Abstract
Objective
To investigate the relationship between android-to-gynoid fat ratio (A/G ratio) and stress urinary incontinence (SUI) in US adult women.
Methods
The study employed data from the National Health and Nutrition Examination Survey (NHANES) database, encompassing the 2005–2006 and 2011–2018 cycles. We utilized multivariable-adjusted logistic regression models to examine the independent effect of the A/G ratio on SUI. Additionally, refined subgroup analyses were conducted to assess the impact of effect modifiers, along with sophisticated smooth curve fitting methods to delve into the potential non-linear relationship between the A/G ratio and SUI.
Results
Of the 6779 women participants with an average age of 40.812, 2565 (37.8%) self-reported SUI. A non-linear relationship between the A/G ratio and SUI has been detected, with the threshold (A/G ratio = 0.446). Above and below this threshold, the effect sizes and the 95% confidence intervals were 1.464 (0.775–2.766) and 10.308 (3.558–29.863), respectively. Subgroup analyses showed that the positive association was more pronounced among young women who were premenopausal, living alone, hypertensive and with normal or low body mass index (BMI).
Conclusion
There was a non-linear relationship between the A/G ratio and SUI. In individuals with a lower A/G ratio, even small increases in the A/G ratio substantially elevated the risk of SUI. Results of this study indicate that A/G ratio, as a measure of fat distribution, is of great importance for clinicians in assessing SUI risk and timely intervening.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12905-025-03998-4.
Keywords: Stress urinary incontinence, Android-to-gynoid fat ratio, Fat distribution, Adult women, NHANES, Cross-sectional study
Introduction
Urinary incontinence (UI) is a widespread dysfunction of the lower urinary tract, defined by any involuntary leakage of urine. Stress urinary incontinence (SUI), as the most prevalent form of UI, occurs when physical activity or coughing raises intra-abdominal pressure [1]. The patient’s quality of life is somewhat diminished by SUI since it may lead to discomfort, embarrassment, and social isolation [2]. Epidemiological data indicate that SUI affects approximately 26.8% of women in North America, with a notably higher prevalence among non-Hispanic white women [3].
Adipose tissue buildup in the abdomen has been linked to increased intra-abdominal pressure, which puts undue strain on the pelvic floor muscles and connective tissues, according to research. The integrity of the bladder and urethra is compromised by the straining and weakening of these supporting structures brought on by this ongoing pressure. Because of this, the pelvic floor becomes less able to sustain urinary continence, which increases the risk of SUI. Traditionally, body mass index (BMI) has served as the fundamental metric for evaluating obesity in clinical and scientific contexts. Nevertheless, BMI is a rudimentary measure that does not take into consideration the distribution of body fat and fails to differentiate between lean muscle mass and adipose tissue [4]. Given that abdominal fat buildup exerts a greater influence on intra-abdominal pressure, this constraint is especially noteworthy, emphasizing the necessity for more precise metrics to evaluate obesity and fat distribution.
Dual-energy X-ray absorptiometry (DXA) is used for precise quantification of body composition and fat distribution [5]. Studies have shown that the accumulation of android fat, mainly distributed in the lower part of the trunk, is linked to adverse metabolic and cardiovascular profiles. Conversely, the accumulation of gynoid fat, mainly in the gluteofemoral region, exerts a protective effect with respect to mortality in individuals with cardiovascular disease. An innovative form of fat distribution measure that illustrates the disparity between the distribution of fat in the gynoid and android regions is the android-to-gynoid fat ratio (A/G ratio). Research has shown that diseases including metabolic syndrome (METS) [6]and insulin resistance (IR) [7]are intimately linked to elevated A/G ratio. Nevertheless, the potential association between the A/G ratio and SUI remains insufficiently investigated in the current literature.
The purpose of this cross-sectional study is to examine their relationship through the analysis of National Health and Nutrition Examination Survey (NHANES) data extracted from 2005 to 2006 and 2011–2018 cycles. Illuminating this association may enhance our understanding of how aberrant fat distribution and metabolic status relate to SUI risk, informing future preventive strategies.
Methods
Study population
This study employed data from the 2005–2006 and 2011–2018 NHANES cycles. NHANES is an ongoing series of nationally representative cross-sectional surveys conducted by the National Center for Health Statistics (NCHS) under the Centers for Disease Control and Prevention (CDC) (http://www.cdc.gov/nchs/nhanes.htm). NHANES adheres to the tenets set forth in the Declaration of Helsinki, and its research protocol has been approved by the NCHS Institutional Review Board. All participants were duly informed and provided their consent. It should be noted that the analyses do not include any individuals with incomplete data on SUI, A/G ratio, BMI and system error (shown in Fig. 1) This study does not necessitate further ethical approval, as the NHANES protocol received prior authorization from the NCHS Institutional Review Board.
Fig. 1.
Study population flowchart
Evaluation of SUI
Respondents were classified as having SUI if they indicated affirmative responses to the subsequent query. The following question was posed to the respondents: “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 exercising?”.
Evaluation of fat distribution
An individual’s body weight in kilograms is divided by the square of height in meters to determine BMI. The android and gynoid regions were precisely defined through DXA scans performed by certified medical imaging technologists utilizing Hologic Discovery model A densitometers, followed by analysis using Hologic APEX software. Subsequently, the A/G ratio was calculated on the base of the above data [8].
Covariates
In order to mitigate the potential bias caused by confounding factors, this study employed a series of covariates for adjustment. The inclusion of covariates was based on conventional adjustments, clinical relevance, similar literature, and inclusion criteria derived from statistical analysis. Demographic variables included age, race/ethnicity (Mexican American, other Hispanic, non-Hispanic White, non-Hispanic Black, other races), educational level (less than high school, high school diploma, more than high school), marital status (living alone or divorced, living with a partner), and the family poverty income ratio (PIR). BMI is categorized as <25, 25–30, or > 30. Regarding reproductive health, menstruation status (yes or no), gravidity categorized as ≤ 3 or > 3, parity represented the number of births a woman has had, including those that were vaginal births as well as those that were cesarean sections, categorized as ≤ 1 or > 1. In biochemical parameters, Non-high-density lipoprotein cholesterol (Non-HDL-C) was calculated by subtracting serum high-density lipoprotein cholesterol from total serum cholesterol. Physical activity was calculated as the total time spent in a series of leisure activities over 7 days, such as walking, cycling, and various recreational activities. In the Physical Activity Questionnaire (PAQ), participants reported the frequency, duration, and intensity (moderate or vigorous) of their weekly exercise. The intensity of physical activity was classified based on the metabolic equivalent (MET) levels: low (< 500 MET-min/week), moderate (500-1,000 MET-min/week), and high (> 1,000 MET-min/week). Participants were categorized according to their history of tobacco consumption, with those who had smoked at least 100 cigarettes during their lifetime designated as smokers. Those who had smoked fewer than 100 cigarettes in their lifetime were classified as non-smokers. According to the Alcohol Use Questionnaire, participants were classified as drinkers or non-drinkers in consideration of their responses to the following inquiry: “Have you had at least 12 drinks in the past year?“. Participants were considered diabetic if a doctor had previously diagnosed them with diabetes, if they were using anti-diabetic medications or insulin, or if they had a glycated hemoglobin level ≥ 6.5%. Participants were considered hypertensive if they had received a diagnosis of the condition from a medical professional before the interview or if they had systolic blood pressure ≥ 140 mmHg, diastolic blood pressure ≥ 90 mmHg, or were taking antihypertensive medication. Participants were considered to have cardiovascular disease (CVD) if they had previously experienced heart failure, angina, myocardial infarction, or coronary heart disease.
Statistical analysis
Considering the complex sampling design features and weighting recommendations of NHANES Strata, clustering, and sample weights were all incorporated into the statistical analysis procedure. Covariate selection was guided by variance inflation factor (VIF) analysis (excluding covariates with VIF ≥ 5), results from univariate analysis, and changes in the regression coefficient of the A/G ratio (Supplementary Tables 1–3). Logistic regression was employed to estimate percentages for categorical variables. For continuous variables, linear regression was utilized to extrapolate means and standard errors (SE). In addition, three models were developed to elucidate the association between the A/G ratio and SUI: model 1 was unadjusted; model 2 adjusted for age, race, education, marital status, and PIR; and model 3 further adjusted for smoking status, alcohol consumption, diabetes, hypertension, CVD, pregnancy history, parity, menopausal status, physical activity, non-HDL cholesterol, and BMI. The A/G ratio was divided into tertiles as ordinal categorical variables to examine potential trends. The random forest imputation method was applied to handle missing values in covariates. Furthermore, we used a two-piecewise linear regression model and smooth curve fitting techniques to assess the non-linear relationship. These methods enabled the detection of potential threshold and the characterization of the shape of the association. Additionally, subgroup analyses were performed to investigate potential heterogeneity in the correlation between the A/G ratio and SUI. All statistical analyses were conducted using R (version 4.2.3; The R Foundation) and Empower (X&Y Solutions Inc) software. Statistical significance was considered by a two-sided p-value of < 0.05.
Results
Basic characteristics
The average age of the 6779 female participants in this study was 40.812 years. Among these, 2,565 women (37.8%) were diagnosed with SUI. In comparison with those who did not present with SUI, women with SUI exhibited a greater prevalence of advanced age and a higher probability of being non-Hispanic White. They exhibited lower education level, were more likely to be married, and engaged less frequently in physical activity. A higher proportion of women with SUI were postmenopausal and had a greater number of pregnancies. Additionally, they exhibited increased rates of smoking, alcohol consumption, diabetes, hypertension, and CVD. In terms of anthropometric and biochemical indicators, women with SUI had higher BMI, elevated A/G ratio, and increased levels of non-HDL-C compared to the non-SUI group (Table 1).
Table 1.
Weighted descriptive statistics of participants stratified by SUI
| Characteristics | Total (n = 6779) |
Non-SUI (n = 4214) |
SUI (n = 2565) |
P-value |
|---|---|---|---|---|
| Age (year), mean (SE) | 40.812 (0.265) | 38.205 (0.301) | 44.719 (0.358) | < 0.001 |
| Age (%) | ||||
| ≤ 50 | 5009 (73.890%) | 3320 (78.785%) | 1689 (65.848%) | < 0.001 |
| >50 | 1770 (26.110%) | 894 (21.215%) | 876 (34.152%) | |
| Race (%) | < 0.001 | |||
| Mexican American | 1053 (15.533%) | 598 (14.191%) | 455 (17.739%) | |
| Other Hispanic | 633 (9.338%) | 395 (9.374%) | 238 (9.279%) | |
| Non-Hispanic White | 2530 (37.321%) | 1427 (33.863%) | 1103 (43.002%) | |
| Non-Hispanic Black | 1584 (23.366%) | 1132 (26.863%) | 452 (17.622%) | |
| Other race | 979 (14.442%) | 662 (15.710%) | 317 (12.359%) | |
| Education level (%) | < 0.001 | |||
| Less than high school | 1156 (17.053%) | 659 (15.638%) | 497 (19.376%) | |
| High school | 1369 (20.195%) | 825 (19.578%) | 544 (21.209%) | |
| Above high school | 4254 (62.753%) | 2730 (64.784%) | 1524 (59.415%) | |
| Marital status (%) | < 0.001 | |||
| Never married or separated | 2877 (42.440%) | 1943 (46.108%) | 934 (36.413%) | |
| Married or living with partner | 3902 (57.560%) | 2271 (53.892%) | 1631 (63.587%) | |
| PIR (%) | 0.705 | |||
| ≤ 1.3 | 2012 (29.680%) | 1238 (29.378%) | 774 (30.175%) | |
| > 1.3, ≤ 3.5 | 2714 (40.035%) | 1702 (40.389%) | 1012 (39.454%) | |
| > 3.5 | 2053 (30.285%) | 1274 (30.233%) | 779 (30.370%) | |
| Smoking status (%) | < 0.001 | |||
| No | 4481 (66.101%) | 2956 (70.147%) | 1525 (59.454%) | |
| Yes | 2298 (33.899%) | 1258 (29.853%) | 1040 (40.546%) | |
| Alcohol drinking status (%) | 0.022 | |||
| No | 2126 (31.362%) | 1364 (32.368%) | 762 (29.708%) | |
| Yes | 4653 (68.638%) | 2850 (67.632%) | 1803 (70.292%) | |
| Menstruation (%) | < 0.001 | |||
| No | 2302 (33.958%) | 1226 (29.093%) | 1076 (41.949%) | |
| Yes | 4477 (66.042%) | 2988 (70.907%) | 1489 (58.051%) | |
| Gravidity (%) | < 0.001 | |||
| ≤ 3 | 4826 (71.190%) | 3174 (75.320%) | 1652 (64.405%) | |
| > 3 | 1953 (28.810%) | 1040 (24.680%) | 913 (35.595%) | |
| Parity (%) | < 0.001 | |||
| ≤ 1 | 2141 (31.583%) | 1566 (37.162%) | 575 (22.417%) | |
| > 1 | 4638 (68.417%) | 2648 (62.838%) | 1990 (77.583%) | |
| Physical activity (%) | < 0.001 | |||
| Low | 2318 (34.194%) | 1369 (32.487%) | 949 (36.998%) | |
| Moderate | 2706 (39.917%) | 1643 (38.989%) | 1063 (41.442%) | |
| High | 1755 (25.889%) | 1202 (28.524%) | 553 (21.559%) | |
| Hypertension (%) | < 0.001 | |||
| No | 4839 (71.382%) | 3197 (75.866%) | 1642 (64.016%) | |
| Yes | 1940 (28.618%) | 1017 (24.134%) | 923 (35.984%) | |
| Diabetes (%) | < 0.001 | |||
| No | 6133 (90.471%) | 3904 (92.644%) | 2229 (86.901%) | |
| Yes | 646 (9.529%) | 310 (7.356%) | 336 (13.099%) | |
| Cardiovascular disease (%) | < 0.001 | |||
| No | 6471 (95.457%) | 4068 (96.535%) | 2403 (93.684%) | |
| Yes | 308 (4.543%) | 146 (3.465%) | 162 (6.316%) | |
| BMI (kg/m²), mean (SE) | 28.992 (0.158) | 28.070 (0.173) | 30.374 (0.205) | < 0.001 |
| BMI level (%) | ||||
| <25.0 | 2233 (32.940%) | 1599 (37.945%) | 634 (24.717%) | < 0.001 |
| 25.0–30.0 | 1792 (26.435%) | 1106 (26.246%) | 686 (26.745%) | |
| >30.0 | 2754 (40.625%) | 1509 (35.809%) | 1245 (48.538%) | |
| A/G ratio, mean (SE) | 0.444 (0.003) | 0.423 (0.003) | 0.477 (0.004) | < 0.001 |
| A/G ratio, (%) | ||||
| Tertile 1 (0.15–0.38) | 2260 (33.338%) | 1660 (39.393%) | 600 (23.392%) | < 0.001 |
| Tertile 2 (0.39–0.51) | 2259 (33.323%) | 1382 (32.795%) | 877 (34.191%) | |
| Tertile 3 (0.52–0.99) | 2260 (33.338%) | 1172 (27.812%) | 1088 (42.417%) | |
| Non-HDL-C level (mg/dL), mean (SE) | 135.780 (0.739) | 131.655 (0.861) | 141.961 (1.152) | < 0.001 |
Logistic regression was used for categorical variables to estimate percentages, while linear regression calculated means and standard errors (SE) for continuous variables
SE standard error, A/G ratio android-to-gynoid fat ratio, BMI body mass index, SUI stress urinary incontinence, PIR poverty income ratio, Non-HDL-C non-high-density lipoprotein cholesterol
Correlations between A/G ratio and covariates
To investigate the correlation between the A/G ratio and covariates, spearman rank correlation analysis was performed (Supplementary Table 4). The A/G ratio exhibited pronounced positive correlations with BMI (r = 0.6189). Non-HDL-C level (r = 0.3551), diabetes (r = 0.2996), hypertension (r = 0.2896), and age (r = 0.2047) were weakly positively associated with the A/G ratio. Menstruation (r = −0.2146) was weakly negatively associated with the A/G ratio. Parity (r = 0.1829), gravidity (r = 0.1513), CVD (r = 0.1004), smoking (r = 0.0716), marital status (r = 0.0406), race (r = −0.1438), PIR (r = −0.1374), alcohol drinking status (r = −0.1097) and physical activity (r = −0.1017) were not associated with the A/G ratio.
Association analysis between A/G ratio and SUI
In order to comprehensively evaluate the relationship between the A/G ratio and SUI, we established three models. Across all models, an increased A/G ratio was significantly positively correlated with SUI, indicating a clear and consistent association irrespective of adjustment levels. In the categorical analysis, the A/G ratio was divided into tertiles: tertile 1 (0.15–0.38) as the reference group, and the risk of SUI progressively increased with higher tertiles. Unadjusted (SUI: OR = 2.568; 95% CI = 2.267–2.909); moderately adjusted (SUI: OR = 2.274; 95% CI = 1.992–2.596); and fully adjusted (SUI: OR = 1.416; 95% CI = 1.194–1.679) (Table 2). The results of multivariable analysis stratified by BMI indicated that the A/G ratio was positively associated with SUI at varying BMI levels, and this association was more pronounced among individuals with normal or low BMI (Supplementary Table 5).
Table 2.
Multivariable adjusted models of A/G ratio associations with SUI
| OR (95%CI), P value | OR (95%CI), P value | OR (95%CI), P value | |
|---|---|---|---|
| Model 1 | Model 2 | Model 3 | |
| A/G ratio | 14.562 (10.314,20.560) < 0.0001 | 10.212 (7.083, 14.724) < 0.0001 | 2.676 (1.661, 4.311) < 0.0001 |
| Categories | |||
| Tertile 1 (0.15–0.38) | Reference | Reference | Reference |
| Tertile 2 (0.39–0.51) | 1.756 (1.548, 1.991) < 0.0001 | 1.677 (1.473, 1.910) < 0.0001 | 1.272 (1.097, 1.476) 0.0015 |
| Tertile 3 (0.52–0.99) | 2.568 (2.267, 2.909) < 0.0001 | 2.274 (1.992, 2.596) < 0.0001 | 1.416 (1.194, 1.679) < 0.0001 |
| p for trend | < 0.0001 | < 0.0001 | 0.0001 |
Model 1: unadjusted
Model 2: adjusted for age, race, education level, marital status, PIR
Model 3: adjusted for age, race, marital status, PIR, education level, smoking status, alcohol drinking status, diabetes, hypertension, CVD, parity, menstruation, gravidity, BMI, Non-HDL-C, and physical activity
A/G ratio android-to-gynoid fat ratio, BMI body mass index, SUI stress urinary incontinence, PIR poverty income ratio, CVD cardiovascular diseases, Non-HDL-C non-high-density lipoprotein cholesterol, OR odds ratio, CI confidence interval
Non-linear relationship analysis
In the full adjusted model, a non-linear relationship was detected between SUI and the A/G ratio (shown in Fig. 2). Subsequently, we utilized two-piecewise linear regression model to find the threshold, which came out to be 0.446. On the left of the threshold, a significant positive link existed between the elevated A/G ratio and the development of SUI (OR = 10.308, 95% CI: 3.558–29.863, P < 0.0001). On the right of the threshold, however, the association was not significant (OR = 1.464; 95% CI: 0.775–2.766; P = 0.2402) (Supplementary Table 6). The fitted curves stratified by menstruation revealed distinct threshold effects across subgroups (Fig. 3) (Supplementary Table 7). A threshold was observed at an A/G ratio of 0.417 among premenopausal women. Below this threshold, the A/G ratio was dramatically pronounced associated with risk of SUI (OR = 22.307, 95% CI: 4.982–99.885, P < 0.0001). Above the threshold, this relationship was less pronounced but still significant (OR = 2.551, 95% CI: 1.137–5.724, P = 0.0232). In postmenopausal women, the curve displayed a breakpoint at a A/G ratio of 0.721. Below this breakpoint, the A/G ratio was positively associated with SUI, though this result was non-significant (OR = 2.054, 95% CI: 0.867–4.866, P = 0.1017). Above the breakpoint, the A/G ratio was negatively associated with SUI but did not reach statistical significance (OR = 0.065, 95% CI: 0.003–1.306, P = 0.0742).
Fig. 2.
Smooth curve fitting of the non-linear relationship between the A/G ratio and SUI
Fig. 3.
Smooth curve fitting of the non-linear relationship between the A/G ratio and SUI stratified by menstruation
Subgroup analysis
We then conducted a subgroup analysis between different subgroups, and it is noteworthy that significant interactions were found in female SUI patients among the following subgroups: marital status, menstruation, and hypertension (shown in Fig. 4).
Fig. 4.
Subgroup analyses of the correlation between the A/G ratio and SUI
Discussion
In this study, we sought to investigate the association between SUI and the A/G ratio by analyzing data from NHANES spanning the 2005–2006 and 2011–2018 cycles. Even after accounting for potential confounders, a positive correlation emerged between the A/G ratio and the incidence of SUI. Further exploration through threshold analysis revealed a non-linear relationship between the A/G ratio and SUI, with the inflection point (A/G ratio = 0.446). On the left side of the threshold, the increased A/G ratio was accompanied by a significant rise in SUI risk. However, on the right side of the threshold, the relationship plateaued, and the increase in SUI risk was no longer statistically significant. These findings provide valuable insights into the association between the A/G ratio and SUI.
A/G ratio is an index derived from direct body composition measurements via DXA. The unique feature of A/G ratio lies in its focus on the fat distribution ratio between the android and gynoid regions, directly reflecting “apple-shaped obesity” (abdominal obesity) and “pear-shaped obesity” (lower-body obesity), rather than general obesity [8, 9]. Additionally, the A/G ratio also reflects metabolic status. Android fat was metabolically active and excessive android fat accumulation was associated with metabolic dysregulation and abnormal hormone levels [10, 11]. The decrease in estrogen levels is linked to the increase in android fat. Furthermore, reduced estrogen leads to a reduction in collagen levels within connective tissues and deterioration of the pelvic floor muscles, thereby increasing the risk of SUI [12]. In contrast, gynoid fat has a protective effect, helping to reduce metabolic risk [13]. The elevated A/G ratio reflects an increase in android fat compared to gynoid fat, implying that risk factors may surpass protective factors. A significant correlation has been identified between elevated A/G ratio level and METS [6]. Abdominal obesity, hypertension, IR, and dyslipidemia are key components of METS, and these factors may directly affect the health of the urinary system. Excessive accumulation of abdominal fat increases intra-abdominal pressure, and this persistent high pressure places additional strain on the bladder and pelvic floor muscles, resulting in deformation and atrophy of pelvic floor tissues and thinning of connective tissue, ultimately triggering SUI [14]. In addition, hypertension may lead to pelvic vascular dysfunction and organ ischemia [15]. Pang et al.‘s study has suggested that hypertension was a risk factor for female SUI [16]. Studies have shown that hyperlipidemia can trigger neurodegeneration, activation of inflammation, apoptosis, and increased levels of oxidative stress [17]. Additionally, hyperlipidemia is associated with overactive bladder and decreased spontaneous urethral tension. These changes may contribute to the occurrence of SUI [18]. IR diminishes the anti-inflammatory actions of insulin, promoting a pro-inflammatory state. Under normal conditions, insulin helps reduce inflammation by downregulating the activity of nuclear factor kappa B (NF-κB) within the nucleus and by decreasing reactive oxygen species (ROS). IR interferes with these processes, further enhancing inflammatory responses [19]. In addition, IR is associated with elevated levels of inflammatory mediators, such as tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6) [20]. TNF-α has been demonstrated to suppress the differentiation of urethral sphincter cells, revealing the connection between inflammatory cytokines and SUI [21].
Furthermore, a non-linear association between the A/G ratio and SUI was found in this investigation, characterized by a critical threshold at an A/G ratio value of 0.446, which has important clinical implications. The pronounced association between A/G ratio and SUI below this threshold suggested that interventions aimed at changing abnormal fat distribution and unhealthy metabolic status could be particularly effective in reducing SUI risk among individuals with lower A/G ratio. In addition, the study population was divided into distinct subgroups to check for the existence of heterogeneity. The findings revealed that menstrual status, marital status, and hypertension act as effect modifiers in this association. The relationship between the risk of SUI and the A/G ratio is more pronounced in the premenopausal female population. This may be attributed to the changes in estrogen levels in women before and after menopause. Barbara et al. showed that reduced levels of oestradiol might be associated with an increased risk of SUI in women [22]. Ahn and colleagues demonstrated that urethral hypermobility, as one of the pathophysiological mechanisms of SUI, was associated with low serum estradiol levels [23]. Further studies have shown that estrogen deficiency resulted in insufficient blood flow to the urethral mucosa, causing mucosal atrophy and compromised sealing capacity, which eventually elevated the risk of SUI [24]. Additionally, in postmenopausal women, estrogen withdrawal reduced the proportion of collagen and elastic fibers in the urethral wall, leading to urethral closure dysfunction and contributing to the development of SUI [24]. In summary, postmenopausal estrogen decline was associated with multiple effects, and the A/G ratio showed a weaker association with SUI risk in this context. Compared with postmenopausal women, premenopausal women had better-preserved lower urinary tract structure and function, potentially related to higher estrogen levels [22]. The A/G ratio exhibited a stronger association with SUI risk in premenopausal women. These findings suggest that menstrual history may be a relevant factor in the assessment of SUI risk in women, and the relationship between the A/G ratio and SUI appeared to vary across physiological conditions. Individuals who lived alone or were divorced exhibited a stronger association between the A/G ratio and SUI risk. Mental health issues and differences in lifestyle behaviors among those living alone may be contributing factors. Individuals living alone may be more prone to anxiety, depression, and reduced healthy behaviors, such as high-fat dietary patterns and lack of physical activity, due to a lack of necessary social support. These combined factors contribute to the risk of SUI [25]. Additionally, the relationship between the A/G ratio and SUI risk was more significant in non-hypertensive populations. Hypertension, as an independent risk factor for SUI [16], may dilute the risks associated with obesity and metabolic factors. The potential mechanisms include: pelvic vascular/microvascular supply was susceptible to changes in blood pressure, thereby impacting the lower urinary tract system [15]. Hypertension may precipitate abrupt elevations in intra-abdominal pressure, which can result in fatigue of the pelvic floor muscles, thus contributing to SUI [26]. Moreover, hypertensive patients often required long-term oral antihypertensive medications, the side effects of which may pose new challenges for patients. Studies have shown that the use of alpha-adrenergic receptor antagonists may cause urethral relaxation, thereby increasing the risk of SUI [27]. In contrast, in non-hypertensive women, pelvic floor muscle, vascular, and urethral functions were usually maintained in a normal state, making the impact of abnormal fat distribution more evident in this context. It was noteworthy that the A/G ratio was positively associated with SUI at different BMI levels, and this association was more pronounced among individuals with BMI < 25 kg/m² according to the results of subgroup analysis. Although BMI is widely used due to its convenience and low cost, it cannot distinguish between lean mass and fat mass, nor can it accurately reflect body fat distribution [28]. A potential explanation for this finding is that individuals with normal or low BMI were at increased risk of visceral fat deposition and metabolic dysfunction, who were identified as metabolically unhealthy normal weight (MUHNW) patients [29]. Compared with their obese counterparts, MUHNW patients more commonly exhibited metabolic abnormalities such as insulin resistance, central fat distribution, hyperlipidemia, and elevated blood pressure, all of which significantly increased the risk of SUI [29]. Additionally, in line with earlier studies, we also observed a stronger association between the A/G ratio and SUI among young women and those with high levels of physical activity. Hannestad et al. reported a gradual increase in the prevalence of SUI symptoms among women aged 20 to 49 years [30]. A recent cross-sectional study also observed a stronger association between metabolic score for visceral fat (METS-VF) and UI in young women [31]. Moreover, women who engaged more frequently in high-intensity activities in daily life may experience repeated and significant fluctuations in intra-abdominal pressure [32, 33]. These changes might lead to weak pelvic floor muscle contractions, making them more susceptible to SUI.
These results suggested that the A/G ratio may serve as a tool for early screening of SUI and for the management of high-risk populations. Specifically, the nonlinear relationship between the A/G ratio and SUI indicated that early interventions targeting abnormal fat distribution and metabolic dysfunction (e.g., body shape management, dietary therapy, and medications for metabolic disorders) at low A/G ratio levels may help reduce the risk of SUI. In addition, subgroup analysis results suggested the significance of tailored interventions in particular high-risk populations for SUI. For instance, premenopausal women with increased A/G ratio may need to focus on pelvic floor muscle training and body shape management, while individuals with normal BMI but significant abdominal obesity may benefit from improving body composition and managing metabolic health. Additionally, these strategies might improve the quality of life for women with SUI and help alleviate the healthcare burden related to SUI.
Our study results are scientific and reliable because we used sample weights when analyzing the nationally representative NHANES database to ensure the study’s representativeness. Furthermore, to enhance the reliability of the results, we adjusted for multiple covariates. However, this study has certain limitations: It is difficult to confirm a causal link between SUI and the A/G ratio since the cross-sectional study’s fundamental design. Furthermore, because SUI assessment relied on a binary response format questionnaire, the veracity of SUI self-reported outcomes was yet to be established with certainty. Besides, the findings of this study cannot be extrapolated to the relationship between the A/G ratio and the severity of SUI, which calls for further investigation in future studies. Moreover, the lack of android/gynoid measurements in recent NHANES cycles has limited our ability to explore the relationship between the A/G ratio and SUI in recent years. Future updates with the latest data may help refine our understanding. Additionally, despite the rigorous efforts to control for potential confounding variables, it is possible that residual confounding factors may still exert an influence on the results of the study. Lastly, this study was conducted with a sample comprising solely adult women with SUI, which precludes the possibility of extrapolating its findings to those with urge urinary incontinence (UUI) or other population groups.
Conclusion
There was a non-linear relationship between the A/G ratio and SUI after adjusting for multiple covariates. In individuals with lower A/G ratio, even small increases in A/G ratio substantially elevated the risk of SUI. The clinical significance is that the assessment of fat distribution derived from imaging may be a valuable component of the risk assessment for SUI, enabling physicians to implement early, tailored interventions.
Supplementary Information
Supplementary material 1. Variance inflation factor analysis
Supplementary material 2. Univariate analysis of covariates and SUI
Supplementary material 3. Changes in the regression coefficient of the A/G ratio when introducing covariates in the basic model and removing covariates in the full model
Supplementary material 4. Spearman rank correlation analysis
Supplementary material 5. Multivariable adjusted models of A/G ratio associations with SUI stratified by BMI
Supplementary material 6. The threshold effect analysis of the A/G ratio on SUI
Supplementary material 7. The threshold effect analysis of the A/G ratio on SUI stratified by menstruation
Acknowledgements
We express our sincere gratitude to the NHANES personnel and all participants in this study.
Abbreviations
- A/G ratio
Android-to-gynoid fat ratio
- UI
Urinary incontinence
- SUI
Stress urinary incontinence
- UUI
Urge urinary incontinence
- PIR
Poverty income ratio
- CVD
Cardiovascular disease
- BMI
Body mass index
- Non-HDL-C
Non-high-density lipoprotein cholesterol
- DXA
Dual-energy X-ray absorptiometry
- METS
Metabolic syndrome
- IR
Insulin resistance
- NHANES
National Health and Nutrition Examination Survey
- NCHS
National Center for Health Statistics
- CDC
Centers for Disease Control and Prevention
- PAQ
Physical Activity Questionnaire
- MET
Metabolic equivalent
- SE
Standard errors
- OR
Odds ratio
- CI
Confidence interval
- NF-κB
Nuclear factor kappa B
- ROS
Reactive oxygen species
- TNF-α
Tumor necrosis factor-α
- IL-6
Interleukin-6
- MUHNW
Metabolically unhealthy normal weight
- METS-VF
Metabolic score for visceral fat
- VIF
Variance inflation factor
Authors’ contributions
Jun Zhang: Writing—review and editing, formal analysis. Yucong Zhang: review and editing, formal analysis. Wen Song: review and editing, formal analysis. Xiaodong Song: Writing—review and editing, supervision. Sheng Xin, Adilijiang Aibibuli, and Jiaquan Mao: Methodology, Investigation, Data curation, Conceptualization. Hui Cheng and Xiaming Liu: Investigation, Data curation. All authors accept responsibility for the accuracy and integrity of the work in its entirety.
Funding
Not applicable.
Data availability
The National Health and Nutrition Examination Survey (NHANES) database contains the datasets that were examined in this investigation. (https://www.cdc.gov/nchs/nhanes/index.htm)
Declarations
Ethics approval and consent to participate
All participants gave written informed consent, and the National Center for Health Statistics' Research Ethics Review Board authorized the study. This cross-sectional study was conducted using the NHANES database and did not contain clinical trials.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Jun Zhang and Yucong Zhang contributed equally to this work.
Contributor Information
Wen Song, Email: songwen@hust.edu.cn.
Xiaodong Song, Email: songxdd@126.com.
References
- 1.Haylen BT, et al. An international urogynecological association (IUGA)/international continence society (ICS) joint report on the terminology for female pelvic floor dysfunction. Neurourol Urodyn. 2010;29:4–20. 10.1002/nau.20798. [DOI] [PubMed] [Google Scholar]
- 2.Özdemir K, Şahin S, Özerdoğan N, Ünsal A. Evaluation of urinary incontinence and quality of life in married women aged between 20 and 49 years (Sakarya, Turkey). Turk J Med Sci. 2018;48:100–9. 10.3906/sag-1605-73. [DOI] [PubMed] [Google Scholar]
- 3.Nygaard I. Prevalence of symptomatic pelvic floor disorders in US women. JAMA. 2008;300:1311–6. 10.1001/jama.300.11.1311. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Yao D et al. (2020) Relationship between Maternal Central Obesity and the Risk of Gestational Diabetes Mellitus: A Systematic Review and Meta-Analysis of Cohort Studies Journal of diabetes research 2020, 6303820. 10.1155/2020/6303820 [DOI] [PMC free article] [PubMed]
- 5.Lukaski HC. Soft tissue composition and bone mineral status: evaluation by dual-energy X-ray absorptiometry. J Nutr. 1993;123:438–43. [PubMed] [Google Scholar]
- 6.Petro JL, et al. Serum levels of myonectin are lower in adults with metabolic syndrome and are negatively correlated with android fat mass. Int J Mol Sci. 2023. 10.3390/ijms24086874. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Samsell L, Regier M, Walton C, Cottrell L. Importance of android/gynoid fat ratio in predicting metabolic and cardiovascular disease risk in normal weight as well as overweight and obese children. J Obes. 2014;2014:2014(846578). 10.1155/2014/846578. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Shepherd JA, et al. A multinational study to develop universal standardization of whole-body bone density and composition using GE healthcare lunar and hologic DXA systems. J Bone Miner Res. 2012;27:2208–16. 10.1002/jbmr.1654. [DOI] [PubMed] [Google Scholar]
- 9.Karastergiou K, Smith SR, Greenberg AS, Fried SK. Sex differences in human adipose tissues - the biology of pear shape. Biol Sex Differ. 2012;3:13. 10.1186/2042-6410-3-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hirschberg AL. Sex hormones, appetite and eating behaviour in women. Maturitas. 2012;71:248–56. 10.1016/j.maturitas.2011.12.016. [DOI] [PubMed] [Google Scholar]
- 11.Palmer BF, Clegg DJ. The sexual dimorphism of obesity. Mol Cell Endocrinol. 2015;402:113–9. 10.1016/j.mce.2014.11.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Blanco ZE, Lilue M, Palacios S. Experience with ospemifene in patients with vulvar and vaginal atrophy and urinary incontinence: case studies. Drugs Context. 2020. 10.7573/dic.2020-3-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Manolopoulos KN, Karpe F, Frayn KN. Gluteofemoral body fat as a determinant of metabolic health. Int J Obes. 2010;34:949–59. 10.1038/ijo.2009.286. [DOI] [PubMed] [Google Scholar]
- 14.Osborn DJ, Strain M, Gomelsky A, Rothschild J, Dmochowski R. Obesity and female stress urinary incontinence. Urology. 2013;82:759–63. 10.1016/j.urology.2013.06.020. [DOI] [PubMed] [Google Scholar]
- 15.John G. Urinary incontinence and cardiovascular disease: a narrative review. Int Urogynecol J. 2020;31:857–63. 10.1007/s00192-019-04058-w. [DOI] [PubMed] [Google Scholar]
- 16.Pang H, et al. The incidence of urinary incontinence in Chinese hypertensive women and the relationship between hypertension and urinary incontinence. Eur J Obstet Gynecol Reprod Biol. 2024;301:210–5. 10.1016/j.ejogrb.2024.08.019. [DOI] [PubMed] [Google Scholar]
- 17.Pillar S, Moisseiev E, Sokolovska J, Grzybowski A. Recent developments in diabetic retinal neurodegeneration: a literature review. J Diabetes Res. 2020;2020:2020(5728674). 10.1155/2020/5728674. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chen D, et al. Hypertriglyceridemia impairs urethral spontaneous tone through down-regulation of ANO1 in mouse urethral smooth muscle cells. Urology. 2022;165:157–63. 10.1016/j.urology.2022.01.042. [DOI] [PubMed] [Google Scholar]
- 19.Dandona P, et al. Insulin inhibits intranuclear nuclear factor kappab and stimulates IkappaB in mononuclear cells in obese subjects: evidence for an anti-inflammatory effect? J Clin Endocrinol Metab. 2001;86:3257–65. 10.1210/jcem.86.7.7623. [DOI] [PubMed] [Google Scholar]
- 20.Akash MSH, et al. Biochemical investigation of gender-specific association between insulin resistance and inflammatory biomarkers in types 2 diabetic patients. Biomed Pharmacother. 2018;106:285–91. 10.1016/j.biopha.2018.06.044. [DOI] [PubMed] [Google Scholar]
- 21.Shinohara M, et al. Tumor necrosis factor-alpha inhibits differentiation of myogenic cells in human urethral rhabdosphincter. Int J Urol. 2017;24(6):461–7. 10.1111/iju.13330. [DOI] [PubMed] [Google Scholar]
- 22.Bodner-Adler B, et al. Role of serum steroid hormones in women with stress urinary incontinence: a case-control study. BJU Int. 2017;120:416–21. 10.1111/bju.13902. [DOI] [PubMed] [Google Scholar]
- 23.Ahn KH, et al. Relationship between serum estradiol and follicle-stimulating hormone levels and urodynamic results in women with stress urinary incontinence. Int Urogynecol J. 2011;22:731–7. 10.1007/s00192-011-1359-6. [DOI] [PubMed] [Google Scholar]
- 24.Yang X, et al. The anatomical pathogenesis of stress urinary incontinence in women. Medicina (B Aires). 2022. 10.3390/medicina59010005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ramage-Morin PL, Gilmour H. Urinary incontinence and loneliness in Canadian seniors. Health Rep. 2013;24:3–10. [PubMed] [Google Scholar]
- 26.Almutairi S, et al. Urinary incontinence among Saudi women: prevalence, risk factors, and impact on quality of life. Eur Rev Med Pharmacol Sci. 2021;25:6311–8. 10.26355/eurrev_202110_27001. [DOI] [PubMed] [Google Scholar]
- 27.Ruby CM, et al. The effect of medication use on urinary incontinence in community-dwelling elderly women. J Am Geriatr Soc. 2010;58:1715–20. 10.1111/j.1532-5415.2010.03006.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.De Lorenzo A, et al. New obesity classification criteria as a tool for bariatric surgery indication. World J Gastroenterol. 2016;22(2):681–703. 10.3748/wjg.v22.i2.681. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Mathew H, Farr OM, Mantzoros CS. Metabolic health and weight: understanding metabolically unhealthy normal weight or metabolically healthy obese patients. Metabolism. 2016;65:73–80. 10.1016/j.metabol.2015.10.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hannestad YS, Rortveit G, Sandvik H, Hunskaar S. A community-based epidemiological survey of female urinary incontinence: the Norwegian EPINCONT study. Epidemiology of incontinence in the County of Nord-Trøndelag. J Clin Epidemiol. 2000;53:1150–7. 10.1016/s0895-4356(00)00232-8. [DOI] [PubMed] [Google Scholar]
- 31.Ye C, et al. Associations between metabolic score for visceral fat and urinary incontinence among US adult women: a population-based cross-sectional study. BMC Public Health. 2025;25:932. 10.1186/s12889-025-21966-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Ree ML, Nygaard I, Bø K. Muscular fatigue in the pelvic floor muscles after strenuous physical activity. Acta Obstet Gynecol Scand. 2007;86:870–6. 10.1080/00016340701417281. [DOI] [PubMed] [Google Scholar]
- 33.Skaug KL, Engh ME, Frawley H, Bø K. Urinary and anal incontinence among female gymnasts and cheerleaders-bother and associated factors. A cross-sectional study. Int Urogynecol J. 2022;33:955–64. 10.1007/s00192-021-04696-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary material 1. Variance inflation factor analysis
Supplementary material 2. Univariate analysis of covariates and SUI
Supplementary material 3. Changes in the regression coefficient of the A/G ratio when introducing covariates in the basic model and removing covariates in the full model
Supplementary material 4. Spearman rank correlation analysis
Supplementary material 5. Multivariable adjusted models of A/G ratio associations with SUI stratified by BMI
Supplementary material 6. The threshold effect analysis of the A/G ratio on SUI
Supplementary material 7. The threshold effect analysis of the A/G ratio on SUI stratified by menstruation
Data Availability Statement
The National Health and Nutrition Examination Survey (NHANES) database contains the datasets that were examined in this investigation. (https://www.cdc.gov/nchs/nhanes/index.htm)