The official NHANES website¹ served as the primary source of information for this study. NHANES is a comprehensive research initiative conducted by the Centers for Disease Control and Prevention (CDC) in collaboration with the National Center for Health Statistics (NCHS). Its purpose is to provide nationally representative statistics regarding the civilian non-institutionalized population in the United States. All participants provided informed consent prior to interviews and testing, and the NHCS Ethics Review Board approved the NHANES data collection procedures. In this survey, a dataset was created by extracting responses from the publicly accessible NHANES data files between 2005 and 2018. A total of 70,190 individuals participated in two consecutive NHANES survey cycles. Given that OAB is primarily an age-related condition, we excluded individuals under the age of 50 (n = 50,495), focusing specifically on the elderly population who are at a higher risk of developing OAB. Epidemiological studies consistently indicate that the prevalence of OAB significantly increases after the age of 50 (16). Subsequently, individuals without calculated GNRI data (n = 913) and those missing OAB symptom data (n = 1,621) were also excluded. To address potential issues associated with missing data in multivariable analysis, we employed multiple imputation techniques to handle missing values of covariates. The final study population comprised 17,161 individuals. Figure 1 displays a flowchart detailing the criteria used for patient selection.

According to the Lorenz formula, the ideal body weight for males is calculated using the formula: height − 100 − [(height − 150)/4]. The specific formula for calculating the GNRI is: GNRI = [41.7 × present body weight (Kg) / ideal body weight (Kg)] + [1.489 serum albumin(g/L)] (17).

If the ratio of current body weight to ideal body weight is one or greater, it is adjusted to equal one for the calculation. Previous studies have calculated the GNRI cut-off values using the critical values of serum albumin and weight loss, thereby determining the GNRI nutritional assessment levels: high nutritional risk (GNRI <98) and low nutritional risk (GNRI ≥98) (11). Patients included in this study were divided into two groups: the high GNRI group (GNRI ≥98) and the low GNRI group (GNRI <98) (18).

According to the definition of OAB, when patients experience urgent urinary incontinence and nocturia, the presence of OAB should be considered (19). At the same time, other potential causes of urination symptoms need to be considered, such as urinary tract infections, prostate diseases, and gynecological diseases. The NHANES database utilizes the Kidney Conditions-Urology questionnaire to assess OAB-related symptoms. Specifically, participants are asked the questions: “During the past 12 months, have you leaked or lost control of your urine, even just a little, because you felt the sudden urge or pressure to urinate and could not get to the toilet in time?” and “How often does this happen?” These two questions evaluate the severity of urgent urinary incontinence (UUI) (20). The severity of nocturia is assessed by the question: “During the past 30 days, how many times do you typically get up to urinate from the time you go to bed at night until you wake up in the morning?” To reinforce the diagnosis of OAB, we used the symptom frequency conversion criteria recorded in NHANES and OABSS scoring to further quantify the symptoms of overactive bladder, with specific scoring criteria shown in Table 1. Finally, the nocturia score and the urgency urinary incontinence score are added together to obtain the overall OABSS score for each participant (21). Individuals with a total score of ≥3 points are considered to be diagnosed with overactive bladder.

Interviewers determined participants’ date of birth and other demographic data through an initial screening questionnaire, reported by the participants themselves. Age, reported in years, is a continuous variable. Gender is reported as male or female. Race is categorized into Mexican Americans, Other Hispanics, Non-Hispanic White, Non-Hispanic Black, or Other. Body Mass Index (BMI) is calculated as follows: BMI = weight(kg) / height(m^2). Educational attainment is assessed through the initial screening questionnaire and is categorized into Less than high school, High school or equivalent, and College or above. Marital status is divided into Married/Living with Partner, never married, Widowed/Divorced/Separated. Household income is stratified based on the poverty income ratio (22) into three categories: PIR < 1.3, 1.3 ≤ PIR < 3.5, and PIR ≥ 3.5 (23). Smoking status is determined by answering the questions: “Have you smoked at least 100 cigarettes in your entire life?” and “Do you now smoke cigarettes?” Classification of an individual’s smoking status is based on the duration and amount of smoking throughout their lifetime. There are three distinct categories of smoking populations: “Never,” “Former,” and “Current” smokers. The study divides participants into two groups based on whether they engage in excessive alcohol consumption daily. Hypertension is diagnosed based on four criteria, and individuals meeting any of these criteria are classified as hypertensive. The diagnostic criteria include self-reported history of hypertension, use of antihypertensive medication, systolic blood pressure (SBP) ≥ 140 mmHg, or diastolic blood pressure (DBP) ≥ 90 mmHg (24). Participants meeting either of the following two criteria are defined as having diabetes. The diagnostic criteria for diabetes are: the subject uses antidiabetic medication; the patient has a self-reported diagnosis of diabetes (25).

This study utilized Mobile Examination Center sample weights and associated home examination sample design variables to create weights. These weights were employed to ensure representative results consistent with the demographic characteristics of the entire U.S. population. All analyses in this study were conducted with appropriate weighting of the samples according to the NHANES Analytic Guidelines. Continuous variables were reported as weighted means with standard errors, and categorical variables were reported as numerical proportions. Weighted t-tests for continuous variables and weighted chi-square tests for categorical variables were used to assess group differences. After adjusting for potential demographic confounding factors, weighted multivariate logistic regression models analyzed the independent association between high GNRI and OAB risk. Subgroup analyses were stratified by age, gender, race, BMI, marital status, education level, poverty rate, smoking, alcohol consumption, and related diseases to examine the correlation of GNRI in different subgroups. To further evaluate the predictive performance of GNRI and other indicators (BMI and serum albumin), we utilized receiver operating characteristic (ROC) curves and the area under the curve (AUC) values. Additionally, causal mediation analysis was conducted to explore the roles of different mediators in this context, estimating the proportions of direct and indirect effects in the association between GNRI and OAB risk. Through causal mediation analysis, the relative contributions of inflammatory cell clusters and cognitive function to the interaction between GNRI and OAB were assessed. Statistical tests were two-tailed, with p < 0.05 considered statistically significant. All analyses were performed using R software version 4.4.1 and Free Statistics software version 1.8.

Table 1 presents the symptom scoring criteria for participants, who were divided into two groups based on the diagnosis of OAB. A total of 17,161 participants were screened for inclusion in this study. In the preliminary analysis, age (p < 0.0001) was identified as a significant factor affecting OAB (Table 2). The mean age of participants with OAB was higher than that of those without OAB, indicating a higher risk of disease in the elderly population. Additionally, there was a significant intergroup difference between females and males (59.85% vs. 40.15%), suggesting a higher risk of disease in the elderly female population. The data indicated a higher proportion of individuals with lower education levels in the OAB group, while the non-OAB group was primarily composed of individuals with a college or higher education level. This suggests that educational attainment may be associated with the risk of OAB (p < 0.0001). Furthermore, a higher proportion of divorced or widowed individuals was observed in the OAB group compared to the non-OAB group, which had a higher proportion of married individuals (p < 0.0001). It is thus hypothesized that changes in marital status may indirectly affect health status.

Regarding economic factors, a higher proportion of individuals with higher poverty income ratios (22) were found in the non-OAB group, while the OAB group was predominantly concentrated in middle and low-income households. Economic hardship may contribute to the occurrence of OAB (p < 0.0001). Analysis of disease factors indicated a higher prevalence of hypertension in the OAB group (69.59% vs. 30.41%, p < 0.0001) and a lower prevalence of diabetes (33.05% vs. 66.95%, p < 0.0001). Analysis of unhealthy habits showed no significant correlation between smoking and alcohol consumption and the occurrence of OAB.

The relationship between the GNRI and the prevalence of OAB was explored using multivariate logistic regression analysis (Table 3), with GNRI divided into two groups. Data analysis revealed that, compared to the low nutritional risk index (Low-GNRI), there was a varying degree of negative correlation between high GNRI levels and OAB across the three models. In the unadjusted model, the results showed a significant association between High-GNRI and OAB prevalence. In the adjusted Model I, demographic parameters such as age, gender, race, and BMI were included as adjustment variables. Adjusted Model II included additional adjustments for education, marital status, PIR, smoking, alcohol consumption, diabetes, and hypertension, in addition to age, gender, race, and BMI. The results were consistent with the unadjusted model. The OR values and p values indicated that for each one-unit increase in GNRI, the risk of OAB prevalence was reduced by 0.75 times (p = 0.002), 0.74 times (p = 0.002), and 0.81 times (p = 0.02) in the three models, respectively. Therefore, it can be preliminarily inferred that there is a negative correlation between GNRI levels and the occurrence of OAB, and a high nutritional risk index may be a protective factor in preventing the development of OAB.

The restricted cubic spline regression model demonstrated a non-linear U-shaped association between GNRI and the risk of OAB in the elderly population (P-nonlinear = 0.0029; Figure 2). The negative correlation was limited to when GNRI was less than 63, with a lower risk of OAB in the elderly. However, after the inflection point, the log odds ratio (log OR) of the OAB risk continued to increase with higher levels of GNRI.

This study employed subgroup analysis to ascertain whether the relationship between OAB and GNRI varied across different subgroups (Figure 3). After adjusting for covariates, we observed no significant differences in the association between OAB and GNRI across subgroups. The relationship between GNRI and the risk of OAB remained consistent across various subgroups, including gender, ethnicity, BMI, PIR, education level, marital status, smoking status, alcohol consumption, hypertension, and diabetes. Specifically, in the elderly population aged 50 to 70 years, the risk of OAB is lower when the GNRI values are below 63. However, beyond this threshold, the risk of OAB increases progressively with higher GNRI values. In contrast, in the population aged 70 to 80 years, GNRI showed a linear negative correlation with the risk of OAB (Figure 4).

The ROC curves for the predictive abilities of GNRI, serum albumin, and BMI for OAB risk are shown in Figure 5. The AUC for GNRI in the ROC analysis was 0.603, which is higher than that for serum albumin (AUC = 0.578) and BMI (AUC = 0.571). This suggests that, compared to albumin or BMI alone, GNRI may be a more suitable predictor of OAB, although the diagnostic accuracy remains limited.

Among the tested inflammatory markers, the systemic immune inflammation index (SII) exhibited the highest mediation proportion (4.8%), followed by basophils (0.9%), monocytes (0.4%), white blood cells (0.1%), and lymphocytes (0.1%) (Table 4). All inflammatory response markers had a significant overall effect on the association between GNRI and OAB (p < 0.001). Cognitive function was measured using the Consortium to Establish a Registry for Alzheimer’s Disease (CERAD), Animal Fluency Test (AFT), and Digit Symbol Substitution Test (DSST) (26, 27). Mediation analysis of cognitive function revealed that the digit symbol score had both direct and indirect effects between GNRI and OAB (p < 0.05). The animal fluency score total had only an indirect effect (p < 0.001). CERAD confirmed that Alzheimer’s disease had no influence on either.