Abstract
This study aims to evaluate the association between the geriatric nutritional risk index (GNRI) and urinary incontinence (UI). This cross-sectional study utilized data derived from 9 cycles (2001–2018) of the National Health and Nutrition Examination Survey. Weighted multivariable logistic regression models were employed to assess the connection between GNRI and UI [stress UI (SUI), urgency UI (UUI), and mixed UI (MUI)]. Smooth curve fitting was conducted to explore the linear association between GNRI and UI. Subgroup analysis was used to assess the robustness of the relationship between GNRI and UI across various stratifications. In the current study, a total of 11,324 participants older than 60 years were enrolled, with 32.2% experiencing SUI, 36.5% complaining of UUI, and 17.8% having MUI. In the multivariable logistic regression, GNRI was negatively associated with 3 types of UI with full adjustment (SUI: odds ratio [OR] = 0.986, 95% confidence interval [CI]: 0.974–0.999; UUI: OR = 0.972, 95% CI: 0.961–0.983; MUI: OR = 0.972, 95% CI: 0.957–0.987, all P < .05). After categorizing GNRI into quartiles, a significant negative relationship between GNRI and UI was stable, particularly evident in the highest GNRI group when compared to the lowest GNRI quartile (SUI: OR = 0.797, 95% CI: 0.672–0.945; UUI: OR = 0.722, 95% CI: 0.615–0.848; MUI: OR = 0.674, 95% CI: 0.538–0.845, all P for trend < .05). Additionally, the smooth curves fitting revealed a negative linear connection between GNRI and UI. Moreover, the negative association between GNRI and SUI, as well as MUI, was more significant in men than in women (P for interaction < .05). In conclusion, GNRI was inversely associated with the prevalence of UI. The association between GNRI and SUI, as well as MUI, was stronger in males compared to females. Further research is needed to confirm causality and explore potential mechanisms.
Keywords: geriatric nutritional risk index, National Health and Nutrition Examination Survey, NHANES, nutrition, urinary incontinence
1. Introduction
Urinary incontinence (UI) is the involuntary leakage of urine, which is common in older adults but can also affect younger adults and significantly impacts both health and quality of life.[1] Approximately 24 to 45% of women reported UI; in women older than 60, approximately 9 to 39% reported UI on a daily basis.[1] And UI has a reported prevalence of 11% in men aged 60 to 64 years and 31% in men ≥ 85 years of age.[2] UI can be classified into stress UI (SUI), urgency UI (UUI), and mixed UI (MUI). SUI indicates the involuntary discharge of urine during physical exertion, coughing, or sneezing.[3] UUI refers to the involuntary voiding of urine accompanied by a sudden urge. MUI constitutes a condition wherein involuntary urine leakage occurs with urgency, as well as with physical activities, coughing, or sneezing.[4] Various risk factors contribute to UI, encompassing but not confined to aging, complications arising from childbirth, medications, sedentary lifestyle, and obesity.[5] Although UI significantly impacts quality of life, its underlying causes and mechanisms are not yet fully understood.
An increasing number of studies have shown an association between UI and nutrient intake. A cross-sectional study conducted by Yuan et al demonstrated that the excessive intake of added sugar among women may increase their risk of SUI and MUI.[6] A propensity scores matching study involving 8090 women showed that UI was associated with high carbohydrate intake.[7] They think it may be associated with obesity and blood sugar fluctuations brought on by carbohydrates. Choi et al revealed that unhealthy dietary habits (such as a preference for salty foods, frequent intake of sweets, and insufficient vegetable consumption) are associated with the occurrence and severity of lower urinary tract symptoms.[8] Xu et al found that women with higher serum albumin levels had a lower risk of SUI, suggesting that higher nutritional status may reduce the risk of developing SUI.[9] Nutrient levels seem to play a role in the occurrence of UI.
The geriatric nutritional risk index (GNRI), a simple and objective method for assessing the nutritional status of older adults, can be easily calculated using serum albumin levels and ideal body weight.[10] Studies have found that GNRI is a better predictor of nutritional status than body mass index (BMI) and serum albumin level alone.[11] Multiple studies have found that GNRI has a significant predictive role in the prognosis of cardiovascular disease and tumors.[12–17] GNRI, as an indicator that better reflects nutritional status, reduces the mutual confounding of BMI and serum albumin. However, to our knowledge, the relationship between GNRI and UI has not been explored. This study will focus on the association using data from the National Health and Nutrition Examination Survey (NHANES) database.
2. Materials and methods
2.1. Study description and population
The data utilized in this study were obtained from the NHANES dataset, a prominent population-based initiative conducted by the Centers for Disease Control and Prevention’s National Center for Health Statistics. The primary objective of NHANES is to assess the health and nutritional status of the United States (U.S.) population, and a distinctive feature of its survey is the combination of interviews and physical examinations. The NHANES employed a sophisticated stratified multistage probability design to gather a nationally representative sample from the noninstitutionalized civilian population of the U.S. More detailed information can be accessed at: https://www.cdc.gov/nchs/nhanes/index.html.
In this study, we obtained data from 9 NHANES cycles (2001–2018) and included a total of 42,247 participants at first. Survey individuals aged ≤ 60 years old (n = 28,951), without GNRI data (n = 1,263), without UI information (n = 650), and with missing covariates (n = 59) were excluded. Finally, 11,324 participants were enrolled in this study (Fig. 1).
Figure 1.
Flow chart of sample selection process. GNRI = geriatric nutritional risk index, MUI = mixed urinary incontinence, n = number of participants, NHANES = National Health and Nutrition Examination Survey, SUI = stress urinary incontinence, UUI = urgency urinary incontinence.
2.2. Measurement of GNRI
GNRI is a risk index related to nutritional status and serves as a strong prognostic indicator for morbidity and mortality, especially in hospitalized older adults. GNRI was calculated as: GNRI = [1.489 × albumin (g/L)] + {41.7 × [present weight (kg)/ ideal weight (kg)]}. The ideal weight was defined as height2 (m2) × 22 (kg/m2).[10] Generally, a higher GNRI meant the higher nutrition-related risk. Based on the quartile of GNRI, GNRI was divided into 4 groups (Q1, Q2, Q3, Q4). GNRI was set as a continuous and categorical variable in the analyses.
2.3. Assessment of UI
The status of UI was evaluated by 2 questions in the “Kidney Condition” in the NHANES. Survey individuals who answered “yes” to the question “During the past 12 months, have you leaked or lost control of even a small amount of urine with an activity like coughing, lifting, or exercise?” were defined as having SUI. Participants were set as having UUI when they said “yes” to the question “During the past 12 months, have you leaked or lost control of even a small amount of urine with an urge or pressure to urinate and you couldn’t get to the toilet fast enough?.” Furthermore, if participants answered “yes” to both questions, they were regarded as having MUI.
2.4. Covariates of interest
In this study, gender, age, race/ethnicity, education level, marital status, the family poverty-to-income ratio (PIR), BMI, smoking status, alcohol intake, vigorous activity, moderate activity, diabetes, and hypertension were designed as covariates of interest. In order to avoid the reduction of the substantial sample size within our study, we treated numerous missing covariates, the family PIR (n = 1024) and alcohol intake (n = 1714), as missing value categories, which were subsequently encoded as dummy variables in our regression models. Participants meeting the criterion of having smoked at least 100 cigarettes throughout their lifetime and presently smoking every day or some days at the time of the interview were categorized as current smokers. Those who acknowledged a history of having smoked at least 100 cigarettes but were not currently smoking at the time of the questionnaire were designated as former smokers. Furthermore, for the classification of survey respondents as nonsmokers, it was stipulated that individuals reporting smoking fewer than 100 cigarettes over their lifetime would be encompassed within this category. Participants were stratified into 2 distinct groups: the “drinkers,” defined as individuals who reported an annual consumption of a minimum of 12 alcoholic beverages, and the “nondrinkers.” Survey participants who had received a medical diagnosis of diabetes prior to the interview or exhibited a fasting plasma glucose level ≥ 126 mg/dL were categorized as individuals with diabetes. Participants were classified as having hypertension if they had received a medical diagnosis from doctors, were under medication for hypertension, or presented with a systolic blood pressure level ≥ 140 mm Hg or a diastolic blood pressure level ≥ 90 mm Hg.
2.5. Statistical analysis
In the current study, we applied the sample weights, stratifications, and clustering in the NHANES dataset in all statistical analyses to consider the multistage and complex sampling design employed to ensure the representation of a noninstitutionalized civilian U.S. population. Continuous variables were represented as weighted mean and standard error, while categorical variables were set as weighted proportions and 95% confidence interval (CI). To assess baseline characteristics across the 4 groups categorized by GNRI quartiles, survey-weighted linear regression was utilized for continuous variables, while a survey-weighted chi-square test was employed for categorical variables. Multivariable logistic regression analysis was used to evaluate the connection between GNRI and UI. In model 1, no covariates were adjusted. In model 2, gender, age, and race were adjusted. In model 3, gender, age, race, education level, marital status, the family PIR, BMI, smoking status, alcohol intake, vigorous activity, moderate activity, diabetes, and hypertension were adjusted. The use of smooth curve fitting and generalized additive models allowed for an examination of the potential segmentation of GNRI into different intervals, thereby enabling an assessment of the nonlinear relationship between the GNRI and UI. Subgroup analysis was employed to explore the stability of the association between GNRI and UI across various stratifications. Statistical significance was established at a 2-sided P value < .05. EmpowerStats (http://www.empowerstats.com, X&Y Solutions, Inc.) and statistical software packages R (http://www.R-project.org; The R Foundation) were used in all statistical analyses.
3. Results
3.1. Participant characteristics
The current study included a sample of 11,324 participants with an average age (standard error) of 70.51 (0.10) years. The weighted proportions of male and female participants were 45.0% and 55.0%, respectively. Of the study individuals, 32.1% disclosed a history of SUI, 36.5% manifested a previous occurrence of UUI, and 17.8% reported symptoms indicative of MUI. Based on the quartiles of GNRI, participants in the highest quartile had a lower risk of 3 types of UI (all P < .05). More detailed baseline characteristics are presented in Table 1.
Table 1.
Baseline characteristics of participants by the GNRI quartiles, weighted.
| Total | Q1 | Q2 | Q3 | Q4 | P value | |
|---|---|---|---|---|---|---|
| Age (yr), mean (SE) | 70.51 (0.10) | 71.66 (0.22) | 70.81 (0.17) | 70.42 (0.22) | 69.80 (0.14) | < .001 |
| BMI (kg/m2), mean (SE) | 29.00 (0.09) | 29.14 (0.23) | 29.83 (0.14) | 28.53 (0.20) | 28.44 (0.12) | < .001 |
| Gender, % (95% CI) | < .001 | |||||
| Male | 45.0 (44.0–45.9) |
34.6 (31.5–37.8) |
41.4 (39.3–43.6) |
45.6 (42.2–49.1) |
52.1 (50.3–53.8) | |
| Female | 55.0 (54.1–56.0) | 65.4 (62.2–68.5) | 58.6 (56.4–60.7) | 54.4 (50.9–57.8) | 47.9 (46.2–49.7) | |
| Age, % (95% CI) | < .001 | |||||
| < 70 | 49.8 (48.4–51.2) | 43.0 (40.1–45.9) | 47.7 (45.2–50.2) | 50.9 (47.6–54.2) | 54.1 (52.0–56.2) | |
| ≥ 70 | 50.2 (48.8–51.6) | 57.0 (54.1–59.9) | 52.3 (49.8–54.8) | 49.1 (45.8–52.5) | 45.9 (43.8–48.0) | |
| Race/Ethnicity, % (95% CI) | < .001 | |||||
| Mexican American | 3.8 (3.0–4.9) | 4.4 (3.3–5.7) | 3.7 (2.8–4.7) | 3.9 (2.8–5.2) | 3.8 (2.9–4.9) | |
| Other Hispanic | 3.2 (2.6–4.0) | 3.0 (2.3–4.1) | 3.5 (2.8–4.4) | 3.3 (2.4–4.4) | 3.1 (2.4–4.0) | |
| Non-Hispanic White | 80.2 (78.1–82.1) | 74.4 (71.4–77.2) | 80.0 (77.7–82.1) | 80.7 (77.7–83.3) | 82.6 (80.3–84.8) | |
| Non-Hispanic Black | 7.9 (6.8–9.1) | 13.2 (11.3–15.4) | 8.4 (7.1–9.9) | 7.4 (6.1–9.0) | 5.3 (4.5–6.3) | |
| Other race | 4.9 (4.2–5.6) | 5.0 (3.8–6.6) | 4.4 (3.7–5.4) | 4.8 (3.7–6.1) | 5.2 (4.2–6.4) | |
| Education level, % (95% CI) | < .001 | |||||
| < high school | 19.8 (18.3–21.4) | 24.1 (21.7–26.6) | 20.3 (18.4–22.4) | 19.4 (16.9–22.2) | 17.7 (15.8–19.7) | |
| High school or GED | 25.8 (24.6–27.1) | 27.4 (24.8–30.1) | 26.8 (24.4–29.3) | 21.8 (18.7–25.2) | 25.9 (23.9–28.0) | |
| > high school | 54.4 (52.6–56.2) | 48.6 (45.5–51.6) | 52.9 (50.1–55.7) | 58.8 (54.8–62.7) | 56.5 (53.6–59.3) | |
| Marital status, % (95% CI) | < .001 | |||||
| Living alone | 36.6 (35.2–38.1) | 46.4 (43.7–49.2) | 38.7 (36.6–40.9) | 32.8 (29.4–36.2) | 32.1 (30.2–34.2) | |
| Married or living with partner | 63.4 (61.9–64.9) | 53.6 (50.9–56.3) | 61.3 (59.1–63.4) | 67.3 (63.8–70.6) | 67.9 (65.8–69.8) | |
| Family PIR, % (95% CI) | < .001 | |||||
| < 1.3 | 16.4 (15.2–17.8) | 21.1 (18.8–23.6) | 17.1 (15.5–18.8) | 16.2 (13.9–18.8) | 14.0 (12.5–15.7) | |
| > 1.3 and ≤ 3.5 | 38.6 (37.0–40.2) | 39.0 (35.6–42.6) | 38.7 (36.4–41.0) | 37.5 (33.9–41.3) | 38.6 (36.3–41.1) | |
| > 3.5 | 37.1 (34.9–39.3) | 30.7 (26.5–35.3) | 36.4 (34.0–38.9) | 37.4 (33.1–41.9) | 40.2 (37.5–42.9) | |
| Unclear | 8.0 (7.1–8.8) | 9.2 (7.2–11.6) | 7.9 (6.7–9.2) | 8.9 (7.0–11.3) | 7.1 (6.1–8.4) | |
| BMI, % (95% CI) | < .001 | |||||
| < 25 | 25.7 (24.5–27.0) | 33.9 (30.9–37.0) | 20.9 (19.1–22.8) | 29.7 (26.5–33.1) | 24.8 (22.9–26.8) | |
| ≥ 25 and < 30 | 37.0 (35.7–38.2) | 24.0 (21.2–27.0) | 36.4 (34.2–38.7) | 34.8 (31.5–38.2) | 43.8 (41.8–45.8) | |
| ≥ 30 | 37.3 (36.0–38.7) | 42.1 (39.1–45.2) | 42.7 (40.5–44.9) | 35.5 (31.8–39.3) | 31.5 (29.6–33.4) | |
| Hypertension, % (95% CI) | .225 | |||||
| No | 30.8 (29.4–32.3) | 28.8 (25.8–31.9) | 30.1 (27.7–32.6) | 33.0 (29.7–36.5) | 31.5 (29.5–33.6) | |
| Yes | 69.2 (67.7–70.6) | 71.3 (68.1–74.2) | 69.9 (67.4–72.3) | 67.0 (63.5–70.3) | 68.5 (66.4–70.5) | |
| Diabetes, % (95% CI) | < .001 | |||||
| No | 77.3 (76.3–78.3) | 73.2 (70.3–76.0) | 76.3 (74.6–77.9) | 79.5 (76.7–82.0) | 79.1 (77.2–80.78) | |
| Yes | 22.7 (21.7–23.7) | 26.8 (24.0–29.7) | 23.7 (22.1–25.5) | 20.5 (18.0–23.3) | 21.0 (19.2–22.8) | |
| Vigorous activity, % (95% CI) | < .001 | |||||
| No | 85.1 (83.9–86.3) | 90.1 (87.8–92.1) | 85.1 (83.1–86.9) | 85.0 (82.2–87.5) | 83.0 (81.2–84.8) | |
| Yes | 14.9 (13.7–16.1) | 9.9 (8.0–12.2) | 14.9 (13.1–16.9) | 15.0 (12.5–17.8) | 17.0 (15.3–18.8) | |
| Moderate activity, % (95% CI) | < .001 | |||||
| No | 59.5 (57.8–61.2) | 68.4 (65.0–71.6) | 59.9 (57.2–62.5) | 59.7 (56.4–62.9) | 55.4 (52.7–58.1) | |
| Yes | 40.5 (38.8–42.2) | 31.6 (28.4–35.0) | 40.2 (37.6–42.8) | 40.4 (37.1–43.7) | 44.6 (41.9–47.3) | |
| Smoking status, % (95% CI) | < .001 | |||||
| Current smokers | 11.0 (10.3–11.9) | 15.4 (13.1–17.9) | 10.9 (9.7–12.3) | 10.9 (8.9–13.4) | 9.4 (8.2–10.7) | |
| Former smokers | 40.0 (38.7–41.4) | 37.8 (35.0–40.6) | 38.4 (36.1–40.7) | 38.5 (35.1–41.9) | 42.9 (40.7–45.0) | |
| Nonsmokers | 49.0 (47.5–50.4) | 46.9 (43.9–49.8) | 50.7 (48.2–53.2) | 50.6 (46.8–54.5) | 47.8 (45.4–50.2) | |
| Alcohol intaking, % (95% CI) | < .001 | |||||
| Nondrinkers | 27.7 (26.0–29.6) | 28.4 (25.5–31.4) | 28.3 (26.2–30.5) | 29.9 (26.8–33.3) | 26.2 (23.8–28.7) | |
| Drinkers | 55.6 (53.3–57.9) | 42.1 (38.4–45.9) | 50.4 (47.7–53.0) | 57.1 (53.1–61.1) | 65.3 (62.3–68.1) | |
| Unclear | 16.6 (14.7–18.7) | 29.5 (25.4–34.1) | 21.4 (18.9–24.1) | 13.0 (10.0–16.7) | 8.6 (6.6–11.1) | |
| SUI, % (95% CI) | < .001 | |||||
| No | 67.8 (66.7–69.0) | 60.5 (57.3–63.6) | 66.1 (63.8–68.3) | 69.4 (66.3–72.4) | 71.8 (69.9–73.6) | |
| Yes | 32.2 (31.0–33.3) | 39.5 (36.4–42.7) | 33.9 (31.7–36.2) | 30.6 (27.6–33.7) | 28.2 (26.4–30.1) | |
| MUI, % (95% CI) | < .001 | |||||
| No | 63.5 (62.1–64.8) | 55.2 (52.5–58.0) | 60.5 (58.2–62.8) | 64.8 (61.4–68.1) | 68.9 (66.5–71.2) | |
| Yes | 36.5 (35.2–38.0) | 44.8 (42.0–47.5) | 39.5 (37.3–41.8) | 35.2 (31.9–38.6) | 31.1 (28.8–33.5) | |
| UUI, % (95% CI) | < .001 | |||||
| No | 82.2 (81.2–83.2) | 75.6 (72.8–78.3) | 80.2 (78.5–81.9) | 82.9 (80.2–85.4) | 86.4 (84.7–88.0) | |
| Yes | 17.8 (16.8–18.8) | 24.4 (21.7–27.2) | 19.8 (18.1–21.5) | 17.1 (14.7–19.8) | 13.6 (12.0–15.3) |
Q1–Q4: quartile 1 to quartile 4.
BMI = body mass index, CI = confidence interval, GED = general educational development, GNRI = geriatric nutritional risk index, MUI = mixed urinary incontinence, PIR = poverty-to-income ratio, SE = standard error, SUI = stress urinary incontinence, UUI = urgency urinary incontinence.
3.2. The association between GNRI and UI
Multivariable logistic regression was used to assess the relationship between GNRI and UI in crude (model 1), minimally adjusted (model 2), and fully adjusted (model 3) models. The results demonstrated that the GNRI was negatively related to the likelihood of UI. In model 3, there was a decrease in the risk of UI with each successive unit increase in GNRI (SUI: odds ratio [OR] = 0.986, 95% CI: 0.974–0.999; UUI: OR = 0.972, 95% CI: 0.961–0.983; MUI: OR = 0.972, 95% CI: 0.957–0.987, all P < .05). After converting GNRI into a categorical variable according to the GNRI quartiles (Q1–Q4), the negative connection still existed. In model 3, survey individuals belonging to the highest GNRI quartile (Q4) had a lower likelihood of UI compared to participants in the lowest group (Q1) (SUI: OR = 0.797, 95% CI: 0.672–0.945; UUI: OR = 0.722, 95% CI: 0.615–0.848; MUI: OR = 0.674, 95% CI: 0.538–0.845, all P for trend < .05). More information is presented in Table 2. Furthermore, the smooth curve fitting revealed a negative linear association between GNRI and 3 types of UI (Fig. 2).
Table 2.
Association between GNRI and urinary incontinence, weighted.
| SUI | OR (95% CI), P value | ||
|---|---|---|---|
| Model 1 | Model 2 | Model 3 | |
| Continuous* | 0.964 (0.953–0.976), < .001 | 0.981 (0.968–0.993), .004 | 0.986 (0.974–0.999), .031 |
| Categories† | |||
| Q1 | Reference | Reference | Reference |
| Q2 | 0.785 (0.672–0.918), .003 | 0.835 (0.708–0.986), .036 | 0.832 (0.704–0.983), .033 |
| Q3 | 0.674 (0.551–0.825), < .001 | 0.756 (0.609–0.939), .013 | 0.791 (0.637–0.980), .035 |
| Q4 | 0.601 (0.507–0.712), < .001 | 0.754 (0.634–0.897), .002 | 0.797 (0.672–0.945), .010 |
| P for trend | < .001 | .004 | .034 |
| UUI | OR (95% CI), P value | ||
|---|---|---|---|
| Model 1 | Model 2 | Model 3 | |
| Continuous* | 0.953 (0.943–0.964), < .001 | 0.964 (0.954–0.975), < .001 | 0.972 (0.961–0.983), < .001 |
| Categories† | |||
| Q1 | Reference | Reference | Reference |
| Q2 | 0.806 (0.700–0.928), .003 | 0.860 (0.739–1.000), .050 | 0.869 (0.742–1.017), .084 |
| Q3 | 0.670 (0.565–0.794), < .001 | 0.742 (0.617–0.893), .002 | 0.805 (0.674–0.961), .018 |
| Q4 | 0.557 (0.481–0.645), < .001 | 0.654 (0.558–0.767), < .001 | 0.722 (0.615–0.848), < .001 |
| P for trend | < .001 | < .001 | < .001 |
| MUI | OR (95% CI), P value | ||
|---|---|---|---|
| Model 1 | Model 2 | Model 3 | |
| Continuous* | 0.949 (0.936–0.963), < .001 | 0.963 (0.948–0.978), < .001 | 0.972 (0.957–0.987), < .001 |
| Categories† | |||
| Q1 | Reference | Reference | Reference |
| Q2 | 0.765 (0.637–0.919), .005 | 0.822 (0.673–1.000), .057 | 0.840 (0.686–1.030), .093 |
| Q3 | 0.640 (0.515–0.795), < .001 | 0.726 (0.577–0.913), .007 | 0.799 (0.638–1.002), .052 |
| Q4 | 0.487 (0.394–0.602), < .001 | 0.601 (0.482–0.750), < .001 | 0.674 (0.538–0.845), < .001 |
| P for trend | < .001 | < .001 | < .001 |
Model 1: Unadjusted.
Model 2: Adjusted for gender, age, and race/ethnicity.
Model 3: Adjusted for gender, age, race/ethnicity, education level, marital status, the family PIR, BMI, smoking status, alcohol intaking, vigorous activity, moderate activity, diabetes, and hypertension.
Q1–Q4: quartile 1 to quartile 4.
BMI = body mass index, CI = confidence interval, GNRI = geriatric nutritional risk index, MUI = mixed urinary incontinence, OR = odds ratio, PIR = poverty-to-income ratio, SUI = stress urinary incontinence, UUI = urgency urinary incontinence.
“Continuous” means that GNRI is designed as a continuous variable.
“Categories” represents that GNRI is set as a categorical variable and divided into 4 groups (quartile 1-quartile 4).
Figure 2.
Smooth curve fitting for the relationship between GNRI and 3 types of UI. (A–C) represent the linear connections between GNRI and SUI, UUI, and MUI, respectively. The area between 2 blue dotted lines represents a 95% CI. The red dotted line suggests the negative linear relationship between GNRI and 3 types of UI. CI = confidence interval, GNRI = geriatric nutritional risk index, MUI = mixed urinary incontinence, SUI = stress urinary incontinence, UI = urinary incontinence, UUI = urgency urinary incontinence.
3.3. Subgroup analysis
Subgroup analysis stratified by gender, age, BMI, hypertension, and diabetes was conducted to explore the stability of the negative relationship between GNRI and UI. In Figure 3A, the negative association between GNRI and SUI was stronger in males compared to females. Additionally, a difference was observed in the relationship between GNRI and MUI as shown in Figure 3C (P for interaction < .05). In UUI, there were no significant differences in subgroups (Fig. 3B). Therefore, gender may be a potential factor influencing the connection. Moreover, tests for interaction were not significant in other stratifications. More detailed data are shown in Figure 3.
Figure 3.
Subgroup analysis for the association between GNRI and 3 types of UI. (A–C) present the subgroup analysis results of the associations between the GNRI index and SUI, UUI, and MUI, respectively, across different stratifications. All stratified factors include gender, age, race/ethnicity, education level, marital status, the family PIR, smoking status, alcohol intake, vigorous activity, moderate activity, diabetes, and hypertension, except the stratified factor itself. BMI = body mass index, CI = confidence interval, GNRI = geriatric nutritional risk index, MUI = mixed urinary incontinence, OR = odds ratio, PIR = poverty-to-income ratio, SUI = stress urinary incontinence, UI = urinary incontinence, UUI = urgency urinary incontinence.
4. Discussion
To our knowledge, this is the first study to examine the link between nutritional status and the occurrence of UI in older adults, based on a large-scale demographic survey of the NHANES database in the U.S. The results of this study show that an increase in GNRI in older adults is associated with a reduced risk of UI after adjusting for covariates, regardless of the type of UI. The smooth curve fitting shows a negative linear relationship between GNRI and UI. Moreover, the subgroup analysis demonstrates a stronger association between GNRI and both SUI and MUI in males than in females. This gender difference warrants deeper reflection.
UI causes distress in the lives of many older people, especially older women. There is no consensus on the risk factors for UI. Current evidence supports that older age, obesity, and post-hysterectomy are important risk factors for UI.[18–22] Some chronic diseases may also be contributing factors to incontinence, such as stroke, diabetes, and hypertension.[5,23] However, the evidence on the relationship between nutritional status and the prevalence of UI among older adults remains insufficient.
The GNRI is a composite indicator of nutritional status that integrates serum albumin levels and BMI, thereby minimizing the confounding effect between them. Several studies have previously explored the relationship between serum albumin and UI. Xu et al reported that elevated serum albumin levels are beneficial in reducing the risk of SUI in women based on a large population-based study.[9] For each 1 g/dL increase in serum albumin, the risk of moderate/severe SUI was reduced by 23.7%. In addition, in a study on overactive bladder syndrome, researchers observed that serum albumin levels tended to be lower in patients with nocturia and those with UUI.[24] These findings suggest that serum albumin levels may reflect the prevalence of UI to some extent, and UI may be linked to nutritional status. The evidence for BMI and obesity as risk factors for UI appears to be stronger. Several studies have demonstrated that BMI can be an independent predictor of UI, especially for female SUI.[21,25–28] Clinical trials have confirmed the therapeutic effect of bariatric surgery for UI.[26,29] A systematic review and meta-analysis suggest that weight loss may be considered a standard treatment for UI.[30] Moderate BMI levels are also an important indicator for assessing nutritional status in older adults. In this study, the GNRI combined patients’ serum albumin levels and BMI to comprehensively assess their nutritional status. Therefore, GNRI may be a predictor of UI development. While the current study primarily focuses on the influence of poor nutritional status on the prevalence and severity of UI among older adults, it is important to acknowledge the potential for a bidirectional relationship. A cross-sectional study in Taiwan revealed that UI was reported by approximately one-fifth of the participants and was significantly connected to more comorbidity, poorer physical function, and worse nutritional status.[31] Rose et al showed that in nursing homes, greater severity of incontinence was strongly associated with poorer nutritional status, increased dementia-related symptoms, and impaired mobility.[32] The potential impact of UI on nutritional status merits further investigation. Several possible mechanisms could underline this reverse relationship. Older people experiencing UI may restrict their fluid and food intake for managing or avoiding incontinent episodes, which can result in dehydration and insufficient nutrient intake. Moreover, UI is often associated with psychosocial consequences such as embarrassment, social withdrawal, and depressive symptoms, all of which can negatively impact appetite and eating behavior. The potential bidirectional relationship between GNRI and UI may affect the interpretation of our findings. While poor nutritional status may increase the risk of UI, UI itself could contribute to nutritional decline through factors such as reduced appetite or intentional fluid and food restriction. Recognizing this complexity highlights the need to pay attention to both nutrition and urinary health in older adults.
It is unclear why nutritional status in older adults affects the development of UI, and there may be several underlying mechanisms. Pelvic floor muscle hypofunction is a possible pathologic mechanism for the development of UI.[33,34] Several studies have shown a correlation between nutritional status and muscle strength.[35–37] Better nutritional status in older adults may affect the strength and function of the pelvic floor muscles and may thereby reduce the risk of UI. Estrogen levels are also an important factor in the progression of UI.[38,39] Nutritional status has also been associated with hormone levels.[40,41] The effect of nutritional status on hormones may also be one of the potential mechanisms for their association with the development of UI. In this study, the subgroup analysis demonstrates that a stronger association between GNRI and UI is observed in male participants compared to females. This sex-specific difference warrants further discussion. There are several possible explanations for this kind of difference. First, malnutrition was related to an increased risk of sarcopenia.[42] Men normally have greater baseline muscle mass and larger pelvic floor muscles. Malnutrition-induced sarcopenia may lead to a more obvious loss of urethral support and detrusor contractility in men. Indeed, Langston et al showed that higher limb muscle strength was associated with milder lower urinary tract symptoms in older men, suggesting that sarcopenia exacerbated male UI.[43] Moreover, a large population survey found that female UI is strongly associated with obesity, physical inactivity, and smoking, whereas male UI is driven by age, frailty, and comorbid diseases.[44] Therefore, a poorer nutritional status would have a larger effect on UI risk in males.
Our study utilized a substantial dataset derived from the NHANES database and accounted for the sampling design and weighting to accurately describe the general population of the U.S. Nevertheless, several limitations are inherent in the present investigation. First, the cross-sectional design of this study limits the exploration of causal association between GNRI and UI but also fails to rule out the potential impact of reverse causation. Specifically, it remains unclear whether poor nutritional status (reflected by low GNRI scores) contributes to the development or progression of UI, or if UI itself leads to reduced dietary intake, weight loss, and subsequent declines in nutritional status; this bidirectional relationship cannot be disentangled with the current cross-sectional data. In addition, the NHANES database reflects only the U.S. population, limiting the generalizability of the findings to other countries. Therefore, the association between GNRI and UI warrants validation through further studies across diverse national populations. Finally, while we adjusted for several potential covariates in the present study, residual confounding remains an unavoidable limitation. It is difficult to eliminate the impact of other conceivable confounding factors. In addition to commonly controlled sociodemographic and clinical factors, several other covariates may influence the relationship between GNRI and UI, such as pelvic floor muscle strength, caregiver support, and hormonal levels. However, due to the limitation of NHANES database, we could not include these covariates in the current study. It is recommended that future studies include these factors to enhance the robustness and generalizability of the present results.
5. Conclusions
To the best of our knowledge, this represents the first cross-sectional investigation delving into the association between GNRI and UI within the old adults (> 60 years old) of the U.S. Our study indicated a negative relationship between GNRI and all 3 types of UI (SUI, UUI, and MUI). Notably, even upon stratification of GNRI into quartiles, the negative connection remained stable. The relation between GNRI and SUI, MUI was stronger in men than in women. Nevertheless, further research is necessary to corroborate and extend our findings.
Acknowledgments
All authors express their gratitude to all participants and personnel involved in the NHANES.
Author contributions
Conceptualization: Juan Xu, Xiang Li, Shu Wen.
Data curation: Juan Xu, Shangqi Cao, Xu Hu.
Formal analysis: Juan Xu, Shangqi Cao.
Funding acquisition: Shu Wen.
Methodology: Juan Xu, Shangqi Cao, Xu Hu.
Resources: Shangqi Cao.
Software: Juan Xu, Xu Hu.
Supervision: Xiang Li, Shu Wen.
Writing – original draft: Juan Xu, Shangqi Cao, Xu Hu.
Writing – review & editing: Juan Xu, Xiang Li, Shu Wen.
Abbreviations:
- BMI
- body mass index
- CI
- confidence interval
- GNRI
- geriatric nutritional risk index
- MUI
- mixed urinary incontinence/mixed UI
- NHANES
- National Health and Nutrition Examination Survey
- OR
- odds ratio
- PIR
- poverty-to-income ratio
- SUI
- stress urinary incontinence/stress UI
- UI
- urinary incontinence
- U.S.
- United States
- UUI
- urgency urinary incontinence/urgency UI
This study was supported by the Natural Science Foundation of Sichuan, China (No.2026NSFSC1541 to S.W.). The funder had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or decision to submit the manuscript for publication.
Consent for publication is not applicable.
The studies involving human participants were reviewed and approved by the Research Ethics Review Board of the National Center for Health Statistics. All methods and procedures were conducted in accordance with the relevant guidelines and regulations (Declaration of Helsinki). The patients/participants provided their written informed consent to participate in this study.
The authors have no conflicts of interest to declare.
The datasets generated during and/or analyzed during the current study are publicly available.
How to cite this article: Xu J, Cao S, Hu X, Li X, Wen S. Generic nutritional risk index and urinary incontinence in U.S. older adults: A cross-sectional analysis of NHANES 2001–2018. Medicine 2026;105:36(e50584).
Contributor Information
Juan Xu, Email: 1456227293@qq.com.
Shangqi Cao, Email: 1343417195@qq.com.
Xu Hu, Email: 979568175@qq.com.
Xiang Li, Email: hx_uro@sina.com.
References
- [1].Leslie SW, Tran LN, Puckett Y. Urinary incontinence. In: StatPearls. Treasure Island (FL) ineligible companies. Disclosure: Linh Tran declares no relevant financial relationships with ineligible companies. Disclosure: Yana Puckett declares no relevant financial relationships with ineligible companies.: StatPearls Publishing Copyright © 2026, StatPearls Publishing LLC; 2026. [Google Scholar]
- [2].Gacci M, Sakalis VI, Karavitakis M, et al. European association of urology guidelines on male urinary incontinence. Eur Urol. 2022;82:387–98. [DOI] [PubMed] [Google Scholar]
- [3].Moris L, Heesakkers J, Nitti V, et al. Prevalence, diagnosis, and management of stress urinary incontinence in women: a collaborative review. Eur Urol. 2025;87:292–301. [DOI] [PubMed] [Google Scholar]
- [4].Roberts K, Mahajan S. Management of urgency and mixed urinary incontinence. Clin Obstet Gynecol. 2021;64:306–13. [DOI] [PubMed] [Google Scholar]
- [5].Vaughan CP, Markland AD. Urinary incontinence in women. Ann Intern Med. 2020;172:ITC17–32. [DOI] [PubMed] [Google Scholar]
- [6].Yuan Y, Huang Y, Huang H, Zeng S, Hu Z. Association between added sugar intake and urinary incontinence in females: a cross-sectional population-based study. Int Urogynecol J. 2024;35:1201–10. [DOI] [PubMed] [Google Scholar]
- [7].Lee JH, Lee HS. Nutrient intake and urinary incontinence in Korean women: a propensity score-matched analysis from the Korea National Health and Nutrition Examination Survey data. Int J Urol. 2017;24:793–7. [DOI] [PubMed] [Google Scholar]
- [8].Choi WS, Heo NJ, Son H. The influence of dietary habits on male lower urinary tract symptoms: differential influence of water intake on voiding symptoms and storage symptoms. World J Mens Health. 2025;43:647–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Xu M, Zhou H, Pan Y, Xu Z, Liu X. Serum albumin levels and stress urinary incontinence in females: a retrospective study based on NHANES 2007-2016. Heliyon. 2023;9:e21757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Bouillanne O, Morineau G, Dupont C, et al. Geriatric Nutritional Risk Index: a new index for evaluating at-risk elderly medical patients. Am J Clin Nutr. 2005;82:777–83. [DOI] [PubMed] [Google Scholar]
- [11].Lidoriki I, Schizas D, Frountzas M, et al. GNRI as a prognostic factor for outcomes in cancer patients: a systematic review of the literature. Nutr Cancer. 2021;73:391–403. [DOI] [PubMed] [Google Scholar]
- [12].Zuo J, Huang Z, Ge Y, Ding X, Wang X, Zhou X. Geriatric Nutrition Risk Index is closely associated with sarcopenia and quality of life in gastric cancer patients: a cross-sectional study. Sci Rep. 2024;14:31545. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Cheng K, Zhang J, Ye LY, et al. Geriatric nutrition risk index in the prediction of all-cause and cardiovascular mortality in older adults with hyperlipidemia: NHANES 1999-2018. BMC Geriatr. 2024;24:634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Jiang Z, Cai X, Lan W, et al. Predictive role of the geriatric nutritional risk index in all-cause and cardiovascular mortality among elderly patients with osteoarthritis. BMC Geriatr. 2025;25:592. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Dai Z, Wu Y, Huang H, et al. Association between preoperative geriatric nutritional risk index and major adverse cardiovascular and cerebrovascular events after non-cardiac surgery in older patients: a retrospective cohort study. BMC Geriatr. 2025;25:741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Wei J, Zeng Q, Liu M. Joint association of physical activity and the geriatric nutritional risk index with survival outcomes among cancer survivors in the United States: a population-based cohort study. Front Nutr. 2024;11:1483507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Grinstead C, Yoon SL. Geriatric Nutritional Risk Index (GNRI) and survival in pancreatic cancer: a retrospective study. Nutrients. 2025;17:509. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Al-Mukhtar Othman J, Åkervall S, Milsom I, Gyhagen M. Urinary incontinence in nulliparous women aged 25-64 years: a national survey. Am J Obstet Gynecol. 2017;216:149.e1–149.e11. [DOI] [PubMed] [Google Scholar]
- [19].Harlow BL, Bavendam TG, Palmer MH, et al. The Prevention of Lower Urinary Tract Symptoms (PLUS) Research Consortium: a transdisciplinary approach toward promoting bladder health and preventing lower urinary tract symptoms in women across the life course. J Womens Health (Larchmt). 2018;27:283–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [20].Kuittinen T, Tulokas S, Rahkola-Soisalo P, et al. Pelvic organ prolapse after hysterectomy: a 10-year national follow-up study. Acta Obstet Gynecol Scand. 2023;102:556–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Townsend MK, Danforth KN, Rosner B, Curhan GC, Resnick NM, Grodstein F. Body mass index, weight gain, and incident urinary incontinence in middle-aged women. Obstet Gynecol. 2007;110(2 Pt 1):346–53. [DOI] [PubMed] [Google Scholar]
- [22].Tulokas S, Mentula M, Härkki P, et al. Stress urinary incontinence after hysterectomy: a 10-year national follow-up study. Arch Gynecol Obstet. 2022;305:1089–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [23].Markland AD, Vaughan CP, Okosun IS, Goode PS, Burgio KL, Johnson TM, 2nd. Cluster analysis of multiple chronic conditions associated with urinary incontinence among women in the USA. BJU Int. 2018;122:1041–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Chuang PH, Chang YH, Hsiao PJ, Chou EC. Diagnostic potential of low serum platelet, albumin and prolong PT-INR for overactive bladder and nocturia in chronic hepatitis-related liver cirrhosis. J Clin Med. 2021;10:2838. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [25].Buchsbaum GM, Chin M, Glantz C, Guzick D. Prevalence of urinary incontinence and associated risk factors in a cohort of nuns. Obstet Gynecol. 2002;100:226–9. [DOI] [PubMed] [Google Scholar]
- [26].Subak LL, Wing R, West DS, et al. Weight loss to treat urinary incontinence in overweight and obese women. N Engl J Med. 2009;360:481–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Townsend MK, Curhan GC, Resnick NM, Grodstein F. BMI, waist circumference, and incident urinary incontinence in older women. Obesity (Silver Spring). 2008;16:881–6. [DOI] [PubMed] [Google Scholar]
- [28].Pang H, Xu T, Li Z, et al. Remission and transition of female urinary incontinence and its subtypes and the impact of body mass index on this progression: a nationwide population-based 4-year longitudinal s1tudy in China. J Urol. 2022;208:360–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Lasserre A, Pelat C, Guéroult V, et al. Urinary incontinence in French women: prevalence, risk factors, and impact on quality of life. Eur Urol. 2009;56:177–83. [DOI] [PubMed] [Google Scholar]
- [30].Vissers D, Neels H, Vermandel A, et al. The effect of non-surgical weight loss interventions on urinary incontinence in overweight women: a systematic review and meta-analysis. Obes Rev. 2014;15:610–7. [DOI] [PubMed] [Google Scholar]
- [31].Wang CJ, Hung CH, Tang TC, et al. Urinary incontinence and its association with frailty among men aged 80 years or older in Taiwan: a cross-sectional study. Rejuvenation Res. 2017;20:111–7. [DOI] [PubMed] [Google Scholar]
- [32].Rose A, Thimme A, Halfar C, Nehen HG, Rübben H. Severity of urinary incontinence of nursing home residents correlates with malnutrition, dementia and loss of mobility. Urol Int. 2013;91:165–9. [DOI] [PubMed] [Google Scholar]
- [33].Aoki Y, Brown HW, Brubaker L, Cornu JN, Daly JO, Cartwright R. Urinary incontinence in women. Nat Rev Dis Primers. 2017;3:17042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [34].Brennen R, Frawley HC, Martin J, Haines TP. Group-based pelvic floor muscle training for all women during pregnancy is more cost-effective than postnatal training for women with urinary incontinence: cost-effectiveness analysis of a systematic review. J Physiother. 2021;67:105–14. [DOI] [PubMed] [Google Scholar]
- [35].Schalk BW, Deeg DJ, Penninx BW, Bouter LM, Visser M. Serum albumin and muscle strength: a longitudinal study in older men and women. J Am Geriatr Soc. 2005;53:1331–8. [DOI] [PubMed] [Google Scholar]
- [36].Snyder CK, Lapidus JA, Cawthon PM, Dam TT, Sakai LY, Marshall LM. Serum albumin in relation to change in muscle mass, muscle strength, and muscle power in older men. J Am Geriatr Soc. 2012;60:1663–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [37].Latronico N, Nisoli E, Eikermann M. Muscle weakness and nutrition in critical illness: matching nutrient supply and use. Lancet Respir Med. 2013;1:589–90. [DOI] [PubMed] [Google Scholar]
- [38].Hendrix SL, Cochrane BB, Nygaard IE, et al. Effects of estrogen with and without progestin on urinary incontinence. JAMA. 2005;293:935–48. [DOI] [PubMed] [Google Scholar]
- [39].Shamliyan TA, Kane RL, Wyman J, Wilt TJ. Systematic review: randomized, controlled trials of nonsurgical treatments for urinary incontinence in women. Ann Intern Med. 2008;148:459–73. [DOI] [PubMed] [Google Scholar]
- [40].Mauvais-Jarvis F, Clegg DJ, Hevener AL. The role of estrogens in control of energy balance and glucose homeostasis. Endocr Rev. 2013;34:309–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [41].Zhang H, Wang L, Li Y, Wang P, Wang C. Background nutrients and bacterial community evolution determine (13)C-17β-estradiol mineralization in lake sediment microcosms. Sci Total Environ. 2019;651(Pt 2):2304–11. [DOI] [PubMed] [Google Scholar]
- [42].Yuan S, Larsson SC. Epidemiology of sarcopenia: prevalence, risk factors, and consequences. Metabolism. 2023;144:155533. [DOI] [PubMed] [Google Scholar]
- [43].Langston ME, Cawthon PM, Lu K, et al. Associations of lower extremity muscle strength, area, and specific force with lower urinary tract symptoms in older men: the Baltimore longitudinal study of aging. J Gerontol A Biol Sci Med Sci. 2024;79:glae008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].Yavuz M, Etiler N. Addressing urinary incontinence by gender: a nationwide population-based study in Turkiye. BMC Urol. 2023;23:205. [DOI] [PMC free article] [PubMed] [Google Scholar]