 29,194).
Fig. 1.
Flowchart for researcher inclusion screening. We excluded participants who were minors and had missing values in the data.
Calculation of the RAR
This study uses the RAR as the exposure variable, computed using the following formula: RAR = red blood cell distribution width (%) divided by serum albumin (g/dL). The concentration of serum albumin was measured using the LX20 method, a bichromatic digital endpoint approach. In this process, albumin interacts with the bromocresol purple (BCP) reagent to form a complex, and the system quantifies changes in absorbance at 600 nm. A complete blood count was acquired using the Beckman Coulter MAXM instrument included within the mobile examination centers (MECs), and this device provided a blood cell distribution for all participants.
Assessment of UI
The survey data from the “Kidney Conditions – Urology” section of the NHANES database (2007–2018) were analyzed. Responses to the KIQ042 and KIQ044 questionnaires, which have a structure similar to that of the 3IQ questionnaire, were used to identify UI and its subtypes. If a person replied “yes” to the KIQ042 question “In the past 12 months, have you/has SP experienced urine leakage or loss of control during activities such as coughing, lifting, or exercising?”, they were considered to have SUI. If a person replied “yes” to the KIQ044 question “In the past 12 months, have you/has SP experienced urine leakage or loss of control when experiencing a strong urge to urinate and were unable to reach the toilet in time?”, they were considered to have UI with urgency. Participants who responded “yes” to both inquiries were considered to have MUI.
Covariates of interest
This study examines various covariates, including demographic variables (sex categories, age, ethnic background, level of education, marital status), socioeconomic factors (family poverty-to-income ratio, smoking status, alcohol consumption), health conditions (diabetes, hypertension), physical activity levels (moderate, vigorous), physical examination metrics (body mass index), and laboratory parameters (albumin, alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, blood urea nitrogen, creatinine, uric acid, globulin, red blood cell distribution width, total cholesterol, HDL cholesterol). Smoking status was determined through the SMQ020 questionnaire, where participants who answered “yes” to the question “Have you smoked at least 100 cigarettes in your lifetime?” were classified as smokers. Alcohol consumption was evaluated using the BPQ020 questionnaire, with participants who answered “yes” to “Have you ever been told you have high blood pressure?” were classified as hypertensive. Other laboratory data were sourced from the laboratory data module. Age, BMI, and the PIR were categorized for subgroup analysis. This classification aimed to streamline and standardize the data, facilitating more effective presentation and interpretation of participant characteristics.
Statistical analysis
A logistic regression model was used to assess the associations between the RAR and the occurrence of three types of UI (SUI, UUI, and MUI). The findings are displayed as odds ratios (ORs) accompanied by 95% confidence intervals (CIs). A stepwise regression method was utilized, followed by a multicollinearity assessment. The variance inflation factors (VIFs) varied from 1 to 2, indicating that the model exhibited a good fit, hence ensuring its stability and reliability. Model 1 included no covariate adjustments; Model 2 adjusted for age, sex, race, marital status, education level, and PIR; Model 3 adjusted for age, sex, race, marital status, education level, PIR, diabetes mellitus, hypertension, vigorous work activity, moderate work activity, smoking, alcohol consumption, alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), creatinine, uric acid, globulin, total cholesterol, triglycerides, and high-density lipoprotein (HDL) cholesterol. Restricted cubic splines were employed to determine the ideal RAR threshold most strongly correlated with UI. The subgroup analysis investigated the association between the RAR and UI across several strata. Data analysis was performed using RStudio 4.30, incorporating R packages such as ggrcs, scitb, and others. The code can be accessed at https://github.com/liuqiang070488. Statistical significance was established when the two-tailed p value was less than 0.05.
Results
Baseline characteristics of the participants
Figure 1 depicts the participant selection procedure from the original 29,194 individuals. The characteristics of the weighted baseline study cohort are shown in Table 1. The Table 1 weighted baseline table shows that the population was 49% male and 51% female, with an average age ± (SD) of 47.48 ± 17.03 years. Among the individuals, 24.00% reported SUI, 20.00% reported UUI, and 9.30% reported MUI. The RAR for each quartile 1–4 was as follows: 2.17–2.88, 2.88–3.10, 3.10–3.41, and 3.41–13.40. The mean RAR was 3.20, accompanied by a standard deviation of 0.26. A markedly greater prevalence of all types of UI was noted in the upper quartiles (Q2, Q3, and Q4) compared to that in quartile 1 (Q1) (all p < 0.001). Moreover, individuals in the upper RAR quartiles presented markedly elevated incidences of hypertension, diabetes, and BMI as well as blood lipid levels (p < 0.001).
Table 1.
Baseline characteristics of participants after grouping by RAR quartiles, weighted.
| Characteristic | Na nd |
Overall N = 201,999,118b n = 29194d |
Q1(2.17–2.88) N = 51,195686b n = 6305d |
Q2(2.88–3.10) N = 50,124,877b n = 6740d |
Q3(3.10–3.41) N = 50,213,095b n = 7379d |
Q4(3.41–13.40) N = 50,465,460b n = 8769d |
P-valuec |
|---|---|---|---|---|---|---|---|
| Age, (years) | 29,194 | 47.48 ± (17.03) | 40.59 ± (15.14) | 46.62 ± (16.19) | 50.15 ± (16.88) | 52.68 ± (17.35) | < 0.001 |
| Gender, n (%) | 29,194 | < 0.001 | |||||
| Male | 14,377 (49%) | 4090 (64%) | 3674 (53%) | 3408 (45%) | 3205 (33%) | ||
| Female | 14,817 (51%) | 2216 (36%) | 3066 (47%) | 3971 (55%) | 5564 (67%) | ||
| Race, n (%) | 29,194 | < 0.001 | |||||
| Mexican American | 4396 (8.5%) | 1024 (8.5%) | 1097 (8.8%) | 1128 (7.6%) | 1147 (8.4%) | ||
| Other Hispanic | 3044 (5.8%) | 605 (5.0%) | 739 (5.7%) | 861 (5.8%) | 839 (6.2%) | ||
| Non-Hispanic white | 12,189 (67%) | 3129 (73%) | 2983 (70.0%) | 3046 (72%) | 3031 (58%) | ||
| Non-Hispanic Black | 6219 (11%) | 661 (4.9%) | 1033 (7.6%) | 1584 (9.7%) | 2941 (20%) | ||
| Other Race | 3346 (7.6%) | 887 (8.5%) | 888 (8.0%) | 760 (6.6%) | 811 (7.3%) | ||
| Education Level, n (%) | 29,194 | < 0.001 | |||||
| Less Than 9th Grade | 2932 (5.0%) | 508 (4.0%) | 187 (3.9%) | 783 (5.3%) | 977 (6.2%) | ||
| 9-11th grade | 4079 (10%) | 796 (8.8%) | 281 (9.5%) | 1021 (10%) | 1361 (13%) | ||
| High School | 6724 (23%) | 1377 (21%) | 502 (22%) | 1726 (23%) | 2128 (26%) | ||
| Some College | 8694 (32%) | 1821 (30%) | 652 (30%) | 2239 (33%) | 2711 (33%) | ||
| College Graduate | 6765 (30%) | 1804 (36%) | 659 (35%) | 1610 (29%) | 1592 (23%) | ||
| PIR | 29,194 | 2.99 ± (1.66) | 3.18 ± (1.65) | 3.13 ± (1.64) | 2.97 ± (1.64) | 2.65 ± (1.64) | < 0.001 |
| Marital status, n (%) | 29,194 | < 0.001 | |||||
| Married | 14,846 (55%) | 3256 (54%) | 3667 (57%) | 3813 (56%) | 4110 (52%) | ||
| Widowed | 2310 (5.7%) | 221 (2.3%) | 364 (3.9%) | 672 (6.6%) | 1053 (10%) | ||
| Divorced | 3222 (10%) | 498 (7.6%) | 685 (10.0%) | 875 (11%) | 1164 (12%) | ||
| Separated | 990(2.4%) | 135 (1.5%) | 200 (2.1%) | 281 (2.8%) | 374 (3.1%) | ||
| Never married | 5428 (19%) | 1612 (26%) | 1215 (18%) | 1169 (16%) | 1432 (15%) | ||
| Living with partner | 2398 (8.2%) | 584 (8.9%) | 609 (9.0%) | 569 (7.6%) | 636 (7.5%) | ||
| Smoke, n (%) | 29,194 | 0.043 | |||||
| Yes | 13,003 (44%) | 2685 (43%) | 2940 (44%) | 3359 (46%) | 4019 (45%) | ||
| No | 16,191 (56%) | 3621 (57%) | 3800 (56%) | 4020 (54%) | 4750 (55%) | ||
| Alcohol drinks, n (%) | 29,194 | < 0.001 | |||||
| Yes | 20,988 (78%) | 4998 (84%) | 5012 (80%) | 5292 (77%) | 5686 (69%) | ||
| No | 8206 (22%) | 1308 (16%) | 1728 (20%) | 2087 (23%) | 3083 (31%) | ||
| Diabetes, n (%) | 29,194 | < 0.001 | |||||
| Yes | 4059 (10%) | 399 (4.2%) | 666 (7.4%) | 1091 (12%) | 1903 (18%) | ||
| No | 25,135 (90%) | 5907 (96%) | 6074 (93%) | 6288 (88%) | 6866 (82%) | ||
| Hypertension, n (%) | 29,194 | < 0.001 | |||||
| Yes | 10,611 (32%) | 1475 (22%) | 2069 (28%) | 2890 (35%) | 4177 (44%) | ||
| No | 18,583 (68%) | 1658 (77%) | 4671(72%) | 4489 (65%) | 4592 (56%) | ||
| Vigorous work activity, n (%) | 29,194 | < 0.001 | |||||
| Yes | 5783 (21%) | 1500 (22%) | 1465 (24%) | 1439 (22%) | 1379 (17%) | ||
| No | 23,411 (79%) | 4806 (78%) | 5275 (76%) | 5940 (78%) | 7390 (83%) | ||
| Moderate work activity, n (%) | 29,194 | < 0.001 | |||||
| Yes | 10,766 (42%) | 2566 (45%) | 2661 (44%) | 2675 (42%) | 2864 (37%) | ||
| No | 18,428 (58%) | 3740 (55%) | 4079 (56%) | 4704 (58%) | 5905 (63%) | ||
| BMI, (kg/m2) | 29,194 | 29.11 ± (6.89) | 26.52 ± (4.97) | 28.10 ± (5.73) | 29.72 ± (6.67) | 32.15 ± (8.39) | < 0.001 |
| Albumin, (g/dL) | 29,194 | 4.26 ± (0.35) | 4.60 ± (0.22) | 4.36 ± (0.18) | 4.17 ± (0.19) | 3.90 ± (0.31) | < 0.001 |
| ALT, (U/L) | 29,194 | 25.05 ± (19.03) | 27.10 ± (18.08) | 26.36 ± (18.08) | 24.23 ± (21.47) | 22.49 ± (17.95) | < 0.001 |
| AST, (U/L) | 29,194 | 25.11 ± (15.28) | 26.24 ± (13.40) | 25.59 ± (13.25) | 24.43 ± (16.51) | 24.17 ± (17.45) | < 0.001 |
| ALP, (U/L) | 29,194 | 68.67 ± (24.56) | 63.47 ± (18.85) | 65.74 ± (20.28) | 68.69 ± (21.51) | 76.82 ± (32.92) | < 0.001 |
| BUN, (mg/dL) | 29,194 | 13.73 ± (5.44) | 13.18 ± (4.30) | 13.52 ± (4.56) | 13.92 ± (5.27) | 14.29 ± (7.11) | < 0.001 |
| Creatinine, (umol/L) | 29,194 | 78.27 ± (32.41) | 78.27 ± (16.56) | 76.95 ± (18.38) | 77.17 ± (27.43) | 80.65 ± (53.24) | < 0.001 |
| Uric acid, (umol/L) | 29,194 | 322.63 ± (83.87) | 330.10 ± (80.86) | 322.29 ± (81.11) | 319.82 ± (82.36) | 318.18 ± (90.34) | < 0.001 |
| Globulin, (g/L) | 29,194 | 28.38 ± (4.36) | 26.94 ± (3.65) | 27.67 ± (3.91) | 28.52 ± (4.11) | 30.42 ± (4.86) | < 0.001 |
| RDW, (%) | 29,194 | 13.25 ± (1.26) | 12.34 ± (0.52) | 12.86 ± (0.54) | 13.26 ± (0.62) | 14.56 ± (1.66) | < 0.001 |
| Total Cholesterol, (mg/dL) | 29,194 | 193.36 ± (41.64) | 195.31 ± (40.74) | 196.11 ± (42.49) | 192.89 ± (40.12) | 189.12 ± (42.81) | < 0.001 |
| HDL-Cholesterol, (mg/dL) | 29,194 | 53.46 ± (16.44) | 53.43 ± (16.17) | 53.80 ± (16.68) | 53.28 ± (16.21) | 53.34 ± (16.68) | 0.514 |
| BMI group, N(%) | 29,194 | < 0.001 | |||||
| BMI < 18.5 | 448 (1.5%) | 141 (2.2%) | 117 (1.6%) | 104 (1.3%) | 86 (0.8%) | ||
| 18.5 ≤ BMI < 25 | 7795 (28%) | 2428 (39%) | 2002 (29%) | 1738 (24%) | 1627 (18%) | ||
| 25 ≤ BMI < 30 | 9570 (33%) | 2316 (37%) | 2437 (37%) | 2453 (32%) | 2364 (27%) | ||
| BMI ≥ 30 | 11,271 (38%) | 1393 (22%) | 2145 (32%) | 3063 (43%) | 4670 (54%) | ||
| Age group, n (%) | 29,194 | < 0.001 | |||||
| < 65 years | 22,170 (81%) | 5602 (92%) | 5392 (85%) | 5337 (77%) | 5839 (71%) | ||
| ≥ 65 years | 7024 (19%) | 704 (7.8%) | 1348 (15%) | 2042 (23%) | 2930 (29%) | ||
| PIR group, n (%) | 29,194 | < 0.001 | |||||
| PIR < 1.3 | 9376 (22%) | 1809 (19%) | 2004 (20%) | 2324 (21%) | 3239 (28%) | ||
| 1.3 ≤ PIR < 3.5 | 11,074 (36%) | 2258 (33%) | 2487 (34%) | 2886 (37%) | 3443 (38%) | ||
| PIR > 3.5 | 8744 (42%) | 2239 (48%) | 2249 (46%) | 2169 (41%) | 2087 (34%) | ||
| SUI | 29,194 | < 0.001 | |||||
| No | 22,432 (76%) | 5375 (85%) | 5416 (80%) | 5518 (74%) | 6123 (67%) | ||
| Yes | 6762 (24%) | 931 (15%) | 1324 (20%) | 1861 (26%) | 2646 (33%) | ||
| UUI | 29,194 | < 0.001 | |||||
| No | 22,440 (80%) | 5553 (90%) | 5480 (84%) | 5538 (77%) | 5869 (69%) | ||
| Yes | 6754 (20%) | 753 (10%) | 1260 (16%) | 1841 (23%) | 2900 (31%) | ||
| MUI | 29,194 | < 0.001 | |||||
| No | 8872 (91%) | 5989 (96%) | 6213 (93%) | 6552 (89%) | 7413 (84%) | ||
| Yes | 998 (9.3%) | 317 (4.5%) | 527 (7.0%) | 827 (11%) | 1356 (16%) |
P values were calculated by Kruskal–Wallis rank sum test or Pearson’s Chi-squared test. Values were represented by mean (SD) for continuous variables and frequency (percentage) for categorical variables.
aN not Missing (weighted)
bMean ± (SD); N (weighted) (%)
cDesign-based Kruskal Wallis test; Pearson’s X2: Rao and Scott adjustment
dn not Missing (unweighted)
Relationship between the RAR and UI
Table 2 illustrates the investigation of the links between the RAR and UI using weighted univariate and multivariate logistic regression models. Models 1, 2, and 3 were the crude model, minimally adjusted model, and completely adjusted model, respectively.
Table 2.
Multiple logistic regression associations of quartiles of red blood cell distribution width to albumin ratio (RAR) with the prevalence of different types of urinary incontinence (UI) among the general adult population in NHANES 2007–2018, weighted.
| SUI | Crude Model (Model1) OR (95% CI) P-value |
Model 2 OR (95% CI) P-value |
Model 3 OR (95% CI) P-value |
|---|---|---|---|
| RAR | 1.96 (1.80, 2.13) < 0.001 | 1.28 (1.16, 1.41) < 0.001 | 1.23 (1.12, 1.35) < 0.001 |
| RAR quartiles | |||
| Q1 (2.17–2.88) | Ref | Ref | Ref |
| Q2 (2.88–3.10) | 1.47 (1.31, 1.66) < 0.001 | 1.11 (0.67, 1.27) 0.140 | 1.10 (0.96, 1.26) 0.200 |
| Q3 (3.10–3.41) | 2.07 (1.84, 2.32) < 0.001 | 1.28 (1.12, 1.46) < 0.001 | 1.24 (1.09, 1.42) 0.001 |
| Q4 (3.41–13.40) | 2.84 (2.54, 3.18) < 0.001 | 1.47(1.28, 1.69) < 0.001 | 1.40 (1.22, 1.60) < 0.001 |
| P for trend | < 0.001 | < 0.001 | < 0.001 |
| UUI | Crude Model (Model1) OR (95% CI) P-value |
Model 1 OR (95% CI) P-value |
Model 2 OR (95% CI) P-value |
|---|---|---|---|
| RAR | 2.23 (2.06, 2.41) < 0.001 | 1.46 (1.34, 1.59) < 0.001 | 1.43 (1.31, 1.56) < 0.001 |
| RAR quartiles | |||
| Q1 (2.17–2.88) | Ref | Ref | Ref |
| Q2 (2.88–3.10) | 1.72 (1.54, 1.91) < 0.001 | 1.25 (1.11, 1.40) < 0.001 | 1.24 (1.10, 1.40) < 0.001 |
| Q3 (3.10–3.41) | 2.62 (2.29, 2.99) < 0.001 | 1.54 (1.33, 1.77) < 0.001 | 1.50 (1.30, 1.74) < 0.001 |
| Q4 (3.41–13.40) | 3.97 (3.51, 4.49) < 0.001 | 1.91 (1.65, 2.21) < 0.001 | 1.85 (1.60, 2.14) < 0.001 |
| P for trend | < 0.001 | < 0.001 | < 0.001 |
| MUI | Crude Model (Model1) OR (95% CI) P-value |
Model 1 OR (95% CI) P-value |
Model 2 OR (95% CI) P-value |
|---|---|---|---|
| RAR | 2.13 (1.94, 2.34) < 0.001 | 1.47 (1.31, 1.66) < 0.001 | 1.44 (1.27, 1.62) < 0.001 |
| RAR quartiles | |||
| Q1 (2.17–2.88) | Ref | Ref | Ref |
| Q2 (2.88–3.10) | 1.59 (1.34, 1.89) < 0.001 | 1.13 (0.93, 1.36)0.200 | 1.12 (0.92, 1.35)0.300 |
| Q3 (3.10–3.41) | 2.50 (2.08, 3.01) < 0.001 | 1.41 (1.15, 1.73)0.100 | 1.38 (1.12, 1.69)0.003 |
| Q4 (3.41–13.40) | 4.07 (3.42, 4.83) < 0.001 | 1.88 (1.53, 2.31) < 0.001 | 1.81 (1.47. 2.21) < 0.001 |
| P for trend | < 0.001 | < 0.001 | < 0.001 |
Model 1: unadjusted.
Model 2: adjust Age, Sex, Race, Marital status, Education level, PIR.
Model 3: adjust Age, Sex, Race, Marital status, Education level, PIR, Diabetic Mellitus, Hypertension, Vigorous work activity, Moderate work activity, Smoke, Alcohol drinks, ALT, AST, ALP, Creatinine, Blood urea nitrogen, Uric acid, Globulin, Total Cholesterol, HDL-Cholesterol.
OR, Odds ratio, CI, Confidence Interval
The results indicate that in all models, an increased RAR was positively associated with an elevated risk of all three types of UI. In Model 3, each unit increase in the RAR significantly increased the risk of UI across all subtypes (SUI: OR 1.23, 95% CI 1.12–1.35; UUI: OR 1.43, 95% CI 1.31–1.56; MUI: OR 1.44, 95% CI 1.27–1.62; all p < 0.05).
Additionally, we performed a reanalysis by integrating the RAR quartiles (Q1–Q4) into the model. In Model 3, participants in the highest quartile (Q4) demonstrated a markedly elevated risk of all types of UI compared with those in the lowest quartile (Q1) (SUI: OR 1.40, 95% CI 1.22–1.60; UUI: OR 1.85, 95% CI 1.60–2.14; MUI: OR 1.81, 95% CI 1.47–2.21; all p < 0.05).
Detection of nonlinear relationships
Nonrestricted cubic splines (RCSs) were utilized to analyze the link between the RAR and different types of UI, as depicted in Fig. 2. The investigation demonstrated a nonlinear, inverted U-shaped relationship between the RAR and UI. This indicates that the impact of the RAR on UI decreases or becomes insignificant when the RAR surpasses a specific threshold (SUI: P for overall < 0.001, P for nonlinear = 0.041, Fig. 2a; UUI: P for overall < 0.001, P for nonlinear < 0.001, Fig. 2b; MUI: P for overall = < 0.001, P for nonlinear = < 0.001, Fig. 2c).
Fig. 2.
RCS of the relationship between RAR and three types of UI. (a), (b), and (c) denote the linear correlation between RAR and SUI, UUI and MUI, respectively. The area of the two pink zones signifies the 95% confidence interval. The red dashed line suggests a nonlinear correlation between RAR and the three UIs. The light blue bar graph illustrates the population distribution over various RAR levels. The x-axis denotes the degree of RAR, whereas the y-axis indicates the chance of UI.
A threshold effect study was subsequently performed, with the findings displayed in Table 3, demonstrating a significant threshold effect between the RAR and each type of UI, as revealed by the likelihood ratio test (all p < 0.05). The two-piecewise model revealed a consistent inflection point for all three types of UI (SUI: 3.231; UUI: 3.468; MUI: 3.538). Below this inflection point, the RAR exhibited a robust correlation with UI (SUI: OR 1.541, 95% CI 1.301–1.827; UUI: OR 2.069, 95% CI 1.780–2.283; MUI: OR 1.902, 95% CI 1.615–2.242; all p < 0.001). However, beyond the inflection point, the significance of this association greatly diminishes (SUI: OR 1.157, 95% CI 1.064–1.258; UUI: OR 1.080, 95% CI 0.987–1.180; MUI: OR 1.132, 95% CI 1.004–1.269; all p < 0.05). These results indicate that the RAR can serve as a predictive factor for UI risk, but beyond this range, the association weakens or disappears. Therefore, the clinical focus should be on the population whose RAR values fall within this specified range.
Table 3.
Threshould effect analysis of RAR on UI.
| Outcome | Model | Variables | Adjusted OR (95%) | P-value |
|---|---|---|---|---|
| SUI | Fitting by standard linear model | RAR | 1.245 (1.169,1.327) | < 0.001 |
| Fitting by two-piecewise model | Inflection point | 3.231 | ||
| RAR < 3.231 | 1.541 (1.301, 1.827) | < 0.001 | ||
| RAR > 3.231 | 1.157 (1.064,1.258) | 0.001 | ||
| likelihood ratio test | 0.008 | |||
| UUI | Fitting by standard linear model | RAR | 1.367 (1.287,1.561) | < 0.001 |
| Fitting by two-piecewise model | Inflection point | 3.486 | ||
| RAR < 3.486 | 2.069 (1.780, 2.283) | < 0.001 | ||
| RAR > 3.486 | 1.080 (0.987, 1.180) | 0.090 | ||
| likelihood ratio test | < 0.001 | |||
| MUI | Fitting by standard linear model | RAR | 1.367(1.267, 1.475) | < 0.001 |
| Fitting by two-piecewise model | Inflection point | 3.538 | ||
| RAR < 3.538 | 1.902 (1.615, 2.242) | < 0.001 | ||
| RAR > 3.538 | 1.132 (1.004, 1.269) | 0.038 | ||
| likelihood ratio test | < 0.001 |
Model: adjust Age, Sex, Race, Marital status, Education level, PIR, Diabetic Mellitus, Hypertension, Vigorous work activity, Moderate work activity, Smoke, Alcohol drinks, ALT, AST, ALP, Creatinine, Blood urea nitrogen, Uric acid, Globulin, Total Cholesterol, HDL-Cholesterol.
Subgroup analysis
Figure 3 illustrates the results from a subgroup analysis conducted to evaluate the strength of the link between the RAR and UI across several strata, including sex, age, PIR, smoking, alcohol use, diabetes, and hypertension. The results suggest that sex may significantly affect the relationship between the RAR and all types of UI. Compared with that in males, the association between the RAR and SUI was stronger in females (p for interaction < 0.05). This result may be related to anatomical and hormonal differences between sexes. Additionally, as shown in Fig. 3c, a stronger association between the RAR and MUI was observed among smokers (p < 0.05), indicating that smoking may exacerbate inflammation or influence red blood cell distribution, thereby increasing the risk of UI.
Fig. 3.
Subgroup investigation regarding the correlation between RAR and three categories of UI. (a), (b), and (c) represents the results of subgroup analyses for the relationships between and SUI, UUI, and MUI in different stratifications, respectively. All stratified factors include gender, age, the family poverty income ratio, smoking status, alcohol intaking, diabetes, hypertension, except the stratified factor itself.
Discussion
This research gathered data on all individuals with UI from the 2007–2018 NHANES cohort. After screening, 29,194 adult subjects were recruited. The RAR and variables were included in a logistic regression model to examine their associations with UI. In addition, the RAR was changed from a continuous to a quartile variable, and research has consistently shown that there is a high correlation between the RAR and UI. RCSs revealed a nonlinear positive associations among the RAR and SUI, UUI, and MUI. According to the threshold effect study, the relationship between the RAR and UI is stronger before the inflection point and weakens after. Finally, subgroup analysis indicated that sex and smoking had a significant interactive effect on the association between the RAR and UI. After crossing the inflection point, the RCSs gradually stabilized. Research by Weigan Xu and others indicated that the RAR also exhibited a plateau phenomenon similar to that of all-cause mortality in sepsis13. Related studies have also shown a similar plateau effect.
RDW is a critical marker of oxidative stress and chronic inflammation20. Typically, nitric oxide promotes the proliferation of erythroid progenitor cells, whereas inflammatory mediators diminish the capacity of endothelial cells to synthesize nitric oxide, consequently hindering red blood cell synthesis. Furthermore, inflammation hinders red blood cell development by interfering with iron metabolism and diminishing erythropoietin (EPO) levels21. This mechanism ultimately diminishes red blood cell longevity, leading to the occurrence of red blood cells of disparate sizes. Under inflammatory conditions, oxidative stress undermines red blood cell membrane integrity, resulting in an increase in RDW22. Thus, an increased RDW is often linked to chronic inflammation in individuals with diabetes and cardiovascular conditions. A 2014 study revealed that elevated RDW was strongly linked with diabetic nephropathy among individuals with type 2 diabetes23. Additional research indicates that an increased RDW is associated with adverse clinical consequences in individuals with heart failure and coronary artery disease23,24. Furthermore, albumin is the predominant circulating antioxidant, alleviating oxidative stress by binding to prostaglandin E2 and activating immune cells. It also plays a vital role in regulating inflammation, immunological responses, hemostasis, and pH levels25. A prospective observational study conducted by Bardan Ghimire and colleagues revealed that the serum albumin can serve as an indicator for identifying patients at elevated risk of mortality and morbidity26,27.
An increased RDW and reduced albumin levels are significantly correlated with increased inflammation23. The RAR was recently established as a novel composite biomarker of systemic inflammation. Its primary advantage lies in its ability to indirectly assess the body’s inflammatory state using readily available biochemical blood tests. The simple calculation of the RAR facilitates its potential use in developing countries. Studies have shown that the RAR predicts poor outcomes in patients with acute kidney injury (AKI) in intensive care units28. Research conducted by Weigan Xu and colleagues revealed a substantial correlation between an increased RAR and worse outcomes in patients with sepsis, with an increased RAR linked to increased death rates at 28 and 90 days and during hospitalization13. Nevertheless, no studies have investigated the correlation between the RAR and UI. This article presents an inaugural inquiry into the link between the RAR and UI.
Inflammation causes OAB, a common type of UI29. In a 2017 case–control study, patients with SUI presented significantly increased inflammation-related protein levels30. A 2022 systematic review revealed that SUI is linked to changes in extracellular matrix (ECM) metabolism, oxidative stress, apoptosis, and myocyte differentiation and contraction protein expression31. A Boston (BACH) epidemiological survey revealed a consistent link between increased CRP levels and OAB in both men and women, suggesting the involvement of inflammation in OAB32. Several studies have demonstrated that proinflammatory cytokines such as IL-1, IL-5, TNF, and IFN can increase persistent vaginal inflammation and discomfort in female patients, causing UI. UI is also affected by tissue inflammation, remodeling, and repair cytokines such as IL-6, IL-10, and TGF33.
UTIs significantly contribute to the onset of UI34. UTIs induce the intracellular colonization of urinary epithelial cells, triggering an inflammatory response marked by cytokine production and increased permeability of the urine epithelium. Bladder sensory neurons are more sensitive to mechanical bladder wall stretching due to epithelial barrier change. Consequently, this sensitivity causes nocturia as well as urinary urgency and frequency34. Fluctuations in the RAR can predict the severity of UTIs and the likelihood of developing UI, providing essential insights for early diagnosis and therapy.
The results we obtained suggest that sex significantly impacts the prevalence of UI, possibly due to variations in sex hormone levels. According to a meta-analysis of 14 randomized controlled trials (RCTs) involving postmenopausal women, applying estrogen locally to the vaginal region can successfully improve UI35. However, some studies report that systemic hormone replacement therapy may worsen UI36.
This study demonstrates the value of the correlation between the RAR and UI, providing a new perspective for its prediction and evaluation. Existing literature examines predominantly the relationship between a single indicator, such as RDW or serum albumin, and various diseases. In contrast, this study integrates both markers and introduces the RAR as a new indicator, exploring for the first time its link to UI. Compared with traditional markers, the RAR combines two key indicators—reflecting inflammation (RDW) and nutritional status (albumin)—providing the advantage of a multidimensional evaluation of the inflammatory state. This study integrates biological and epidemiological data to improve our understanding of the processes of UI. These findings not only advance our knowledge of the potential mechanisms driving UI but also provide valuable insights for its early detection and management in clinical settings.
This study used cross-sectional data to examine the relationship between the RAR and UI. First, an inability to identify the issue of the temporal order of exposure and outcome was a central limitation of our study, preventing our cross-sectional study from being able to explore causality and highlighting the need for larger, multicenter longitudinal cohort studies. Second, self-reported data from the NHANES may introduce recall bias and information gaps, thereby underestimating the prevalence of UI. In addition, binary questionnaires may not reflect the complexity and severity of urinary incontinence. We are unable to accurately and objectively diagnose UI, and future studies of objective indicators such as voiding diaries, urinary pad tests, and urodynamic tests will improve accuracy and reliability. In addition, because the NHANES database uses data from the United States, more studies are needed to confirm the association of the RAR with UI in different populations in other countries and regions. Finally, the RAR has the potential for intraday or seasonal fluctuations that produce measurement bias, and more accurate measures are needed in the future to capture the dynamic associations between exposures and outcomes.
This study underscores the critical importance of the RAR in predicting inflammation-associated illnesses. Its use provides an essential direction for the prompt identification and management of inflammation-associated disorders.
Conclusion
All three types of UI were positively linked with the RAR according to our findings. The RAR was more strongly associated with SUI in women than in men. The present research provides systematic epidemiologic evidence for the effect of the novel inflammatory marker RAR on UI, but further exploration is needed to explore the potential mechanisms underlying its effect on UI.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
We would like to express our special thanks to Dr. Han and Dr. Liu from the Central Laboratory of Shengli Oilfield Central Hospital for their invaluable technical support in this study and their careful review of the manuscript.
Abbreviations
- PIR
Family poverty-to-income ratio
- SUI
Stress urinary incontinence
- UUI
Urgency urinary incontinence
- MUI
Mixed urinary incontinence
- UTI
Urinary tract infection
- BMI
Body mass index
- RDW
Red blood cell distribution width
- RAR
RDW-to-albumin ratio
- ALT
Alanine transaminase
- AST
Aspartate transaminase
- ALP
Alkaline phosphatase
- BUN
Blood urea nitrogen
- NHANES
National Health and Nutrition Examination Survey
Author contributions
LTX and LYH contributed equally to this work and were responsible for the conceptualization, data analysis, and drafting of the manuscript. YJH provided oversight, contributed to the methodology development, and critically revised the manuscript. WMY contributed to data collection and preprocessing. CB supported the experimental design and provided technical guidance throughout the study. All authors reviewed and approved the final manuscript and agree to be accountable for all aspects of the work.
Data availability
The datasets generated and analyzed in the current study are available at NHANES website: https://www.cdc.gov/nchs/nhanes/index.htm.
Declarations
Competing interests
The authors declare no competing interests.
Ethics declarations
Studies involving human participants were reviewed and approved by the NCHS Research Ethics Review Committee. The patients/participants provided written informed consent to participate in this study.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Mingyue Wang and Cheng Bo contributed equally to this work.
Contributor Information
Jinhui Yang, Email: yjhmx1987@126.com.
Mingyue Wang, Email: myzxyy@outlook.com.
References
- 1.Jw, R. Consensus conference urinary incontinence in adults. JAMA261, 2685–2690. 10.1001/jama.1989.03420180109040 (1989). [DOI] [PubMed] [Google Scholar]
- 2.Gacci, M. et al. European association of urology guidelines on male urinary incontinence. Eur. Urol.82, 387–398. 10.1016/j.eururo.2022.05.012 (2022). [DOI] [PubMed] [Google Scholar]
- 3.Vaughan, C. P. & Markland, A. D. Urinary incontinence in women. Ann. Intern. Med.172, 17–32. 10.7326/aitc202002040 (2020). [DOI] [PubMed] [Google Scholar]
- 4.Ebbesen, M. H., Hunskaar, S., Rortveit, G. & Hannestad, Y. S. Prevalence, incidence and remission of urinary incontinence in women: Longitudinal data from the Norwegian HUNT study (EPINCONT). BMC Urol.13, 27. 10.1186/1471-2490-13-27 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wu, J. M., Hundley, A. F., Fulton, R. G. & Myers, E. R. Forecasting the prevalence of pelvic floor disorders in U.S. women: 2010 to 2050. Obstet. Gynecol.114, 1278–1283. 10.1097/AOG.0b013e3181c2ce96 (2009). [DOI] [PubMed] [Google Scholar]
- 6.Malmsten, U. G., Molander, U., Peeker, R., Irwin, D. E. & Milsom, I. Urinary incontinence, overactive bladder, and other lower urinary tract symptoms: A longitudinal population-based survey in men aged 45–103 years. Eur. Urol.58, 149–156. 10.1016/j.eururo.2010.03.014 (2010). [DOI] [PubMed] [Google Scholar]
- 7.Thom, D. H. et al. Differences in prevalence of urinary incontinence by race/ethnicity. J. Urol.175, 259–264. 10.1016/s0022-5347(05)00039-x (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Zhu, L. et al. The epidemiological study of women with urinary incontinence and risk factors for stress urinary incontinence in China. Menopause16, 831–836. 10.1097/gme.0b013e3181967b5d (2009). [DOI] [PubMed] [Google Scholar]
- 9.Aoki, Y. et al. Urinary incontinence in women. Nat. Rev. Dis. Primers3, 17042. 10.1038/nrdp.2017.42 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Weiss, G., Ganz, T. & Goodnough, L. T. Anemia of inflammation. Blood133, 40–50. 10.1182/blood-2018-06-856500 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Eckart, A. et al. Relationship of nutritional status, inflammation, and serum albumin levels during acute illness: A prospective study. Am. J. Med.133, 713-722.e717. 10.1016/j.amjmed.2019.10.031 (2020). [DOI] [PubMed] [Google Scholar]
- 12.Levitt, D. G. & Levitt, M. D. Human serum albumin homeostasis: A new look at the roles of synthesis, catabolism, renal and gastrointestinal excretion, and the clinical value of serum albumin measurements. Int. J. Gener. Med.9, 229–255. 10.2147/ijgm.S102819 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Xu, W. et al. Association between red blood cell distribution width to albumin ratio and prognosis of patients with sepsis: A retrospective cohort study. Front. Nutr.9, 1019502. 10.3389/fnut.2022.1019502 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Li, D., Ruan, Z. & Wu, B. Association of red blood cell distribution width-albumin ratio for acute myocardial infarction patients with mortality: A retrospective cohort study. Clin. Appl. Thromb. Hemost.28, 10760296221121286. 10.1177/10760296221121286 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Kimura, H. et al. Impact of red blood cell distribution width-albumin ratio on prognosis of patients with CKD. Sci. Rep.13, 15774. 10.1038/s41598-023-42986-2 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Hong, J. et al. Impact of red cell distribution width and red cell distribution width/albumin ratio on all-cause mortality in patients with type 2 diabetes and foot ulcers: A retrospective cohort study. Cardiovasc. Diabetol.21, 91. 10.1186/s12933-022-01534-4 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Ni, Q., Wang, X., Wang, J. & Chen, P. The red blood cell distribution width-albumin ratio: A promising predictor of mortality in heart failure patients—A cohort study. Clinica Chimica Acta Int. J. Clin. Chem.527, 38–46. 10.1016/j.cca.2021.12.027 (2022). [DOI] [PubMed] [Google Scholar]
- 18.Zhao, N. et al. The red blood cell distribution width-albumin ratio: A promising predictor of mortality in stroke patients. Int. J. Gener. Med.14, 3737–3747. 10.2147/ijgm.S322441 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Chen, C. et al. Relationship between the ratio of red cell distribution width to albumin and 28-day mortality among Chinese patients over 80 years with atrial fibrillation. Gerontology69, 1471–1481. 10.1159/000534259 (2023). [DOI] [PubMed] [Google Scholar]
- 20.Hao, M. et al. Ratio of red blood cell distribution width to albumin level and risk of mortality. JAMA Netw. Open7, e2413213. 10.1001/jamanetworkopen.2024.13213 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Chen, C. et al. Oxybutynin ameliorates LPS-induced inflammatory response in human bladder epithelial cells. J. Biochem. Mol. Toxicol.38, e23584. 10.1002/jbt.23584 (2024). [DOI] [PubMed] [Google Scholar]
- 22.Namazi, G. et al. Relationship between red cell distribution width and oxidative stress indexes in patients with coronary artery disease. Rep. Biochem. Mol. Biol.12, 241–250. 10.61186/rbmb.12.2.241 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Magri, C. J. & Fava, S. Red blood cell distribution width and diabetes-associated complications. Diabetes Metabol. Syndrome8, 13–17. 10.1016/j.dsx.2013.10.012 (2014). [DOI] [PubMed] [Google Scholar]
- 24.Tsuboi, S. et al. Impact of red blood cell distribution width on long-term mortality in diabetic patients after percutaneous coronary intervention. Circ. J.77, 456–461. 10.1253/circj.cj-12-0730 (2013). [DOI] [PubMed] [Google Scholar]
- 25.Pompili, E., Zaccherini, G., Baldassarre, M., Iannone, G. & Caraceni, P. Albumin administration in internal medicine: A journey between effectiveness and futility. Eur. J. Intern. Med.117, 28–37. 10.1016/j.ejim.2023.07.003 (2023). [DOI] [PubMed] [Google Scholar]
- 26.Domínguez de Villota, E. et al. Association of a low serum albumin with infection and increased mortality in critically ill patients. Intensive Care Med.7, 19–22. 10.1007/bf01692917 (1980). [DOI] [PubMed] [Google Scholar]
- 27.Ghimire, B. et al. Serial estimations of serum albumin levels as a prognostic marker in critically ill patients admitted in ICU in tertiary center: An observational study. Medicine102, e35979. 10.1097/md.0000000000035979 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Gao, C. & Peng, L. Association and prediction of red blood cell distribution width to albumin ratio in all-cause mortality of acute kidney injury in critically ill patients. Front. Med.10, 1047933. 10.3389/fmed.2023.1047933 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Funahashi, Y. et al. Bladder overactivity and afferent hyperexcitability induced by prostate-to-bladder cross-sensitization in rats with prostatic inflammation. J. Physiol.597, 2063–2078. 10.1113/jp277452 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Koch, M. et al. Serum proteomic pattern in female stress urinary incontinence. Electrophoresis39, 1071–1078. 10.1002/elps.201700423 (2018). [DOI] [PubMed] [Google Scholar]
- 31.Post, W. M. et al. Molecular processes in stress urinary incontinence: A systematic review of human and animal studies. Int. J. Mol. Sci.10.3390/ijms23063401 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Kupelian, V. et al. Association of overactive bladder and C-reactive protein levels. Results from the Boston Area Community Health (BACH) Survey. BJU Int.110, 401–407. 10.1111/j.1464-410X.2011.10769.x (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Chen, Y. et al. Cytokine modulation in pelvic organ prolapse and urinary incontinence: From molecular insights to therapeutic targets. Mol. Med.30, 214. 10.1186/s10020-024-00989-3 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Mansfield, K. J., Chen, Z., Moore, K. H. & Grundy, L. Urinary tract infection in overactive bladder: An update on pathophysiological mechanisms. Front. Physiol.13, 886782. 10.3389/fphys.2022.886782 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Weber, M. A. et al. Local oestrogen for pelvic floor disorders: A systematic review. PLoS ONE10, e0136265. 10.1371/journal.pone.0136265 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Cody, J. D., Richardson, K., Moehrer, B., Hextall, A. & Glazener, C. M. Oestrogen therapy for urinary incontinence in post-menopausal women. Cochrane Database System. Rev.10.1002/14651858.CD001405.pub2 (2009). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and analyzed in the current study are available at NHANES website: https://www.cdc.gov/nchs/nhanes/index.htm.